An Exploratory Study of Environmental Stress in Four High Violent Crime Cities: What Sets Them Apart?

Abstract

Understanding the social contexts of violent crime remains controversial in the literature. In the current study, we examine common social contexts in four cities (Houston, TX, Baltimore, MD, Jackson, MS, Wilmington, DE). Data were examined in two studies. In the first, each city was compared to national county-level data on health outcomes. In the second, communities within the four cities were examined for correlates of crime. Results suggest that some common social contexts emerge: high STD rates, air pollution, single-parent homes, insufficient food resources and sleep, residential segregation, housing cost burdens, comparatively few older adults and comparatively more females, were common among the four high-violence cities. By contrast, all four cities unexpectedly had uncommonly low suicide rates compared to the national average. At the community level, unemployment, community stress, median household income, and population density all correlated with criminal outcomes. High-violence cities tended to have higher proportions of Black residents, however community level evaluations suggested that class-related issues, not race per se, was correlated with violent crime.

Keywords

violent crime economics social stress race

Understanding social correlates of violent crime remains an important task. We sought to conduct an ecological investigation comparing factors correlated with crime in the Center for Justice and Research’s national study of several high-crime US cities. We investigate several different factors in this research and their interaction with violent and property crime. While we don’t wish to be limited to one theoretical approach to our work, there are subgroupings of work commenting on the effect of certain theoretical concepts (such as informal social controls; Sampson, 1986) from which we borrow regarding this endeavor. We first explain our approach and why we chose the factors we did, our methods in greater detail and what we found in light of the literature.

The ecological and factor prevention approach flows from social science which evolved a few decades ago (Bronfenbrenner, 1992). This first looks to the environment and identifiers for the area to determine certain attributes simultaneously occurring with or preceding crime. This approach was partially wrought from public health paradigm that predated its inception in criminology. Loeber and Farrington (1998) presented a detailed exposition of this paradigm as applied to serious and violent juvenile offenders. Empirical research continues to suggest that violent and property offenders face different factors from non-offenders and intervention efforts aimed at addressing these factors is critical for reducing initial crime or recidivism. For instance, regarding public health the wealth of literature, whether it be rooted in public health or criminology, suggest that neighborhoods with a higher proportion of persons who are healthier and engage in preferable health practice (Heckman et al., 2013; Otsu & Yuen, 2020) experience less crime. The same is true for some social factors. For example, the informal social controls and social cohesion that accompany less drinking (Snowden et al., 2020) and more two-parent families (Carr, 2003; Sampson, 1985, 1986, 1987; Sampson & Groves, 1989) relate to neighborhoods which experience less crime as well. There is a continuing body of research, stemming from the work of Sampson (1985) and others (Sharkey, 2006) to examine the connections between informal controls and crime. Research further comments on the associations of all three: health, social controls, and crime. High levels of informal controls within a community are associated with medically dangerous sexual behavior, mortality, eating behavior, and asthma (Browning & Cagney, 2002; Franzini et al., 2005; Lidfeldt et al., 2007; Locher et al., 2005).

More research is needed to better understand the activities and places in neighborhood environments that promote these social processes so that this knowledge can be integrated into neighborhood-level health promotion strategies. Sampson himself has commented on the relationship between these three and the importance of determining which of these variables is most predictive for social policy purposes (Sampson, 2003). In other words, by understanding the ecological contexts in which crime occurs, we may have a better understanding of how to prevent it in the future.

This exploratory approach concerns itself less with causes and more with correlates that appear with crime or other risk factors. In other words, the factors suggested by the research to accompany or precede crime are essentially that and beneficial for that designation without implying causal properties. This is exemplified in quite a few studies over the last few decades.

Individual risk factors of crime have been extensively researched. Most of these studies include consideration of age, gender, and race. Much of the literature suggests that violent crime begins between 14 and 18 years (Wolf et al., 2015). However, some studies have found the age of entry to be significantly younger, noting an average age of entry of 11 years. Regarding race, Black and Latino Americans tend to be overrepresented as perpetrators of violent crimes (Beck, 2021) although this may relate to social stress and class factors rather than race per se.

Structural causes have been studied as well. Crime is associated with disadvantaged neighborhoods, and the literature suggests a relationship between community factors and violent as well as property crime. The presence of violence and disorganization in the neighborhood also is associated with increases in crime involvement at the individual level. Likewise, the presence of adult support (as in two parent households, etc.) in a neighborhood and feelings of safety in a community are associated with less crime (Merrin et al., 2015). However, much of this literature can grow redundant or circular in reasoning and so other domains of risk have been explored from a greater ecological contextualism. These can be classified as (1) Health Practice and Health Outcomes as well as (2) Social, Economic Factors, and Physical environment.

Healthy Practice and Outcomes

Physical health outcomes have been found to correlate with crime. Unhealthy environments are associated with adverse mental health and psychological distress (Beck et al., 2013). Cross-sectional evidence from nationally representative data has explained that lower levels of physical and mental well-being has been associated with crime (Sundaresh et al., 2020). Further, lower levels of well-being (mental and physical) have been found to correlate with higher levels of exposure to police stops, arrests, and of course, incarceration (Hobbs et al., 2021). More specifically, longer durations of incarceration and multiple incarcerations have been associated with progressively lower well-being. Those who have been stopped and frisked by the police have reported low well-being similar to that of those who had been incarcerated (Sundaresh et al., 2020) though, again, these associations are correlative in nature, not necessarily causal. Further empirical work is needed to determine if well-being is associated with any exposure to police and similarly, more involved exposure suggests lower well-being and other impairments such as respiratory trouble, obesity, drug abuse, etc.

One of these impairments may be asthma morbidity. A relationship between asthma impairment and crime has been found in research regarding criminal offenders (Beck et al., 2013). Another may be obesity; higher neighborhood crime is longitudinally, positively associated with greater body mass index among neighborhood residents (Richardson et al., 2017).

