![]()
ONE
through the looking glass: the complex relationship between crime and the economy
Willie Sutton was a notorious bank robber. During his 40-year criminal career he allegedly stole over $2 million.1 He spent about half of his adult life behind bars, although he managed to escape from prison twice. He is best known for his answer when asked why he robbed banks: â⌠because thatâs where the money is.â In his partly ghost-written autobiography, Where the Money Was: The Memoirs of a Bank Robber, Sutton denied having uttered the phrase. He credited some enterprising reporter. Nevertheless, the words attributed to him seem to contain a pearl of wisdom: when trying to understand some phenomenon, begin with the obvious. This sage advice is sometimes referred to as âSuttonâs Lawâ.
At first glance, the basic focus of this book â the relationship between the economy and crime â might seem particularly well suited for the application of Suttonâs Law. There appears to be nothing particularly profound or surprising in the proposition that the ups and downs of the economy and criminal activity go hand in hand. It turns out, however, that Suttonâs Law serves only as a useful starting point in inquiring about the economy and crime. The picture becomes murkier as we peer through the looking glass. As self-evident and straightforward as the connection between crime and the economy might seem, the relationship is actually quite complex. A half century ago, a thorough review of the academic literature concluded that âthe general relations of economic conditions and criminality are so indefinite that no clear or definite conclusions can be drawnâ (Vold, 1958: 181). Our reading of the research literature is not quite as pessimistic, but we agree that the accumulated evidence defies simple conclusions.
In this book we review the research on the relationship between crime and the economy and consider the theoretical perspectives that have been advanced to explain the complexities in that relationship. We then offer our own take on the relationship, especially as it is manifested in contemporary advanced industrial societies. Our review covers all kinds of criminal acts, violent as well as property offences and so-called white-collar as well as street crimes. Our perspective on the crimeâeconomy relationship considers the modern market economy as a social institution that has powerful, if not always obvious or direct, consequences for criminal activity. The key concept in our institutional analysis is the market: the exchange of goods and services for money.
To set the stage for the analyses to follow, it is useful to begin by considering one particularly salient outcome of the market economy, specifically, economic deprivation or poverty. Here is where Suttonâs Law is most likely be invoked. To the extent that the economy has any connection with crime, it seems almost commonsensical to surmise that poor economic circumstances will promote high levels of criminal activity. Yet, as we shall see, the relationship between economic deprivation and crime proves to be more complex than it initially appears.
Socioeconomic status and street crime
The phrase âItâs the economy, stupidâ was popularized during Bill Clintonâs 1992 US Presidential campaign. Advisor James Carville used the phrase to focus the campaign on the recession that occurred during the Presidential term of Clintonâs opponent, George H. W. Bush. âSo goes the economy, so goes the electionâ has become a staple of political science research on the impact of the economy on electoral politics (Dolan et al., 2008; Fair, 2009). When we look at statistics on the socio-economic status of persons caught up in the criminal justice system, economics also predict the outcome: the poor get prison.2
The poor are over-represented among those who are arrested for crimes, sentenced to prison, and even those who are victimized by crime. Surveys of the inmates in local jails, including persons awaiting trial and those convicted of crimes, consistently find that jail inmates have less education, higher unemployment rates, and lower incomes than the general population of adults. For example, in 2002, 44% of jail inmates did not have a high school diploma or GED, compared with 20% of the US adult population.3 Fully 29% of the jail inmates were unemployed during the month before they were arrested, compared with an unemployment rate of 5.8% in the general population. During the month before their arrest, 59% of jail inmates had incomes less than the federal poverty threshold for a family of two; most of those did not have enough money to support themselves above the poverty line. The poverty rate for the general population was just over 12%. The socioeconomic portrait of persons convicted for serious crimes and sentenced to state and federal prisons, if anything, is even bleaker than that of inmates in local jails (Petersilia, 2003; Western, 2006).
The picture of the socioeconomic characteristics of offenders drawn from official statistics on arrest and incarceration is of course imperfect. Criminologists have long recognized that the processing of cases by criminal justice officials â from police to prosecutors to juries and judges â can introduce systematic biases, often based on socio-economic status (Cooney, 2009; Reiman and Leighton, 2009). But even âself-reportâ surveys which ask persons if they have committed crimes find that those of lower socioeconomic status report greater involvement in serious violent and property crime (e.g., Loeber and Farrington, 1998).
Another vantage point from which to observe the relationship between socioeconomic status and crime is through an examination of criminal victimization. Residential patterns in the US tend to be segregated by class (and race). In addition, criminological researchers have documented that offenders tend not to travel far from home when selecting their targets for victimization (Brantingham and Brantingham, 1993; Chainey and Ratcliffe, 2005). It is therefore not surprising that persons who live in close proximity to criminal offenders tend to have higher rates of victimization, regardless of their lifestyle and other characteristics (Sampson and Lauritsen, 1990). If socioeconomic status is in fact associated with criminal offending, we would expect that low-income residents will be at comparatively high risk of victimization.
