Scribd
Upload
3 views

Table 1: Panel Regression Findings Table 2

The document contains 7 tables that analyze the relationship between various crime rates and firearms availability at the city and county level using quarterly panel data and controlling for other relevant factors. Table 1 provides descriptive statistics for the variables. Tables 2-7 use fixed effects panel regressions to examine the effects of lagged firearms on different types of violent crimes, both overall and firearm-involved specifically, while controlling for other crime types. All reported coefficients are statistically significant. The notes discuss methodological considerations like using fixed effects and address potential issues.

Uploaded by

William Zheng
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
3 views

Table 1: Panel Regression Findings Table 2

The document contains 7 tables that analyze the relationship between various crime rates and firearms availability at the city and county level using quarterly panel data and controlling for other relevant factors. Table 1 provides descriptive statistics for the variables. Tables 2-7 use fixed effects panel regressions to examine the effects of lagged firearms on different types of violent crimes, both overall and firearm-involved specifically, while controlling for other crime types. All reported coefficients are statistically significant. The notes discuss methodological considerations like using fixed effects and address potential issues.

Uploaded by

William Zheng
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Table 1

Variable
Firearms
Totpop

Mean (StD)
44.9 (71.3)
106762 (109630)

Fondling
Theft
Mvtheft
Drugs
Simassault

23.2 (30.7)
1535.4 (1915.5)
276.9 (614.4)
249.9 (327.0)
924.8 (1393.7)

Variable
Homicide
Homicide_f
kidnap
kidnap_f
Rape
Rape_f
Robbery
Robbery_f
Aggassault
Aggassault_f
Burglary

Mean (StD)
4.9 (12.7)
3.4 (9.6)
12.4 (20.4)
1.4 (3.0)
30.6 (40.1)
1.2 (3.0)
159.2 (340.8)
88.1 (215.0)
213.4 (443.0)
63.6 (163.5)
556.2 (916.9)

Descriptive statistics for quarterly city data


Note: Here we divide all kinds of crimes into two columns. Crimes on the left column are the ones
defined as other crimes that are not involved with display of firearms in any case. While crimes on
the right column are defined as violent crimes that may be involved with firearms.

Ln
firearms(t-1
)
Ln
fondling(t)
Ln
simassault(t
)
Ln theft(t)
Ln
mvtheft(t)
Ln drugs(t)
Quarter
fixed effect
City fixed
effect

Panel Regression Findings


Table 2
Baseline model, city-level data
Ln(ho Ln(ho Ln(ki Ln(ki
micid micid dnap) dnap)
e)
e)
Yes
Yes
Yes
Yes

Ln(ra
pe)
Yes

Ln(r
ape)
Yes

Ln(R Ln(R Ln(a


obber obber ggass
y)
y)
ault)
Yes
Yes
Yes

Ln(a
ggass
ault)
Yes

Ln(burgl
ary)
Yes

Yes
Yes

Yes

Yes

Yes

Yes
Yes

Yes

Yes
Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes
Yes
Yes

Yes

Yes
Yes

Yes

Yes

Yes

Yes
Yes

Yes

Yes
Yes

Yes

Yes

Yes

Yes

Note: Here we use both quarter and city fixed effect because it could be a serious issue using

random effect when other minor variables are correlated with the main variable, which is lagged
firearms. That will lead to biased and inconsistent results, which could be confusing. Hence, using
fixed effect which is generally more robust would be safer.

Table 3
Effect of stolen firearms on homicide
Ln(homicide)

Ln(homicide)

Ln(homicide_f)

Ln[Firearms(t-1)]
Ln(mvtheft)
Ln(drugs)

0.2008(0.0083)

0.1003(0.0090)
0.1500(0.0100)
0.0766(0.0080)

0.0952(0.0086)
0.1105(0.0096)
0.0579(0.0076)

Quarter Fixed
County Fixed

Yes
Yes

Yes
Yes

Yes
Yes

Adjusted R^2

0.8162

0.8242

0.8093

Ln(kidnap)

Ln(kidnap)

Ln(kidnap_f)

Ln[Firearms(t-1)]
Ln(mvtheft)
Ln(drugs)

0.2654(0.1001)

0.0828(0.0106)
0.2661(0.0118)
0.1454(0.0094)

0.0600(0.0085)
0.1012(0.0094)
0.0244(0.0075)

Quarter Fixed
County Fixed

Yes
Yes

Yes
Yes

Yes
Yes

Adjusted R^2

0.8116

0.8299

0.6689

Ln(rape)

Ln(rape)

Ln(rape_f)

Ln[Firearms(t-1)]
Ln(simassault)
Ln(drugs)

0.2333(0.0078)

0.0604(0.0079)
0.2779(0.0089)
0.0814(0.0081)

0.0692(0.0084)
0.0776(0.0094)
0.0176(0.0087)

Quarter Fixed
County Fixed

Yes
Yes

Yes
Yes

Yes
Yes

Adjusted R^2

0.8600

0.8829

0.6678

Ln(robbery)

Ln(robbery)

Ln(robbery_f)

Ln[Firearms(t-1)]
Ln(theft)
Ln(drugs)

0.4244(0.0092)

0.1528(0.0085)
0.4215(0.0095)
0.1435(0.0088)

0.1658(0.00104)
0.3180(0.0106)
0.1250(0.0107)

Quarter Fixed
County Fixed

Yes
Yes

Yes
Yes

Yes
Yes

Adjusted R^2

0.9224

0.9451

0.9219

Table 4
Effect of stolen firearms on kidnap

Table 5
Effect of stolen firearms on rape

Table 6
Effect of stolen firearms on robbery

Table 7
Effect of stolen firearms on aggassault
Ln(aggassault)

Ln(aggassault)

Ln(aggassault_f)

Ln[Firearms(t-1)]
Ln(fondling)
Ln(simassault)
Ln(theft)
Ln(mvtheft)
Ln(drugs)

0.4247(0.0084)

0.0702(0.0070)
-0.0067(0.0058)
0.1799(0.0096)
0.2873(0.0123)
0.2251(0.0098)
0.0971(0.0072)

0.1349(0.0096)
0.0264(0.0080)
0.2130(0.0132)
0.1030(0.0169)
0.2004(0.0135)
0.0345(0.0099)

Quarter Fixed
County Fixed

Yes
Yes

Yes
Yes

Yes
Yes

Adjusted R^2

0.9007

0.9466

0.9217

Note: All of the variables in table 1-7, are statistically significant in 10%, 5%, 1% level
respectively.

Fixed effect, endogeneity. Perfect multicollinearity, generate ln(rape-rapef) and study firearms and
non-gun crimes. All viriables +1. Pick the biggest ratio

You might also like