However, it may not be that obesity is causing greater crime; this seems unlikely. Crime and obesity and other unhealthy practices/outcomes may be results of the same phenomenon. For example, greater community property crime correlates with obesity and physical inactivity (Yu & Lippert, 2016). Decreases in healthy food availability has been significantly associated with a higher violent crime rate and increased availability of supercenters and fast-food restaurants has been associated with a higher violent crime rate (McGuirt et al., 2019). Alcohol availability use is another factor. Myers et al. (2016) found a positive relationship between alcohol outlet density within neighborhoods and crime within neighborhoods after their review of the American Community Survey. Study at the local or county level for research on the role of the retail food environment in community violence when controlling for other sociological and health-related factors has shown a positive relationship between unhealthy food establishments and violent crime (Singleton et al., 2019).

Similarly, insufficient sleep is another common presence in certain circles that can have its own relationship with crime, both direct and indirect. Sleep disorders correlate with significant acute and long-term negative health outcomes. Further, insufficient sleep is associated with adverse physical, behavioral, and psychosocial consequences among youth and associated with a greater increase in criminal outcomes (Hildenbrand et al., 2013). Sleep problems may be important contributors to aggression, especially in the domains involving intimate partner violence, school and cyber bullying, and institutional aggression within psychiatric and correctional institutions (Krizan & Herlache, 2016).

Another community problem correlated with crime is mental health. Although arguments regarding causality may be made for both as causal agents in the relationship (Lorenc et al., 2012) and the effect of local crime on mental health (Dustmann & Fasani, 2016), reciprocity and the cycle it creates, when exacerbated by low income may be the most likely conclusion. Community crime has been associated with alcohol abuse, depression and anxiety symptoms after controlling for other community covariates such as income, community crime explains 4% to 5% of variability in treatment outcomes. Further, patients living in low-income areas required a higher number of treatment sessions to benefit from therapy. Patients living in economically deprived neighborhoods tend to have poorer depression and anxiety treatment outcomes and require lengthier interventions. They may have less access to clinical care than comparable urban and rural areas (Finegan et al., 2020).

Crime has been associated with certain communicable, biological pathogens as well. For example, Chlamydia has been associated with violent and non-violent crime (Marotta, 2017). Similarly, HIV and other STD’s have been suggested as correlates with violent and non-violent crime (Bonar et al., 2016). Exactly how this relationship between STD rates and violent crime rates in a community evolves is poorly understood. Some suggest the reasons for this relationship is not just the direct effects of illegal drug use on STD transmission but also the presence of both in similar counter cultures suggesting direct and indirect effects (Taylor, 2011). However, once again, we caution that issues such as STD rates may represent community “red flags,” rather than causal influences.

Empirical study suggests that access to certain health improving sites or agencies can reduce unhealthy practice and diminish unhealthy outcomes such as crime. Improving access to care and reducing involvement in the criminal justice system for people with mental illness correlates with less crime (McGinty et al., 2018). Increases in the number of open mental healthcare offices has been found to correlate with reductions in unhealthy behavior. These findings suggest an unintended benefit from expanding the office-based mental healthcare workforce on reductions in crime (Deza et al., 2020).

Social, Economic Factors and Physical Environment

Crime and location research has demonstrated that crime is not homoscedastic across space, even when reviewed in nesting sites suggested to attract crime such as red-light districts or liquor establishments (Eck et al., 2007). It is rather concentrated at relatively few locations, suggesting a particular inclusion of similar characteristics with physical and social descriptions determining the intra-community variation in crime (Sherman et al., 1989; Weisburd, 2015). We therefore examine the extent to which neighborhood features are associated with violent, property, and drug crime at low-income housing developments.

Location of and status of housing may play a role. Dunworth and Saiger (1994), looked at crime in public housing developments across Washington, DC, Los Angeles, and Phoenix and found that the rates of violent and drug offenses within the housing developments were consistently higher than nearby neighborhood rates, and substantially higher than citywide rates. Similarly, Holzman et al. (2005) examined crime in and around public housing in three unidentified cities and reported that the risk of aggravated assault in public housing developments was higher than in the 300 buffers surrounding the developments, which in turn were more dangerous than the jurisdiction as a whole. Parallel with this line of research are suggestions that air pollution and overcrowding may affect the crime rate as well. Bondy et al. (2020) found that air pollution has a positive and statistically significant relationship with overall crime, particularly property crime. They report the greatest effect within pollution levels below regulatory standards and correlations appear to be unevenly distributed across income groups. Further, many studies suggest that overcrowding may be an issue that encourages crime although evidence within the United States has been inconsistent (McCarty, 2010; Roncek et al., 1981).

Income inequality may be one of the greatest correlates with crime. Most of the literature on the effect of income inequality on crime and violent acts finds that the impact is a positive relationship between the two (Ehrlich, 1973; Enamorado et al., 2016; Kelly, 2000; Sampson, 1984). The reported positive relationship between inequality and crime may be largely driven by economic segregation across communities, rather than simple aggregate measures (Kang, 2016). Looking within counties or cities may be more beneficial since much street crime involves a perpetrator and victim from the same “street” (Kang, 2016).

Feelings of economic security in the community tend to mitigate crime and feelings of insecurity spawn crime through retaliatory tactics or gang membership (Merrin et al., 2015). Other prevalent factors include the aforementioned single-parent households. Single-parent families are associated with increased criminal behavior among youth raised in such households, although this appears to interact with other community factors (Wright & Younts, 2009).

Research on segregation that involve multilevel models of crime suggest that, net of city-level and neighborhood characteristics, racial segregation within a city is significantly related to greater violent and property crime (Smith et al., 2019). Multi-level model studies have found racial segregation of low income from high income households is connected with higher crime, particularly neighborhood violence (Krivo et al., 2015). This may be the result of the nature of social and political processes within U.S. cities that are difficult to operationalize in even ecological studies. For example, Houston, a city with a large non-European, White population, has a long history of structural fiscal inequality and institutional difficulties potentially encouraging serious disinvestment and political neglect that encourage decline in such communities. Similarly, other similar analysis has found economic segregation is as common and crucial as racial segregation in cities. It is associated with greater crime and other social problems for neighborhoods with high concentrations of disadvantaged and minority residents (Browning et al., 2010; Kang, 2016; Krivo et al., 2009).

