Download Official Statistics and Crime Rates: An Analysis of Uniform Crime Reports in the US and more Slides Statistics in PDF only on Docsity!
C H A P T E R
A weary English bobby (a popular nickname for British police officers) patrolling his foot beat on a chilly November night hears the unmistakable sounds of sexual activity from the dark entranceway of a closed greengrocer’s shop. He smiles to himself and tiptoes toward the sound. When he reaches the entranceway, he switches on his flashlight and booms out the favorite line of the stereotypical bobby: “What’s goin’ on ‘ere then?” The squeaking couple immediately come to attention and adjust their dress before the young man— obviously still in a state of arousal—stammers, “Why, nothing, constable.” The officer recog- nizes the woman as a local “slapper” (prostitute) and he vaguely recognizes the man (more of a boy of around 17, really) as a local supermarket worker. The constable reasons that he should arrest both parties for public indecency, but that would entail about an hour of paperwork (an hour in the warm police station with a nice cup of tea sounds good, though) and lead to the profound embarrassment of the poor boy. He finally decides to give the boy some sound advice about sexually transmitted diseases and a stern warning to the woman and sends them both on their way. This short story illustrates that official statistics are measuring police behavior as much as they are measuring crime. Sir Josiah Stamp, director of the Bank of England in the 1920s, cynically stated this criticism: “The government are very keen on amassing statistics. They collect them, raise them to the nth power, take the cube root and prepare wonderful dia- grams. But you must never forget that every one of these figures comes in the first instance from the village watchmen, who just puts down what he damn pleases” (quoted in Nettler, 1984, p. 39). I don’t recommend this kind of cynicism, but I do counsel that you keep a healthy skepticism about statistics as you read this chapter.
Measuring Crime
and Criminal Behavior
Measuring Crime
and Criminal Behavior
22 CRIMINOLOGY: THE ESSENTIALS
y (^) Categorizing and Measuring Crime and Criminal Behavior When attempting to understand, predict, and control any social problem, including the crime problem, the first step is to determine its extent. Gauging the extent of the problem means discovering how much of it there is, where and when it occurs most often, and among what social categories it occurs most frequently. It also helps our endeavors if we have knowledge of the patterns and trends of the problem over time. Note that we did not address “why” questions (why does crime occur, why is it increasing/decreasing, who com- mits it and why, and so on); such questions can only be adequately addressed after we have reliable data about the extent of the problem. However, all social statistics are suspect to some extent, and crime statistics are perhaps the most suspect of all. They have been collected from many different sources in many different ways and have passed through many sieves of judgment before being recorded. There is a wide variety of data provided by government and private sources to help us to come to grips with America’s crime problem, all with their particular strengths and weaknesses. The major data sources that we have can be grouped into three categories: official statistics, victimization survey data, and self-reported data. Official statistics are those derived from the routine functioning of the criminal justice system. The most basic category of official statistics comes from the calls made to police by vic- tims or witnesses and from crimes that the police discover on patrol. Other major categories of official crime data consist of information about arrests, about convictions, and about correctional (prison, probation/parole) populations.
y (^) The Uniform Crime Reports: Counting Crime Officially The primary source of official crime statistics in the United States is the annual Uniform Crime Reports (UCR) compiled by the Federal Bureau of Investigation (FBI). The UCR reports crimes known to the nation’s police and sheriff ’s depart- ments and the number of arrests made by these agencies; federal crimes are not included. Offenses known to the police are recorded whether or not an arrest is made or if an arrested person is subsequently prosecuted and con- victed. Participation in the UCR reporting pro- gram is voluntary, and thus all agencies do not participate. This is unfortunate for anyone hop- ing for comprehensive crime data. In 2009, law enforcement agencies active in the UCR program represented more than 295 million U. S. inhabit- ants—96.3% of the total population (FBI, 2010). This means that crimes committed in the juris- dictions of agencies representing about 4% of the population (about 12.5 million people) were not included in the UCR data.
▲ Photo 2.1 The J. Edgar Hoover building, headquarters of the FBI, in Washington, D.C. Annual Uniform Crime Reports are compiled by the FBI after local, county, and state criminal justice agencies send in their annual crime data.
