Endogenous Variables: All endogenous variables are the natural logs of the crime rate per 100,000 people
Exogenous
Variables ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -0.0490 -0.0850 -0.0527 -0.0701 -0.0221 0.0269 0.00048 0.03342 0.0714
Issue Law
Adopted (5.017) (4.650) (4.305) (6.137) (1.661) (3.745) (0.063) (3.763) (6.251)
Dummy
1% 2% 1% 1% .3% 1% .02% 1% 1%
Arrest -0.00048 -0.00139 -0.00081 -0.000896 -0.00057 -0.000759 -0.0024 -0.00018 -0.00018
Rate for
the crime (77.257) (37.139) (47.551) (69.742) (88.984) (96.996) (90.189) (77.616) (74.972)
category
appropriate
endogenous
Variable 9% 7% 4% 9% 4% 10% 11% 4% 3%
(e.g.,
violent
crimes,
murders,
and so
on).
Population 0.00006 -0.00002 -0.00002 5.76E-06 0.000316 4.83E-06 -0.00007 0.000037 0.00048
per
Square Mile (3.684) (0.942) (1.022) (0.320) (15.117) (0.428) (5.605) (2.651) (26.722)
5% 1% 1% .4% 17% 1% 9% 4% 36%
Real Per 7.92E-06 0.0000163 -5.85E-06 4.71E-06 4.73E-06 -0.0000102 -0.0000184 -0.0000123 0.000015
Capita
Personal (2.883) (3.623) (1.669) (1.467) (1.244) (5.118) (8.729) (4.981) (4.689)
Income
1% 2% 1% 1% 1% 3% 4% 2% 2%
Real Per -0.00022 -0.00046 -0.00047 -0.00019 0.00007 0.00038 0.00060 0.00019 0.00021
Capita
Unemploymen (3.970) (5.260) (6.731) (2.904) (0.898) (9.468) (14.003) (3.706) (3.316)
t Ins.
.07% 1% 1% .05% .01% 2% 3% .08% .06%
Real Per -0.0000699 0.00025 -0.00017 0.000139 -0.00032 0.00019 0.00039 0.00002 0.00033
Capita
Income (0.841) (1.928) (1.634) (1.438) (2.840) (3.107) (6.219) (0.320) (3.452)
Maintenance
.3% 1% .7% .7% 1% 2% 4% .1% 2%
Real Per -1.97E-06 -0.000013 -2.37E-06 -6.81E-06 -5.50E-06 -8.65E-06 -0.0000106 -6.34E-06 -9.27E-06
Capita
Retirement (0.895) (3.713) (0.861) (2.651) (1.835) (5.371) (6.273) (3.186) (3.613)
Payments
per person .5% 3% .4% 2% 1% 4% 7% 2% 2%
over 65
Population 8.59E-08 -3.44E-08 -2.94E-07 4.54E-08 -6.10E-08 -2.18E-07 -2.14E-07 -3.10E-07 -4.06E-09
(4.283) (1.109) (11.884) (1.947) (2.271) (15.063) (14.060) (17.328) (0.177)
1% .4% 3% .06% .06% 6% 5% 6% .05%
% of Pop 0.05637 0.1134 0.04108 0.0900695 0.10548 0.1287 0.074 0.1710 0.0513
Black Male
Between (1.293) (1.515) (0.722) (1.767) (1.752) (4.068) (2.214) (4.366) (1.007)
10-19
5% 8% 3% 7% 5% 22% 11% 22% 4%
% of Pop 0.0009 0.0663 0.0794 -0.0528 -0.0060 -0.0143 -0.0203 -0.0057 0.00665
Black Male
Between (0.035) (1.514) (2.366) (1.749) (0.168) (0.759) (1.022) (0.245) (0.220)
20-29
% of Pop 0.0419 0.1085 -0.0832 0.2024 0.0061 0.04126 -0.0074 0.0044 0.14955
Black Male
Between (1.063) (1.640) (1.617) (4.424) (0.111) (1.445) (0.246) (0.124) (3.254)
30-39
Table 3 Continued
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
% of Pop -0.0243 -0.33549 0.9029 -0.3654 -0.00867 -0.02391 -0.03132 0.18939 -0.6846
Black Male
Between (0.300) (2.498) (8.562) (3.860) (0.077) (0.406) (0.506) (2.601) (7.235)
40-49
% of Pop 0.1816 -0.34753 -0.1509 0.2861 -0.00706 -0.0519 0.09135 -0.1318 0.05626
Black Male
Between (2.159) (2.518) (1.381) (2.889) (0.060) (0.843) (1.409) (1.730) (0.569)
50-64
% of Pop 0.12165 -0.14275 0.4373 0.1053 0.17053 -0.0367 0.06132 -0.0965 -0.3384
Black Male
Over 65 (1.377) (0.971) (3.742) (1.014) (1.379) (0.567) (0.900) (1.204) (3.254)
% of Pop -0.00394 0.0374 0.0368 -0.0692 -0.18307 0.0836 0.0217 0.1564 -0.1766
Black
Female
Between (0.088) (0.490) (0.630) (1.321) (2.957) (2.570) (0.631) (3.883) (3.372)
10-19
% of Pop -0.0993 -0.2247 0.1751 -0.1938 -0.2167 -0.0996 -0.1688 -0.0075 -0.2481
Black
Female
Between (3.094) (4.312) (4.280) (5.219) (4.986) (4.307) (6.936) (0.264) (6.711)
20-29
% of Pop 0.1218 -0.0828 0.1489 0.0947 0.3808 0.13409 0.2721 0.0944 0.1701
Black
Female
Between (3.383) (1.409) (3.228) (2.265) (7.691) (5.137) (9.909) (2.923) (4.072)
30-39
% of Pop 0.0107 0.59197 -0.7396 0.26946 -0.06891 0.05958 -0.05022 -0.0342 0.4816
Black
Female
Between (0.158) (5.321) (8.431) (3.387) (0.738) (1.213) (0.970) (0.562) (6.093)
40-49
% of Pop -0.2105 0.20188 0.1044 -0.0532 0.07078 -0.0241 -0.21799 0.0100 0.1153
Black
Female
Between (2.826) (1.648) (1.076) (0.612) (0.684) (0.443) (3.817) (0.149) (1.321)
50-64
% of Pop -0.2035 0.3071 -0.5164 -0.1557 -0.36915 -0.2035 -0.3877 -0.1234 0.2433
Black
Female
Over 65 (3.229) (2.969) (6.278) (2.104) (4.212) (4.406) (7.968) (2.160) (3.283)
% of Pop -0.0060 -0.0271 0.0056 0.03998 0.00219 -0.0066 -0.0062 0.00027 -0.0568
White Male
Between (0.382) (0.935) (0.265) (2.208) (0.098) (0.593) (0.523) (0.020) (3.152)
10-19
% of Pop 0.00842 0.0598 0.03779 0.0219 0.0426 0.00456 0.01738 0.00377 -0.0200
White Male
Between (0.729) (3.023) (2.528) (1.623) (2.636) (0.542) (1.958) (0.362) (1.487)
20-29
% of Pop -0.006 -0.01289 -0.0376 0.0739 -0.0706 -0.0520 -0.0268\ -0.0579 -0.0592
White Male