The Current Study

Given various threads of research examining correlates with crime, we wished to examine for patterns in several high-violent crime cities to examine which of the health, physical, and social factors were related. The four cities chosen were Houston, TX, Baltimore, MD, Jackson, MS, and Wilmington, DE; these were selected because they were part of a broader study to examine motivations of youth gun possession in cities where Historically Black Universities were located, all of which served a co-PI’s on that initial project. These cities were useful in examining these issues given their historic history of high violent crime as well as their geographic diversity. These cities will be examined with two analyses. The first will compare the four cities against national rates of social stress indicators. The second will examine communities within these four cities to see what factors are most associated with crime.

Our study is a fundamentally exploratory one and, as such, not necessarily committed to any single theoretical model. Nonetheless, we were particularly interested in variables related to informal controls and social ecology (e.g., Sampson, 1986) as these might help us to understand the degree to which these may be related to race disparities in the commission of crime (Beck, 2021). Although not a formal hypothesis, we wished to examine the degree to which social indicators and informal controls as opposed to race per se might correlate with violent crime.

To accomplish this, we have set out to do two studies, each with a different level of analysis. In the first study, we examine the counties in which four big American cities with relatively high violent crime are situated (Houston, Baltimore, Jackson MS, Wilmington DE). This analysis gives us a fairly macro-level view, examining how these regions differ from the mean US county on social indicators of crime. From such an analysis we can obtain a broad picture of variables correlated with violent crime at the between-county level.

However, some of these variables may also correlate with each other. Further, within counties, neighborhoods likely differ significantly from each other, some high crime, some low crime. In our second analysis, we drill down to the neighborhood level using zip codes to examine correlates of crime between neighborhoods within those four cities. Doing so allows us to employ regression informed by the first study, to disentangle relationships between correlated variables in order to examine which are most important in understanding the correlates of crime.

As such, our two studies work together and inform one another. Naturally, correlates at two levels (county and zip code) will operate in different ways. However, this is part of our argument for why examining two levels of analysis is variable. Looking at broader levels (county) can inform our neighborhood (zip code) level analysis, while also giving us the opportunity to examine for general consistencies across levels of analysis.

Study 1

Methods

Materials and procedure

In order to determine the degree and extent of public health, we utilize the County Health Rankings Dataset (CHRD). This dataset has been compiled yearly by the Robert Wood Johnson Foundation and the University of Wisconsin Pollution Health Institute since 2010 (Remington et al., 2015). A collection of 30 county-level measures, the CHRD provides an understanding of the collective health of its members (health outcomes—mortality and morbidity) and those factors found to impact future health (health factors—health behaviors, clinical care, social and economic factors, and the physical environment).

Examining the factors that contribute to health and overall health outcomes, the County Health Rankings Dataset allows us to determine the level of health in over 3,000 U.S. counties. More importantly, we are able to supplement our analysis by including measures of teen birth rate, percent of adolescents living in poverty and the degree and extent to which disparities exists in the counties under question. The county level analysis allows us to overcome the challenges associated with the more often used aggregate level national data and its associated challenges.

For the current study the following variables were considered: county-level rates of poor health, adult smoking, adult obesity, excessive drinking, chlamydia diagnosis, teen births, HS Graduation, unemployment, children in poverty, income inequality, single parent households, violent crime, injury deaths, air pollution, severe housing problems, physical distress. Mental distress, diabetes prevalence, HIV prevalence, food insecurity, drug overdose mortality rate, insufficient sleep, math scores, median household income, proportion of children eligible for free or reduced price lunch, residential segregation Black/White, residential segregation non-White/White, homicide rate, suicide rate, firearm fatalities, juvenile arrest rate, percent housing cost burden, percent under 18, percent over 65, percent Black, percent non-English Speaking, percent female. Included in the dataset were data from 3,195 US counties.

Analyses used were one-sample t-tests. Target values were each of the values for the county in which the four target cities were located. Separate t-tests were run for each city comparing the US counties average against the value for that city for each variable. For instance, for suicide rates, 4 one-sample t-tests were run comparing the US counties mean value to the target values for Houston, Baltimore, Wilmington, and Jackson.

Exploratory analyses with such large datasets carry significant rates for spurious results to be “statistically significant.” This can result in misidentification of trivial correlates as important and inflated Type I error rates. This was addressed in two ways. First, a minimal p-value standard was employed (p < .005; Benjamin et al., 2018). Second, effect sizes need to reach a minimum of r = .10. This was selected as values below this are typically difficult to distinguish from noise due to the limits of precision in social science research (Ferguson & Heene, 2021). Together, these procedures should indicate which social correlates were both most consistent and strongest in this database. We certify that we are reporting all analyses conducted whether significant or not.

Results

Results from the t-test analyzes are presented in Table 1. In most cases, social stress variables were hypothesized to be higher in the four high-crime cities. Where the outcome was inverse of this, the outcome was coded as non-significant, essentially making these one-tailed tests. There were some exceptions such as demographic variables (percent female, for example) which were kept as two-tailed and reported as such. Also, one social stress variable (suicides) had strong and consistent evidence in an unexpected direct and, as such, was worth considering. In most other cases, opposing results were indicative of lack of consistency between cities and is interpreted as such.

Table 1.

T-Test Results Comparing Four Cities to US County Data on Social Stress Indicators.

Indicator	Houston	Baltimore	Wilmington	Jackson
Poor health	−12.88	x	x	−36.83
Adult smoking	x	x	x	−24.02
Adult obesity	x	x	x	−43.52
Excessive drinking	−44.33	x	−26.39	x
Chlamydia	−37.03	−26.42	−31.08	−31.08
Teen births	−16.48	x	x	−16.48
HS graduation	13.01	x	x	114.16
Unemployment	−10.23	x	x	−17.79
Children in poverty	−25.20	x	x	−50.07
Income inequality	−36.05	23.50	−6.28	−36.05
Single parent household	−19.36	−8.69	−19.36	−136.69
Violent crime	−135.11	−72.94	−84.29	−142.21
Injury deaths	−168.26	−44.99	−19.20	x
Air pollution	−85.40	−53.97	−28.25	−48.25
Severe housing problems	−75.33	−25.87	x	−87.70
Frequency of physical distress	x	x	x	−21.06
Frequency of mental distress	x	x	x	−28.7
Diabetes prevalence	x	x	x	−26.75
HIV prevalence	−107.31	−63.13	−56.23	−195.38
Food insecurity	−256.5	−23.69	−14.1	−11.63
Drug overdose	x	−93.59	−52.56	x
Insufficient sleep	−11.81	−12.84	−53.69	−67.31
Math scores	x	x	16.85	64.85
Median household income	x	x	x	33.24
Children/reduced price lunch	−30.93	x	NA	−104.98
Res. Segregation Black/White	−7.75	−34.87	x	−32.07
Res. Segregation non-White/White	−20.2	−76.03	−27.81	−96.12
Homicides	−16.11	−4.33	−11.65	−143.41
Suicides	50.53	53.62	47.2	53.62
Firearm fatalities	x	x	x	−105.26
Juvenile arrests	x	−37	−49.46	x
Percent housing cost burden	−68.17	−42.45	−42.45	−89.65
Percent under 18	−74.52	7.75	9.39	−33.06
Percent over 65	104.86	24.36	43.64	61.71
Percent Black	−38.57	−79.21	−62.95	−251.33
Percent non-English speaking	−203.97	−5.75	−5.75	x
Percent female	−10.58	−67.32	−39.93	−84.75

