posc data

Results Section

Towson University

POSC Research 1

April 27th, 2026

To test the hypotheses about the rates of violent crime in the U.S. states I performed three

bivariate regressions, and one multiple regression. The rate of violent crime per 100,000

residents was the dependent variable in all the models. The independent variables included

income inequality (Gini coefficient), poverty rate and police expenditure per capita. According to

these tests, it is possible to find out whether each of the factors is separately related to violent

crime and whether the relations are present when all the factors are considered.

The first bivariate regression was adopted in order to conduct the test on the

association between the income inequality and the violent crime rates. The findings showed that

there was a positive and significant relationship. The slope of the Gini coefficient was positive

and it implied that the higher the income inequality of a state, the higher was the likelihood of

the state reporting a higher violent crime. The t-score of the slope was above the customary level

of significance and the p-value was less than.05 and it suggested that the association was not

probably based on chance. The R2 that indicated that income disparity was the key variable in

the varying rates of violent crime among states was observed. These results support the initial

hypothesis according to which increasing income inequality is linked to an increase in violent

crime. This is also in line with the point made that the skewed income distribution can lead to

undermining of the social cohesion and disappointment or relative deprivation, which is a state

that can lead to crimes (Zhuang et al., 2025).

With the help of the second bivariate regression, the connection between the rate

of poverty and violent crime was tested. A positive and statistically significant slope coefficient

was also obtained in this analysis. Such high rates of violent crime were also more likely to be

high in the states where a larger percentage of the population was below the federal poverty line.

The t-score was significant and less than.05 was the p-value. One of the measures that poverty

was a moderate predictor of the state level variance in violent crime was the R 2 value. This way,

the second hypothesis was also proven to be correct. This result is in line with the literature that

has taken time to consider structural disadvantage as one of the major determinants of the crime

rate particularly in areas where economic distress is high (Ulmer et al., 2012).

The third bivariate regression was on the police expenditure per capita and violent

crimes rates. The slope coefficient had a negative value and, therefore, a smaller expenditure on

policing in a state would tend to provide relatively lower levels of violent crime. The relationship

was however not as strong as the two earlier models. In other cases, the coefficient was large yet

less compared to the impacts of inequality and poverty. The R 2 was also very low and this fact

means that the proportion of the variation in violent crime that was accounted for by the police

expenditure alone was much lower than the proportion of the variation in violent crime that was

accounted for by the socioeconomic variables. This partly supports the third hypothesis.

Although the policing resources are thought to lead to a decrease in crime, when they are looked

at in isolation then they are found to be not as effective as the overall state of structure.

To simultaneously test all the hypotheses, I estimated a multiple regression model,

which included income inequality, poverty rate, and police spending per capita. The intercept of

the regression line gives the value of the violent crime rate which is likely to occur in a scenario

in which all the independent variables are zero. This value has little or no practical meaning

since a zero value is not viable as a measure of such, but it is the constant of the model.

The entire model was positively correlated between the income inequality and the violent

crime, and the correlation was found to be statistically significant. Keeping poverty and

expenditure on the police constant, an increase in the Gini coefficient was related to an increase

in the violent rate of crime. This implies that inequality has an independent impact other than

poverty. The time element of this relation implies that the absence of absolute deprivation is not

the only element that can dictate the patterns of crime, yet resource inequity also could be used to

define crime.

The multivariate model also found a positive relationship between the level of poverty

and violent crime with a coefficient of slightly less than that between the bivariate regression.

This is an indication that there is also an intersection of an element on the impact of poverty with

the impact of inequality as even the poorer states can be characterized by the increased degree of

income inequality. However, the statistical significance of the effect of poverty was also

significant, which suggests that concentrated hardship is a significant predictor of violent crime.

The police per capita expenditure was not positively correlated with the multiple

regression but the correlation was not high and significant relative to the socioeconomic

variables. Further police spending after the taming of inequality and poverty was connected to

the lowest violent crime rate by an insignificant amount, although the outcomes were poor. This

renders the strategies of enforcement as potentially significant, but the effects of more

fundamental economic and social conditions are not entirely offset.

The multiple regression model had a significant R 2 when compared to bivariate

regressions R 2. This means that the three variables explained a higher percentage of change in

the rates of violent crime in the states than did a single variable. That is, violent crime can best

be considered a multidimensional phenomenon, which can be influenced by both economic

inequality and policy decisions.

In general, the results are the most persuasive to the hypothesis of income

inequality. The greater the degree of inequality, the greater the level of violent crime with the

bivariate regression as well as with the multiple regression. Another very popular hypothesis is

the poverty hypothesis but to some degree some of its impacts were similar to inequality. The

hypothesis on police spending was less strong; high spending was linked to a slightly lower

crime and the correlation was less strong and less enduring than the structural economic factors.