Between (0.322) (0.371) (1.444) (3.206) (2.507) (3.633) (1.779) (3.268) (2.583)
30-39
% of Pop -0.0095 -0.02078 0.0898 -0.0406 -0.11188 -0.14626 -0.0995 -0.1271 -0.0962
White Male
Between (0.375) (0.462) (2.685) (1.369) (3.099) (7.981) (5.147) (5.600) (3.265)
40-49
% of Pop -0.00575 -0.0458 0.0397 -0.0904 -0.14195 -0.1282 -0.0729 -0.1071 -0.2749
White Male
Between (0.236) (1.074) (1.237) (3.184) (4.104) (7.309) (3.942) (4.929) (9.771)
50-64
Table 3 Continued
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
% of Pop -0.1291 0.02336 0.0441 -0.1651 0.0421 -0.1442 -0.1194 -0.13975 -0.1104
White Male
Over 65 (6.065) (0.618) (1.547) (6.627) (1.370) (7.635) (8.887) (6.264) (5.651)
% of Pop 0.02346 0.0452 0.0741 -0.00863 0.0561 0.0824 0.0816 0.0865 0.0866
White
Female
Between (1.410) (1.473) (3.307) (0.448) (2.359) (6.907) (6.474) (5.863) (4.513)
10-19
% of Pop 0.0128 -0.0405 0.0551 0.03926 0.01327 -0.0086 -0.0421 0.02928 -0.0289
White
Female
Between (0.896) (1.673) (2.999) (2.348) (0.669) (0.828) (3.832) (2.272) (1.739)
20-29
% of Pop 0.01878 0.0447 0.14127 0.0299 -0.0079 0.0388 0.0171 0.06611 -0.1017
White
Female
Between (0.890) (1.209) (5.092) (1.215) (0.265) (2.545) (1.065) (3.502) (4.165)
30-39
% of Pop -0.0901 -0.00077 -0.0689 -0.0031 -0.02258 0.0584 -0.0354 0.0741 -0.0172
White
Female
Between (3.553) (0.017) (2.061) (0.106) (0.626) (3.193) (1.833) (3.270) (0.585)
40-49
% of Pop 0.00332 0.0119 0.0213 0.07882 0.03094 0.1044 0.06396 0.1100 0.10687
White
Female
Between (0.163) (0.335) (0.794) (3.313) (1.072) (7.103) (4.126) (6.042) (4.534)
50-64
% of Pop 0.0558 -0.0681 0.0578 0.0836 -0.0870 0.02027 0.0483 0.03631 -0.0459
White
Female
Over 65 (3.719) (2.588) (2.904) (4.761) (4.046) (1.867) (4.218) (2.701) (2.636)
% of Pop 0.2501 0.6624 0.5572 0.1872 0.5360 0.1587 0.2708 0.1487 0.6039
Other Male
Between (2.179) (3.022) (3.546) (1.389) (3.124) (1.917) (3.100) (1.451) (4.532)
10-19
% of Pop -0.1229 0.14495 -0.1656 -0.0573 0.0129 0.0786 0.0007 0.2037 -0.4066
Other Male
Between (1.966) (1.367) (2.065) (0.794) (0.149) (1.748) (0.015) (3.661) (5.667)
20-29
% of Pop 0.23126 -0.2958 -0.1907 0.4015 -0.1021 -0.1779 -0.4257 -0.0415 0.64667
Other Male
Between (1.866) (1.370) (1.161) (2.777) (0.572) (1.996) (4.532) (0.376) (4.525)
30-39
% of Pop 0.12678 -0.35775 -0.2406 -0.1903 0.77753 0.0287 0.2356 -0.2320 0.4640
Other Male
Between (0.824) (1.341) (1.180) (1.060) (3.538) (0.261) (2.027) (1.700) (2.620)
40-49
% of Pop -0.0904 -0.1572 0.2403 -0.2829 -0.39616 -0.0211 0.2676 -0.1952 -0.4198
Other Male
Between (0.605) (0.623) (1.240) (1.612) (1.869) (0.194) (2.330) (1.449) (2.411)
50-64
% of Pop 0.3469 -0.2585 0.8709 1.0193 -0.267 -0.0785 0.1863 -0.2342 -0.1792
Other Male
Over 65 (2.222) (1.019) (4.389) (5.566) (1.237) (0.688) (1.549) (1.659) (0.985)
% of Pop -0.0303 -0.7299 -0.1095 0.1207 -0.3461 -0.1769 -0.2861 -0.2304 -0.2739
Other
Female
Between (0.253) (3.185) (0.670) (0.857) (1.936) (2.049) (3.140) (2.155) (1.971)
10-19
Table 3 Continued
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
% of Pop -0.1323 -0.3293 0.2093 0.0933 -0.3033 -0.1464 -0.3243 -0.3334 -0.5646
Other
Female
Between (1.253) (2.145) (1.670) (0.557) (1.535) (1.849) (3.366) (2.435) (4.768)
20-29
% of Pop -0.2187 -0.1103 0.1556 -0.1674 -0.2158 -0.0874 0.2703 -0.2838 -0.7516
Other
Female
Between (1.823) (0.531) (0.988) (1.189) (1.253) (1.005) (2.949) (2.638) (5.395)
30-39
% of Pop -0.1413 0.56562 0.07877 0.1831 -0.48132 0.2452 -0.2767 0.6971 -0.1461
Other
Female
Between (1.011) (2.343) (0.429) (1.116) (2.407) (2.432) (2.600) (5.574) (0.901)
40-49
% of Pop -0.0972 0.4354 -0.6588 -0.2700 0.36585 -0.0491 -0.4901 0.1615 0.3078
Other
Female
Between (0.607) (1.612) (3.184) (1.439) (1.620) (0.424) (4.006) (1.125) (1.659)
50-64
% of Pop -0.4376 0.0569 -0.3715 -0.4428 -0.3596 -0.1052 -0.1408 -0.0478 -0.587
Other
Female
Over 65 (3.489) (0.277) (2.324) (3.012) (2.058) (1.148) (1.458) (0.422) (4.020)
Intercept 5.8905 2.0247 0.4189 4.2648 5.4254 9.1613 8.7058 7.596 8.332
(15.930) (3.326) (0.890) (9.857) (10.623) (33.945) (30.614) (22.751) (19.372)
Observation 43451 26458 33865 43445 34949 45940 45769 45743 43589
s =
F-statistic 115.11 37.95 44.93 70.47 131.75 87.22 82.16 59.33 116.35
=
Adjusted 0.8925 0.8060 0.8004 0.8345 0.9196 0.8561 0.8490 0.8016 0.8931
R2 =
Table 4: Questions of Aggregating the Data: National State Level Cross-Sectional Time-Series Evidence (Except for the use of state dummies in place of county dummies, the control variables are the same as those used in Table 3 including year dummies, though they are not all reported. Absolute t-statistics are in parentheses, and the percentage reported below that for some of the numbers is the percent of a standard deviation change in the endogenous variable that can be explained by a one standard deviation change in the exogenous variable. All regressions use weighted least squares where the weighting is each state's population)