Note. x = did not meet threshold for “significance”; NA = data not available.

Summarizing the results from the table, the following social stress indicators were consistent across all four cities: Chlamydia rate, single parent households, violent crime rate, air pollution, HIV prevalence, food insecurity, insufficient sleep, residential segregation white/nonwhite, homicides, suicides, percent housing cost burden, percent over 65 (all four cities were “younger” than the national average), percent Black, percent female (all four cities had proportionally more females than the national average.) Taken together, economic and stress factors appear consistent across the four cities (income inequality was also significant for three of the four cities). Target cities also tended to be younger, had higher proportions of Black residents, had more females, and experienced more segregation than other parts of the US.

Brief Discussion of Study 1

A few variables warrant a closer look at this point. First, interestingly enough, regarding STD rates, both Chlamydia and HIV are higher in the target cities. It is unlikely that such effects are necessarily causal but may be indicative of a general culture of risk taking in these cities, higher than the national average. This may indicate that these cities suffer from wider public health issues that are impacting numerous outcomes.

Second, the results for suicide are interesting. All four target cities had considerably lower suicide rates than the national average. This may seem counterintuitive given the expectation that both violent crime and suicide may result from stress. However, this result is consistent with a recent cross-national study which found that suicides are higher in wealthier nations, whereas homicides were more common in countries with higher income inequality (Ferguson & Smith, 2021). As such, violent crime and suicide may be best conceptualized as existing along divergent stress paths.

Third, we note that our analysis confirmed other data (Beck, 2021) which finds violent crime higher among Black Americans. However, we note that this issue is complex, and it may not be that race, per se, is the relevant factor but rather than class-based issues involving economic disadvantage and social stress may be more salient. This, however, can only be tested in multivariate analysis and it is to this that we now turn.

Study 2

Methods

Materials and procedure

The most obvious critique of study 1 is that the data is county level and as such treats even of the four target cities (and surrounding municipalities) as whole units. However, much community variation is to be found within cities as well as between. Also, our previous analyses were univariate, not multivariate.

To test for variations within cities, we sought to examine communities at the level of zip code (the narrowest level we had available). In this case, the use of zip codes can help us examine neighborhood level correlates of crime, which may be more precise than county-level analysis given the significant variance in crime within broad regions such as counties. Using zip codes also gives us a wider arrange of data and variance allowing us to employ multiple regression with correlates which may help to disentangle relationships among the correlates themselves. Or put simply, this may help us to better understand which correlates are stronger than others. To do this, we took data from two sources. The first of these was the Area Deprivation Index (ADI; Kind & Buckingham, 2018) which provides state and national level rankings of communities based on an analysis of education, employment, housing, and income. The second was the Livability Index which is compiled from US Census Data and FBI crime data (Areavibes, 2021). From this, we were able to get US government data compiled for violent crime, property crime, mean household income, unemployment, academic test scores, percent Black residents, and population density.

From this data, we were able to employ multiple regression analyses. OLS regression models were compiled with pairwise deletion for missing data. ADI state and national ranks were collinear and, as such, were considered in separate models. Independent variables included ADI (both state and national rankings, in separate models), mean household income, unemployment, academic test scores, percept Black residents, and population density. Outcome variables were property and violent crime rates.

We note also that it would be reasonable to assume that zip codes are nested within cities and so the analyses reported below were retested with hierarchical linear modeling. This did not substantially impact our results so the OLS models are reported. As the models for ADI state and national were similar for all other variables and overall model statistics, the main statistics for ADI state models are reported, with the regression weights for ADI national added to the Table.

Results

Note, all results are reported in Table 2. As noted above, the standardized regression weights for ADI state and national are reported together, though separate models were run for each due to collinearity issues.

Table 2.

Regression Results for Study 2, Standardized Regression Coefficients and p-Values.

Correlate variable	Violent crimes	Property crimes
Unemployment	.228, p = .002	.368, p = .001
Median household income	−.341, p < .001	−.003, p = .979
Test scores	−.151, p = .009	−.039, p = .663
Percent Black	.097, p = .041	.058, p = .338
Population density	.283, p < .001	.057, p = .434
ADI state rank	.144, p = .014	−.341, p < .001
ADI national rank	.080, p = .208	−.280, p = .002

Property crimes

For property crimes, the overall regression models were statistically significant [ADI State model, R = .365, R²_adj = .107, F(6, 201) = 5.15, p < .001]. As can be seen in Table 2, both state and national ADI as well as unemployment were significant correlates with property crimes. All other variables were non-significant.

Violent crimes

For violent crimes, the overall regression models were statistically significant [ADI State model, R = .754, R²_adj = .556, F(6, 201) = 44.27, p < .001]. As can be seen in Table 2, most of the significant correlates were unrelated to ADI (state ADI had a small correlation with violent crimes, albeit in the opposite direction from expected). Unemployment, median household income, academic test scores, and population density were all significant. Percent Black reached statistical significance, but the effect size was smaller than the r = .10 standard we adopted for interpretation. Thus, such a result is judged as trivial and potentially noise.

As indicated above, retesting these models with zip codes nested within cities using HLM did not substantially change these results.