Some evidence supports this hypothesis. Each year, the US Bureau of Justice Statistics, in conjunction with the Census Bureau, conducts an ambitious data collection effort, the National Crime Victimization Survey (NCVS). The NCVS asks representative samples of the population, age 12 and older, whether they have been the victim of a crime during the past six months and, if so, what type of crime. The NCVS also collects information on the characteristics of the households in the sample, including household income. Table 1.1 reports the burglary victimization rates per 1,000 households in 2009 according to household income. The data reveal a clear negative relationship between household income and burglary. The lowest two income categories exhibit the highest burglary rates, and the rates decline steadily as incomes rise.
Whether we look at official statistics on arrest and incarceration, self-report studies of criminal offending, or surveys of crime victims, the same pattern emerges: lower socioeconomic status is associated with greater involvement with the criminal justice system, higher rates of criminal offending, and higher rates of various forms of victimization. The relationship between socioeconomic deprivation and involvement in crime and the justice system holds not only for individuals, as we have seen, but also for neighbourhoods. Extensive research confirms the perception of nearly all urban residents: there are readily identifiable âbadâ neighbourhoods where the risk of becoming a victim of crime is high. These neighbourhoods are typically characterized by a host of disadvantages, including pervasive poverty.
Table 1.1 NCVS Burglary Rates per 1,000 Households, 2009
| Under $7,500 | 44.4 |
| $7,500 to $14,999 | 46.3 |
| $15,000 to $24,999 | 35.3 |
| $25,000 to $34,999 | 32.3 |
| $35,000 to $49,999 | 26.7 |
| $50,000 to $74,999 | 19.3 |
| $75,000 and over | 15.1 |
Source: Truman and Rand, 2010
An example comes from research on the city of St. Louis (Rosenfeld et al., 1999). The map displayed in Figure 1.1 divides the city into 588 census âblock groupsâ, small geographic areas with an average population of 675 residents per block group. The block groups are shaded according to their score on an index of disadvantage, consisting of the rate of poverty, public assistance income, and female-headed households in each block group in 1990. The dark-shaded areas have very high levels of disadvantage, the lighter-shaded areas exhibit moderate levels of disadvantage, and the areas with no shading are the least disadvantaged. Superimposed on the block groups are crosses and circles representing homicides committed between 1985 and 1995, coded as âgang-motivatedâ and âgang-affiliatedâ, respectively. The researchers defined a homicide as gang-motivated if it resulted from gang behaviour or relationships, such as an initiation ritual, the âthrowingâ of gang signs, or a gang fight. A homicide was defined as gang-affiliated if it involved a suspect or victim who was a gang member but did not arise from gang activity.
Regardless of the type of homicide, the map reveals a striking association between an areaâs level of disadvantage and the frequency of killings. Homicides are heavily concentrated in the northern section of the city, where socioeconomic disadvantage is most pronounced, and are virtually absent in the least disadvantaged southwestern neighbourhoods.
| Figure 1.1 | The Relationship Between Gang Homicides and Neighbourhood Socioeconomic Disadvantage in St. Louis, 1985â95. |
Source: Rosenfeld et al., 1999
St. Louis is not alone in exhibiting this strong spatial association between gang homicides and socioeconomic disadvantage. Nor is the relationship between disadvantage and crime confined to homicides or gang crimes. A century of research on neighbourhoods and crime consistently reveals the same pattern: crime tends to be concentrated in disadvantaged places (Bursik and Grasmick, 1993; Pratt and Cullen, 2005; Sampson et al., 2002).
The evidence presented thus far would seem to confirm the obvious: there is a straightforward relationship between crime and economic outcomes. Low socioeconomic status is associated with increased risk of criminal offending and victimization, and thus by extension, the level of criminal activity should reflect the extent to which the economy effectively âdelivers the goodsâ in society. It turns out, however, that the relationship between economic conditions and crime â even crimes committed for economic gain â is more complex than it appears at first glance. To understand these complexities, it is necessary to introduce some important conceptual and analytic distinctions.
Complicating the picture
Up to this point, we have framed the basic topic of inquiry for this book in rather loose language, by referring to the ârelationshipâ between the economy and crime. Our illustrations of a definite connection between socioeconomic deprivation and crime have in fact been based on empirical relationships and the statistical associations that represent them. However, such associations are typically of interest insofar as they allow us to make inferences about the presence and nature of causal processes. Do features of the economy actually produce or inhibit criminal activity in meaningful ways, and if so, how? It turns out that drawing causal inferences on the basis of empirical associations is a highly challenging task that ultimately requires theoretical guidance.
A full exposition of the intricacies of causal inference goes well beyond the scope of our discussion, but we can indicate here some of the complexities of interpreting statistical associations with illustrations from criminological research. This will serve as a useful starting point for our investigation into the complex relationship between crime and the economy. Figure 1.2 displays different...
