econ, keep the project (HW5)

Nan Zhou

Econ 490

Professor Nancy

3/23/2017

-Need a more complete interpretation of you empirical result, in which you discuss magnitude and economic as well as statistical significance.

-Given Table 4, you need to run same models with fixed effects and report those.

Does the economic have an affect on the crime

Introduction

The paper will research the economic project, which looks into the effect of the median income in the determination of crime. therefore, the paper will look into the literature review, the model section and empirical results. The other part of the report shall look into the data to be used in the sample study. Finally, the paper will consider the empirical results, which include, regression results for at least one given model. Consequently, the paper will provide interpretation and discussion of the results obtained from the regression models.

The hypothesis of the project involves that an economy is able to affect the crime rate. In other words, the paper will test to see if the economic conditions are able to affect the crime rates for citizen. As a result, there are several variables that will be evaluated in the model to determine the stand in the above hypothesis. The variables include, economics(median income), unemployment rate, race, and age. The types of crimes that are violent crimes and property crimes.

Literature Review

there is the evaluation of the past studies that have been done and they are apparently similar or related to the current study. According to Goulas and Zervoyianni, (Zervoyianni, 2013) they argue that where is increased crime rates there’s an asymmetrical impact on the development in the economy. As seen, the report looks at the other angle of crime affecting economics and hence provides insights on the counterargument that can be developed in the research.

The other paper that we can refer to the argument by Randi and Lance (Lochner, 2012), that propose that the education levels are far more impactful on the crime than ever considered. They argue that the initial policies to eradicate crime focus on the punishment and retribution instead of embarking on education policies that can be more efficient in the determination of lower crime levels. They conclude that there is very close correlation of the crime and the education levels of a particular set of population statistics.

The other research paper that is relevant in the study is the discussion by Benjamin, Manish and Nair (Benjamin Powell, 2010). This is a similar paper to the analysis we had in the first paper. The counterargument that the crime is the variable that relates to the economy. They are of opinion that certain crimes such as corruption can both increase or decrease economic growth. However, the comparison between the different nations was difficult, owing to the fact that it is difficult to ascertain what is crime in one nation and what is crime in another nation. This was the biggest challenge in the research.

The Regression Model.

The use of the model is to enable, generalization of the crime in terms of the variables that are considered to affect it. It allows the calculation of any given crime rate provided the coefficients affecting the impact of the variables in the given model. The model is described below:

In order to justify the variables, we only need to look into the effect the said variables could have on the crime rate. As we have seen, from the past results, in the studies covered by other statisticians, there is a lot significant involvement in the chosen variables in determination of the crime rates. ( Need to discuss economic theory behind some of your perdictions for signs of

The coefficient estimates.)

Variable

Dependent Variable:

Crime

Key Independent Variables:

Economic- (real GDP& median income)

Other Control Variables:

U= Unemployment rate

R=race: black, Asian, white

A=Age

%age 15-29

%age 30-54

%age 55-69

Model Prediction table:

Model

Predictions

y/ x

2y/ x2

2y/ xz

Dependent Variable

Crime

n.a.

n.a.

Key Independent Variables:

Economic - (real GDP)

1>0

n.a.

n.a.

Other control Variables

n.a.

n.a.

Education background

>0

n.a.

n.a.

Income

>0

n.a.

n.a.

Unemployment

>0

n.a.

n.a.

Race

>0

n.a.

n.a.

Age

>0

n.a.

n.a.

Table 2

Mean

Std-Dev

Min

25th%

median

75th%

max

Violent Crimes

355.504

129.8843

99.3

255.55

329.85

245.55

705.2

Property Crimes

2746.686

599.4182

1406.6

2226.43

2682

3255.22

4015.3

Real GDP

46925.79

8955.848

31169

40137.25

45474

51910.50

73464

Median Income (dollars)

52804.3

8703.047

36796

46655

50723.5

58297.25

74551

Social welfare

Unemployment Rate

0.103474

0.453161

0.029

0.07

0.079

0.09

8.55

%White

0.802063

0.115777

0.269

0.719

0.815

0.88825

0.972

%Black

0.111174

0.095412

0.006

0.035

0.083

0.161

0.38

%Asian

0.04538

0.076554

0.006

0.017

0.027

0.047

0.573

Age 15-29

0.070126

0.003803

0.057

0.203

0.07

0.213

0.082

Age 30-54

0.064129

0.004696

0.054

0.328

0.064

0.349

0.079

Age 55-69

0.071577

0.00562

0.052

0.154

0.072

0.172

0.088

2-3 decimal places only.

