------------------------------------------------------------------------------------------- log: C:\Llama\GrowthPoles.txt log type: text opened on: 23 Aug 2011, 17:54:20 . . /* Open */ . clear . set more off . set mem 100m (102400k) . set matsize 800 . . /* Merge datasets */ . clear . insheet using "C:\Llama\enroll.csv" (3 vars, 38 obs) . ren ccode code . sort code . save "C:\Llama\Temp\enroll.dta", replace file C:\Llama\Temp\enroll.dta saved . use "C:\Llama\Chicago.dta", clear . sort code . merge code using "C:\Llama\Temp\enroll.dta" variable code does not uniquely identify observations in the master data . drop _merge . sort code year . save "C:\Llama\Temp\raw1.dta", replace file C:\Llama\Temp\raw1.dta saved . . /* Generate indices */ . preserve . keep if year == 1971 (1669 observations deleted) . gsort -index_constusd_HPg . group1d index_constusd_HPg, max(3) gen(bindex=3) Partitions of 19 data up to 3 groups 1 group: sum of squares 225.20 Group Size First Last Mean SD 1 19 1 13.6748 19 -.46662 2.12 3.44 2 groups: sum of squares 77.70 Group Size First Last Mean SD 2 14 6 3.03673 19 -.46662 0.45 1.08 1 5 1 13.6748 5 4.1478 6.78 3.50 3 groups: sum of squares 10.08 Group Size First Last Mean SD 3 12 8 .754723 19 -.46662 0.04 0.42 2 6 2 5.93255 7 2.81237 4.35 1.15 1 1 1 13.6748 1 13.6748 13.67 0.00 Groups Sums of squares 1 225.20 2 77.70 3 10.08 . replace bindex = 1 if bindex == 2 (6 real changes made) . replace bindex = 0 if bindex == 3 (12 real changes made) . gsort -index_constusd_migr_HPg . group1d index_constusd_migr_HPg, max(3) gen(bindex3=3) Partitions of 18 data up to 3 groups 1 group: sum of squares 224.18 Group Size First Last Mean SD 1 18 1 14.6557 18 -.564739 2.11 3.53 2 groups: sum of squares 57.57 Group Size First Last Mean SD 2 17 2 5.06765 18 -.564739 1.37 1.84 1 1 1 14.6557 1 14.6557 14.66 0.00 3 groups: sum of squares 9.84 Group Size First Last Mean SD 3 12 7 1.53188 18 -.564739 0.29 0.72 2 5 2 5.06765 6 2.82198 3.97 0.85 1 1 1 14.6557 1 14.6557 14.66 0.00 Groups Sums of squares 1 224.18 2 57.57 3 9.84 . replace bindex3 = 1 if bindex3 == 2 (5 real changes made) . replace bindex3 = 0 if bindex3 == 3 (12 real changes made) . keep code year bindex* . sort year code . save "C:\Llama\Temp\bindex0.dta", replace file C:\Llama\Temp\bindex0.dta saved . restore . local i = 1976 . while `i' <= 2006 { 2. preserve 3. keep if year == `i' 4. gsort -index_constusd_HPg 5. group1d index_constusd_HPg, max(3) gen(bindex=3) 6. replace bindex = 1 if bindex == 2 7. replace bindex = 0 if bindex == 3 8. gsort -index_constusd_migr_HPg 9. group1d index_constusd_migr_HPg, max(3) gen(bindex3=3) 10. replace bindex3 = 1 if bindex3 == 2 11. replace bindex3 = 0 if bindex3 == 3 12. keep code year bindex* 13. sort year code 14. append using "C:\Llama\Temp\bindex0.dta" 15. save "C:\Llama\Temp\bindex0.dta", replace 16. local i = `i' + 5 17. restore 18. } (1669 observations deleted) Partitions of 86 data up to 3 groups 1 group: sum of squares 290.56 Group Size First Last Mean SD 1 86 1 13.3489 86 -.530635 0.15 1.84 2 groups: sum of squares 78.18 Group Size First Last Mean SD 2 82 5 3.22045 86 -.530635 -0.20 0.55 1 4 1 13.3489 4 4.28404 7.26 3.65 3 groups: sum of squares 20.03 Group Size First Last Mean SD 3 81 6 1.41019 86 -.530635 -0.24 0.40 2 4 2 6.82169 5 3.22045 4.73 1.31 1 1 1 13.3489 1 13.3489 13.35 0.00 Groups Sums of squares 1 290.56 2 78.18 3 20.03 (4 real changes made) (81 real changes made) Partitions of 81 data up to 3 groups 1 group: sum of squares 317.60 Group Size First Last Mean SD 1 81 1 14.6678 81 -.635383 0.20 1.98 2 groups: sum of squares 95.76 Group Size First Last Mean SD 2 79 3 4.12923 81 -.635383 -0.07 0.89 1 2 1 14.6678 2 6.52474 10.60 4.07 3 groups: sum of squares 28.16 Group Size First Last Mean SD 3 75 7 1.12333 81 -.635383 -0.24 0.46 2 5 2 6.52474 6 2.13962 3.82 1.59 1 1 1 14.6678 1 14.6678 14.67 0.00 Groups Sums of squares 1 317.60 2 95.76 3 28.16 (5 real changes made) (75 real changes made) file C:\Llama\Temp\bindex0.dta saved (1669 observations deleted) Partitions of 93 data up to 3 groups 1 group: sum of squares 149.00 Group Size First Last Mean SD 1 93 1 10.346 93 -.525478 -0.09 1.27 2 groups: sum of squares 36.58 Group Size First Last Mean SD 2 91 3 1.72527 93 -.525478 -0.26 0.45 1 2 1 10.346 2 4.29889 7.32 3.02 3 groups: sum of squares 11.95 Group Size First Last Mean SD 3 85 9 .503461 93 -.525478 -0.36 0.19 2 7 2 4.29889 8 .738979 1.68 1.14 1 1 1 10.346 1 10.346 10.35 0.00 Groups Sums of squares 1 149.00 2 36.58 3 11.95 (7 real changes made) (85 real changes made) Partitions of 90 data up to 3 groups 1 group: sum of squares 178.41 Group Size First Last Mean SD 1 90 1 11.2968 90 -.718702 -0.07 1.41 2 groups: sum of squares 47.69 Group Size First Last Mean SD 2 89 2 3.84723 90 -.718702 -0.20 0.73 1 1 1 11.2968 1 11.2968 11.30 0.00 3 groups: sum of squares 10.71 Group Size First Last Mean SD 3 81 10 .596794 90 -.718702 -0.40 0.25 2 8 2 3.84723 9 .933647 1.85 0.84 1 1 1 11.2968 1 11.2968 11.30 0.00 Groups Sums of squares 1 178.41 2 47.69 3 10.71 (8 real changes made) (81 real changes made) file C:\Llama\Temp\bindex0.dta saved (1667 observations deleted) Partitions of 92 data up to 3 groups 1 group: sum of squares 340.89 Group Size First Last Mean SD 1 92 1 14.0398 92 -.594895 0.10 1.92 2 groups: sum of squares 76.10 Group Size First Last Mean SD 2 90 3 4.89346 92 -.594895 -0.15 0.84 1 2 1 14.0398 2 8.92105 11.48 2.56 3 groups: sum of squares 30.71 Group Size First Last Mean SD 3 86 7 1.1591 92 -.594895 -0.31 0.37 2 4 3 4.89346 6 1.81677 3.14 1.21 1 2 1 14.0398 2 8.92105 11.48 2.56 Groups Sums of squares 1 340.89 2 76.10 3 30.71 (4 real changes made) (86 real changes made) Partitions of 90 data up to 3 groups 1 group: sum of squares 423.69 Group Size First Last Mean SD 1 90 1 16.9686 90 -1.01857 0.11 2.17 2 groups: sum of squares 114.14 Group Size First Last Mean SD 2 88 3 4.82854 90 -1.01857 -0.17 0.91 1 2 1 16.9686 2 7.8462 12.41 4.56 3 groups: sum of squares 41.32 Group Size First Last Mean SD 3 82 9 1.16491 90 -1.01857 -0.39 0.37 2 7 2 7.8462 8 1.72411 3.45 2.07 1 1 1 16.9686 1 16.9686 16.97 0.00 Groups Sums of squares 1 423.69 2 114.14 3 41.32 (7 real changes made) (82 real changes made) file C:\Llama\Temp\bindex0.dta saved (1667 observations deleted) Partitions of 99 data up to 3 groups 1 group: sum of squares 154.17 Group Size First Last Mean SD 1 99 1 8.42354 99 -.697538 -0.04 1.25 2 groups: sum of squares 35.94 Group Size First Last Mean SD 2 95 5 1.64075 99 -.697538 -0.27 0.46 1 4 1 8.42354 4 2.8924 5.28 1.99 3 groups: sum of squares 15.08 Group Size First Last Mean SD 3 86 14 .157175 99 -.697538 -0.40 0.15 2 10 4 2.8924 13 .457074 1.21 0.70 1 3 1 8.42354 3 4.84468 6.08 1.66 Groups Sums of squares 1 154.17 2 35.94 3 15.08 (10 real changes made) (86 real changes made) Partitions of 94 data up to 3 groups 1 group: sum of squares 178.23 Group Size First Last Mean SD 1 94 1 9.51661 94 -.717996 -0.03 1.38 2 groups: sum of squares 47.30 Group Size First Last Mean SD 2 90 5 1.83514 94 -.717996 -0.28 0.53 1 4 1 9.51661 4 3.32748 5.57 2.36 3 groups: sum of squares 25.59 Group Size First Last Mean SD 3 80 15 .353345 94 -.717996 -0.44 0.22 2 11 4 3.32748 14 .571652 1.25 0.75 1 3 1 9.51661 3 4.43556 6.32 2.27 Groups Sums of squares 1 178.23 2 47.30 3 25.59 (11 real changes made) (80 real changes made) file C:\Llama\Temp\bindex0.dta saved (1667 observations deleted) Partitions of 119 data up to 3 groups 1 group: sum of squares 238.14 Group Size First Last Mean SD 1 119 1 13.7127 119 -.700572 -0.10 1.41 2 groups: sum of squares 45.72 Group Size First Last Mean SD 2 118 2 3.50999 119 -.700572 -0.22 0.62 1 1 1 13.7127 1 13.7127 13.71 0.00 3 groups: sum of squares 10.99 Group Size First Last Mean SD 3 108 12 .422389 119 -.700572 -0.38 0.18 2 10 2 3.50999 11 .757721 1.56 0.86 1 1 1 13.7127 1 13.7127 13.71 0.00 Groups Sums of squares 1 238.14 2 45.72 3 10.99 (10 real changes made) (108 real changes made) Partitions of 119 data up to 3 groups 1 group: sum of squares 335.65 Group Size First Last Mean SD 1 119 1 15.6841 119 -3.84747 -0.17 1.68 2 groups: sum of squares 82.07 Group Size First Last Mean SD 2 118 2 4.09306 119 -3.84747 -0.31 0.83 1 1 1 15.6841 1 15.6841 15.68 0.00 3 groups: sum of squares 38.50 Group Size First Last Mean SD 3 107 13 .402495 119 -3.84747 -0.50 0.51 2 11 2 4.09306 12 .572931 1.59 0.97 1 1 1 15.6841 1 15.6841 15.68 0.00 Groups Sums of squares 1 335.65 2 82.07 3 38.50 (11 real changes made) (107 real changes made) file C:\Llama\Temp\bindex0.dta saved (1667 observations deleted) Partitions of 127 data up to 3 groups 1 group: sum of squares 150.34 Group Size First Last Mean SD 1 127 1 9.90412 127 -.51535 -0.15 1.09 2 groups: sum of squares 41.54 Group Size First Last Mean SD 2 125 3 2.56744 127 -.51535 -0.26 0.46 1 2 1 9.90412 2 4.43884 7.17 2.73 3 groups: sum of squares 16.00 Group Size First Last Mean SD 3 115 13 .256827 127 -.51535 -0.38 0.16 2 11 2 4.43884 12 .639234 1.42 1.09 1 1 1 9.90412 1 9.90412 9.90 0.00 Groups Sums of squares 1 150.34 2 41.54 3 16.00 (11 real changes made) (115 real changes made) Partitions of 127 data up to 3 groups 1 group: sum of squares 222.14 Group Size First Last Mean SD 1 127 1 12.398 127 -.676271 -0.13 1.32 2 groups: sum of squares 64.01 Group Size First Last Mean SD 2 126 2 4.71495 127 -.676271 -0.23 0.71 1 1 1 12.398 1 12.398 12.40 0.00 3 groups: sum of squares 20.68 Group Size First Last Mean SD 3 116 12 .56446 127 -.676271 -0.40 0.26 2 10 2 4.71495 11 .755923 1.77 1.12 1 1 1 12.398 1 12.398 12.40 0.00 Groups Sums of squares 1 222.14 2 64.01 3 20.68 (10 real changes made) (116 real changes made) file C:\Llama\Temp\bindex0.dta saved (1668 observations deleted) Partitions of 125 data up to 3 groups 1 group: sum of squares 143.49 Group Size First Last Mean SD 1 125 1 9.58728 125 -.482181 -0.15 1.07 2 groups: sum of squares 19.40 Group Size First Last Mean SD 2 123 3 .999809 125 -.482181 -0.28 0.31 1 2 1 9.58728 2 5.73739 7.66 1.92 3 groups: sum of squares 10.10 Group Size First Last Mean SD 3 109 17 .022542 125 -.482181 -0.38 0.13 2 14 3 .999809 16 .097343 0.49 0.25 1 2 1 9.58728 2 5.73739 7.66 1.92 Groups Sums of squares 1 143.49 2 19.40 3 10.10 (14 real changes made) (109 real changes made) Partitions of 125 data up to 3 groups 1 group: sum of squares 170.05 Group Size First Last Mean SD 1 125 1 9.39154 125 -.584192 -0.14 1.17 2 groups: sum of squares 27.57 Group Size First Last Mean SD 2 123 3 2.32105 125 -.584192 -0.27 0.45 1 2 1 9.39154 2 7.07762 8.23 1.16 3 groups: sum of squares 11.41 Group Size First Last Mean SD 3 103 23 .032724 125 -.584192 -0.43 0.16 2 20 3 2.32105 22 .086562 0.55 0.56 1 2 1 9.39154 2 7.07762 8.23 1.16 Groups Sums of squares 1 170.05 2 27.57 3 11.41 (20 real changes made) (103 real changes made) file C:\Llama\Temp\bindex0.dta saved . preserve . keep if year == 1981 (1669 observations deleted) . gsort -index_ppp_HPg . group1d index_ppp_HPg, max(3) gen(bindex2=3) Partitions of 91 data up to 3 groups 1 group: sum of squares 293.15 Group Size First Last Mean SD 1 91 1 15.0517 91 -2.40286 -0.02 1.79 2 groups: sum of squares 63.43 Group Size First Last Mean SD 2 90 2 4.95303 91 -2.40286 -0.19 0.84 1 1 1 15.0517 1 15.0517 15.05 0.00 3 groups: sum of squares 19.15 Group Size First Last Mean SD 3 88 4 1.72298 91 -2.40286 -0.29 0.46 2 2 2 4.95303 3 3.97424 4.46 0.49 1 1 1 15.0517 1 15.0517 15.05 0.00 Groups Sums of squares 1 293.15 2 63.43 3 19.15 . replace bindex2 = 1 if bindex2 == 2 (2 real changes made) . replace bindex2 = 0 if bindex2 == 3 (88 real changes made) . keep code year bindex2 . sort year code . save "C:\Llama\Temp\bindex2.dta", replace file C:\Llama\Temp\bindex2.dta saved . restore . local i = 1986 . while `i' <= 2006 { 2. preserve 3. keep if year == `i' 4. gsort -index_ppp_HPg 5. group1d index_ppp_HPg, max(3) gen(bindex2=3) 6. replace bindex2 = 1 if bindex2 == 2 7. replace bindex2 = 0 if bindex2 == 3 8. keep code year bindex2 9. sort year code 10. append using "C:\Llama\Temp\bindex2.dta" 11. save "C:\Llama\Temp\bindex2.dta", replace 12. local i = `i' + 5 13. restore 14. } (1667 observations deleted) Partitions of 90 data up to 3 groups 1 group: sum of squares 425.30 Group Size First Last Mean SD 1 90 1 15.4275 90 -.666607 0.20 2.17 2 groups: sum of squares 90.38 Group Size First Last Mean SD 2 87 4 4.15705 90 -.666607 -0.15 0.71 1 3 1 15.4275 3 5.84304 10.59 3.91 3 groups: sum of squares 35.03 Group Size First Last Mean SD 3 84 7 1.29981 90 -.666607 -0.26 0.40 2 4 3 5.84304 6 2.08925 3.60 1.52 1 2 1 15.4275 2 10.5046 12.97 2.46 Groups Sums of squares 1 425.30 2 90.38 3 35.03 (4 real changes made) (84 real changes made) file C:\Llama\Temp\bindex2.dta saved (1667 observations deleted) Partitions of 97 data up to 3 groups 1 group: sum of squares 185.85 Group Size First Last Mean SD 1 97 1 9.21808 97 -.804864 0.03 1.38 2 groups: sum of squares 42.97 Group Size First Last Mean SD 2 93 5 1.97692 97 -.804864 -0.22 0.51 1 4 1 9.21808 4 3.22318 5.88 2.15 3 groups: sum of squares 17.39 Group Size First Last Mean SD 3 85 13 .529589 97 -.804864 -0.36 0.19 2 9 4 3.22318 12 .594881 1.49 0.76 1 3 1 9.21808 3 5.46496 6.77 1.73 Groups Sums of squares 1 185.85 2 42.97 3 17.39 (9 real changes made) (85 real changes made) file C:\Llama\Temp\bindex2.dta saved (1667 observations deleted) Partitions of 117 data up to 3 groups 1 group: sum of squares 289.17 Group Size First Last Mean SD 1 117 1 15.155 117 -.678888 -0.03 1.57 2 groups: sum of squares 56.59 Group Size First Last Mean SD 2 116 2 3.98558 117 -.678888 -0.16 0.70 1 1 1 15.155 1 15.155 15.16 0.00 3 groups: sum of squares 13.65 Group Size First Last Mean SD 3 106 12 .541845 117 -.678888 -0.35 0.19 2 10 2 3.98558 11 .983202 1.82 0.98 1 1 1 15.155 1 15.155 15.16 0.00 Groups Sums of squares 1 289.17 2 56.59 3 13.65 (10 real changes made) (106 real changes made) file C:\Llama\Temp\bindex2.dta saved (1667 observations deleted) Partitions of 125 data up to 3 groups 1 group: sum of squares 171.86 Group Size First Last Mean SD 1 125 1 10.5595 125 -.462668 -0.09 1.17 2 groups: sum of squares 48.17 Group Size First Last Mean SD 2 123 3 2.94932 125 -.462668 -0.22 0.51 1 2 1 10.5595 2 4.86249 7.71 2.85 3 groups: sum of squares 18.78 Group Size First Last Mean SD 3 113 13 .321187 125 -.462668 -0.35 0.17 2 11 2 4.86249 12 .77191 1.62 1.19 1 1 1 10.5595 1 10.5595 10.56 0.00 Groups Sums of squares 1 171.86 2 48.17 3 18.78 (11 real changes made) (113 real changes made) file C:\Llama\Temp\bindex2.dta saved (1668 observations deleted) Partitions of 124 data up to 3 groups 1 group: sum of squares 206.95 Group Size First Last Mean SD 1 124 1 11.2169 124 -.454964 -0.04 1.29 2 groups: sum of squares 29.30 Group Size First Last Mean SD 2 122 3 1.82039 124 -.454964 -0.19 0.42 1 2 1 11.2169 2 7.40559 9.31 1.91 3 groups: sum of squares 12.36 Group Size First Last Mean SD 3 108 17 .150812 124 -.454964 -0.32 0.16 2 14 3 1.82039 16 .315243 0.84 0.41 1 2 1 11.2169 2 7.40559 9.31 1.91 Groups Sums of squares 1 206.95 2 29.30 3 12.36 (14 real changes made) (108 real changes made) file C:\Llama\Temp\bindex2.dta saved . use "C:\Llama\Temp\bindex0.dta", clear . sort code year . save "C:\Llama\Temp\bindex1.dta", replace file C:\Llama\Temp\bindex1.dta saved . use "C:\Llama\Temp\bindex2.dta", clear . sort code year . merge code year using "C:\Llama\Temp\bindex1.dta" . drop _merge . sort code year . save "C:\Llama\Temp\bindex.dta", replace file C:\Llama\Temp\bindex.dta saved . use "C:\Llama\Temp\raw1.dta", clear . merge code year using "C:\Llama\Temp\bindex.dta" . drop _merge . . gen lindex = ln(1.8+index_constusd_HPg) (1146 missing values generated) . summarize index_constusd_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- index_~d_HPg | 760 -4.69e-10 1.534652 -.7005721 14.03978 . gen index_rescale = 100/(r(max)-r(min))*index_constusd_HPg-(r(min)*100/(r(max)-r(min))) (1146 missing values generated) . gen lindex_rescale = ln(1+index_rescale) (1146 missing values generated) . gen lindex2 = ln(3.5+index_ppp_HPg) (1262 missing values generated) . summarize index_ppp_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- index_~p_HPg | 644 7.87e-10 1.566281 -2.402863 15.4275 . gen index2_rescale = 100/(r(max)-r(min))*index_ppp_HPg-(r(min)*100/(r(max)-r(min))) (1262 missing values generated) . gen lindex2_rescale = ln(1+index2_rescale) (1262 missing values generated) . gen lindex3 = ln(4.9+index_constusd_migr_HPg) (1162 missing values generated) . summarize index_constusd_migr_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- i~d_migr_HPg | 744 7.32e-10 1.698255 -3.847472 16.96864 . gen index3_rescale = 100/(r(max)-r(min))*index_constusd_migr_HPg-(r(min)*100/(r(max)-r(mi > n))) (1162 missing values generated) . gen lindex3_rescale = ln(1+index3_rescale) (1162 missing values generated) . . /* Generate proximate controls */ . gen init_gdppc_temp = avg5_gdppc_new (614 missing values generated) . replace init_gdppc_temp = 0 if year ~= 1971 (1669 real changes made) . egen init_gdppc = mean(init_gdppc_temp), by(code) . replace init_gdppc = . if init_gdppc == 0 (954 real changes made, 954 to missing) . drop init_gdppc_temp . gen linit_gdppc = ln(init_gdppc) (954 missing values generated) . . gen liy = ln(1+I_over_Y) (810 missing values generated) . gen lpop = ln(avg5_pop_wdi) (260 missing values generated) . gen ln = ln(1.1+avg5_pop_growth) (261 missing values generated) . gen lenroll_p = ln(avg5_enroll_pri_grs) (593 missing values generated) . gen lenroll_s = ln(0.5+avg5_enroll_sec_grs) (651 missing values generated) . gen lgdppc = ln(avg5_gdppc_new) (614 missing values generated) . gen lroads = ln(1+avg5_roads_paved) (1341 missing values generated) . gen lcell = ln(1+avg5_cell_subscriptions) (328 missing values generated) . gen lmort_ufive = ln(avg5_mort_ufive) (481 missing values generated) . gen llifeexp = ln(avg5_lifeexp) (336 missing values generated) . gen ldepratio1 = ln(avg5_depratio1) (394 missing values generated) . gen ldepratio2 = ln(0.4+avg5_depratio2) (1213 missing values generated) . gen ldepratioold = ln(avg5_depratio_old) (394 missing values generated) . gen ldepratioyoung = ln(avg5_depratio_young) (394 missing values generated) . gen lgovt_Y3 = ln(1+G_over_Y) (764 missing values generated) . . /* Generate fundamental controls */ . pca avg5_ICRG_govstab avg5_ICRG_corrupt avg5_ICRG_law avg5_ICRG_bureau Principal components/correlation Number of obs = 684 Number of comp. = 4 Trace = 4 Rotation: (unrotated = principal) Rho = 1.0000 -------------------------------------------------------------------------- Component | Eigenvalue Difference Proportion Cumulative -------------+------------------------------------------------------------ Comp1 | 2.50201 1.56952 0.6255 0.6255 Comp2 | .93249 .642065 0.2331 0.8586 Comp3 | .290426 .015354 0.0726 0.9312 Comp4 | .275072 . 0.0688 1.0000 -------------------------------------------------------------------------- Principal components (eigenvectors) -------------------------------------------------------------------- Variable | Comp1 Comp2 Comp3 Comp4 | Unexplained -------------+----------------------------------------+------------- avg5_ICRG_~b | 0.3097 0.8819 0.0376 0.3535 | 0 avg5_ICRG~pt | 0.5215 -0.4222 -0.3800 0.6367 | 0 avg5_ICRG_~w | 0.5671 0.0920 -0.4593 -0.6775 | 0 avg5_ICRG_~u | 0.5573 -0.1887 0.8020 -0.1028 | 0 -------------------------------------------------------------------- . predict comp1 comp2, score (2 components skipped) Scoring coefficients sum of squares(column-loading) = 1 ------------------------------------------------------ Variable | Comp1 Comp2 Comp3 Comp4 -------------+---------------------------------------- avg5_ICRG_~b | 0.3097 0.8819 0.0376 0.3535 avg5_ICRG~pt | 0.5215 -0.4222 -0.3800 0.6367 avg5_ICRG_~w | 0.5671 0.0920 -0.4593 -0.6775 avg5_ICRG_~u | 0.5573 -0.1887 0.8020 -0.1028 ------------------------------------------------------ . gen inst = 0.6152*comp1 + 0.2306*comp2 (1222 missing values generated) . pca avg5_ICRG_* Principal components/correlation Number of obs = 684 Number of comp. = 12 Trace = 12 Rotation: (unrotated = principal) Rho = 1.0000 -------------------------------------------------------------------------- Component | Eigenvalue Difference Proportion Cumulative -------------+------------------------------------------------------------ Comp1 | 6.2724 4.95782 0.5227 0.5227 Comp2 | 1.31458 .2365 0.1095 0.6322 Comp3 | 1.07808 .353912 0.0898 0.7221 Comp4 | .72417 .13554 0.0603 0.7824 Comp5 | .58863 .133883 0.0491 0.8315 Comp6 | .454746 .0202198 0.0379 0.8694 Comp7 | .434527 .0951791 0.0362 0.9056 Comp8 | .339348 .0828849 0.0283 0.9339 Comp9 | .256463 .0404953 0.0214 0.9552 Comp10 | .215967 .0373092 0.0180 0.9732 Comp11 | .178658 .0362337 0.0149 0.9881 Comp12 | .142424 . 0.0119 1.0000 -------------------------------------------------------------------------- Principal components (eigenvectors) ------------------------------------------------------------------------------------ Variable | Comp1 Comp2 Comp3 Comp4 Comp5 Comp6 Comp7 -------------+---------------------------------------------------------------------- avg5_ICRG_~u | 0.3286 -0.2978 -0.1999 -0.0001 -0.0615 -0.0550 0.0354 avg5_ICRG~pt | 0.2807 -0.4260 0.0969 -0.2069 -0.2636 -0.1840 0.4022 avg5_ICRG~oc | 0.2816 -0.2667 0.0199 0.5380 -0.1790 0.5189 0.1494 avg5_ICRG~ic | 0.2580 0.2447 0.3157 -0.4853 -0.0187 0.5958 -0.2360 a~xtconflict | 0.2678 0.2634 0.2434 0.4000 -0.3693 -0.4172 -0.4246 avg5_ICRG_~b | 0.2056 0.6032 -0.2827 -0.0109 -0.1272 -0.0438 0.5213 a~ntconflict | 0.3404 0.2312 0.1453 -0.0857 -0.0955 -0.1028 -0.1232 avg5_ICRG_~f | 0.2865 0.2228 -0.4344 0.2625 0.3765 0.1587 -0.0064 avg5_ICRG_~w | 0.3381 -0.0102 -0.0251 -0.3299 -0.1850 -0.1475 0.1601 avg5_ICRG_~y | 0.3319 -0.1678 -0.0030 0.0809 0.0339 0.1187 -0.2429 avg5_ICRG_~s | 0.2082 0.0266 0.6210 0.1641 0.6245 -0.1714 0.3204 avg5_ICRG~ec | 0.2959 -0.1829 -0.3396 -0.2316 0.4069 -0.2490 -0.3233 ------------------------------------------------------------------------------------ ------------------------------------------------------------------------------ Variable | Comp8 Comp9 Comp10 Comp11 Comp12 | Unexplained -------------+--------------------------------------------------+------------- avg5_ICRG_~u | 0.1572 0.0846 -0.7695 -0.2359 -0.2750 | 0 avg5_ICRG~pt | 0.2373 0.3306 0.4607 0.0354 -0.2080 | 0 avg5_ICRG~oc | 0.1132 -0.3035 0.1093 -0.1114 0.3270 | 0 avg5_ICRG~ic | 0.3112 0.1388 -0.0583 0.0862 -0.0214 | 0 a~xtconflict | 0.2973 0.0907 -0.0204 0.2259 0.0301 | 0 avg5_ICRG_~b | 0.0128 0.2754 -0.0934 -0.1157 0.3548 | 0 a~ntconflict | -0.3074 -0.2819 0.2392 -0.6798 -0.2721 | 0 avg5_ICRG_~f | 0.0886 -0.0267 0.2295 0.2883 -0.5471 | 0 avg5_ICRG_~w | -0.2683 -0.5820 -0.1146 0.5154 0.0709 | 0 avg5_ICRG_~y | -0.6851 0.5140 -0.0005 0.1550 0.1527 | 0 avg5_ICRG_~s | 0.0124 -0.0072 -0.1324 0.0371 0.0481 | 0 avg5_ICRG~ec | 0.2773 -0.0697 0.1748 -0.1506 0.4929 | 0 ------------------------------------------------------------------------------ . predict comp1a comp2a comp3a, score (9 components skipped) Scoring coefficients sum of squares(column-loading) = 1 ------------------------------------------------------------------------------------ Variable | Comp1 Comp2 Comp3 Comp4 Comp5 Comp6 Comp7 -------------+---------------------------------------------------------------------- avg5_ICRG_~u | 0.3286 -0.2978 -0.1999 -0.0001 -0.0615 -0.0550 0.0354 avg5_ICRG~pt | 0.2807 -0.4260 0.0969 -0.2069 -0.2636 -0.1840 0.4022 avg5_ICRG~oc | 0.2816 -0.2667 0.0199 0.5380 -0.1790 0.5189 0.1494 avg5_ICRG~ic | 0.2580 0.2447 0.3157 -0.4853 -0.0187 0.5958 -0.2360 a~xtconflict | 0.2678 0.2634 0.2434 0.4000 -0.3693 -0.4172 -0.4246 avg5_ICRG_~b | 0.2056 0.6032 -0.2827 -0.0109 -0.1272 -0.0438 0.5213 a~ntconflict | 0.3404 0.2312 0.1453 -0.0857 -0.0955 -0.1028 -0.1232 avg5_ICRG_~f | 0.2865 0.2228 -0.4344 0.2625 0.3765 0.1587 -0.0064 avg5_ICRG_~w | 0.3381 -0.0102 -0.0251 -0.3299 -0.1850 -0.1475 0.1601 avg5_ICRG_~y | 0.3319 -0.1678 -0.0030 0.0809 0.0339 0.1187 -0.2429 avg5_ICRG_~s | 0.2082 0.0266 0.6210 0.1641 0.6245 -0.1714 0.3204 avg5_ICRG~ec | 0.2959 -0.1829 -0.3396 -0.2316 0.4069 -0.2490 -0.3233 ------------------------------------------------------------------------------------ ---------------------------------------------------------------- Variable | Comp8 Comp9 Comp10 Comp11 Comp12 -------------+-------------------------------------------------- avg5_ICRG_~u | 0.1572 0.0846 -0.7695 -0.2359 -0.2750 avg5_ICRG~pt | 0.2373 0.3306 0.4607 0.0354 -0.2080 avg5_ICRG~oc | 0.1132 -0.3035 0.1093 -0.1114 0.3270 avg5_ICRG~ic | 0.3112 0.1388 -0.0583 0.0862 -0.0214 a~xtconflict | 0.2973 0.0907 -0.0204 0.2259 0.0301 avg5_ICRG_~b | 0.0128 0.2754 -0.0934 -0.1157 0.3548 a~ntconflict | -0.3074 -0.2819 0.2392 -0.6798 -0.2721 avg5_ICRG_~f | 0.0886 -0.0267 0.2295 0.2883 -0.5471 avg5_ICRG_~w | -0.2683 -0.5820 -0.1146 0.5154 0.0709 avg5_ICRG_~y | -0.6851 0.5140 -0.0005 0.1550 0.1527 avg5_ICRG_~s | 0.0124 -0.0072 -0.1324 0.0371 0.0481 avg5_ICRG~ec | 0.2773 -0.0697 0.1748 -0.1506 0.4929 ---------------------------------------------------------------- . gen inst2 = 0.50477*comp1a + 0.1087*comp2a + 0.0910*comp3a (1222 missing values generated) . gen inst3 = 0.25*avg5_ICRG_govstab + 0.25*avg5_ICRG_corrupt + 0.25*avg5_ICRG_law + 0.25*a > vg5_ICRG_bureau (1222 missing values generated) . pca avg5_ICRG_bureau avg5_ICRG_corrupt avg5_ICRG_ethnic avg5_ICRG_extconflict avg5_ICRG_g > ovstab avg5_ICRG_intconflict avg5_ICRG_invprof avg5_ICRG_law avg5_ICRG_military avg5_ICRG > _religious avg5_ICRG_socioec Principal components/correlation Number of obs = 684 Number of comp. = 11 Trace = 11 Rotation: (unrotated = principal) Rho = 1.0000 -------------------------------------------------------------------------- Component | Eigenvalue Difference Proportion Cumulative -------------+------------------------------------------------------------ Comp1 | 5.81958 4.56573 0.5291 0.5291 Comp2 | 1.25385 .17607 0.1140 0.6430 Comp3 | 1.07778 .450971 0.0980 0.7410 Comp4 | .626807 .0617609 0.0570 0.7980 Comp5 | .565046 .128992 0.0514 0.8494 Comp6 | .436054 .090249 0.0396 0.8890 Comp7 | .345805 .0466954 0.0314 0.9204 Comp8 | .29911 .0790867 0.0272 0.9476 Comp9 | .220023 .0319341 0.0200 0.9676 Comp10 | .188089 .0202297 0.0171 0.9847 Comp11 | .167859 . 0.0153 1.0000 -------------------------------------------------------------------------- Principal components (eigenvectors) ------------------------------------------------------------------------------------ Variable | Comp1 Comp2 Comp3 Comp4 Comp5 Comp6 Comp7 -------------+---------------------------------------------------------------------- avg5_ICRG_~u | 0.3354 -0.3280 -0.1916 0.0390 -0.1222 0.0420 0.0855 avg5_ICRG~pt | 0.2858 -0.4680 0.1102 -0.2190 -0.2168 0.4333 0.1727 avg5_ICRG~ic | 0.2751 0.2018 0.3164 -0.5520 0.4168 -0.3799 0.1669 a~xtconflict | 0.2782 0.2874 0.2346 0.2488 -0.6704 -0.3042 0.3577 avg5_ICRG_~b | 0.2240 0.5924 -0.2948 -0.1487 -0.0317 0.5132 -0.0388 a~ntconflict | 0.3586 0.2016 0.1436 -0.0928 -0.0839 -0.0872 -0.1828 avg5_ICRG_~f | 0.2971 0.2250 -0.4411 0.3724 0.2496 -0.0540 -0.0009 avg5_ICRG_~w | 0.3551 -0.0696 -0.0185 -0.3269 -0.0525 0.2026 -0.1231 avg5_ICRG_~y | 0.3406 -0.1878 0.0018 0.1433 -0.0604 -0.2831 -0.7434 avg5_ICRG_~s | 0.2161 0.0265 0.6211 0.5173 0.3823 0.3588 0.0357 avg5_ICRG~ec | 0.3112 -0.2590 -0.3284 0.1400 0.3034 -0.2232 0.4508 ------------------------------------------------------------------------------------ -------------------------------------------------------------------- Variable | Comp8 Comp9 Comp10 Comp11 | Unexplained -------------+----------------------------------------+------------- avg5_ICRG_~u | 0.2679 -0.7200 0.2672 -0.2369 | 0 avg5_ICRG~pt | 0.3066 0.4890 -0.1535 -0.1342 | 0 avg5_ICRG~ic | 0.3517 -0.0568 -0.0695 0.0333 | 0 a~xtconflict | 0.0537 0.0035 -0.1080 0.1966 | 0 avg5_ICRG_~b | 0.1800 0.0486 0.3514 0.2565 | 0 a~ntconflict | -0.4269 0.1901 0.2938 -0.6699 | 0 avg5_ICRG_~f | 0.1937 0.0836 -0.5753 -0.2953 | 0 avg5_ICRG_~w | -0.5614 -0.2887 -0.4657 0.2959 | 0 avg5_ICRG_~y | 0.2041 0.1711 0.1219 0.3255 | 0 avg5_ICRG_~s | -0.0320 -0.1137 0.0408 0.0869 | 0 avg5_ICRG~ec | -0.3124 0.2613 0.3383 0.2885 | 0 -------------------------------------------------------------------- . gen inst4 = 0.5128*comp1 + 0.1118*comp2 + 0.0992*comp3 (1222 missing values generated) . gen linst = ln(4+inst) (1222 missing values generated) . gen linst2 = ln(4+inst2) (1224 missing values generated) . gen linst3 = ln(4+inst3) (1222 missing values generated) . gen linst4 = ln(4+inst4) (1222 missing values generated) . . gen elf = 0.5*avg5_frac_ethnic + 0.5*avg5_frac_lang (474 missing values generated) . gen elrf = 0.33*avg5_frac_ethnic + 0.33*avg5_frac_lang + 0.33*avg5_frac_rel (474 missing values generated) . gen lgeog = ln(1+avg5_JID_disteq) (1082 missing values generated) . gen ltrade = ln(avg5_trade_ratio_nom) (864 missing values generated) . gen ldemoc = ln(avg5_ICRG_democ) (1229 missing values generated) . gen lelf = ln(1+elf) (474 missing values generated) . gen lelrf = ln(1+elrf) (474 missing values generated) . gen lenroll1900 = ln(enroll1900) (1609 missing values generated) . gen ltrust1 = ln(1+avg5_Rel_wv00trust) (1362 missing values generated) . gen ltrust2 = ln(1+avg5_Rel_is98trust) (1674 missing values generated) . gen lsch = ln(1+avg5_yr_sch) (760 missing values generated) . reg avg5_JID_democ1995 avg5_JID_democ1990 avg5_JID_democ1985 avg5_JID_rgdppc avg5_JID_pmf > gap, robust Linear regression Number of obs = 744 F( 4, 739) = 381.84 Prob > F = 0.0000 R-squared = 0.6749 Root MSE = 2.0329 ------------------------------------------------------------------------------ | Robust avg5_J~c1995 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- avg5_JID_d~0 | .6375556 .0310482 20.53 0.000 .5766024 .6985087 avg5_JI~1985 | -.0455533 .0178245 -2.56 0.011 -.080546 -.0105606 avg5_JID_r~c | 5.57e-06 7.32e-06 0.76 0.447 -8.81e-06 .0000199 avg5_JID_~ap | -.0667665 .0125146 -5.34 0.000 -.0913348 -.0421981 _cons | 3.329651 .2186876 15.23 0.000 2.900328 3.758973 ------------------------------------------------------------------------------ . predict pldemoc (option xb assumed; fitted values) (1146 missing values generated) . . encode code, gen(code2) . xtset code2 countperiod panel variable: code2 (unbalanced) time variable: countperiod, 1 to 8 . save "C:\Llama\poledet.dta", replace file C:\Llama\poledet.dta saved . . /* Provide descriptive statistics */ . gsort -index_constusd_HPg . gen rank = _n if index_constusd_HPg ~= . (1146 missing values generated) . gen lrank = ln(rank) if rank ~= . (1146 missing values generated) . graph twoway bar index_constusd_HPg rank . graph twoway (lfit lindex lrank, lcolor(maroon)) || scatter lindex lrank, note("Source: A > uthors' calculations.") ytitle("Logarithm of synthetic growth polarity index") xtitle("Lo > garithm of rank") mcolor(navy) msymbol (o) graphregion(fcolor(white) lcolor(white) ifcolo > r(white) ilcolor(white)) legend(off) . graph export "C:\Llama\Output\powerdist.eps", replace (file C:\Llama\Output\powerdist.eps written in EPS format) . reg lrank lindex Source | SS df MS Number of obs = 760 -------------+------------------------------ F( 1, 758) = 5758.58 Model | 647.075407 1 647.075407 Prob > F = 0.0000 Residual | 85.1743621 758 .112367232 R-squared = 0.8837 -------------+------------------------------ Adj R-squared = 0.8835 Total | 732.249769 759 .964755954 Root MSE = .33521 ------------------------------------------------------------------------------ lrank | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex | -2.391519 .0315149 -75.89 0.000 -2.453386 -2.329652 _cons | 6.756131 .0190948 353.82 0.000 6.718646 6.793616 ------------------------------------------------------------------------------ . histogram index_constusd_HPg, kdensity note("Source: Authors' calculations.") ytitle("Den > sity function") xtitle("Synthetic growth polarity index") legend (off) graphregion(fcolor > (white) lcolor(white) ifcolor(white) ilcolor(white)) (bin=27, start=-.70057207, width=.54593881) . histogram lindex, kdensity note("Source: Authors' calculations.") ytitle("Density functio > n") xtitle("Logarithm of synthetic growth polarity index") legend (off) graphregion(fcolo > r(white) lcolor(white) ifcolor(white) ilcolor(white)) (bin=27, start=.09478997, width=.09880497) . graph export "C:\Llama\Output\histkdensity.eps", replace (file C:\Llama\Output\histkdensity.eps written in EPS format) . pwcorr index_constusd_HPg index_constusd_migr_HPg GDP_constusd growth_constusd, sig | in~d_HPg index_.. GDP_co~d growth~d -------------+------------------------------------ index_~d_HPg | 1.0000 | | i~d_migr_HPg | 0.9770 1.0000 | 0.0000 | GDP_constusd | 0.7699 0.7757 1.0000 | 0.0000 0.0000 | growth_con~d | 0.0272 0.0582 -0.0314 1.0000 | 0.4558 0.1129 0.2628 | . pwcorr index_ppp_HPg GDP_constusd growth_ppp, sig | in~p_HPg GDP_co~d growth~p -------------+--------------------------- index_~p_HPg | 1.0000 | | GDP_constusd | 0.8129 1.0000 | 0.0000 | growth_ppp | 0.0225 -0.0293 1.0000 | 0.5684 0.3651 | . . /* Proximate correlates regressions */ . /* Benchmark proximates */ . reg lindex ln liy lenroll_p period2-period8, robust Linear regression Number of obs = 676 F( 10, 665) = 3.44 Prob > F = 0.0002 R-squared = 0.0927 Root MSE = .3582 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .2561551 .2597062 0.99 0.324 -.2537878 .7660979 liy | .4560867 .2286743 1.99 0.047 .0070761 .9050974 lenroll_p | .0976361 .0358748 2.72 0.007 .0271947 .1680776 period2 | -.4350856 .203478 -2.14 0.033 -.8346223 -.0355489 period3 | -.5202363 .2001335 -2.60 0.010 -.9132061 -.1272665 period4 | -.480546 .2023253 -2.38 0.018 -.8778193 -.0832727 period5 | -.5077557 .200395 -2.53 0.012 -.9012389 -.1142726 period6 | -.5275161 .1993057 -2.65 0.008 -.9188604 -.1361718 period7 | -.5391351 .1988339 -2.71 0.007 -.9295529 -.1487172 period8 | -.5513245 .198217 -2.78 0.006 -.940531 -.162118 _cons | .3964628 .2695969 1.47 0.142 -.1329009 .9258266 ------------------------------------------------------------------------------ . testparm period2-period8 /* Test for whether time FE should be kept */ ( 1) period2 = 0 ( 2) period3 = 0 ( 3) period4 = 0 ( 4) period5 = 0 ( 5) period6 = 0 ( 6) period7 = 0 ( 7) period8 = 0 F( 7, 665) = 1.77 Prob > F = 0.0910 . estimates store P1 . xtserial lindex ln liy lenroll_p period2-period8 /* Test for presence of autocorrelation > */ Wooldridge test for autocorrelation in panel data H0: no first order autocorrelation F( 1, 110) = 13.869 Prob > F = 0.0003 . xtreg lindex ln liy lenroll_p period2-period8, fe Fixed-effects (within) regression Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.1894 Obs per group: min = 1 between = 0.0056 avg = 5.1 overall = 0.0524 max = 8 F(10,533) = 12.45 corr(u_i, Xb) = 0.0332 Prob > F = 0.0000 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.079705 .0421903 -1.89 0.059 -.1625845 .0031746 liy | .7076098 .1751349 4.04 0.000 .3635705 1.051649 lenroll_p | -.080648 .055313 -1.46 0.145 -.1893062 .0280102 period2 | -.4253131 .0545463 -7.80 0.000 -.5324652 -.3181611 period3 | -.4930247 .0543049 -9.08 0.000 -.5997025 -.3863468 period4 | -.4432516 .0544329 -8.14 0.000 -.550181 -.3363222 period5 | -.4654278 .0542921 -8.57 0.000 -.5720805 -.3587751 period6 | -.4650212 .0541912 -8.58 0.000 -.5714757 -.3585668 period7 | -.4619254 .0543583 -8.50 0.000 -.5687082 -.3551426 period8 | -.4777168 .0543851 -8.78 0.000 -.5845522 -.3708815 _cons | 1.151675 .2531614 4.55 0.000 .654358 1.648991 -------------+---------------------------------------------------------------- sigma_u | .29884469 sigma_e | .14675966 rho | .80569176 (fraction of variance due to u_i) ------------------------------------------------------------------------------ F test that all u_i=0: F(132, 533) = 25.97 Prob > F = 0.0000 . xttest3 /* Test for heteroskedasticity in panel */ Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (133) = 8.1e+30 Prob>chi2 = 0.0000 . estimates store fixed . xtreg lindex ln liy lenroll_p period2-period8, re Random-effects GLS regression Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.1876 Obs per group: min = 1 between = 0.0114 avg = 5.1 overall = 0.0639 max = 8 Random effects u_i ~ Gaussian Wald chi2(10) = 118.08 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0548398 .0429894 -1.28 0.202 -.1390975 .0294178 liy | .6101123 .1651478 3.69 0.000 .2864286 .9337961 lenroll_p | -.0371451 .0501488 -0.74 0.459 -.135435 .0611448 period2 | -.407725 .0551262 -7.40 0.000 -.5157705 -.2996796 period3 | -.4781888 .0548559 -8.72 0.000 -.5857043 -.3706733 period4 | -.4313929 .0549657 -7.85 0.000 -.5391237 -.323662 period5 | -.4538608 .0547929 -8.28 0.000 -.5612528 -.3464687 period6 | -.4561739 .0546317 -8.35 0.000 -.56325 -.3490978 period7 | -.45544 .0547245 -8.32 0.000 -.562698 -.348182 period8 | -.4710961 .0547397 -8.61 0.000 -.5783839 -.3638083 _cons | .9246886 .2309812 4.00 0.000 .4719738 1.377403 -------------+---------------------------------------------------------------- sigma_u | .25193093 sigma_e | .14675966 rho | .7466297 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store random . hausman fixed random /* Test for choice between FE and RE */ ---- Coefficients ---- | (b) (B) (b-B) sqrt(diag(V_b-V_B)) | fixed random Difference S.E. -------------+---------------------------------------------------------------- ln | -.079705 -.0548398 -.0248651 . liy | .7076098 .6101123 .0974975 .0582961 