Brief Discussion of Study 2

Results from this study found that, at the community level as defined by zip code, multiple community stress factors were associated with violent crime. For property crime, however, only unemployment and ADI were related to higher rates of property crime. This suggests that property crimes may be more narrowly related to economic factors, whereas violent crimes are related to more general community stress.

As to the issue of race, race was not correlated with either crime outcome, particularly using the r = .10 cutoff. This suggests that community factors related to stress and class are correlated with crime perpetration, rather than race per se.

Discussion

Numerous factors have been proposed as possible correlates of crime. It can be valuable to look for common patterns related to community stressors as they relate to crime. We sought to address this in two studies of four high violent crime cities. Our results suggest that community stress as relates to health and economics are associated with violent crime.

Our results suggest that economic factors (such as state ADI rank, unemployment, median household income, population density) correlate with crime. Throughout our findings this is a consistent outcome, and it is consistent with the preponderance of the literature on criminal causality in criminal justice, criminology, psychology, sociology, and public health. This consistency buttresses the proposition that economic factors are risk factors for property and violent crime. Some research suggests that the primary risk for early onset of offending before the age of 20 includes socio-economic factors such as large family sizes and low family income (Smith et al., 2019) (Farrington, 2003). The research suggests an abundance of causes, particularly when looking at youth.

There is discussion in the literature regarding whether unemployment affects the crime rate or must act in accord with other demographic characteristics to affect crime. Raphael and Winter-Ebmer (2001) found an effect of unemployment on property crime rates. The estimates suggest that the decline in property crime rates during the 1990s is attributable to the decrease in the unemployment rate (Raphael & Winter-Ebmer, 2001). However, Britt (1997) found that the effects of unemployment (by measuring unemployment) on broad arrest rate (robbery, arson, theft) are contingent on age group and the time period. One of the aspects of our own findings, suggest that race has no significant effect on crime, when controlling for unemployment, and other economic factors. In effect, in discussions of race and crime, it may help to understand that race functions mainly as a proxy for class issues and addressing those class economic issues may decrease crime propensity for all ethnic groups.

Among our findings was the observation that single-parent households are associated with higher violent crime. This is consistent with other research finding that growing up in a single-parent household is associated with an increased risk of involvement in a crime (Kroese et al., 2020). Criminal behavior, particularly in adolescents, corresponds with chronic deprivation of parental acceptance; thus, the abandonment of a fatherly figure furthers the aggressiveness due to family turmoil (Fagan, 1995). One significant cause of poverty is fatherless families (Fagan, 1995). For example, the absence of a father leads to a heightened risk of lower intellectual development, illegitimate parenting in teenage years, and higher welfare dependency from the single mother (Fagan, 1995). As a result, this increases the risk of adolescent criminal behavior. On the other hand, a distant relationship between a mother and child can lead to neglectful and abusive relationships that can cause an adolescent to become emotionally detached and become delinquent (Fagan, 1995).

There are mixed results on the impact of food insecurity and violence. Caughron (2016) found that a 1% increase in food insecurity leads to a 12% increase in violent crime rates when holding other correlates of violent crime constant. However, the impact of food insecurity on crime rates depends on the income level and population of the county (Caughron, 2016). On the other hand, Blankenberger (2016) found no statistically significant relationship between food insecurity and violence. The overall effect of food insecurity various when it comes to an individual’s income and living location, such as urban or rural areas. Food insecurity is a health problem that does cause violence in undeveloped nations (Blankenberger, 2016). In the United States though, this may be a step in the chain between impoverished areas and crime.

Another aspect potentially addressing the connection between poverty and crime may be a lack of sleep. Adolescents tend to have a rising number of sleeping problems and an inadequate amount of sleep (Backman et al., 2015). A loss of sleep is often related to diminished behavioral responses and an increase in irritability and hostility (Backman et al., 2015). Therefore, sleep difficulties and an inadequate amount of sleep are associated with delinquent behavior, particularly in adolescents (Backman et al., 2015). Research also suggests that a lack of sleep from a partner increased aggressiveness and the likelihood of domestic violence (Hoshino et al., 2009). However, other research indicates that sleep deprivation is more associated with property crime than violent crime (Backman et al., 2015). The connection between lack of sleep and depression is well-documented. (Voelker, 2004) as is the bi-directionality of the association (Al-Abri, 2015). Our supports support the string of literature that explains stress, sleep disorder, and substance abuse as factors in depression (Voelker, 2004).

Our findings here show that sexual risk behavior may be associated with these other neighborhood disadvantages, whether economic, or environmentally based. Such risk has been associated with social disorder, environment, and social disorganization before (Bowleg et al., 2014). Tobin et al. (2012) looked at sex exchange norms and drug use in Baltimore, Maryland, and how it connects to STD and HIV infections. Baltimore ranks as one of the highest cities nationally in sexually transmitted infections and sexually transmitted diseases. Tobin et al. (2012) found evidence of a more significant spatial clustering of individuals in low-income areas who report risky sex exchange behavior. This research also suggested a relationship between sexually risky behavior and STD infections. As such, rates of STDs may be a risk marker for other community problems.

Polluting industries are more likely to be located in black neighborhoods than in white communities (Obach, 2013). Thus, air pollution typically affects individuals living in poverty or are disadvantaged. Research suggests that there is a relationship between air pollution (PM 2.5) and increased crime, such as assault and theft (Mapou et al., 2017). Air pollution in low-income areas creates a hazardous environment for the inhabitants, which causes stress, especially to children growing up in these neighborhoods (Evans et al., 2011). The increased amount of stress and health problems due to air pollution can lead to criminal activity.

Our study contributes to a robust literature on social correlates of violent crime. We believe our analysis contributes to this literature in two important ways. First, by considering four cities at two levels of analysis (county and neighborhood), we are able to examine multiple levels of analysis, allowing broader trends to inform regression analysis to examine which correlates are strongest and, as such, perhaps best targeted for policy. Second, we also believe our analyses are important for ongoing discussions and debates regarding the complex nexus of race, crime, and policing. Much of that narrative has focused on race and that racial minorities are both disproportionately involved in criminal activity but also, though this was not a subject of our current analysis, experiencing violence by police. Our analysis suggests that, at least on the issue of perpetrating crimes, it may be better to conceptualize these issues in relation to social correlates rather than racial correlates. As such, policies that attempt to alleviate the negative social environments correlated with crime, regardless of the ethnic composition of the neighborhood in question, may be fruitful in reducing crime.

