------------------------------------------------------------------------------------------ log: C:\Documents and Settings\Angela Fonseca\Mis documentos\聲gela\Ciclos\Composi > tion Mar2008\outputs\voting_07_08_todofin_wa.txt log type: text opened on: 16 Jul 2008, 13:10:36 . . set more off . . . **Running regressions for composition paper (new table 4)** . clear . use "C:\Documents and Settings\Angela Fonseca\Mis documentos\聲gela\Ciclos\Composition M > ar2008\datos\populat02_08fin2wa.dta", clear . . **NOTE that in the original paper the interaction variables don't have the '2' because t > he dataset they use, populat2, are named different* . regress pctv pcv_lag deficit deficit_coinp2 invers invers_coinp2 gfunci gfunci_coinp2 gd > pgr gdpgr_coinp2, robust Linear regression Number of obs = 2032 F( 9, 2022) = 64.15 Prob > F = 0.0000 R-squared = 0.2298 Root MSE = 26.416 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .5039794 .0278381 18.10 0.000 .4493851 .5585737 deficit | .0257958 .0188063 1.37 0.170 -.0110859 .0626776 deficit_co~2 | -.0466457 .0228301 -2.04 0.041 -.0914188 -.0018726 invers | -3.546657 1.557267 -2.28 0.023 -6.600672 -.4926429 invers_coi~2 | 6.059115 1.775257 3.41 0.001 2.577591 9.540639 gfunci | -1.240801 1.824009 -0.68 0.496 -4.817934 2.336332 gfunci_coi~2 | 1.80937 1.786598 1.01 0.311 -1.694394 5.313135 gdpgr | 140.7787 39.27133 3.58 0.000 63.76216 217.7951 gdpgr_coinp2 | -94.97985 44.36398 -2.14 0.032 -181.9837 -7.975957 _cons | 29.15835 2.341918 12.45 0.000 24.56552 33.75117 ------------------------------------------------------------------------------ . test deficit+deficit_coinp2=0 ( 1) deficit + deficit_coinp2 = 0 F( 1, 2022) = 2.59 Prob > F = 0.1075 . test invers+invers_coinp2=0 ( 1) invers + invers_coinp2 = 0 F( 1, 2022) = 7.09 Prob > F = 0.0078 . test gfunci+gfunci_coinp2=0 ( 1) gfunci + gfunci_coinp2 = 0 F( 1, 2022) = 0.25 Prob > F = 0.6141 . . regress pctvd pcv_lagd deficitpd deficit_coinp2d inverspd invers_coinp2d gfuncipd gfunci > _coinp2d gdpgrpd gdpgr_coinp2d, robust noconstant Linear regression Number of obs = 2032 F( 9, 2023) = 50.14 Prob > F = 0.0000 R-squared = 0.1949 Root MSE = 25.987 ------------------------------------------------------------------------------ | Robust pctvd | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lagd | .3894856 .0261661 14.89 0.000 .3381703 .4408009 deficitpd | .0280636 .0171535 1.64 0.102 -.0055767 .061704 deficit_c~2d | -.0486255 .0205647 -2.36 0.018 -.0889557 -.0082954 inverspd | -.913553 1.12326 -0.81 0.416 -3.11642 1.289314 invers_co~2d | 3.367228 1.423543 2.37 0.018 .5754643 6.158991 gfuncipd | .1295557 1.399213 0.09 0.926 -2.614493 2.873604 gfunci_co~2d | -1.504388 1.383988 -1.09 0.277 -4.218578 1.209803 gdpgrpd | 73.83646 31.74945 2.33 0.020 11.57143 136.1015 gdpgr_coi~2d | 22.00643 33.45855 0.66 0.511 -43.61038 87.62325 ------------------------------------------------------------------------------ . test deficitpd+deficit_coinp2d=0 ( 1) deficitpd + deficit_coinp2d = 0 F( 1, 2023) = 3.26 Prob > F = 0.0712 . test inverspd+invers_coinp2d=0 ( 1) inverspd + invers_coinp2d = 0 F( 1, 2023) = 6.99 Prob > F = 0.0083 . test gfuncipd+gfunci_coinp2d=0 ( 1) gfuncipd + gfunci_coinp2d = 0 F( 1, 2023) = 1.79 Prob > F = 0.1807 . . clear . use "C:\Documents and Settings\Angela Fonseca\Mis documentos\聲gela\Ciclos\Composition M > ar2008\datos\regmiggastos02_08fin2wa.dta", clear . regress pctv pcv_lag deficit deficit_coin2 invers invers_coin2 gdpgr gdpgr_coin2, robust Linear regression Number of obs = 2052 F( 7, 2044) = 80.19 Prob > F = 0.0000 R-squared = 0.2280 Root MSE = 26.449 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .5243948 .0264984 19.79 0.000 .4724281 .5763615 deficit | .0232511 .0151901 1.53 0.126 -.0065387 .0530408 deficit_co~2 | -.0371689 .0191342 -1.94 0.052 -.0746934 .0003556 invers | -4.305668 2.56838 -1.68 0.094 -9.342583 .7312475 invers_coin2 | 9.382664 3.129825 3.00 0.003 3.244685 15.52064 gdpgr | 115.8134 40.03789 2.89 0.004 37.29405 194.3327 gdpgr_coin2 | -60.92867 45.68363 -1.33 0.182 -150.52 28.66265 _cons | 28.86951 2.111162 13.67 0.000 24.72926 33.00977 ------------------------------------------------------------------------------ . test deficit+deficit_coin2=0 ( 1) deficit + deficit_coin2 = 0 F( 1, 2044) = 1.43 Prob > F = 0.2323 . test invers+invers_coin2=0 ( 1) invers + invers_coin2 = 0 F( 1, 2044) = 4.43 Prob > F = 0.0354 . . *** New Table 5 *** . **** Column 1 **** . use "C:\Documents and Settings\Angela Fonseca\Mis documentos\聲gela\Ciclos\Composition M > ar2008\datos\populat02_08fin2wa.dta", clear . . **Column 1, without vars interacted with incumbent dummy, run only for incumbents** . regress pctv pcv_lag deficit invers gfunci gdpgr if coin2==1, robust Linear regression Number of obs = 1369 F( 5, 1363) = 86.01 Prob > F = 0.0000 R-squared = 0.2268 Root MSE = 25.124 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .6848419 .0342706 19.98 0.000 .6176132 .7520707 deficit | -.0216748 .0129091 -1.68 0.093 -.0469986 .0036491 invers | 2.642349 .9365175 2.82 0.005 .805177 4.479521 gfunci | -.5989457 1.080036 -0.55 0.579 -2.71766 1.519768 gdpgr | 55.75378 23.92438 2.33 0.020 8.821186 102.6864 _cons | 12.90268 3.172554 4.07 0.000 6.679063 19.1263 ------------------------------------------------------------------------------ . . . ***** Column 2: swing1a (swing2c not included) ***** . . foreach var in swing1a swing2c { 2. regress pctv pcv_lag `var' deficit deficit_`var' invers invers_`var' gfunci gfunci_`v > ar' gdpgr if coin2==1, robust 3. test deficit+deficit_`var'=0 4. test invers+invers_`var'=0 5. test gfunci+gfunci_`var'=0 6. } Linear regression Number of obs = 1369 F( 9, 1359) = 118.99 Prob > F = 0.0000 R-squared = 0.3966 Root MSE = 22.226 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .4979103 .0347695 14.32 0.000 .4297026 .5661179 swing1a | -29.85477 2.196062 -13.59 0.000 -34.16281 -25.54673 deficit | -.0113502 .0093337 -1.22 0.224 -.0296603 .0069599 deficit_s~1a | -.0419182 .0226027 -1.85 0.064 -.0862582 .0024217 invers | -.6618803 1.010132 -0.66 0.512 -2.643467 1.319706 invers_sw~1a | 9.899197 1.868114 5.30 0.000 6.234496 13.5639 gfunci | 2.964862 1.40792 2.11 0.035 .2029299 5.726794 gfunci_sw~1a | -6.452738 1.833672 -3.52 0.000 -10.04987 -2.855604 gdpgr | 53.9362 20.31407 2.66 0.008 14.08587 93.78653 _cons | 41.40985 3.577378 11.58 0.000 34.39206 48.42763 ------------------------------------------------------------------------------ ( 1) deficit + deficit_swing1a = 0 F( 1, 1359) = 6.70 Prob > F = 0.0098 ( 