The study of growth and decline of violent crime and property crimes has become increasingly important. Despite reports of decline in crime rates according to some reports, there still were about 9,063,173 property crimes in 2011; that is, an estimated rate of 2908.7 per 100,000 residents, amounting to an estimated loss of $15.6 billion (FBI, 2011). Also, despite a 4.1% reduction in property crimes in 2013, a total of an estimated 8,632,512 occurred in 2013, which amounted to 2730.7 per 100,000 residents, with an estimated total cost of $16.6 billion. Specifically, 69.6% of the overall property crime in 2013 was larceny, burglary amounted to 22.3%, and vehicle theft amounted to 8.1% (FBI, 2013). Looking at the long-term trend of property crimes, one would notice a seemingly important decrease in the phenomenon. Particularly, property crimes in the United States declined to about 2.6% between 2014 and 2015. It still amounted to a loss of $14.3 billion. A 10-year trend also revealed a 20.2% decline in property crime between 2005 and 2015. Yet, the total number of property crimes that occurred within the period remained high. Currently, based on 2018 data, it is estimated that 7196.045 incidences of property crimes occurred in the United States, with a 6.3% decline as compared to the 2017 total estimate. The 10-year trend also showed a decline of 22.9% in property crimes. The estimated total cost of property crimes in 2018 was $16.4 billion (FBI, 2018).

Violent crimes negatively impact the health, economic, and social well-being of individuals, families, governments, and institutions. According to the preliminary Uniform Crimes Report for 2020 (FBI, 2020), offenses such as nonnegligent manslaughter and murder increased by 14.8%, in the first 6 months of 2020, while aggravated assault also increased by 4.6%. In all, cities with less populations (<10,000) reported decreases in violent crimes (7.2%) while those with larger populations (100,000 to 249,000) had lower decline in violent crimes (0.3%). Despite the reported decreases in violent crimes for the time frame under consideration (first 6 months of 2020), a large portion of the US population continues to struggle with the direct and indirect consequences of violent crimes. It is specifically troubling that what might be considered the most heinous forms of violent crimes—murder and nonnegligent manslaughter—have continued to be on the increase. Although existing information on total crime estimates in the United States is not current, the total cost is estimated at $2.6 trillion. This amounts to $620 billion cash costs including $1.95 trillion towards losses to quality of life, with an uncertainty interval of 95% ($2.2 – $3.0 trillion, Miller et al., 2020). According to Miller et al. (2020), violent offenses resulted in 85% cost of crimes in the United States, with violent crimes accounting for 85% of the costs.

Scholars have highlighted the relationship between violent crime and property crime. For example, Tcherni, Davies, Lopes, and Lizotte (2016) have noted that “one indirect consequence of property crimes occurring online could be a drop in the violent crime rate” (p. 907). Also, examinations of patterns of criminal behavior have exposed some connections between violent crime and property crimes (Lussier et al., 2005; Shover, 1996). In fact, most offenders do have prior histories that are not limited to a category of crime (Piquero, Farrington, & Blumstein, 2003). A phrase captured by previous scholars “offense versatility” seems to connote the associations between and among different forms of crime (Deane et al., 2005; Shover, 1996; Steffensmeier & Ulmer, 2005; Sullivan et al., 2006). Although both violent crime and property crimes are offenses, the differences between the two are quite clear. But interestingly, studies assessing how violent crime predicts property crimes are not many. More so, studies investigating growth trajectories for property crimes based on violent crime, and their effects being mediated by a marker of prevalence of poverty within a geographical setting are scarce. In terms of public nutritional assistance, no study known to the author has explored its association with violence and crime. Public nutritional assistance is administered in the United States. According to US Food and Nutrition Service (2022), “SNAP provides nutrition benefits to supplement the food budget of needy families so they can purchase healthy food and move towards self-sufficiency.” Some show associations between hunger and violence and other forms of crime (Islam, 2016). It is also important to explore whether an intervention that is aimed at ameliorating hunger would negatively correlate crime.