Limitations

As with all studies, ours has limitations that must be addressed. First, our data is based strictly in correlation and, as such, causal conclusions cannot be reached. Second, our first study, considered data at the county level. However, considerable heterogeneity exists between communities within those counties. This was somewhat accounted for in our second study, which considered zip code level data. Though we believe examining data at these two levels provides strength to our analyses, we also recognize that phenomena can operate differently at different levels of analysis. By drilling down to smaller units of analysis (i.e., neighborhoods), this can help us get a fuller picture of issues operating at larger units of analysis such as counties or cities. Third, our first study examines data which is, fundamentally, bivariate. It is possible that some effects may be spurious and would not be significant in better controlled, multivariate analyses.

Conclusion

In conclusion, our data provide evidence for some economic and social stress indicators that appear to be consistent across four high violent crime cities. These particular factors may be particularly useful for highlighting for public policy efforts, particularly with experimental designs. We hope that our efforts will be a positive contribution to this area.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Supported by the National Collaborative on Gun Violence Research and the Thurgood Marshall College Fund.

ORCID iD

Christopher J. Ferguson

Author Biographies

Sven Smith is a professor of sociology at Stetson University and attorney. His research interests include the social learning and Weberian theory with regard to crime and courts.

Christopher J. Ferguson is professor of psychology at Stetson University. His research interests include the effect (or lack of) violent media on crime, and race, crime and policing.

Howard Henderson is professor of justice administration and director of the Center for Justice Research at Texas Southern University. His research informs the design of equitable criminal justice policy. He currently serves as senior fellow at the Brookings Institution and University of Virginia’s Institute for Advanced Studies in Culture.

References

Al-Abri

M. A.

(2015). Sleep deprivation and depression: A bi-directional association. Sultan Qaboos University Medical Journal, 15(1), e4.

Areavibes. (2021). Livability index. https://www.areavibes.com/methodology/

Backman

Laajasalo

Saukkonen

Salmi

Kivivuori

Aronen

E. T.

(2015). Are qualitative and quantitative sleep problems associated with delinquency when controlling for psychopathic features and parental supervision? Journal of Sleep Research, 24(5), 543–548.

Beck

A. F.

Moncrief

Huang

Simmons

J. M.

Sauers

Chen

Kahn

R. S.

(2013). Inequalities in neighborhood child asthma admission rates and underlying community characteristics in one US county. The Journal of Pediatrics, 163(2), 574–580.

Beck

A. J.

(2021). Race and ethnicity of violent crime offenders and arrestees, 2018. Bureau of Justice Statistics.

Benjamin

D. J.

Berger

J. O.

Johannesson

Nosek

B. A.

Wagenmakers

E. J.

Berk

Bollen

K. A.

Brembs

Brown

Camerer

Cesarini

Chambers

C. D.

Clyde

Cook

T. D.

De Boeck

Dienes

Dreber

Easwaran

Efferson

, . . . Johnson

V. E.

(2018, July 22) Redefine statistical significance. Nature Human Behaviour, 2, 6–10. https://doi.org/10.1038/s41562-017-0189-z

Blankenberger

(2016). Feeding peace: An investigation of the relationship between food insecurity and violence. Stevenson Center for Community and Economic Development to Stevenson Center for Community and Economic Development—Student Research. 20. https://ir.library.illinoisstate.edu/scced/20

Bonar

E. E.

Whiteside

L. K.

Walton

M. A.

Zimmerman

M. A.

Booth

B. M.

Blow

F. C.

Cunningham

(2016). Prevalence and correlates of HIV risk among adolescents and young adults reporting drug use: Data from an urban emergency department in the United States. Journal of HIV/AIDS & Social Services, 15(1), 3–28.

Bondy

Roth

Sager

(2020). Crime is in the air: The contemporaneous relationship between air pollution and crime. Journal of the Association of Environmental and Resource Economists, 7(3), 555–585.

10.

Bowleg

Neilands

T. B.

Tabb

L. P.

Burkholder

G. J.

Malebranche

D. J.

Tschann

J. M.

(2014). Neighborhood context and black heterosexual men’s sexual HIV risk behaviors. AIDS and Behavior, 18, 2207–2218. https://doi.org/10.1007/s10461-014-0803-2

11.

Britt

C. L.

(1997). Reconsidering the unemployment and crime relationship: Variation by age group and historical period. Journal of Quantitative Criminology, 13(4), 405–428. https://www-jstor-org-443.web.bisu.edu.cn/stable/23366008

12.

Bronfenbrenner

(1992). Ecological systems theory. Jessica Kingsley Publishers.

13.

Browning

C. R.

Byron

R. A.

Calder

C. A.

Krivo

L. J.

Kwan

M. P.

Lee

J. Y.

Peterson

R. D.

(2010). Commercial density, residential concentration, and crime: Land use patterns and violence in neighborhood context. Journal of Research in Crime and Delinquency, 47(3), 329–357.

14.

Browning

C. R.

Cagney

K. A.

(2002). Neighborhood structural disadvantage, collective efficacy, and self-rated physical health in an urban setting. Journal of Health and Social Behavior, 43, 383–399.

15.

Carr

(2003). The new parochialism: The implications of the beltway case for arguments concerning informal social control. American Journal of Sociology, 108(6), 1249–1291.

16.

Caughron

(2016). An examination of food insecurity and its impact on violent crime in American communities. All theses. 2565. https://tigerprints.clemson.edu/all_theses/2565

17.

Deza

Maclean

J. C.

Solomon

K. T.

(2020). Local access to mental healthcare and crime (No. w27619). National Bureau of Economic Research.

18.

Dunworth

Saiger

A. J.

(1994). Drugs and crime in public housing: A three-city analysis. US Department of Justice, Office of Justice Programs, National Institute of Justice.

19.

Dustmann

Fasani

(2016). The effect of local area crime on mental health. The Economic Journal, 126(593), 978–1017.

20.