lenroll_p | -.080648 -.0371451 -.0435029 .0233371 period2 | -.4253131 -.407725 -.0175881 . period3 | -.4930247 -.4781888 -.0148359 . period4 | -.4432516 -.4313929 -.0118588 . period5 | -.4654278 -.4538608 -.011567 . period6 | -.4650212 -.4561739 -.0088474 . period7 | -.4619254 -.45544 -.0064854 . period8 | -.4777168 -.4710961 -.0066207 . ------------------------------------------------------------------------------ b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(10) = (b-B)'[(V_b-V_B)^(-1)](b-B) = -6.46 chi2<0 ==> model fitted on these data fails to meet the asymptotic assumptions of the Hausman test; see suest for a generalized test . xtreg lindex ln liy lenroll_p period2-period8, re robust Random-effects GLS regression Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.1876 Obs per group: min = 1 between = 0.0114 avg = 5.1 overall = 0.0639 max = 8 Random effects u_i ~ Gaussian Wald chi2(10) = 41.00 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0548398 .08509 -0.64 0.519 -.2216131 .1119335 liy | .6101123 .1357749 4.49 0.000 .3439983 .8762263 lenroll_p | -.0371451 .0458314 -0.81 0.418 -.126973 .0526829 period2 | -.407725 .1430985 -2.85 0.004 -.6881929 -.1272571 period3 | -.4781888 .138623 -3.45 0.001 -.7498848 -.2064927 period4 | -.4313929 .1397451 -3.09 0.002 -.7052882 -.1574976 period5 | -.4538608 .1396985 -3.25 0.001 -.7276648 -.1800567 period6 | -.4561739 .1392829 -3.28 0.001 -.7291632 -.1831845 period7 | -.45544 .1398157 -3.26 0.001 -.7294738 -.1814062 period8 | -.4710961 .1396776 -3.37 0.001 -.7448592 -.1973331 _cons | .9246886 .2440333 3.79 0.000 .4463922 1.402985 -------------+---------------------------------------------------------------- sigma_u | .25193093 sigma_e | .14675966 rho | .7466297 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . xtregar lindex ln liy lenroll_p period2-period8, re RE GLS regression with AR(1) disturbances Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.1841 Obs per group: min = 1 between = 0.0150 avg = 5.1 overall = 0.0681 max = 8 Wald chi2(11) = 105.09 corr(u_i, Xb) = 0 (assumed) Prob > chi2 = 0.0000 ------------------- theta -------------------- min 5% median 95% max 0.5235 0.6268 0.7197 0.7338 0.7338 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0271698 .039597 -0.69 0.493 -.1047784 .0504388 liy | .6154396 .1592373 3.86 0.000 .3033403 .9275389 lenroll_p | -.0129988 .0538079 -0.24 0.809 -.1184604 .0924628 period2 | -.3330604 .0492547 -6.76 0.000 -.4295978 -.236523 period3 | -.4074602 .0507798 -8.02 0.000 -.5069868 -.3079336 period4 | -.3596056 .051545 -6.98 0.000 -.460632 -.2585791 period5 | -.3868265 .051591 -7.50 0.000 -.4879431 -.2857099 period6 | -.3859475 .0516258 -7.48 0.000 -.4871322 -.2847627 period7 | -.3871208 .0518426 -7.47 0.000 -.4887303 -.2855112 period8 | -.4021654 .051957 -7.74 0.000 -.5039992 -.3003315 _cons | .7399573 .2463776 3.00 0.003 .2570661 1.222849 -------------+---------------------------------------------------------------- rho_ar | .4211582 (estimated autocorrelation coefficient) sigma_u | .2973745 sigma_e | .14617479 rho_fov | .80539774 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . xtreg lindex ln liy lenroll_p period2-period8, re robust cluster(code2) /* rE regression > with heterskedasticity and autocorrelation correction */ Random-effects GLS regression Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.1876 Obs per group: min = 1 between = 0.0114 avg = 5.1 overall = 0.0639 max = 8 Random effects u_i ~ Gaussian Wald chi2(10) = 40.27 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 133 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0548398 .0672477 -0.82 0.415 -.1866429 .0769633 liy | .6101123 .1697574 3.59 0.000 .277394 .9428307 lenroll_p | -.0371451 .0554399 -0.67 0.503 -.1458054 .0715152 period2 | -.407725 .1264897 -3.22 0.001 -.6556403 -.1598098 period3 | -.4781888 .1380414 -3.46 0.001 -.748745 -.2076325 period4 | -.4313929 .1358347 -3.18 0.001 -.6976239 -.1651618 period5 | -.4538608 .1383801 -3.28 0.001 -.7250808 -.1826407 period6 | -.4561739 .1368568 -3.33 0.001 -.7244083 -.1879395 period7 | -.45544 .1372229 -3.32 0.001 -.724392 -.186488 period8 | -.4710961 .1401349 -3.36 0.001 -.7457555 -.1964368 _cons | .9246886 .3075776 3.01 0.003 .3218477 1.52753 -------------+---------------------------------------------------------------- sigma_u | .25193093 sigma_e | .14675966 rho | .7466297 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . predict epsilon, e (1230 missing values generated) . qnorm epsilon, note("Source: Authors' calculations.") ytitle("Residuals") legend (off) gr > aphregion(fcolor(white) lcolor(white) ifcolor(white) ilcolor(white)) /* Test for normalit > y of residuals */ . graph export "C:\Llama\Output\qplotp2.eps", replace (file C:\Llama\Output\qplotp2.eps written in EPS format) . estimates store P2 . xtabond2 lindex ln liy lenroll_p period2-period8, gmm(l(1/4).lindex ln liy lenroll_p) iv( > period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 676 Time variable : countperiod Number of groups = 133 Number of instruments = 152 Obs per group: min = 1 F(10, 132) = 3.02 avg = 5.08 Prob > F = 0.002 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .03186 .14058 0.23 0.821 -.2462212 .3099412 liy | .585234 .4241886 1.38 0.170 -.2538529 1.424321 lenroll_p | .3068326 .1295251 2.37 0.019 .0506191 .5630462 period2 | -.5653957 .2173193 -2.60 0.010 -.9952749 -.1355166 period3 | -.6545969 .2289636 -2.86 0.005 -1.107509 -.2016843 period4 | -.6241266 .2327378 -2.68 0.008 -1.084505 -.1637483 period5 | -.6545002 .2395013 -2.73 0.007 -1.128257 -.180743 period6 | -.6729057 .2384101 -2.82 0.006 -1.144504 -.201307 period7 | -.6924882 .2403104 -2.88 0.005 -1.167846 -.2171304 period8 | -.7133352 .2418963 -2.95 0.004 -1.19183 -.2348405 _cons | -.4139855 .6313157 -0.66 0.513 -1.66279 .8348194 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.60 Pr > z = 0.110 Arellano-Bond test for AR(2) in first differences: z = -0.21 Pr > z = 0.834 Hansen test of overid. restrictions: chi2(141) = 128.94 Prob > chi2 = 0.758 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(102) = 100.34 Prob > chi2 = 0.528 Difference (null H = exogenous): chi2(39) = 28.60 Prob > chi2 = 0.890 ivstyle(period*) Hansen test excluding group: chi2(134) = 126.54 Prob > chi2 = 0.664 Difference (null H = exogenous): chi2(7) = 2.40 Prob > chi2 = 0.935 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store P3 . . reg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robust Linear regression Number of obs = 665 F( 14, 650) = 12.11 Prob > F = 0.0000 R-squared = 0.3258 Root MSE = .31174 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .2585858 .2360579 1.10 0.274 -.2049422 .7221138 liy | .1634637 .2448335 0.67 0.505 -.3172964 .6442238 lenroll_p | -.1319527 .0415896 -3.17 0.002 -.2136189 -.0502865 lgdppc | .141732 .0201165 7.05 0.000 .1022309 .1812332 llifeexp | -.4850222 .1278525 -3.79 0.000 -.736076 -.2339685 ldepratio1 | -.2705976 .0638208 -4.24 0.000 -.3959173 -.1452779 lgovt_Y3 | -.7679464 .2137831 -3.59 0.000 -1.187735 -.3481575 period2 | -.3154427 .1730069 -1.82 0.069 -.6551624 .0242771 period3 | -.3931896 .1700367 -2.31 0.021 -.727077 -.0593021 period4 | -.3553096 .1721649 -2.06 0.039 -.6933762 -.017243 period5 | -.3906771 .1702617 -2.29 0.022 -.7250065 -.0563477 period6 | -.408231 .1698985 -2.40 0.017 -.7418472 -.0746149 period7 | -.4356818 .1696048 -2.57 0.010 -.7687213 -.1026424 period8 | -.4803464 .1690017 -2.84 0.005 -.8122016 -.1484912 _cons | 3.547987 .5965069 5.95 0.000 2.376674 4.7193 ------------------------------------------------------------------------------ . testparm period2-period8 ( 1) period2 = 0 ( 2) period3 = 0 ( 3) period4 = 0 ( 4) period5 = 0 ( 5) period6 = 0 ( 6) period7 = 0 ( 7) period8 = 0 F( 7, 650) = 2.56 Prob > F = 0.0133 . estimates store P4 . xtserial lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8 Wooldridge test for autocorrelation in panel data H0: no first order autocorrelation F( 1, 109) = 8.793 Prob > F = 0.0037 . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, fe Fixed-effects (within) regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2724 Obs per group: min = 1 between = 0.2600 avg = 5.2 overall = 0.2880 max = 8 F(14,522) = 13.96 corr(u_i, Xb) = -0.1681 Prob > F = 0.0000 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0852773 .0408055 -2.09 0.037 -.1654404 -.0051141 liy | .5360184 .1783167 3.01 0.003 .1857119 .886325 lenroll_p | -.0050618 .0594153 -0.09 0.932 -.1217843 .1116607 lgdppc | .1480444 .0369096 4.01 0.000 .0755348 .2205541 llifeexp | -.3017728 .2027695 -1.49 0.137 -.7001174 .0965718 ldepratio1 | -.3237556 .0870572 -3.72 0.000 -.4947811 -.15273 lgovt_Y3 | -1.064784 .2403887 -4.43 0.000 -1.537032 -.592536 period2 | -.4573118 .0535454 -8.54 0.000 -.5625027 -.3521209 period3 | -.531583 .0545586 -9.74 0.000 -.6387645 -.4244016 period4 | -.4943195 .056621 -8.73 0.000 -.6055525 -.3830864 period5 | -.5424648 .0585807 -9.26 0.000 -.6575478 -.4273819 period6 | -.5650856 .0602019 -9.39 0.000 -.6833533 -.4468179 period7 | -.5948167 .0634682 -9.37 0.000 -.7195012 -.4701322 period8 | -.6507884 .06727 -9.67 0.000 -.7829415 -.5186353 _cons | 2.532751 .9616134 2.63 0.009 .6436434 4.421859 -------------+---------------------------------------------------------------- sigma_u | .27515385 sigma_e | .14048539 rho | .79322142 (fraction of variance due to u_i) ------------------------------------------------------------------------------ F test that all u_i=0: F(128, 522) = 20.93 Prob > F = 0.0000 . xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (129) = 1.0e+06 Prob>chi2 = 0.0000 . estimates store fixed . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2680 Obs per group: min = 1 between = 0.2655 avg = 5.2 overall = 0.3001 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 238.14 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0474598 .0412718 -1.15 0.250 -.1283511 .0334315 liy | .5005293 .1681421 2.98 0.003 .1709769 .8300818 lenroll_p | -.040251 .0536953 -0.75 0.453 -.1454918 .0649899 lgdppc | .11866 .0179667 6.60 0.000 .083446 .1538741 llifeexp | -.4192052 .1690496 -2.48 0.013 -.7505362 -.0878742 ldepratio1 | -.2735264 .0737398 -3.71 0.000 -.4180538 -.128999 lgovt_Y3 | -.9979316 .2232987 -4.47 0.000 -1.435589 -.5602741 period2 | -.4146266 .0529985 -7.82 0.000 -.5185018 -.3107514 period3 | -.4815418 .0530066 -9.08 0.000 -.5854328 -.3776507 period4 | -.43844 .0537865 -8.15 0.000 -.5438596 -.3330203 period5 | -.4779716 .054215 -8.82 0.000 -.5842311 -.3717122 period6 | -.4944398 .0545979 -9.06 0.000 -.6014497 -.3874299 period7 | -.5159391 .0555596 -9.29 0.000 -.624834 -.4070442 period8 | -.5626319 .0565286 -9.95 0.000 -.673426 -.4518378 _cons | 3.125001 .757269 4.13 0.000 1.640781 4.60922 -------------+---------------------------------------------------------------- sigma_u | .21689823 sigma_e | .14048539 rho | .70446461 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store random . hausman fixed random ---- Coefficients ---- | (b) (B) (b-B) sqrt(diag(V_b-V_B)) | fixed random Difference S.E. -------------+---------------------------------------------------------------- ln | -.0852773 -.0474598 -.0378175 . liy | .5360184 .5005293 .0354891 .0593724 lenroll_p | -.0050618 -.040251 .0351891 .0254361 lgdppc | .1480444 .11866 .0293844 .0322415 llifeexp | -.3017728 -.4192052 .1174324 .111972 ldepratio1 | -.3237556 -.2735264 -.0502292 .0462752 lgovt_Y3 | -1.064784 -.9979316 -.0668525 .089019 period2 | -.4573118 -.4146266 -.0426852 .0076331 period3 | -.531583 -.4815418 -.0500413 .0129205 period4 | -.4943195 -.43844 -.0558795 .0176904 period5 | -.5424648 -.4779716 -.0644932 .0221909 period6 | -.5650856 -.4944398 -.0706458 .025364 period7 | -.5948167 -.5159391 -.0788776 .0306813 period8 | -.6507884 -.5626319 -.0881565 .0364659 ------------------------------------------------------------------------------ b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(14) = (b-B)'[(V_b-V_B)^(-1)](b-B) = -25.71 chi2<0 ==> model fitted on these data fails to meet the asymptotic assumptions of the Hausman test; see suest for a generalized test . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re rob > ust Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2680 Obs per group: min = 1 between = 0.2655 avg = 5.2 overall = 0.3001 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 85.85 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0474598 .0837088 -0.57 0.571 -.211526 .1166065 liy | .5005293 .1577204 3.17 0.002 .1914029 .8096557 lenroll_p | -.040251 .0440764 -0.91 0.361 -.1266391 .0461372 lgdppc | .11866 .030624 3.87 0.000 .0586381 .178682 llifeexp | -.4192052 .1738528 -2.41 0.016 -.7599504 -.07846 ldepratio1 | -.2735264 .087064 -3.14 0.002 -.4441687 -.102884 lgovt_Y3 | -.9979316 .2759072 -3.62 0.000 -1.5387 -.4571634 period2 | -.4146266 .1333837 -3.11 0.002 -.6760539 -.1531993 period3 | -.4815418 .1307677 -3.68 0.000 -.7378418 -.2252417 period4 | -.43844 .1327595 -3.30 0.001 -.6986438 -.1782362 period5 | -.4779716 .1334465 -3.58 0.000 -.7395219 -.2164214 period6 | -.4944398 .1343395 -3.68 0.000 -.7577403 -.2311393 period7 | -.5159391 .1361622 -3.79 0.000 -.7828122 -.249066 period8 | -.5626319 .1369513 -4.11 0.000 -.8310516 -.2942122 _cons | 3.125001 .675305 4.63 0.000 1.801427 4.448574 -------------+---------------------------------------------------------------- sigma_u | .21689823 sigma_e | .14048539 rho | .70446461 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . xtregar lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re RE GLS regression with AR(1) disturbances Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2674 Obs per group: min = 1 between = 0.2676 avg = 5.2 overall = 0.2991 max = 8 Wald chi2(15) = 195.29 corr(u_i, Xb) = 0 (assumed) Prob > chi2 = 0.0000 ------------------- theta -------------------- min 5% median 95% max 0.4828 0.5969 0.6983 0.7136 0.7136 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.027992 .0386652 -0.72 0.469 -.1037744 .0477905 liy | .5135198 .1654312 3.10 0.002 .1892806 .837759 lenroll_p | -.021671 .0575619 -0.38 0.707 -.1344902 .0911482 lgdppc | .1130338 .0203372 5.56 0.000 .0731735 .152894 llifeexp | -.3701149 .1887416 -1.96 0.050 -.7400417 -.0001881 ldepratio1 | -.2614544 .0826941 -3.16 0.002 -.4235319 -.099377 lgovt_Y3 | -.8655861 .2281145 -3.79 0.000 -1.312682 -.41849 period2 | -.3510713 .0482025 -7.28 0.000 -.4455466 -.2565961 period3 | -.422526 .0499964 -8.45 0.000 -.5205171 -.3245349 period4 | -.3786611 .0517004 -7.32 0.000 -.479992 -.2773302 period5 | -.420747 .0526177 -8.00 0.000 -.5238758 -.3176183 period6 | -.4356128 .0534109 -8.16 0.000 -.5402961 -.3309294 period7 | -.4581637 .054806 -8.36 0.000 -.5655814 -.3507459 period8 | -.5031315 .0561716 -8.96 0.000 -.6132257 -.3930373 _cons | 2.74471 .8423567 3.26 0.001 1.093722 4.395699 -------------+---------------------------------------------------------------- rho_ar | .3876865 (estimated autocorrelation coefficient) sigma_u | .25406379 sigma_e | .14153234 rho_fov | .76316566 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re rob > ust cluster(code2) Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2680 Obs per group: min = 1 between = 0.2655 avg = 5.2 overall = 0.3001 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 75.28 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0474598 .064711 -0.73 0.463 -.174291 .0793715 liy | .5005293 .1754761 2.85 0.004 .1566026 .8444561 lenroll_p | -.040251 .0442764 -0.91 0.363 -.127031 .0465291 lgdppc | .11866 .0345613 3.43 0.001 .0509212 .1863989 llifeexp | -.4192052 .2399453 -1.75 0.081 -.8894894 .051079 ldepratio1 | -.2735264 .1060201 -2.58 0.010 -.481322 -.0657307 lgovt_Y3 | -.9979316 .4115618 -2.42 0.015 -1.804578 -.1912852 period2 | -.4146266 .1186779 -3.49 0.000 -.647231 -.1820222 period3 | -.4815418 .1273349 -3.78 0.000 -.7311136 -.2319699 period4 | -.43844 .124371 -3.53 0.000 -.6822027 -.1946773 period5 | -.4779716 .1278107 -3.74 0.000 -.728476 -.2274672 period6 | -.4944398 .1296448 -3.81 0.000 -.748539 -.2403406 period7 | -.5159391 .1332324 -3.87 0.000 -.7770699 -.2548084 period8 | -.5626319 .1393989 -4.04 0.000 -.8358487 -.2894151 _cons | 3.125001 1.015089 3.08 0.002 1.135462 5.114539 -------------+---------------------------------------------------------------- sigma_u | .21689823 sigma_e | .14048539 rho | .70446461 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . drop epsilon . predict epsilon, e (1241 missing values generated) . qnorm epsilon, note("Source: Authors' calculations.") ytitle("Residuals") legend (off) gr > aphregion(fcolor(white) lcolor(white) ifcolor(white) ilcolor(white)) . graph export "C:\Llama\Output\qplotp5.eps", replace (file C:\Llama\Output\qplotp5.eps written in EPS format) . estimates store P5 . xtabond2 lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, gmm > (l(1/4).lindex ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(period* > ) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 665 Time variable : countperiod Number of groups = 129 Number of instruments = 74 Obs per group: min = 1 F(14, 128) = 3.32 avg = 5.16 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.042566 .0814696 -0.52 0.602 -.2037676 .1186355 liy | .6391906 .3261416 1.96 0.052 -.0061362 1.284518 lenroll_p | -.2646663 .1471627 -1.80 0.074 -.5558528 .0265202 lgdppc | .1651671 .0617187 2.68 0.008 .0430461 .2872881 llifeexp | -.5694345 .5638056 -1.01 0.314 -1.68502 .5461511 ldepratio1 | -.5690924 .2162794 -2.63 0.010 -.9970382 -.1411466 lgovt_Y3 | -1.467291 .7978098 -1.84 0.068 -3.045894 .1113118 period2 | -.4759565 .1563786 -3.04 0.003 -.7853782 -.1665347 period3 | -.5577233 .1609805 -3.46 0.001 -.8762508 -.2391958 period4 | -.5163149 .1602885 -3.22 0.002 -.833473 -.1991568 period5 | -.5713913 .1668423 -3.42 0.001 -.9015173 -.2412654 period6 | -.6001106 .1625167 -3.69 0.000 -.9216777 -.2785436 period7 | -.6436595 .1685225 -3.82 0.000 -.97711 -.310209 period8 | -.7162655 .178832 -4.01 0.000 -1.070115 -.3624159 _cons | 5.8025 2.645024 2.19 0.030 .5688683 11.03613 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.32 Pr > z = 0.187 Arellano-Bond test for AR(2) in first differences: z = -0.62 Pr > z = 0.537 Hansen test of overid. restrictions: chi2(59) = 57.71 Prob > chi2 = 0.523 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 41.49 Prob > chi2 = 0.735 Difference (null H = exogenous): chi2(11) = 16.23 Prob > chi2 = 0.133 ivstyle(period*) Hansen test excluding group: chi2(52) = 55.32 Prob > chi2 = 0.350 Difference (null H = exogenous): chi2(7) = 2.39 Prob > chi2 = 0.935 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store P6 . gen lgdppcXlgovt_Y3 = lgdppc*lgovt_Y3 /* Explore interaction between government and GDP p > er capita */ (782 missing values generated) . reg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lgdppcXlgovt_Y3 period2-p > eriod8, robust Linear regression Number of obs = 665 F( 15, 649) = 12.26 Prob > F = 0.0000 R-squared = 0.3259 Root MSE = .31195 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .258759 .2358841 1.10 0.273 -.2044292 .7219472 liy | .1768599 .2480974 0.71 0.476 -.3103106 .6640304 lenroll_p | -.1280246 .0454466 -2.82 0.005 -.2172647 -.0387845 lgdppc | .1323587 .0393343 3.36 0.001 .0551208 .2095966 llifeexp | -.4856014 .1275401 -3.81 0.000 -.7360424 -.2351604 ldepratio1 | -.2708286 .063967 -4.23 0.000 -.3964358 -.1452213 lgovt_Y3 | -1.24327 1.147038 -1.08 0.279 -3.495623 1.009083 lgdppcXlgo~3 | .0604071 .1634651 0.37 0.712 -.2605772 .3813914 period2 | -.3158379 .1731941 -1.82 0.069 -.6559262 .0242505 period3 | -.3938882 .1703206 -2.31 0.021 -.7283342 -.0594423 period4 | -.3557988 .172378 -2.06 0.039 -.6942848 -.0173128 period5 | -.3918903 .1707494 -2.30 0.022 -.7271782 -.0566023 period6 | -.4093831 .1704458 -2.40 0.017 -.7440749 -.0746914 period7 | -.4367995 .1701071 -2.57 0.010 -.7708262 -.1027729 period8 | -.4813363 .1693575 -2.84 0.005 -.8138911 -.1487815 _cons | 3.60376 .5926843 6.08 0.000 2.439949 4.76757 ------------------------------------------------------------------------------ . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lgdppcXlgovt_Y3 period2 > -period8, re robust cluster(code2) Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2970 Obs per group: min = 1 between = 0.2375 avg = 5.2 overall = 0.2800 max = 8 Random effects u_i ~ Gaussian Wald chi2(15) = 74.89 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0384099 .0636716 -0.60 0.546 -.163204 .0863842 liy | .4512275 .1718062 2.63 0.009 .1144935 .7879615 lenroll_p | -.0612117 .0481106 -1.27 0.203 -.1555067 .0330833 lgdppc | .2065554 .0664232 3.11 0.002 .0763683 .3367424 llifeexp | -.3934091 .2215998 -1.78 0.076 -.8277366 .0409185 ldepratio1 | -.2534659 .1027074 -2.47 0.014 -.4547687 -.052163 lgovt_Y3 | 3.505409 1.790412 1.96 0.050 -.0037345 7.014553 lgdppcXlgo~3 | -.5933737 .2768558 -2.14 0.032 -1.136001 -.0507463 period2 | -.391237 .1170694 -3.34 0.001 -.6206889 -.1617852 period3 | -.4534912 .1241427 -3.65 0.000 -.6968064 -.210176 period4 | -.4105573 .1222912 -3.36 0.001 -.6502437 -.1708709 period5 | -.4437589 .1251658 -3.55 0.000 -.6890793 -.1984385 period6 | -.4601409 .1272768 -3.62 0.000 -.7095988 -.210683 period7 | -.4800888 .1312881 -3.66 0.000 -.7374087 -.222769 period8 | -.529016 .1369251 -3.86 0.000 -.7973843 -.2606476 _cons | 2.354414 .7852163 3.00 0.003 .8154185 3.89341 -------------+---------------------------------------------------------------- sigma_u | .217331 sigma_e | .13768056 rho | .71360764 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . . estimates drop fixed random . estout * using "C:\Llama\Output\benchmark1.tex", replace style(tex) varlabels(_cons Const > ant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p sargan sarganp ar2 a > r2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Hansen J" "J p-value" "AR(2)" "AR(2) > p" "Instruments" "N") fmt(%8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels(* 0.10 * > * 0.05 *** 0.01) & P1 & P2 & P3 & P4 & P5 > & P6 \\ & b/se & b/se & b/se & b/se & b/se > & b/se \\ ln & 0.256 & -0.055 & 0.032 & 0.259 & -0.047 > & -0.043 \\ & (0.26) & (0.07) & (0.14) & (0.24) & (0.06) > & (0.08) \\ liy & 0.456 & 0.610 & 0.585 & 0.163 & 0.501 > & 0.639 \\ & (0.23)** & (0.17)***& (0.42) & (0.24) & (0.18)** > *& (0.33)* \\ lenroll_p & 0.098 & -0.037 & 0.307 & -0.132 & -0.040 > & -0.265 \\ & (0.04)***& (0.06) & (0.13)** & (0.04)***& (0.04) > & (0.15)* \\ period2 & -0.435 & -0.408 & -0.565 & -0.315 & -0.415 > & -0.476 \\ & (0.20)** & (0.13)***& (0.22)** & (0.17)* & (0.12)** > *& (0.16)***\\ period3 & -0.520 & -0.478 & -0.655 & -0.393 & -0.482 > & -0.558 \\ & (0.20)***& (0.14)***& (0.23)***& (0.17)** & (0.13)** > *& (0.16)***\\ period4 & -0.481 & -0.431 & -0.624 & -0.355 & -0.438 > & -0.516 \\ & (0.20)** & (0.14)***& (0.23)***& (0.17)** & (0.12)** > *& (0.16)***\\ period5 & -0.508 & -0.454 & -0.655 & -0.391 & -0.478 > & -0.571 \\ & (0.20)** & (0.14)***& (0.24)***& (0.17)** & (0.13)** > *& (0.17)***\\ period6 & -0.528 & -0.456 & -0.673 & -0.408 & -0.494 > & -0.600 \\ & (0.20)***& (0.14)***& (0.24)***& (0.17)** & (0.13)** > *& (0.16)***\\ period7 & -0.539 & -0.455 & -0.692 & -0.436 & -0.516 > & -0.644 \\ & (0.20)***& (0.14)***& (0.24)***& (0.17)** & (0.13)** > *& (0.17)***\\ period8 & -0.551 & -0.471 & -0.713 & -0.480 & -0.563 > & -0.716 \\ & (0.20)***& (0.14)***& (0.24)***& (0.17)***& (0.14)** > *& (0.18)***\\ lgdppc & & & & 0.142 & 0.119 > & 0.165 \\ & & & & (0.02)***& (0.03)** > *& (0.06)***\\ llifeexp & & & & -0.485 & -0.419 > & -0.569 \\ & & & & (0.13)***& (0.24)* > & (0.56) \\ ldepratio1 & & & & -0.271 & -0.274 > & -0.569 \\ & & & & (0.06)***& (0.11)** > *& (0.22)***\\ lgovt_Y3 & & & & -0.768 & -0.998 > & -1.467 \\ & & & & (0.21)***& (0.41)** > & (0.80)* \\ Constant & 0.396 & 0.925 & -0.414 & 3.548 & 3.125 > & 5.802 \\ & (0.27) & (0.31)***& (0.63) & (0.60)***& (1.02)** > *& (2.65)** \\ Adjusted R^2& 0.079 & & & 0.311 & > & \\ F & 3.440 & & 3.025 & 12.112 & > & 3.315 \\ F p-value & & & 0.002 & & > & 0.000 \\ Hansen J & & & 128.942 & & > & 57.713 \\ J p-value & & & 0.758 & & > & 0.523 \\ AR(2) & & & -0 & & > & -1 \\ AR(2) p & & & 1 & & > & 1 \\ Instruments & & & 152 & & > & 74 \\ N & 676 & 676 & 676 & 665 & 665 > & 665 \\ . estimates clear . . /* Alternative proximates */ . reg lindex2 ln liy lenroll_p period2-period8, robust Linear regression Number of obs = 587 F( 8, 578) = 4.27 Prob > F = 0.0001 R-squared = 0.0570 Root MSE = .22616 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -4.292034 1.114106 -3.85 0.000 -6.480224 -2.103843 liy | .0913773 .1541842 0.59 0.554 -.2114523 .394207 lenroll_p | .0488189 .023449 2.08 0.038 .0027633 .0948745 period2 | (dropped) period3 | -.0354956 .0461714 -0.77 0.442 -.1261797 .0551885 period4 | (dropped) period5 | -.0342598 .0420013 -0.82 0.415 -.1167536 .0482341 period6 | -.0549049 .0404124 -1.36 0.175 -.1342781 .0244682 period7 | -.0676634 .0394895 -1.71 0.087 -.1452238 .009897 period8 | -.0631852 .0390232 -1.62 0.106 -.1398298 .0134595 _cons | 1.474736 .1851568 7.96 0.000 1.111073 1.838398 ------------------------------------------------------------------------------ . estimates store AP1 . xtreg lindex2 ln liy lenroll_p period2-period8, re robust cluster(code2) note: period2 dropped due to collinearity note: period4 dropped due to collinearity Random-effects GLS regression Number of obs = 587 Group variable (i): code2 Number of groups = 131 R-sq: within = 0.0392 Obs per group: min = 1 between = 0.0168 avg = 4.5 overall = 0.0323 max = 6 Random effects u_i ~ Gaussian Wald chi2(8) = 24.65 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0018 (Std. Err. adjusted for 131 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.9724679 1.215811 -0.80 0.424 -3.355414 1.410479 liy | .3002532 .1129194 2.66 0.008 .0789352 .5215711 lenroll_p | .0184257 .0372163 0.50 0.621 -.054517 .0913683 period3 | -.051654 .0159105 -3.25 0.001 -.082838 -.02047 period5 | -.0212037 .014602 -1.45 0.146 -.0498231 .0074157 period6 | -.0242293 .0175773 -1.38 0.168 -.0586801 .0102216 period7 | -.032454 .0215461 -1.51 0.132 -.0746835 .0097756 period8 | -.0281105 .0227666 -1.23 0.217 -.0727322 .0165111 _cons | 1.179884 .1427774 8.26 0.000 .9000453 1.459722 -------------+---------------------------------------------------------------- sigma_u | .18656204 sigma_e | .09155031 rho | .80592567 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store AP2 . xtabond2 lindex2 ln liy lenroll_p period2-period8, gmm(l(1/4).lindex2 ln liy lgdppc lenro > ll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(period*) robust small orthog noleveleq Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step difference GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 446 Time variable : countperiod Number of groups = 122 Number of instruments = 58 Obs per group: min = 0 F(8, 122) = 3.01 avg = 3.66 Prob > F = 0.004 max = 5 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -1.198774 1.879023 -0.64 0.525 -4.918488 2.520939 liy | .4308181 .2305673 1.87 0.064 -.0256129 .887249 lenroll_p | .0201956 .0727783 0.28 0.782 -.1238763 .1642676 period3 | -.0271037 .0215358 -1.26 0.211 -.069736 .0155286 period4 | .0307491 .0253645 1.21 0.228 -.0194625 .0809607 period5 | .0088916 .0161611 0.55 0.583 -.0231009 .0408841 period6 | .0074829 .0113292 0.66 0.510 -.0149443 .0299101 period7 | -.0018045 .0067697 -0.27 0.790 -.0152058 .0115968 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.78 Pr > z = 0.075 Arellano-Bond test for AR(2) in first differences: z = -0.71 Pr > z = 0.477 Hansen test of overid. restrictions: chi2(50) = 56.25 Prob > chi2 = 0.252 Difference-in-Sargan tests of exogeneity of instrument subsets: ivstyle(period*) Hansen test excluding group: chi2(45) = 49.75 Prob > chi2 = 0.290 Difference (null H = exogenous): chi2(5) = 6.50 Prob > chi2 = 0.261 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AP3 . reg lindex2 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robust Linear regression Number of obs = 574 F( 12, 561) = 10.15 Prob > F = 0.0000 R-squared = 0.2237 Root MSE = .2076 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -1.361205 1.407328 -0.97 0.334 -4.125481 1.40307 liy | .1344403 .1604861 0.84 0.403 -.1807868 .4496674 lenroll_p | -.0526735 .0281754 -1.87 0.062 -.1080157 .0026687 lgdppc | .0904038 .0155728 5.81 0.000 .0598156 .120992 llifeexp | -.3662432 .0855149 -4.28 0.000 -.5342117 -.1982747 ldepratio1 | -.1097917 .0563896 -1.95 0.052 -.2205523 .0009688 lgovt_Y3 | -.6245103 .1658691 -3.77 0.000 -.9503106 -.29871 period2 | (dropped) period3 | -.0510396 .0453311 -1.13 0.261 -.140079 .0379999 period4 | (dropped) period5 | -.02719 .0383668 -0.71 0.479 -.10255 .0481701 period6 | -.0411597 .0374544 -1.10 0.272 -.1147278 .0324083 period7 | -.0575462 .0371069 -1.55 0.122 -.1304317 .0153393 period8 | -.0721994 .0382425 -1.89 0.060 -.1473154 .0029167 _cons | 2.984401 .4177127 7.14 0.000 2.163929 3.804873 ------------------------------------------------------------------------------ . estimates store AP4 . xtreg lindex2 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re ro > bust cluster(code2) note: period2 dropped due to collinearity note: period4 dropped due to collinearity Random-effects GLS regression Number of obs = 574 Group variable (i): code2 Number of groups = 127 R-sq: within = 0.1173 Obs per group: min = 1 between = 0.1668 avg = 4.5 overall = 0.1726 max = 6 Random effects u_i ~ Gaussian Wald chi2(12) = 51.07 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 127 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0417464 2.249453 -0.02 0.985 -4.450594 4.367101 liy | .2964107 .0973391 3.05 0.002 .1056295 .4871918 lenroll_p | .01465 .0304107 0.48 0.630 -.0449539 .0742539 lgdppc | .0284286 .0241435 1.18 0.239 -.0188919 .075749 llifeexp | -.1677662 .1042768 -1.61 0.108 -.3721451 .0366126 ldepratio1 | -.3140858 .1015634 -3.09 0.002 -.5131463 -.1150252 lgovt_Y3 | -.1196815 .2721438 -0.44 0.660 -.6530736 .4137106 period3 | -.0425165 .0139214 -3.05 0.002 -.069802 -.0152309 period5 | -.0302994 .0158632 -1.91 0.056 -.0613908 .000792 period6 | -.0465337 .0222583 -2.09 0.037 -.0901592 -.0029081 period7 | -.0720589 .0291205 -2.47 0.013 -.1291342 -.0149837 period8 | -.0882567 .0350227 -2.52 0.012 -.1568999 -.0196136 _cons | 2.932698 .533433 5.50 0.000 1.887189 3.978208 -------------+---------------------------------------------------------------- sigma_u | .17123097 sigma_e | .08747203 rho | .79304663 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store AP5 . xtabond2 lindex2 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, gm > m(l(1/4).lindex2 ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(perio > d*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 574 Time variable : countperiod Number of groups = 127 Number of instruments = 70 Obs per group: min = 1 F(12, 126) = 4.56 avg = 4.52 Prob > F = 0.000 max = 6 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -4.763755 2.609769 -1.83 0.070 -9.928412 .4009016 liy | .478972 .2431628 1.97 0.051 -.00224 .960184 lenroll_p | -.0800487 .0683413 -1.17 0.244 -.215294 .0551966 lgdppc | .1280366 .0398429 3.21 0.002 .0491887 .2068846 llifeexp | -.5393418 .3868939 -1.39 0.166 -1.304994 .2263099 ldepratio1 | .0541513 .0889644 0.61 0.544 -.1219066 .2302091 lgovt_Y3 | -.9989634 .4576862 -2.18 0.031 -1.904711 -.0932159 period3 | -.010295 .0216908 -0.47 0.636 -.0532205 .0326304 period4 | .0612658 .0267729 2.29 0.024 .008283 .1142485 period5 | .0356919 .0181912 1.96 0.052 -.0003079 .0716918 period6 | .0247824 .0126425 1.96 0.052 -.0002369 .0498016 period7 | .0112873 .0090686 1.24 0.216 -.006659 .0292337 _cons | 3.155498 1.417436 2.23 0.028 .3504343 5.960561 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.43 Pr > z = 0.153 Arellano-Bond test for AR(2) in first differences: z = -0.98 Pr > z = 0.325 Hansen test of overid. restrictions: chi2(57) = 63.36 Prob > chi2 = 0.262 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(46) = 37.28 Prob > chi2 = 0.817 Difference (null H = exogenous): chi2(11) = 26.08 Prob > chi2 = 0.006 ivstyle(period*) Hansen test excluding group: chi2(52) = 61.15 Prob > chi2 = 0.180 Difference (null H = exogenous): chi2(5) = 2.21 Prob > chi2 = 0.819 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AP6 . . estout * using "C:\Llama\Output\benchmark1a.