其他
多期 DID 经典论文-视频+操作应用：银行管制与收入差距
多期 DID 经典论文-视频+操作应用：银行管制与收入差距
来源：Beck T, Levine R, Levkov A. Big bad banks? The winners and losers from bank deregulation in the United States[J]. The Journal of Finance, 2010, 65(5): 1637-1667
简介
20世纪，美国学者认为，大型银行的成立会减少贫困人口的经济机会，从而拉大了收入差距，因此贫困人口经常与银行设立分支机构做斗争。
我们评估了银行放松管制对美国收入分配的影响州。从20世纪70年代到90年代，大多数州都取消了对州内分支机构，加剧了银行竞争，提高了银行绩效。利用部门放松管制时机的跨州、跨时间差异，我们发现放松管制通过提高收入分配较低部分的收入，而对中位数以上的收入几乎没有影响，从而实质性地收紧了收入分配。放松对银行的管制，提高了非技术工人的相对工资率和工作时间，从而收紧了收入分配。
2、变量以及模型介绍
我们使用DID模型来评估之间的关系 分行放松管制与收入分配，基于以下回归分析 Yst = α + βDst + δXst + As + Bt + εst,s = 1,..., 49; t = 1976,..., 2006（1)
式(1)中，Yst是t年s州收入分配的测度， As和Bt是截面个体向量和年份虚拟变量反映个体和年份固定效应， Xst是一组时变的控制变量， εst 是误差项。 我们感兴趣的变量是Dst，它是一个虚拟变量，等于s在国家放松管制后的几年里是1，在其他方面是0。 系数β反映分支机构放松管制对收入分配的影响。β值为正且显著，表明放松管制具有正向效应在收入不平等程度上，而β值为负且显著 放松管制降低了收入不平等。
我们包括特定年份的虚拟变量来控制全国性的冲击和趋势随着时间的推移会影响收入分配，比如随着经济周期的变化，国家法规和法律的变化，长期趋势在收入分配方面，以及女性劳动参与率的变化。
我们用特定状态的虚拟变量来控制时不变的、不可观察的 各州的特征决定了各州之间的收入分配。我们估计式(1)允许状态级的误差聚类，即允许对于误差项在状态内随时间变化的相关性。
3、DID模型结果
回归结果如表2所示， 放松成立分行规制至少在5%的显著性水平上降低了收入差距。 例如，从Panel A部分的逻辑---基尼系数来看（第1列）， 放松规制使得逻辑基尼系数显著下降了3.9%；
use "C:\Users\Metrics\Desktop\macro_workfile.dta", clear
(Data for 'Big Bad Banks?' paper.)
. do "C:\Users\Metrics\AppData\Local\Temp\STD4ae4_000000.tmp"
. #delimit;
delimiter now ;
. label var _intra "Bank deregulation";
. tsset statefip wrkyr;
Panel variable: statefip (strongly balanced)
Time variable: wrkyr, 1976 to 2006
Delta: 1 unit
. //声明数据为statefip层面的时间序列数据的集合
>
> tabulate wrkyr, gen(wrkyr_dumm);
Year | Freq. Percent Cum.
------------+-----------------------------------
1976 | 49 3.23 3.23
1977 | 49 3.23 6.45
1978 | 49 3.23 9.68
1979 | 49 3.23 12.90
1980 | 49 3.23 16.13
1981 | 49 3.23 19.35
1982 | 49 3.23 22.58
1983 | 49 3.23 25.81
1984 | 49 3.23 29.03
1985 | 49 3.23 32.26
1986 | 49 3.23 35.48
1987 | 49 3.23 38.71
1988 | 49 3.23 41.94
1989 | 49 3.23 45.16
1990 | 49 3.23 48.39
1991 | 49 3.23 51.61
1992 | 49 3.23 54.84
1993 | 49 3.23 58.06
1994 | 49 3.23 61.29
1995 | 49 3.23 64.52
1996 | 49 3.23 67.74
1997 | 49 3.23 70.97
1998 | 49 3.23 74.19
1999 | 49 3.23 77.42
2000 | 49 3.23 80.65
2001 | 49 3.23 83.87
2002 | 49 3.23 87.10
2003 | 49 3.23 90.32
2004 | 49 3.23 93.55
2005 | 49 3.23 96.77
2006 | 49 3.23 100.00
------------+-----------------------------------
Total | 1,519 100.00
. tabulate statefip, gen(state_dumm);
State FIPS |
code | Freq. Percent Cum.