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -0.1447 -0.0962 -0.0883 -.04468 -0.1372 -0.0527 -.1076 -0.0416 -0.045097
Issue Law
Adopted (4.025) (2.206) (1.468) (4.003) (2.852) (1.942) (3.268) (1.598) (1.056)
Dummy
7.6% 4.9% 4.7% 8.2% 5.3% 4.1% 7.9% 3.4% 2%
Arrest -0.000548 -0.000643 -0.000326 -0.002398 -0.009559 -0.00144 -0.002145 -0.005051 -0.001060
Rate for
the crime (2.035) (3.810) (3.8130) (5.566) (15.679) (4.431) (4.674) (4.385) (3.078)
category
correspondi
ng to the
appropriate 1.6% 4.6% 3.9% 5.6% 12.7% 1.3% 1.8% 5.5% 4.5%
endogenous
variable.
Intercept 2.9217 0.3820 3.3256 3.0062 0.7310 10.2591 8.5195 9.9704 8.1055
(1.479) (0.159) (1.000) (1.457) (0.276) (6.881) (4.687) (6.973) (3.446)
Observation 810 808 807 810 810 810 810 810 810
s =
F-statistic 137.38 100.896 58.523 119.518 154.604 58.612 60.234 59.948 176.584
=
Adjusted 0.9483 0.9309 0.8860 0.9410 0.9539 0.8857 0.8885 0.8880 0.9594
R2 =
Table 5: The Effect of Concealed Handguns on Victim Costs: What if All States Had Adopted "Shall Issue" Laws
(Using Miller et. al.'s 1996 estimates of the costs of crime in 1992 dollars)
Change in number of Change in Victim
crimes Costs from
if the states if the states
without "Shall without "Shall
Issue Laws" Issue Laws"
in 1992 had adopted in 1992 had adopted
the law the law
. .
Crime Category Estimates Using Estimates Using Estimates Using Estimates Using
County Level Data State Level Data County Level Data State Level Data
Murder -1,570 -1,777 -$4,753,977,904 -$5,379,921,760
Rape -4,177 -7,000 -$374,277,659 -$627,205,629
Aggravated Assault -60,363 -128,906 -$1,405,042,403 -$3,000,497,114
Robbery -11,898 -73,865 -$98,033,414 -$608,605,630
Burglary 1,052 -235,823 $1,516,890 -$340,036,068
Larceny 191,743 -238,674 $73,068,706 -$90,953,267
Auto Theft 89,928 -56,799 $342,694,264 -$216,449,345
Total Change in -$6,214,051,520 -$10,263,669,813
Victim CostsTable 6: Questions of Aggregating the Data: Does Law Enforcement and "Shall Issue" Laws have the Same Effect in High and Low Crime Areas? (The control variables are the same as those used in Table 3 including year and county dummies, though they are not reported. Absolute t-statistics are in parentheses. All regressions use weighted least squares where the weighting is each state's population)
A) Sample Where County Crime Rates are Above the Median
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -0.0597 -0.1021 -0.0719 -.04468 -0.0342 0.0161 0.0036 0.0296 0.0524
Issue Law
Adopted (7.007) (7.870) (7.415) (4.411) (3.012) (2.943) (0.533) (5.474) (5.612)
Dummy
Arrest -0.000523 -0.00105 -0.000326 -0.00063 -0.00294 -0.005354 -0.00565 -0.00596 -0.00133
Rate for
the crime (-17.661) (29.291) (3.8130) (18.456) (9.381) (33.669) (27.390) (41.585) (11.907)
category
correspondi
ng to the
appropriate
endogenous
variable.
B) Sample Where County Crime Rates are Below the Median
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -0.0369 -0.0761 -0.0304 -0.0025 -0.0787 0.0881 0.0297 0.0874 0.07226
Issue Law
Adopted (1.934) (1.753) (0.978) (0.013) (2.978) (5.801) (2.110) (5.246) (3.276)
Dummy
Arrest -0.0005242 -0.0008799 -0.000656 -.00068 -0.0003699 -0.001354 -0.0027135 -0.000998 -0.0001412
Rate for
the crime (30.302) (11.647) (31.542) (37.306) (9.018) (39.101) (41.603) (37.559) (62.596)
category
correspondi
ng to the
appropriate
endogenous
variable.
Table 7: Controlling for the fact that Larger Changes in Crime Rates are Expected in the More Populous Counties Where the Change in the Law Constituted a Bigger Break with Past Policies (The control variables are the same as those used in Table 3 including year and county dummies, though they are not reported since the coefficient estimates are very similar to those reported earlier. Absolute t-statistics are in parentheses. All regressions use weighted least squares where the weighting is each county's population)
Exogenous
Variables ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
ed Theft
Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -9.41E-08 -2.07E-07 -7.83E-08 -1.06E-07 -2.29E-08 5.18E-08 6.96E-09 4.90E-08 1.40E-07
Issue Law
Adopted (6.001) (7.388) (4.043) (5.784) (1.295) (4.492) (0.572) (3.432) (7.651)
Dummy
*County
Population
Arrest -0.000475 -0.00139 -0.000807 -0.000895 -0.000575 -0.000759 -0.002429 -0.000177 -0.0001754
Rate for
the crime (77.222) (37.135) (47.535) (69.663) (88.980) (97.027) (90.185) (77.620) (75.013)
category
correspondi
ng to the
appropriate
endogenous
variable.
