II. Homicide Trends Over Time
One of the most remarkable findings in the study of urban violence is that homicide rates fell sharply in U.S. cities in the 1990s. In fact, homicides plunged to their lowest point in 35 years, making this drop critical to any discussion of homicide. That is, any effort to understand homicide requires an examination of homicide trends over time, particularly this rather remarkable, unexpected crime drop of the 1990s. To that end, this research paper will provide statistical information on urban homicide trends since the 1980s, drawing specifically on SHRs. After documenting some important changes, some of the leading explanations for the crime drop will then be outlined, to give the reader an understanding of the level and nature of work being conducted to understand this precipitous decline.
A. The Crime Drop
Researchers use time series data of total homicide rates to document the crime drop. As stated in the introduction, homicide is the most accurately measured and reported offense, making it the best benchmark when trying to illustrate changes in criminal offending over time. In addition, homicide is the most serious crime, leading it to be the most widely used among academicians. For these reasons and others, homicides provide a useful and accurate account of crime trends.
Time series data show that homicides averaged 19.0 per 100,000 population in 1980, and rates tended to fluctuate between 19 and 22.5 until 1991, when homicides peaked (22.5 per 100,000 population). That is, examination of SHR data reveals that homicides dropped by 20% from 1980 to 1985, but then rose by 47% from 1985 to 1991. Starting in 1991, homicides started a steady decline until 2000, falling to their lowest rate in 2000, or a drop of 46%. Since 2000, the rate of homicide has been largely stable until 2006 or so when an increase was observed. Overall, SHR data have documented a dramatic rise in homicides in the late 1980s, followed by the precipitous decline in the 1990s. This incredible crime drop has gained widespread attention as scholars have searched for answers (FBI, 2008).
B. Explanations for the Crime Drop
The drop in homicide rates occurred without warning, leading to an explosion of newspaper articles, TV reports, and other media accounts. Scholarly attention soon followed with a list of potential explanations, including greater police presence, prison expansion, reduced handgun availability, tapering drug (specifically crack cocaine) markets, gains in the economy, and age shifts in the population (Blumstein & Wallman, 2001). While the list continues to grow, some of the explanations receiving the greatest attention in the literature are outlined below.
Rise in Imprisonment Rates. An understanding of the changes in crime rates cannot occur without some consideration of the political and legal context of the time period. The enormous growth in “get tough on crime” policies that began in the 1970s is no exception. The expansion of the incarcerated population started in the mid-1970s, and by 2000 more than 2 million persons were incarcerated— 4 times the prison population of 1970. Because the rise in incarceration rates corresponds closely with the decline in homicide rates, some researchers linked the two. For example, while homicides were dropping from 1991 to 2001 in large cities, the Bureau of Justice Statistics reports incarceration rates rose by 54.2% during this time period (a rate change of 310 to 478 per 100,000 residents nationally). The rise in incarceration, backed by structured sentencing (i.e., “get tough” on violent and drug-related criminal offenders) and other conservative criminal justice policies, is one of the longest trends documented in the literature. Given the steady and prolonged trends in both rates of violence and incarcerations, it is not surprising that a number of scholars argue for the association between the two.
Increase in Police Presence. One response to rising crime rates is to hire more police officers. There is evidence that this was indeed a response to crime trends based on annual figures in the UCR. These reports tell of more police on the street, particularly in the 1990s when the FBI reports an extra 50,000–60,000 officers nationally (Levitt, 2004). On average, the police force size was 236.1 per 100,000 city residents in 2000, up from 206.9 per 100,000 persons in 1980 in large U.S. cities (Parker, 2008).These trends provide scholars with reasons to argue that increasing police presence is a likely predictor of the crime decline in the 1990s.
Diminishing Drug Markets. A link between violence and illicit drug markets is another major theme in the crime drop debate. Crack cocaine markets, which grew throughout the mid-1980s and peaked in the early 1990s, were related to homicide trends during this same time period (Blumstein, 1995). In fact, researchers found that drug markets contribute to violence, and studies have pointed to crack cocaine patterns specifically as related to trends in urban violence (Blumstein & Rosenfeld, 1998; Cook & Laub, 1998; Goldstein 1985). While determining how to best capture the impact of drug markets has hindered much of this research, police arrests for drug (specifically cocaine) sales represent one way to tap the level of drug activity in a given area or city. The UCR has shown that drug arrests for sales/manufacturing have exploded, growing by two and a half times from 1982 to 2003 alone (from 137,900 to 330,600). Thus, evidence of the waning crack market in the 1990s, or at least the growing enforcement of drug sales in recent times, has placed drug markets at the forefront of the crime drop debate.
Improving Economy of the 1990s. The link between economic factors and crime cannot be understated, so it comes as no surprise that the economic improvements of the 1990s have gained attention as a plausible explanation for the crime decline. In fact, in accord with labor statistics, the unemployment rate rose during the recessions of the early 1980s and early 1990s, recovering after both periods. On the other hand, the unemployment rate steadily declined throughout the 1990s, where employment gains for males and females correspond to the crime drop of this period. The unemployment rate alone fell from 6.8 in 1991 to 4.8 in 2001 (a drop of 30% in 10 years). Other economic performance indicators suggest better times for many Americans in the 1990s, as well as growth in major industries like information technology and services. Given that the 1990s mark a time of sustained economic growth and prosperity, these economic improvements are likely contributors to the crime drop.
Guns and Gun Control Policies. Finally, while explanations derived from guns and gun control policies drew a lot of attention early in the crime drop debate, mainly because such a large percentage of homicides are gun-related (Cook & Laub, 1998), interest in this explanation has diminished over time. The early interest in the relationship between violent crime and firearms made sense—the rate of violent crimes committed with firearms rose in the 1980s and 1990s and subsequently dropped. But over time, scholars have downplayed the degree to which gun control and concealed weapon laws contributed to the crime drop (Levitt, 2004). For example, some researchers found that the percentage of total killings by young males remained stable during the time of the crime drop, which was troubling since young males are much more likely to use a gun in a homicide than others, and other researchers discovered that the passage of the Brady Act gun control legislation in 1993 had no influence on homicide trends. Adding to the downfall of this explanation, researchers evaluating gun buyback programs and other gun control policies found that these programs also had little to do with reduction in gun violence. Even the highly publicized concealed weapon laws link to lower violent crime came under scrutiny (Lott & Mustard, 1997) when researchers revealed that the decline in crime actually predated the passage of many concealed weapon laws.
Since many of these explanations represent early responses to the crime drop, the research paper will now turn to more recent trends in the study of homicide. Clearly understanding homicide trends, particularly the crime drop of the 1990s, remains a critical focus. Moving beyond time series data of total homicide rates, scholars have acknowledged that since homicide trends differ across groups (Blumstein & Rosenfeld, 1998; Cook & Laub, 2002; Parker, 2008), these characteristics need to be accounted for in the crime drop debate. Current examples include Heimer and Lauritsen’s (in press) examination of trends in violence against women; LaFree, O’Brien, and Baumer’s (2006) exploration into racial patterns in arrest rates for multiple violent offenses; and Parker’s (2008) effort to account for the role of local labor markets in the study of race-specific homicide trends since the 1980s. All of these efforts acknowledge the diversity in the American population, including the differential levels of involvement in violence by the various groups, and argue that accounting for the differences across groups will advance understanding of the crime drop. To illustrate, a closer look at homicide trends is offered, involving two specific characteristics— racial groups and intimate partners.