------------+-----------------------------------
1 | 31 2.04 2.04
2 | 31 2.04 4.08
4 | 31 2.04 6.12
5 | 31 2.04 8.16
6 | 31 2.04 10.20
8 | 31 2.04 12.24
9 | 31 2.04 14.29
11 | 31 2.04 16.33
12 | 31 2.04 18.37
13 | 31 2.04 20.41
15 | 31 2.04 22.45
16 | 31 2.04 24.49
17 | 31 2.04 26.53
18 | 31 2.04 28.57
19 | 31 2.04 30.61
20 | 31 2.04 32.65
21 | 31 2.04 34.69
22 | 31 2.04 36.73
23 | 31 2.04 38.78
24 | 31 2.04 40.82
25 | 31 2.04 42.86
26 | 31 2.04 44.90
27 | 31 2.04 46.94
28 | 31 2.04 48.98
29 | 31 2.04 51.02
30 | 31 2.04 53.06
31 | 31 2.04 55.10
32 | 31 2.04 57.14
33 | 31 2.04 59.18
34 | 31 2.04 61.22
35 | 31 2.04 63.27
36 | 31 2.04 65.31
37 | 31 2.04 67.35
38 | 31 2.04 69.39
39 | 31 2.04 71.43
40 | 31 2.04 73.47
41 | 31 2.04 75.51
42 | 31 2.04 77.55
44 | 31 2.04 79.59
45 | 31 2.04 81.63
47 | 31 2.04 83.67
48 | 31 2.04 85.71
49 | 31 2.04 87.76
50 | 31 2.04 89.80
51 | 31 2.04 91.84
53 | 31 2.04 93.88
54 | 31 2.04 95.92
55 | 31 2.04 97.96
56 | 31 2.04 100.00
------------+-----------------------------------
Total | 1,519 100.00
. replace p10 = 1 if p10==0;
(1 real change made)
. generate logistic_gini = log(gini/(1-gini));
. generate log_gini = log(gini);
. generate log_theil = log(theil);
. generate log_9010 = log(p90)-log(p10);
. generate log_7525 = log(p75)-log(p25);
. local Xs gsp_pc_growth prop_blacks prop_dropouts prop_female_headed unemploymentrate;
. xtreg logistic_gini _intra wrkyr_dumm*, fe i(statefip) robust cluster(statefip);
note: wrkyr_dumm31 omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 1,519
Group variable: statefip Number of groups = 49
R-squared: Obs per group:
Within = 0.3552 min = 31
Between = 0.0074 avg = 31.0
Overall = 0.2537 max = 31
F(31,48) = 42.65
corr(u_i, Xb) = -0.0120 Prob > F = 0.0000
(Std. err. adjusted for 49 clusters in statefip)
------------------------------------------------------------------------------
| Robust
logistic_g~i | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
_intra | -.0388911 .0131655 -2.95 0.005 -.0653622 -.01242
wrkyr_dumm1 | -.1584755 .0229805 -6.90 0.000 -.2046809 -.1122701
wrkyr_dumm2 | -.1543535 .0212529 -7.26 0.000 -.1970853 -.1116216
wrkyr_dumm3 | -.0746086 .0203929 -3.66 0.001 -.1156114 -.0336059
wrkyr_dumm4 | -.0287208 .020239 -1.42 0.162 -.0694139 .0119724
wrkyr_dumm5 | -.0551676 .0194951 -2.83 0.007 -.094365 -.0159701
wrkyr_dumm6 | -.0312674 .0209108 -1.50 0.141 -.0733114 .0107765
wrkyr_dumm7 | .1227202 .0219675 5.59 0.000 .0785515 .1668889
wrkyr_dumm8 | -.0089456 .0211688 -0.42 0.674 -.0515083 .0336171
wrkyr_dumm9 | -.0204623 .0190874 -1.07 0.289 -.05884 .0179155
wrkyr_dumm10 | -.0274459 .0195157 -1.41 0.166 -.0666848 .011793
wrkyr_dumm11 | -.0207858 .0203435 -1.02 0.312 -.061689 .0201175
wrkyr_dumm12 | -.0461078 .0197006 -2.34 0.023 -.0857185 -.0064972
wrkyr_dumm13 | -.0625037 .0200984 -3.11 0.003 -.1029143 -.0220931