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p sargan sarganp ar2 > ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Hansen J" "J p-value" "AR(2)" "AR(2) > p" "Instruments" "N") fmt(%8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starle > vels(* 0.10 ** 0.05 *** 0.01) & AP1 & AP2 & AP3 & AP4 & AP5 > & AP6 \\ & b/se & b/se & b/se & b/se & b/se > & b/se \\ ln & -4.292 & -0.972 & -1.199 & -1.361 & -0.042 > & -4.764 \\ & (1.11)***& (1.22) & (1.88) & (1.41) & (2.25) > & (2.61)* \\ liy & 0.091 & 0.300 & 0.431 & 0.134 & 0.296 > & 0.479 \\ & (0.15) & (0.11)***& (0.23)* & (0.16) & (0.10)** > *& (0.24)* \\ lenroll_p & 0.049 & 0.018 & 0.020 & -0.053 & 0.015 > & -0.080 \\ & (0.02)** & (0.04) & (0.07) & (0.03)* & (0.03) > & (0.07) \\ period2 & 0.000 & & & 0.000 & > & \\ & (0.00) & & & (0.00) & > & \\ period3 & -0.035 & -0.052 & -0.027 & -0.051 & -0.043 > & -0.010 \\ & (0.05) & (0.02)***& (0.02) & (0.05) & (0.01)** > *& (0.02) \\ period4 & 0.000 & & 0.031 & 0.000 & > & 0.061 \\ & (0.00) & & (0.03) & (0.00) & > & (0.03)** \\ period5 & -0.034 & -0.021 & 0.009 & -0.027 & -0.030 > & 0.036 \\ & (0.04) & (0.01) & (0.02) & (0.04) & (0.02)* > & (0.02)* \\ period6 & -0.055 & -0.024 & 0.007 & -0.041 & -0.047 > & 0.025 \\ & (0.04) & (0.02) & (0.01) & (0.04) & (0.02)** > & (0.01)* \\ period7 & -0.068 & -0.032 & -0.002 & -0.058 & -0.072 > & 0.011 \\ & (0.04)* & (0.02) & (0.01) & (0.04) & (0.03)** > & (0.01) \\ period8 & -0.063 & -0.028 & & -0.072 & -0.088 > & \\ & (0.04) & (0.02) & & (0.04)* & (0.04)** > & \\ lgdppc & & & & 0.090 & 0.028 > & 0.128 \\ & & & & (0.02)***& (0.02) > & (0.04)***\\ llifeexp & & & & -0.366 & -0.168 > & -0.539 \\ & & & & (0.09)***& (0.10) > & (0.39) \\ ldepratio1 & & & & -0.110 & -0.314 > & 0.054 \\ & & & & (0.06)* & (0.10)** > *& (0.09) \\ lgovt_Y3 & & & & -0.625 & -0.120 > & -0.999 \\ & & & & (0.17)***& (0.27) > & (0.46)** \\ Constant & 1.475 & 1.180 & & 2.984 & 2.933 > & 3.155 \\ & (0.19)***& (0.14)***& & (0.42)***& (0.53)** > *& (1.42)** \\ Adjusted R^2& 0.044 & & & 0.207 & > & \\ F & 4.273 & & 3.007 & 10.147 & > & 4.563 \\ F p-value & & & 0.004 & & > & 0.000 \\ Hansen J & & & 56.250 & & > & 63.362 \\ J p-value & & & 0.252 & & > & 0.262 \\ AR(2) & & & -0.711 & & > & -0.984 \\ AR(2) p & & & 0.477 & & > & 0.325 \\ Instruments & & & 58 & & > & 70 \\ N & 587 & 587 & 446 & 574 & 574 > & 574 \\ . estimates clear . . reg lindex3 ln liy lenroll_p period2-period8, robust Linear regression Number of obs = 667 F( 10, 656) = 2.13 Prob > F = 0.0205 R-squared = 0.0755 Root MSE = .21295 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .1744057 .1560075 1.12 0.264 -.1319285 .48074 liy | .1585352 .1161807 1.36 0.173 -.0695957 .386666 lenroll_p | .0016747 .0240529 0.07 0.945 -.0455553 .0489047 period2 | -.1760096 .0968868 -1.82 0.070 -.3662553 .014236 period3 | -.2226801 .094566 -2.35 0.019 -.4083686 -.0369916 period4 | -.2034208 .0961856 -2.11 0.035 -.3922895 -.0145521 period5 | -.220408 .0943203 -2.34 0.020 -.4056141 -.0352019 period6 | -.2624014 .0953412 -2.75 0.006 -.4496121 -.0751907 period7 | -.2349784 .0935251 -2.51 0.012 -.4186232 -.0513337 period8 | -.2404395 .0931201 -2.58 0.010 -.4232888 -.0575902 _cons | 1.718492 .1469778 11.69 0.000 1.429888 2.007095 ------------------------------------------------------------------------------ . estimates store BP1 . xtreg lindex3 ln liy lenroll_p period2-period8, re robust cluster(code2) Random-effects GLS regression Number of obs = 667 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.0982 Obs per group: min = 1 between = 0.0174 avg = 5.0 overall = 0.0550 max = 8 Random effects u_i ~ Gaussian Wald chi2(10) = 36.11 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0001 (Std. Err. adjusted for 133 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0006695 .0401743 -0.02 0.987 -.0794096 .0780706 liy | .2532045 .0998332 2.54 0.011 .0575349 .448874 lenroll_p | -.0413189 .0371284 -1.11 0.266 -.1140892 .0314514 period2 | -.1787922 .0652301 -2.74 0.006 -.3066408 -.0509436 period3 | -.2171202 .0696258 -3.12 0.002 -.3535843 -.0806561 period4 | -.193714 .069499 -2.79 0.005 -.3299296 -.0574984 period5 | -.2097657 .0690849 -3.04 0.002 -.3451696 -.0743618 period6 | -.2405733 .0706064 -3.41 0.001 -.3789593 -.1021872 period7 | -.2069392 .0699859 -2.96 0.003 -.344109 -.0697693 period8 | -.2154069 .0714589 -3.01 0.003 -.3554638 -.07535 _cons | 1.88433 .1786064 10.55 0.000 1.534268 2.234392 -------------+---------------------------------------------------------------- sigma_u | .1290771 sigma_e | .11546001 rho | .55551302 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store BP2 . xtabond2 lindex3 ln liy lenroll_p period2-period8, gmm(l(1/4).lindex3 ln liy lgdppc lenro > ll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 667 Time variable : countperiod Number of groups = 133 Number of instruments = 74 Obs per group: min = 1 F(10, 132) = 3.23 avg = 5.02 Prob > F = 0.001 max = 8 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0360692 .0655964 -0.55 0.583 -.1658254 .0936871 liy | .5816951 .1861907 3.12 0.002 .2133915 .9499988 lenroll_p | .0586023 .055574 1.05 0.294 -.0513285 .1685331 period2 | -.3062505 .1291366 -2.37 0.019 -.5616956 -.0508055 period3 | -.3548114 .1352504 -2.62 0.010 -.6223501 -.0872728 period4 | -.3283025 .1348867 -2.43 0.016 -.5951217 -.0614834 period5 | -.3492812 .1394825 -2.50 0.013 -.6251913 -.0733711 period6 | -.3880723 .1398813 -2.77 0.006 -.6647713 -.1113733 period7 | -.3651863 .1418457 -2.57 0.011 -.6457712 -.0846015 period8 | -.3776852 .1444522 -2.61 0.010 -.663426 -.0919444 _cons | 1.529622 .2202344 6.95 0.000 1.093976 1.965267 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.98 Pr > z = 0.048 Arellano-Bond test for AR(2) in first differences: z = -0.93 Pr > z = 0.353 Hansen test of overid. restrictions: chi2(63) = 64.70 Prob > chi2 = 0.417 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(52) = 51.95 Prob > chi2 = 0.476 Difference (null H = exogenous): chi2(11) = 12.75 Prob > chi2 = 0.310 ivstyle(period*) Hansen test excluding group: chi2(56) = 62.85 Prob > chi2 = 0.247 Difference (null H = exogenous): chi2(7) = 1.85 Prob > chi2 = 0.968 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BP3 . reg lindex3 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robust Linear regression Number of obs = 657 F( 14, 642) = 7.28 Prob > F = 0.0000 R-squared = 0.2163 Root MSE = .19792 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .1742685 .146405 1.19 0.234 -.113222 .461759 liy | .0111825 .1337991 0.08 0.933 -.2515542 .2739192 lenroll_p | -.109266 .0326446 -3.35 0.001 -.1733691 -.0451629 lgdppc | .0660655 .0141547 4.67 0.000 .0382704 .0938606 llifeexp | -.1939324 .0816942 -2.37 0.018 -.3543524 -.0335123 ldepratio1 | -.1110154 .0563388 -1.97 0.049 -.221646 -.0003848 lgovt_Y3 | -.4880694 .1550813 -3.15 0.002 -.7925973 -.1835416 period2 | -.1229989 .0851678 -1.44 0.149 -.29024 .0442422 period3 | -.1633273 .083139 -1.96 0.050 -.3265846 -.0000701 period4 | -.1450255 .0851428 -1.70 0.089 -.3122174 .0221664 period5 | -.164732 .0830455 -1.98 0.048 -.3278056 -.0016585 period6 | -.2072113 .0835426 -2.48 0.013 -.3712611 -.0431615 period7 | -.1862928 .0835319 -2.23 0.026 -.3503215 -.0222641 period8 | -.2065583 .0830903 -2.49 0.013 -.36972 -.0433966 _cons | 3.028289 .4443248 6.82 0.000 2.155784 3.900795 ------------------------------------------------------------------------------ . estimates store BP4 . xtreg lindex3 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re ro > bust cluster(code2) Random-effects GLS regression Number of obs = 657 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.1434 Obs per group: min = 1 between = 0.1829 avg = 5.1 overall = 0.1947 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 48.76 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .0054889 .0407136 0.13 0.893 -.0743082 .085286 liy | .2025658 .1054152 1.92 0.055 -.0040443 .4091758 lenroll_p | -.0647479 .0417712 -1.55 0.121 -.146618 .0171221 lgdppc | .0565117 .0185939 3.04 0.002 .0200684 .092955 llifeexp | -.2048833 .1094575 -1.87 0.061 -.4194162 .0096495 ldepratio1 | -.1271047 .0536832 -2.37 0.018 -.2323217 -.0218876 lgovt_Y3 | -.6323454 .1981768 -3.19 0.001 -1.020765 -.2439259 period2 | -.1762518 .0634638 -2.78 0.005 -.3006386 -.051865 period3 | -.2095684 .0665894 -3.15 0.002 -.3400813 -.0790555 period4 | -.1867906 .0660119 -2.83 0.005 -.3161716 -.0574096 period5 | -.211756 .0669742 -3.16 0.002 -.343023 -.080489 period6 | -.2485593 .068833 -3.61 0.000 -.3834694 -.1136492 period7 | -.2241954 .0700258 -3.20 0.001 -.3614434 -.0869473 period8 | -.2474402 .0727041 -3.40 0.001 -.3899376 -.1049429 _cons | 3.052415 .489705 6.23 0.000 2.092611 4.012219 -------------+---------------------------------------------------------------- sigma_u | .11956045 sigma_e | .11338641 rho | .52648547 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store BP5 . xtabond2 lindex3 ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, gm > m(l(1/4).lindex3 ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(perio > d*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 657 Time variable : countperiod Number of groups = 129 Number of instruments = 74 Obs per group: min = 1 F(14, 128) = 2.87 avg = 5.09 Prob > F = 0.001 max = 8 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .0049562 .0480495 0.10 0.918 -.0901181 .1000304 liy | .3407144 .1848511 1.84 0.068 -.025045 .7064738 lenroll_p | -.1795805 .0997855 -1.80 0.074 -.3770231 .0178621 lgdppc | .0798951 .0362295 2.21 0.029 .0082087 .1515814 llifeexp | -.2840496 .3346046 -0.85 0.398 -.9461219 .3780227 ldepratio1 | -.3768803 .1484844 -2.54 0.012 -.6706821 -.0830786 lgovt_Y3 | -1.363804 .5769926 -2.36 0.020 -2.505483 -.2221259 period2 | -.2089406 .0854383 -2.45 0.016 -.3779949 -.0398863 period3 | -.2497531 .0864454 -2.89 0.005 -.4208001 -.0787061 period4 | -.2320712 .0850049 -2.73 0.007 -.4002679 -.0638746 period5 | -.2679686 .0894021 -3.00 0.003 -.444866 -.0910712 period6 | -.3234213 .0884881 -3.65 0.000 -.4985101 -.1483324 period7 | -.317526 .0925503 -3.43 0.001 -.5006525 -.1343994 period8 | -.3616904 .099875 -3.62 0.000 -.5593102 -.1640707 _cons | 4.925026 1.641737 3.00 0.003 1.676569 8.173484 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.78 Pr > z = 0.076 Arellano-Bond test for AR(2) in first differences: z = -0.81 Pr > z = 0.420 Hansen test of overid. restrictions: chi2(59) = 43.46 Prob > chi2 = 0.935 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 30.79 Prob > chi2 = 0.975 Difference (null H = exogenous): chi2(11) = 12.67 Prob > chi2 = 0.316 ivstyle(period*) Hansen test excluding group: chi2(52) = 41.93 Prob > chi2 = 0.840 Difference (null H = exogenous): chi2(7) = 1.54 Prob > chi2 = 0.981 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BP6 . . estout * using "C:\Llama\Output\benchmark1b.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p sargan sarganp ar2 > ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Hansen J" "J p-value" "AR(2)" "AR(2) > p" "Instruments" "N") fmt(%8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starle > vels(* 0.10 ** 0.05 *** 0.01) & BP1 & BP2 & BP3 & BP4 & BP5 > & BP6 \\ & b/se & b/se & b/se & b/se & b/se > & b/se \\ ln & 0.174 & -0.001 & -0.036 & 0.174 & 0.005 > & 0.005 \\ & (0.16) & (0.04) & (0.07) & (0.15) & (0.04) > & (0.05) \\ liy & 0.159 & 0.253 & 0.582 & 0.011 & 0.203 > & 0.341 \\ & (0.12) & (0.10)** & (0.19)***& (0.13) & (0.11)* > & (0.18)* \\ lenroll_p & 0.002 & -0.041 & 0.059 & -0.109 & -0.065 > & -0.180 \\ & (0.02) & (0.04) & (0.06) & (0.03)***& (0.04) > & (0.10)* \\ period2 & -0.176 & -0.179 & -0.306 & -0.123 & -0.176 > & -0.209 \\ & (0.10)* & (0.07)***& (0.13)** & (0.09) & (0.06)** > *& (0.09)** \\ period3 & -0.223 & -0.217 & -0.355 & -0.163 & -0.210 > & -0.250 \\ & (0.09)** & (0.07)***& (0.14)***& (0.08)** & (0.07)** > *& (0.09)***\\ period4 & -0.203 & -0.194 & -0.328 & -0.145 & -0.187 > & -0.232 \\ & (0.10)** & (0.07)***& (0.13)** & (0.09)* & (0.07)** > *& (0.09)***\\ period5 & -0.220 & -0.210 & -0.349 & -0.165 & -0.212 > & -0.268 \\ & (0.09)** & (0.07)***& (0.14)** & (0.08)** & (0.07)** > *& (0.09)***\\ period6 & -0.262 & -0.241 & -0.388 & -0.207 & -0.249 > & -0.323 \\ & (0.10)***& (0.07)***& (0.14)***& (0.08)** & (0.07)** > *& (0.09)***\\ period7 & -0.235 & -0.207 & -0.365 & -0.186 & -0.224 > & -0.318 \\ & (0.09)** & (0.07)***& (0.14)** & (0.08)** & (0.07)** > *& (0.09)***\\ period8 & -0.240 & -0.215 & -0.378 & -0.207 & -0.247 > & -0.362 \\ & (0.09)** & (0.07)***& (0.14)***& (0.08)** & (0.07)** > *& (0.10)***\\ lgdppc & & & & 0.066 & 0.057 > & 0.080 \\ & & & & (0.01)***& (0.02)** > *& (0.04)** \\ llifeexp & & & & -0.194 & -0.205 > & -0.284 \\ & & & & (0.08)** & (0.11)* > & (0.33) \\ ldepratio1 & & & & -0.111 & -0.127 > & -0.377 \\ & & & & (0.06)** & (0.05)** > & (0.15)** \\ lgovt_Y3 & & & & -0.488 & -0.632 > & -1.364 \\ & & & & (0.16)***& (0.20)** > *& (0.58)** \\ Constant & 1.718 & 1.884 & 1.530 & 3.028 & 3.052 > & 4.925 \\ & (0.15)***& (0.18)***& (0.22)***& (0.44)***& (0.49)** > *& (1.64)***\\ Adjusted R^2& 0.061 & & & 0.199 & > & \\ F & 2.130 & & 3.226 & 7.280 & > & 2.871 \\ F p-value & & & 0.001 & & > & 0.001 \\ Hansen J & & & 64.698 & & > & 43.462 \\ J p-value & & & 0.417 & & > & 0.935 \\ AR(2) & & & -0.929 & & > & -0.806 \\ AR(2) p & & & 0.353 & & > & 0.420 \\ Instruments & & & 74 & & > & 74 \\ N & 667 & 667 & 667 & 657 & 657 > & 657 \\ . estimates clear . . reg bindex ln liy lenroll_p period2-period8, robust Linear regression Number of obs = 676 F( 10, 665) = 2.14 Prob > F = 0.0199 R-squared = 0.0429 Root MSE = .29079 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .1226388 .1403473 0.87 0.383 -.1529383 .398216 liy | .3088053 .2053419 1.50 0.133 -.0943912 .7120018 lenroll_p | .0426721 .0260874 1.64 0.102 -.0085516 .0938958 period2 | -.3011105 .1464642 -2.06 0.040 -.5886984 -.0135225 period3 | -.2814539 .1467827 -1.92 0.056 -.5696672 .0067595 period4 | -.2997479 .1457945 -2.06 0.040 -.5860209 -.0134749 period5 | -.231835 .1478411 -1.57 0.117 -.5221267 .0584566 period6 | -.26808 .1463881 -1.83 0.068 -.5555185 .0193584 period7 | -.2613057 .1463145 -1.79 0.075 -.5485998 .0259884 period8 | -.2386069 .146879 -1.62 0.105 -.5270095 .0497956 _cons | .0861411 .2069721 0.42 0.677 -.3202564 .4925386 ------------------------------------------------------------------------------ . estimates store CP1 . xtreg bindex ln liy lenroll_p period2-period8, re robust cluster(code2) Random-effects GLS regression Number of obs = 676 Group variable (i): code2 Number of groups = 133 R-sq: within = 0.0839 Obs per group: min = 1 between = 0.0000 avg = 5.1 overall = 0.0229 max = 8 Random effects u_i ~ Gaussian Wald chi2(10) = 16.51 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0859 (Std. Err. adjusted for 133 clusters in code2) ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0262532 .0503843 -0.52 0.602 -.1250046 .0724981 liy | .4846708 .217162 2.23 0.026 .0590411 .9103004 lenroll_p | -.0745135 .0505933 -1.47 0.141 -.1736746 .0246475 period2 | -.27304 .102205 -2.67 0.008 -.4733582 -.0727218 period3 | -.2393722 .1018234 -2.35 0.019 -.4389423 -.0398021 period4 | -.2482603 .1077453 -2.30 0.021 -.4594371 -.0370835 period5 | -.1769811 .1023455 -1.73 0.084 -.3775746 .0236123 period6 | -.1987592 .1022845 -1.94 0.052 -.3992332 .0017147 period7 | -.1882653 .100812 -1.87 0.062 -.3858532 .0093225 period8 | -.1683538 .1019789 -1.65 0.099 -.3682287 .0315212 _cons | .5289094 .2905117 1.82 0.069 -.0404831 1.098302 -------------+---------------------------------------------------------------- sigma_u | .20682001 sigma_e | .16967731 rho | .59770241 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store CP2 . xtabond2 bindex ln liy lenroll_p period2-period8, gmm(l(1/3).lindex3 ln liy lgdppc lenrol > l_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 676 Time variable : countperiod Number of groups = 133 Number of instruments = 73 Obs per group: min = 1 F(10, 132) = 2.00 avg = 5.08 Prob > F = 0.038 max = 8 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0710123 .0799056 -0.89 0.376 -.2290734 .0870489 liy | .4673649 .264847 1.76 0.080 -.0565285 .9912584 lenroll_p | -.0128735 .0679072 -0.19 0.850 -.1472007 .1214537 period2 | -.4282313 .1692261 -2.53 0.013 -.7629773 -.0934854 period3 | -.406003 .1644536 -2.47 0.015 -.7313084 -.0806976 period4 | -.4201881 .1714303 -2.45 0.016 -.759294 -.0810821 period5 | -.3554603 .1697541 -2.09 0.038 -.6912508 -.0196699 period6 | -.3855728 .1699694 -2.27 0.025 -.721789 -.0493566 period7 | -.3794138 .1712437 -2.22 0.028 -.7181508 -.0406769 period8 | -.3568612 .1721098 -2.07 0.040 -.6973114 -.016411 _cons | .4509758 .3224973 1.40 0.164 -.1869558 1.088907 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.27 Pr > z = 0.023 Arellano-Bond test for AR(2) in first differences: z = 1.11 Pr > z = 0.266 Hansen test of overid. restrictions: chi2(62) = 37.39 Prob > chi2 = 0.994 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(52) = 23.15 Prob > chi2 = 1.000 Difference (null H = exogenous): chi2(10) = 14.24 Prob > chi2 = 0.162 ivstyle(period*) Hansen test excluding group: chi2(55) = 35.25 Prob > chi2 = 0.982 Difference (null H = exogenous): chi2(7) = 2.14 Prob > chi2 = 0.952 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CP3 . reg bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robust Linear regression Number of obs = 665 F( 14, 650) = 7.02 Prob > F = 0.0000 R-squared = 0.2136 Root MSE = .26637 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .12008 .1260393 0.95 0.341 -.1274134 .3675733 liy | .1559606 .2316207 0.67 0.501 -.2988545 .6107757 lenroll_p | -.106323 .0303282 -3.51 0.000 -.1658761 -.0467698 lgdppc | .1033743 .0143888 7.18 0.000 .0751202 .1316284 llifeexp | -.3820514 .0938112 -4.07 0.000 -.566261 -.1978419 ldepratio1 | -.1379645 .0572487 -2.41 0.016 -.2503791 -.0255499 lgovt_Y3 | -.3501186 .1720296 -2.04 0.042 -.6879193 -.0123179 period2 | -.2249184 .1267807 -1.77 0.077 -.4738676 .0240308 period3 | -.1972298 .1272419 -1.55 0.122 -.4470845 .052625 period4 | -.2121018 .1261882 -1.68 0.093 -.4598876 .0356839 period5 | -.1446162 .128121 -1.13 0.259 -.3961972 .1069648 period6 | -.1767391 .1270164 -1.39 0.165 -.4261512 .072673 period7 | -.1787841 .1274356 -1.40 0.161 -.4290193 .0714511 period8 | -.1742924 .1284719 -1.36 0.175 -.4265625 .0779776 _cons | 2.127862 .5256361 4.05 0.000 1.095713 3.160012 ------------------------------------------------------------------------------ . estimates store CP4 . xtreg bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re rob > ust cluster(code2) Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.1369 Obs per group: min = 1 between = 0.2193 avg = 5.2 overall = 0.2039 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 29.32 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0095 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0272597 .0485821 -0.56 0.575 -.1224789 .0679595 liy | .447642 .223202 2.01 0.045 .0101741 .88511 lenroll_p | -.0979844 .0459096 -2.13 0.033 -.1879656 -.0080031 lgdppc | .1215018 .0293316 4.14 0.000 .064013 .1789906 llifeexp | -.4875183 .1988528 -2.45 0.014 -.8772626 -.0977739 ldepratio1 | -.0745087 .0800915 -0.93 0.352 -.2314851 .0824676 lgovt_Y3 | -.4261514 .2704077 -1.58 0.115 -.9561408 .103838 period2 | -.2813717 .0947899 -2.97 0.003 -.4671566 -.0955869 period3 | -.2362027 .0885017 -2.67 0.008 -.4096628 -.0627426 period4 | -.2373275 .0933115 -2.54 0.011 -.4202147 -.0544402 period5 | -.1675654 .086025 -1.95 0.051 -.3361713 .0010405 period6 | -.1916337 .0872354 -2.20 0.028 -.3626119 -.0206554 period7 | -.1911023 .0874276 -2.19 0.029 -.3624573 -.0197473 period8 | -.1878556 .0896609 -2.10 0.036 -.3635877 -.0121235 _cons | 2.116235 .936592 2.26 0.024 .2805488 3.951922 -------------+---------------------------------------------------------------- sigma_u | .18788779 sigma_e | .16557028 rho | .56288976 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store CP5 . xtabond2 bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, gmm > (l(1/4).lindex3 ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3, collapse) iv(period > *) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 665 Time variable : countperiod Number of groups = 129 Number of instruments = 74 Obs per group: min = 1 F(14, 128) = 2.02 avg = 5.16 Prob > F = 0.021 max = 8 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0275907 .0567722 -0.49 0.628 -.1399241 .0847428 liy | .5747534 .2996878 1.92 0.057 -.0182301 1.167737 lenroll_p | -.2723646 .1113194 -2.45 0.016 -.492629 -.0521001 lgdppc | .1326943 .0477443 2.78 0.006 .038224 .2271646 llifeexp | -.4029189 .382793 -1.05 0.295 -1.16034 .3545025 ldepratio1 | -.185204 .1783156 -1.04 0.301 -.5380318 .1676238 lgovt_Y3 | -.4800127 .5411138 -0.89 0.377 -1.550699 .5906734 period2 | -.3040803 .1137781 -2.67 0.009 -.5292097 -.0789508 period3 | -.2713647 .1066163 -2.55 0.012 -.4823233 -.0604061 period4 | -.2740338 .1138463 -2.41 0.018 -.4992981 -.0487695 period5 | -.2124929 .1111919 -1.91 0.058 -.4325052 .0075193 period6 | -.2399758 .1111994 -2.16 0.033 -.4600028 -.0199487 period7 | -.2440631 .1137933 -2.14 0.034 -.4692227 -.0189036 period8 | -.2500566 .1171234 -2.13 0.035 -.4818052 -.0183079 _cons | 2.960215 1.843277 1.61 0.111 -.6870244 6.607454 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.28 Pr > z = 0.023 Arellano-Bond test for AR(2) in first differences: z = 1.17 Pr > z = 0.244 Hansen test of overid. restrictions: chi2(59) = 38.85 Prob > chi2 = 0.980 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 19.23 Prob > chi2 = 1.000 Difference (null H = exogenous): chi2(11) = 19.62 Prob > chi2 = 0.051 ivstyle(period*) Hansen test excluding group: chi2(52) = 35.14 Prob > chi2 = 0.965 Difference (null H = exogenous): chi2(7) = 3.71 Prob > chi2 = 0.812 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CP6 . . estout * using "C:\Llama\Output\benchmark1c.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p sargan sarganp ar2 > ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Hansen J" "J p-value" "AR(2)" "AR(2) > p" "Instruments" "N") fmt(%8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starle > vels(* 0.10 ** 0.05 *** 0.01) & CP1 & CP2 & CP3 & CP4 & CP5 > & CP6 \\ & b/se & b/se & b/se & b/se & b/se > & b/se \\ ln & 0.123 & -0.026 & -0.071 & 0.120 & -0.027 > & -0.028 \\ & (0.14) & (0.05) & (0.08) & (0.13) & (0.05) > & (0.06) \\ liy & 0.309 & 0.485 & 0.467 & 0.156 & 0.448 > & 0.575 \\ & (0.21) & (0.22)** & (0.26)* & (0.23) & (0.22)** > & (0.30)* \\ lenroll_p & 0.043 & -0.075 & -0.013 & -0.106 & -0.098 > & -0.272 \\ & (0.03) & (0.05) & (0.07) & (0.03)***& (0.05)** > & (0.11)** \\ period2 & -0.301 & -0.273 & -0.428 & -0.225 & -0.281 > & -0.304 \\ & (0.15)** & (0.10)***& (0.17)** & (0.13)* & (0.09)** > *& (0.11)***\\ period3 & -0.281 & -0.239 & -0.406 & -0.197 & -0.236 > & -0.271 \\ & (0.15)* & (0.10)** & (0.16)** & (0.13) & (0.09)** > *& (0.11)** \\ period4 & -0.300 & -0.248 & -0.420 & -0.212 & -0.237 > & -0.274 \\ & (0.15)** & (0.11)** & (0.17)** & (0.13)* & (0.09)** > & (0.11)** \\ period5 & -0.232 & -0.177 & -0.355 & -0.145 & -0.168 > & -0.212 \\ & (0.15) & (0.10)* & (0.17)** & (0.13) & (0.09)* > & (0.11)* \\ period6 & -0.268 & -0.199 & -0.386 & -0.177 & -0.192 > & -0.240 \\ & (0.15)* & (0.10)* & (0.17)** & (0.13) & (0.09)** > & (0.11)** \\ period7 & -0.261 & -0.188 & -0.379 & -0.179 & -0.191 > & -0.244 \\ & (0.15)* & (0.10)* & (0.17)** & (0.13) & (0.09)** > & (0.11)** \\ period8 & -0.239 & -0.168 & -0.357 & -0.174 & -0.188 > & -0.250 \\ & (0.15) & (0.10)* & (0.17)** & (0.13) & (0.09)** > & (0.12)** \\ lgdppc & & & & 0.103 & 0.122 > & 0.133 \\ & & & & (0.01)***& (0.03)** > *& (0.05)***\\ llifeexp & & & & -0.382 & -0.488 > & -0.403 \\ & & & & (0.09)***& (0.20)** > & (0.38) \\ ldepratio1 & & & & -0.138 & -0.075 > & -0.185 \\ & & & & (0.06)** & (0.08) > & (0.18) \\ lgovt_Y3 & & & & -0.350 & -0.426 > & -0.480 \\ & & & & (0.17)** & (0.27) > & (0.54) \\ Constant & 0.086 & 0.529 & 0.451 & 2.128 & 2.116 > & 2.960 \\ & (0.21) & (0.29)* & (0.32) & (0.53)***& (0.94)** > & (1.84) \\ Adjusted R^2& 0.029 & & & 0.197 & > & \\ F & 2.138 & & 2.004 & 7.018 & > & 2.020 \\ F p-value & & & 0.038 & & > & 0.021 \\ Hansen J & & & 37.391 & & > & 38.853 \\ J p-value & & & 0.994 & & > & 0.980 \\ AR(2) & & & 1.113 & & > & 1.165 \\ AR(2) p & & & 0.266 & & > & 0.244 \\ Instruments & & & 73 & & > & 74 \\ N & 676 & 676 & 676 & 665 & 665 > & 665 \\ . estimates clear . . /* Robustness proximates */ . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lcell period2-period8, > fe robust cluster(code2) Fixed-effects (within) regression Number of obs = 658 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2571 Obs per group: min = 1 between = 0.2188 avg = 5.1 overall = 0.2592 max = 8 F(15,128) = 4.19 corr(u_i, Xb) = -0.0038 Prob > F = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0589242 .0614633 -0.96 0.340 -.1805398 .0626915 liy | .6268371 .1579465 3.97 0.000 .314313 .9393612 lenroll_p | .027696 .0456337 0.61 0.545 -.0625981 .1179901 lgdppc | .0621278 .0421213 1.47 0.143 -.0212163 .1454719 llifeexp | -.0554416 .1645186 -0.34 0.737 -.3809697 .2700865 ldepratio1 | -.3872734 .1434971 -2.70 0.008 -.6712069 -.1033398 lgovt_Y3 | -.6412825 .2464199 -2.60 0.010 -1.128866 -.1536986 lcell | .0076251 .0132954 0.57 0.567 -.0186821 .0339322 period2 | -.4226127 .1334563 -3.17 0.002 -.6866788 -.1585466 period3 | -.4848849 .140075 -3.46 0.001 -.7620473 -.2077226 period4 | -.4526969 .1376462 -3.29 0.001 -.7250535 -.1803404 period5 | -.5030098 .1414924 -3.56 0.001 -.7829768 -.2230429 period6 | -.5319645 .1424416 -3.73 0.000 -.8138095 -.2501194 period7 | -.5763234 .1463095 -3.94 0.000 -.8658217 -.2868251 period8 | -.6336394 .155716 -4.07 0.000 -.9417501 -.3255287 _cons | 2.170675 .9995297 2.17 0.032 .1929345 4.148415 -------------+---------------------------------------------------------------- sigma_u | .26824922 sigma_e | .12475489 rho | .82217175 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R1 . xtabond2 lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lcell period2-period > 8, gmm(l(1/4).lindex ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3 lcell, collapse > ) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 658 Time variable : countperiod Number of groups = 129 Number of instruments = 80 Obs per group: min = 1 F(15, 128) = 4.03 avg = 5.10 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0510187 .0922769 -0.55 0.581 -.2336044 .131567 liy | .4304281 .2828056 1.52 0.130 -.1291511 .9900073 lenroll_p | -.224453 .1225132 -1.83 0.069 -.4668663 .0179604 lgdppc | .1951542 .0676766 2.88 0.005 .0612445 .329064 llifeexp | -.7556796 .5249805 -1.44 0.152 -1.794443 .283084 ldepratio1 | -.457004 .2139624 -2.14 0.035 -.8803651 -.0336429 lgovt_Y3 | -1.719799 .9064762 -1.90 0.060 -3.513418 .0738186 lcell | -.0585322 .0251562 -2.33 0.022 -.1083079 -.0087564 period2 | -.4826863 .1838127 -2.63 0.010 -.8463912 -.1189814 period3 | -.5347186 .1867655 -2.86 0.005 -.904266 -.1651713 period4 | -.4838571 .1836853 -2.63 0.009 -.8473098 -.1204044 period5 | -.5203438 .1896442 -2.74 0.007 -.8955872 -.1451003 period6 | -.4975748 .1795496 -2.77 0.006 -.8528443 -.1423052 period7 | -.4452783 .178268 -2.50 0.014 -.798012 -.0925447 period8 | -.4319977 .1867725 -2.31 0.022 -.801559 -.0624364 _cons | 5.740179 2.692017 2.13 0.035 .4135629 11.0668 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.29 Pr > z = 0.196 Arellano-Bond test for AR(2) in first differences: z = -2.13 Pr > z = 0.033 Hansen test of overid. restrictions: chi2(64) = 65.78 Prob > chi2 = 0.415 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(52) = 40.22 Prob > chi2 = 0.883 Difference (null H = exogenous): chi2(12) = 25.56 Prob > chi2 = 0.012 ivstyle(period*) Hansen test excluding group: chi2(57) = 62.96 Prob > chi2 = 0.274 Difference (null H = exogenous): chi2(7) = 2.82 Prob > chi2 = 0.901 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R2 . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lroads period2-period8, > re robust cluster(code2) note: period2 dropped due to collinearity note: period3 dropped due to collinearity note: period4 dropped due to collinearity note: period8 dropped due to collinearity Random-effects GLS regression Number of obs = 340 Group variable (i): code2 Number of groups = 121 R-sq: within = 0.1876 Obs per group: min = 1 between = 0.2702 avg = 2.8 overall = 0.2758 max = 4 Random effects u_i ~ Gaussian Wald chi2(11) = 51.83 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 121 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | 1.92431 1.387199 1.39 0.165 -.7945497 4.64317 liy | .5739221 .2356186 2.44 0.015 .1121182 1.035726 lenroll_p | .0360475 .0371067 0.97 0.331 -.0366803 .1087753 lgdppc | .1105075 .0310086 3.56 0.000 .0497318 .1712831 llifeexp | -.4980661 .1804965 -2.76 0.006 -.8518328 -.1442995 ldepratio1 | -.3803343 .1495744 -2.54 0.011 -.6734948 -.0871738 lgovt_Y3 | -1.247314 .3415959 -3.65 0.000 -1.91683 -.5777983 lroads | -.0049753 .0227956 -0.22 0.827 -.0496539 .0397033 period5 | .0976257 .0398412 2.45 0.014 .0195383 .1757131 period6 | .0714133 .024893 2.87 0.004 .0226239 .1202027 period7 | .0446413 .0176096 2.54 0.011 .0101271 .0791555 _cons | 2.877396 1.026054 2.80 0.005 .8663676 4.888424 -------------+---------------------------------------------------------------- sigma_u | .2314463 sigma_e | .10280767 rho | .83520513 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R3 . xtabond2 lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 lroads period2-perio > d8, gmm(l(1/4).lindex ln liy lgdppc lenroll_p llifeexp ldepratio1 lgovt_Y3 lroads, collap > se) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period4 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 340 Time variable : countperiod Number of groups = 121 Number of instruments = 74 Obs per group: min = 1 F(11, 120) = 3.93 avg = 2.81 Prob > F = 0.000 max = 4 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -6.95847 3.303678 -2.11 0.037 -13.49952 -.4174161 liy | 1.098539 .4971988 2.21 0.029 .1141203 2.082958 lenroll_p | -.0949553 .1249827 -0.76 0.449 -.3424124 .1525018 lgdppc | .186943 .0494897 3.78 0.000 .0889569 .284929 llifeexp | -.5238124 .3653528 -1.43 0.154 -1.247185 .1995607 ldepratio1 | -.2971177 .2240186 -1.33 0.187 -.7406588 .1464235 lgovt_Y3 | -1.550999 .9417551 -1.65 0.102 -3.415608 .3136108 lroads | -.1988336 .0979183 -2.03 0.045 -.3927051 -.0049622 period5 | .1123443 .0488215 2.30 0.023 .0156811 .2090074 period6 | .0764301 .0392709 1.95 0.054 -.0013235 .1541838 period7 | .0572387 .0262081 2.18 0.031 .0053486 .1091289 _cons | 4.247657 2.146878 1.98 0.050 -.0030126 8.498327 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = 0.82 Pr > z = 0.409 Arellano-Bond test for AR(2) in first differences: z = -1.69 Pr > z = 0.091 Hansen test of overid. restrictions: chi2(62) = 59.26 Prob > chi2 = 0.575 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(50) = 51.65 Prob > chi2 = 0.409 Difference (null H = exogenous): chi2(12) = 7.61 Prob > chi2 = 0.815 ivstyle(period*) Hansen test excluding group: chi2(59) = 58.05 Prob > chi2 = 0.511 Difference (null H = exogenous): chi2(3) = 1.21 Prob > chi2 = 0.750 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R4 . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 avg5_JID_safrica avg5_J > ID_laam avg5_JID_asiae period2-period8, re robust cluster(code2) Random-effects GLS regression Number of obs = 512 Group variable (i): code2 Number of groups = 84 R-sq: within = 0.3057 Obs per group: min = 2 between = 0.3291 avg = 6.1 overall = 0.3269 max = 8 Random effects u_i ~ Gaussian Wald chi2(17) = 110.57 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 84 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0810035 .0676996 -1.20 0.231 -.2136923 .0516852 liy | .8850254 .1817014 4.87 0.000 .5288971 1.241154 lenroll_p | .0267811 .0594248 0.45 0.652 -.0896893 .1432515 lgdppc | .092147 .0291181 3.16 0.002 .0350767 .1492174 llifeexp | -.191902 .1649276 -1.16 0.245 -.5151542 .1313502 ldepratio1 | -.3989283 .1489981 -2.68 0.007 -.6909592 -.1068974 lgovt_Y3 | -.5612538 .2733259 -2.05 0.040 -1.096963 -.0255449 avg5_JID_s~a | .0599414 .0757637 0.79 0.429 -.0885526 .2084355 avg5_JID_~am | -.1142754 .0519841 -2.20 0.028 -.2161623 -.0123885 avg5_JID_a~e | .0376579 .0606124 0.62 0.534 -.0811403 .1564561 period2 | -.479159 .1479482 -3.24 0.001 -.7691321 -.1891858 period3 | -.5490278 .1550905 -3.54 0.000 -.8529996 -.2450561 period4 | -.5040574 .1525108 -3.31 0.001 -.802973 -.2051417 period5 | -.5472036 .1570625 -3.48 0.000 -.8550405 -.2393667 period6 | -.5688621 .1598785 -3.56 0.000 -.8822181 -.2555061 period7 | -.6120148 .1651505 -3.71 0.000 -.9357039 -.2883257 period8 | -.6700613 .1730698 -3.87 0.000 -1.009272 -.3308508 _cons | 2.585263 .9446998 2.74 0.006 .733685 4.43684 -------------+---------------------------------------------------------------- sigma_u | .26089107 sigma_e | .13042577 rho | .80004848 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R5 . xtabond2 lindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 avg5_JID_safrica avg > 5_JID_laam avg5_JID_asiae period2-period8, gmm(l(1/4).lindex ln liy lgdppc lenroll_p llif > eexp ldepratio1 lgovt_Y3, collapse) iv(period* avg5_JID_safrica avg5_JID_laam avg5_JID_as > iae) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 512 Time variable : countperiod Number of groups = 84 Number of instruments = 77 Obs per group: min = 2 F(17, 83) = 5.01 avg = 6.10 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0680228 .1022052 -0.67 0.508 -.2713048 .1352593 liy | .9806722 .4109994 2.39 0.019 .163211 1.798133 lenroll_p | -.2536856 .225426 -1.13 0.264 -.7020489 .1946776 lgdppc | .1917489 .085637 2.24 0.028 .0214203 .3620775 llifeexp | -1.538543 1.065194 -1.44 0.152 -3.657172 .580085 ldepratio1 | -.7443474 .2734927 -2.72 0.008 -1.288313 -.2003815 lgovt_Y3 | -1.42234 .8686632 -1.64 0.105 -3.150076 .3053959 avg5_JID_s~a | -.0760434 .1902041 -0.40 0.690 -.4543517 .3022649 avg5_JID_~am | -.039854 .0787503 -0.51 0.614 -.1964853 .1167772 avg5_JID_a~e | .0421873 .0847682 0.50 0.620 -.1264133 .210788 period2 | -.6462225 .2067918 -3.12 0.002 -1.057523 -.234922 period3 | -.6987423 .2090855 -3.34 0.001 -1.114605 -.2828797 period4 | -.633072 .2103036 -3.01 0.003 -1.051357 -.2147866 period5 | -.6706724 .2166473 -3.10 0.003 -1.101575 -.2397697 period6 | -.7040025 .2170864 -3.24 0.002 -1.135779 -.2722264 period7 | -.7467799 .2213357 -3.37 0.001 -1.187008 -.3065521 period8 | -.8043164 .2278478 -3.53 0.001 -1.257496 -.3511364 _cons | 10.42702 4.987607 2.09 0.040 .5068667 20.34717 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.61 Pr > z = 0.107 Arellano-Bond test for AR(2) in first differences: z = -1.44 Pr > z = 0.149 Hansen test of overid. restrictions: chi2(59) = 66.75 Prob > chi2 = 0.228 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 51.54 Prob > chi2 = 0.337 Difference (null H = exogenous): chi2(11) = 15.21 Prob > chi2 = 0.173 ivstyle(period* avg5_JID_safrica avg5_JID_laam avg5_JID_asiae) Hansen test excluding group: chi2(49) = 52.61 Prob > chi2 = 0.336 Difference (null H = exogenous): chi2(10) = 14.13 Prob > chi2 = 0.167 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R6 . . xtreg lindex ln liy lenroll_s lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re rob > ust cluster(code2) Random-effects GLS regression Number of obs = 635 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2780 Obs per group: min = 1 between = 0.2614 avg = 4.9 overall = 0.3008 max = 8 Random effects u_i ~ Gaussian Wald chi2(14) = 66.62 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0492586 .0655777 -0.75 0.453 -.1777885 .0792713 liy | .4977529 .1932433 2.58 0.010 .1190029 .8765029 lenroll_s | .0120417 .0229305 0.53 0.599 -.0329012 .0569846 lgdppc | .1236137 .0357752 3.46 0.001 .0534956 .1937319 llifeexp | -.5402326 .2409954 -2.24 0.025 -1.012575 -.0678903 ldepratio1 | -.2879212 .1019663 -2.82 0.005 -.4877716 -.0880709 lgovt_Y3 | -1.068985 .4274506 -2.50 0.012 -1.906772 -.231197 period2 | -.4132855 .1189872 -3.47 0.001 -.6464962 -.1800748 period3 | -.4763765 .1279841 -3.72 0.000 -.7272208 -.2255322 period4 | -.4395525 .1248146 -3.52 0.000 -.6841845 -.1949204 period5 | -.4771609 .1286296 -3.71 0.000 -.7292703 -.2250516 period6 | -.496141 .1308096 -3.79 0.000 -.7525231 -.2397588 period7 | -.5210608 .1345437 -3.87 0.000 -.7847616 -.2573601 period8 | -.570743 .1409475 -4.05 0.000 -.8469949 -.294491 _cons | 3.435372 1.049862 3.27 0.001 1.37768 5.493063 -------------+---------------------------------------------------------------- sigma_u | .22187156 sigma_e | .14315734 rho | .70605669 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R7 . xtabond2 lindex ln liy lenroll_s lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, gmm > (l(1/4).lindex ln liy lgdppc lenroll_s llifeexp ldepratio1 lgovt_Y3, collapse) iv(period* > ) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 635 Time variable : countperiod Number of groups = 129 Number of instruments = 74 Obs per group: min = 1 F(14, 128) = 3.63 avg = 4.92 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0413831 .0910996 -0.45 0.650 -.2216391 .138873 liy | .8659705 .3444653 2.51 0.013 .1843871 1.547554 lenroll_s | -.1056306 .0795099 -1.33 0.186 -.2629546 .0516934 lgdppc | .213092 .0643715 3.31 0.001 .0857219 .340462 llifeexp | -1.14792 .6455013 -1.78 0.078 -2.425154 .1293148 ldepratio1 | -.7017975 .2413163 -2.91 0.004 -1.179283 -.2243119 lgovt_Y3 | -1.421127 .8384368 -1.69 0.093 -3.080118 .2378632 period2 | -.4777816 .1695418 -2.82 0.006 -.813249 -.1423142 period3 | -.5233013 .1717612 -3.05 0.003 -.8631602 -.1834424 period4 | -.4816752 .1707057 -2.82 0.006 -.8194457 -.1439048 period5 | -.5241859 .1761666 -2.98 0.003 -.8727617 -.1756102 period6 | -.5456882 .1736927 -3.14 0.002 -.8893688 -.2020076 period7 | -.5954293 .1821311 -3.27 0.001 -.9558068 -.2350517 period8 | -.6753829 .1919578 -3.52 0.001 -1.055204 -.2955615 _cons | 7.538285 3.014842 2.50 0.014 1.572905 13.50367 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.33 Pr > z = 0.183 Arellano-Bond test for AR(2) in first differences: z = -0.20 Pr > z = 0.840 Hansen test of overid. restrictions: chi2(59) = 56.87 Prob > chi2 = 0.555 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 49.76 Prob > chi2 = 0.403 Difference (null H = exogenous): chi2(11) = 7.11 Prob > chi2 = 0.790 ivstyle(period*) Hansen test excluding group: chi2(52) = 55.02 Prob > chi2 = 0.361 Difference (null H = exogenous): chi2(7) = 1.85 Prob > chi2 = 0.968 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R8 . xtreg lindex ln liy lenroll_p lgdppc lmort_ufive ldepratio1 lgovt_Y3 period2-period8, fe > robust cluster(code2) Fixed-effects (within) regression Number of obs = 662 Group variable (i): code2 Number of groups = 128 R-sq: within = 0.2550 Obs per group: min = 1 between = 0.2727 avg = 5.2 overall = 0.2818 max = 8 F(14,127) = 4.43 corr(u_i, Xb) = -0.1813 Prob > F = 0.0000 (Std. Err. adjusted for 128 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0701765 .0611775 -1.15 0.253 -.1912358 .0508827 liy | .5994152 .1776418 3.37 0.001 .2478941 .9509363 lenroll_p | -.0113366 .0451395 -0.25 0.802 -.1006594 .0779863 lgdppc | .1712677 .0822785 2.08 0.039 .0084534 .3340819 lmort_ufive | .0817992 .0629643 1.30 0.196 -.0427958 .2063941 ldepratio1 | -.3834823 .1513857 -2.53 0.013 -.6830473 -.0839173 lgovt_Y3 | -.9748188 .4401302 -2.21 0.029 -1.845757 -.1038806 period2 | -.4106813 .1308694 -3.14 0.002 -.6696482 -.1517145 period3 | -.485771 .1370271 -3.55 0.001 -.7569229 -.2146191 period4 | -.440109 .1340597 -3.28 0.001 -.705389 -.1748289 period5 | -.4841438 .138326 -3.50 0.001 -.757866 -.2104216 period6 | -.4996491 .1409264 -3.55 0.001 -.778517 -.2207812 period7 | -.5253922 .1460233 -3.60 0.000 -.814346 -.2364384 period8 | -.5737347 .1539242 -3.73 0.000 -.8783228 -.2691465 _cons | .9889616 .9442172 1.05 0.297 -.8794738 2.857397 -------------+---------------------------------------------------------------- sigma_u | .27149905 sigma_e | .13905629 rho | .79218713 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R9 . xtabond2 lindex ln liy lenroll_p lgdppc lmort_ufive ldepratio1 lgovt_Y3 period2-period8, > gmm(l(1/4).lindex ln liy lgdppc lenroll_p lmort_ufive ldepratio1 lgovt_Y3 lmort_ufive, co > llapse) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 662 Time variable : countperiod Number of groups = 128 Number of instruments = 74 Obs per group: min = 1 F(14, 127) = 3.75 avg = 5.17 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0256956 .0891352 -0.29 0.774 -.2020781 .1506869 liy | .2244914 .4810373 0.47 0.642 -.7273945 1.176377 lenroll_p | -.3976512 .1623337 -2.45 0.016 -.7188803 -.0764221 lgdppc | .2558766 .0870244 2.94 0.004 .0836709 .4280823 lmort_ufive | .209607 .124982 1.68 0.096 -.0377098 .4569238 ldepratio1 | -.8466762 .3374185 -2.51 0.013 -1.514367 -.1789858 lgovt_Y3 | -1.565875 .9807016 -1.60 0.113 -3.506506 .3747569 period2 | -.4737237 .1684145 -2.81 0.006 -.8069856 -.1404619 period3 | -.5431459 .1680218 -3.23 0.002 -.8756307 -.2106611 period4 | -.4973339 .1653123 -3.01 0.003 -.8244571 -.1702107 period5 | -.5386523 .1710104 -3.15 0.002 -.877051 -.2002535 period6 | -.5510088 .1706999 -3.23 0.002 -.8887932 -.2132245 period7 | -.5862295 .1782804 -3.29 0.001 -.9390142 -.2334447 period8 | -.6381879 .1875416 -3.40 0.001 -1.009299 -.2670768 _cons | 3.784251 1.622669 2.33 0.021 .5732815 6.99522 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.09 Pr > z = 0.276 Arellano-Bond test for AR(2) in first differences: z = -1.11 Pr > z = 0.268 Hansen test of overid. restrictions: chi2(59) = 53.97 Prob > chi2 = 0.661 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 38.96 Prob > chi2 = 0.821 Difference (null H = exogenous): chi2(11) = 15.01 Prob > chi2 = 0.182 ivstyle(period*) Hansen test excluding group: chi2(52) = 51.69 Prob > chi2 = 0.486 Difference (null H = exogenous): chi2(7) = 2.28 Prob > chi2 = 0.943 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R10 . xtreg lindex ln liy lenroll_p lgdppc llifeexp ldepratioyoung ldepratioold lgovt_Y3 period > 2-period8, re robust cluster(code2) Random-effects GLS regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 R-sq: within = 0.2523 Obs per group: min = 1 between = 0.2690 avg = 5.2 overall = 0.2998 max = 8 Random effects u_i ~ Gaussian Wald chi2(15) = 79.74 corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000 (Std. Err. adjusted for 129 clusters in code2) ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0499118 .067432 -0.74 0.459 -.1820761 .0822525 liy | .5236918 .1815665 2.88 0.004 .1678281 .8795555 lenroll_p | -.050971 .0398251 -1.28 0.201 -.1290267 .0270847 lgdppc | .1257827 .033547 3.75 0.000 .0600318 .1915336 llifeexp | -.3012457 .2555822 -1.18 0.239 -.8021776 .1996863 ldepratioy~g | -.1345435 .0704807 -1.91 0.056 -.2726831 .0035961 ldepratioold | -.0485333 .0811532 -0.60 0.550 -.2075907 .1105241 lgovt_Y3 | -1.046705 .4228175 -2.48 0.013 -1.875412 -.2179985 period2 | -.4164202 .1202042 -3.46 0.001 -.6520162 -.1808243 period3 | -.4832262 .1283124 -3.77 0.000 -.734714 -.2317385 period4 | -.4387967 .1251566 -3.51 0.000 -.6840991 -.1934943 period5 | -.4763409 .1281779 -3.72 0.000 -.7275651 -.2251168 period6 | -.4899053 .1296122 -3.78 0.000 -.7439406 -.23587 period7 | -.5084061 .1323008 -3.84 0.000 -.767711 -.2491013 period8 | -.5536202 .1379129 -4.01 0.000 -.8239244 -.2833159 _cons | 2.119501 .8532737 2.48 0.013 .4471157 3.791887 -------------+---------------------------------------------------------------- sigma_u | .21247172 sigma_e | .14170842 rho | .69212543 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . estimates store R11 . xtabond2 lindex ln liy lenroll_p lgdppc llifeexp ldepratioyoung ldepratioold lgovt_Y3 per > iod2-period8, gmm(l(1/4).lindex ln liy lgdppc lenroll_p llifeexp ldepratioyoung ldepratio > old lgovt_Y3, collapse) iv(period*) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 665 Time variable : countperiod Number of groups = 129 Number of instruments = 82 Obs per group: min = 1 F(15, 128) = 3.94 avg = 5.16 Prob > F = 0.000 max = 8 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | -.0477212 .0865069 -0.55 0.582 -.21889 .1234475 liy | .8090975 .3163205 2.56 0.012 .1832034 1.434992 lenroll_p | -.169238 .1288112 -1.31 0.191 -.4241129 .085637 lgdppc | .2037694 .0707673 2.88 0.005 .0637442 .3437947 llifeexp | -.8372446 .5400308 -1.55 0.124 -1.905788 .2312987 ldepratioy~g | -.2881008 .2078564 -1.39 0.168 -.6993802 .1231786 ldepratioold | -.1076047 .1496155 -0.72 0.473 -.4036445 .1884351 lgovt_Y3 | -1.563243 .8684091 -1.80 0.074 -3.281538 .155053 period2 | -.495681 .1625869 -3.05 0.003 -.817387 -.173975 period3 | -.5641487 .164309 -3.43 0.001 -.8892621 -.2390353 period4 | -.5024304 .1619697 -3.10 0.002 -.8229151 -.1819457 period5 | -.5433429 .1665417 -3.26 0.001 -.8728742 -.2138117 period6 | -.555166 .1608582 -3.45 0.001 -.8734515 -.2368806 period7 | -.591205 .1664003 -3.55 0.001 -.9204565 -.2619535 period8 | -.6594114 .1762791 -3.74 0.000 -1.00821 -.3106131 _cons | 5.139344 2.723322 1.89 0.061 -.2492136 10.5279 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.31 Pr > z = 0.189 Arellano-Bond test for AR(2) in first differences: z = -0.50 Pr > z = 0.618 Hansen test of overid. restrictions: chi2(66) = 72.50 Prob > chi2 = 0.272 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(54) = 54.88 Prob > chi2 = 0.441 Difference (null H = exogenous): chi2(12) = 17.62 Prob > chi2 = 0.128 ivstyle(period*) Hansen test excluding group: chi2(59) = 67.74 Prob > chi2 = 0.204 Difference (null H = exogenous): chi2(7) = 4.77 Prob > chi2 = 0.689 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store R12 . . estout * using "C:\Llama\Output\robust1.tex", replace style(tex) varlabels(_cons Constant > ) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p sargan sarganp ar2 ar2p > zcols N, labels("Adjusted R^2" "F" "F p-value" "Hansen J" "J p-value" "AR(2)" "AR(2) p" > "Instruments" "N") fmt(%8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels > (* 0.10 ** 0.05 *** 0.01) & R1 & R2 & R3 & R4 & R5 > & R6 & R7 & R8 & R9 & R10 & > R11 & R12 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se & > b/se & b/se \\ ln & -0.059 & -0.051 & 1.924 & -6.958 & -0.081 > & -0.068 & -0.049 & -0.041 & -0.070 & -0.026 & - > 0.050 & -0.048 \\ & (0.06) & (0.09) & (1.39) & (3.30)** & (0.07) > & (0.10) & (0.07) & (0.09) & (0.06) & (0.09) & ( > 0.07) & (0.09) \\ liy & 0.627 & 0.430 & 0.574 & 1.099 & 0.885 > & 0.981 & 0.498 & 0.866 & 0.599 & 0.224 & > 0.524 & 0.809 \\ & (0.16)***& (0.28) & (0.24)** & (0.50)** & (0.18)** > *& (0.41)** & (0.19)** & (0.34)** & (0.18)***& (0.48) & ( > 0.18)***& (0.32)** \\ lenroll_p & 0.028 & -0.224 & 0.036 & -0.095 & 0.027 > & -0.254 & & & -0.011 & -0.398 & - > 0.051 & -0.169 \\ & (0.05) & (0.12)* & (0.04) & (0.12) & (0.06) > & (0.23) & & & (0.05) & (0.16)** & ( > 0.04) & (0.13) \\ lgdppc & 0.062 & 0.195 & 0.111 & 0.187 & 0.092 > & 0.192 & 0.124 & 0.213 & 0.171 & 0.256 & > 0.126 & 0.204 \\ & (0.04) & (0.07)***& (0.03)***& (0.05)***& (0.03)** > *& (0.09)** & (0.04)***& (0.06)***& (0.08)** & (0.09)***& ( > 0.03)***& (0.07)***\\ llifeexp & -0.055 & -0.756 & -0.498 & -0.524 & -0.192 > & -1.539 & -0.540 & -1.148 & & & - > 0.301 & -0.837 \\ & (0.16) & (0.52) & (0.18)***& (0.37) & (0.16) > & (1.07) & (0.24)** & (0.65)* & & & ( > 0.26) & (0.54) \\ ldepratio1 & -0.387 & -0.457 & -0.380 & -0.297 & -0.399 > & -0.744 & -0.288 & -0.702 & -0.383 & -0.847 & > & \\ & (0.14)***& (0.21)** & (0.15)** & (0.22) & (0.15)** > *& (0.27)***& (0.10)***& (0.24)***& (0.15)** & (0.34)** & > & \\ lgovt_Y3 & -0.641 & -1.720 & -1.247 & -1.551 & -0.561 > & -1.422 & -1.069 & -1.421 & -0.975 & -1.566 & - > 1.047 & -1.563 \\ & (0.25)** & (0.91)* & (0.34)***& (0.94) & (0.27)** > & (0.87) & (0.43)** & (0.84)* & (0.44)** & (0.98) & ( > 0.42)** & (0.87)* \\ lcell & 0.008 & -0.059 & & & > & & & & & & > & \\ & (0.01) & (0.03)** & & & > & & & & & & > & \\ period2 & -0.423 & -0.483 & & & -0.479 > & -0.646 & -0.413 & -0.478 & -0.411 & -0.474 & - > 0.416 & -0.496 \\ & (0.13)***& (0.18)***& & & (0.15)** > *& (0.21)***& (0.12)***& (0.17)***& (0.13)***& (0.17)***& ( > 0.12)***& (0.16)***\\ period3 & -0.485 & -0.535 & & & -0.549 > & -0.699 & -0.476 & -0.523 & -0.486 & -0.543 & - > 0.483 & -0.564 \\ & (0.14)***& (0.19)***& & & (0.16)** > *& (0.21)***& (0.13)***& (0.17)***& (0.14)***& (0.17)***& ( > 0.13)***& (0.16)***\\ period4 & -0.453 & -0.484 & & & -0.504 > & -0.633 & -0.440 & -0.482 & -0.440 & -0.497 & - > 0.439 & -0.502 \\ & (0.14)***& (0.18)***& & & (0.15)** > *& (0.21)***& (0.12)***& (0.17)***& (0.13)***& (0.17)***& ( > 0.13)***& (0.16)***\\ period5 & -0.503 & -0.520 & 0.098 & 0.112 & -0.547 > & -0.671 & -0.477 & -0.524 & -0.484 & -0.539 & - > 0.476 & -0.543 \\ & (0.14)***& (0.19)***& (0.04)** & (0.05)** & (0.16)** > *& (0.22)***& (0.13)***& (0.18)***& (0.14)***& (0.17)***& ( > 0.13)***& (0.17)***\\ period6 & -0.532 & -0.498 & 0.071 & 0.076 & -0.569 > & -0.704 & -0.496 & -0.546 & -0.500 & -0.551 & - > 0.490 & -0.555 \\ & (0.14)***& (0.18)***& (0.02)***& (0.04)* & (0.16)** > *& (0.22)***& (0.13)***& (0.17)***& (0.14)***& (0.17)***& ( > 0.13)***& (0.16)***\\ period7 & -0.576 & -0.445 & 0.045 & 0.057 & -0.612 > & -0.747 & -0.521 & -0.595 & -0.525 & -0.586 & - > 0.508 & -0.591 \\ & (0.15)***& (0.18)** & (0.02)** & (0.03)** & (0.17)** > *& (0.22)***& (0.13)***& (0.18)***& (0.15)***& (0.18)***& ( > 0.13)***& (0.17)***\\ period8 & -0.634 & -0.432 & & & -0.670 > & -0.804 & -0.571 & -0.675 & -0.574 & -0.638 & - > 0.554 & -0.659 \\ & (0.16)***& (0.19)** & & & (0.17)** > *& (0.23)***& (0.14)***& (0.19)***& (0.15)***& (0.19)***& ( > 0.14)***& (0.18)***\\ lroads & & & -0.005 & -0.199 & > & & & & & & > & \\ & & & (0.02) & (0.10)** & > & & & & & & > & \\ avg5_JID_safrica& & & & & 0.0 > 60 & -0.076 & & & & & > & \\ & & & & & (0.08) > & (0.19) & & & & & > & \\ avg5_JID_laam& & & & & -0.114 > & -0.040 & & & & & > & \\ & & & & & (0.05)** > & (0.08) & & & & & > & \\ avg5_JID_asiae& & & & & 0.038 > & 0.042 & & & & & > & \\ & & & & & (0.06) > & (0.08) & & & & & > & \\ lenroll_s & & & & & > & & 0.012 & -0.106 & & & > & \\ & & & & & > & & (0.02) & (0.08) & & & > & \\ lmort_ufive & & & & & > & & & & 0.082 & 0.210 & > & \\ & & & & & > & & & & (0.06) & (0.12)* & > & \\ ldepratioyoung& & & & & > & & & & & & > -0.135 & -0.288 \\ & & & & & > & & & & & & ( > 0.07)* & (0.21) \\ ldepratioold& & & & & > & & & & & & - > 0.049 & -0.108 \\ & & & & & > & & & & & & ( > 0.08) & (0.15) \\ Constant & 2.171 & 5.740 & 2.877 & 4.248 & 2.585 > & 10.427 & 3.435 & 7.538 & 0.989 & 3.784 & > 2.120 & 5.139 \\ & (1.00)** & (2.69)** & (1.03)***& (2.15)* & (0.94)** > *& (4.99)** & (1.05)***& (3.01)** & (0.94) & (1.62)** & ( > 0.85)** & (2.72)* \\ Adjusted R^2& 0.240 & & & & > & & & & 0.239 & & > & \\ F & 4.187 & 4.033 & & 3.925 & > & 5.009 & & 3.630 & 4.434 & 3.746 & > & 3.944 \\ F p-value & & 0.000 & & 0.000 & > & 0.000 & & 0.000 & & 0.000 & > & 0.000 \\ Hansen J & & 65.783 & & 59.264 & > & 66.745 & & 56.867 & & 53.972 & > & 72.504 \\ J p-value & & 0.415 & & 0.575 & > & 0.228 & & 0.555 & & 0.661 & > & 0.272 \\ AR(2) & & -2.133 & & -1.693 & > & -1.445 & & -0.203 & & -1.107 & > & -0.499 \\ AR(2) p & & 0.033 & & 0.091 & > & 0.149 & & 0.840 & & 0.268 & > & 0.618 \\ Instruments & & 80 & & 74 & > & 77 & & 74 & & 74 & > & 82 \\ N & 658 & 658 & 340 & 340 & 512 > & 512 & 635 & 635 & 662 & 662 & > 665 & 665 \\ . estimates clear . . /* Fundamental determinants regressions */ . /* Benchmark fundamentals */ . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex lgeog (linst4 ltrade = avg5_JID_logem4 avg5_JID_logfrankromr), robust ffirs > t Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 2, 38) P-value linst4 | 0.3430 | 0.3887 | 6.99 0.0026 ltrade | 0.2953 | 0.3347 | 9.80 0.0004 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 14.08 0.0002 Cragg-Donald N*CDEV statistic 16.73 0.0000 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 7.57 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(2,38)= 2.07 P-val=0.1402 Anderson-Rubin test Chi-sq(2)=4.58 P-val=0.1015 Stock-Wright S statistic Chi-sq(2)=4.24 P-val=0.1198 NB: Test statistics heteroskedasticity-robust Number of observations N = 42 Number of regressors K = 4 Number of instruments L = 4 Number of excluded instruments L1 = 2 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 42 F( 3, 38) = 1.77 Prob > F = 0.1700 Total (centered) SS = 5.271066734 Centered R2 = 0.1270 Total (uncentered) SS = 12.25133364 Uncentered R2 = 0.6244 Residual SS = 4.601418695 Root MSE = .331 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 2.148406 .9032722 2.38 0.017 .3780252 3.918787 ltrade | -.4690561 .2390265 -1.96 0.050 -.9375394 -.0005727 lgeog | -.0786451 .0701923 -1.12 0.263 -.2162195 .0589293 _cons | -2.72825 1.267908 -2.15 0.031 -5.213303 -.243196 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 14.084 Chi-sq(1) P-val = 0.0002 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 7.569 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store F1 . drop epsilon . predict epsilon, resid (164 missing values generated) . sktest epsilon /* Test for normality of residuals */ Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.104 0.214 4.30 0.1167 . ivreg2 lindex lgeog (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankr > omr), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 69) P-value linst4 | 0.1047 | 0.1883 | 3.31 0.0250 ltrade | 0.2131 | 0.3832 | 13.07 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 7.87 0.0195 Cragg-Donald N*CDEV statistic 8.31 0.0157 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 2.58 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,69)= 1.19 P-val=0.3195 Anderson-Rubin test Chi-sq(3)=3.83 P-val=0.2801 Stock-Wright S statistic Chi-sq(3)=2.99 P-val=0.3934 NB: Test statistics heteroskedasticity-robust Number of observations N = 74 Number of regressors K = 4 Number of instruments L = 5 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 74 F( 3, 70) = 2.09 Prob > F = 0.1098 Total (centered) SS = 8.65972659 Centered R2 = -0.5704 Total (uncentered) SS = 25.43445996 Uncentered R2 = 0.4653 Residual SS = 13.598879 Root MSE = .4287 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 3.548113 1.686808 2.10 0.035 .2420305 6.854195 ltrade | -.5785547 .2972209 -1.95 0.052 -1.161097 .0039876 lgeog | -.2134089 .1600236 -1.33 0.182 -.5270494 .1002315 _cons | -4.39571 2.167599 -2.03 0.043 -8.644125 -.1472945 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 7.875 Chi-sq(2) P-val = 0.0195 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 2.583 Stock-Yogo weak ID test critical values: 10% maximal IV size 13.43 15% maximal IV size 8.18 20% maximal IV size 6.40 25% maximal IV size 5.45 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.821 Chi-sq(1) P-val = 0.3650 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store F2 . drop epsilon . predict epsilon, resid (164 missing values generated) . sktest epsilon Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.287 0.553 1.53 0.4648 . restore . xthtaylor lindex linst4 ltrade lgeog period2-period8, endog(linst4 ltrade) note: period2 dropped due to collinearity note: period3 dropped due to collinearity note: period4 dropped due to collinearity Hausman-Taylor estimation Number of obs = 366 Group variable (i): code2 Number of groups = 79 Obs per group: min = 1 avg = 4.6 max = 5 Random effects u_i ~ i.i.d. Wald chi2(7) = 29.38 Prob > chi2 = 0.0001 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- TVexogenous | period5 | -.02691 .0226514 -1.19 0.235 -.0713059 .0174859 period6 | -.0648502 .0275112 -2.36 0.018 -.1187712 -.0109292 period7 | -.0884755 .0264787 -3.34 0.001 -.1403728 -.0365782 period8 | -.1185606 .0294295 -4.03 0.000 -.1762414 -.0608798 TVendogenous | linst4 | .2018437 .085305 2.37 0.018 .0346489 .3690385 ltrade | .1190104 .0474486 2.51 0.012 .0260129 .2120079 TIexogenous | lgeog | .1023419 .0453341 2.26 0.024 .0134888 .1911951 | _cons | .0430257 .1718533 0.25 0.802 -.2938005 .379852 -------------+---------------------------------------------------------------- sigma_u | .35564487 sigma_e | .13205471 rho | .8788338 (fraction of variance due to u_i) ------------------------------------------------------------------------------ note: TV refers to time varying; TI refers to time invariant. . xtabond2 lindex l.lindex linst4 ltrade lgeog period2-period8, gmm(l(3/4).lindex linst4 lt > rade) iv(period* lgeog avg5_JID_eurfrac avg5_JID_engfrac) robust small twostep Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation > . Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, two-step system GMM ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 352 Time variable : countperiod Number of groups = 78 Number of instruments = 66 Obs per group: min = 0 F(8, 77) = 147.91 avg = 4.51 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Corrected lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex | L1. | .8280095 .050618 16.36 0.000 .7272161 .9288029 linst4 | .1534964 .0734439 2.09 0.040 .007251 .2997418 ltrade | .0269168 .0214039 1.26 0.212 -.0157038 .0695375 lgeog | -.0020748 .0066374 -0.31 0.755 -.0152915 .011142 period4 | .0661616 .0277355 2.39 0.020 .0109332 .12139 period5 | .0314442 .0125941 2.50 0.015 .0063661 .0565222 period6 | .0092924 .0220587 0.42 0.675 -.0346321 .0532168 period7 | -.0070051 .016681 -0.42 0.676 -.0402213 .0262111 _cons | -.1264248 .0834112 -1.52 0.134 -.2925178 .0396682 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.42 Pr > z = 0.016 Arellano-Bond test for AR(2) in first differences: z = -1.65 Pr > z = 0.100 Hansen test of overid. restrictions: chi2(57) = 60.17 Prob > chi2 = 0.362 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(41) = 48.73 Prob > chi2 = 0.190 Difference (null H = exogenous): chi2(16) = 11.44 Prob > chi2 = 0.781 ivstyle(period* lgeog avg5_JID_eurfrac avg5_JID_engfrac) Hansen test excluding group: chi2(49) = 47.36 Prob > chi2 = 0.540 Difference (null H = exogenous): chi2(8) = 12.81 Prob > chi2 = 0.119 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store F3 . . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex lgeog lelf (linst4 ltrade ldemoc = avg5_JID_logem4 avg5_JID_logfrankromr pl > democ), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 32) P-value linst4 | 0.3669 | 0.4266 | 3.98 0.0161 ltrade | 0.1975 | 0.3438 | 6.38 0.0016 ldemoc | 0.1698 | 0.2626 | 5.19 0.0049 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 5.28 0.0216 Cragg-Donald N*CDEV statistic 5.66 0.0173 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 1.59 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,32)= 1.89 P-val=0.1513 Anderson-Rubin test Chi-sq(3)=6.73 P-val=0.0811 Stock-Wright S statistic Chi-sq(3)=4.68 P-val=0.1971 NB: Test statistics heteroskedasticity-robust Number of observations N = 38 Number of regressors K = 6 Number of instruments L = 6 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 38 F( 5, 32) = 1.28 Prob > F = 0.2970 Total (centered) SS = 5.206912839 Centered R2 = 0.1788 Total (uncentered) SS = 11.9211273 Uncentered R2 = 0.6413 Residual SS = 4.276038156 Root MSE = .3355 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 2.255008 .8614226 2.62 0.009 .5666513 3.943366 ltrade | -.5074877 .2988024 -1.70 0.089 -1.09313 .0781543 ldemoc | -.1686806 .2894507 -0.58 0.560 -.7359935 .3986324 lgeog | -.0995381 .08212 -1.21 0.225 -.2604904 .0614141 lelf | .4127145 .3275239 1.26 0.208 -.2292205 1.054649 _cons | -2.748697 1.198434 -2.29 0.022 -5.097585 -.3998086 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 5.277 Chi-sq(1) P-val = 0.0216 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 1.589 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store F4 . drop epsilon . predict epsilon, resid (165 missing values generated) . sktest epsilon Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.065 0.146 5.36 0.0687 . ivreg2 lindex lgeog lelf (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfrac avg5_J > ID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 62) P-value linst4 | 0.0572 | 0.2731 | 5.03 0.0014 ltrade | 0.1353 | 0.4293 | 10.02 0.0000 ldemoc | 0.1005 | 0.3228 | 6.66 0.0002 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 3.26 0.1960 Cragg-Donald N*CDEV statistic 3.34 0.1885 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.75 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,62)= 2.41 P-val=0.0587 Anderson-Rubin test Chi-sq(4)=10.73 P-val=0.0298 Stock-Wright S statistic Chi-sq(4)=6.51 P-val=0.1641 NB: Test statistics heteroskedasticity-robust Number of observations N = 69 Number of regressors K = 6 Number of instruments L = 7 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 69 F( 5, 63) = 1.19 Prob > F = 0.3226 Total (centered) SS = 8.49124626 Centered R2 = -1.1307 Total (uncentered) SS = 24.97815506 Uncentered R2 = 0.2757 Residual SS = 18.09201881 Root MSE = .5121 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 4.66091 2.815182 1.66 0.098 -.8567443 10.17856 ltrade | -.7036274 .4278042 -1.64 0.100 -1.542108 .1348535 ldemoc | -.4593008 .5535636 -0.83 0.407 -1.544266 .625664 lgeog | -.2474184 .2104208 -1.18 0.240 -.6598356 .1649988 lelf | .6755554 .5006586 1.35 0.177 -.3057173 1.656828 _cons | -5.513889 3.164038 -1.74 0.081 -11.71529 .6875111 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 3.259 Chi-sq(2) P-val = 0.1960 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.750 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.208 Chi-sq(1) P-val = 0.6483 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store F5 . drop epsilon . predict epsilon, resid (165 missing values generated) . sktest epsilon Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.219 0.614 1.83 0.4012 . restore . xthtaylor lindex linst4 ltrade ldemoc lgeog lelf period2-period8, endog(linst4 ltrade lde > moc) note: period2 dropped due to collinearity note: period3 dropped due to collinearity note: period4 dropped due to collinearity Hausman-Taylor estimation Number of obs = 361 Group variable (i): code2 Number of groups = 78 Obs per group: min = 1 avg = 4.6 max = 5 Random effects u_i ~ i.i.d. Wald chi2(9) = 30.81 Prob > chi2 = 0.0003 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- TVexogenous | period5 | -.027395 .0229925 -1.19 0.233 -.0724595 .0176694 period6 | -.0661399 .0279286 -2.37 0.018 -.1208789 -.0114008 period7 | -.0891192 .0267606 -3.33 0.001 -.141569 -.0366693 period8 | -.1223348 .0299441 -4.09 0.000 -.1810241 -.0636454 TVendogenous | linst4 | .1778204 .0915715 1.94 0.052 -.0016565 .3572973 ltrade | .1081706 .0491048 2.20 0.028 .0119269 .2044143 ldemoc | .0424804 .0352592 1.20 0.228 -.0266263 .1115872 TIexogenous | lgeog | .0889057 .0542734 1.64 0.101 -.0174681 .1952796 lelf | -.0627959 .2800307 -0.22 0.823 -.611646 .4860542 | _cons | .0734815 .255176 0.29 0.773 -.4266542 .5736172 -------------+---------------------------------------------------------------- sigma_u | .35394549 sigma_e | .13264247 rho | .87685414 (fraction of variance due to u_i) ------------------------------------------------------------------------------ note: TV refers to time varying; TI refers to time invariant. . xtabond2 lindex l.lindex linst4 ltrade ldemoc lgeog lelf period2-period8, gmm(l(4/5).lind > ex ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1 avg5_JID_engfrac avg5_JID_eurfrac > avg5_JID_logfrankromr) robust small orthog twostep Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation > . Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, two-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 194 Time variable : countperiod Number of groups = 43 Number of instruments = 77 Obs per group: min = 0 F(10, 42) = 54.35 avg = 4.51 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Corrected lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex | L1. | .8053925 .0562516 14.32 0.000 .6918721 .9189128 linst4 | .155015 .0927545 1.67 0.102 -.0321713 .3422012 ltrade | -.0182018 .0317439 -0.57 0.569 -.0822636 .0458601 ldemoc | .0263478 .0793643 0.33 0.742 -.1338158 .1865114 lgeog | -.0091093 .0133033 -0.68 0.497 -.0359565 .0177379 lelf | .0656128 .0751841 0.87 0.388 -.0861148 .2173404 period4 | .1013866 .0558906 1.81 0.077 -.0114052 .2141783 period5 | .0239398 .030099 0.80 0.431 -.0368024 .084682 period6 | .0406903 .0472982 0.86 0.395 -.0547613 .1361419 period7 | .0088686 .0399527 0.22 0.825 -.0717592 .0894965 _cons | -.1941449 .1544392 -1.26 0.216 -.5058157 .117526 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.66 Pr > z = 0.008 Arellano-Bond test for AR(2) in first differences: z = -1.65 Pr > z = 0.099 Hansen test of overid. restrictions: chi2(66) = 32.58 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(48) = 29.93 Prob > chi2 = 0.981 Difference (null H = exogenous): chi2(18) = 2.65 Prob > chi2 = 1.000 gmmstyle(l(4/5).lindex ltrade linst4 ldemoc) Hansen test excluding group: chi2(0) = 0.00 Prob > chi2 = . Difference (null H = exogenous): chi2(66) = 32.58 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1 avg5_JID_engfrac avg5_JID_eurfrac avg5_JID_logfrankrom > r) Hansen test excluding group: chi2(55) = 28.09 Prob > chi2 = 0.999 Difference (null H = exogenous): chi2(11) = 4.49 Prob > chi2 = 0.953 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store F6 . . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex lgeog ltrust1 (linst4 ltrade = avg5_JID_logem4 avg5_JID_logfrankromr), robu > st ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 2, 13) P-value linst4 | 0.1452 | 0.4447 | 6.07 0.0137 ltrade | 0.1349 | 0.4132 | 4.64 0.0302 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 2.13 0.1448 Cragg-Donald N*CDEV statistic 2.26 0.1331 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.81 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(2,13)= 1.31 P-val=0.3040 Anderson-Rubin test Chi-sq(2)=3.62 P-val=0.1637 Stock-Wright S statistic Chi-sq(2)=2.33 P-val=0.3114 NB: Test statistics heteroskedasticity-robust Number of observations N = 18 Number of regressors K = 5 Number of instruments L = 5 Number of excluded instruments L1 = 2 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 18 F( 4, 13) = 1.22 Prob > F = 0.3488 Total (centered) SS = 4.639390671 Centered R2 = -0.7648 Total (uncentered) SS = 9.690325114 Uncentered R2 = 0.1551 Residual SS = 8.187401951 Root MSE = .6744 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 5.465474 2.375504 2.30 0.021 .8095723 10.12138 ltrade | -1.278256 .694427 -1.84 0.066 -2.639307 .0827964 lgeog | -.2855713 .3202248 -0.89 0.373 -.9132003 .3420577 ltrust1 | .0605327 2.356129 0.03 0.980 -4.557394 4.67846 _cons | -7.698443 3.37372 -2.28 0.022 -14.31081 -1.086073 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 2.126 Chi-sq(1) P-val = 0.1448 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.815 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog ltrust1 Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr ------------------------------------------------------------------------------ . ivreg2 lindex lgeog lelf ltrust1 (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfra > c avg5_JID_logfrankromr pldemoc), ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 32) P-value linst4 | 0.0335 | 0.5218 | 8.73 0.0001 ltrade | 0.1293 | 0.5590 | 10.14 0.0000 ldemoc | 0.0098 | 0.1148 | 1.04 0.4034 Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 0.33 0.8494 Cragg-Donald N*CDEV statistic 0.33 0.8488 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.07 See main output for Cragg-Donald weak id test critical values Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,32)= 5.50 P-val=0.0017 Anderson-Rubin test Chi-sq(4)=27.48 P-val=0.0000 Stock-Wright S statistic Chi-sq(4)=16.29 P-val=0.0027 Number of observations N = 40 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Number of obs = 40 F( 6, 33) = 0.18 Prob > F = 0.9809 Total (centered) SS = 6.996875177 Centered R2 = -12.3656 Total (uncentered) SS = 21.55105173 Uncentered R2 = -3.3394 Residual SS = 93.51761211 Root MSE = 1.529 ------------------------------------------------------------------------------ lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 9.742437 13.04902 0.75 0.455 -15.83317 35.31804 ltrade | -1.080139 1.389811 -0.78 0.437 -3.804119 1.643841 ldemoc | -4.007807 7.961956 -0.50 0.615 -19.61295 11.59734 lgeog | .3899692 1.079992 0.36 0.718 -1.726776 2.506714 lelf | .7095477 1.870237 0.38 0.704 -2.956049 4.375144 ltrust1 | -4.211003 6.482472 -0.65 0.516 -16.91641 8.494407 _cons | -9.048384 10.53032 -0.86 0.390 -29.68744 11.59067 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 0.326 Chi-sq(2) P-val = 0.8494 ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.066 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Sargan statistic (overidentification test of all instruments): 0.024 Chi-sq(1) P-val = 0.8764 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf ltrust1 Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store F7 . drop epsilon . predict epsilon, resid (197 missing values generated) . sktest epsilon Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.005 0.012 11.46 0.0032 . restore . xtabond2 lindex l.lindex linst4 ltrade ldemoc lgeog lelf ltrust1 period2-period8, gmm(l(4 > /6).lindex ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 199 Time variable : countperiod Number of groups = 44 Number of instruments = 75 Obs per group: min = 1 F(11, 43) = 77.25 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex | L1. | .7215153 .0561605 12.85 0.000 .6082569 .8347737 linst4 | .1957727 .1192264 1.64 0.108 -.0446702 .4362156 ltrade | -.0424748 .0566883 -0.75 0.458 -.1567977 .0718481 ldemoc | .059381 .0414811 1.43 0.160 -.0242736 .1430357 lgeog | -.017839 .0211767 -0.84 0.404 -.0605459 .0248679 lelf | .028445 .0939737 0.30 0.764 -.1610711 .2179611 ltrust1 | .0574563 .1122066 0.51 0.611 -.16883 .2837425 period4 | .1199913 .0484485 2.48 0.017 .0222856 .217697 period5 | .0309625 .0351602 0.88 0.383 -.0399448 .1018698 period6 | .0428269 .04037 1.06 0.295 -.0385869 .1242407 period7 | .0109058 .0326159 0.33 0.740 -.0548704 .0766819 _cons | -.2441992 .2085686 -1.17 0.248 -.6648178 .1764195 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.21 Pr > z = 0.027 Arellano-Bond test for AR(2) in first differences: z = -1.60 Pr > z = 0.109 Hansen test of overid. restrictions: chi2(63) = 27.63 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(44) = 30.73 Prob > chi2 = 0.935 Difference (null H = exogenous): chi2(19) = -3.10 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(55) = 32.76 Prob > chi2 = 0.993 Difference (null H = exogenous): chi2(8) = -5.13 Prob > chi2 = 1.000 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store F8 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex lgeog (linst4 ltrade lsch = avg5_JID_logem4 avg5_JID_logfrankromr lenroll19 > 00), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 10) P-value linst4 | 0.4463 | 0.7061 | 12.51 0.0010 ltrade | 0.3138 | 0.4807 | 5.47 0.0174 lsch | 0.3968 | 0.8832 | 23.60 0.0001 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 4.46 0.0347 Cragg-Donald N*CDEV statistic 5.19 0.0227 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 1.15 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,10)= 0.85 P-val=0.4995 Anderson-Rubin test Chi-sq(3)=3.81 P-val=0.2830 Stock-Wright S statistic Chi-sq(3)=2.91 P-val=0.4054 NB: Test statistics heteroskedasticity-robust Number of observations N = 15 Number of regressors K = 5 Number of instruments L = 5 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 15 F( 4, 10) = 1.06 Prob > F = 0.4249 Total (centered) SS = 4.520053037 Centered R2 = 0.2199 Total (uncentered) SS = 9.247733312 Uncentered R2 = 0.6187 Residual SS = 3.526302156 Root MSE = .4849 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .6143283 2.184087 0.28 0.778 -3.666404 4.895061 ltrade | -1.070942 .6330931 -1.69 0.091 -2.311781 .169898 lsch | 1.444427 .9949391 1.45 0.147 -.5056174 3.394472 lgeog | -.5606026 .5602607 -1.00 0.317 -1.658693 .5374883 _cons | -2.738414 1.707249 -1.60 0.109 -6.084561 .6077337 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 4.459 Chi-sq(1) P-val = 0.0347 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 1.154 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade lsch Included instruments: lgeog Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr lenroll1900 ------------------------------------------------------------------------------ . ivreg2 lindex lgeog lelf (linst4 ltrade ldemoc lsch = avg5_JID_eurfrac avg5_JID_engfrac a > vg5_JID_logfrankromr pldemoc lenroll1900), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 5, 25) P-value linst4 | 0.1273 | 0.3238 | 2.88 0.0344 ltrade | 0.4514 | 0.6613 | 14.43 0.0000 ldemoc | 0.0817 | 0.0946 | 0.62 0.6828 lsch | 0.2525 | 0.7574 | 12.02 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 1.61 0.4476 Cragg-Donald N*CDEV statistic 1.65 0.4387 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.25 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(5,25)= 1.63 P-val=0.1896 Anderson-Rubin test Chi-sq(5)=10.74 P-val=0.0569 Stock-Wright S statistic Chi-sq(5)=5.46 P-val=0.3628 NB: Test statistics heteroskedasticity-robust Number of observations N = 33 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 5 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 33 F( 6, 26) = 1.30 Prob > F = 0.2932 Total (centered) SS = 6.434999212 Centered R2 = -0.2728 Total (uncentered) SS = 19.05975704 Uncentered R2 = 0.5703 Residual SS = 8.190617225 Root MSE = .4982 ------------------------------------------------------------------------------ | Robust lindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .9227412 2.350374 0.39 0.695 -3.683907 5.529389 ltrade | -.4300793 .3084879 -1.39 0.163 -1.034704 .1745458 ldemoc | 3.304411 2.775759 1.19 0.234 -2.135976 8.744799 lsch | .7448414 .7862164 0.95 0.343 -.7961145 2.285797 lgeog | -.3643294 .6260509 -0.58 0.561 -1.591367 .8627079 lelf | .6739536 .6171389 1.09 0.275 -.5356164 1.883524 _cons | -7.198017 4.751358 -1.51 0.130 -16.51051 2.114473 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 1.608 Chi-sq(2) P-val = 0.4476 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.250 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 1.022 Chi-sq(1) P-val = 0.3120 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc lsch Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc lenroll1900 ------------------------------------------------------------------------------ . estimates store F9 . drop epsilon . predict epsilon, resid (166 missing values generated) . sktest epsilon Skewness/Kurtosis tests for Normality ------- joint ------ Variable | Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2 -------------+------------------------------------------------------- epsilon | 0.000 0.002 22.59 0.0000 . restore . xtabond2 lindex l.lindex linst4 ltrade ldemoc lsch lgeog lelf period2-period8, gmm(l(1/2) > .lindex ltrade linst4 ldemoc lsch) iv(period* lgeog lelf) robust small orthog twostep Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation > . Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, two-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 350 Time variable : countperiod Number of groups = 77 Number of instruments = 117 Obs per group: min = 1 F(11, 76) = 222.54 avg = 4.55 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Corrected lindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex | L1. | .8371984 .0338107 24.76 0.000 .7698586 .9045383 linst4 | .1409265 .0658492 2.14 0.036 .0097765 .2720765 ltrade | .0084272 .0256383 0.33 0.743 -.042636 .0594904 ldemoc | .0219137 .0202003 1.08 0.281 -.0183187 .0621461 lsch | .0092913 .0357063 0.26 0.795 -.061824 .0804066 lgeog | .0008198 .0076344 0.11 0.915 -.0143854 .016025 lelf | .0777038 .0478886 1.62 0.109 -.0176745 .1730822 period4 | .0816508 .0339997 2.40 0.019 .0139346 .1493671 period5 | .0362407 .0194863 1.86 0.067 -.0025697 .0750511 period6 | .0170007 .0220798 0.77 0.444 -.026975 .0609764 period7 | -.0034133 .0151594 -0.23 0.822 -.0336058 .0267792 _cons | -.2137098 .1123014 -1.90 0.061 -.4373774 .0099579 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.54 Pr > z = 0.011 Arellano-Bond test for AR(2) in first differences: z = -1.67 Pr > z = 0.096 Hansen test of overid. restrictions: chi2(105) = 63.33 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(77) = 65.41 Prob > chi2 = 0.824 Difference (null H = exogenous): chi2(28) = -2.08 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf) Hansen test excluding group: chi2(99) = 61.17 Prob > chi2 = 0.999 Difference (null H = exogenous): chi2(6) = 2.16 Prob > chi2 = 0.904 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store F10 . . estout * using "C:\Llama\Output\benchmark2.tex", replace style(tex) varlabels(_cons Const > ant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p idstat idp cdf sarga > n sarganp ar2 ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Anderson LR" "LR p-val > ue" "Cragg-Donald F" "Hansen J" "J p-value" "AR(2)" "AR(2) p" "Instruments" "N") fmt(%8.3 > f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels(* 0.10 * > * 0.05 *** 0.01) & F1 & F2 & F3 & F4 & F5 > & F6 & F7 & F8 & F9 & F10 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se \\ linst4 & 2.148 & 3.548 & 0.153 & 2.255 & 4.661 > & 0.155 & 9.742 & 0.196 & 0.923 & 0.141 \\ & (0.90)** & (1.69)** & (0.07)** & (0.86)***& (2.82)* > & (0.09) & (13.05) & (0.12) & (2.35) & (0.07)** \\ ltrade & -0.469 & -0.579 & 0.027 & -0.507 & -0.704 > & -0.018 & -1.080 & -0.042 & -0.430 & 0.008 \\ & (0.24)** & (0.30)* & (0.02) & (0.30)* & (0.43) > & (0.03) & (1.39) & (0.06) & (0.31) & (0.03) \\ lgeog & -0.079 & -0.213 & -0.002 & -0.100 & -0.247 > & -0.009 & 0.390 & -0.018 & -0.364 & 0.001 \\ & (0.07) & (0.16) & (0.01) & (0.08) & (0.21) > & (0.01) & (1.08) & (0.02) & (0.63) & (0.01) \\ L.lindex & & & 0.828 & & > & 0.805 & & 0.722 & & 0.837 \\ & & & (0.05)***& & > & (0.06)***& & (0.06)***& & (0.03)***\\ period4 & & & 0.066 & & > & 0.101 & & 0.120 & & 0.082 \\ & & & (0.03)** & & > & (0.06)* & & (0.05)** & & (0.03)** \\ period5 & & & 0.031 & & > & 0.024 & & 0.031 & & 0.036 \\ & & & (0.01)** & & > & (0.03) & & (0.04) & & (0.02)* \\ period6 & & & 0.009 & & > & 0.041 & & 0.043 & & 0.017 \\ & & & (0.02) & & > & (0.05) & & (0.04) & & (0.02) \\ period7 & & & -0.007 & & > & 0.009 & & 0.011 & & -0.003 \\ & & & (0.02) & & > & (0.04) & & (0.03) & & (0.02) \\ ldemoc & & & & -0.169 & -0.459 > & 0.026 & -4.008 & 0.059 & 3.304 & 0.022 \\ & & & & (0.29) & (0.55) > & (0.08) & (7.96) & (0.04) & (2.78) & (0.02) \\ lelf & & & & 0.413 & 0.676 > & 0.066 & 0.710 & 0.028 & 0.674 & 0.078 \\ & & & & (0.33) & (0.50) > & (0.08) & (1.87) & (0.09) & (0.62) & (0.05) \\ ltrust1 & & & & & > & & -4.211 & 0.057 & & \\ & & & & & > & & (6.48) & (0.11) & & \\ lsch & & & & & > & & & & 0.745 & 0.009 \\ & & & & & > & & & & (0.79) & (0.04) \\ Constant & -2.728 & -4.396 & -0.126 & -2.749 & -5.514 > & -0.194 & -9.048 & -0.244 & -7.198 & -0.214 \\ & (1.27)** & (2.17)** & (0.08) & (1.20)** & (3.16)* > & (0.15) & (10.53) & (0.21) & (4.75) & (0.11)* \\ Adjusted R^2& 0.058 & -0.638 & & 0.050 & -1.300 > & & -14.796 & & -0.567 & \\ F & 1.767 & 2.087 & 147.909 & 1.279 & 1.194 > & 54.354 & 0.178 & 77.248 & 1.297 & 222.541 \\ F p-value & & & 0.000 & & > & 0.000 & & 0.000 & & 0.000 \\ Anderson LR & 14.084 & 7.875 & & 5.277 & 3.259 > & & 0.326 & & 1.608 & \\ LR p-value & 0.000 & 0.019 & & 0.022 & 0.196 > & & 0.849 & & 0.448 & \\ Cragg-Donald F& 7.569 & 2.583 & & 1.589 & 0.750 > & & 0.066 & & 0.250 & \\ Hansen J & & & 60.170 & & > & 32.579 & 0.024 & 27.631 & & 63.330 \\ J p-value & & & 0.362 & & > & 1.000 & 0.876 & 1.000 & & 1.000 \\ AR(2) & & & -1.647 & & > & -1.650 & & -1.601 & & -1.667 \\ AR(2) p & & & 0.100 & & > & 0.099 & & 0.109 & & 0.096 \\ Instruments & & & 66 & & > & 77 & & 75 & & 117 \\ N & 42 & 74 & 352 & 38 & 69 > & 194 & 40 & 199 & 33 & 350 \\ . estimates clear . . /* Alternative fundamentals */ . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex2 lgeog (linst4 ltrade = avg5_JID_logem4 avg5_JID_logfrankromr), robust ffir > st Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 2, 38) P-value linst4 | 0.3430 | 0.3887 | 6.99 0.0026 ltrade | 0.2953 | 0.3347 | 9.80 0.0004 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 14.08 0.0002 Cragg-Donald N*CDEV statistic 16.73 0.0000 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 7.57 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(2,38)= 1.56 P-val=0.2231 Anderson-Rubin test Chi-sq(2)=3.45 P-val=0.1782 Stock-Wright S statistic Chi-sq(2)=3.45 P-val=0.1783 NB: Test statistics heteroskedasticity-robust Number of observations N = 42 Number of regressors K = 4 Number of instruments L = 4 Number of excluded instruments L1 = 2 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 42 F( 3, 38) = 1.30 Prob > F = 0.2888 Total (centered) SS = 2.413499571 Centered R2 = 0.1507 Total (uncentered) SS = 61.72245544 Uncentered R2 = 0.9668 Residual SS = 2.049897948 Root MSE = .2209 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.301846 .633596 2.05 0.040 .0600209 2.543671 ltrade | -.302674 .1713578 -1.77 0.077 -.6385292 .0331811 lgeog | -.0467779 .0442139 -1.06 0.290 -.1334355 .0398797 _cons | -.7296204 .9015902 -0.81 0.418 -2.496705 1.037464 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 14.084 Chi-sq(1) P-val = 0.0002 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 7.569 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store AF1 . ivreg2 lindex2 lgeog (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrank > romr), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 69) P-value linst4 | 0.1047 | 0.1883 | 3.31 0.0250 ltrade | 0.2131 | 0.3832 | 13.07 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 7.87 0.0195 Cragg-Donald N*CDEV statistic 8.31 0.0157 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 2.58 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,69)= 1.17 P-val=0.3267 Anderson-Rubin test Chi-sq(3)=3.77 P-val=0.2872 Stock-Wright S statistic Chi-sq(3)=3.11 P-val=0.3743 NB: Test statistics heteroskedasticity-robust Number of observations N = 74 Number of regressors K = 4 Number of instruments L = 5 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 74 F( 3, 70) = 1.76 Prob > F = 0.1619 Total (centered) SS = 3.597109054 Centered R2 = -0.7429 Total (uncentered) SS = 114.4055858 Uncentered R2 = 0.9452 Residual SS = 6.269546724 Root MSE = .2911 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 2.41348 1.184095 2.04 0.042 .092697 4.734264 ltrade | -.39003 .2108975 -1.85 0.064 -.8033815 .0233214 lgeog | -.1559203 .1118616 -1.39 0.163 -.375165 .0633243 _cons | -2.054957 1.52387 -1.35 0.177 -5.041687 .9317721 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 7.875 Chi-sq(2) P-val = 0.0195 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 2.583 Stock-Yogo weak ID test critical values: 10% maximal IV size 13.43 15% maximal IV size 8.18 20% maximal IV size 6.40 25% maximal IV size 5.45 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.749 Chi-sq(1) P-val = 0.3866 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store AF2 . restore . xtabond2 lindex2 l.lindex2 linst4 ltrade lgeog period2-period8, gmm(l(1/4).lindex2 linst4 > ltrade) iv(period* lgeog) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 359 Time variable : countperiod Number of groups = 79 Number of instruments = 66 Obs per group: min = 1 F(8, 78) = 185.47 avg = 4.54 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex2 | L1. | .8083485 .0424569 19.04 0.000 .7238233 .8928738 linst4 | .116351 .0475037 2.45 0.017 .0217784 .2109235 ltrade | .0063785 .0175956 0.36 0.718 -.0286516 .0414086 lgeog | .0010424 .0038559 0.27 0.788 -.006634 .0087188 period4 | .048967 .0165105 2.97 0.004 .0160971 .0818369 period5 | -.000312 .0123592 -0.03 0.980 -.0249174 .0242934 period6 | -.0056083 .0197059 -0.28 0.777 -.0448397 .0336231 period7 | -.0198048 .0129127 -1.53 0.129 -.045512 .0059025 _cons | .0735192 .0548911 1.34 0.184 -.0357606 .1827989 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.88 Pr > z = 0.060 Arellano-Bond test for AR(2) in first differences: z = -1.39 Pr > z = 0.166 Hansen test of overid. restrictions: chi2(57) = 64.12 Prob > chi2 = 0.241 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(38) = 48.48 Prob > chi2 = 0.119 Difference (null H = exogenous): chi2(19) = 15.64 Prob > chi2 = 0.681 ivstyle(period* lgeog) Hansen test excluding group: chi2(51) = 56.03 Prob > chi2 = 0.292 Difference (null H = exogenous): chi2(6) = 8.09 Prob > chi2 = 0.231 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AF3 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex2 lgeog lelf (linst4 ltrade ldemoc = avg5_JID_logem4 avg5_JID_logfrankromr p > ldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 32) P-value linst4 | 0.3669 | 0.4266 | 3.98 0.0161 ltrade | 0.1975 | 0.3438 | 6.38 0.0016 ldemoc | 0.1698 | 0.2626 | 5.19 0.0049 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 5.28 0.0216 Cragg-Donald N*CDEV statistic 5.66 0.0173 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 1.59 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,32)= 1.32 P-val=0.2843 Anderson-Rubin test Chi-sq(3)=4.71 P-val=0.1943 Stock-Wright S statistic Chi-sq(3)=3.67 P-val=0.2990 NB: Test statistics heteroskedasticity-robust Number of observations N = 38 Number of regressors K = 6 Number of instruments L = 6 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 38 F( 5, 32) = 0.94 Prob > F = 0.4698 Total (centered) SS = 2.391837051 Centered R2 = 0.2109 Total (uncentered) SS = 56.71960533 Uncentered R2 = 0.9667 Residual SS = 1.88730993 Root MSE = .2229 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.351587 .6130798 2.20 0.027 .1499727 2.553201 ltrade | -.3180617 .1995085 -1.59 0.111 -.7090912 .0729677 ldemoc | -.077899 .168489 -0.46 0.644 -.4081313 .2523333 lgeog | -.0595994 .0513143 -1.16 0.245 -.1601736 .0409748 lelf | .2465666 .1914751 1.29 0.198 -.1287177 .6218509 _cons | -.7455817 .8600683 -0.87 0.386 -2.431285 .9401213 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 5.277 Chi-sq(1) P-val = 0.0216 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 1.589 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store AF4 . ivreg2 lindex2 lgeog lelf (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfrac avg5_ > JID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 62) P-value linst4 | 0.0572 | 0.2731 | 5.03 0.0014 ltrade | 0.1353 | 0.4293 | 10.02 0.0000 ldemoc | 0.1005 | 0.3228 | 6.66 0.0002 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 3.26 0.1960 Cragg-Donald N*CDEV statistic 3.34 0.1885 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.75 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,62)= 2.05 P-val=0.0979 Anderson-Rubin test Chi-sq(4)=9.14 P-val=0.0577 Stock-Wright S statistic Chi-sq(4)=6.18 P-val=0.1861 NB: Test statistics heteroskedasticity-robust Number of observations N = 69 Number of regressors K = 6 Number of instruments L = 7 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 69 F( 5, 63) = 0.94 Prob > F = 0.4591 Total (centered) SS = 3.544377625 Centered R2 = -1.3946 Total (uncentered) SS = 108.0707522 Uncentered R2 = 0.9215 Residual SS = 8.487190901 Root MSE = .3507 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 3.176322 1.954915 1.62 0.104 -.655241 7.007886 ltrade | -.4803005 .3001262 -1.60 0.110 -1.068537 .107936 ldemoc | -.341456 .3811872 -0.90 0.370 -1.088569 .4056571 lgeog | -.1756379 .146793 -1.20 0.232 -.463347 .1120711 lelf | .4542345 .352204 1.29 0.197 -.2360725 1.144542 _cons | -2.796141 2.205898 -1.27 0.205 -7.119623 1.52734 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 3.259 Chi-sq(2) P-val = 0.1960 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.750 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.323 Chi-sq(1) P-val = 0.5698 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store AF5 . restore . xtabond2 lindex2 l.lindex2 linst4 ltrade ldemoc lgeog lelf period2-period8, gmm(l(2/4).li > ndex2 ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 199 Time variable : countperiod Number of groups = 44 Number of instruments = 73 Obs per group: min = 0 F(10, 43) = 88.94 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex2 | L1. | .6249594 .0614068 10.18 0.000 .5011207 .7487981 linst4 | .1965442 .1079176 1.82 0.076 -.0210924 .4141808 ltrade | -.0456653 .0444224 -1.03 0.310 -.1352516 .0439209 ldemoc | .0569608 .030817 1.85 0.071 -.0051876 .1191093 lgeog | -.013095 .0176204 -0.74 0.461 -.0486298 .0224399 lelf | .0163212 .0793417 0.21 0.838 -.1436867 .176329 period4 | .0676485 .0272825 2.48 0.017 .0126281 .1226689 period5 | -.0061463 .0236949 -0.26 0.797 -.0539317 .0416391 period6 | -.0073494 .0325277 -0.23 0.822 -.0729478 .058249 period7 | -.0218168 .0218663 -1.00 0.324 -.0659145 .0222808 _cons | .1176348 .1615024 0.73 0.470 -.2080659 .4433354 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.78 Pr > z = 0.075 Arellano-Bond test for AR(2) in first differences: z = -1.48 Pr > z = 0.138 Hansen test of overid. restrictions: chi2(62) = 35.10 Prob > chi2 = 0.998 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(43) = 32.15 Prob > chi2 = 0.887 Difference (null H = exogenous): chi2(19) = 2.95 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(54) = 34.82 Prob > chi2 = 0.980 Difference (null H = exogenous): chi2(8) = 0.28 Prob > chi2 = 1.000 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AF6 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex2 lgeog lelf ltrust1 (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfr > ac avg5_JID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 32) P-value linst4 | 0.0335 | 0.5218 | 8.07 0.0001 ltrade | 0.1293 | 0.5590 | 10.36 0.0000 ldemoc | 0.0098 | 0.1148 | 1.98 0.1218 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 0.33 0.8494 Cragg-Donald N*CDEV statistic 0.33 0.8488 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.07 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,32)= 1.80 P-val=0.1538 Anderson-Rubin test Chi-sq(4)=8.98 P-val=0.0615 Stock-Wright S statistic Chi-sq(4)=5.79 P-val=0.2157 NB: Test statistics heteroskedasticity-robust Number of observations N = 40 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 40 F( 6, 33) = 0.74 Prob > F = 0.6202 Total (centered) SS = 3.068445723 Centered R2 = -14.6177 Total (uncentered) SS = 70.34901708 Uncentered R2 = 0.3188 Residual SS = 47.92215825 Root MSE = 1.095 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 6.878006 5.924294 1.16 0.246 -4.733396 18.48941 ltrade | -.7672607 .5516064 -1.39 0.164 -1.848389 .3138679 ldemoc | -2.914309 4.194888 -0.69 0.487 -11.13614 5.307521 lgeog | .2871012 .5556102 0.52 0.605 -.8018748 1.376077 lelf | .4792373 1.70091 0.28 0.778 -2.854485 3.81296 ltrust1 | -3.022325 2.690552 -1.12 0.261 -8.29571 2.251059 _cons | -5.413323 3.839934 -1.41 0.159 -12.93946 2.11281 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 0.326 Chi-sq(2) P-val = 0.8494 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.066 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.028 Chi-sq(1) P-val = 0.8662 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf ltrust1 Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store AF7 . restore . xtabond2 lindex2 l.lindex2 linst4 ltrade ldemoc lgeog lelf ltrust1 period2-period8, gmm(l > (2/4).lindex2 ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 199 Time variable : countperiod Number of groups = 44 Number of instruments = 73 Obs per group: min = 1 F(11, 43) = 82.92 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex2 | L1. | .6227275 .0635885 9.79 0.000 .4944891 .750966 linst4 | .1793784 .1001857 1.79 0.080 -.0226653 .381422 ltrade | -.0475896 .0446823 -1.07 0.293 -.1377001 .0425209 ldemoc | .05847 .0310015 1.89 0.066 -.0040505 .1209905 lgeog | -.0159439 .0193821 -0.82 0.415 -.0550316 .0231439 lelf | .0082306 .0805718 0.10 0.919 -.1542579 .1707191 ltrust1 | .0543544 .0983446 0.55 0.583 -.1439765 .2526852 period4 | .0659462 .0274893 2.40 0.021 .0105087 .1213837 period5 | -.0068134 .0235913 -0.29 0.774 -.0543898 .0407631 period6 | -.006153 .0317881 -0.19 0.847 -.0702597 .0579537 period7 | -.0218256 .0217987 -1.00 0.322 -.0657868 .0221357 _cons | .1397792 .1602557 0.87 0.388 -.1834072 .4629656 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.56 Pr > z = 0.120 Arellano-Bond test for AR(2) in first differences: z = -1.50 Pr > z = 0.133 Hansen test of overid. restrictions: chi2(61) = 36.84 Prob > chi2 = 0.994 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(42) = 30.57 Prob > chi2 = 0.905 Difference (null H = exogenous): chi2(19) = 6.27 Prob > chi2 = 0.997 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(53) = 35.81 Prob > chi2 = 0.966 Difference (null H = exogenous): chi2(8) = 1.03 Prob > chi2 = 0.998 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AF8 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex2 lgeog lelf (linst4 ltrade ldemoc lsch = avg5_JID_eurfrac avg5_JID_engfrac > avg5_JID_logfrankromr pldemoc lenroll1900), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 5, 25) P-value linst4 | 0.1273 | 0.3238 | 2.88 0.0344 ltrade | 0.4514 | 0.6613 | 14.43 0.0000 ldemoc | 0.0817 | 0.0946 | 0.62 0.6828 lsch | 0.2525 | 0.7574 | 12.02 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 1.61 0.4476 Cragg-Donald N*CDEV statistic 1.65 0.4387 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.25 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(5,25)= 1.69 P-val=0.1745 Anderson-Rubin test Chi-sq(5)=11.13 P-val=0.0488 Stock-Wright S statistic Chi-sq(5)=5.52 P-val=0.3557 NB: Test statistics heteroskedasticity-robust Number of observations N = 33 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 5 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 33 F( 6, 26) = 0.84 Prob > F = 0.5499 Total (centered) SS = 2.88854664 Centered R2 = -0.3991 Total (uncentered) SS = 59.41295408 Uncentered R2 = 0.9320 Residual SS = 4.041277064 Root MSE = .3499 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .8551562 1.584744 0.54 0.589 -2.250885 3.961197 ltrade | -.295771 .2246683 -1.32 0.188 -.7361127 .1445707 ldemoc | 2.290541 1.898559 1.21 0.228 -1.430565 6.011648 lsch | .4498163 .5509001 0.82 0.414 -.629928 1.529561 lgeog | -.3180839 .4220006 -0.75 0.451 -1.14519 .5090221 lelf | .380304 .4397184 0.86 0.387 -.4815282 1.242136 _cons | -4.035329 3.277425 -1.23 0.218 -10.45897 2.388306 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 1.608 Chi-sq(2) P-val = 0.4476 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.250 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.943 Chi-sq(1) P-val = 0.3314 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc lsch Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc lenroll1900 ------------------------------------------------------------------------------ . estimates store AF9 . restore . xtabond2 lindex2 l.lindex2 linst4 ltrade ldemoc lsch lgeog lelf period2-period8, gmm(l(2/ > 4).lindex2 ltrade linst4 ldemoc lsch) iv(period* lgeog lelf) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 352 Time variable : countperiod Number of groups = 77 Number of instruments = 100 Obs per group: min = 1 F(11, 76) = 109.48 avg = 4.57 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex2 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex2 | L1. | .6331789 .0473665 13.37 0.000 .5388403 .7275174 linst4 | .1603087 .0764557 2.10 0.039 .0080339 .3125835 ltrade | -.0096755 .0203461 -0.48 0.636 -.0501983 .0308473 ldemoc | .0223083 .0116844 1.91 0.060 -.0009631 .0455797 lsch | .0135412 .0303812 0.45 0.657 -.0469682 .0740505 lgeog | .006312 .0068888 0.92 0.362 -.0074082 .0200322 lelf | .0409822 .0545157 0.75 0.455 -.0675953 .1495596 period4 | .057064 .0176496 3.23 0.002 .0219118 .0922163 period5 | .0089466 .0128239 0.70 0.488 -.0165945 .0344877 period6 | -.0071758 .0206776 -0.35 0.730 -.0483588 .0340073 period7 | -.0188351 .0128523 -1.47 0.147 -.0444327 .0067625 _cons | .1285131 .1166361 1.10 0.274 -.1037879 .3608141 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.77 Pr > z = 0.077 Arellano-Bond test for AR(2) in first differences: z = -1.50 Pr > z = 0.133 Hansen test of overid. restrictions: chi2(88) = 69.30 Prob > chi2 = 0.930 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(64) = 64.25 Prob > chi2 = 0.468 Difference (null H = exogenous): chi2(24) = 5.05 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf) Hansen test excluding group: chi2(82) = 72.19 Prob > chi2 = 0.772 Difference (null H = exogenous): chi2(6) = -2.89 Prob > chi2 = 1.000 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store AF10 . . estout * using "C:\Llama\Output\benchmark2a.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p idstat idp cdf sarg > an sarganp ar2 ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Anderson LR" "LR p-va > lue" "Cragg-Donald F" "Hansen J" "J p-value" "AR(2)" "AR(2) p" "Instruments" "N") fmt(%8. > 3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels(* 0.10 > ** 0.05 *** 0.01) & AF1 & AF2 & AF3 & AF4 & AF5 > & AF6 & AF7 & AF8 & AF9 & AF10 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se \\ linst4 & 1.302 & 2.413 & 0.116 & 1.352 & 3.176 > & 0.197 & 6.878 & 0.179 & 0.855 & 0.160 \\ & (0.63)** & (1.18)** & (0.05)** & (0.61)** & (1.95) > & (0.11)* & (5.92) & (0.10)* & (1.58) & (0.08)** \\ ltrade & -0.303 & -0.390 & 0.006 & -0.318 & -0.480 > & -0.046 & -0.767 & -0.048 & -0.296 & -0.010 \\ & (0.17)* & (0.21)* & (0.02) & (0.20) & (0.30) > & (0.04) & (0.55) & (0.04) & (0.22) & (0.02) \\ lgeog & -0.047 & -0.156 & 0.001 & -0.060 & -0.176 > & -0.013 & 0.287 & -0.016 & -0.318 & 0.006 \\ & (0.04) & (0.11) & (0.00) & (0.05) & (0.15) > & (0.02) & (0.56) & (0.02) & (0.42) & (0.01) \\ L.lindex2 & & & 0.808 & & > & 0.625 & & 0.623 & & 0.633 \\ & & & (0.04)***& & > & (0.06)***& & (0.06)***& & (0.05)***\\ period4 & & & 0.049 & & > & 0.068 & & 0.066 & & 0.057 \\ & & & (0.02)***& & > & (0.03)** & & (0.03)** & & (0.02)***\\ period5 & & & -0.000 & & > & -0.006 & & -0.007 & & 0.009 \\ & & & (0.01) & & > & (0.02) & & (0.02) & & (0.01) \\ period6 & & & -0.006 & & > & -0.007 & & -0.006 & & -0.007 \\ & & & (0.02) & & > & (0.03) & & (0.03) & & (0.02) \\ period7 & & & -0.020 & & > & -0.022 & & -0.022 & & -0.019 \\ & & & (0.01) & & > & (0.02) & & (0.02) & & (0.01) \\ ldemoc & & & & -0.078 & -0.341 > & 0.057 & -2.914 & 0.058 & 2.291 & 0.022 \\ & & & & (0.17) & (0.38) > & (0.03)* & (4.19) & (0.03)* & (1.90) & (0.01)* \\ lelf & & & & 0.247 & 0.454 > & 0.016 & 0.479 & 0.008 & 0.380 & 0.041 \\ & & & & (0.19) & (0.35) > & (0.08) & (1.70) & (0.08) & (0.44) & (0.05) \\ ltrust1 & & & & & > & & -3.022 & 0.054 & & \\ & & & & & > & & (2.69) & (0.10) & & \\ lsch & & & & & > & & & & 0.450 & 0.014 \\ & & & & & > & & & & (0.55) & (0.03) \\ Constant & -0.730 & -2.055 & 0.074 & -0.746 & -2.796 > & 0.118 & -5.413 & 0.140 & -4.035 & 0.129 \\ & (0.90) & (1.52) & (0.05) & (0.86) & (2.21) > & (0.16) & (3.84) & (0.16) & (3.28) & (0.12) \\ Adjusted R^2& 0.084 & -0.818 & & 0.088 & -1.585 > & & -17.457 & & -0.722 & \\ F & 1.299 & 1.764 & 185.470 & 0.938 & 0.944 > & 88.935 & 0.741 & 82.925 & 0.841 & 109.482 \\ F p-value & & & 0.000 & & > & 0.000 & & 0.000 & & 0.000 \\ Anderson LR & 14.084 & 7.875 & & 5.277 & 3.259 > & & 0.326 & & 1.608 & \\ LR p-value & 0.000 & 0.019 & & 0.022 & 0.196 > & & 0.849 & & 0.448 & \\ Cragg-Donald F& 7.569 & 2.583 & & 1.589 & 0.750 > & & 0.066 & & 0.250 & \\ Hansen J & & & 64.121 & & > & 35.099 & & 36.839 & & 69.302 \\ J p-value & & & 0.241 & & > & 0.998 & & 0.994 & & 0.930 \\ AR(2) & & & -1.387 & & > & -1.482 & & -1.501 & & -1.501 \\ AR(2) p & & & 0.166 & & > & 0.138 & & 0.133 & & 0.133 \\ Instruments & & & 66 & & > & 73 & & 73 & & 100 \\ N & 42 & 74 & 359 & 38 & 69 > & 199 & 40 & 199 & 33 & 352 \\ . estimates clear . . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex3 lgeog (linst4 ltrade = avg5_JID_logem4 avg5_JID_logfrankromr), robust ffir > st Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 2, 38) P-value linst4 | 0.3430 | 0.3887 | 6.99 0.0026 ltrade | 0.2953 | 0.3347 | 9.80 0.0004 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 14.08 0.0002 Cragg-Donald N*CDEV statistic 16.73 0.0000 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 7.57 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(2,38)= 1.73 P-val=0.1905 Anderson-Rubin test Chi-sq(2)=3.83 P-val=0.1473 Stock-Wright S statistic Chi-sq(2)=3.61 P-val=0.1644 NB: Test statistics heteroskedasticity-robust Number of observations N = 42 Number of regressors K = 4 Number of instruments L = 4 Number of excluded instruments L1 = 2 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 42 F( 3, 38) = 1.55 Prob > F = 0.2167 Total (centered) SS = 2.054141402 Centered R2 = 0.1728 Total (uncentered) SS = 101.9477719 Uncentered R2 = 0.9833 Residual SS = 1.699144719 Root MSE = .2011 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.2316 .560602 2.20 0.028 .1328405 2.33036 ltrade | -.295423 .1499933 -1.97 0.049 -.5894045 -.0014414 lgeog | -.0423267 .0407515 -1.04 0.299 -.1221982 .0375448 _cons | -.2847979 .7931562 -0.36 0.720 -1.839356 1.26976 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 14.084 Chi-sq(1) P-val = 0.0002 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 7.569 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store BF1 . ivreg2 lindex3 lgeog (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrank > romr), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 69) P-value linst4 | 0.1047 | 0.1883 | 3.31 0.0250 ltrade | 0.2131 | 0.3832 | 13.07 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 7.87 0.0195 Cragg-Donald N*CDEV statistic 8.31 0.0157 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 2.58 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,69)= 1.09 P-val=0.3609 Anderson-Rubin test Chi-sq(3)=3.49 P-val=0.3215 Stock-Wright S statistic Chi-sq(3)=2.67 P-val=0.4460 NB: Test statistics heteroskedasticity-robust Number of observations N = 74 Number of regressors K = 4 Number of instruments L = 5 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 74 F( 3, 70) = 1.74 Prob > F = 0.1677 Total (centered) SS = 2.600800933 Centered R2 = -0.7184 Total (uncentered) SS = 182.8135603 Uncentered R2 = 0.9756 Residual SS = 4.469145645 Root MSE = .2458 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.976803 .9667628 2.04 0.041 .081983 3.871623 ltrade | -.3541811 .1794467 -1.97 0.048 -.7058902 -.0024721 lgeog | -.1316246 .0925966 -1.42 0.155 -.3131105 .0498613 _cons | -1.137339 1.248346 -0.91 0.362 -3.584053 1.309374 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 7.875 Chi-sq(2) P-val = 0.0195 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 2.583 Stock-Yogo weak ID test critical values: 10% maximal IV size 13.43 15% maximal IV size 8.18 20% maximal IV size 6.40 25% maximal IV size 5.45 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 1.527 Chi-sq(1) P-val = 0.2166 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store BF2 . restore . xtabond2 lindex3 l.lindex3 linst4 ltrade lgeog period2-period8, gmm(l(1/3).lindex3 linst4 > ltrade) iv(period* lgeog) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 352 Time variable : countperiod Number of groups = 79 Number of instruments = 79 Obs per group: min = 1 F(8, 78) = 261.20 avg = 4.46 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex3 | L1. | .8264674 .0394632 20.94 0.000 .7479022 .9050326 linst4 | .0577021 .0274559 2.10 0.039 .0030416 .1123626 ltrade | -.0034255 .0236881 -0.14 0.885 -.050585 .043734 lgeog | -.0009838 .0034363 -0.29 0.775 -.007825 .0058573 period4 | .029847 .0168165 1.77 0.080 -.0036321 .0633261 period5 | -.0082625 .0124975 -0.66 0.510 -.0331431 .0166181 period6 | .0063805 .0123892 0.52 0.608 -.0182844 .0310454 period7 | -.0045905 .0103648 -0.44 0.659 -.0252253 .0160443 _cons | .1829079 .064989 2.81 0.006 .0535247 .3122912 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.95 Pr > z = 0.051 Arellano-Bond test for AR(2) in first differences: z = -0.93 Pr > z = 0.355 Hansen test of overid. restrictions: chi2(70) = 74.70 Prob > chi2 = 0.328 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(47) = 51.34 Prob > chi2 = 0.308 Difference (null H = exogenous): chi2(23) = 23.37 Prob > chi2 = 0.440 ivstyle(period* lgeog) Hansen test excluding group: chi2(64) = 70.03 Prob > chi2 = 0.282 Difference (null H = exogenous): chi2(6) = 4.67 Prob > chi2 = 0.587 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BF3 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex3 lgeog lelf (linst4 ltrade ldemoc = avg5_JID_logem4 avg5_JID_logfrankromr p > ldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 32) P-value linst4 | 0.3669 | 0.4266 | 3.98 0.0161 ltrade | 0.1975 | 0.3438 | 6.38 0.0016 ldemoc | 0.1698 | 0.2626 | 5.19 0.0049 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 5.28 0.0216 Cragg-Donald N*CDEV statistic 5.66 0.0173 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 1.59 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,32)= 1.42 P-val=0.2562 Anderson-Rubin test Chi-sq(3)=5.04 P-val=0.1686 Stock-Wright S statistic Chi-sq(3)=3.87 P-val=0.2760 NB: Test statistics heteroskedasticity-robust Number of observations N = 38 Number of regressors K = 6 Number of instruments L = 6 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 38 F( 5, 32) = 1.42 Prob > F = 0.2449 Total (centered) SS = 2.037458656 Centered R2 = 0.2825 Total (uncentered) SS = 93.15274204 Uncentered R2 = 0.9843 Residual SS = 1.461874478 Root MSE = .1961 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.277005 .5323491 2.40 0.016 .2336203 2.320391 ltrade | -.288258 .170491 -1.69 0.091 -.6224143 .0458982 ldemoc | -.063138 .1507447 -0.42 0.675 -.3585923 .2323162 lgeog | -.0434336 .0424438 -1.02 0.306 -.126622 .0397548 lelf | .3156656 .1658155 1.90 0.057 -.0093269 .640658 _cons | -.3506331 .7487633 -0.47 0.640 -1.818182 1.116916 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 5.277 Chi-sq(1) P-val = 0.0216 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 1.589 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store BF4 . ivreg2 lindex3 lgeog lelf (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfrac avg5_ > JID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 62) P-value linst4 | 0.0572 | 0.2731 | 5.03 0.0014 ltrade | 0.1353 | 0.4293 | 10.02 0.0000 ldemoc | 0.1005 | 0.3228 | 6.66 0.0002 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 3.26 0.1960 Cragg-Donald N*CDEV statistic 3.34 0.1885 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.75 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,62)= 1.58 P-val=0.1896 Anderson-Rubin test Chi-sq(4)=7.05 P-val=0.1331 Stock-Wright S statistic Chi-sq(4)=4.99 P-val=0.2888 NB: Test statistics heteroskedasticity-robust Number of observations N = 69 Number of regressors K = 6 Number of instruments L = 7 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 69 F( 5, 63) = 0.97 Prob > F = 0.4428 Total (centered) SS = 2.571801272 Centered R2 = -1.5681 Total (uncentered) SS = 171.717765 Uncentered R2 = 0.9615 Residual SS = 6.604542041 Root MSE = .3094 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 2.82251 1.62886 1.73 0.083 -.3699974 6.015017 ltrade | -.4527181 .2537759 -1.78 0.074 -.9501098 .0446736 ldemoc | -.3302687 .3300077 -1.00 0.317 -.977072 .3165346 lgeog | -.1522266 .1243134 -1.22 0.221 -.3958763 .0914231 lelf | .498778 .3078287 1.62 0.105 -.1045551 1.102111 _cons | -2.030708 1.836789 -1.11 0.269 -5.630749 1.569333 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 3.259 Chi-sq(2) P-val = 0.1960 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.750 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.218 Chi-sq(1) P-val = 0.6404 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store BF5 . restore . xtabond2 lindex3 l.lindex3 linst4 ltrade ldemoc lgeog lelf period2-period8, gmm(l(5/6).li > ndex3 ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 196 Time variable : countperiod Number of groups = 44 Number of instruments = 69 Obs per group: min = 0 F(10, 43) = 7.97 avg = 4.45 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex3 | L1. | .5615942 .091597 6.13 0.000 .3768711 .7463172 linst4 | .1413725 .0816635 1.73 0.091 -.0233176 .3060625 ltrade | -.030157 .0399685 -0.75 0.455 -.1107612 .0504473 ldemoc | .0174411 .0259057 0.67 0.504 -.0348028 .069685 lgeog | .0024087 .0140315 0.17 0.865 -.0258884 .0307058 lelf | .0899523 .077884 1.15 0.254 -.0671158 .2470204 period4 | .065509 .0255515 2.56 0.014 .0139796 .1170384 period5 | -.0045601 .0184268 -0.25 0.806 -.0417213 .0326011 period6 | .0150841 .0195089 0.77 0.444 -.0242594 .0544275 period7 | .0048484 .0162696 0.30 0.767 -.0279623 .0376591 _cons | .4055889 .2249597 1.80 0.078 -.0480857 .8592635 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.71 Pr > z = 0.087 Arellano-Bond test for AR(2) in first differences: z = -1.63 Pr > z = 0.102 Hansen test of overid. restrictions: chi2(58) = 24.40 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(42) = 29.60 Prob > chi2 = 0.925 Difference (null H = exogenous): chi2(16) = -5.20 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(50) = 33.80 Prob > chi2 = 0.962 Difference (null H = exogenous): chi2(8) = -9.39 Prob > chi2 = 1.000 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BF6 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex3 lgeog lelf ltrust1 (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfr > ac avg5_JID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 32) P-value linst4 | 0.0335 | 0.5218 | 8.07 0.0001 ltrade | 0.1293 | 0.5590 | 10.36 0.0000 ldemoc | 0.0098 | 0.1148 | 1.98 0.1218 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 0.33 0.8494 Cragg-Donald N*CDEV statistic 0.33 0.8488 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.07 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,32)= 1.10 P-val=0.3723 Anderson-Rubin test Chi-sq(4)=5.51 P-val=0.2386 Stock-Wright S statistic Chi-sq(4)=4.05 P-val=0.3997 NB: Test statistics heteroskedasticity-robust Number of observations N = 40 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 40 F( 6, 33) = 0.78 Prob > F = 0.5933 Total (centered) SS = 2.234061375 Centered R2 = -15.1258 Total (uncentered) SS = 107.2020176 Uncentered R2 = 0.6639 Residual SS = 36.02593358 Root MSE = .949 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 5.839215 4.996444 1.17 0.243 -3.953635 15.63207 ltrade | -.6779713 .4711224 -1.44 0.150 -1.601354 .2454115 ldemoc | -2.567904 3.595678 -0.71 0.475 -9.615303 4.479495 lgeog | .2687917 .47949 0.56 0.575 -.6709914 1.208575 lelf | .5315279 1.49196 0.36 0.722 -2.392661 3.455716 ltrust1 | -2.558939 2.269066 -1.13 0.259 -7.006227 1.888349 _cons | -4.071906 3.193911 -1.27 0.202 -10.33186 2.188044 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 0.326 Chi-sq(2) P-val = 0.8494 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.066 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.020 Chi-sq(1) P-val = 0.8866 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf ltrust1 Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store BF7 . restore . xtabond2 lindex3 l.lindex3 linst4 ltrade ldemoc lgeog lelf ltrust1 period2-period8, gmm(l > (1/3).lindex3 ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 196 Time variable : countperiod Number of groups = 44 Number of instruments = 94 Obs per group: min = 1 F(11, 43) = 197.36 avg = 4.45 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex3 | L1. | .8346227 .0304726 27.39 0.000 .7731689 .8960765 linst4 | .0619153 .0438195 1.41 0.165 -.0264552 .1502857 ltrade | -.0160456 .0293444 -0.55 0.587 -.0752242 .043133 ldemoc | .0137517 .0193565 0.71 0.481 -.0252843 .0527878 lgeog | -.0067822 .0069982 -0.97 0.338 -.0208955 .0073311 lelf | .0317851 .033136 0.96 0.343 -.03504 .0986102 ltrust1 | .0061125 .0313629 0.19 0.846 -.0571367 .0693618 period4 | .0615316 .0278956 2.21 0.033 .0052748 .1177884 period5 | -.0156566 .0184194 -0.85 0.400 -.0528029 .0214898 period6 | .0215954 .02172 0.99 0.326 -.0222071 .0653978 period7 | .0057077 .0176179 0.32 0.748 -.0298221 .0412375 _cons | .1360581 .0920179 1.48 0.147 -.0495137 .3216298 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.87 Pr > z = 0.062 Arellano-Bond test for AR(2) in first differences: z = -0.84 Pr > z = 0.401 Hansen test of overid. restrictions: chi2(82) = 35.99 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(55) = 36.96 Prob > chi2 = 0.971 Difference (null H = exogenous): chi2(27) = -0.97 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(74) = 30.52 Prob > chi2 = 1.000 Difference (null H = exogenous): chi2(8) = 5.47 Prob > chi2 = 0.706 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BF8 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 lindex3 lgeog lelf (linst4 ltrade ldemoc lsch = avg5_JID_eurfrac avg5_JID_engfrac > avg5_JID_logfrankromr pldemoc lenroll1900), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 5, 25) P-value linst4 | 0.1273 | 0.3238 | 2.88 0.0344 ltrade | 0.4514 | 0.6613 | 14.43 0.0000 ldemoc | 0.0817 | 0.0946 | 0.62 0.6828 lsch | 0.2525 | 0.7574 | 12.02 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 1.61 0.4476 Cragg-Donald N*CDEV statistic 1.65 0.4387 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.25 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(5,25)= 1.13 P-val=0.3684 Anderson-Rubin test Chi-sq(5)=7.49 P-val=0.1870 Stock-Wright S statistic Chi-sq(5)=4.20 P-val=0.5209 NB: Test statistics heteroskedasticity-robust Number of observations N = 33 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 5 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 33 F( 6, 26) = 0.97 Prob > F = 0.4644 Total (centered) SS = 2.196868921 Centered R2 = -0.3125 Total (uncentered) SS = 89.36879179 Uncentered R2 = 0.9677 Residual SS = 2.883413948 Root MSE = .2956 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .8430739 1.329119 0.63 0.526 -1.761951 3.448098 ltrade | -.3120716 .1918618 -1.63 0.104 -.6881138 .0639707 ldemoc | 1.992965 1.564102 1.27 0.203 -1.072619 5.058548 lsch | .3540331 .4543699 0.78 0.436 -.5365155 1.244582 lgeog | -.2960485 .3467403 -0.85 0.393 -.975647 .38355 lelf | .4832716 .3702503 1.31 0.192 -.2424056 1.208949 _cons | -3.093236 2.6563 -1.16 0.244 -8.299488 2.113017 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 1.608 Chi-sq(2) P-val = 0.4476 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.250 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.484 Chi-sq(1) P-val = 0.4867 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc lsch Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc lenroll1900 ------------------------------------------------------------------------------ . estimates store BF9 . restore . xtabond2 lindex3 l.lindex3 linst4 ltrade ldemoc lsch lgeog lelf period2-period8, gmm(l(2/ > 3).lindex3 ltrade linst4 ldemoc lsch) iv(period* lgeog lelf) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 345 Time variable : countperiod Number of groups = 77 Number of instruments = 112 Obs per group: min = 1 F(11, 76) = 97.33 avg = 4.48 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust lindex3 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lindex3 | L1. | .7916781 .0448462 17.65 0.000 .7023591 .8809971 linst4 | .075342 .0444114 1.70 0.094 -.0131109 .163795 ltrade | -.0099716 .015364 -0.65 0.518 -.0405716 .0206284 ldemoc | .0098354 .0114676 0.86 0.394 -.0130043 .0326751 lsch | .0004343 .0182203 0.02 0.981 -.0358546 .0367232 lgeog | .0011599 .0038652 0.30 0.765 -.0065383 .008858 lelf | .0432929 .0251741 1.72 0.090 -.0068457 .0934316 period4 | .0363456 .017554 2.07 0.042 .0013838 .0713074 period5 | -.0045428 .0106212 -0.43 0.670 -.0256968 .0166112 period6 | .0045518 .0138586 0.33 0.743 -.0230501 .0321536 period7 | -.0049464 .01032 -0.48 0.633 -.0255004 .0156076 _cons | .1738842 .0960477 1.81 0.074 -.0174114 .3651798 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -1.87 Pr > z = 0.061 Arellano-Bond test for AR(2) in first differences: z = -1.00 Pr > z = 0.316 Hansen test of overid. restrictions: chi2(100) = 65.54 Prob > chi2 = 0.997 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(73) = 63.44 Prob > chi2 = 0.780 Difference (null H = exogenous): chi2(27) = 2.10 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf) Hansen test excluding group: chi2(94) = 63.18 Prob > chi2 = 0.994 Difference (null H = exogenous): chi2(6) = 2.36 Prob > chi2 = 0.884 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store BF10 . . estout * using "C:\Llama\Output\benchmark2b.