1) invers + invers_swing1a = 0 F( 1, 1359) = 34.98 Prob > F = 0.0000 ( 1) gfunci + gfunci_swing1a = 0 F( 1, 1359) = 8.41 Prob > F = 0.0038 Linear regression Number of obs = 1369 F( 9, 1359) = 96.22 Prob > F = 0.0000 R-squared = 0.3633 Root MSE = 22.832 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .4634292 .0365194 12.69 0.000 .3917886 .5350698 swing2c | -24.82675 2.335661 -10.63 0.000 -29.40864 -20.24485 deficit | -.0057345 .0073187 -0.78 0.433 -.0200916 .0086226 deficit_s~2c | -.0516211 .0219264 -2.35 0.019 -.0946345 -.0086078 invers | -.1263568 1.047401 -0.12 0.904 -2.181056 1.928342 invers_sw~2c | 3.487817 1.630722 2.14 0.033 .2888115 6.686823 gfunci | .9278178 1.527831 0.61 0.544 -2.069346 3.924981 gfunci_sw~2c | -2.486237 2.010677 -1.24 0.216 -6.430604 1.45813 gdpgr | 52.85224 21.56934 2.45 0.014 10.53942 95.16506 _cons | 43.54811 3.749282 11.62 0.000 36.1931 50.90312 ------------------------------------------------------------------------------ ( 1) deficit + deficit_swing2c = 0 F( 1, 1359) = 7.70 Prob > F = 0.0056 ( 1) invers + invers_swing2c = 0 F( 1, 1359) = 7.22 Prob > F = 0.0073 ( 1) gfunci + gfunci_swing2c = 0 F( 1, 1359) = 1.34 Prob > F = 0.2474 . . ***** Not included: swing1a_3 swing2c_3 ***** . . foreach var in swing1a_3 swing2c_3 { 2. regress pctv pcv_lag `var' deficit deficit_`var' invers invers_`var' gfunci gfunci_`v > ar' gdpgr if coin2==1, robust 3. test deficit+deficit_`var'=0 4. test invers+invers_`var'=0 5. test gfunci+gfunci_`var'=0 6. } Linear regression Number of obs = 1330 F( 9, 1320) = 113.94 Prob > F = 0.0000 R-squared = 0.3934 Root MSE = 22.28 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .5053234 .0356383 14.18 0.000 .4354096 .5752373 swing1a_3 | -29.06296 2.234024 -13.01 0.000 -33.44559 -24.68034 deficit | -.010908 .0090455 -1.21 0.228 -.0286532 .0068372 deficit_~a_3 | -.0447383 .0227413 -1.97 0.049 -.0893513 -.0001252 invers | -.560586 1.012164 -0.55 0.580 -2.546211 1.425039 invers_s~a_3 | 9.741183 1.909557 5.10 0.000 5.995084 13.48728 gfunci | 3.006765 1.421375 2.12 0.035 .2183637 5.795166 gfunci_s~a_3 | -6.317567 1.850298 -3.41 0.001 -9.947414 -2.687721 gdpgr | 58.95909 20.76836 2.84 0.005 18.2165 99.70169 _cons | 40.56973 3.660622 11.08 0.000 33.38846 47.751 ------------------------------------------------------------------------------ ( 1) deficit + deficit_swing1a_3 = 0 F( 1, 1320) = 7.12 Prob > F = 0.0077 ( 1) invers + invers_swing1a_3 = 0 F( 1, 1320) = 32.40 Prob > F = 0.0000 ( 1) gfunci + gfunci_swing1a_3 = 0 F( 1, 1320) = 7.47 Prob > F = 0.0064 Linear regression Number of obs = 1330 F( 9, 1320) = 96.91 Prob > F = 0.0000 R-squared = 0.3721 Root MSE = 22.668 ------------------------------------------------------------------------------ | Robust pctv | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .4690939 .0372114 12.61 0.000 .3960939 .5420938 swing2c_3 | -25.5477 2.318303 -11.02 0.000 -30.09566 -20.99974 deficit | -.004915 .0067429 -0.73 0.466 -.018143 .008313 deficit_~c_3 | -.0534828 .0219012 -2.44 0.015 -.0964476 -.0105179 invers | -.4905676 .9812817 -0.50 0.617 -2.415609 1.434474 invers_s~c_3 | 3.820909 1.597997 2.39 0.017 .6860178 6.9558 gfunci | 1.70473 1.499719 1.14 0.256 -1.237363 4.646824 gfunci_s~c_3 | -3.142322 2.014049 -1.56 0.119 -7.093408 .8087642 gdpgr | 55.49846 21.62558 2.57 0.010 13.07421 97.92271 _cons | 44.13702 3.775591 11.69 0.000 36.7302 51.54383 ------------------------------------------------------------------------------ ( 1) deficit + deficit_swing2c_3 = 0 F( 1, 1320) = 7.86 Prob > F = 0.0051 ( 1) invers + invers_swing2c_3 = 0 F( 1, 1320) = 6.94 Prob > F = 0.0085 ( 1) gfunci + gfunci_swing2c_3 = 0 F( 1, 1320) = 1.09 Prob > F = 0.2973 . . . *** New Table 6 *** . * PROBIT * . . ** Column 1 (mfx) ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1, robust Iteration 0: log pseudolikelihood = -955.11313 Iteration 1: log pseudolikelihood = -881.49933 Iteration 2: log pseudolikelihood = -881.34979 Iteration 3: log pseudolikelihood = -881.34979 Probit regression Number of obs = 1462 Wald chi2(5) = 144.67 Prob > chi2 = 0.0000 Log pseudolikelihood = -881.34979 Pseudo R2 = 0.0772 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0208743 .0017601 11.86 0.000 .0174245 .0243241 deficit | -.0005973 .0004926 -1.21 0.225 -.0015628 .0003683 invers | .068907 .0471474 1.46 0.144 -.0235001 .1613141 gfunci | -.0373901 .0631017 -0.59 0.553 -.1610672 .086287 gdpgr | .3104799 1.220123 0.25 0.799 -2.080917 2.701876 _cons | -1.322896 .156481 -8.45 0.000 -1.629593 -1.016198 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .64955989 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0077354 .00065 11.87 0.000 .006458 .009012 79.7966 deficit | -.0002213 .00018 -1.21 0.225 -.000579 .000136 -.270509 invers | .0255348 .01747 1.46 0.144 -.008709 .059778 -.004475 gfunci | -.0138556 .02338 -0.59 0.553 -.059681 .031969 -.903845 gdpgr | .115054 .45217 0.25 0.799 -.771186 1.00129 .024738 ------------------------------------------------------------------------------ . . . . ***** SWING1A: : included, Columns 2 & 3 ***** . . *** Swing *** . ** Column 2 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing1a==1, robust Iteration 0: log pseudolikelihood = -430.56562 Iteration 1: log pseudolikelihood = -404.84559 Iteration 2: log pseudolikelihood = -404.66217 Iteration 3: log pseudolikelihood = -404.66186 Probit regression Number of obs = 652 Wald chi2(5) = 51.67 Prob > chi2 = 0.0000 Log pseudolikelihood = -404.66186 Pseudo R2 = 0.0602 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0158243 .0025411 6.23 0.000 .0108439 .0208047 deficit | -.0025249 .0012165 -2.08 0.038 -.0049093 -.0001406 invers | .2535571 .0848826 2.99 0.003 .0871904 .4199239 gfunci | -.1317689 .0903935 -1.46 0.145 -.3089368 .045399 gdpgr | -3.53952 1.881637 -1.88 0.060 -7.22746 .1484196 _cons | -1.526382 .2172254 -7.03 0.000 -1.952136 -1.100628 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .36474429 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0059466 .00095 6.23 0.000 .004075 .007818 72.5663 deficit | -.0009488 .00046 -2.08 0.038 -.001844 -.000053 -1.37383 invers | .0952838 .03183 2.99 0.003 .03289 .157677 -.015445 gfunci | -.0495172 .03399 -1.46 0.145 -.116132 .017098 -.905798 gdpgr | -1.33011 .70723 -1.88 0.060 -2.71626 .056038 .024479 ------------------------------------------------------------------------------ . . . *** Non-Swing *** . . ** Column 3 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing1a==0, robust Iteration 0: log pseudolikelihood = -332.55107 Iteration 