Abstract
The balance between crime control methods and individual liberties is always problematic, creating tension, because in order to investigate crime, and adjudicate and punish offenders, it is necessary to make reasonable intrusions into the liberty of citizens. This study uses data from 40 countries to examine the crime control measures (police per capita and conviction rates) that reflect government investments in criminal justice apparatus to control crime and criminals, as well as the use of these crime control measures through government intervention in the lives of its citizens (formal citizen contacts with police, prosecution rate, and detention rate), to examine their impact on crime victimization rates (homicide rates and crimes included in the international crime victim survey). The purpose is to examine whether these government interventions have any impact on crime rates across countries, controlling other independent variables that might help to explain any observed relationships among these variables (such as measures of civil liberties, democracy, human development, available information and communications technologies, political rights, corruption perceptions, education, economic freedom, freedom of the press, and prosperity).
Keywords
The balance between crime control methods and individual liberties is always problematic, creating tension, because in order to investigate crime, and adjudicate and punish offenders, it is necessary to make reasonable intrusions into the liberty of citizens. These intrusions include formal police contacts with citizens, prosecution, conviction, and detention.
Besides individual liberty interests, another constraint on these criminal justice interventions into the lives of citizens is the resources available. That is, a country’s financial capacity to fund police, courts, and corrections impacts its ability for effective crime control and its impacts on individual liberty. Most importantly, however, does the extent of government interventions have any impact on crime rates across countries?
This research is designed to generate an empirical comparison of multiple nations to examine the extent to which crime control efforts, and investments in criminal justice, impact crime rates. Important social, economic and political variables are also examined to assess the extent they have any significant influence on the relationship between crime control efforts and crime rates.
Literature Review
There is a body of research which attempts to assess the impact of government policy and practice on crime. This research is designed to inform governments in ways to best allocate public funds to achieve the greatest impact on public safety and to learn from the experiences of other jurisdictions.
There are differences in approach, of course, with some researchers taking a more critical approach that questions the motives of governments to improve public safety in general versus protecting the privileged classes or those holding power (Liska, 1993). Analyses of the general deterrent effect of government efforts to control and prevent crime run from the macro (social level) to micro (individual level). Some work has concluded that the criminal justice system does a good job in identifying and punishing offenders who break the law frequently, but the mechanisms to make this work more effectively are not always clear (Barnes, 2014). A study in Brazil, for example, attributed a drop in homicides to more effective policing methods and stricter gun control, without solving underlying socioeconomic problems (Goertzel & Tulio, 2009). A study in Guatemala concluded that poor efficiency in police departments resulted in low clearance rates, which would have to be increased to improve crime control (Alda, 2014). An analysis in the United States found no relationship between apprehension risk and crime rates, using crimes reported to police as the measure of crime. But when using victimization survey data, strong deterrent effects were found. At the same time, police resource levels were found to impact apprehension risk for most property crimes, except the most common one: larceny (Zedlewski, 1989).
These mixed results have led many researchers to examine noncriminal justice factors to assess their possible impact on crime rates. In most cases, the research has examined the comparative impact of legal versus nonlegal factors on crime rates.
A Combination of Legal and Extralegal Factors
Studies that have examined the impact noncriminal justice–related factors on crime rates have reported a range of different findings. An analysis of 169 countries found income inequality was related to violence in both low- and middle-income countries, arguing for reduced income equality as a violence reduction approach (Wolf, Gray, & Fazel, 2014). But another study found the association between income inequality and violent crime rates was limited to murder rates (Crutchfield, 1989). Other work has found similar impacts of inequality on homicide rates (Messner, Raffalovich, & Shrock, 2002).
Still other research finds other factors at work. A study by Pridemore (2011) found that when poverty and inequality are both measured, there was a significant poverty–homicide association, but the inequality–homicide association disappeared. Similarly in Sweden, the socially and/or economically disadvantaged were much more likely to experience violence than others (Stickley & Carlson, 2010).
More recent work has uncovered some similarities in their results. Baumer and Wolff (2014) conducted a cross-national analysis of homicide trends over 20 years and found declines were linked to reductions in poverty, urbanization, and the ratio of older to younger persons. Higher imprisonment rates were not found to be associated with declines in homicide rates. A large study of international homicide trends, using cross-sectional analyses, found lethal violence is dependent on the rule of law, the quality governance, level of democracy, and social and economic equality. High rates and imprisonment and long sentences were found to correspond with high homicide rates (Lappi-Seppala & Lehti, 2014). A comprehensive review of the findings of 70 previous cross-national homicide studies found consistent evidence connecting homicide rates and resource deprivation cross-nationally (Trent & Pridemore, 2011).