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p idstat idp cdf sarg > an sarganp ar2 ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Anderson LR" "LR p-va > lue" "Cragg-Donald F" "Hansen J" "J p-value" "AR(2)" "AR(2) p" "Instruments" "N") fmt(%8. > 3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels(* 0.10 > ** 0.05 *** 0.01) & BF1 & BF2 & BF3 & BF4 & BF5 > & BF6 & BF7 & BF8 & BF9 & BF10 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se \\ linst4 & 1.232 & 1.977 & 0.058 & 1.277 & 2.823 > & 0.141 & 5.839 & 0.062 & 0.843 & 0.075 \\ & (0.56)** & (0.97)** & (0.03)** & (0.53)** & (1.63)* > & (0.08)* & (5.00) & (0.04) & (1.33) & (0.04)* \\ ltrade & -0.295 & -0.354 & -0.003 & -0.288 & -0.453 > & -0.030 & -0.678 & -0.016 & -0.312 & -0.010 \\ & (0.15)** & (0.18)** & (0.02) & (0.17)* & (0.25)* > & (0.04) & (0.47) & (0.03) & (0.19) & (0.02) \\ lgeog & -0.042 & -0.132 & -0.001 & -0.043 & -0.152 > & 0.002 & 0.269 & -0.007 & -0.296 & 0.001 \\ & (0.04) & (0.09) & (0.00) & (0.04) & (0.12) > & (0.01) & (0.48) & (0.01) & (0.35) & (0.00) \\ L.lindex3 & & & 0.826 & & > & 0.562 & & 0.835 & & 0.792 \\ & & & (0.04)***& & > & (0.09)***& & (0.03)***& & (0.04)***\\ period4 & & & 0.030 & & > & 0.066 & & 0.062 & & 0.036 \\ & & & (0.02)* & & > & (0.03)** & & (0.03)** & & (0.02)** \\ period5 & & & -0.008 & & > & -0.005 & & -0.016 & & -0.005 \\ & & & (0.01) & & > & (0.02) & & (0.02) & & (0.01) \\ period6 & & & 0.006 & & > & 0.015 & & 0.022 & & 0.005 \\ & & & (0.01) & & > & (0.02) & & (0.02) & & (0.01) \\ period7 & & & -0.005 & & > & 0.005 & & 0.006 & & -0.005 \\ & & & (0.01) & & > & (0.02) & & (0.02) & & (0.01) \\ ldemoc & & & & -0.063 & -0.330 > & 0.017 & -2.568 & 0.014 & 1.993 & 0.010 \\ & & & & (0.15) & (0.33) > & (0.03) & (3.60) & (0.02) & (1.56) & (0.01) \\ lelf & & & & 0.316 & 0.499 > & 0.090 & 0.532 & 0.032 & 0.483 & 0.043 \\ & & & & (0.17)* & (0.31) > & (0.08) & (1.49) & (0.03) & (0.37) & (0.03)* \\ ltrust1 & & & & & > & & -2.559 & 0.006 & & \\ & & & & & > & & (2.27) & (0.03) & & \\ lsch & & & & & > & & & & 0.354 & 0.000 \\ & & & & & > & & & & (0.45) & (0.02) \\ Constant & -0.285 & -1.137 & 0.183 & -0.351 & -2.031 > & 0.406 & -4.072 & 0.136 & -3.093 & 0.174 \\ & (0.79) & (1.25) & (0.06)***& (0.75) & (1.84) > & (0.22)* & (3.19) & (0.09) & (2.66) & (0.10)* \\ Adjusted R^2& 0.108 & -0.792 & & 0.170 & -1.772 > & & -18.058 & & -0.615 & \\ F & 1.553 & 1.735 & 261.203 & 1.416 & 0.970 > & 7.967 & 0.778 & 197.358 & 0.971 & 97.326 \\ F p-value & & & 0.000 & & > & 0.000 & & 0.000 & & 0.000 \\ Anderson LR & 14.084 & 7.875 & & 5.277 & 3.259 > & & 0.326 & & 1.608 & \\ LR p-value & 0.000 & 0.019 & & 0.022 & 0.196 > & & 0.849 & & 0.448 & \\ Cragg-Donald F& 7.569 & 2.583 & & 1.589 & 0.750 > & & 0.066 & & 0.250 & \\ Hansen J & & & 74.703 & & > & 24.404 & & 35.995 & & 65.539 \\ J p-value & & & 0.328 & & > & 1.000 & & 1.000 & & 0.997 \\ AR(2) & & & -0.926 & & > & -1.635 & & -0.840 & & -1.003 \\ AR(2) p & & & 0.355 & & > & 0.102 & & 0.401 & & 0.316 \\ Instruments & & & 79 & & > & 69 & & 94 & & 112 \\ N & 42 & 74 & 352 & 38 & 69 > & 196 & 40 & 196 & 33 & 345 \\ . estimates clear . . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 bindex lgeog (linst4 ltrade = avg5_JID_logem4 avg5_JID_logfrankromr), robust ffirs > t Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 2, 38) P-value linst4 | 0.3430 | 0.3887 | 6.99 0.0026 ltrade | 0.2953 | 0.3347 | 9.80 0.0004 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 14.08 0.0002 Cragg-Donald N*CDEV statistic 16.73 0.0000 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 7.57 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(2,38)= 1.26 P-val=0.2944 Anderson-Rubin test Chi-sq(2)=2.79 P-val=0.2476 Stock-Wright S statistic Chi-sq(2)=2.38 P-val=0.3043 NB: Test statistics heteroskedasticity-robust Number of observations N = 42 Number of regressors K = 4 Number of instruments L = 4 Number of excluded instruments L1 = 2 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 42 F( 3, 38) = 1.08 Prob > F = 0.3692 Total (centered) SS = 1.904761905 Centered R2 = 0.1425 Total (uncentered) SS = 2 Uncentered R2 = 0.1834 Residual SS = 1.633249815 Root MSE = .1972 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .933632 .5452101 1.71 0.087 -.1349603 2.002224 ltrade | -.2336895 .1295942 -1.80 0.071 -.4876894 .0203103 lgeog | -.0152644 .0286732 -0.53 0.594 -.0714628 .040934 _cons | -1.388717 .7883609 -1.76 0.078 -2.933876 .156442 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 14.084 Chi-sq(1) P-val = 0.0002 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 7.569 Stock-Yogo weak ID test critical values: 10% maximal IV size 7.03 15% maximal IV size 4.58 20% maximal IV size 3.95 25% maximal IV size 3.63 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store CF1 . ivreg2 bindex lgeog (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankr > omr), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 69) P-value linst4 | 0.1047 | 0.1883 | 3.31 0.0250 ltrade | 0.2131 | 0.3832 | 13.07 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 7.87 0.0195 Cragg-Donald N*CDEV statistic 8.31 0.0157 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 2.58 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,69)= 1.38 P-val=0.2569 Anderson-Rubin test Chi-sq(3)=4.43 P-val=0.2184 Stock-Wright S statistic Chi-sq(3)=3.59 P-val=0.3091 NB: Test statistics heteroskedasticity-robust Number of observations N = 74 Number of regressors K = 4 Number of instruments L = 5 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 74 F( 3, 70) = 2.07 Prob > F = 0.1123 Total (centered) SS = 7.905405405 Centered R2 = -0.3079 Total (uncentered) SS = 9 Uncentered R2 = -0.1488 Residual SS = 10.33951688 Root MSE = .3738 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 2.618042 1.450967 1.80 0.071 -.225802 5.461886 ltrade | -.4081589 .2309556 -1.77 0.077 -.8608236 .0445058 lgeog | -.139893 .1359792 -1.03 0.304 -.4064074 .1266214 _cons | -3.513541 1.842049 -1.91 0.056 -7.123891 .0968083 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 7.875 Chi-sq(2) P-val = 0.0195 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 2.583 Stock-Yogo weak ID test critical values: 10% maximal IV size 13.43 15% maximal IV size 8.18 20% maximal IV size 6.40 25% maximal IV size 5.45 Test statistic(s) not robust Source: Stock-Yogo (2005). Reproduced by permission. ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.174 Chi-sq(1) P-val = 0.6763 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Included instruments: lgeog Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ . estimates store CF2 . restore . xtabond2 bindex l.bindex linst4 ltrade lgeog period2-period8, gmm(l(1/3).bindex linst4 lt > rade) iv(period* lgeog) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 357 Time variable : countperiod Number of groups = 79 Number of instruments = 72 Obs per group: min = 1 F(8, 78) = 10.26 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | L1. | .5613476 .1278973 4.39 0.000 .3067235 .8159716 linst4 | .2459888 .117462 2.09 0.039 .0121399 .4798377 ltrade | .0977338 .0613312 1.59 0.115 -.0243672 .2198348 lgeog | .0174785 .0162041 1.08 0.284 -.0147814 .0497385 period4 | -.0176946 .0272391 -0.65 0.518 -.0719234 .0365342 period5 | .0778075 .0365207 2.13 0.036 .0051004 .1505146 period6 | -.0520766 .035352 -1.47 0.145 -.122457 .0183038 period7 | -.029589 .0259642 -1.14 0.258 -.0812797 .0221017 _cons | -.2485776 .1497699 -1.66 0.101 -.5467466 .0495914 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.82 Pr > z = 0.005 Arellano-Bond test for AR(2) in first differences: z = 0.76 Pr > z = 0.444 Hansen test of overid. restrictions: chi2(63) = 67.32 Prob > chi2 = 0.332 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(40) = 27.86 Prob > chi2 = 0.926 Difference (null H = exogenous): chi2(23) = 39.46 Prob > chi2 = 0.018 ivstyle(period* lgeog) Hansen test excluding group: chi2(57) = 59.16 Prob > chi2 = 0.397 Difference (null H = exogenous): chi2(6) = 8.16 Prob > chi2 = 0.227 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CF3 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 bindex lgeog lelf (linst4 ltrade ldemoc = avg5_JID_logem4 avg5_JID_logfrankromr pl > democ), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 3, 32) P-value linst4 | 0.3669 | 0.4266 | 3.98 0.0161 ltrade | 0.1975 | 0.3438 | 6.38 0.0016 ldemoc | 0.1698 | 0.2626 | 5.19 0.0049 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(1) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 5.28 0.0216 Cragg-Donald N*CDEV statistic 5.66 0.0173 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 1.59 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(3,32)= 0.89 P-val=0.4544 Anderson-Rubin test Chi-sq(3)=3.19 P-val=0.3635 Stock-Wright S statistic Chi-sq(3)=2.57 P-val=0.4625 NB: Test statistics heteroskedasticity-robust Number of observations N = 38 Number of regressors K = 6 Number of instruments L = 6 Number of excluded instruments L1 = 3 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 38 F( 5, 32) = 0.80 Prob > F = 0.5594 Total (centered) SS = 1.894736842 Centered R2 = 0.3460 Total (uncentered) SS = 2 Uncentered R2 = 0.3804 Residual SS = 1.239214283 Root MSE = .1806 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | .8969404 .4940322 1.82 0.069 -.071345 1.865226 ltrade | -.1410054 .1231424 -1.15 0.252 -.3823601 .1003494 ldemoc | .0867472 .0995952 0.87 0.384 -.1084558 .2819501 lgeog | .0105291 .0372283 0.28 0.777 -.0624371 .0834954 lelf | .4321631 .2508804 1.72 0.085 -.0595535 .9238796 _cons | -1.581653 .7856527 -2.01 0.044 -3.121504 -.0418024 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 5.277 Chi-sq(1) P-val = 0.0216 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 1.589 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.000 (equation exactly identified) ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_logem4 avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store CF4 . ivreg2 bindex lgeog lelf (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfrac avg5_J > ID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 62) P-value linst4 | 0.0572 | 0.2731 | 5.03 0.0014 ltrade | 0.1353 | 0.4293 | 10.02 0.0000 ldemoc | 0.1005 | 0.3228 | 6.66 0.0002 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 3.26 0.1960 Cragg-Donald N*CDEV statistic 3.34 0.1885 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.75 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,62)= 1.66 P-val=0.1709 Anderson-Rubin test Chi-sq(4)=7.39 P-val=0.1169 Stock-Wright S statistic Chi-sq(4)=5.44 P-val=0.2450 NB: Test statistics heteroskedasticity-robust Number of observations N = 69 Number of regressors K = 6 Number of instruments L = 7 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 69 F( 5, 63) = 1.26 Prob > F = 0.2932 Total (centered) SS = 7.826086957 Centered R2 = -0.5047 Total (uncentered) SS = 9 Uncentered R2 = -0.3084 Residual SS = 11.77554801 Root MSE = .4131 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 3.128271 2.405647 1.30 0.193 -1.586711 7.843253 ltrade | -.4653441 .3438514 -1.35 0.176 -1.139281 .2085924 ldemoc | -.2680607 .470272 -0.57 0.569 -1.189777 .6536555 lgeog | -.1514389 .1763553 -0.86 0.390 -.4970889 .1942111 lelf | .352983 .3962321 0.89 0.373 -.4236176 1.129584 _cons | -3.968357 2.650354 -1.50 0.134 -9.162955 1.226241 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 3.259 Chi-sq(2) P-val = 0.1960 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.750 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.250 Chi-sq(1) P-val = 0.6169 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store CF5 . restore . xtabond2 bindex l.bindex linst4 ltrade ldemoc lgeog lelf period2-period8, gmm(l(5/6).bind > ex ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 199 Time variable : countperiod Number of groups = 44 Number of instruments = 69 Obs per group: min = 0 F(10, 43) = 10.04 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | L1. | .5637015 .1967427 2.87 0.006 .166932 .9604709 linst4 | .2119291 .2168978 0.98 0.334 -.2254869 .6493452 ltrade | .1802362 .0924851 1.95 0.058 -.0062777 .3667502 ldemoc | .008689 .0580115 0.15 0.882 -.1083022 .1256803 lgeog | .0201866 .0234069 0.86 0.393 -.0270179 .067391 lelf | .0423231 .1908939 0.22 0.826 -.3426511 .4272974 period4 | -.06856 .0495271 -1.38 0.173 -.1684409 .031321 period5 | .0792416 .0628169 1.26 0.214 -.0474408 .2059241 period6 | -.0427255 .0588559 -0.73 0.472 -.1614198 .0759688 period7 | -.0407041 .0493584 -0.82 0.414 -.1402448 .0588366 _cons | -.1198839 .3322866 -0.36 0.720 -.7900037 .5502359 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.39 Pr > z = 0.017 Arellano-Bond test for AR(2) in first differences: z = 0.67 Pr > z = 0.503 Hansen test of overid. restrictions: chi2(58) = 23.65 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(42) = 20.36 Prob > chi2 = 0.998 Difference (null H = exogenous): chi2(16) = 3.29 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(50) = 25.39 Prob > chi2 = 0.999 Difference (null H = exogenous): chi2(8) = -1.74 Prob > chi2 = 1.000 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CF6 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 bindex lgeog lelf ltrust1 (linst4 ltrade ldemoc = avg5_JID_eurfrac avg5_JID_engfra > c avg5_JID_logfrankromr pldemoc), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 4, 32) P-value linst4 | 0.0335 | 0.5218 | 8.07 0.0001 ltrade | 0.1293 | 0.5590 | 10.36 0.0000 ldemoc | 0.0098 | 0.1148 | 1.98 0.1218 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 0.33 0.8494 Cragg-Donald N*CDEV statistic 0.33 0.8488 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.07 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(4,32)= 2.10 P-val=0.1040 Anderson-Rubin test Chi-sq(4)=10.49 P-val=0.0329 Stock-Wright S statistic Chi-sq(4)=6.63 P-val=0.1566 NB: Test statistics heteroskedasticity-robust Number of observations N = 40 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 4 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 40 F( 6, 33) = 0.69 Prob > F = 0.6562 Total (centered) SS = 6.975 Centered R2 = -4.0810 Total (uncentered) SS = 9 Uncentered R2 = -2.9378 Residual SS = 35.4397883 Root MSE = .9413 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 6.258342 6.364843 0.98 0.325 -6.21652 18.7332 ltrade | -.7182094 .5970655 -1.20 0.229 -1.888436 .4520175 ldemoc | -2.175019 4.153417 -0.52 0.601 -10.31557 5.965529 lgeog | .2117926 .5228845 0.41 0.685 -.8130421 1.236627 lelf | .4411828 1.38892 0.32 0.751 -2.281051 3.163417 ltrust1 | -2.547815 2.972016 -0.86 0.391 -8.37286 3.277229 _cons | -6.506418 4.547301 -1.43 0.152 -15.41896 2.406127 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 0.326 Chi-sq(2) P-val = 0.8494 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.066 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 0.080 Chi-sq(1) P-val = 0.7772 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc Included instruments: lgeog lelf ltrust1 Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc ------------------------------------------------------------------------------ . estimates store CF7 . restore . xtabond2 bindex l.bindex linst4 ltrade ldemoc lgeog lelf ltrust1 period2-period8, gmm(l(1 > /3).bindex ltrade linst4 ldemoc) iv(period* lgeog lelf ltrust1) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 199 Time variable : countperiod Number of groups = 44 Number of instruments = 88 Obs per group: min = 1 F(11, 43) = 11.29 avg = 4.52 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | L1. | .5652317 .1346384 4.20 0.000 .2937075 .8367558 linst4 | .1552133 .1903455 0.82 0.419 -.228655 .5390815 ltrade | .1817365 .0841162 2.16 0.036 .0121 .3513729 ldemoc | .0036618 .0517761 0.07 0.944 -.1007548 .1080783 lgeog | .0192643 .0293057 0.66 0.514 -.0398363 .078365 lelf | .0253799 .1962423 0.13 0.898 -.3703804 .4211402 ltrust1 | .0902056 .2742703 0.33 0.744 -.4629131 .6433244 period4 | -.0718022 .0493312 -1.46 0.153 -.1712881 .0276837 period5 | .0797846 .0556603 1.43 0.159 -.0324651 .1920343 period6 | -.0344249 .0586061 -0.59 0.560 -.1526154 .0837656 period7 | -.0382107 .0486244 -0.79 0.436 -.1362712 .0598498 _cons | -.0448122 .3057787 -0.15 0.884 -.6614737 .5718493 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.24 Pr > z = 0.025 Arellano-Bond test for AR(2) in first differences: z = 0.50 Pr > z = 0.616 Hansen test of overid. restrictions: chi2(76) = 23.88 Prob > chi2 = 1.000 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(49) = 22.52 Prob > chi2 = 1.000 Difference (null H = exogenous): chi2(27) = 1.35 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf ltrust1) Hansen test excluding group: chi2(69) = 21.93 Prob > chi2 = 1.000 Difference (null H = exogenous): chi2(7) = 1.95 Prob > chi2 = 0.963 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CF8 . preserve . keep if year==2001 (1667 observations deleted) . ivreg2 bindex lgeog lelf (linst4 ltrade ldemoc lsch = avg5_JID_eurfrac avg5_JID_engfrac a > vg5_JID_logfrankromr pldemoc lenroll1900), robust ffirst Summary results for first-stage regressions ------------------------------------------- Variable | Shea Partial R2 | Partial R2 | F( 5, 25) P-value linst4 | 0.1273 | 0.3238 | 2.88 0.0344 ltrade | 0.4514 | 0.6613 | 14.43 0.0000 ldemoc | 0.0817 | 0.0946 | 0.62 0.6828 lsch | 0.2525 | 0.7574 | 12.02 0.0000 NB: first-stage F-stat heteroskedasticity-robust Underidentification tests Ho: matrix of reduced form coefficients has rank=K-1 (underidentified) Ha: matrix has rank=K (identified) Chi-sq(2) P-value Anderson canon. corr. -N*ln(1-CCEV) LR stat. 1.61 0.4476 Cragg-Donald N*CDEV statistic 1.65 0.4387 Weak identification tests Ho: equation is weakly identified Cragg-Donald (N-L)*CDEV/L1 F-statistic 0.25 See main output for Cragg-Donald weak id test critical values NB: Anderson and Cragg-Donald under- and weak identification stats not robust Weak-instrument-robust inference Tests of joint significance of endogenous regressors B1 in main equation Ho: B1=0 and overidentifying restrictions are valid Anderson-Rubin test F(5,25)= 1.76 P-val=0.1589 Anderson-Rubin test Chi-sq(5)=11.58 P-val=0.0410 Stock-Wright S statistic Chi-sq(5)=6.21 P-val=0.2860 NB: Test statistics heteroskedasticity-robust Number of observations N = 33 Number of regressors K = 7 Number of instruments L = 8 Number of excluded instruments L1 = 5 IV (2SLS) estimation -------------------- Statistics robust to heteroskedasticity Number of obs = 33 F( 6, 26) = 1.37 Prob > F = 0.2633 Total (centered) SS = 5.515151515 Centered R2 = -0.0711 Total (uncentered) SS = 7 Uncentered R2 = 0.1561 Residual SS = 5.907268418 Root MSE = .4231 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 1.549507 2.321768 0.67 0.505 -3.001074 6.100088 ltrade | -.1653648 .228219 -0.72 0.469 -.6126658 .2819363 ldemoc | 1.789295 2.889652 0.62 0.536 -3.874319 7.452909 lsch | .3501237 .792625 0.44 0.659 -1.203393 1.90364 lgeog | -.2955479 .6126154 -0.48 0.629 -1.496252 .9051562 lelf | .5457284 .5699742 0.96 0.338 -.5714004 1.662857 _cons | -5.155019 4.50393 -1.14 0.252 -13.98256 3.672523 ------------------------------------------------------------------------------ Anderson canon. corr. LR statistic (underidentification test): 1.608 Chi-sq(2) P-val = 0.4476 Test statistic(s) not robust ------------------------------------------------------------------------------ Cragg-Donald F statistic (weak identification test): 0.250 Stock-Yogo weak ID test critical values: ------------------------------------------------------------------------------ Hansen J statistic (overidentification test of all instruments): 2.274 Chi-sq(1) P-val = 0.1316 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade ldemoc lsch Included instruments: lgeog lelf Excluded instruments: avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr pldemoc lenroll1900 ------------------------------------------------------------------------------ . estimates store CF9 . restore . xtabond2 bindex l.bindex linst4 ltrade ldemoc lsch lgeog lelf period2-period8, gmm(l(2/3) > .bindex ltrade linst4 ldemoc lsch) iv(period* lgeog lelf) robust small orthog Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm > . period2 dropped due to collinearity period3 dropped due to collinearity period8 dropped due to collinearity Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate robust weighting matrix for Hansen test. Difference-in-Sargan statistics may be negative. Warning: Number of instruments may be large relative to number of observations. Dynamic panel-data estimation, one-step system GMM Orthogonal deviations transform ------------------------------------------------------------------------------ Group variable: code2 Number of obs = 350 Time variable : countperiod Number of groups = 77 Number of instruments = 109 Obs per group: min = 1 F(11, 76) = 16.50 avg = 4.55 Prob > F = 0.000 max = 5 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | L1. | .7052784 .1040836 6.78 0.000 .497978 .9125789 linst4 | .1248353 .1027963 1.21 0.228 -.0799013 .3295718 ltrade | .0748847 .0482529 1.55 0.125 -.0212192 .1709887 ldemoc | -.0256777 .0356482 -0.72 0.474 -.0966772 .0453217 lsch | .0739106 .0502832 1.47 0.146 -.026237 .1740581 lgeog | .0158923 .0181471 0.88 0.384 -.0202507 .0520353 lelf | .0610911 .1021503 0.60 0.552 -.1423588 .264541 period4 | -.0208216 .0320531 -0.65 0.518 -.0846608 .0430176 period5 | .0840213 .0407608 2.06 0.043 .0028391 .1652035 period6 | -.0447245 .0364911 -1.23 0.224 -.1174029 .0279538 period7 | -.0275465 .0309343 -0.89 0.376 -.0891576 .0340646 _cons | -.2415326 .1604112 -1.51 0.136 -.5610192 .0779539 ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -2.65 Pr > z = 0.008 Arellano-Bond test for AR(2) in first differences: z = 0.78 Pr > z = 0.437 Hansen test of overid. restrictions: chi2(97) = 67.74 Prob > chi2 = 0.990 Difference-in-Sargan tests of exogeneity of instrument subsets: GMM instruments for levels Hansen test excluding group: chi2(71) = 64.11 Prob > chi2 = 0.706 Difference (null H = exogenous): chi2(26) = 3.63 Prob > chi2 = 1.000 ivstyle(period* lgeog lelf) Hansen test excluding group: chi2(91) = 61.45 Prob > chi2 = 0.993 Difference (null H = exogenous): chi2(6) = 6.29 Prob > chi2 = 0.391 Warning: Sargan/Hansen tests are weak when instruments are many. ------------------------------------------------------------------------------ . estimates store CF10 . . estout * using "C:\Llama\Output\benchmark2c.tex", replace style(tex) varlabels(_cons Cons > tant) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_a F F_p idstat idp cdf sarg > an sarganp ar2 ar2p zcols N, labels("Adjusted R^2" "F" "F p-value" "Anderson LR" "LR p-va > lue" "Cragg-Donald F" "Hansen J" "J p-value" "AR(2)" "AR(2) p" "Instruments" "N") fmt(%8. > 3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %8.3f %6.0f %6.0f)) starlevels(* 0.10 > ** 0.05 *** 0.01) & CF1 & CF2 & CF3 & CF4 & CF5 > & CF6 & CF7 & CF8 & CF9 & CF10 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se \\ linst4 & 0.934 & 2.618 & 0.246 & 0.897 & 3.128 > & 0.212 & 6.258 & 0.155 & 1.550 & 0.125 \\ & (0.55)* & (1.45)* & (0.12)** & (0.49)* & (2.41) > & (0.22) & (6.36) & (0.19) & (2.32) & (0.10) \\ ltrade & -0.234 & -0.408 & 0.098 & -0.141 & -0.465 > & 0.180 & -0.718 & 0.182 & -0.165 & 0.075 \\ & (0.13)* & (0.23)* & (0.06) & (0.12) & (0.34) > & (0.09)* & (0.60) & (0.08)** & (0.23) & (0.05) \\ lgeog & -0.015 & -0.140 & 0.017 & 0.011 & -0.151 > & 0.020 & 0.212 & 0.019 & -0.296 & 0.016 \\ & (0.03) & (0.14) & (0.02) & (0.04) & (0.18) > & (0.02) & (0.52) & (0.03) & (0.61) & (0.02) \\ L.bindex & & & 0.561 & & > & 0.564 & & 0.565 & & 0.705 \\ & & & (0.13)***& & > & (0.20)***& & (0.13)***& & (0.10)***\\ period4 & & & -0.018 & & > & -0.069 & & -0.072 & & -0.021 \\ & & & (0.03) & & > & (0.05) & & (0.05) & & (0.03) \\ period5 & & & 0.078 & & > & 0.079 & & 0.080 & & 0.084 \\ & & & (0.04)** & & > & (0.06) & & (0.06) & & (0.04)** \\ period6 & & & -0.052 & & > & -0.043 & & -0.034 & & -0.045 \\ & & & (0.04) & & > & (0.06) & & (0.06) & & (0.04) \\ period7 & & & -0.030 & & > & -0.041 & & -0.038 & & -0.028 \\ & & & (0.03) & & > & (0.05) & & (0.05) & & (0.03) \\ ldemoc & & & & 0.087 & -0.268 > & 0.009 & -2.175 & 0.004 & 1.789 & -0.026 \\ & & & & (0.10) & (0.47) > & (0.06) & (4.15) & (0.05) & (2.89) & (0.04) \\ lelf & & & & 0.432 & 0.353 > & 0.042 & 0.441 & 0.025 & 0.546 & 0.061 \\ & & & & (0.25)* & (0.40) > & (0.19) & (1.39) & (0.20) & (0.57) & (0.10) \\ ltrust1 & & & & & > & & -2.548 & 0.090 & & \\ & & & & & > & & (2.97) & (0.27) & & \\ lsch & & & & & > & & & & 0.350 & 0.074 \\ & & & & & > & & & & (0.79) & (0.05) \\ Constant & -1.389 & -3.514 & -0.249 & -1.582 & -3.968 > & -0.120 & -6.506 & -0.045 & -5.155 & -0.242 \\ & (0.79)* & (1.84)* & (0.15) & (0.79)** & (2.65) > & (0.33) & (4.55) & (0.31) & (4.50) & (0.16) \\ Adjusted R^2& 0.075 & -0.364 & & 0.244 & -0.624 > & & -5.005 & & -0.318 & \\ F & 1.080 & 2.068 & 10.257 & 0.798 & 1.258 > & 10.045 & 0.694 & 11.289 & 1.371 & 16.500 \\ F p-value & & & 0.000 & & > & 0.000 & & 0.000 & & 0.000 \\ Anderson LR & 14.084 & 7.875 & & 5.277 & 3.259 > & & 0.326 & & 1.608 & \\ LR p-value & 0.000 & 0.019 & & 0.022 & 0.196 > & & 0.849 & & 0.448 & \\ Cragg-Donald F& 7.569 & 2.583 & & 1.589 & 0.750 > & & 0.066 & & 0.250 & \\ Hansen J & & & 67.318 & & > & 23.649 & & 23.877 & & 67.741 \\ J p-value & & & 0.332 & & > & 1.000 & & 1.000 & & 0.990 \\ AR(2) & & & 0.765 & & > & 0.669 & & 0.502 & & 0.778 \\ AR(2) p & & & 0.444 & & > & 0.503 & & 0.616 & & 0.437 \\ Instruments & & & 72 & & > & 69 & & 88 & & 109 \\ N & 42 & 74 & 357 & 38 & 69 > & 199 & 40 & 199 & 33 & 350 \\ . estimates clear . . /* Growth pole determinants regressions */ . logit bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robust Iteration 0: log pseudolikelihood = -212.86542 Iteration 1: log pseudolikelihood = -158.64283 Iteration 2: log pseudolikelihood = -150.12709 Iteration 3: log pseudolikelihood = -149.59748 Iteration 4: log pseudolikelihood = -134.02159 Iteration 5: log pseudolikelihood = -128.24051 Iteration 6: log pseudolikelihood = -126.75997 Iteration 7: log pseudolikelihood = -126.69365 Iteration 8: log pseudolikelihood = -126.69339 Iteration 9: log pseudolikelihood = -126.69339 Logistic regression Number of obs = 665 Wald chi2(14) = 69.27 Prob > chi2 = 0.0000 Log pseudolikelihood = -126.69339 Pseudo R2 = 0.4048 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .5162235 .4710165 1.10 0.273 -.4069519 1.439399 liy | 3.269239 4.99318 0.65 0.513 -6.517213 13.05569 lenroll_p | -.272948 1.660167 -0.16 0.869 -3.526816 2.98092 lgdppc | 1.99481 .3626019 5.50 0.000 1.284123 2.705497 llifeexp | -6.942286 3.480584 -1.99 0.046 -13.7641 -.1204667 ldepratio1 | -4.371167 1.617248 -2.70 0.007 -7.540915 -1.201419 lgovt_Y3 | -6.679209 4.315599 -1.55 0.122 -15.13763 1.77921 period2 | -2.197102 .9909155 -2.22 0.027 -4.139261 -.2549436 period3 | -1.989166 .9403056 -2.12 0.034 -3.832131 -.1462011 period4 | -2.807107 1.036732 -2.71 0.007 -4.839064 -.7751497 period5 | -1.451672 1.005098 -1.44 0.149 -3.421627 .5182832 period6 | -2.043808 .9688431 -2.11 0.035 -3.942705 -.14491 period7 | -2.231852 .9700288 -2.30 0.021 -4.133074 -.3306308 period8 | -2.134397 .9802251 -2.18 0.029 -4.055603 -.2131912 _cons | 30.52143 19.08528 1.60 0.110 -6.88503 67.92789 ------------------------------------------------------------------------------ . probit bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, robus > t Iteration 0: log pseudolikelihood = -212.86542 Iteration 1: log pseudolikelihood = -144.83148 Iteration 2: log pseudolikelihood = -131.85958 Iteration 3: log pseudolikelihood = -128.21964 Iteration 4: log pseudolikelihood = -127.66747 Iteration 5: log pseudolikelihood = -127.65087 Iteration 6: log pseudolikelihood = -127.65086 Probit regression Number of obs = 665 Wald chi2(14) = 70.09 Prob > chi2 = 0.0000 Log pseudolikelihood = -127.65086 Pseudo R2 = 0.4003 ------------------------------------------------------------------------------ | Robust bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .3151972 .308883 1.02 0.308 -.2902024 .9205967 liy | 2.922312 2.384363 1.23 0.220 -1.750954 7.595578 lenroll_p | -.3292521 .7271737 -0.45 0.651 -1.754486 1.095982 lgdppc | 1.008423 .1763418 5.72 0.000 .6627989 1.354046 llifeexp | -3.665844 1.705737 -2.15 0.032 -7.009028 -.3226607 ldepratio1 | -2.195968 .7770773 -2.83 0.005 -3.719012 -.6729249 lgovt_Y3 | -3.814777 2.126692 -1.79 0.073 -7.983016 .353462 period2 | -1.260594 .5445057 -2.32 0.021 -2.327806 -.1933829 period3 | -1.090917 .5220663 -2.09 0.037 -2.114148 -.0676857 period4 | -1.479602 .566799 -2.61 0.009 -2.590507 -.368696 period5 | -.6905675 .5299144 -1.30 0.193 -1.729181 .3480457 period6 | -1.072496 .5237016 -2.05 0.041 -2.098932 -.0460601 period7 | -1.159705 .5265609 -2.20 0.028 -2.191745 -.1276644 period8 | -1.041443 .5384705 -1.93 0.053 -2.096826 .0139392 _cons | 16.67981 9.39149 1.78 0.076 -1.727176 35.08679 ------------------------------------------------------------------------------ . estimates store B1 . xtprobit bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re Fitting comparison model: Iteration 0: log likelihood = -212.86542 Iteration 1: log likelihood = -144.83148 Iteration 2: log likelihood = -131.85958 Iteration 3: log likelihood = -128.21964 Iteration 4: log likelihood = -127.66747 Iteration 5: log likelihood = -127.65087 Iteration 6: log likelihood = -127.65086 Fitting full model: rho = 0.0 log likelihood = -127.65086 rho = 0.1 log likelihood = -113.37336 rho = 0.2 log likelihood = -104.8591 rho = 0.3 log likelihood = -99.236723 rho = 0.4 log likelihood = -95.487316 rho = 0.5 log likelihood = -93.206855 rho = 0.6 log likelihood = -92.236514 rho = 0.7 log likelihood = -92.624369 Iteration 0: log likelihood = -92.365093 Iteration 1: log likelihood = -75.052522 Iteration 2: log likelihood = -74.226497 Iteration 3: log likelihood = -74.219033 Iteration 4: log likelihood = -74.219026 Random-effects probit regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 5.2 max = 8 Wald chi2(14) = 48.04 Log likelihood = -74.219026 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .0984984 1.139513 0.09 0.931 -2.134907 2.331904 liy | 13.61305 5.086422 2.68 0.007 3.643843 23.58225 lenroll_p | -2.329498 1.973247 -1.18 0.238 -6.196992 1.537996 lgdppc | 2.144846 .4544374 4.72 0.000 1.254165 3.035527 llifeexp | -7.282735 5.496048 -1.33 0.185 -18.05479 3.489321 ldepratio1 | -3.915394 2.005978 -1.95 0.051 -7.84704 .0162513 lgovt_Y3 | -14.57341 8.548855 -1.70 0.088 -31.32886 2.182034 period2 | -3.043351 1.165655 -2.61 0.009 -5.327994 -.7587086 period3 | -2.018227 1.121973 -1.80 0.072 -4.217253 .1807985 period4 | -3.021064 1.228051 -2.46 0.014 -5.428001 -.6141281 period5 | -.9875583 1.130993 -0.87 0.383 -3.204264 1.229147 period6 | -1.987479 1.160375 -1.71 0.087 -4.261773 .2868149 period7 | -1.99725 1.175919 -1.70 0.089 -4.30201 .3075093 period8 | -1.736256 1.174851 -1.48 0.139 -4.038921 .5664093 _cons | 36.66657 24.34913 1.51 0.132 -11.05686 84.38999 -------------+---------------------------------------------------------------- /lnsig2u | 1.673281 .254514 1.174443 2.17212 -------------+---------------------------------------------------------------- sigma_u | 2.308599 .2937853 1.798983 2.962578 rho | .8420128 .0338573 .7639472 .8977178 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 106.86 Prob >= chibar2 = 0.000 . estimates store B2 . xtlogit bindex ln liy lenroll_p lgdppc llifeexp ldepratio1 lgovt_Y3 period2-period8, re Fitting comparison model: Iteration 0: log likelihood = -212.86542 Iteration 1: log likelihood = -158.64283 Iteration 2: log likelihood = -150.12709 Iteration 3: log likelihood = -149.59748 Iteration 4: log likelihood = -134.02159 Iteration 5: log likelihood = -128.24051 Iteration 6: log likelihood = -126.75997 Iteration 7: log likelihood = -126.69365 Iteration 8: log likelihood = -126.69339 Iteration 9: log likelihood = -126.69339 Fitting full model: tau = 0.0 log likelihood = -126.69339 tau = 0.1 log likelihood = -120.35835 tau = 0.2 log likelihood = -114.64287 tau = 0.3 log likelihood = -109.49051 tau = 0.4 log likelihood = -104.82258 tau = 0.5 log likelihood = -100.57804 tau = 0.6 log likelihood = -96.747658 tau = 0.7 log likelihood = -93.42953 tau = 0.8 log likelihood = -90.974853 Iteration 0: log likelihood = -93.432234 Iteration 1: log likelihood = -82.051806 Iteration 2: log likelihood = -76.933939 Iteration 3: log likelihood = -76.596684 Iteration 4: log likelihood = -76.59481 Iteration 5: log likelihood = -76.59481 Random-effects logistic regression Number of obs = 665 Group variable (i): code2 Number of groups = 129 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 5.2 max = 8 Wald chi2(14) = 47.51 Log likelihood = -76.59481 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- ln | .1998222 1.757691 0.11 0.909 -3.245189 3.644833 liy | 19.93482 8.131571 2.45 0.014 3.997236 35.87241 lenroll_p | -3.570069 3.165012 -1.13 0.259 -9.77338 2.633241 lgdppc | 3.529105 .742533 4.75 0.000 2.073767 4.984442 llifeexp | -11.65478 8.754167 -1.33 0.183 -28.81263 5.503072 ldepratio1 | -6.346387 3.186261 -1.99 0.046 -12.59134 -.1014301 lgovt_Y3 | -20.95158 13.13173 -1.60 0.111 -46.6893 4.786135 period2 | -4.777205 1.841971 -2.59 0.009 -8.387401 -1.167008 period3 | -3.260778 1.781406 -1.83 0.067 -6.75227 .2307148 period4 | -4.824984 1.93068 -2.50 0.012 -8.609046 -1.040921 period5 | -1.744461 1.801322 -0.97 0.333 -5.274988 1.786067 period6 | -3.223961 1.839329 -1.75 0.080 -6.82898 .3810578 period7 | -3.30186 1.860404 -1.77 0.076 -6.948185 .3444653 period8 | -2.946344 1.864811 -1.58 0.114 -6.601307 .7086192 _cons | 57.61098 39.15817 1.47 0.141 -19.13763 134.3596 -------------+---------------------------------------------------------------- /lnsig2u | 2.446486 .239589 1.976901 2.916072 -------------+---------------------------------------------------------------- sigma_u | 3.398191 .4070845 2.687067 4.297511 rho | .7782744 .0413443 .6869832 .8488006 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 100.20 Prob >= chibar2 = 0.000 . estimates store B3 . preserve . . keep if year==2001 (1667 observations deleted) . ivprobit bindex lgeog (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfran > kromr), twostep first /* Wald test insig, suggests unnecessary to instrument both */ invalid 'suggests' r(198); . . ivprobit bindex lgeog lelf ldemoc (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5 > _JID_logfrankromr), twostep first Checking reduced-form model... First stage regressions Source | SS df MS Number of obs = 73 -------------+------------------------------ F( 6, 66) = 10.73 Model | .872422295 6 .145403716 Prob > F = 0.0000 Residual | .894059907 66 .013546362 R-squared = 0.4939 -------------+------------------------------ Adj R-squared = 0.4479 Total | 1.7664822 72 .024534475 Root MSE = .11639 ------------------------------------------------------------------------------ linst4 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- avg5_J~rfrac | -.0260495 .0407245 -0.64 0.525 -.1073587 .0552596 avg5_J~gfrac | .1694234 .0580227 2.92 0.005 .0535774 .2852695 avg5_JID_~mr | .0648364 .0245507 2.64 0.010 .0158195 .1138534 lgeog | .0506073 .017972 2.82 0.006 .014725 .0864895 lelf | -.1609407 .0969457 -1.66 0.102 -.354499 .0326176 ldemoc | .0688669 .0383782 1.79 0.077 -.0077577 .1454915 _cons | 1.020286 .1216673 8.39 0.000 .7773698 1.263203 ------------------------------------------------------------------------------ Source | SS df MS Number of obs = 73 -------------+------------------------------ F( 6, 66) = 7.21 Model | 7.98522306 6 1.33087051 Prob > F = 0.0000 Residual | 12.1837878 66 .184602846 R-squared = 0.3959 -------------+------------------------------ Adj R-squared = 0.3410 Total | 20.1690109 72 .280125152 Root MSE = .42965 ------------------------------------------------------------------------------ ltrade | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- avg5_J~rfrac | -.1074961 .1503364 -0.72 0.477 -.4076523 .1926602 avg5_J~gfrac | .3212299 .2141933 1.50 0.138 -.1064209 .7488806 avg5_JID_~mr | .557718 .0906298 6.15 0.000 .3767698 .7386662 lgeog | -.0783678 .0663444 -1.18 0.242 -.2108286 .0540929 lelf | .2285273 .357879 0.64 0.525 -.4860013 .9430559 ldemoc | .1489999 .1416748 1.05 0.297 -.133863 .4318629 _cons | -2.538791 .4491401 -5.65 0.000 -3.435528 -1.642054 ------------------------------------------------------------------------------ Two-step probit with endogenous regressors Number of obs = 73 Wald chi2(5) = 4.81 Prob > chi2 = 0.4397 ------------------------------------------------------------------------------ bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 7.955602 7.085635 1.12 0.262 -5.931986 21.84319 ltrade | -1.977965 1.243494 -1.59 0.112 -4.415168 .4592385 lgeog | 1.041415 1.479187 0.70 0.481 -1.857738 3.940568 lelf | 1.144242 2.541112 0.45 0.653 -3.836246 6.124729 ldemoc | 1.552137 2.535043 0.61 0.540 -3.416457 6.52073 _cons | -21.03334 10.37463 -2.03 0.043 -41.36725 -.6994334 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Instruments: lgeog lelf ldemoc avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ Wald test of exogeneity: chi2(2) = 2.63 Prob > chi2 = 0.2691 . . ivprobit bindex lgeog ltrade lelf ldemoc (linst4 = avg5_JID_eurfrac avg5_JID_engfrac), ro > bust first Fitting exogenous probit model Iteration 0: log likelihood = -27.260001 Iteration 1: log likelihood = -19.88618 Iteration 2: log likelihood = -18.931205 Iteration 3: log likelihood = -18.613227 Iteration 4: log likelihood = -18.487309 Iteration 5: log likelihood = -18.476504 Iteration 6: log likelihood = -18.476413 Iteration 7: log likelihood = -18.476413 Fitting full model Iteration 0: log pseudolikelihood = 39.024705 Iteration 1: log pseudolikelihood = 39.253307 Iteration 2: log pseudolikelihood = 40.502785 Iteration 3: log pseudolikelihood = 40.720425 Iteration 4: log pseudolikelihood = 40.734631 Iteration 5: log pseudolikelihood = 40.734726 Iteration 6: log pseudolikelihood = 40.734726 Probit model with endogenous regressors Number of obs = 73 Wald chi2(5) = 