wrkyr_dumm14 | -.0502186 .0184222 -2.73 0.009 -.0872589 -.0131782
wrkyr_dumm15 | -.0693075 .0170128 -4.07 0.000 -.1035141 -.0351009
wrkyr_dumm16 | -.0674973 .0148464 -4.55 0.000 -.097348 -.0376467
wrkyr_dumm17 | -.0582625 .0161039 -3.62 0.001 -.0906415 -.0258835
wrkyr_dumm18 | -.0476189 .0147039 -3.24 0.002 -.0771829 -.0180548
wrkyr_dumm19 | -.0512824 .0147865 -3.47 0.001 -.0810127 -.0215521
wrkyr_dumm20 | .0065045 .0149502 0.44 0.665 -.0235549 .036564
wrkyr_dumm21 | -.0262409 .013878 -1.89 0.065 -.0541446 .0016627
wrkyr_dumm22 | -.0054051 .0145519 -0.37 0.712 -.0346637 .0238535
wrkyr_dumm23 | -.0327607 .0143069 -2.29 0.026 -.0615267 -.0039947
wrkyr_dumm24 | -.043271 .0142403 -3.04 0.004 -.071903 -.014639
wrkyr_dumm25 | .0030401 .0149872 0.20 0.840 -.0270937 .0331738
wrkyr_dumm26 | -.0154153 .0118423 -1.30 0.199 -.0392259 .0083953
wrkyr_dumm27 | -.0163578 .0109749 -1.49 0.143 -.0384242 .0057087
wrkyr_dumm28 | -.000445 .0105167 -0.04 0.966 -.0215903 .0207003
wrkyr_dumm29 | -.0155752 .00943 -1.65 0.105 -.0345356 .0033852
wrkyr_dumm30 | .0187134 .0103208 1.81 0.076 -.002038 .0394648
wrkyr_dumm31 | 0 (omitted)
_cons | -.2181624 .0194521 -11.22 0.000 -.2572734 -.1790514
-------------+----------------------------------------------------------------
sigma_u | .0511354
sigma_e | .065767
rho | .37676956 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. estimates store m1, title(Logistic Gini);
. xtreg log_gini _intra wrkyr_dumm*, fe i(statefip) robust cluster(statefip);
note: wrkyr_dumm31 omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 1,519
Group variable: statefip Number of groups = 49
R-squared: Obs per group:
Within = 0.3490 min = 31
Between = 0.0075 avg = 31.0
Overall = 0.2490 max = 31
F(31,48) = 45.99
corr(u_i, Xb) = -0.0117 Prob > F = 0.0000
(Std. err. adjusted for 49 clusters in statefip)
------------------------------------------------------------------------------
| Robust
log_gini | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
_intra | -.0219874 .0075254 -2.92 0.005 -.0371183 -.0068565
wrkyr_dumm1 | -.0911798 .0133276 -6.84 0.000 -.1179768 -.0643828
wrkyr_dumm2 | -.0885007 .0121072 -7.31 0.000 -.1128439 -.0641576
wrkyr_dumm3 | -.0416721 .0115299 -3.61 0.001 -.0648544 -.0184897
wrkyr_dumm4 | -.0156937 .0115146 -1.36 0.179 -.0388454 .007458
wrkyr_dumm5 | -.0306987 .0110288 -2.78 0.008 -.0528735 -.0085239
wrkyr_dumm6 | -.0171464 .0117795 -1.46 0.152 -.0408306 .0065378
wrkyr_dumm7 | .0664589 .0120134 5.53 0.000 .0423044 .0906134
wrkyr_dumm8 | -.0049492 .0119127 -0.42 0.680 -.0289014 .019003
wrkyr_dumm9 | -.0113038 .0107746 -1.05 0.299 -.0329675 .0103599
wrkyr_dumm10 | -.0150011 .0110063 -1.36 0.179 -.0371308 .0071286
wrkyr_dumm11 | -.0114927 .0114851 -1.00 0.322 -.034585 .0115997
wrkyr_dumm12 | -.026008 .0111867 -2.32 0.024 -.0485004 -.0035156
wrkyr_dumm13 | -.0355916 .0114459 -3.11 0.003 -.0586051 -.0125781