1: log pseudolikelihood = -321.83971 Iteration 2: log pseudolikelihood = -321.75885 Iteration 3: log pseudolikelihood = -321.75883 Probit regression Number of obs = 809 Wald chi2(5) = 22.60 Prob > chi2 = 0.0004 Log pseudolikelihood = -321.75883 Pseudo R2 = 0.0325 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0102393 .0030809 3.32 0.001 .0042008 .0162779 deficit | -.0000259 .0005859 -0.04 0.965 -.0011742 .0011224 invers | -.0561261 .0850902 -0.66 0.510 -.2228999 .1106477 gfunci | .1270715 .1116456 1.14 0.255 -.0917498 .3458929 gdpgr | 5.524972 1.979454 2.79 0.005 1.645313 9.404632 _cons | .2002129 .2933066 0.68 0.495 -.3746575 .7750832 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .86433698 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0022306 .00067 3.34 0.001 .00092 .003541 85.6395 deficit | -5.65e-06 .00013 -0.04 0.965 -.000256 .000245 .426939 invers | -.012227 .01853 -0.66 0.509 -.048545 .024091 .004074 gfunci | .0276824 .02437 1.14 0.256 -.020083 .075448 -.902463 gdpgr | 1.203609 .42735 2.82 0.005 .366025 2.04119 .024946 ------------------------------------------------------------------------------ . . ***** SWING2C: not included ***** . . *** Swing *** . ** 1 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing2c==1, robust Iteration 0: log pseudolikelihood = -497.00318 Iteration 1: log pseudolikelihood = -475.992 Iteration 2: log pseudolikelihood = -475.8176 Iteration 3: log pseudolikelihood = -475.81723 Probit regression Number of obs = 721 Wald chi2(5) = 41.28 Prob > chi2 = 0.0000 Log pseudolikelihood = -475.81723 Pseudo R2 = 0.0426 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0140329 .0023593 5.95 0.000 .0094087 .0186572 deficit | -.0027404 .0012719 -2.15 0.031 -.0052333 -.0002474 invers | .0161646 .0652993 0.25 0.804 -.1118196 .1441488 gfunci | -.1047292 .0748562 -1.40 0.162 -.2514447 .0419863 gdpgr | .7440496 1.751714 0.42 0.671 -2.689247 4.177346 _cons | -1.25049 .1960701 -6.38 0.000 -1.63478 -.8661993 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .45349781 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0055603 .00093 5.95 0.000 .003729 .007392 72.5284 deficit | -.0010858 .0005 -2.16 0.031 -.002073 -.000099 -.832079 invers | .0064049 .02587 0.25 0.804 -.044306 .057116 -.066629 gfunci | -.0414967 .02966 -1.40 0.162 -.099623 .016629 -.922094 gdpgr | .294814 .69403 0.42 0.671 -1.06547 1.65509 .024327 ------------------------------------------------------------------------------ . . *** Non-Swing *** . ** 2** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing2c==0, robust Iteration 0: log pseudolikelihood = -348.52773 Iteration 1: log pseudolikelihood = -335.89428 Iteration 2: log pseudolikelihood = -335.82916 Iteration 3: log pseudolikelihood = -335.82911 Probit regression Number of obs = 740 Wald chi2(5) = 28.26 Prob > chi2 = 0.0000 Log pseudolikelihood = -335.82911 Pseudo R2 = 0.0364 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .014539 .0032354 4.49 0.000 .0081977 .0208803 deficit | .0000581 .0007108 0.08 0.935 -.0013349 .0014512 invers | .0603567 .0765851 0.79 0.431 -.0897474 .2104607 gfunci | .173786 .1169238 1.49 0.137 -.0553804 .4029525 gdpgr | -.9578322 1.845022 -0.52 0.604 -4.57401 2.658345 _cons | -.1403331 .3125096 -0.45 0.653 -.7528407 .4721745 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .82851341 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0036997 .00082 4.51 0.000 .002092 .005308 86.8954 deficit | .0000148 .00018 0.08 0.935 -.00034 .000369 .067005 invers | .0153588 .01948 0.79 0.430 -.022814 .053531 .055764 gfunci | .0442229 .0298 1.48 0.138 -.01418 .102626 -.886274 gdpgr | -.2437373 .46934 -0.52 0.604 -1.16362 .676149 .025137 ------------------------------------------------------------------------------ . . ***** SWING1A_3: not included ***** . . *** Swing *** . ** 1 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing1a_3==1, robust Iteration 0: log pseudolikelihood = -415.75775 Iteration 1: log pseudolikelihood = -389.61751 Iteration 2: log pseudolikelihood = -389.40433 Iteration 3: log pseudolikelihood = -389.40387 Probit regression Number of obs = 624 Wald chi2(5) = 52.78 Prob > chi2 = 0.0000 Log pseudolikelihood = -389.40387 Pseudo R2 = 0.0634 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0163145 .0025909 6.30 0.000 .0112365 .0213925 deficit | -.0027634 .0012741 -2.17 0.030 -.0052606 -.0002662 invers | .2485657 .0880779 2.82 0.005 .0759362 .4211953 gfunci | -.1303722 .0929241 -1.40 0.161 -.3125001 .0517558 gdpgr | -3.750629 1.947058 -1.93 0.054 -7.566792 .0655339 _cons | -1.519793 .2192087 -6.93 0.000 -1.949434 -1.090151 ------------------------------------------------------------------------------ . mfx compute Marginal effects after probit y = Pr(win_coin) (predict) = .37685679 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .006196 .00098 6.29 0.000 .004267 .008125 72.3928 deficit | -.0010495 .00048 -2.17 0.030 -.001997 -.000102 -1.37924 invers | .0944009 .03339 2.83 0.005 .028954 .159848 -.019174 gfunci | -.0495131 .03531 -1.40 0.161 -.118717 .019691 -.908656 gdpgr | -1.424423 .73961 -1.93 0.054 -2.87403 .025187 .024666 ------------------------------------------------------------------------------ . . . *** Non-Swing *** . . ** 2 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing1a_3==0, robust Iteration 0: log pseudolikelihood = -320.03912 Iteration 1: log pseudolikelihood = -308.83089 Iteration 2: log pseudolikelihood = -308.73932 Iteration 3: log pseudolikelihood = -308.7393 Probit regression Number of obs = 786 Wald chi2(5) = 23.39 Prob > chi2 = 0.0003 Log pseudolikelihood = -308.7393 Pseudo R2 = 0.0353 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0108574 .0031455 3.45 0.001 .0046924 .0170224 deficit | -7.11e-06 .0006157 -0.01 0.991 -.0012139 .0011997 invers | -.0335426 .0818318 -0.41 0.682 -.19393 .1268448 gfunci | .1157252 .1115795 1.04 0.300 -.1029666 .334417 gdpgr | 5.831706 2.001239 2.91 0.004 1.909349 9.754063 _cons | .1387775 .2981522 0.47 0.642 -.4455901 .723145 ------------------------------------------------------------------------------ . mfx compute Marginal effects after probit y = Pr(win_coin) (predict) = .86711964 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0023318 .00067 3.47 0.001 .001015 .003649 85.9245 deficit | -1.53e-06 .00013 -0.01 0.991 -.000261 .000258 .395957 invers | -.0072039 .01758 -0.41 0.682 -.041655 .027247 .007926 gfunci | .0248543 .02403 1.03 0.301 -.022246 .071954 -.900183 gdpgr | 1.252476 .42564 2.94 0.003 .418234 2.08672 .024972 ------------------------------------------------------------------------------ . . ***** SWING2C_3: not included ***** . . *** Swing *** . ** 1 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing2c_3==1, robust Iteration 0: log pseudolikelihood = -488.83688 Iteration 1: log pseudolikelihood = -466.96914 Iteration 2: log pseudolikelihood = -466.77111 Iteration 3: log pseudolikelihood = -466.77063 Probit regression Number of obs = 708 Wald chi2(5) = 42.77 Prob > chi2 = 0.0000 Log pseudolikelihood = -466.77063 Pseudo R2 = 0.0451 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0145073 .0023902 6.07 0.000 .0098225 .019192 deficit | -.0028558 .0012943 -2.21 0.027 -.0053925 -.000319 invers | .0188747 .065945 0.29 0.775 -.1103752 .1481246 gfunci | -.1068402 .0760121 -1.41 0.160 -.2558211 .0421406 gdpgr | .8386173 1.780231 0.47 0.638 -2.650572 4.327806 _cons | -1.271628 .1982893 -6.41 0.000 -1.660268 -.8829885 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .46063758 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0057594 .00095 6.07 0.000 .0039 .007619 72.5802 deficit | -.0011337 .00051 -2.21 0.027 -.00214 -.000127 -.779648 invers | .0074932 .02618 0.29 0.775 -.043819 .058805 -.065058 gfunci | -.0424154 .03017 -1.41 0.160 -.101556 .016725 -.9208 gdpgr | .3329301 .7067 0.47 0.638 -1.05218 1.71804 .024424 ------------------------------------------------------------------------------ . . *** Non-Swing *** . ** 2 ** . probit win_coin pcv_lag deficit invers gfunci gdpgr if coin2==1 & swing2c_3==0, robust Iteration 0: log pseudolikelihood = -313.05449 Iteration 1: log pseudolikelihood = -300.46366 Iteration 2: log pseudolikelihood = -300.3617 Iteration 3: log pseudolikelihood = -300.36147 Iteration 4: log pseudolikelihood = -300.36147 Probit regression Number of obs = 702 Wald chi2(5) = 28.12 Prob > chi2 = 0.0000 Log pseudolikelihood = -300.36147 Pseudo R2 = 0.0405 ------------------------------------------------------------------------------ | Robust win_coin | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- pcv_lag | .0144883 .0034202 4.24 0.000 .0077847 .0211918 deficit | .0000714 .0008553 0.08 0.933 -.0016049 .0017477 invers | .0874998 .0778651 1.12 0.261 -.0651131 .2401126 gfunci | .2142033 .120043 1.78 0.074 -.0210767 .4494832 gdpgr | -1.308021 1.91773 -0.68 0.495 -5.066702 2.45066 _cons | -.0302613 .3270113 -0.09 0.926 -.6711917 .610669 ------------------------------------------------------------------------------ . mfx Marginal effects after probit y = Pr(win_coin) (predict) = .84551145 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- pcv_lag | .0034448 .00081 4.26 0.000 .00186 .00503 87.3546 deficit | .000017 .0002 0.08 0.933 -.000382 .000416 .003658 invers | .0208046 .01849 1.13 0.261 -.015439 .057049 .057444 gfunci | .0509307 .02859 1.78 0.075 -.005099 .10696 -.886921 gdpgr | -.3110054 .45577 -0.68 0.495 -1.2043 .58229 .025252 ------------------------------------------------------------------------------ . . . log close log: C:\Documents and Settings\Angela Fonseca\Mis documentos\聲gela\Ciclos\Composi > tion Mar2008\outputs\voting_07_08_todofin_wa.txt log type: text closed on: 16 Jul 2008, 13:10:46 ------------------------------------------------------------------------------------------