Eck

J. E.

Clarke

R. V.

Guerette

R. T.

(2007). Risky facilities: Crime concentration in homogeneous sets of establishments and facilities. Crime Prevention Studies, 21, 225.

21.

Ehrlich

(1973). Participation in illegitimate activities: A theoretical and empirical investigation. Journal of Political Economy, 81(3), 521–565.

22.

Enamorado

López-Calva

L. F.

Rodríguez-Castelán

Winkler

(2016). Income inequality and violent crime: Evidence from Mexico’s drug war. Journal of Development Economics, 120, 128–143.

23.

Evans

Brooks-Gunn

Kato Klebanov

(2011). Stressing out the poor. Community Investments, 23(2), 22–27. http://cpi.stanford.edu/_media/pdf/pathways/winter_2011/PathwaysWinter11_Evans.pdf

24.

Fagan

(1995). The real root causes of violent crime: The breakdown of marriage, family, and community. Crime and justice, the heritage foundation report. https://www.heritage.org

25.

Farrington

D. P.

(2003). Developmental and life-course criminology: Key theoretical and empirical issues—The 2002 Sutherland award address. Criminology, 41(2), 221–255. https://doi.org/10.1111/j.1745-9125.2003.tb00987.x

26.

Ferguson

C. J.

Heene

(2021). Providing a lower-bound estimate for psychology’s “crud factor”: The case of aggression. Professional Psychology Research and Practice. Advance online publication. https://doi.org/10.1037/pro0000386

27.

Ferguson

C. J.

Smith

(2021). Examining homicides and suicides cross-nationally: Economic factors, guns and video games. International Journal of Psychology, 56, 812–823.

28.

Finegan

Firth

Delgadillo

(2020). Adverse impact of neighbourhood socioeconomic deprivation on psychological treatment outcomes: The role of area-level income and crime. Psychotherapy Research, 30(4), 546–554.

29.

Franzini

Caughy

Spears

Eugenia Fernandez Esquer

(2005). Neighborhood economic conditions, social processes, and self-rated health in low-income neighborhoods in Texas: A multilevel latent variables model. Social Science & Medicine, 61(6), 1135–1150.

30.

Heckman

Pinto

Savelyev

(2013). Understanding the mechanisms through which an influential early childhood program boosted adult outcomes. American Economic Review, 103(6), 2052–2086.

31.

Hildenbrand

A. K.

Daly

B. P.

Nicholls

Brooks-Holliday

Kloss

J. D.

(2013). Increased risk for school violence-related behaviors among adolescents with insufficient sleep. Journal of School Health, 83(6), 408–414.

32.

Hobbs

Kingham

Wiki

Marek

Campbell

(2021). Unhealthy environments are associated with adverse mental health and psychological distress: Cross-sectional evidence from nationally representative data in New Zealand. Preventive Medicine, 145, 106416.

33.

Holzman

H. R.

Hyatt

R. A.

Kudrick

T. R.

(2005). Measuring crime in and around public housing using GIS. In Wang

(Ed.), Geographic information systems and crime analysis (pp. 311–329). IGI Global.

34.

Hoshino

Pasqualini

J. C.

D’Oliveira

É. P.

Silva

C. P.

Modesto

Â. E.

Silveira

R. S. M.

(2009). Is sleep deprivation involved in domestic violence? Sleep Science, 2(1), 14–20. https://sleepscience.org.br/details/147/en-US

35.

Kang

(2016). Inequality and crime revisited: Effects of local inequality and economic segregation on crime. Journal of Population Economics, 29(2), 593–626.

36.

Kelly

(2000). Inequality and crime. The Review of Economics and Statistics, 82(4), 530–539.

37.

Kind

A. J.

Buckingham

W. R.

(2018). Making neighborhood-disadvantage metrics accessible—The neighborhood atlas. New England Journal of Medicine, 378, 2456–2458. https://doi.org/10.1056/NEJMp1802313

38.

Krivo

Peterson

Kuhl

(2009). Segregation, racial structure, and neighborhood violent crime. American Journal of Sociology, 114(6), 1765–1802.

39.

Krivo

L. J.

Byron

R. A.

Calder

C. A.

Peterson

R. D.

Browning

C. R.

Kwan

M. P.

Lee

J. Y.

(2015). Patterns of local segregation: Do they matter for neighborhood crime? Social Science Research, 54, 303–318.

40.

Krizan

Herlache

A. D.

(2016). Sleep disruption and aggression: Implications for violence and its prevention. Psychology of Violence, 6(4), 542–552.

41.

Kroese

Bernasco

Liefbroer

A. C.

Rouwendal

(2020). Growing up in single-parent families and the criminal involvement of adolescents: A systematic. Psychology, Crime & Law, 27(1), 61–75. https://doi.org/10.1080/1068316X.2020.1774589

42.

Lidfeldt

T. Y.

F. B.

Manson

J. E.

Kawachi

(2007). A prospective study of childhood and adult socioeconomic status and incidence of type 2 diabetes in women. American Journal of Epidemiology, 165(8), 882–889.

43.

Locher

J. L.

Ritchie

C. S.

Roth

D. L.

Baker

P. S.

Bodner

E. V.

Allman

R. M.

(2005). Social isolation, support, and capital and nutritional risk in an older sample: Ethnic and gender differences. Social Science & Medicine, 60(4), 747–761.

44.

Loeber

Farrington

D. P.

(eds) (1998). Serious & violent juvenile offenders: Risk factors and successful interventions. Sage Publications, Inc.

45.

Lorenc

Clayton

Neary

Whitehead

Petticrew

Thomson

Cummins

Sowden

Renton

(2012). Crime, fear of crime, environment, and mental health and wellbeing: Mapping review of theories and causal pathways. Health & Place, 18(4), 757–765.

46.

Mapou

Shendell

Ohman-Strickland

Madrigano

Meng

Whytlaw

Miller

(2017). Environmental factors and fluctuations in daily crime rates. Journal of Environmental Health, 80(5), 8–22. https://www-jstor-org-443.web.bisu.edu.cn/stable/10.2307/26329867

47.

Marotta

(2017). Assessing spatial relationships between rates of crime and rates of gonorrhea and chlamydia in Chicago, 2012. Journal of Urban Health, 94(2), 276–288.

48.