Observation 43451 26458 33865 43445 34949 45940 45769 45743 43589
s =
F-statistic 115.15 38.02 44.92 70.46 131.74 87.23 82.16 59.33 116.41
=
Adjusted 0.8925 0.8062 0.8004 0.8345 0.9196 0.8561 0.8490 0.8016 0.8931
R2 =
Implied Percent Change in Crime Rate: The Effect of the "Shall Issue" Interaction Coefficient Evaluated at Different Levels of County Populations
Violent Murder Rape Aggravated Robbery Property Auto Burglary Larceny
Population Crimes Assault Crimes Theft
1/2 Mean -.36% -.78% -.3% -.4% -.1% .2% .03% .2% .5%
37,887
Mean -.71 -1.6 -.6 -.8 -.2 .4 .05 .4 1.1
75,773
Plus 1 -3.1 -6.8 -2.6 -3.5 -.7 1.7 .23 1.6 4.6
Standard
Dev.
326,123
Plus 2 -5.4 -11.9 -4.5 -6.1 -1.3 2.99 .4 2.8 8.1
Standard
Dev.
576,474
Percent of a one standard deviation change in corresponding crime rate that can be explained by a one standard deviation change in the arrest rate for that crime.
Violent Murder Rape Aggravated Robbery Property Auto Burglary Larceny
Crimes Assault Crimes Theft
9% 7% 4% 9% 4% 10% 11% 4% 3%
Table
8: Using Other Crime Rates that are Relatively Unrelated to Changes in "Shall
Issue" Rules as an Method of Controlling for Other Changes in the Legal
Environment: Controlling for Robbery and Burglary Rates (While not all
the coefficient estimates are reported, all the control variables are the same
as those used in Table 3, including year and county dummies. Absolute
t-statistics are in parentheses. All regressions use weighted least squares
where the weighting is each county's population. Net violent and property
crime rates are respectively net of robbery and burglary crime rates to avoid
producing any artificial collinearity. Likewise, the arrest rates for those
values subtract out that portion of the corresponding arrest rates do to
arrests for robbery and burglary.)Endogenous Variables
Controlling for Robbery Rates
Exogenous ln(Net ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Burglary ln(Larceny ln(Auto
Violent ed Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -1.03E-07 -1.72E-07 -7.73E-08 -1.03E-07 . . . 5.61E-08 -3.50E-09 5.35E-08 1.47E-07
Issue Law
Adopted (6.318) (7.253) (4.049) (5.777) (5.206) (0.304) (3.911) (8.844)
Dummy
*County
Population
Arrest -0.0003792 -0.0013449 -0.00073 -0.000776 . . . -0.0006448 -0.0020339 -0.0001547 -0.0001382
Rate for
the
crime (57.644) (36.240) (42.672) (60.834) (86.517) (77.992) (69.968) (63.888)
category
correspondi
ng to the
appropriate
endogenous
variable.
Ln(Robbery 0.1083118 0.116406 0.0983088 0.1196466 . . . 0.1176149 0.1135451 0.1164045 0.2173908
Rate)
(46.370) (24.616) (30.363) (47.469) (78.825) (70.826) (61.762) (92.212)
Observation 43197 26458 33865 43445 . . . 45940 45769 45743 43589
s =
F-statistic 81.93 39.19 46.55 75.09 . . . 101.83 93.39 65.82 143.54
=
Adjusted 0.8555 0.8111 0.8062 0.8433 . . . 0.8744 0.8649 0.8179 0.9117
R2 =
Controlling for Burglary Rates
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Net ln(Burglary ln(Larceny ln(Auto
ed Prop. Theft
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
Shall -9.52E-08 -1.73E-07 -8.03E-08 -1.03E-07 -1.47E-08 7.23E-08 . . . 5.50E-08 1.45E-07
Issue Law
Adopted (6.937) (7.434) (4.356) (6.072) (0.759) (6.854) (4.769) (8.943)
Dummy
*County
Population
Arrest -0.00026 -0.00128 -0.00051 -0.00054 -0.000429 -0.000469 . . . -0.000102 -0.000116
Rate for
the
crime (44.982) (35.139) (30.010) (42.883) (69.190) (61.478) (53.545) (53.961)
category
correspondi
ng to the
appropriate
endogenous
variable.
Ln(Burglary 0.5667123 0.4459916 0.4916113 0.5302516 0.6719892 0.5773792 . . . 0.6009071 0.6416852
Rate)
(110.768) (37.661) (56.461) (83.889) (78.531) (155.849) (150.635) (106.815)
Observation 43451 26458 33865 43445 34949 45813 . . . 45743 43589
s =
F-statistic 154.04 40.78 50.59 84.97 159.18 123.99 . . . 98.08 152.82
=
Adjusted 0.9176 0.8173 0.8191 0.8591 0.9327 0.8949 . . . 0.8706 0.9167
R2 =
Table
9: Rerunning the Regressions on Differences (The variables for income;
population; racial, sex, and age compositions of the population; and density
are all in terms of first differences. While not all the coefficient estimates
are reported, all the control variables used in Table 3 are used here,
including year and county dummies. Absolute t-statistics are in parentheses.
All regressions use weighted least squares where the weighting is each county's
population.)
All Endogenous Variables are in Terms of First Differences
All Variables Except for the "Shall Issued" Dummy Differenced:
Exogenous [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l
n(Violent n(Murder n(Rape n(Aggravat n(Robbery n(Property n(Burglary n(Larceny n(Auto
ed Theft
Variables Crime Rate) Rate) Assault Rate) Crime Rate) Rate) Rate)
Rate) Rate) Rate)
Using the
Shall
Issue
Dummy
Shall -0.021589 -0.025933 -.052034 -.0456251 .0331607 .0526532 .0352582 .0522435 .128475
Issue Law
Adopted (1.689) (0.841) (2.761) (2.693) (1.593) (4.982) (3.16) (4.049) (5.324)
Dummy
First -.0004919 -.0015482 -.0008641 -.0009272 -.0005725 -.0007599 -.0024482 -.0001748 -.0001831
Difference
s in the
Arrest (75.713) (25.967) (46.509) (67.782) (82.38) (91.259) (88.38) (75.969) (53.432)
Rate for
the
crime
category
correspond
ing to
the
appropriat
e
endogenous
variable.