wrkyr_dumm14 | -.0283805 .0104317 -2.72 0.009 -.0493547 -.0074062
wrkyr_dumm15 | -.0393801 .0096183 -4.09 0.000 -.058719 -.0200412
wrkyr_dumm16 | -.0381164 .0084287 -4.52 0.000 -.0550634 -.0211695
wrkyr_dumm17 | -.0330203 .0090847 -3.63 0.001 -.0512864 -.0147542
wrkyr_dumm18 | -.0269633 .0082764 -3.26 0.002 -.0436042 -.0103224
wrkyr_dumm19 | -.0289675 .00835 -3.47 0.001 -.0457563 -.0121787
wrkyr_dumm20 | .0037781 .0083932 0.45 0.655 -.0130975 .0206537
wrkyr_dumm21 | -.0149576 .0078631 -1.90 0.063 -.0307674 .0008523
wrkyr_dumm22 | -.0031038 .0081802 -0.38 0.706 -.0195513 .0133437
wrkyr_dumm23 | -.0183246 .0081121 -2.26 0.028 -.034635 -.0020142
wrkyr_dumm24 | -.0244602 .0081826 -2.99 0.004 -.0409123 -.008008
wrkyr_dumm25 | .0017328 .0085132 0.20 0.840 -.0153841 .0188497
wrkyr_dumm26 | -.0085793 .0066574 -1.29 0.204 -.0219649 .0048063
wrkyr_dumm27 | -.0089367 .0061134 -1.46 0.150 -.0212285 .0033551
wrkyr_dumm28 | -.0000603 .0059824 -0.01 0.992 -.0120888 .0119682
wrkyr_dumm29 | -.0082868 .0053009 -1.56 0.125 -.0189449 .0023714
wrkyr_dumm30 | .0109113 .0057286 1.90 0.063 -.0006068 .0224294
wrkyr_dumm31 | 0 (omitted)
_cons | -.8090041 .0110005 -73.54 0.000 -.8311221 -.7868861
-------------+----------------------------------------------------------------
sigma_u | .02915639
sigma_e | .03757708
rho | .37579374 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. estimates store m2, title(Log Gini);
. xtreg log_theil _intra wrkyr_dumm*, fe i(statefip) robust cluster(statefip);
note: wrkyr_dumm31 omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 1,519
Group variable: statefip Number of groups = 49
R-squared: Obs per group:
Within = 0.4318 min = 31
Between = 0.0176 avg = 31.0
Overall = 0.3376 max = 31
F(31,48) = 78.16
corr(u_i, Xb) = -0.0018 Prob > F = 0.0000
(Std. err. adjusted for 49 clusters in statefip)
------------------------------------------------------------------------------
| Robust
log_theil | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
_intra | -.0407613 .0158468 -2.57 0.013 -.0726235 -.0088991
wrkyr_dumm1 | -.2417574 .0282783 -8.55 0.000 -.2986147 -.1849
wrkyr_dumm2 | -.2390445 .0258751 -9.24 0.000 -.2910699 -.1870192
wrkyr_dumm3 | -.1361627 .0255308 -5.33 0.000 -.1874958 -.0848295
wrkyr_dumm4 | -.0745215 .0254613 -2.93 0.005 -.1257148 -.0233282
wrkyr_dumm5 | -.1105896 .024418 -4.53 0.000 -.1596853 -.061494
wrkyr_dumm6 | -.0815275 .0259175 -3.15 0.003 -.1336382 -.0294168
wrkyr_dumm7 | .1055342 .0263444 4.01 0.000 .0525651 .1585032
wrkyr_dumm8 | -.0602201 .0261134 -2.31 0.025 -.1127245 -.0077156
wrkyr_dumm9 | -.07262 .023714 -3.06 0.004 -.1203003 -.0249398
wrkyr_dumm10 | -.0799455 .0243985 -3.28 0.002 -.1290021 -.030889
wrkyr_dumm11 | -.0743835 .0252309 -2.95 0.005 -.1251136 -.0236534
wrkyr_dumm12 | -.1062868 .0242935 -4.38 0.000 -.1551322 -.0574415
