-------------------------------------------------------------------------------------- name: log: /Users/justinmccrary/Dropbox/Chalfin_McCrary/final/master/table3.log log type: text opened on: 30 May 2013, 14:06:01 . . clear all . use base (Chalfin-McCrary data, 242 cities, 1960-2010) . d Contains data from base.dta obs: 10,589 Chalfin-McCrary data, 242 cities, 1960-2010 vars: 23 30 May 2013 14:05 size: 1,016,544 -------------------------------------------------------------------------------------- storage display value variable name type format label variable label -------------------------------------------------------------------------------------- STATE str2 %9s state abbreviation numcities byte %9.0g # cities in given (STATE,year) pair ORI7 str7 %9s FBI ORI code, unique city identifier cityid int %9.0g roughly sequential city numeric identifier year int %9.0g 4 digit year stateyear int %9.0g group(STATE year) Y1 float %9.0g growth rate, murder Y2 float %9.0g growth rate, rape Y3 float %9.0g growth rate, robbery Y4 float %9.0g growth rate, aggravated assault Y5 float %9.0g growth rate, burglary Y6 float %9.0g growth rate, larceny Y7 float %9.0g growth rate, motor vehicle theft Y8 float %9.0g growth rate, sum of violent crimes Y9 float %9.0g growth rate, sum of property crimes Y10 double %10.0g growth rate, cost-weighted sum of violent Y11 double %10.0g growth rate, cost-weighted sum of property Y12 double %10.0g growth rate, cost-weighted sum, all crimes S float %9.0g growth rate, sworn officers (UCR) Z float %9.0g growth rate, sworn officers (ASG) C1 float %9.0g growth rate, city population (UCR) C2 float %9.0g growth rate, city population (ASG) W float %9.0g 2010 city population (UCR) -------------------------------------------------------------------------------------- Sorted by: cityid year . . //Column 1 . areg Y12 S C1 C2 [aw=W], absorb(year) robust (sum of wgt is 3.2408e+09) Linear regression, absorbing indicators Number of obs = 10589 F( 3, 10537) = 10.90 Prob > F = 0.0000 R-squared = 0.0733 Adj R-squared = 0.0688 Root MSE = 0.2611 ------------------------------------------------------------------------------ | Robust Y12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- S | -.2125238 .0537142 -3.96 0.000 -.3178137 -.1072339 C1 | 1.059309 .4124542 2.57 0.010 .2508206 1.867797 C2 | -.1514903 .3334545 -0.45 0.650 -.8051241 .5021436 _cons | .0156198 .0029688 5.26 0.000 .0098003 .0214392 -------------+---------------------------------------------------------------- year | absorbed (49 categories) . . //Column 2 . areg Y12 S C1 C2 [aw=W], absorb(stateyear) robust (sum of wgt is 3.2408e+09) Linear regression, absorbing indicators Number of obs = 10589 F( 3, 8502) = 5.13 Prob > F = 0.0015 R-squared = 0.2572 Adj R-squared = 0.0749 Root MSE = 0.2603 ------------------------------------------------------------------------------ | Robust Y12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- S | -.1436565 .0713888 -2.01 0.044 -.2835959 -.0037171 C1 | 1.029104 .6628764 1.55 0.121 -.2702953 2.328502 C2 | -.0275893 .4826253 -0.06 0.954 -.9736522 .9184735 _cons | .0138222 .0031408 4.40 0.000 .0076654 .0199789 -------------+---------------------------------------------------------------- stateyear | absorbed (2084 categories) . . //Column 3 . areg Y12 Z C1 C2 [aw=W], absorb(year) robust (sum of wgt is 3.2408e+09) Linear regression, absorbing indicators Number of obs = 10589 F( 3, 10537) = 9.10 Prob > F = 0.0000 R-squared = 0.0725 Adj R-squared = 0.0680 Root MSE = 0.2612 ------------------------------------------------------------------------------ | Robust Y12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- Z | -.1115387 .0335331 -3.33 0.001 -.1772699 -.0458075 C1 | .9905286 .4120622 2.40 0.016 .1828089 1.798248 C2 | -.1582338 .3339652 -0.47 0.636 -.8128688 .4964013 _cons | .0146505 .0029024 5.05 0.000 .0089612 .0203398 -------------+---------------------------------------------------------------- year | absorbed (49 categories) . . //Column 4 . areg Y12 Z C1 C2 [aw=W], absorb(stateyear) robust (sum of wgt is 3.2408e+09) Linear regression, absorbing indicators Number of obs = 10589 F( 3, 8502) = 5.48 Prob > F = 0.0009 R-squared = 0.2572 Adj R-squared = 0.0750 Root MSE = 0.2603 ------------------------------------------------------------------------------ | Robust Y12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- Z | -.0985641 .0407758 -2.42 0.016 -.1784946 -.0186336 C1 | 1.008809 .6607152 1.53 0.127 -.2863534 2.303971 C2 | -.0257543 .4804493 -0.05 0.957 -.9675517 .9160432 _cons | .0132958 .0030746 4.32 0.000 .0072689 .0193228 -------------+---------------------------------------------------------------- stateyear | absorbed (2084 categories) . . //Column 5 . foreach var in Y12 S Z { 2. quietly areg `var' C1 C2 [aw=W], absorb(year) 3. predict double r`var', resid 4. } . ivreg rY12 (rS=rZ) [aw=W], robust (sum of wgt is 3.2408e+09) Instrumental variables (2SLS) regression Number of obs = 10589 F( 1, 10587) = 10.74 Prob > F = 0.0010 R-squared = . Root MSE = .26136 ------------------------------------------------------------------------------ | Robust rY12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- rS | -.6047148 .1844811 -3.28 0.001 -.9663324 -.2430971 _cons | -3.21e-19 .0028036 -0.00 1.000 -.0054956 .0054956 ------------------------------------------------------------------------------ Instrumented: rS Instruments: rZ ------------------------------------------------------------------------------ . capture drop rY12 rS rZ . . //Column 6 . //Note: these standards errors are a bit too small . //because they ignore the dof adjustment for absorbing . //so many effects. The standard errors in the paper . //reflect the correct dof adjustment. . foreach var in Y12 S Z { 2. quietly areg `var' C1 C2 [aw=W], absorb(stateyear) 3. predict double r`var', resid 4. } . ivreg rY12 (rS=rZ) [aw=W], robust (sum of wgt is 3.2408e+09) Instrumental variables (2SLS) regression Number of obs = 10589 F( 1, 10587) = 7.10 Prob > F = 0.0077 R-squared = . Root MSE = .23406 ------------------------------------------------------------------------------ | Robust rY12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- rS | -.6140692 .2305345 -2.66 0.008 -1.06596 -.1621783 _cons | 2.32e-19 .0020622 0.00 1.000 -.0040422 .0040422 ------------------------------------------------------------------------------ Instrumented: rS Instruments: rZ ------------------------------------------------------------------------------ . capture drop rY12 rS rZ . . //Column 7 . foreach var in Y12 S Z { 2. quietly areg `var' C1 C2 [aw=W], absorb(year) 3. predict double r`var', resid 4. } . ivreg rY12 (rZ=rS) [aw=W], robust (sum of wgt is 3.2408e+09) Instrumental variables (2SLS) regression Number of obs = 10589 F( 1, 10587) = 15.69 Prob > F = 0.0001 R-squared = . Root MSE = .26303 ------------------------------------------------------------------------------ | Robust rY12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- rZ | -.5831252 .1472314 -3.96 0.000 -.8717265 -.2945239 _cons | 1.29e-18 .0027751 0.00 1.000 -.0054397 .0054397 ------------------------------------------------------------------------------ Instrumented: rZ Instruments: rS ------------------------------------------------------------------------------ . capture drop rY12 rZ rS . . //Column 8 . //Note: these standards errors are a bit too small . //(see dof discussion above) . foreach var in Y12 S Z { 2. quietly areg `var' C1 C2 [aw=W], absorb(stateyear) 3. predict double r`var', resid 4. } . ivreg rY12 (rZ=rS) [aw=W], robust (sum of wgt is 3.2408e+09) Instrumental variables (2SLS) regression Number of obs = 10589 F( 1, 10587) = 5.19 Prob > F = 0.0227 R-squared = . Root MSE = .234 ------------------------------------------------------------------------------ | Robust rY12 | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- rZ | -.402984 .1768752 -2.28 0.023 -.7496926 -.0562753 _cons | 9.02e-19 .0020643 0.00 1.000 -.0040463 .0040463 ------------------------------------------------------------------------------ Instrumented: rZ Instruments: rS ------------------------------------------------------------------------------ . //Now I do not drop these because I am going to use them for . //column (9) . //capture drop rY12 rZ rS . . //Column 9 . //Reproduces forward IV . //gmm ( F: rY12 - {b}*rS ) [aw=W], instruments(F: rZ) . //Reproduces reflected IV . //gmm ( R: rY12 - {b}*rZ ) [aw=W], instruments(R: rS) . //Produces the pooled estimate . //Note: these standards errors are a bit too small . //(see dof discussion above) . gmm ( F: rY12 - {b}*rS ) ( R: rY12 - {b}*rZ ) [aw=W], /// > instruments(F: rZ) instruments(R: rS) /// > winitial(identity) twostep (sum of wgt is 3.241e+09) Step 1 Iteration 0: GMM criterion Q(b) = 2.044e-07 Iteration 1: GMM criterion Q(b) = 8.440e-09 Iteration 2: GMM criterion Q(b) = 8.440e-09 Step 2 Iteration 0: GMM criterion Q(b) = .00007261 Iteration 1: GMM criterion Q(b) = .00006781 Iteration 2: GMM criterion Q(b) = .00006781 GMM estimation Number of parameters = 1 Number of moments = 4 Initial weight matrix: Identity Number of obs = 10589 GMM weight matrix: Robust ------------------------------------------------------------------------------ | Robust | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- /b | -.4729982 .1570909 -3.01 0.003 -.7808906 -.1651057 ------------------------------------------------------------------------------ Instruments for equation 1: rZ _cons Instruments for equation 2: rS _cons . . . log close name: log: /Users/justinmccrary/Dropbox/Chalfin_McCrary/final/master/table3.log log type: text closed on: 30 May 2013, 14:06:04 --------------------------------------------------------------------------------------