Need to discuss some key summary

statistics from Table 2.

In this section, the paper will research the description of data. [From these regression results, the interesting points is that the real GDP and income have a negative coefficient on the model. This is meant to reveal that the increase in these variables are meant to reduce the crime rates particularly the property crime rates. The meaning that the median income and real GDP are but not statistically significant determinant of property crime. The comparison of the different data from the two models, (one for violent and another for the property) the property crime is important than violent crime. The economy is a not a very strong determinant of violent crime as compared to the property crime.

Table 3. Empirical Results

Dependent Variable: Property Crime

Coefficient

t Stat

P-value

Real GDP

-0.008966

-0.08

0.933

Median Income (dollars)

-42.5936

-0.83805

0.165

Unemployment Rate

674.3145

1.16944

0.243055

%White

541.4287

0.808612

0.41931

%Black

-249.709

-0.02876

0.977076

%Asian

-13911.7

-1.87635

0.061474

Age 15-29

-107.56

-1.19

0.241

Age 30-54

151.98

-1.23

0.226

Age 55-69

-258.23

-2.04

0.047

Number of Observations

350

R2

0.7539

Report coefficient estimates with standard error below in parantheses.

To compare the results, we can undertake comparison with the violent crime as the dependent variable and the same input independent variables. The output is as follows:

Dependent Variable: Violent Crime

Coefficient

t Stat

P-value

Real GDP

-0.0006

-0.66904

0.503931

Median Income (dollars)

-2.49009

-0.22415

0.822776

Unemployment Rate

674.904

-1.9987

0.046446

%White

-113.21

-7.78393

%Black

-735.21

-3.87626

0.000128

%Asian

-50.31

-3.1318

0.00189

Age 15-29

13766.16

4.827639

0.00109

Age 30-54

-229.2

-5.51522

0.001253

Age 55-69

-99.85

-3.17054

0.001662

Number of Observations

350

R2

0.499185

In terms of the hypothesis, it can be well noted that there is a negative correlation between the crime and the economics. It is apparent that the increase in the GDP and the median income results to decline in the rates of crime of property. It also shows that the property crime is more controlled by the economy as compared to the violent.

Table 4: *Need to interpret magnitude of coefficient estimates, and discuss economic

And statistical significance.

Hausman Test

Chi^2

P value

conclusion

Ho= re=fe

45.38

reject

F-Test

F stat

P value

conclusion

Ho:all FE coef are zero

41.74

reject

Need to discuss the fact that Table 4 implies your models should include fixed effects.

Table 5:

Dependent Variable: Property Crime

Alternative model 2

Alternative model 3

Median Income (dollars)

-.0000175

-.8792437

Real GDP

1.75e-06

1.98e-06

Unemployment Rate

-.0249207

-.0247894

%White

.6818503

.3872131

%Black

-1.846826

-2.035659

%Asian

2.50183

2.58868

Age 15-39

4.246971

4.097063

Age 40-69

-7.093295

-6.959051

Number of Observations

350

350

R2

0.4695

0.4639

Need to explain these models and interpret the results as compared to your earlier results

Also need to report standard errors.

*At least some of your models need to be run with fixed effects.

References

Benjamin Powell, G. M. (2010). Corruption, Crime and Economic Growth . Benson Print.

Lochner, R. H. (2012). The Impact of Education on Crime: International Evidence . Dice Report.

Zervoyianni, E. G. (2013). Economic growth and crime: does uncertainty matter? Retrieved from Econ Papers: http://econpapers.repec.org/article/tafapeclt/v_3a20_3ay_3a2013_3ai_3a5_3ap_3a420-427.htm