wrkyr_dumm13 | -.1245003 .024896 -5.00 0.000 -.1745572 -.0744435
wrkyr_dumm14 | -.1113052 .0224464 -4.96 0.000 -.1564367 -.0661738
wrkyr_dumm15 | -.1363589 .0207384 -6.58 0.000 -.1780562 -.0946616
wrkyr_dumm16 | -.1393295 .0184578 -7.55 0.000 -.1764414 -.1022176
wrkyr_dumm17 | -.1286445 .0195312 -6.59 0.000 -.1679146 -.0893744
wrkyr_dumm18 | -.1179508 .0184843 -6.38 0.000 -.1551161 -.0807856
wrkyr_dumm19 | -.121553 .018759 -6.48 0.000 -.1592705 -.0838355
wrkyr_dumm20 | -.0044449 .0198309 -0.22 0.824 -.0443177 .0354279
wrkyr_dumm21 | -.0424344 .017776 -2.39 0.021 -.0781754 -.0066934
wrkyr_dumm22 | -.0104915 .0204919 -0.51 0.611 -.0516933 .0307102
wrkyr_dumm23 | -.0502053 .0188477 -2.66 0.010 -.088101 -.0123095
wrkyr_dumm24 | -.0841599 .0189625 -4.44 0.000 -.1222866 -.0460332
wrkyr_dumm25 | .0260963 .0208109 1.25 0.216 -.0157469 .0679395
wrkyr_dumm26 | -.0128339 .0156976 -0.82 0.418 -.044396 .0187282
wrkyr_dumm27 | -.0227308 .0147873 -1.54 0.131 -.0524627 .0070011
wrkyr_dumm28 | -.0089447 .0149929 -0.60 0.554 -.03909 .0212006
wrkyr_dumm29 | -.0304297 .0134518 -2.26 0.028 -.0574764 -.0033831
wrkyr_dumm30 | .0208075 .01406 1.48 0.145 -.0074621 .0490771
wrkyr_dumm31 | 0 (omitted)
_cons | -1.025716 .024343 -42.14 0.000 -1.074661 -.9767708
-------------+----------------------------------------------------------------
sigma_u | .05685927
sigma_e | .08108198
rho | .32965122 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. estimates store m3, title(Log Theil);
. xtreg log_9010 _intra wrkyr_dumm*, fe i(statefip) robust cluster(statefip);
note: wrkyr_dumm31 omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 1,519
Group variable: statefip Number of groups = 49
R-squared: Obs per group:
Within = 0.7351 min = 31
Between = 0.0219 avg = 31.0
Overall = 0.6772 max = 31
F(31,48) = 56.74
corr(u_i, Xb) = 0.0001 Prob > F = 0.0000
(Std. err. adjusted for 49 clusters in statefip)
------------------------------------------------------------------------------
| Robust
log_9010 | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
_intra | -.1343047 .058278 -2.30 0.026 -.2514803 -.017129
wrkyr_dumm1 | .0887786 .0617709 1.44 0.157 -.0354202 .2129774
wrkyr_dumm2 | .094809 .0618358 1.53 0.132 -.0295202 .2191383
wrkyr_dumm3 | .5081806 .0761709 6.67 0.000 .3550286 .6613325
wrkyr_dumm4 | .9967736 .0836182 11.92 0.000 .8286479 1.164899
wrkyr_dumm5 | .8161614 .0737209 11.07 0.000 .6679355 .9643872
wrkyr_dumm6 | .7631135 .0749965 10.18 0.000 .6123231 .913904
wrkyr_dumm7 | 2.876993 .2208164 13.03 0.000 2.433012 3.320974
wrkyr_dumm8 | .7533073 .0730019 10.32 0.000 .606527 .9000875
wrkyr_dumm9 | .6311457 .0669676 9.42 0.000 .4964983 .765793
wrkyr_dumm10 | .6000623 .0640934 9.36 0.000 .4711938 .7289308
wrkyr_dumm11 | .6183907 .0638836 9.68 0.000 .4899441 .7468373
wrkyr_dumm12 | .4649765 .0585198 7.95 0.000 .3473147 .5826384