Intercept -.073928 -.0402018 -.014342 -.0522417 -.1203331 -.0952347 -0770997 -.1062443 -.2604944
(6.049) (1.554) (0.904) (3.68) (6.925) (10.8) (8.312) (9.872) (13.009)
Observatio 37611 20420 26269 37694 27999 40901 40686 40671 37581
ns =
F-statisti 3.80 0.69 2.56 4.03 4.05 4.36 6.62 3.1 10.34
c =
Adjusted 0.1867 -0.0379 0.1389 0.1972 0.2283 0.2047 .3018 0.1386 0.4338
R2 =
All
Variables
Difference
d:
Exogenous [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l [[Delta]]l
n(Violent n(Murder n(Rape n(Aggravat n(Robbery n(Property n(Burglary n(Larceny n(Auto
ed Theft
Variables Crime Rate) Rate) Assault Rate) Crime Rate) Rate) Rate)
Rate) Rate) Rate)
Using the
Shall
Issue
Dummy
First -0.026959 -0.0363798 -.0394318 -0.0540946 .0071132 .0481937 .0072487 .0623146 .2419118
Difference
s in the
Shall (2.57) (1.826) (2.887) (4.414) (0.471) (6.303) (0.898) (6.676) (13.884)
Issue Law
Adopted
Dummy
First -.0004919 -.0015481 -.0008642 -.0009275 -.0005724 -.0007598 -.002448 -.0001748 -.0001829
Difference
s in the
Arrest (75.728) (25.968) (46.519) (67.819) (82.371) (91.266) (88.362) (75.978) (53.495)
Rate for
the
crime
category
correspond
ing to
the
appropriat
e
endogenous
variable.
Intercept -.0758797 -.042305 -.0188927 -.0562624 -.1176478 -.0907433 -.0742121 -.1016434 -.248623
(6.241) (1.642) (1.196) (3.983) (6.801) (10.341) (8.038) (9.494) (12.506)
Observatio 37611 20420 26269 37694 27999 40901 40686 40671 37581
ns =
F-statisti 3.8 0.69 2.56 4.04 4.05 4.37 6.62 3.11 10.45
c =
Adjusted 0.1868 -0.0378 0.1389 0.1975 0.2282 0.205 .3016 0.1393 0.4365
R2 =
Table 10: Allowing the Change in the "Shall Issue" Law and the Arrest Rate to be Endogenous Using 2SLS (While not all the coefficient estimates are reported, all the control variables are the same as those used in Table 3, including year and county dummies. Absolute t-statistics are in parentheses, and the percentage reported below that for some of the numbers is the percent of a standard deviation change in the endogenous variable that can be explained by a one standard deviation change in the exogenous variable.)
Endogenous Variables are in Crimes per 100,000 Population
Exogenous
Variables ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Auto ln(Burglary ln(Larceny
ed Theft
Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Rate) Rate) Rate)
Rate)
A) Using
the
predicted
values of
arrest
rates
similar to
Ehrlich's
(1973)
study
Shall -1.262 -1.1063 -1.059 -1.3192 -0.8744 -1.1182 -0.7668 -0.7603 -1.122
Issue Law
Adopted (21.731) (5.7598) (-4.4884) (18.5277) (7.4979) (15.3716) (11.435) (19.328) (25.479)
Dummy
10.5% 7.5% 6.4% 10.1% 4.9% 7.67% 11.4% 10.6% 13.5%
Arrest -0.002324 -0.00094 -0.0359 -0.002176 -0.00241 -0.01599 -0.002759 -0.01783 -0.0124
Rate for
the crime (9.6892) (1.8436) (9.667) (7.1883) (4.481) (33.26) (2.989) (14.36) (31.814)
category
correspondi
ng to the
appropriate 60.7% 5.2% 60.1% 44.6% 36.9% 80.1% 21.3% 79.6% 80.6%
endogenous
variable.
Observation 31129 31129 31129 31129 31129 31129 31129 31129 31129
s =
F-statistic 61.97 19.07 22.3 39.81 63.71 60.78 84.21 46.48 38.37
=
Adjusted 0.8592 0.644 0.6807 0.7953 0.8626 0.8568 0.8893 0.8199 0.7891
R2 =
B)
Including
the change
in crime
rates when
estimating
the
predicted
values of
the arrest
rates
Shall -.26104 -.5732 -.1992 -.29881 -.0054 -.20994 -.2774 -.1153 -.2623
Issue Law
Adopted (20.12) (18.21) (9.6317) (15.4465) (0.2935) (29.4242) (32.5051) (13.397) (32.4253)
Dummy
2.2% 3.9% 1.2% 2.3% 0.3% 3.3% 2.1% 1.6% 3.2%
Arrest -0.007827 -0.024 -0.02626 -0.01028 -0.00716 -0.00933 -0.01233 -0.03839 -0.0101
Rate for
the crime (746.74) (687.7) (1047) (582) (901.8) (820.7) (1242.7) (796.8) (956.14)
category
correspondi
ng to the
appropriate 104% 95% 117% 88% 109% 95% 95.1% 71% 101%
endogenous
variable.
Observation 31129 31129 31129 31129 31129 31129 31129 31129 31129
s =
F-statistic 1723 1260.9 4909.6 797.5 3614.86 1671.49 6424 1389 1625.8
=
Adjusted 0.9942 0.9921 0.9980 0.9876 0.9972 0.9941 0.9984 0.9929 0.9939
R2 =
Table 10 continued
First stage estimates of Shall Issue Law (Absolute t-statistics are in parentheses. The sample is limited because the data on police employment used in producing the predicted arrest rates were only available from 1982 to 1992. While the estimates from the first specification were used in the above regressions, the logit estimates are provided for comparison. Not all the variables that were controlled for are shown. These additional variables included: year and regional dummies (South, Northeast, and Midwest) and the state's population.)
Exogenous Variables .
Endogenou ln(Violen [[Delta]] ln(Prop. Nat % Rep. % Rep. % Rep. % Rep. % Pop. % Pop.
s t ln(Violen [[Delta]] Rifle Pres. Pres. Pres. Pres.
t ln(Prop. Assoc.