Reports of decline in property crimes have been noted both in the United States and internationally (Button & Cross, 2017; FBI, 2011, 2013, 2015, 2018; Tcherni et al., 2016; Van Dijk, Tseloni, & Farrell, 2012). In the United States, some decline in property crimes has been observed since the 1980s (Bureau of Justice Statistics, 1994; Walters, Moore, Berzofsky, & Langton, 2013). Violent crimes have also been observed as declining nationally since the 1990s (Truman & Planty, 2012). However, online crimes have increased (Caneppele & Aebi, 2017). Van Dijk et al. (2012) see the decline in property crimes because of the increase in cybercrimes. It is not easily dismissible because the online environment has increasingly created a new environment for human activities (Caneppele & Aebi, 2017), thereby possibly creating new forms of crime (Button & Cross, 2017), Levi (2017). Levi pondered whether some of the decline is due to the overlap between offenders and previous offenders outside the online environment, or as Tcherni et al. (2016) would suggest, due to an increase in online property crimes. Farrell and Birks (2018) reviewed available evidence related to the reasons for decline in crime and concluded that cybercrime is not responsible for the decline in property crimes. Rather, they suggested that crimes, crime decline, or rise are dependent-less trends resulting from wide-range adjustments to the structures upon which crime opportunities exist. The authors pointed to improvements in security as the reason for decline in physical crime, consistent with the findings of Kriven and Ziersch (2007) and Farrell et al. (2014).

In 2011, 2013, and 2015, the setting of the study reported important trajectories in terms of violent crime and property crime incidences. In those years, property crime rates per 1,000 residents in the study’s location were 38.5, 33.5, and 31.8, respectively. Violent crime rates per 1,000 residents in the county for the same years were 3.9, 4.4, and 4.2, respectively. Based on the most recent crime statistics for the study’s setting, there was a 4.9% increase in property crime in 2019 and an 11.5% increase in violent crimes in the study’s setting, in 2019. Although the national trend in property and violent crimes seem to be declining, except for murder and aggravated assault and arson (FBI, 2020), the increasing trend of violent crime and property crimes in the study’s setting remained a concern.

Scholars of several backgrounds have applied several frameworks in the study of violent crime and property crime, and their correlates (Cohen & Felson, 1979 Pratt, Holtfreter, & Reisig, 2010). Many of these frameworks have their foundations in criminology, sociology, psychology, and economics. This study considers the empowerment perspective in exploring how poverty relates to violent crime and property crimes. This framework is used for two main reasons: First, it is a perspective that aligns with the field of social work. Second, the empowerment perspective has been used in the study of community-level factors across various disciplines (Webster & Perkins, 2000; Zimmerman & Rappaport, 1988); as such, it allows scholars and policy practitioners to assess individuals and their challenges from a strengths-based perspective. Third, it aligns with the grand challenges of social work, which include ending family violence, reducing extreme economic disequilibrium, and encouraging decarceration in a smart way. Fourth, it aligns with the 2022 educational policy and accreditation standards (Council on social work Education, CSWE, 2023), which recognizes the need to align social work curriculum to current challenges facing humanity including violence and social determinants of health. The empowerment perspective has been described as contextually situated and nonstatic (Perkins & Zimmerman, 1995).

Public nutritional assistance is an empowerment-based policy that is recognized in social work. It is largely aimed at empowering individuals that are poor whose benefits from receiving the assistance may be far beyond nutrition. This is partly demonstrated in Lord and Hutchisons’ (1993) study, which sought to understand the lived experience of disempowered community members and the stages they passed through to achieve empowerment. They found that nearly all the participants identified unresponsive services and systems (or resources) as being a source of their problems. The authors reported two types of resource-related problems. The first problem was described as systemic failures via “neglect” while the second problem was described as failure by “inappropriate interventions” (Lord & Hutchison, 1993, p. 8). Interestingly, factors which the authors identified as disempowering included poverty. As endorsed by Keifer (1984), empowerment is concerned with an individual’s capacity to feel capable and competent to participate in community, among other things.