wrkyr_dumm13 | .3801245 .054339 7.00 0.000 .2708687 .4893803
wrkyr_dumm14 | .374755 .0449456 8.34 0.000 .2843859 .4651241
wrkyr_dumm15 | .3192522 .0406805 7.85 0.000 .2374586 .4010459
wrkyr_dumm16 | .2826086 .0338632 8.35 0.000 .214522 .3506952
wrkyr_dumm17 | .2942029 .0339505 8.67 0.000 .2259408 .3624651
wrkyr_dumm18 | .2903669 .0339683 8.55 0.000 .222069 .3586648
wrkyr_dumm19 | .2951438 .035523 8.31 0.000 .22372 .3665675
wrkyr_dumm20 | .222676 .0326756 6.81 0.000 .1569774 .2883747
wrkyr_dumm21 | .1278076 .0267669 4.77 0.000 .0739891 .1816261
wrkyr_dumm22 | .1131149 .0317177 3.57 0.001 .0493422 .1768876
wrkyr_dumm23 | .1009421 .0269933 3.74 0.000 .0466685 .1552158
wrkyr_dumm24 | .1031534 .0299788 3.44 0.001 .0428771 .1634297
wrkyr_dumm25 | .0548704 .0211746 2.59 0.013 .0122961 .0974447
wrkyr_dumm26 | .0741486 .0200022 3.71 0.001 .0339315 .1143658
wrkyr_dumm27 | .0824645 .021409 3.85 0.000 .0394188 .1255101
wrkyr_dumm28 | .1235081 .0207046 5.97 0.000 .0818787 .1651374
wrkyr_dumm29 | .0880528 .0158094 5.57 0.000 .0562659 .1198397
wrkyr_dumm30 | .1019528 .0189623 5.38 0.000 .0638265 .1400792
wrkyr_dumm31 | 0 (omitted)
_cons | 2.461968 .0680267 36.19 0.000 2.325191 2.598745
-------------+----------------------------------------------------------------
sigma_u | .18873827
sigma_e | .33587948
rho | .23998124 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. estimates store m4, title(Log 90/10);
. xtreg log_7525 _intra wrkyr_dumm*, fe i(statefip) robust cluster(statefip);
note: wrkyr_dumm31 omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 1,519
Group variable: statefip Number of groups = 49
R-squared: Obs per group:
Within = 0.6015 min = 31
Between = 0.0174 avg = 31.0
Overall = 0.4526 max = 31
F(31,48) = 76.07
corr(u_i, Xb) = -0.0076 Prob > F = 0.0000
(Std. err. adjusted for 49 clusters in statefip)
------------------------------------------------------------------------------
| Robust
log_7525 | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
_intra | -.0768279 .0194977 -3.94 0.000 -.1160306 -.0376251
wrkyr_dumm1 | .0210062 .0275623 0.76 0.450 -.0344115 .0764239
wrkyr_dumm2 | .0207275 .026096 0.79 0.431 -.031742 .073197
wrkyr_dumm3 | .1191752 .0266094 4.48 0.000 .0656735 .1726769
wrkyr_dumm4 | .204538 .0287532 7.11 0.000 .146726 .2623501
wrkyr_dumm5 | .1737033 .0275693 6.30 0.000 .1182714 .2291352
wrkyr_dumm6 | .1842971 .0257969 7.14 0.000 .1324289 .2361653
wrkyr_dumm7 | .4505584 .035072 12.85 0.000 .3800415 .5210753
wrkyr_dumm8 | .2038444 .0284442 7.17 0.000 .1466535 .2610353
wrkyr_dumm9 | .1651283 .0248228 6.65 0.000 .1152188 .2150378
wrkyr_dumm10 | .1577708 .0246889 6.39 0.000 .1081304 .2074111
wrkyr_dumm11 | .1799656 .0275585 6.53 0.000 .1245556 .2353757
wrkyr_dumm12 | .1303745 .0240155 5.43 0.000 .0820882 .1786609
wrkyr_dumm13 | .0936169 .0236676 3.96 0.000 .0460301 .1412037
wrkyr_dumm14 | .1020452 .022636 4.51 0.000 .0565325 .147558