A large number of additional studies have been conducted, using different measures of social factors, different combinations of countries and jurisdictions, over different time periods, and using different measures of crime. These multiple differences likely account for the wide variation in results.
For example, resource deprivation and a larger youth population have been found to increase homicide rates over time (McCall, Parker, & MacDonald, 2008). Childhood economic disadvantage has also been found to be related to criminal involvement (Fergusson, Swain-Campbell, & Horwood, 2004). In Japan, measures of economic stress, population age structure, and certainty of punishment were found to be predictors of postwar trends in violent crime (Roberts & LaFree, 2004). Many other studies have assessed other social, economic, and political factors that might impact crime rates and victimization risk, including housing, health, racial heterogeneity, social support, and gender (Livingston, Kearns, & Bannister, 2014; Lee & Ousey, 2005; Pratt & Godsey, 2003; Steffensmeier, 2000; Stickley, Koyanagi, Roberts, Rotman, & McKee, 2013).
The Present Study
In the present study, an effort is made to examine a large number of countries, using available measures of crime, criminal justice, and social conditions in recent years. In this way, an effort is made to examine these factors in multiple countries over a common time period.
Specific research questions include:
Figure 1 illustrates the approach taken in this study. It outlines how government investments in crime control (as measured by police per capita and conviction rates) combine with government intervention in the lives of citizens (as measured by formal contacts with police, prosecution rate, and detention rate) in an effort to impact crime rates. 1 Figure 2 diagrams the logic that the number of police per capita and their formal contacts with the public might lead to increased rates of prosecution and conviction, resulting in higher detention rates, and ultimately in lower crime rates.
Does government crime control investment and intervention impact crime rates?
Logic model of potential criminal justice impact on crime rates.
Data Sources
The United Nations Office on Drugs and Crime (UNODC) has produced a global statistical series on crime and criminal justice via a Crime Trend Survey, which requires the responses of individual governments to provide these data. Statistics on formal police contacts, prosecutions, convictions, and detention among multiple countries (called UN Member States) are included in this data gathering effort. Of course, these data are limited in that they reflect only crimes known to governments, and there is a great deal of missing data resulting from many UN Member States that do not provide data regularly to the UNODC.
The first limitation is addressed by use of the International Crime Victimization Survey, which is carried out periodically in multiple nations. These data provide a more comprehensive count of all crime because crimes
The data are for 2010, excluding missing data, leaving 40 countries for analysis (see Table 1 for a listing of the countries included). Given that there are 193 countries in the United Nations, the missing data problem is large, but data that are not gathered or reported cannot be analyzed.
Countries Included in the Study.
Results
Research Question 1 was whether the criminal justice–related factors, presented in Figures 1 and 2, have a significant impact on crime. That is to say, do crime control measures plus government interventions in the lives of its citizens impact crime and homicide rates? The impact on crime victimization rates (less homicide) was significant. (R 2 = .682, analysis of variance [ANOVA] sig. p < .003) with the detention/incarceration rate explaining nearly all the variance. These are somewhat surprising results, especially in that the detention rate was, by far, the best predictor of crime and homicide rates. 3 The crime victimization rates measured are those counted by the International Crime Victim Survey, which includes car theft, bicycle theft, burglary, robbery, larceny, assault, and sex offenses. The impact on homicide rates was also significant. (R 2 = .714; ANOVA sig. p < .001). 4 The R 2 measures the proportion of variation in the crime rates and homicide rates explained by the independent variables. Findings (of .682 and .714) imply that 88% and 71% of the variation in data for crime rates and homicide rates, respectively, is explained by the regression. This is strong explanatory power in social science research, although it must be kept in mind that no statistic establishes causation, only the logical exclusion of alternate explanations does. This is also a cross-sectional analysis, and trend data over a longer period time would be required to draw stronger conclusions.
This left the question of whether other, noncriminal justice–related variables can better explain crime and homicide rates across countries. This question is illustrated in Figure 3.
Potential missing variables in explaining crime and homicide rates.
Independent variables were added in an effort to explain differences in crime and homicide rates among countries and the importance of social, political, and economic factors in this explanation. Ten different indexes were used to measure different dimensions of social, political, and economic life. These included a civil liberties ranking, a democracy index, human development index, information and communications technologies index, political rights index, corruption perceptions index, education index, index of economic freedom, press freedom index, and prosperity index. These measures are each summarized in Table 2.
Measures of Social, Political, and Economic Life.
Note. ISP = Internet service provider.