60.39 Log pseudolikelihood = 40.734726 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Robust | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | linst4 | 6.440193 2.894237 2.23 0.026 .7675932 12.11279 lgeog | .645262 .8264351 0.78 0.435 -.974521 2.265045 ltrade | -.8813605 .3620637 -2.43 0.015 -1.590992 -.1717287 lelf | 1.138824 1.361742 0.84 0.403 -1.530141 3.807789 ldemoc | .9349488 1.199669 0.78 0.436 -1.416359 3.286257 _cons | -15.20782 2.073794 -7.33 0.000 -19.27238 -11.14326 -------------+---------------------------------------------------------------- linst4 | lgeog | .0594386 .0230005 2.58 0.010 .0143584 .1045187 ltrade | .0817685 .030633 2.67 0.008 .0217289 .141808 lelf | -.1852661 .0951442 -1.95 0.052 -.3717452 .001213 ldemoc | .0522456 .0285704 1.83 0.067 -.0037513 .1082424 avg5_J~rfrac | -.009692 .0407744 -0.24 0.812 -.0896082 .0702243 avg5_J~gfrac | .1307914 .0716261 1.83 0.068 -.0095933 .271176 _cons | 1.289159 .1189893 10.83 0.000 1.055944 1.522374 -------------+---------------------------------------------------------------- /lnsigma | -2.215085 .0792464 -27.95 0.000 -2.370405 -2.059765 /athrho | -.8804082 .6621498 -1.33 0.184 -2.178198 .4173816 -------------+---------------------------------------------------------------- sigma | .1091442 .0086493 .0934429 .1274839 rho | -.7066238 .3315271 -.9746757 .3947223 ------------------------------------------------------------------------------ Instrumented: linst4 Instruments: lgeog ltrade lelf ldemoc avg5_JID_eurfrac avg5_JID_engfrac ------------------------------------------------------------------------------ Wald test of exogeneity (/athrho = 0): chi2(1) = 1.77 Prob > chi2 = 0.1836 . . estimates store B4 . . ivprobit bindex lgeog linst4 lelf ldemoc (ltrade = avg5_JID_logfrankromr), robust first Fitting exogenous probit model Iteration 0: log likelihood = -29.306449 Iteration 1: log likelihood = -21.24491 Iteration 2: log likelihood = -20.225129 Iteration 3: log likelihood = -19.899307 Iteration 4: log likelihood = -19.770888 Iteration 5: log likelihood = -19.759747 Iteration 6: log likelihood = -19.759637 Iteration 7: log likelihood = -19.759637 Fitting full model Iteration 0: log pseudolikelihood = -58.482726 Iteration 1: log pseudolikelihood = -56.976886 Iteration 2: log pseudolikelihood = -55.903939 Iteration 3: log pseudolikelihood = -55.830006 Iteration 4: log pseudolikelihood = -55.829265 Iteration 5: log pseudolikelihood = -55.829265 Probit model with endogenous regressors Number of obs = 74 Wald chi2(5) = 27.61 Log pseudolikelihood = -55.829265 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Robust | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- bindex | ltrade | -1.200514 .4893094 -2.45 0.014 -2.159543 -.2414854 lgeog | 1.231966 .8836221 1.39 0.163 -.499902 2.963833 linst4 | 2.066064 2.265744 0.91 0.362 -2.374712 6.506839 lelf | .0498602 1.546009 0.03 0.974 -2.980261 3.079981 ldemoc | 1.262424 .9269924 1.36 0.173 -.5544477 3.079296 _cons | -11.40943 2.845029 -4.01 0.000 -16.98558 -5.833273 -------------+---------------------------------------------------------------- ltrade | lgeog | -.1136472 .0827997 -1.37 0.170 -.2759316 .0486373 linst4 | .9084507 .4585016 1.98 0.048 .009804 1.807097 lelf | .4283349 .3144806 1.36 0.173 -.1880358 1.044706 ldemoc | .0814203 .1368147 0.60 0.552 -.1867316 .3495723 avg5_JID_~mr | .4988313 .0791241 6.30 0.000 .3437509 .6539117 _cons | -3.520417 .5723042 -6.15 0.000 -4.642112 -2.398721 -------------+---------------------------------------------------------------- /lnsigma | -.9196918 .0845832 -10.87 0.000 -1.085472 -.7539117 /athrho | .5647157 .3339765 1.69 0.091 -.0898663 1.219298 -------------+---------------------------------------------------------------- sigma | .3986419 .0337184 .3377424 .4705224 rho | .5114679 .2466085 -.0896252 .8394469 ------------------------------------------------------------------------------ Instrumented: ltrade Instruments: lgeog linst4 lelf ldemoc avg5_JID_logfrankromr ------------------------------------------------------------------------------ Wald test of exogeneity (/athrho = 0): chi2(1) = 2.86 Prob > chi2 = 0.0909 . . estimates store B5 . . restore . . ivprobit bindex lgeog lelf ldemoc (linst4 ltrade = avg5_JID_eurfrac avg5_JID_engfrac avg5 > _JID_logfrankromr) period2-period8, twostep first note: period2 dropped due to collinearity note: period3 dropped due to collinearity note: period4 dropped due to collinearity Checking reduced-form model... First stage regressions Source | SS df MS Number of obs = 356 -------------+------------------------------ F( 10, 345) = 41.02 Model | 8.70052362 10 .870052362 Prob > F = 0.0000 Residual | 7.31699323 345 .021208676 R-squared = 0.5432 -------------+------------------------------ Adj R-squared = 0.5299 Total | 16.0175168 355 .045119766 Root MSE = .14563 ------------------------------------------------------------------------------ linst4 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- avg5_J~rfrac | -.0255608 .0226591 -1.13 0.260 -.0701282 .0190067 avg5_J~gfrac | .1457439 .0326772 4.46 0.000 .0814724 .2100155 avg5_JID_~mr | .0682725 .0140676 4.85 0.000 .0406035 .0959416 lgeog | .0390948 .01053 3.71 0.000 .0183838 .0598058 lelf | -.1158403 .0571582 -2.03 0.043 -.2282628 -.0034178 ldemoc | .2299016 .0212945 10.80 0.000 .1880181 .271785 period5 | .0460857 .024646 1.87 0.062 -.0023897 .094561 period6 | .1391108 .024904 5.59 0.000 .0901281 .1880936 period7 | .0879595 .0247114 3.56 0.000 .0393555 .1365635 period8 | .0302138 .0251319 1.20 0.230 -.0192171 .0796448 _cons | .7176779 .0730202 9.83 0.000 .5740571 .8612987 ------------------------------------------------------------------------------ Source | SS df MS Number of obs = 356 -------------+------------------------------ F( 10, 345) = 35.72 Model | 51.9889339 10 5.19889339 Prob > F = 0.0000 Residual | 50.2149398 345 .14555055 R-squared = 0.5087 -------------+------------------------------ Adj R-squared = 0.4944 Total | 102.203874 355 .287898236 Root MSE = .38151 ------------------------------------------------------------------------------ ltrade | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- avg5_J~rfrac | -.138072 .0593599 -2.33 0.021 -.254825 -.0213191 avg5_J~gfrac | .3680848 .0856041 4.30 0.000 .1997131 .5364564 avg5_JID_~mr | .5829038 .0368528 15.82 0.000 .5104194 .6553882 lgeog | -.1417134 .0275852 -5.14 0.000 -.1959698 -.087457 lelf | .061744 .149737 0.41 0.680 -.2327683 .3562562 ldemoc | .2291761 .0557851 4.11 0.000 .1194544 .3388977 period5 | .0598998 .0645649 0.93 0.354 -.0670906 .1868902 period6 | .1306541 .0652408 2.00 0.046 .0023344 .2589738 period7 | .1978392 .0647363 3.06 0.002 .0705117 .3251666 period8 | .3352375 .0658377 5.09 0.000 .2057437 .4647314 _cons | -2.682966 .1912904 -14.03 0.000 -3.059208 -2.306724 ------------------------------------------------------------------------------ Two-step probit with endogenous regressors Number of obs = 356 Wald chi2(9) = 33.95 Prob > chi2 = 0.0001 ------------------------------------------------------------------------------ bindex | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- linst4 | 10.78206 3.663162 2.94 0.003 3.602391 17.96172 ltrade | -1.188774 .5234056 -2.27 0.023 -2.21463 -.1629176 lgeog | -.6220511 .2756701 -2.26 0.024 -1.162355 -.0817476 lelf | -.0601853 .9231132 -0.07 0.948 -1.869454 1.749083 ldemoc | -.6953559 .9494458 -0.73 0.464 -2.556236 1.165524 period5 | .6755453 .4436288 1.52 0.128 -.1939511 1.545042 period6 | -.4149876 .6167957 -0.67 0.501 -1.623885 .7939097 period7 | .2222373 .4929207 0.45 0.652 -.7438696 1.188344 period8 | 1.162462 .4511048 2.58 0.010 .2783125 2.046611 _cons | -15.26102 3.800572 -4.02 0.000 -22.71001 -7.812038 ------------------------------------------------------------------------------ Instrumented: linst4 ltrade Instruments: lgeog lelf ldemoc period5 period6 period7 period8 avg5_JID_eurfrac avg5_JID_engfrac avg5_JID_logfrankromr ------------------------------------------------------------------------------ Wald test of exogeneity: chi2(2) = 6.36 Prob > chi2 = 0.0416 . . estimates store B6 . estout * using "C:\Llama\Output\binary1.tex", replace style(tex) varlabels(_cons Constant > ) cells (b(fmt(%9.3f)) se(star par fmt(%9.2f))) stats (r2_p chi2 p chi2_exog p_exog N, la > bels("Pseudo R^2" "Wald Chi^2" "Chi^2 p-value" "Chi^2 exog" "Exog p-value" "N") fmt(%8.3f > %8.3f %8.3f %8.3f %8.3f %6.0f)) starlevels(* 0.10 ** 0.05 *** 0.01) & B1 & B2 & B3 & B4 & B5 > & B6 \\ & b/se & b/se & b/se & b/se & b/se > & b/se \\ main & & & & & > & \\ ln & 0.315 & 0.098 & 0.200 & & > & \\ & (0.31) & (1.14) & (1.76) & & > & \\ liy & 2.922 & 13.613 & 19.935 & & > & \\ & (2.38) & (5.09)***& (8.13)** & & > & \\ lenroll_p & -0.329 & -2.329 & -3.570 & & > & \\ & (0.73) & (1.97) & (3.17) & & > & \\ lgdppc & 1.008 & 2.145 & 3.529 & & > & \\ & (0.18)***& (0.45)***& (0.74)***& & > & \\ llifeexp & -3.666 & -7.283 & -11.655 & & > & \\ & (1.71)** & (5.50) & (8.75) & & > & \\ ldepratio1 & -2.196 & -3.915 & -6.346 & & > & \\ & (0.78)***& (2.01)* & (3.19)** & & > & \\ lgovt_Y3 & -3.815 & -14.573 & -20.952 & & > & \\ & (2.13)* & (8.55)* & (13.13) & & > & \\ period2 & -1.261 & -3.043 & -4.777 & & > & \\ & (0.54)** & (1.17)***& (1.84)***& & > & \\ period3 & -1.091 & -2.018 & -3.261 & & > & \\ & (0.52)** & (1.12)* & (1.78)* & & > & \\ period4 & -1.480 & -3.021 & -4.825 & & > & \\ & (0.57)***& (1.23)** & (1.93)** & & > & \\ period5 & -0.691 & -0.988 & -1.744 & & > & 0.676 \\ & (0.53) & (1.13) & (1.80) & & > & (0.44) \\ period6 & -1.072 & -1.987 & -3.224 & & > & -0.415 \\ & (0.52)** & (1.16)* & (1.84)* & & > & (0.62) \\ period7 & -1.160 & -1.997 & -3.302 & & > & 0.222 \\ & (0.53)** & (1.18)* & (1.86)* & & > & (0.49) \\ period8 & -1.041 & -1.736 & -2.946 & & > & 1.162 \\ & (0.54)* & (1.17) & (1.86) & & > & (0.45)***\\ linst4 & & & & 6.440 & 2.066 > & 10.782 \\ & & & & (2.89)** & (2.27) > & (3.66)***\\ lgeog & & & & 0.645 & 1.232 > & -0.622 \\ & & & & (0.83) & (0.88) > & (0.28)** \\ ltrade & & & & -0.881 & -1.201 > & -1.189 \\ & & & & (0.36)** & (0.49)** > & (0.52)** \\ lelf & & & & 1.139 & 0.050 > & -0.060 \\ & & & & (1.36) & (1.55) > & (0.92) \\ ldemoc & & & & 0.935 & 1.262 > & -0.695 \\ & & & & (1.20) & (0.93) > & (0.95) \\ Constant & 16.680 & 36.667 & 57.611 & -15.208 & -11.409 > & -15.261 \\ & (9.39)* & (24.35) & (39.16) & (2.07)***& (2.85)** > *& (3.80)***\\ lnsig2u & & & & & > & \\ Constant & & 1.673 & 2.446 & & > & \\ & & (0.25)***& (0.24)***& & > & \\ linst4 & & & & & > & \\ lgeog & & & & 0.059 & > & \\ & & & & (0.02)***& > & \\ ltrade & & & & 0.082 & > & \\ & & & & (0.03)***& > & \\ lelf & & & & -0.185 & > & \\ & & & & (0.10)* & > & \\ ldemoc & & & & 0.052 & > & \\ & & & & (0.03)* & > & \\ avg5_JID_eurfrac& & & & -0.010 & > & \\ & & & & (0.04) & > & \\ avg5_JID_engfrac& & & & 0.131 & > & \\ & & & & (0.07)* & > & \\ Constant & & & & 1.289 & > & \\ & & & & (0.12)***& > & \\ athrho & & & & & > & \\ Constant & & & & -0.880 & 0.565 > & \\ & & & & (0.66) & (0.33)* > & \\ lnsigma & & & & & > & \\ Constant & & & & -2.215 & -0.920 > & \\ & & & & (0.08)***& (0.08)** > *& \\ ltrade & & & & & > & \\ lgeog & & & & & -0.114 > & \\ & & & & & (0.08) > & \\ linst4 & & & & & 0.908 > & \\ & & & & & (0.46)** > & \\ lelf & & & & & 0.428 > & \\ & & & & & (0.31) > & \\ ldemoc & & & & & 0.081 > & \\ & & & & & (0.14) > & \\ avg5_JID_logfrankromr& & & & & > 0.499 & \\ & & & & & (0.08)** > *& \\ Constant & & & & & -3.520 > & \\ & & & & & (0.57)** > *& \\ Pseudo R^2 & 0.400 & & & & > & \\ Wald Chi^2 & 70.087 & 48.045 & 47.513 & 60.387 & 27.613 > & 33.951 \\ Chi^2 p-value& 0.000 & 0.000 & 0.000 & 0.000 & 0.000 > & 0.000 \\ Chi^2 exog & & & & 1.768 & 2.859 > & 6.358 \\ Exog p-value& & & & 0.184 & 0.091 > & 0.042 \\ N & 665 & 665 & 665 & 73 & 74 > & 356 \\ . estimates clear . . /* Calculate distributions */ . do "C:\Llama\poledist.do" . /* Compute polarity shares */ . . use "C:\Llama\pca_HP_complete_noEMU", clear . . summarize index_constusd_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- index_~d_HPg | 4024 -1.18e-09 1.444593 -9.01976 14.37756 . gen index_rescale = 100/(r(max)-r(min))*index_constusd_HPg-(r(min)*100/(r(max)-r(min))) (7265 missing values generated) . summarize index_ppp_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- index_~p_HPg | 3385 -7.92e-11 1.543835 -4.170114 17.96434 . gen index2_rescale = 100/(r(max)-r(min))*index_ppp_HPg-(r(min)*100/(r(max)-r(min))) (7904 missing values generated) . summarize index_constusd_migr_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- i~d_migr_HPg | 3947 -3.72e-10 1.619788 -7.074695 16.66891 . gen index3_rescale = 100/(r(max)-r(min))*index_constusd_migr_HPg-(r(min)*100/(r(max)-r(mi > n))) (7342 missing values generated) . summarize index_ppp_migr_HPg Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- i~p_migr_HPg | 3000 -3.48e-10 1.697146 -3.998665 19.09534 . gen index4_rescale = 100/(r(max)-r(min))*index_ppp_migr_HPg-(r(min)*100/(r(max)-r(min))) (8289 missing values generated) . . keep if year > 1967 (1575 observations deleted) . gen poles = 15 . preserve . local indexlist "index_rescale index2_rescale index3_rescale index4_rescale" . foreach x of local indexlist { 2. keep if year == 1968 3. gsort - `x' 4. keep if _n <= poles 5. egen sharemax_`x' = sum(`x') 6. egen sharemean_`x' = mean(`x') 7. gen polarityshare_`x' = `x'/sharemax_`x' 8. egen herf_`x' = sum((polarityshare_`x')^2) 9. egen theil0_`x' = sum((`x'/sharemean_`x')*ln(`x'/sharemean_`x')) 10. gen normherf_`x' = (herf_`x' - 1/poles)/(1 - 1/poles) 11. gen theil_`x' = theil0_`x'/poles 12. gen normtheil_`x' = theil_`x'/ln(poles) 13. inequal2 polarityshare_`x' 14. gen gini_`x' = $S_4 15. } (9489 observations deleted) (210 observations deleted) (2 missing values generated) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .14022004 coefficient of variation .37005023 standard deviation of logs .31429162 Gini coefficient .17412976 Mehran measure .22375892 Piesch measure .14931518 Kakwani measure .03098363 Theil entropy measure .05562037 Theil mean log deviation measure .05071629 Entropy measure GE -1 .04774875 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (2 missing values generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .15584194 coefficient of variation .43764531 standard deviation of logs .36010872 Gini coefficient .20423912 Mehran measure .26602032 Piesch measure .17334852 Kakwani measure .04092855 Theil entropy measure .07520972 Theil mean log deviation measure .06760777 Entropy measure GE -1 .06384567 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) . collapse (mean) year *herf* *theil* gini* . drop theil0* . save "C:\Llama\Temp\conc", replace file C:\Llama\Temp\conc.dta saved . restore . . local i = 1969 . while `i' <= 2008 { 2. preserve 3. local indexlist "index_rescale index2_rescale index3_rescale index4_rescale" 4. foreach x of local indexlist { 5. keep if year == `i' 6. gsort - `x' 7. keep if _n <= poles 8. egen sharemax_`x' = sum(`x') 9. egen sharemean_`x' = mean(`x') 10. gen polarityshare_`x' = `x'/sharemax_`x' 11. egen herf_`x' = sum((polarityshare_`x')^2) 12. egen theil0_`x' = sum((`x'/sharemean_`x')*ln(`x'/sharemean_`x')) 13. gen normherf_`x' = (herf_`x' - 1/poles)/(1 - 1/poles) 14. gen theil_`x' = theil0_`x'/poles 15. gen normtheil_`x' = (theil_`x')/ln(poles) 16. inequal2 polarityshare_`x' 17. gen gini_`x' = $S_4 18. } 19. collapse (mean) year *herf* *theil* gini* 20. drop theil0* 21. append using "C:\Llama\Temp\conc" 22. sort year 23. save "C:\Llama\Temp\conc", replace 24. restore 25. local i = `i'+1 26. } (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .11724469 coefficient of variation .31513357 standard deviation of logs .26928551 Gini coefficient .14915192 Mehran measure .19464933 Piesch measure .1264032 Kakwani measure .02288496 Theil entropy measure .04095354 Theil mean log deviation measure .03738405 Entropy measure GE -1 .03513076 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .13463599 coefficient of variation .39686437 standard deviation of logs .31996996 Gini coefficient .17981087 Mehran measure .23418876 Piesch measure .15262192 Kakwani measure .03311528 Theil entropy measure .0613754 Theil mean log deviation measure .05431258 Entropy measure GE -1 .05038599 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .10603484 coefficient of variation .29514066 standard deviation of logs .25000035 Gini coefficient .13705248 Mehran measure .17876931 Piesch measure .11619405 Kakwani measure .01987393 Theil entropy measure .03569985 Theil mean log deviation measure .03237469 Entropy measure GE -1 .03020903 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11970097 coefficient of variation .37395645 standard deviation of logs .29384405 Gini coefficient .16253455 Mehran measure .2104836 Piesch measure .13856002 Kakwani measure .02862171 Theil entropy measure .05362385 Theil mean log deviation measure .04660542 Entropy measure GE -1 .0423928 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09854499 coefficient of variation .28524714 standard deviation of logs .23775822 Gini coefficient .12856349 Mehran measure .16682469 Piesch measure .1094329 Kakwani measure .01823124 Theil entropy measure .03298255 Theil mean log deviation measure .02958832 Entropy measure GE -1 .02731705 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11061875 coefficient of variation .36661639 standard deviation of logs .28147009 Gini coefficient .15263889 Mehran measure .19747826 Piesch measure .13021921 Kakwani measure .02679452 Theil entropy measure .05075288 Theil mean log deviation measure .04344267 Entropy measure GE -1 .0389374 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09186265 coefficient of variation .28040661 standard deviation of logs .23086553 Gini coefficient .12255771 Mehran measure .15686591 Piesch measure .10540362 Kakwani measure .01739624 Theil entropy measure .03162959 Theil mean log deviation measure .02814206 Entropy measure GE -1 .0257576 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10511322 coefficient of variation .36353891 standard deviation of logs .27544316 Gini coefficient .14809335 Mehran measure .18968345 Piesch measure .1272983 Kakwani measure .02601224 Theil entropy measure .04953249 Theil mean log deviation measure .04202946 Entropy measure GE -1 .03730134 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08663592 coefficient of variation .26530492 standard deviation of logs .21915575 Gini coefficient .11586374 Mehran measure .14856393 Piesch measure .09951365 Kakwani measure .01566908 Theil entropy measure .02841774 Theil mean log deviation measure .02531512 Entropy measure GE -1 .02315194 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .1147477 coefficient of variation .36853081 standard deviation of logs .28205477 Gini coefficient .1525006 Mehran measure .19205888 Piesch measure .13272146 Kakwani measure .02712371 Theil entropy measure .05132311 Theil mean log deviation measure .04383298 Entropy measure GE -1 .03906982 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08759084 coefficient of variation .26785079 standard deviation of logs .2180192 Gini coefficient .11455052 Mehran measure .14489221 Piesch measure .09937967 Kakwani measure .01572466 Theil entropy measure .02869582 Theil mean log deviation measure .02532336 Entropy measure GE -1 .02294114 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .1161189 coefficient of variation .37266331 standard deviation of logs .28162807 Gini coefficient .15082148 Mehran measure .18757893 Piesch measure .13244275 Kakwani measure .02736429 Theil entropy measure .05206672 Theil mean log deviation measure .04411992 Entropy measure GE -1 .03901567 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08957534 coefficient of variation .26389655 standard deviation of logs .21775459 Gini coefficient .11545098 Mehran measure .14769381 Piesch measure .09932958 Kakwani measure .0154999 Theil entropy measure .0281116 Theil mean log deviation measure .02502127 Entropy measure GE -1 .0228495 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11473643 coefficient of variation .36386796 standard deviation of logs .27860675 Gini coefficient .14980573 Mehran measure .18803898 Piesch measure .1306891 Kakwani measure .02647969 Theil entropy measure .0500744 Theil mean log deviation measure .04276435 Entropy measure GE -1 .03808495 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09197157 coefficient of variation .26998797 standard deviation of logs .22089143 Gini coefficient .11595539 Mehran measure .14681427 Piesch measure .10052594 Kakwani measure .01606409 Theil entropy measure .02925001 Theil mean log deviation measure .02590098 Entropy measure GE -1 .02354584 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11807884 coefficient of variation .37135361 standard deviation of logs .28240496 Gini coefficient .1514089 Mehran measure .189464 Piesch measure .13238135 Kakwani measure .0273425 Theil entropy measure .05188965 Theil mean log deviation measure .04414555 Entropy measure GE -1 .03920463 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08219255 coefficient of variation .26260261 standard deviation of logs .21024596 Gini coefficient .10800763 Mehran measure .13507867 Piesch measure .09447212 Kakwani measure .01483872 Theil entropy measure .02728309 Theil mean log deviation measure .02381896 Entropy measure GE -1 .02135242 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10812155 coefficient of variation .3662103 standard deviation of logs .27203011 Gini coefficient .1436741 Mehran measure .17920176 Piesch measure .12591028 Kakwani measure .02589202 Theil entropy measure .04969716 Theil mean log deviation measure .04164318 Entropy measure GE -1 .03644385 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07978405 coefficient of variation .25224482 standard deviation of logs .20278379 Gini coefficient .10272928 Mehran measure .1273603 Piesch measure .09041377 Kakwani measure .01379418 Theil entropy measure .02528324 Theil mean log deviation measure .02211577 Entropy measure GE -1 .01982136 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10104798 coefficient of variation .34382886 standard deviation of logs .25634206 Gini coefficient .13185171 Mehran measure .16242127 Piesch measure .11656693 Kakwani measure .02305678 Theil entropy measure .04407013 Theil mean log deviation measure .03695274 Entropy measure GE -1 .03221178 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07492368 coefficient of variation .23536964 standard deviation of logs .19143323 Gini coefficient .09532232 Mehran measure .11721679 Piesch measure .08437508 Kakwani measure .01221435 Theil entropy measure .0222324 Theil mean log deviation measure .01957223 Entropy measure GE -1 .01760047 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .18521803 coefficient of variation .61732671 standard deviation of logs .79119686 Gini coefficient .26162811 Mehran measure .34977357 Piesch measure .21755537 Kakwani measure .07916575 Theil entropy measure .15636889 Theil mean log deviation measure .20163137 Entropy measure GE -1 .51485106 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09738788 coefficient of variation .32651367 standard deviation of logs .24737744 Gini coefficient .12697702 Mehran measure .15613056 Piesch measure .11240025 Kakwani measure .02125554 Theil entropy measure .04025343 Theil mean log deviation measure .03405652 Entropy measure GE -1 .02985399 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07151256 coefficient of variation .22990696 standard deviation of logs .18713661 Gini coefficient .09260056 Mehran measure .11403406 Piesch measure .08188381 Kakwani measure .01167762 Theil entropy measure .02123757 Theil mean log deviation measure .01869982 Entropy measure GE -1 .01680504 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .18166509 coefficient of variation .60753837 standard deviation of logs .40433861 Gini coefficient .23137131 Mehran measure .28308752 Piesch measure .20551319 Kakwani measure .06033107 Theil entropy measure .12413739 Theil mean log deviation measure .09902664 Entropy measure GE -1 .08612016 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09578097 coefficient of variation .32066773 standard deviation of logs .24485548 Gini coefficient .12625397 Mehran measure .15570397 Piesch measure .11152896 Kakwani measure .02070368 Theil entropy measure .0390476 Theil mean log deviation measure .03318761 Entropy measure GE -1 .02919643 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .0673062 coefficient of variation .22833142 standard deviation of logs .1847017 Gini coefficient .08984518 Mehran measure .11050746 Piesch measure .07951403 Kakwani measure .01144323 Theil entropy measure .02086647 Theil mean log deviation measure .01829861 Entropy measure GE -1 .01637602 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .18840919 coefficient of variation .61805745 standard deviation of logs .41308701 Gini coefficient .23688321 Mehran measure .28834536 Piesch measure .21115212 Kakwani measure .06281191 Theil entropy measure .12887568 Theil mean log deviation measure .10316799 Entropy measure GE -1 .09002253 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09330834 coefficient of variation .31572076 standard deviation of logs .24067178 Gini coefficient .12291143 Mehran measure .15129647 Piesch measure .10871891 Kakwani measure .02005167 Theil entropy measure .03783304 Theil mean log deviation measure .0321107 Entropy measure GE -1 .02818928 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07260496 coefficient of variation .2481934 standard deviation of logs .19765642 Gini coefficient .09651886 Mehran measure .11846608 Piesch measure .08554526 Kakwani measure .01324857 Theil entropy measure .02436168 Theil mean log deviation measure .02117388 Entropy measure GE -1 .01883059 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .19101183 coefficient of variation .61627334 standard deviation of logs .41766476 Gini coefficient .24046698 Mehran measure .29252426 Piesch measure .21443832 Kakwani measure .0636437 Theil entropy measure .12953692 Theil mean log deviation measure .10457791 Entropy measure GE -1 .09182224 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09824016 coefficient of variation .33463769 standard deviation of logs .25220188 Gini coefficient .12969609 Mehran measure .15865732 Piesch measure .11521548 Kakwani measure .02217494 Theil entropy measure .04211303 Theil mean log deviation measure .03552871 Entropy measure GE -1 .03108518 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (15 missing values generated) (15 missing values generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation . coefficient of variation . standard deviation of logs . Gini coefficient . Mehran measure . Piesch measure . Kakwani measure . Theil entropy measure . Theil mean log deviation measure . Entropy measure GE -1 . ------------------------------------------------------------------------------ (15 missing values generated) file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08335064 coefficient of variation .27751649 standard deviation of logs .21735461 Gini coefficient .10877413 Mehran measure .13316823 Piesch measure .09657708 Kakwani measure .0161762 Theil entropy measure .03003816 Theil mean log deviation measure .02587563 Entropy measure GE -1 .02290481 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .19587044 coefficient of variation .61694615 standard deviation of logs .42504497 Gini coefficient .24519356 Mehran measure .29905402 Piesch measure .21826332 Kakwani measure .06515749 Theil entropy measure .13142927 Theil mean log deviation measure .10720251 Entropy measure GE -1 .09493497 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .12424519 coefficient of variation .4236541 standard deviation of logs .30288106 Gini coefficient .16197984 Mehran measure .19845046 Piesch measure .14374453 Kakwani measure .03287821 Theil entropy measure .06449897 Theil mean log deviation measure .05297225 Entropy measure GE -1 .04583859 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .21969258 coefficient of variation .71533525 standard deviation of logs .47350193 Gini coefficient .27877231 Mehran measure .34047896 Piesch measure .247919 Kakwani measure .08117529 Theil entropy measure .16847655 Theil mean log deviation measure .13531404 Entropy measure GE -1 .12071214 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09327721 coefficient of variation .30461551 standard deviation of logs .23701852 Gini coefficient .1221669 Mehran measure .15024335 Piesch measure .10812867 Kakwani measure .01925329 Theil entropy measure .0359326 Theil mean log deviation measure .03087546 Entropy measure GE -1 .02736902 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .2063111 coefficient of variation .63683981 standard deviation of logs .44918858 Gini coefficient .26190651 Mehran measure .32212841 Piesch measure .23179556 Kakwani measure .07112841 Theil entropy measure .14201372 Theil mean log deviation measure .11779819 Entropy measure GE -1 .10638173 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .13541499 coefficient of variation .44920229 standard deviation of logs .32057766 Gini coefficient .17411308 Mehran measure .21354373 Piesch measure .15439777 Kakwani measure .03671268 Theil entropy measure .07220461 Theil mean log deviation measure .0593515 Entropy measure GE -1 .05161241 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .23092166 coefficient of variation .73023279 standard deviation of logs .49308984 Gini coefficient .29126903 Mehran measure .35696911 Piesch measure .25841898 Kakwani measure .08640368 Theil entropy measure .1776738 Theil mean log deviation measure .1447562 Entropy measure GE -1 .1312401 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09933898 coefficient of variation .31158116 standard deviation of logs .24582404 Gini coefficient .12936006 Mehran measure .16031652 Piesch measure .11388183 Kakwani measure .02046108 Theil entropy measure .03794095 Theil mean log deviation measure .03289976 Entropy measure GE -1 .02942807 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .20940891 coefficient of variation .61615443 standard deviation of logs .45376851 Gini coefficient .26452622 Mehran measure .32863919 Piesch measure .23246973 Kakwani measure .07055338 Theil entropy measure .13766095 Theil mean log deviation measure .11707914 Entropy measure GE -1 .1076685 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .14031149 coefficient of variation .4526782 standard deviation of logs .3275763 Gini coefficient .17982644 Mehran measure .22127241 Piesch measure .15910345 Kakwani measure .0378829 Theil entropy measure .07399417 Theil mean log deviation measure .06137165 Entropy measure GE -1 .05383546 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .23096715 coefficient of variation .7068532 standard deviation of logs .49440688 Gini coefficient .29145138 Mehran measure .35910369 Piesch measure .25762525 Kakwani measure .08502146 Theil entropy measure .17140567 Theil mean log deviation measure .14244644 Entropy measure GE -1 .13077252 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .10125463 coefficient of variation .30617348 standard deviation of logs .24669226 Gini coefficient .13123467 Mehran measure .16411056 Piesch measure .11479671 Kakwani measure .02028166 Theil entropy measure .03720875 Theil mean log deviation measure .03268283 Entropy measure GE -1 .02954753 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .20833344 coefficient of variation .59109568 standard deviation of logs .44984734 Gini coefficient .2608281 Mehran measure .32696488 Piesch measure .22775971 Kakwani measure .06793457 Theil entropy measure .13023876 Theil mean log deviation measure .11279362 Entropy measure GE -1 .10493774 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .14055476 coefficient of variation .44323631 standard deviation of logs .32677299 Gini coefficient .18009013 Mehran measure .22253749 Piesch measure .15886643 Kakwani measure .03721096 Theil entropy measure .07194729 Theil mean log deviation measure .0603295 Entropy measure GE -1 .053356 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .22628489 coefficient of variation .67969053 standard deviation of logs .48734643 Gini coefficient .28624306 Mehran measure .35416666 Piesch measure .25228126 Kakwani measure .08161253 Theil entropy measure .16213557 Theil mean log deviation measure .1364567 Entropy measure GE -1 .12591832 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09953151 coefficient of variation .29043427 standard deviation of logs .23961388 Gini coefficient .12788397 Mehran measure .16103494 Piesch measure .11130849 Kakwani measure .01881889 Theil entropy measure .03409895 Theil mean log deviation measure .0303728 Entropy measure GE -1 .0277448 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .20406482 coefficient of variation .56272301 standard deviation of logs .4396795 Gini coefficient .2532711 Mehran measure .31985446 Piesch measure .21997943 Kakwani measure .06389033 Theil entropy measure .12070978 Theil mean log deviation measure .106012 Entropy measure GE -1 .09935544 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .13551825 coefficient of variation .42298379 standard deviation of logs .31703243 Gini coefficient .17391623 Mehran measure .21510608 Piesch measure .15332131 Kakwani measure .03470934 Theil entropy measure .06645137 Theil mean log deviation measure .05621091 Entropy measure GE -1 .049927 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) (1 missing value generated) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .21733066 coefficient of variation .65034389 standard deviation of logs .47229391 Gini coefficient .27587053 Mehran measure .34151713 Piesch measure .24304723 Kakwani measure .07644952 Theil entropy measure .15052807 Theil mean log deviation measure .1273021 Entropy measure GE -1 .11718048 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .09235384 coefficient of variation .26355134 standard deviation of logs .22368525 Gini coefficient .11950996 Mehran measure .15168939 Piesch measure .10342025 Kakwani measure .01608328 Theil entropy measure .02870993 Theil mean log deviation measure .02599681 Entropy measure GE -1 .02402207 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .18848807 coefficient of variation .51368711 standard deviation of logs .4139592 Gini coefficient .23730872 Mehran measure .30256074 Piesch measure .20468271 Kakwani measure .05564323 Theil entropy measure .10330311 Theil mean log deviation measure .09211613 Entropy measure GE -1 .08684247 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11358051 coefficient of variation .33263461 standard deviation of logs .27095784 Gini coefficient .14773678 Mehran measure .1862327 Piesch measure .12848882 Kakwani measure .02413939 Theil entropy measure .04414033 Theil mean log deviation measure .03910983 Entropy measure GE -1 .03577372 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .18930241 coefficient of variation .53294134 standard deviation of logs .42260331 Gini coefficient .24331908 Mehran measure .30857495 Piesch measure .21069113 Kakwani measure .05834603 Theil entropy measure .10940955 Theil mean log deviation measure .09678773 Entropy measure GE -1 .09104284 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .08347802 coefficient of variation .23644444 standard deviation of logs .20569117 Gini coefficient .10916948 Mehran measure .13953501 Piesch measure .09398672 Kakwani measure .01337309 Theil entropy measure .0235641 Theil mean log deviation measure .0216441 Entropy measure GE -1 .02019463 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .16817221 coefficient of variation .46450974 standard deviation of logs .383309 Gini coefficient .2174456 Mehran measure .27957016 Piesch measure .18638332 Kakwani measure .04718397 Theil entropy measure .08633521 Theil mean log deviation measure .07781758 Entropy measure GE -1 .07347679 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10758214 coefficient of variation .30407504 standard deviation of logs .25422053 Gini coefficient .13798146 Mehran measure .17463029 Piesch measure .11965706 Kakwani measure .02091438 Theil entropy measure .03768946 Theil mean log deviation measure .03388431 Entropy measure GE -1 .03126322 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .17405131 coefficient of variation .4874038 standard deviation of logs .39550428 Gini coefficient .22544877 Mehran measure .28709822 Piesch measure .19462404 Kakwani measure .05066478 Theil entropy measure .0935972 Theil mean log deviation measure .08365811 Entropy measure GE -1 .07871297 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .0781109 coefficient of variation .21328905 standard deviation of logs .18833019 Gini coefficient .09956799 Mehran measure .12780039 Piesch measure .08545179 Kakwani measure .01109025 Theil entropy measure .01939458 Theil mean log deviation measure .01796709 Entropy measure GE -1 .01686284 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15436461 coefficient of variation .42061467 standard deviation of logs .35310222 Gini coefficient .19936299 Mehran measure .25777809 Piesch measure .17015543 Kakwani measure .03972411 Theil entropy measure .07189649 Theil mean log deviation measure .065295 Entropy