Variable Crime Crime Crime Crime Membershi in State in State in State in State Black White
Rate) Rate) p as
Rate) Rate) % of Vote Vote Vote Vote for for
State Pop 80*Year 84*Year 88*Year 92*Yr
Dum 79-82 Dum 83-86 Dum 87-90 Dum 91-92 State State
Least
Squares
Estimate
(1) -.01817 .00825 -.02889 .0094 .000107 .0061 .0034 .01702 .0299 .00518 .0031
Shall
Issue (9.710) (5.031) (8.748) (2.577) (19.383) (5.485) (4.986) (22.844) (27.317) (13.06) (8.470)
Law
F-statist adjusted- Obs. =
ic = R2 = 31137
209.85 .1436
Logit
(2) -.0797 .038249 -.2095 .08119 .0004344 .0567 .01456 .09976 .12249 .0409 .0364
Shall
Issue (6.003) (3.294) (8.657) (3.121) (10.329) (6.227) (2.437) (16.203) (16.273) (10.090) (9.131)
Law
Chi-squar Pseudo Obs. =
ed = R2 = 31137
5007.44 .1687
First stage estimates of the Probability of Arrest Using: Reporting only the estimates for violent and property crime rates (Absolute t-statistics are in parentheses. The sample is limited because the data on police employment were only available from 1982 to 1992. Not all the variables that were controlled for are shown. These additional variables included: the number of police with arrest powers divided by the number of violent crimes; the number of police with arrest powers divided by the number of property crimes; the number of police without arrest powers divided by the number of violent crimes; the number of police without arrest powers divided by the number of property crimes; these preceding variables using payrolls; the breakdown of the county's population by age, sex, and race used in Table 3; year and county dummies; the measures of income reported in Table 3; and the state's population. The estimates also using the change in crime rates are available from the authors.)
Exogenous Variables .
Endogenous ln(Violent ln(Propert # of # of Nat Rifle Population % Rep. % Rep. % Rep. % Rep.
y Police in Police in Assoc. Pres. Pres. Pres. Pres.
St. St.
Variable Crime Crime Employed Employed Membership Density in State in State in State in State
Rate) Rate) with without as
lagged lagged power of power of % of per Vote Vote Vote Vote 92*Yr
arrest/ arrest/ State Pop square 80*Year 84*Year 88*Year
State State mile Dum 79-82 Dum 83-86 Dum 87-90 Dum 91-92
population population
A) The
predicted
values of
arrest
rates
that most
closely
correspond
to
Ehrlich's
(1973)
2SLS
estimates
(1) -2.224 . . . -14093.61 95.085 .01463 .0739 -6.936 -4.293 -3.3467 -3.4316
Arrest
Rate for (1.441) (3.065) (2.206) (1.940) (6.418) (9.975) (8.270) (5.865) (4.967)
Violent F-statisti adjusted-R Obs. =
Crimes c = 1.83 2 = .0814 28954
(2) . . . .90203 -2805.2 -1.3057 .01045 .00415 -1.5931 -.9155 -1.1778 -1.2009
Arrest
Rate for (0.738) (1.173) (0.059) (1.305) (0.697) (4.434) (3.420) (4.004) (3.416)
Property F-statisti adjusted-R Obs. =
Crimes c = 1.08 2 = .0084 30814
B)
Including the change in crime rates in addition to those already noted when
estimating the predicted values of the arrest rates (the coefficients on the
percentage of the state voting Republican in presidential elections is similar
to those reported in the preceding section).
Exogenous Variables .
Endogenous ln(Violent [[Delta]]ln ln(Property [[Delta]]ln # of # of Nat Rifle
(Violent (Property Police in Police in Assoc.
St. St.
Variable Crime Rate) Crime Rate) Crime Rate) Crime Rate) Employed Employed Membership Density
with without as
lagged lagged power of power of % of State per square County
arrest/ arrest/
State State Population mile Population
population population
A) The
predicted
values of
arrest
rates that
correspond
to
Ehrlich's
(1973)
2SLS
estimates
(1) Arrest -128.4 -123.64 . . . . . . -12194 96.3244 .0009 .0646 -.0000726
Rate for (39.86) (44.17) (2.750) (2.317) (0.060) (5.824) (4.877)
Violent F-statistic adjusted-R2 Obs. =
Crimes = 2.59 = .1458 28954
(2) Arrest . . . . . . -109.69 -106.92 -1394 -1.9891 -.0072 .0083 -.0000111
Rate for (49.342) (58.21) (0.618) (0.095) (0.949) (1.473) (1.522)
Property F-statistic adjusted-R2 Obs. =
Crimes = 2.30 = .1165 30814
Table 11: Changes in Murder Methods for Counties Over 100,000 from 1982 to 1991 (While not all the coefficient estimates are reported, all the control variables are the same as those used in Table 3, including the year and county dummies. Absolute t-statistics are in parentheses. All regressions use weighted least squares where the weighting is each county's population. The first column uses the UCR numbers for counties over 100,000, while the second column uses the numbers on total gun deaths available from the Mortality Detail Records and the third column takes the difference between the UCR numbers for total murders and Mortality Detail Records of gun deaths.)
Endogenous Variables are in Murders per 100,000 Population
Exogenous ln(Total ln(Murder with ln(Murders by
Variables Murders) Guns) Nongun Methods)
Using the Shall Issue Dummy
Shall Issue Law -.09704 -.09045 -.08854
Adopted Dummy (3.183) (1.707) (1.689)
Arrest Rate for -.00151 -.00102 -.00138
Murder (26.15) (6.806) (7.931)
Intercept .63988 -8.7993 -7.51556
(0.436) (2.136) (2.444)
Observations = 12740 12759 8712
F-statistic = 21.40 6.60 4.70
Adjusted R2 = 0.8127 0.5432 0.5065
Table 12: Changes in Composition of Murder Victims Using Annual State Level Data from the Uniform Crime Reports Supplementary Homicide Reports from the period 1977 to 1992 (While not all the coefficient estimates are reported, all the control variables are the same as those used in Table 4, including the year and state dummies. Absolute t-statistics are in parentheses. All regressions use weighted least squares where the weighting is each state's population.)
Endogenous Variables are in Percentage Points
By Victim's Sex . By Victim's Race . By Victim's Relationship With Offender .