This author adds that empowered individuals and communities are able to stand against stigma of all forms, especially poverty and mental illness. Two assumptions are important for this study, based on the empowerment perspective: (1) individuals are capable of being empowered and (2) community goals including safety will be achieved as a result of individuals being empowered and coming to work together (Haddad & Toney-Butler, 2023). Hence, the author hypothesizes that public nutritional assistance will have an inverse relationship with violent crime and property crime. Poverty is a key social determinant of mental health. Existing evidence suggests that poverty can be both a cause and effect of mental illness (Knifton & Inglis, 2020; Shim et al., 2014). Evidence suggests that social and economic factors do affect human development, behavior, and relationships (Allen, Balfour, Bell, & Marmot, 2014), and these outcomes have implications for the mental health for adults and children (Cooper & Steward, 2015). Evidence also shows that, despite all of the above, stigmatization of individuals with mental illness and those seeking assistance remains a problem (Turan et al., 2019) such that stigma is also a source of health inequality (Hatzenbuehler, 2013). Also, empowerment has been shown to negatively associate with crime and violence (Hockley & Barden, 2023), while stigmatization of people experiencing poverty may play a negative role in terms of self-image, traumatic response, and management at the individual level. We know that many individuals that are involved in crimes have violent and traumatic backgrounds. In a study by Jaggi, Mezuk, Watkins, and Jackson (2016), it was found that trauma exposure was associated with violence and crime. The study also found that mental health and its comorbidities were correlates of incarceration. These realities have necessitated calls for exploration of the linkages between trauma and criminal involvement from an empowering perspective. The main question for this study was: How does prevalence of public nutritional assistance, which is an indicator of empowerment relate to the prevalence of violent crime and property crime? The author used logistic regression to assess how prevalence of violent crime and prevalence of public nutritional assistance predicted property crime.

This study was based on a secondary analysis of data from the Quality-of-Life, (City of Charlotte, 2019) project. Due to the nature of the data, individual-level descriptions were unavailable for analysis. Given this, the study was based on NPAs. The study’s setting had 4.9% increase in property crime in 2019 and an 11.5% increase in violent crimes in the setting, in 2019. Specifically, all forms of violent crimes showed an increase compared to 2018 data: homicide (n = 108, 89.5%), robberies (n = 2,000, 12.7%), and aggravated assaults increased by (n = 4,559, 11.2%) except rape, which recorded a decrease of 5.6%, with a total of 288 incidences. All forms of property crimes except arson (n =148, −5.1%) showed increase compared to 2018 data: burglary (n = 5,428, 0.3%), motor vehicle theft (n = 2,922, 2.1%), and larceny (n = 27,577, 6.2%) (City of Charlotte, 2019). The current increase in both violent crime and property crimes in the location of this study brings about the importance of investigating the relationships between violent crime and property crimes, and how that relationship might be linked with the growth or decrease in property crimes.

Kline (2015) also indicates a ratio of 20 cases to 1 item, while Schreiber et al. (2006) recommend 10 cases to 1 item ratio as enough sample size for studies involving factorial analysis and structural equations. Other scholars have also indicated lower ratios as acceptable. For example, Bentler and Chou (1987) accept a ratio of five cases to one item acceptable. The sample size for this analysis was 452. Five cases were removed from the analysis due to missing more than 70% responses within the dataset. The overall dataset had missing cases and maximum likelihood, a method that ensures all nonmissing data points were considered in the analysis as recommended by Little and Rubin (1987).

The dataset contained more than 80 variables at the NPA level, including violent crime and property crimes. The NPAs were treated as nominal and identification variables, property crime was treated as a continuous variable comprising the annual total for arson, attempted burglary and burglary, motor vehicle theft, and larceny, divided by the population and multiplied by 1,000. According to the Federal Bureau of Investigations’ (FBI) Uniform Crimes Report (FBI, 2011, 2013, 2015, 2019), offenses, such as larceny-theft, burglary, arson, and motor vehicle theft, are considered property crimes. Furthermore,

Similarly, violent crime was treated as a continuous variable comprising annual total for armed robbery, aggravated assault, homicide, armed robbery, rape and attempted rape, and negligent manslaughter. The crimes data were originally sourced from six coordinating police departments. Public nutritional assistance (PNA) was also applied as an independent variable, with violent crimes, to assess economic mediator roles between violent crime and property crimes. The variable PNA was derived by dividing the total percentage of the population receiving PNA by the total population. Job density at baseline was used to assess the mediational impact on the relationship between violent crime and property crime. Job density was obtained by dividing the number of total jobs by the total land area.

In conducting the logistic regression, the continuous variables were recoded into binary forms, using the mean as a cutoff between a high and low category.