McCarty

W. P.

(2010). Trailers and trouble? An examination of crime in mobile home communities. Cityscape, 12, 127–144. https://www.manufacturedhomelivingnews.com/wp-content/uploads/2016/09/TrailersAndTroubleExaminationOfCrimeInMobileHomeCommunitiesWilliamPMcCartySept12010-postedMobileManufacturedHomeLivingNews-.pdf

49.

McGinty

E. E.

Goldman

H. H.

Pescosolido

B. A.

Barry

C. L.

(2018). Communicating about mental illness and violence: Balancing stigma and increased support for services. Journal of Health Politics Policy and Law, 43(2), 185–228.

50.

McGuirt

Huebner

Ward

Pitts

S. J.

(2019). P36 examining the healthiness of food and beverage options at highway rest areas in North Carolina. Journal of Nutrition Education and Behavior, 51(7), S48–S49.

51.

Merrin

G. J.

Hong

J. S.

Espelage

D. L.

(2015). Are the risk and protective factors similar for gang-involved, pressured-to-join, and non-gang-involved youth? A social-ecological analysis. American Journal of Orthopsychiatry, 85(6), 522–535.

52.

Myers

C. A.

Denstel

K. D.

Broyles

S. T.

(2016). The context of context: Examining the associations between healthy and unhealthy measures of neighborhood food, physical activity, and social environments. Preventive Medicine, 93, 21–26.

53.

Obach

(2013). Ten lessons in introductory sociology: Class and intersectionality (1st ed.). Oxford University Press.

54.

Otsu

Yuen

C. Y.

(2020). Health, crime, and the labor market: Theory and policy analysis. Journal of Interpersonal Violence.

55.

Raphael

Winter-Ebmer

(2001). Identifying the effect of unemployment on crime. The Journal of Law and Economics, 44(1), 259–283. https://doi.org/10.1086/320275

56.

Remington

P. L.

Catlin

B. B.

Gennuso

K. P.

(2015). The county health rankings: Rationale and methods. Population Health Metrics, 13, 11. https://doi.org/10.1186/s12963-015-0044-2

57.

Richardson

A. S.

Troxel

W. M.

Ghosh-Dastidar

Hunter

G. P.

Beckman

Colabianchi

Collins

R. L.

Dubowitz

(2017). Pathways through which higher neighborhood crime is longitudinally associated with greater body mass index. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 155.

58.

Roncek

D. W.

Bell

Francik

J. M. A.

(1981). Housing projects and crime: Testing a proximity hypothesis. Social Problems, 29(2), 151–166.

59.

Sampson

R. J.

(1984). Group size, heterogeneity, and intergroup conflict: A test of Blau’s inequality and heterogeneity. Social Forces, 62(3), 618–639.

60.

Sampson

R. J.

(1985). Neighborhood and crime: The structural determinants of personal victimization. Journal of Research in Crime and Delinquency, 22(1), 7–40.

61.

Sampson

R. J.

(1986). Neighborhood family structure and the risk of personal victimization. In Byrne

J. M.

Sampson

R. J.

(Eds.), The social ecology of crime (pp. 25–46). Springer-Verlag.

62.

Sampson

R. J.

(1987). Communities and crime. In Gottfredson

Hirschi

(Eds.), Positive criminology (pp. 91–114). SAGE.

63.

Sampson

R. J.

(2003). Neighborhood-level context and health: Lessons from sociology (pp. 132–146). Oxford University Press.

64.

Sampson

R. J.

Groves

W. B.

(1989). Community structure and crime: Testing social-disorganization theory. American Journal of Sociology, 94(4), 774–802.

65.

Sharkey

P. T.

(2006). Navigating dangerous streets: The sources and consequences of street efficacy. American Sociological Review, 71(5), 826–846.

66.

Sherman

L. W.

Gartin

P. R.

Buerger

M. E.

(1989). Hot spots of predatory crime: Routine activities and the criminology of place. Criminology, 27(1), 27–56.

67.

Singleton

Adams

McLafferty

Sheehan

Zenk

(2019). P34 county-level associations between food retail outlet availability and violent crime rate. Journal of Nutrition Education and Behavior, 51(7), S47–S48.

68.

Smith

Gomez Auyong

Z. E.

Ferguson

(2019). Social learning, social disorganization, and psychological risk factors for criminal gangs in a British youth context. Deviant Behavior, 40(6), 722–731.

69.

Snowden

A. J.

Hockin

Pridemore

W. A.

(2020). The neighborhood-level association between alcohol outlet density and female criminal victimization rates. Journal of Interpersonal Violence, 35(15–16), 2639–2662.

70.

Sundaresh

Roy

Riley

Wildeman

Wang

E. A.

(2020). Exposure to the US criminal legal system and well-being: A 2018 cross-sectional study. American Journal of Public Health, 110(S1), S116–S122.

71.

Taylor

O. D.

(2011). The sexual victimization of women: Substance abuse, HIV, prostitution, and intimate partner violence as underlying correlates. Journal of Human Behavior in the Social Environment, 21(7), 834–848.

72.

Tobin

K. E.

Hester

Davey-Rothwell

M. A.

Latkin

C. A.

(2012). An examination of spatial concentrations of sex exchange and sex exchange norms among drug users in Baltimore, Maryland. Annals of the Association of American Geographers, 102(5), 1058–1066. https://www-jstor-org.web.bisu.edu.cn/stable/23275580

73.

Voelker

(2004). Stress, sleep loss, and substance abuse create potent recipe for college depression. The Journal of the American Medical Association, 291(18), 2177–2179.

74.

Weisburd

(2015). The law of crime concentration and the criminology of place. Criminology, 53(2), 133–157.

75.

Wolf

A. M.

Lipman

A. D. P.

Boggan

Glesmann

Castro

(2015). Saving lives: Alternative approaches to reducing gun violence. Int Sci Index, 9(6).

76.

Wright

B. R. E.

Younts

C. W.

(2009). Reconsidering the relationship between race and crime: Positive and negative correlates predictors of crime among African American youth. Journal of Research in Crime and Delinquency, 46(3), 327–352.

77.

Lippert

A. M.

(2016). Neighborhood crime rate, weight-related behaviors, and obesity: A systematic review of the literature. Sociology Compass, 10(3), 187–207.