measure GE -1 .06166331 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10232143 coefficient of variation .28260079 standard deviation of logs .23900888 Gini coefficient .1288698 Mehran measure .16401634 Piesch measure .11129653 Kakwani measure .01834956 Theil entropy measure .03285853 Theil mean log deviation measure .02972387 Entropy measure GE -1 .02751732 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16577207 coefficient of variation .45267707 standard deviation of logs .37108458 Gini coefficient .21091282 Mehran measure .26987885 Piesch measure .18142979 Kakwani measure .04441408 Theil entropy measure .08149594 Theil mean log deviation measure .07311703 Entropy measure GE -1 .06869195 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07332012 coefficient of variation .19852625 standard deviation of logs .17528946 Gini coefficient .09185489 Mehran measure .11821399 Piesch measure .07867533 Kakwani measure .00960399 Theil entropy measure .01679691 Theil mean log deviation measure .01555375 Entropy measure GE -1 .01458785 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14607972 coefficient of variation .39308193 standard deviation of logs .33043947 Gini coefficient .18618672 Mehran measure .2420716 Piesch measure .1582443 Kakwani measure .03470226 Theil entropy measure .06277391 Theil mean log deviation measure .05697987 Entropy measure GE -1 .05371946 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09750104 coefficient of variation .27125941 standard deviation of logs .22789718 Gini coefficient .12170157 Mehran measure .15490207 Piesch measure .10510131 Kakwani measure .01676934 Theil entropy measure .03012401 Theil mean log deviation measure .0271232 Entropy measure GE -1 .0249965 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15845787 coefficient of variation .43501187 standard deviation of logs .35319868 Gini coefficient .20033461 Mehran measure .25719851 Piesch measure .17190267 Kakwani measure .04044695 Theil entropy measure .07460472 Theil mean log deviation measure .06647127 Entropy measure GE -1 .06206411 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07138963 coefficient of variation .19966108 standard deviation of logs .17299348 Gini coefficient .08895893 Mehran measure .11398007 Piesch measure .07644837 Kakwani measure .00948974 Theil entropy measure .01674953 Theil mean log deviation measure .01532677 Entropy measure GE -1 .01423397 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14386314 coefficient of variation .3959038 standard deviation of logs .32338572 Gini coefficient .1806098 Mehran measure .23356295 Piesch measure .15413322 Kakwani measure .03386346 Theil entropy measure .06218155 Theil mean log deviation measure .05546146 Entropy measure GE -1 .05159632 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09560679 coefficient of variation .27607258 standard deviation of logs .22684422 Gini coefficient .11973139 Mehran measure .15161311 Piesch measure .10379053 Kakwani measure .01689405 Theil entropy measure .03068818 Theil mean log deviation measure .027254 Entropy measure GE -1 .02483625 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15655196 coefficient of variation .44348017 standard deviation of logs .34872834 Gini coefficient .19649183 Mehran measure .25070267 Piesch measure .16938642 Kakwani measure .0402899 Theil entropy measure .07557302 Theil mean log deviation measure .06604351 Entropy measure GE -1 .06076768 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07139335 coefficient of variation .2185111 standard deviation of logs .1819374 Gini coefficient .09235147 Mehran measure .11623273 Piesch measure .08041084 Kakwani measure .01082654 Theil entropy measure .01948394 Theil mean log deviation measure .01739753 Entropy measure GE -1 .01583379 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14508016 coefficient of variation .43851069 standard deviation of logs .33575416 Gini coefficient .18901724 Mehran measure .23967768 Piesch measure .16368703 Kakwani measure .03807082 Theil entropy measure .07241339 Theil mean log deviation measure .06221124 Entropy measure GE -1 .05642694 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10153052 coefficient of variation .33535066 standard deviation of logs .25372321 Gini coefficient .13262636 Mehran measure .16422526 Piesch measure .11682691 Kakwani measure .02242318 Theil entropy measure .04254894 Theil mean log deviation measure .03599717 Entropy measure GE -1 .03162988 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16511316 coefficient of variation .53311017 standard deviation of logs .37832524 Gini coefficient .21648908 Mehran measure .26990225 Piesch measure .18978251 Kakwani measure .05058539 Theil entropy measure .100501 Theil mean log deviation measure .08293962 Entropy measure GE -1 .07366072 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07039149 coefficient of variation .24431816 standard deviation of logs .19431307 Gini coefficient .09583027 Mehran measure .11873103 Piesch measure .0843799 Kakwani measure .01279764 Theil entropy measure .02356491 Theil mean log deviation measure .0204641 Entropy measure GE -1 .01820072 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14497982 coefficient of variation .4908645 standard deviation of logs .34907827 Gini coefficient .19594546 Mehran measure .24389454 Piesch measure .17197092 Kakwani measure .04330959 Theil entropy measure .08573949 Theil mean log deviation measure .07059756 Entropy measure GE -1 .06212324 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10319831 coefficient of variation .36494253 standard deviation of logs .2655444 Gini coefficient .13724151 Mehran measure .1681343 Piesch measure .12179512 Kakwani measure .02527137 Theil entropy measure .04896417 Theil mean log deviation measure .04049893 Entropy measure GE -1 .03497601 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16698971 coefficient of variation .58318649 standard deviation of logs .39063409 Gini coefficient .22348967 Mehran measure .27480386 Piesch measure .19783259 Kakwani measure .05613452 Theil entropy measure .11509645 Theil mean log deviation measure .09198622 Entropy measure GE -1 .07991282 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07608519 coefficient of variation .27104657 standard deviation of logs .2081586 Gini coefficient .10109901 Mehran measure .12320671 Piesch measure .09004516 Kakwani measure .01507896 Theil entropy measure .02827373 Theil mean log deviation measure .02405059 Entropy measure GE -1 .02104073 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15727638 coefficient of variation .54422827 standard deviation of logs .36906595 Gini coefficient .20824798 Mehran measure .25513639 Piesch measure .18480376 Kakwani measure .05000844 Theil entropy measure .10160618 Theil mean log deviation measure .08155396 Entropy measure GE -1 .07059224 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10887918 coefficient of variation .39340764 standard deviation of logs .2778719 Gini coefficient .14263498 Mehran measure .17256684 Piesch measure .12766906 Kakwani measure .0282578 Theil entropy measure .0556491 Theil mean log deviation measure .04526759 Entropy measure GE -1 .03862883 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .17902842 coefficient of variation .63052865 standard deviation of logs .40717815 Gini coefficient .23407683 Mehran measure .2843825 Piesch measure .208924 Kakwani measure .06250549 Theil entropy measure .13075703 Theil mean log deviation measure .10256162 Entropy measure GE -1 .08812351 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07807423 coefficient of variation .28511016 standard deviation of logs .21302188 Gini coefficient .10066747 Mehran measure .12140366 Piesch measure .09029938 Kakwani measure .01613414 Theil entropy measure .03068411 Theil mean log deviation measure .02567422 Entropy measure GE -1 .02215157 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15892195 coefficient of variation .56849932 standard deviation of logs .36957864 Gini coefficient .20589674 Mehran measure .25115608 Piesch measure .18326707 Kakwani measure .05169943 Theil entropy measure .10751442 Theil mean log deviation measure .08421906 Entropy measure GE -1 .07157345 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11124194 coefficient of variation .40942359 standard deviation of logs .28227933 Gini coefficient .14356342 Mehran measure .1723921 Piesch measure .12914909 Kakwani measure .02968101 Theil entropy measure .05922764 Theil mean log deviation measure .0475131 Entropy measure GE -1 .04011027 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .18100892 coefficient of variation .654113 standard deviation of logs .40843099 Gini coefficient .23314972 Mehran measure .28180247 Piesch measure .20882334 Kakwani measure .06439297 Theil entropy measure .13725727 Theil mean log deviation measure .10564679 Entropy measure GE -1 .08956655 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9489 observations deleted) (210 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07792353 coefficient of variation .28923129 standard deviation of logs .21356383 Gini coefficient .09907427 Mehran measure .11840175 Piesch measure .08941054 Kakwani measure .0163745 Theil entropy measure .0313268 Theil mean log deviation measure .02602468 Entropy measure GE -1 .02231 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15720362 coefficient of variation .57274963 standard deviation of logs .36565578 Gini coefficient .20106968 Mehran measure .24313319 Piesch measure .18003792 Kakwani measure .05142539 Theil entropy measure .10790313 Theil mean log deviation measure .08360139 Entropy measure GE -1 .07029101 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .11116196 coefficient of variation .41399414 standard deviation of logs .28215878 Gini coefficient .1422821 Mehran measure .16917455 Piesch measure .12883589 Kakwani measure .02993015 Theil entropy measure .0600861 Theil mean log deviation measure .04787352 Entropy measure GE -1 .04017359 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .17965582 coefficient of variation .65800738 standard deviation of logs .4043512 Gini coefficient .22962497 Mehran measure .27549006 Piesch measure .20669243 Kakwani measure .06401021 Theil entropy measure .13751799 Theil mean log deviation measure .10483642 Entropy measure GE -1 .08804016 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07685566 coefficient of variation .28325015 standard deviation of logs .2106601 Gini coefficient .09833878 Mehran measure .11786912 Piesch measure .08857362 Kakwani measure .01584717 Theil entropy measure .03020149 Theil mean log deviation measure .02519491 Entropy measure GE -1 .02166958 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15450164 coefficient of variation .55518709 standard deviation of logs .36008382 Gini coefficient .19906487 Mehran measure .24084185 Piesch measure .17817638 Kakwani measure .04938702 Theil entropy measure .1026626 Theil mean log deviation measure .08020732 Entropy measure GE -1 .06774534 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .1091112 coefficient of variation .40864914 standard deviation of logs .2806039 Gini coefficient .14273146 Mehran measure .17124872 Piesch measure .12847282 Kakwani measure .02940573 Theil entropy measure .05882308 Theil mean log deviation measure .04707005 Entropy measure GE -1 .03966103 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .1754756 coefficient of variation .64245674 standard deviation of logs .40126227 Gini coefficient .22885087 Mehran measure .27684036 Piesch measure .20485613 Kakwani measure .06220387 Theil entropy measure .13254836 Theil mean log deviation measure .10193954 Entropy measure GE -1 .08621756 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07306121 coefficient of variation .26606857 standard deviation of logs .20170478 Gini coefficient .09484336 Mehran measure .11471595 Piesch measure .08490706 Kakwani measure .014328 Theil entropy measure .02702615 Theil mean log deviation measure .0227967 Entropy measure GE -1 .01977304 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14570677 coefficient of variation .5193 standard deviation of logs .34752144 Gini coefficient .19213594 Mehran measure .23537463 Piesch measure .1705166 Kakwani measure .04500081 Theil entropy measure .09195799 Theil mean log deviation measure .07304785 Entropy measure GE -1 .06238416 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10538033 coefficient of variation .39150355 standard deviation of logs .27336787 Gini coefficient .13911885 Mehran measure .16778904 Piesch measure .12478376 Kakwani measure .02758826 Theil entropy measure .05467332 Theil mean log deviation measure .044155 Entropy measure GE -1 .03744367 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16806214 coefficient of variation .61190558 standard deviation of logs .39174305 Gini coefficient .22316364 Mehran measure .27222583 Piesch measure .19863255 Kakwani measure .05827044 Theil entropy measure .12250958 Theil mean log deviation measure .09542683 Entropy measure GE -1 .08135929 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .06815186 coefficient of variation .2422374 standard deviation of logs .18820875 Gini coefficient .08759467 Mehran measure .10573389 Piesch measure .07852507 Kakwani measure .01226666 Theil entropy measure .02282506 Theil mean log deviation measure .01952769 Entropy measure GE -1 .01710927 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .13250908 coefficient of variation .46551718 standard deviation of logs .32182809 Gini coefficient .17501235 Mehran measure .21553103 Piesch measure .15475301 Kakwani measure .03793693 Theil entropy measure .07600036 Theil mean log deviation measure .06134264 Entropy measure GE -1 .05270119 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10101725 coefficient of variation .36867548 standard deviation of logs .26306697 Gini coefficient .13336848 Mehran measure .16117233 Piesch measure .11946657 Kakwani measure .02519537 Theil entropy measure .04930953 Theil mean log deviation measure .04030063 Entropy measure GE -1 .03443559 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15764151 coefficient of variation .56791765 standard deviation of logs .37246923 Gini coefficient .20964858 Mehran measure .25672606 Piesch measure .18610985 Kakwani measure .05206608 Theil entropy measure .10783729 Theil mean log deviation measure .08494058 Entropy measure GE -1 .07260799 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .06861711 coefficient of variation .22629972 standard deviation of logs .182391 Gini coefficient .08599236 Mehran measure .10435381 Piesch measure .07681163 Kakwani measure .01120737 Theil entropy measure .02046011 Theil mean log deviation measure .01789765 Entropy measure GE -1 .01596686 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .13624384 coefficient of variation .43600522 standard deviation of logs .3190088 Gini coefficient .17327456 Mehran measure .21309848 Piesch measure .15336259 Kakwani measure .03598345 Theil entropy measure .06974353 Theil mean log deviation measure .05816587 Entropy measure GE -1 .05107536 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10157273 coefficient of variation .35066991 standard deviation of logs .25835516 Gini coefficient .13176965 Mehran measure .15966213 Piesch measure .1178234 Kakwani measure .02377144 Theil entropy measure .04569685 Theil mean log deviation measure .03805168 Entropy measure GE -1 .03296512 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16088814 coefficient of variation .53957996 standard deviation of logs .3700125 Gini coefficient .20769711 Mehran measure .25406474 Piesch measure .18451329 Kakwani measure .05004016 Theil entropy measure .10091351 Theil mean log deviation measure .08150864 Entropy measure GE -1 .07075462 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07013871 coefficient of variation .21983953 standard deviation of logs .18215535 Gini coefficient .08708126 Mehran measure .10635003 Piesch measure .07744688 Kakwani measure .01095464 Theil entropy measure .01971388 Theil mean log deviation measure .01754006 Entropy measure GE -1 .01586262 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14060356 coefficient of variation .42638113 standard deviation of logs .32263515 Gini coefficient .1744617 Mehran measure .21520493 Piesch measure .15409007 Kakwani measure .03613052 Theil entropy measure .06864533 Theil mean log deviation measure .05838367 Entropy measure GE -1 .05190779 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10141286 coefficient of variation .33162204 standard deviation of logs .25468716 Gini coefficient .13287995 Mehran measure .16412986 Piesch measure .11725501 Kakwani measure .02240056 Theil entropy measure .0421281 Theil mean log deviation measure .03595883 Entropy measure GE -1 .03177916 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15795016 coefficient of variation .50777027 standard deviation of logs .36486992 Gini coefficient .20469838 Mehran measure .25322959 Piesch measure .18043277 Kakwani measure .04721441 Theil entropy measure .09275204 Theil mean log deviation measure .07691404 Entropy measure GE -1 .06796684 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07030741 coefficient of variation .21667196 standard deviation of logs .18198903 Gini coefficient .08634214 Mehran measure .10478863 Piesch measure .0771189 Kakwani measure .01084094 Theil entropy measure .01936308 Theil mean log deviation measure .01736997 Entropy measure GE -1 .01580025 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14004084 coefficient of variation .41898594 standard deviation of logs .32113874 Gini coefficient .17199824 Mehran measure .21112582 Piesch measure .15243443 Kakwani measure .03567598 Theil entropy measure .0671826 Theil mean log deviation measure .05753158 Entropy measure GE -1 .05124454 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09948713 coefficient of variation .30641602 standard deviation of logs .24590443 Gini coefficient .12809686 Mehran measure .15939579 Piesch measure .11244739 Kakwani measure .02029123 Theil entropy measure .03724433 Theil mean log deviation measure .03261863 Entropy measure GE -1 .02935327 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15394784 coefficient of variation .46919205 standard deviation of logs .35193072 Gini coefficient .19445379 Mehran measure .24052808 Piesch measure .17141665 Kakwani measure .04300193 Theil entropy measure .0822897 Theil mean log deviation measure .06983017 Entropy measure GE -1 .06238187 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07125505 coefficient of variation .22010052 standard deviation of logs .18502407 Gini coefficient .08602621 Mehran measure .10373207 Piesch measure .07717327 Kakwani measure .01121457 Theil entropy measure .02001035 Theil mean log deviation measure .01795913 Entropy measure GE -1 .01633113 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .1436899 coefficient of variation .43288171 standard deviation of logs .32931804 Gini coefficient .17422984 Mehran measure .21279169 Piesch measure .15494892 Kakwani measure .03783293 Theil entropy measure .0714543 Theil mean log deviation measure .06093442 Entropy measure GE -1 .05404669 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10176421 coefficient of variation .30581525 standard deviation of logs .24772569 Gini coefficient .12840602 Mehran measure .15920716 Piesch measure .11300544 Kakwani measure .02049909 Theil entropy measure .03741208 Theil mean log deviation measure .03294703 Entropy measure GE -1 .0297444 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .15686739 coefficient of variation .47456225 standard deviation of logs .358133 Gini coefficient .19753551 Mehran measure .24326107 Piesch measure .17467273 Kakwani measure .04452365 Theil entropy measure .08480541 Theil mean log deviation measure .0722283 Entropy measure GE -1 .06458918 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .0713378 coefficient of variation .21982755 standard deviation of logs .18439188 Gini coefficient .08457721 Mehran measure .10135365 Piesch measure .076189 Kakwani measure .01116191 Theil entropy measure .01993385 Theil mean log deviation measure .01786504 Entropy measure GE -1 .01622117 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .14736223 coefficient of variation .4430839 standard deviation of logs .33507276 Gini coefficient .17659257 Mehran measure .21393964 Piesch measure .15791904 Kakwani measure .0393444 Theil entropy measure .07453539 Theil mean log deviation measure .06336699 Entropy measure GE -1 .05608902 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .10303768 coefficient of variation .30392927 standard deviation of logs .24882624 Gini coefficient .12894231 Mehran measure .16037165 Piesch measure .11322765 Kakwani measure .02053318 Theil entropy measure .03726406 Theil mean log deviation measure .03302872 Entropy measure GE -1 .02996546 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .16089784 coefficient of variation .47969447 standard deviation of logs .36585518 Gini coefficient .2028909 Mehran measure .25101397 Piesch measure .17882938 Kakwani measure .04610742 Theil entropy measure .08735918 Theil mean log deviation measure .07491476 Entropy measure GE -1 .06741168 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .07188654 coefficient of variation .22551731 standard deviation of logs .18554553 Gini coefficient .08321713 Mehran measure .09887573 Piesch measure .07538783 Kakwani measure .01146653 Theil entropy measure .02067805 Theil mean log deviation measure .01831411 Entropy measure GE -1 .01646862 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .15642895 coefficient of variation .49279433 standard deviation of logs .35327411 Gini coefficient .1866006 Mehran measure .2225208 Piesch measure .1686405 Kakwani measure .04532149 Theil entropy measure .08840404 Theil mean log deviation measure .07298188 Entropy measure GE -1 .0633468 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09971409 coefficient of variation .29243049 standard deviation of logs .23951622 Gini coefficient .121103 Mehran measure .14906926 Piesch measure .10711988 Kakwani measure .0190578 Theil entropy measure .03454659 Theil mean log deviation measure .03060695 Entropy measure GE -1 .02770859 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .1724248 coefficient of variation .50597131 standard deviation of logs .38088917 Gini coefficient .21087447 Mehran measure .26032956 Piesch measure .18614692 Kakwani measure .05044993 Theil entropy measure .09625293 Theil mean log deviation measure .08201671 Entropy measure GE -1 .07356211 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .06939117 coefficient of variation .22921518 standard deviation of logs .18371833 Gini coefficient .0811099 Mehran measure .09457759 Piesch measure .07437606 Kakwani measure .01145198 Theil entropy measure .02093938 Theil mean log deviation measure .01825083 Entropy measure GE -1 .01620509 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .16520872 coefficient of variation .54390744 standard deviation of logs .37825238 Gini coefficient .20913549 Mehran measure .25376272 Piesch measure .18682188 Kakwani measure .05228582 Theil entropy measure .1042498 Theil mean log deviation measure .0848886 Entropy measure GE -1 .07378712 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .09222204 coefficient of variation .28056429 standard deviation of logs .22507434 Gini coefficient .11149793 Mehran measure .13405757 Piesch measure .10021811 Kakwani measure .01707377 Theil entropy measure .03128578 Theil mean log deviation measure .02735345 Entropy measure GE -1 .0244875 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .17847855 coefficient of variation .53164047 standard deviation of logs .3949132 Gini coefficient .2226954 Mehran measure .27448258 Piesch measure .19680181 Kakwani measure .05429518 Theil entropy measure .10462132 Theil mean log deviation measure .08866051 Entropy measure GE -1 .07969131 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved (9488 observations deleted) (211 observations deleted) inequality measures of polarityshare_index_rescale ------------------------------------------------------------------------------ relative mean deviation .06636342 coefficient of variation .22515922 standard deviation of logs .17940365 Gini coefficient .07846199 Mehran measure .09069345 Piesch measure .07234627 Kakwani measure .01096796 Theil entropy measure .02011658 Theil mean log deviation measure .01746708 Entropy measure GE -1 .01545865 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index2_rescale ------------------------------------------------------------------------------ relative mean deviation .16916691 coefficient of variation .55434047 standard deviation of logs .38027874 Gini coefficient .20795771 Mehran measure .24836604 Piesch measure .18775354 Kakwani measure .05355197 Theil entropy measure .10740108 Theil mean log deviation measure .08677048 Entropy measure GE -1 .07481995 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index3_rescale ------------------------------------------------------------------------------ relative mean deviation .089889 coefficient of variation .27627865 standard deviation of logs .22086289 Gini coefficient .1097721 Mehran measure .13223341 Piesch measure .09854145 Kakwani measure .01646887 Theil entropy measure .03024247 Theil mean log deviation measure .02638172 Entropy measure GE -1 .02358096 ------------------------------------------------------------------------------ (0 observations deleted) (0 observations deleted) inequality measures of polarityshare_index4_rescale ------------------------------------------------------------------------------ relative mean deviation .1847484 coefficient of variation .54823572 standard deviation of logs .40603665 Gini coefficient .22969844 Mehran measure .28257615 Piesch measure .20325957 Kakwani measure .05743302 Theil entropy measure .11090176 Theil mean log deviation measure .09394035 Entropy measure GE -1 .08458926 ------------------------------------------------------------------------------ file C:\Llama\Temp\conc.dta saved . use "C:\Llama\Temp\conc", clear . replace herf_index2_rescale = . if year < 1981 (13 real changes made, 13 to missing) . replace normherf_index2_rescale = . if year < 1981 (13 real changes made, 13 to missing) . replace herf_index4_rescale = . if year < 1983 (15 real changes made, 15 to missing) . replace normherf_index4_rescale = . if year < 1983 (15 real changes made, 15 to missing) . replace theil_index2_rescale = . if year < 1981 (13 real changes made, 13 to missing) . replace normtheil_index2_rescale = . if year < 1981 (13 real changes made, 13 to missing) . replace theil_index4_rescale = . if year < 1983 (15 real changes made, 15 to missing) . replace normtheil_index4_rescale = . if year < 1983 (15 real changes made, 15 to missing) . replace gini_index2_rescale = . if year < 1981 (2 real changes made, 2 to missing) . replace gini_index4_rescale = . if year < 1983 (0 real changes made) . local indexlist "index_rescale index2_rescale index3_rescale index4_rescale" . . label variable herf_index_rescale "Herfindahl index, constant USD growth rate, without mi > gration subindex" . label variable normherf_index_rescale "Normalized Herfindahl index, constant USD growth r > ate, without migration subindex" note: label truncated to 80 characters . label variable herf_index2_rescale "Herfindahl index, PPP growth rate, without migration > subindex" . label variable normherf_index_rescale "Normalized Herfindahl index, PPP growth rate, with > out migration subindex" . label variable herf_index3_rescale "Herfindahl index, constant USD growth rate, with migr > ation subindex" . label variable normherf_index3_rescale "Normalized Herfindahl index, constant USD growth > rate, with migration subindex" . label variable herf_index4_rescale "Herfindahl index, PPP growth rate, with migration sub > index" . label variable normherf_index4_rescale "Normalized Herfindahl index, PPP growth rate, wit > h migration subindex" . . label variable theil_index_rescale "Theil index, constant USD growth rate, without migrat > ion subindex" . label variable normtheil_index_rescale "Normalized Theil index, constant USD growth rate, > without migration subindex" . label variable theil_index2_rescale "Theil index, PPP growth rate, without migration subi > ndex" . label variable normtheil_index_rescale "Normalized Theil index, PPP growth rate, without > migration subindex" . label variable theil_index3_rescale "Theil index, constant USD growth rate, with migratio > n subindex" . label variable normtheil_index3_rescale "Normalized Theil index, constant USD growth rate > , with migration subindex" . label variable theil_index4_rescale "Theil index, PPP growth rate, with migration subinde > x" . label variable normtheil_index4_rescale "Normalized Theil index, PPP growth rate, with mi > gration subindex" . . label variable gini_index_rescale "Gini index, constant USD growth rate, without migratio > n subindex" . label variable gini_index2_rescale "Gini index, PPP growth rate, without migration subind > ex" . label variable gini_index3_rescale "Gini index, constant USD growth rate, with migration > subindex" . label variable gini_index4_rescale "Gini index, PPP growth rate, with migration subindex" . . tsset year time variable: year, 1968 to 2008 . sort year . save "C:\Llama\poledist.dta", replace file C:\Llama\poledist.dta saved . . reg normherf_index_rescale year, robust /* Sig only at 10 percent */ Linear regression Number of obs = 41 F( 1, 39) = 3.76 Prob > F = 0.0599 R-squared = 0.1869 Root MSE = .00263 ------------------------------------------------------------------------------ | Robust normherf_i.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0001038 .0000535 -1.94 0.060 -.000212 4.53e-06 _cons | .2110779 .1067975 1.98 0.055 -.0049404 .4270961 ------------------------------------------------------------------------------ . estimates store D1 . reg normherf_index2_rescale year, robust Linear regression Number of obs = 28 F( 1, 26) = 10.21 Prob > F = 0.0036 R-squared = 0.2626 Root MSE = .00462 ------------------------------------------------------------------------------ | Robust normherf_i.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0003292 .000103 -3.20 0.004 -.000541 -.0001174 _cons | .6744787 .2055687 3.28 0.003 .2519262 1.097031 ------------------------------------------------------------------------------ . estimates store D2 . reg normherf_index3_rescale year, robust Linear regression Number of obs = 41 F( 1, 39) = 83.46 Prob > F = 0.0000 R-squared = 0.5448 Root MSE = .00343 ------------------------------------------------------------------------------ | Robust normherf_i.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0003092 .0000338 -9.14 0.000 -.0003776 -.0002407 _cons | .6262784 .067492 9.28 0.000 .4897628 .7627939 ------------------------------------------------------------------------------ . estimates store D3 . reg normherf_index4_rescale year, robust Linear regression Number of obs = 26 F( 1, 24) = 12.94 Prob > F = 0.0014 R-squared = 0.3477 Root MSE = .00773 ------------------------------------------------------------------------------ | Robust normherf_i.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0007227 .0002009 -3.60 0.001 -.0011374 -.0003081 _cons | 1.465673 .4019522 3.65 0.001 .6360845 2.295262 ------------------------------------------------------------------------------ . estimates store D4 . reg normtheil_index_rescale year, robust Linear regression Number of obs = 41 F( 1, 39) = 26.22 Prob > F = 0.0000 R-squared = 0.3572 Root MSE = .00214 ------------------------------------------------------------------------------ | Robust normtheil_.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0001317 .0000257 -5.12 0.000 -.0001837 -.0000796 _cons | .2716479 .0512608 5.30 0.000 .1679631 .3753327 ------------------------------------------------------------------------------ . estimates store D5 . reg normtheil_index2_rescale year, robust Linear regression Number of obs = 28 F( 1, 26) = 13.53 Prob > F = 0.0011 R-squared = 0.3354 Root MSE = .00772 ------------------------------------------------------------------------------ | Robust normtheil_.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0006538 .0001778 -3.68 0.001 -.0010193 -.0002884 _cons | 1.339747 .3546312 3.78 0.001 .6107923 2.068702 ------------------------------------------------------------------------------ . estimates store D6 . reg normtheil_index3_rescale year, robust Linear regression Number of obs = 41 F( 1, 39) = 6.99 Prob > F = 0.0117 R-squared = 0.0916 Root MSE = .004 ------------------------------------------------------------------------------ | Robust normtheil_.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0001048 .0000396 -2.64 0.012 -.000185 -.0000246 _cons | .225698 .0787739 2.87 0.007 .0663627 .3850333 ------------------------------------------------------------------------------ . estimates store D7 . reg normtheil_index4_rescale year, robust Linear regression Number of obs = 26 F( 1, 24) = 10.56 Prob > F = 0.0034 R-squared = 0.2646 Root MSE = .00889 ------------------------------------------------------------------------------ | Robust normtheil_.. | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0006829 .0002102 -3.25 0.003 -.0011166 -.0002491 _cons | 1.404652 .4203013 3.34 0.003 .5371928 2.272111 ------------------------------------------------------------------------------ . estimates store D8 . reg gini_index_rescale year, robust Linear regression Number of obs = 41 F( 1, 39) = 48.16 Prob > F = 0.0000 R-squared = 0.5832 Root MSE = .01337 ------------------------------------------------------------------------------ | Robust gi~x_rescale | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0013037 .0001879 -6.94 0.000 -.0016837 -.0009237 _cons | 2.697413 .3750885 7.19 0.000 1.938725 3.456101 ------------------------------------------------------------------------------ . estimates store D9 . reg gini_index2_rescale year, robust Linear regression Number of obs = 28 F( 1, 26) = 29.57 Prob > F = 0.0000 R-squared = 0.5798 Root MSE = .01996 ------------------------------------------------------------------------------ | Robust gi~2_rescale | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.002797 .0005143 -5.44 0.000 -.0038542 -.0017398 _cons | 5.786646 1.025641 5.64 0.000 3.678411 7.894882 ------------------------------------------------------------------------------ . estimates store D10 . reg gini_index3_rescale year, robust Linear regression Number of obs = 41 F( 1, 39) = 26.24 Prob > F = 0.0000 R-squared = 0.3528 Root MSE = .01683 ------------------------------------------------------------------------------ | Robust gi~3_rescale | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.001024 .0001999 -5.12 0.000 -.0014284 -.0006196 _cons | 2.178807 .3989099 5.46 0.000 1.371936 2.985679 ------------------------------------------------------------------------------ . estimates store D11 . reg gini_index4_rescale year, robust Linear regression Number of obs = 26 F( 1, 24) = 18.62 Prob > F = 0.0002 R-squared = 0.4863 Root MSE = .02222 ------------------------------------------------------------------------------ | Robust gi~4_rescale | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- year | -.0027692 .0006417 -4.32 0.000 -.0040937 -.0014447 _cons | 5.755469 1.282217 4.49 0.000 3.109103 8.401835 ------------------------------------------------------------------------------ . estimates store D12 . . estout * using "C:\Llama\Output\timeseries.tex", replace style(tex) varlabels(_cons Const > ant) cells (b(star fmt(%9.3f)) se(par fmt(%9.2f))) stats (r2_a N, labels("Adjusted R^2" " > N") fmt(%8.3f %6.0f)) starlevels(* 0.10 ** 0.05 *** 0.01) & D1 & D2 & D3 & D4 & D5 > & D6 & D7 & D8 & D9 & D10 & > D11 & D12 \\ & b/se & b/se & b/se & b/se & b/se > & b/se & b/se & b/se & b/se & b/se & > b/se & b/se \\ year & -0.000* & -0.000***& -0.000***& -0.001***& -0.000** > *& -0.001***& -0.000** & -0.001***& -0.001***& -0.003***& - > 0.001***& -0.003***\\ & (0.00) & (0.00) & (0.00) & (0.00) & (0.00) > & (0.00) & (0.00) & (0.00) & (0.00) & (0.00) & ( > 0.00) & (0.00) \\ Constant & 0.211* & 0.674***& 0.626***& 1.466***& 0.272** > *& 1.340***& 0.226***& 1.405***& 2.697***& 5.787***& > 2.179***& 5.755***\\ & (0.11) & (0.21) & (0.07) & (0.40) & (0.05) > & (0.35) & (0.08) & (0.42) & (0.38) & (1.03) & ( > 0.40) & (1.28) \\ Adjusted R^2& 0.166 & 0.234 & 0.533 & 0.321 & 0.341 > & 0.310 & 0.068 & 0.234 & 0.573 & 0.564 & > 0.336 & 0.465 \\ N & 41 & 28 & 41 & 26 & 41 > & 28 & 41 & 26 & 41 & 28 & > 41 & 26 \\ . estimates clear . /* > preserve > keep if year == 1971 > graph twoway bar index3_rescale lgdppc > restore > preserve > keep if year == 1991 > graph twoway bar index3_rescale lgdppc > restore > preserve > keep if year == 2006 > graph twoway bar index3_rescale lgdppc > restore end of do-file . . /* Close */ . . clear . set more on . set mem 1m (1024k) . log close log: C:\Llama\GrowthPoles.txt log type: text closed on: 23 Aug 2011, 17:55:39 -------------------------------------------------------------------------------------------