Summary statistics were calculated for public nutritional assistance (PNA, an indicator of poverty), Job density (JD), property crime (PC), and violent crime (VC). The observations for VC had an average of 1.65 (SD = 1.07, SEM = 0.05). The observations for PC had an average of 3.75 (SD = 1.10, SE_M = 0.05). The observations for PNA had an average of 2.44 (SD = 1.04, SE_M = 0.05). The observations for JOBD had an average of 2.38 (SD = 6.85, SE_M = 0.32). The summary statistics can be found in Table 1.

Summary statistics were calculated for public nutritional assistance (PNA, an indicator of poverty), Job density (JD), property crime (PC), and violent crime (VC) for logistic regression purposes. The most frequently observed category of PC within the 0 category of PNA was 0 (n = 156, 68.12%). The most frequently observed category of PC within the 1 category of PNA was 1 (n = 158, 71.49%). The most frequently observed category of JOBD within the 0 category of PNA was 0 (n = 130, 56.77%). The most frequently observed category of JOBD within the 1 category of PNA was 1 (n = 116, 52.49%). The most frequently observed category of PNA within the 0 category of PNA was 0 (n = 229, 100.00%). The most frequently observed category of PNA within the 1 category of PNA was 1 (n = 221, 100.00%). The most frequently observed category of VC within the 0 category of PNA was 0 (n = 184, 80.35%). The most frequently observed category of VC within the 1 category of PNA was 1 (n = 181, 81.90%). Frequencies and percentages are presented in Table 1.

In this analysis, the author sought to answer the following question: Is there a statistically significant association between property crime and violent crime, while controlling for public nutritional assistance? A partial correlation analysis was conducted for property crime and violent crime while controlling for public nutritional assistance and filtered by job density. Cohen’s (1988) standard was applied in assessing the coefficients. According to Cohen (1988), the strength of coefficients or effect sizes range from small (0.02) to medium (0.15) to large (0.35). For this first analytical group filtered by low job density (0), findings showed that the relationship between violent crime and property crimes was statistically significant (r = 0.43, p < 0.001, 95% CI [0.32, 0.54]), with a medium effect size. This suggests that both violent crime and property crimes increased, controlling for Public Nutrition Assistance. For the high category of the filtered group (job density), a partial correlation analysis was conducted between property crime at wave 1 and violent crime, controlling for public nutritional assistance, and filtered by job density (1). For this first analytical group filtered by low job density (1), findings showed that the relationship between violent crime and property crimes was statistically significant (r = 0.53, p < 0.001, 95% CI [0.43, 0.62]), with a larger effect size compared to the first analytical group (job density, 0). This suggests that both violent crime and property crimes had a higher increase, controlling for public nutritional assistance.

The model was assessed based on an a priori of 0.05. The following equation was applied in the logistic regression: ln(p1−p)=β0+β1x1+β2x2+β1x1×β2x2 Where ln(p/1−p)= outcome: property crime, β0= constant, β1x1= public nutritional assistance, β2x2= violent crime, β1x1*β2x2= interaction term for public nutritional assistance and violent crime.

The overall model was significant, χ2(3)=230.98, p<0.001, indicating that public nutritional assistance and violent crime at the baseline had a significant effect on the odds of observing high prevalence (1 category) of property crime. McFadden’s R-squared was calculated to examine the model fit, where values greater than 0.2 was indicative of models with excellent fit (Rolfe, Bennette, & Louviere et al., 2000). The McFadden R-squared value calculated for this model was 0.39. The regression coefficient for the high category of public nutritional assistance at baseline was not significant, B=−0.47, OR=0.62, p=0.408, indicating that high prevalence of public nutritional assistance (1 category) did not have a significant effect on the odds of observing high prevalence (1 category) of property crimes. The regression coefficient for violent crime (1 category) was significant, B=2.46, OR=11.70, p<0.001, indicating that for a one unit increase in high prevalence of violent crimes, the odds of observing high prevalence of property crime grew by approximately 1,070%. The interaction between public nutritional assistance and violent crime had a significant effect on property crime, B=1.61, χ21=5.42, p=0.020. This suggests that moving from the low (0) to high (1) category of public nutritional assistance strengthened the relationship of high property crime to high violent crime. Table 2 summarizes the results of the regression model.