Each of these 10 indexes measures important social, political, and economic factors that not only influence the quality of life in a society but also help to define the relationship between citizens and their government. There is some overlap among these indexes, but they each measure multiple variables in unique combinations, so they are measuring somewhat different aspects of social, political, and economic life. Correlations among these independent variables are presented in Table 3. It can be seen that there are statistically significant correlations between these indexes, because they measure overlapping concepts, but the strength of the correlations ranges from .896 to .451, suggesting some differences in what they measure.
Zero-Order Correlations.
Note. ICVS = International Crime Victim Survey.
*p < .05.
These 10 measures also have the attribute of being in existence over many years, so their methodology has been refined over time, and they are widely used to assess conditions around the world (Glatzer, Camfield, Møller, & Rojas, 2015; Land, Michalos, & Sirgy, 2012; Miringhoff & Miringhoff, 1999). The civil liberties ranking measures freedom of expression and belief, associational and organizational rights, rule of law, and personal autonomy and individual rights. It has been published annually since 1972 and provides data on nearly 200 countries. The indicators are drawn from the Universal Declaration of Human Rights (Freedom House, 2015a; United Nations, 2015).
The democracy index is gathered by the Economist Intelligence Unit and was first published in 2007. It classifies 165 countries based on five categories: electoral process and pluralism, civil liberties, the functioning of government, political participation, and political culture. Countries are placed within one of the four types of regimes: full democracies, flawed democracies, hybrid regimes, and authoritarian regimes (Economist Intelligence Unit, 2014).
The human development index was established by the United Nations Development Programme. It provides a summary measure on three development dimensions: life expectancy, educational attainment, and standard of living (United Nations Development Programme, 2015a). The information and communications technologies index was developed by United Nations Conference on Trade and Development. It developed indices to comparatively assess across nations the levels of Internet connectivity, numbers of users, and the digital divide (United Nations Conference on Trade and Development, 2013).
The political rights index is based on data in three categories: electoral process, political pluralism and participation, and functioning of government. Countries are ranked based on multiple indicators for each of these criteria (Freedom House, 2015b). The corruptions perceptions index is published by Transparency International, and it ranks countries based on the level of corruption in their public section as perceived by citizens, businesspersons, and country-level experts. It relies on a combination of polls of multiple constituencies (Transparency International, 2015).
The education index was developed by the United Nations Development Programme combining average years of schooling and expected years of schooling, used in measuring economic development and quality of life—a key factor determining whether a country is a developed, developing, or underdeveloped nation (United Nations Development Programme, 2015b). The index of economic freedom was developed by the Heritage Foundation and combines indicators in 10 areas for nearly all countries in the world. The indicators include multiple measures of open markets, regulatory efficiency, limited government, and rule of law (Heritage Foundation, 2015).
The press freedom index was developed by Reporters Without Borders, and it ranks countries annually based on each country’s press freedom record during the previous year. It reflects the comparative degree of freedom that journalists, news organizations, and Internet reporters enjoy in each country, and the efforts made by government to protect this freedom (Reporters Without Borders, 2015). The prosperity index was developed by the Legatum Institute, and it is based on multiple indicators including wealth, economic growth, education, health, personal well-being, and quality of life (Legatum Prosperity Index, 2015).
It is important to remember that these indexes measure qualities that are difficult to quantify, yet this is precisely what they do. Moreover, methodologies can change year to year in order to enhance their validity, but most of these indexes describe their methodology in their reports, so their measurement efforts are transparent.
Figure 4 illustrates how these multiple indexes on a range of social, political, and economic issues were used as independent variables to determine their influence on the relationship between criminal justice factors and crime and homicide rates.
Assessing the influence of social, political, and economic factors.
The impact of these 10 multifactor indexes on crime victimization rates was not significant (R 2 = .103, ANOVA sig. p < .965), although the detention rate remained the strongest factor in explaining the variance. However, the impact on homicide rates was significant (R 2 = .654; ANOVA sig. p < .000). Economic freedom and corruption perceptions had the strongest associations with homicide rates, which corresponds with observed differences in other studies described below.
As a result, government investment in crime control and intervention in the lives of citizens is shown to have measurable impacts on crime victimization rates and homicide rates across 40 countries. Social, political, and economic factors do not have a strong impact on crime rates, although they have a significant impact on homicide rates. The strongest factors were found to be economic freedom (rule of law, limited government, regulatory efficiency, open markets), and low corruption (the extent of corruption in the public sector). Longitudinal studies covering more countries over a longer time period will provide more insight into causal connections beyond mere association.
Several unknowns must be addressed by future research. First, it is not clear what are the precise ways by which government crime control and economic factors impact crime or homicide rates. These require closer analysis in individual countries (i.e., microlevel analysis using in-country observations and interviews with key stakeholders). This microlevel variation might be masked in this study which relies on macro-level data.