At a minimum, technical results show that it is not only the amounts of policing

that dictate the violent crime rate. Although factoring police spending, violent crime is higher in

those states with higher economic inequalities and poverty. These results are consistent with the

majority of the literature that social and economic disadvantage can be an even better predictor

of crime compared to enforcement (Ulmer et al., 2012). Hence, the policies to alleviate poverty

and inequality are equally critical in alleviating violent crime in the U.S. states, just like criminal

justice spending.

References

Crimmins, E. M. (2025). Life expectancy and health expectancy in the twenty-first

century: The unthinkable, the inconceivable, and the unknowable.

Demography. https://doi.org/10.1215/00703370-12185960

Dieleman, J. L., Weil, M., Beauchamp, M., Bisignano, C., Crosby, S. W., DeJarnatt,

D., Lescinsky, H., Mokdad, A. H., Ostroff, S., Paul, H., Pollock, I., Sahu,

M., Scott, J. W., Taylor, K. V., Thomson, A., Weaver, M. R., Wilner, L.

B., & Murray, C. J. L. (2025). Drivers of variation in health care

spending across US counties. JAMA Health Forum, 6(2), e245220.

https://doi.org/10.1001/jamahealthforum.2024.5220

Ding, D., Nguyen, B., Nau, T., Luo, M., del Pozo Cruz, B., Dempsey, P. C., Munn,

Z., Jefferis,B. J., Sherrington, C., Calleja, E. A., Hau Chong, K., Davis,

R., Francois, M. E., Tiedemann, A., Biddle, S. J. H., Okely, A., Bauman,

A., Ekelund, U., Clare, P., & Owen, K. (2025). Daily steps and health

outcomes in adults: A systematic review and dose- response

meta-analysis. The Lancet Public Health, 10(8).

https://doi.org/10.1016/s2468- 2667(25)00164-1

Guth, M., Garfield, R., & Rudowitz, R. (2020). The effects of Medicaid expansion

under the

ACA: Studies from January 2014 to January 2020 | KFF. KFF.

https://www.kff.org/medicaid/the-effects-of-medicaid-expansion-under-t

he-aca-updated-findings-from-a-literature-review/

Lescinsky, H., Beauchamp, M., Bisignano, C., Crosby, S. W., DeJarnatt, D., Sahu,

M., Taylor, K. V., Thomson, A., Weil, M., Mokdad, A. H., Scott, J. W.,

Murray, C. J. L., & Dieleman, J. L. (2025). Evaluating US county

health-care system performance and key associated factors (2014–19): A

triple aim framework analysis. The Lancet Public Health, 10(9),

e741–e750. https://doi.org/10.1016/S2468-2667(25)00173-2

Lescinsky, H., Sahu, M., Beauchamp, M., Crosby, S., Johnson, E., McHugh, T. A.,

Scott, J. W., Schulman, K., Thomson, A., Weil, M., Dieleman, J. L., &

Milstein, A. (2025). Exploring state‐level change in health care value

over three decades in the United States, 1991–2020. Health Services

Research. https://doi.org/10.1111/1475-6773.70054

Ulmer, J. T., Harris, C. T., & Steffensmeier, D. (2012). Racial and ethnic disparities

in structural disadvantage and crime: White, Black, and Hispanic

comparisons*. Social Science Quarterly, 93(3), 799–819.

https://doi.org/10.1111/j.1540-6237.2012.00868.x

Wagner, D. A., Nair, T., Thapa, A., & Kumar, A. (2025). The Gini learning index: A

new framework to measure learning inequality across contexts.

International Journal of Educational Development, 119, 103433.

https://doi.org/10.1016/j.ijedudev.2025.103433

Woolf, S. H. (2023). Falling Behind: The growing gap in life expectancy between the

United States and other countries, 1933-2021. American Journal of

Public Health, e1–e11. https://doi.org/10.2105/ajph.2023.307310

Wu, Z., Qian, F., Zou, S., Zou, X., Zhang, R., Guo, X., & Li, H. (2025). Trends and

disparities in health status and health care in the United States during

COVID-19 pandemic. PubMed, 23(1), 478–478.

https://doi.org/10.1186/s12916-025-04315-4

Zhuang, Z., Xie, X., Huang, L., Yan, J., Cheng, S., & Xu, A. (2025). The impact of

relative deprivation on mental health among middle-aged and older

adults in China: A multiple chain mediation model. Scientific Reports,

15(1). https://doi.org/10.1038/s41598-025- 20182-8