Exogenous % of % of % of % of % of % of % of % of % of % Victims
Victims Victims Victims Victims Victims Victims
Variables Male Female Sex is Victims Victims Victims where the where the where the where the
not that that that Offender
Using the
Shall
Issue
Dummy
Identified are White are Black are is Known Offender Offender relationsh
Hispanic to Victim is in is ip
Shall but is the Family is a Unkown
Issue not in Stranger
Family
Law 0.3910 -.4381 0.0476 0.0137 0.7031 -.8659 2.5824 -.2503 0.5438 -2.8755
Adopted
Dummy (0.388) (0.439) (0.399) (0.017) (0.575) (0.609) (1.567) (0.210) (0.459) (1.464)
Arrest 0.00068 -.001385 0.000703 -.0202 0.0132 0.00327 0.0174 -.0145 0.0079 -.0108
Rate
for Murder (0.141) (0.289) (1.227) (2.316) (2.244) (0.478) (2.198) (2.541) (1.394) (1.141)
Intercept 102.20 -3.2763 1.0558 152.19 -30.948 -7.7863 -73.4677 165.1719 89.843 -81.55
(1.718) (0.056) (0.150) (1.418) (0.428) (0.093) (0.755) (2.345) (165.17) (0.703)
Observatio 804 804 804 804 804 804 804 804 804 804
ns =
F-statisti 14.27 14.51 1.06 45.47 125.09 35.94 14.96 12.87 7.84 26.06
c =
Adjusted 0.6409 0.6450 0.0077 0.8568 0.9435 0.8245 0.6525 0.6150 0.4790 0.7712
R2 =
Table
14: Using Pennsylvania Data on the Number of Permits Issued to Measure the
Differential Impact of Pennsylvania's 1989 "Shall Issue" Law on Different
Counties: Data for Counties with Populations Over 200,000 (Absolute
t-statistics are in parentheses, and the percentage reported below that is the
percent of a standard deviation change in the endogenous variable that can be
explained by a one standard deviation change in the exogenous variable. While
not all the coefficient estimates are reported, all the control variables are
the same as those used in Table 3, including year and county dummies. All
regressions use weighted least squares where the weighting is each county's
population. The use of SHALL*POPULATION variable that was used in the earlier
regressions instead of the change in right-to-carry permits variable was tried
here and produced very similar results. We also tried controlling for either
the robbery or burglary rates, but we obtained very similar results.)
Endogenous
Variables
are in
Crimes per
100,000
Population
Exogenous ln(Violent ln(Murder ln(Rape ln(Aggravat ln(Robbery ln(Property ln(Auto ln(Burglary ln(Larceny
ed
Variables Crime Rate) Rate) Rate) Assault Rate) Crime Rate) Theft Rate) Rate) Rate)
Rate)
Change in -.05613 -0.1123 -0.0741 -0.06499 0.00199 -0.01836 0.01015 -0.0354 0.01659
the (Number
Right-to-Ca (2.159) (2.005) (1.725) (1.656) (0.054) (0.481) (0.365) (2.171) (0.271)
rry Pistol
Permits/Pop
ulation
over 21) 12% 14% 16% 15% 3% 7% 1% 13% 6%
between
1988
and each
year since
the
Law was
implemented
Arrest -.00802 -.00352 -.000459 -.00796 -.008191 -.0041 -.00062 -.01107 .0003095
Rate for
the crime (7.656) (6.201) (0.380) (6.870) (6.898) (2.057) (1.135) (5.057) (0.154)
category
correspondi
ng to the
appropriate 29% 23% 3% 38% 46% 9% 4% 24% 6%
endogenous
variable.
Population -.000117 0.00306 0.000987 -0.00039 0.0005395 0.00037 -0.000171 .000518 0.00077
per
square mile (0.246) (2.243) (1.087) (0.600) (0.835) (1.283) (0.275) (1.442) (2.601)
Real Per .0000302 -.000058 0.000066 .0000197 0.000047 -.0000485 -0.000067 -0.000034 -.00004
Capita
Personal (0.942) (0.614) (1.071) (0.452) (1.055) (2.611) (1.599) (1.396) (2.025)
Income
Intercept -13.352 118.93 -67.015 34.752 -52.529 -10.31 27.816 -29.40 6.2484
(0.348) (1.069) (0.889) (0.671) (0.993) (0.467) (0.557) (1.016) (0.269)
Observation 279 279 279 279 279 279 279 279 279
s =
F-statistic 219.4 38.08 41.06 75.54 223.51 109.68 216.03 87.49 76.11
=
Adjusted 0.9841 0.9133 0.9193 0.9549 0.9844 0.9686 0.9839 0.9609 0.9552
R2 =
Table
15: Using Oregon Data on the Number of Permits Issued, the Conviction Rate,
and Prison Sentence Lengths (Absolute t-statistics are in parentheses, and
the percentage reported below that is the percent of a standard deviation
change in the endogenous variable that can be explained by a one standard
deviation change in the exogenous variable. We also controlled for Prison
Sentence length but the different reporting practices used by Oregon over this
period makes its use somewhat problematic. To deal with this problem the
prison sentence length variable was interacted with year dummy variables. Thus
while the variable is not consistent over time its is still valuable in
distinguishing penalties across counties at a particular point in time. While
not all the coefficient estimates are reported, all the remaining control
variables are the same as those used in Table 3, including year and county
dummies. The categories for violent and property crimes are eliminated because
the mean prison sentence data supplied by Oregon did not allow us to use these
two categories. All regressions use weighted least squares where the weighting
is each county's population.)
Endogenous Variables are in Crimes per 100,000 Population
Exogenous ln(Murder ln(Rape ln(Aggravated ln(Robbery ln(Auto ln(Burglary ln(Larceny
Variables Rate) Rate) Assault Rate) Rate) Theft Rate) Rate) Rate)
Change in -.3747 -.0674 -.0475 -.04664 0.1172 0.02655 -.0936
the (Number
Right-to-Carr (1.598) (0.486) (0.272) (0.385) (1.533) (0.536) (2.328)
y Pistol
Permits/Popul
ation
over 21) 3% 1% 0.5% 0.28% 1% 1% 3%
between 1988
and each
year since
the
Law was
implemented
Arrest Rate -.00338 -.00976 -.00442 -.00363 -.00036 -.00679 -.00936
for
the crime (6.785) (9.284) (7.279) (4.806) (1.481) (4.458) (6.764)
category
corresponding
to the
appropriate 17% 19% 19% 9% 3% 16% 16%
endogenous
variable.