Second, it must be determined through analysis on additional data whether these findings will hold up over time with data from the coming years. Trend data are required to establish relationships over time. Third, Asian and African countries are underrepresented in this group of 40 countries due to lack of data provided to the United Nations from the Member States. Therefore, there might be variables of interest in those regions which have not yet been captured. That is to say, the true relationship between important variables, such as detention rate and crime victimization rate, might be concealed, given the skewed sample of countries reporting these data and able to be included in this analysis.
Discussion and Conclusion
All countries experience crime, yet the evidence is not clear how best to invest limited resources for maximum impact. Every nation invests in police resources and court systems to produce arrests and convictions and thereby impact crime. Nations make use of available crime control measures in formal contacts of citizens with the police, the prosecution rate for those arrests, and the detention rate for those convicted of crimes. This study looked at data on these variables for 40 countries for whom data were available, finding that higher detention/incarceration rates were associated with crime victimization rates. It is not clear, however, whether it was certainty, severity, or other factors behind the detention that drives this relationship. Other comparative studies of homicide rates in particular have found no relationship between incarceration rates and homicide. As Lappi-Seppälä and Lehti found in their cross-national study of homicide, “high rates of imprisonment, and extensive use of life sentences are usually associated with high and increasing homicide rates—and not the other way around” (2014, p. 159). Somewhat different results were found by Baumer and Wolff, who offer an explanation for it. we concluded that growth in imprisonment was unrelated to homicide trends, but this finding describes the overall pooled pattern that emerges across our sample of nations. Drawing insights from theories of legitimacy and procedural justice, we acknowledge the possibility that this null effect could mask offsetting effects across nations: increases in imprisonment could yield homicide reductions in nations where the justice system is considered legitimate, while they could yield increases where there is deep suspicion regarding government authority. (Baumer & Wolff, 2014)
There is reason to believe that noncriminal justice–related factors may also influence crime victimization rates. An effort was made to examine the impact of these factors by controlling for 10 important social, political, and economic influences in each country that exists outside the criminal justice system. These factors did, indeed, make a significant difference in homicide rates (see also Rogers & Pridemore, 2013). Similar to the findings in the current study, Lappi-Seppälä and Lehti’s cross-national study concluded, “homicide is lower under more effective governments and in less corrupt environments” (2014, p. 175). It remains for future research to untangle the reasons for this, which might rely on more specific measures of rule of law, legitimacy, public attitudes, and corruption levels in individual countries.
It should be noted that the problem of reaching a satisfactory explanation of variations in crime rates has been experienced by a variety of researchers. Contemporary studies attempting to explain crime drops and increases around the world suffer from the same data problems. For example, burglary rates across nations have been found to be impacted by gross domestic product and consumer confidence (Rosenfeld & Messner, 2009), and conviction rates have been related to crime rates in some countries (Farrington, 2015). An analysis of homicide rates over time found they “went up and are now going down in most parts of the world, but this has taken place at different times, with different backgrounds, and probably for different reasons” (Lappi-Seppälä & Lehti, 2014, p. 155). The impact of unmeasured variables, better measures of the same variables, and the reasons behind these observed relationships are difficult to answer in macro-level studies. This is because gathering data of consistent reliability and validity comparatively across jurisdictions is even more difficult than obtaining such data within single countries. In addition, macro-level studies have difficulty in being sensitive to local effects and interaction effects that unmeasured, or poorly measured, variables produce.
Studies that are designed to test particular theories of why crimes rates vary also have very mixed results, due to either inadequate theories or the data to test them appropriately. In the case of criminology, it might be both. As LaFree and Drass (2002) concluded, “These results support the utility of recognizing crime booms as an important criminological concept while retaining a healthy scepticism toward any specific claims made about their existence” (p. 790). Most of the best data exist in developed countries, as many developing countries do not have the infrastructure or resources for reliable data collection. As an example, the 40 countries in the current study, listed in Table 1, reveal a disproportionate number of developed countries. Therefore, existing explanations might be missing important variables, simply due to the nature of the sample of countries selected. These problems have been noted in multiple studies (Altheimer, 2013; Farrington, 2015; Pridemore & Grubesic, 2013; Tonry, 2014).
Therefore, it might be a better path to look first at variations within nations in the effort to locate important influences and variables to be measured before moving to multinational-level studies of available data, which often mask these local variations and the reasons for them. As Pridemore (2005) has observed, “An in-depth case study of a single country provides a better understanding of data sets and insight into nation-specific influences on rates of interpersonal violence” (p. 754). As a result, the conceptual breakthroughs are likely to come from observations from the bottom-up (microlevel) investigations, rather than from macro-level studies which have the limitations noted here.
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) received no financial support for the research, authorship, and/or publication of this article.