wrkyr_dumm15 | .0831775 .0194381 4.28 0.000 .0440945 .1222604
wrkyr_dumm16 | .0934973 .0175063 5.34 0.000 .0582985 .1286962
wrkyr_dumm17 | .0948733 .0180991 5.24 0.000 .0584827 .131264
wrkyr_dumm18 | .113412 .0137319 8.26 0.000 .0858022 .1410218
wrkyr_dumm19 | .0856912 .0140211 6.11 0.000 .0574999 .1138826
wrkyr_dumm20 | .0678684 .0148139 4.58 0.000 .0380831 .0976538
wrkyr_dumm21 | .0358818 .0165231 2.17 0.035 .0026599 .0691036
wrkyr_dumm22 | .0320199 .0128136 2.50 0.016 .0062564 .0577833
wrkyr_dumm23 | .0133195 .0134367 0.99 0.327 -.0136969 .0403359
wrkyr_dumm24 | .0067095 .0121773 0.55 0.584 -.0177745 .0311935
wrkyr_dumm25 | -.02706 .013558 -2.00 0.052 -.0543202 .0002001
wrkyr_dumm26 | -.0191736 .0118726 -1.61 0.113 -.0430451 .0046978
wrkyr_dumm27 | -.0184561 .0115157 -1.60 0.116 -.04161 .0046979
wrkyr_dumm28 | .0210567 .0121785 1.73 0.090 -.0034298 .0455432
wrkyr_dumm29 | .007493 .0103327 0.73 0.472 -.0132824 .0282684
wrkyr_dumm30 | .0139605 .0106559 1.31 0.196 -.0074646 .0353857
wrkyr_dumm31 | 0 (omitted)
_cons | 1.182593 .0246446 47.99 0.000 1.133042 1.232144
-------------+----------------------------------------------------------------
sigma_u | .08527458
sigma_e | .09459264
rho | .44833345 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. estimates store m5, title(Log 75/25);
.
end of do-file
. do "C:\Users\Metrics\AppData\Local\Temp\STD4ae4_000000.tmp"
. #delimit;
delimiter now ;
. estout m1 m2 m3 m4 m5 , replace
> keep(_intra)
> cells(b(star fmt(3)) se(par) p(fmt(3) par({ }))) stats(r2 N, labels("R-squared" "Observations") fmt(2 0))
> legend label collabel(none)
> prehead("Table II" "The Impact of Deregulation on Income Inequality")
> posthead("Panel A: No controls")
> postfoot("")
> starlevel(* 0.10 ** 0.05 *** 0.01);
Table II
The Impact of Deregulation on Income Inequality
----------------------------------------------------------------------------------------------------
Logistic G~i Log Gini Log Theil Log 90/10 Log 75/25
----------------------------------------------------------------------------------------------------
Panel A: No controls
Bank deregulation -0.039*** -0.022*** -0.041** -0.134** -0.077***
(0.013) (0.008) (0.016) (0.058) (0.019)
{0.005} {0.005} {0.013} {0.026} {0.000}
----------------------------------------------------------------------------------------------------
R-squared 0.36 0.35 0.43 0.74 0.60
Observations 1519 1519 1519 1519 1519
----------------------------------------------------------------------------------------------------
* p<0.10, ** p<0.05, *** p<0.01
.
end of do-file
.
4、动态效应分析
接下来，我们将研究放松管制与不平等之间动态效应，进行平行趋势检验。
结果为：
说明了两个关键点:收入分配方面的创新没有先于放松管制，而放松管制对不平等的影响很快就会显现出来。
如图所示，解除管制的系数在放松管制之前的所有年份，变量都与零无显著差异，在部门放松管制之前没有不平等趋势。接下来,注意放松管制后，不平等立即下降，以至于D+1为负 在5%的水平上是显著的。因此，必须尽快建立银行放松管制与收入分配之间的特殊机制和途径。放松管制对不平等的影响将持续8年左右放松管制后，效果趋于平稳，表明稳定下降基尼系数约为4%