Conviction -.00208 -.00093 -.01511 -.00190 -.00373 -.00274 -.00859
Rate
conditional
on arrest (6.026) (7.668) (2.150) (4.465) (3.031) (4.297) (3.140)
for the
crime
category
corresponding
to the
appropriate 11% 10% 6% 4% 4% 10% 20%
endogenous
variable.
Population -.00333 0.0063 0.01177 0.0079 0.00062 0.00425 -.00030
per
square mile (0.415) (0.059) (2.430) (2.551) (0.367) (3.937) (0.319)
Real Per -.000138 -.000038 -.000162 0.000108 .000037 .000021 8.29 e-6
Capita
Personal (0.769) (0.463) (1.301) (1.542) (0.965) (0.816) (0.407)
Income
Intercept 6.1725 8.2432 84.464 -16.303 2.6213 -11.2489 20.047
(0.342) (0.496) (3.131) (1.114) (0.326) (2.169) (4.748)
Observations 250 393 239 337 403 487 422
=
F-statistic = 5.74 16.61 38.79 97.94 156.02 89.90 86.81
Adjusted R2 = 0.6620 .8113 .9439 .9677 .9766 .9522 .9569
Table
16: Using the 1990 to 1995 Arizona Data on the Number of Permits Issued, the
Conviction Rate, and Prison Sentence Lengths (Absolute t-statistics are in
parentheses, and the percentage reported below that is the percent of a
standard deviation change in the endogenous variable that can be explained by a
one standard deviation change in the exogenous variable. All variables, except
for the county's population and the year and county dummies, have been
reported. The categories for violent and property crimes are eliminated
because the mean prison sentence data supplied by Oregon did not allow us to
use these two categories. All regressions use weighted least squares where the
weighting is each county's population.)
Endogenous Variables are in Crimes per 100,000 Population
Exogenous ln(Aggravated ln(Robbery ln(Auto Theft ln(Burglary
Variables ln(Murder Rate) ln(Rape Rate) Assault Rate) Rate) Rate) Rate) ln(Larceny Rate)
Change .0016 .0025 -.0803 -.0095 .0051 -.00516 .0037 .0039 -.0019 -.0076 .0006 0.0007 -.0003 -.0005
in the
(Number
Right-t (0.209) (0.311) (1.397) (0.334) (1.265) (1.291) (0.574) (0.551) (0.222) (0.940) (0.210) (0.225) (0.094) (0.185)
o-Carry
Pistol
Permits
/Popula
tion)
from 1.7% 2.7% 8% 2% 9% 9% 3% 3% 2% 9% 8% 9% 1% 1%
the
zero
allowed
before
the
law
and
each
year
since
the
Law
was
impleme
nted,
the
numbers
for
1994
were
multipl
ied by
.5
Convict -.0039 -.00399 -.0055 -.0053 -.0453 -.0429 -.0111 -.0110 -.1373 -.1605 -.10032 -.1037 -.325 -.3298
ion
Rate
for
the (7.677) (6.798) (7.558) (7.014) (13.51) (12.18) (9.553) (9.391) (1.678) (1.879) (14.44) (14.62) (12.1) (13.80)
crime
categor
y
corresp
onding
to the
appropr 29% 30% 27% 26% 72% 67% 21% 20% 37% 43% 28% 29% 60% 60%
iate
endogen
ous
variabl
e.
Mean -.01033 . . . .0052 . . . -.0261 . . . -.0095 . . . -.0087 . . . -.0084 . . . -.018 . . .
Prison
Sentenc
e
Length (1.457) (0.364) (1.155) (0.629) (.055) (1.759) (0.936)
for
those
Sentenc
ed
to 5% 2% 6% 1% .2% .7% 3%
Prison
in
that
Year
Time . . . .0041 . . . -.0178 . . . -.0170 . . . -.0221 . . . 0.0317 . . . -.0119 . . . -.0952
Served
for
those
ending (0.18) (0.602) (0.464) (0.871) (0.463) (0.405) (3.479)
their
prison
terms
in
that
Year
4% 2% 2% 2% 2% .8% 11%
Populat -.1014 -.0791 -.4748 -.4459 -.1424 -.1361 -.1411 -.1514 -.413 -.4019 -.0835 -.0798 -.0313 -.00030
ion
per
square (0.826) (0.569) (3.595) (3.274) (2.164) (1.942) (1.288) (1.477) (2.603) (2.433) (1.759) (1.670) (0.631) (0.319)
mile
Interce 1.208 0.926 1.4750 1.477 4.341 4.365 1.838 1.753 3.432 2.5099 5.467 5.4296 6.621 6.873
pt
(3.594) (1.765) (5.095) (5.262) (28.46) (26.30) (5.157) (4.203) (5.061) (7.094) (38.66) (5.430) (53.03) (57.475
)
Observa 74 70 78 75 89 86 64 68 60 89 84 84 85 84
tions =
F-stati 17.26 14.50 27.64 24.86 56.48 38.79 81.33 76.67 32.12 39.60 109.61 101.18 99.75 118.24
stic =
Adjuste 0.8367 0.8182 .8925 .8856 .9380 .9439 .9656 .9629 .9239 .9330 .9691 .9666 .9658 .9713
d R2 =
Table 17: Did Carrying Concealed Handguns Increase the Number of Accidental Deaths?: Using 1982-91 County Level Data (While not all the coefficient estimates are reported, all the control variables are the same as those used in Table 3, including year and county dummies. Absolute t-statistics are in parentheses. All regressions weight the data by each county's population.)
Endogenous Variables are in Deaths per 100,000 population
Ordinary Least Tobit
Squares
Exogenous ln(Accidental ln(Accidental Accidental Accidental Deaths
Deaths
Variables Deaths from from Nonhandgun Deaths from from Nonhandgun
Using the Shall
Issue Dummy
Handguns) Sources) Handguns Sources
Shall Issue Law 0.00478 .0980 0.574 1.331
Adopted Dummy (0.096) (1.606) (0.743) (0.840)
Population per -.0007 0.000856 -.0000436 -.0001635
square mile (6.701) (7.063) (0.723) (1.083)
Real Per Capita 0.0000267 -.000057 .0000436 -.009046
Personal Income (1.559) (2.882) (1.464) (6.412)
Intercept or -3.376 -8.7655 7.360841 29.36
Ancillary (1.114) (2.506) (44.12) (201.7)
Parameter
Observations = 23271 23271 23271 23271
F-statistic = 3.98 3.91
Adjusted R2 = 0.2896 0.2846
Log Likelihood = -7424.6 -109310.6
Left-censored
Observations = 21897 680