24 CRIMINOLOGY: THE ESSENTIALS
Table 2.1 (Continued)
Male Female Offense 1999 2008 Change 1999 2008 Change Burglary 149,875 157,341 +5.0 23,106 29,055 +25. Larceny-theft 461,632 431,212 - 6.6 254,629 308,011 +21. Motor vehicle theft 60,540 43,801 - 27.6 11,331 9,344 - 17. Arson 8,317 7,116 - 14.4 1,373 1,291 - 6. Violent crime 305,271 286,255 - 6.2 63,085 63,943 +1. Property crime 680,364 639,470 - 6.0 290,439 347,701 +19. Other assaults 564,655 560,226 - 0.8 169,665 196,577 +15. Forgery and counterfeiting 38,570 31,947 - 17.2 24,253 19,631 - 19. Fraud 114,020 80,973 - 29.0 96,234 65,637 - 31. Embezzlement 5,768 6,575 (^) +14.0 5,634 7,039 (^) +24. Stolen property; buying, receiving, possessing
Vandalism 135,146 137,165 +1.5 24,317 28,213 +16. Weapons; carrying, possessing, etc. 87,790 93,112 +6.1 7,492 7,480 - 0. Prostitution and commercialized vice 19,762 12,133 - 38.6 25,240 25,164 - 0. Sex offenses (except forcible rape and prostitution)
Drug abuse violations 711,384 784,561 +10.3 155,256 185,201 +19. Gambling 4,481 2,227 - 50.3 800 350 - 56. Offenses against the family and children 65,172 51,268 - 21.3 18,481 17,303 - 6. Driving under the influence 714,457 667,017 - 6.6 134,279 181,391 +35. Liquor laws 298,874 241,328 - 19.3 84,444 89,999 +6. Drunkenness 370,924 347,399 - 6.3 55,670 66,883 +20. Disorderly conduct 262,713 243,865 - 7.2 82,332 89,530 +8. Vagrancy 13,302 12,037 - 9.5 3,130 3,550 +13. All other offenses (except traffic) 1,687,374 1,724,690 +2.2 444,312 517,358 +16. Suspicion 3,563 (^790) - 77.8 945 197 - 79. Curfew and loitering law violations 56,843 40,275 - 29.1 25,370 19,935 - 21. Runaways 37,359 26,923 - 27.9 54,048 34,363 - 36.
Source: FBI (2010a).
Chapter 2. Measuring Crime and Criminal Behavior 25
decreases) provide interesting information regarding why these gender differences exist. Figure 2.1, the FBI’s famous crime clock, further helps to put crime figures into perspective by indicating how often on an average day one of these crimes is committed. Remember that these are only rough estimates.
Cleared Offenses If a person is arrested and charged for a Part I offense, the UCR records the crime as cleared by arrest. A crime may also be cleared by exceptional means when the police have identified a suspect and have enough evidence to support arrest, but the suspect could not be taken into custody immediately or at all. Such circumstances exist when the suspect dies or is in a location where the police cannot presently gain custody. For instance, he or she is in custody on other charges in another jurisdiction or is residing in a country with no extradition treaty with the United States.
Figure 2.2 Percentage of Crimes Cleared by Arrest or Exceptional Means in 2009
Murder and Nonnegligent Manslaughter Forcible Rape Robbery Aggravated Assault Burglary Larceny-theft Motor Vehicle Theft
Violent Crime Property Crime
Source: FBI (2010a).
Figure 2.1 The FBI Crime Clock
One person is murdered every 31 minutes. One person is raped every 2.7 minutes. One person is assaulted every 7.2 seconds.
One home is burglarized every 9.1 seconds. One women is victimized by an intimatepartner every 1.3 minutes.One man is victimized every 6.7 minutes. One child is reported abused or neglectedevery 35 seconds. One person is killed in an alcohol-relatedtraffic crash every 29 minutes. One person becomes a victim of identitytheft every 8.7 seconds. One elderly person is victimizedevery 2.7 minutes. One hate crime is reported to the policeevery 73 minutes.
Homicide Rape Assault TheftOne home is victimized by theftevery 2.3 seconds. Burglary Domestic Violence Child Abuse and Neglect Drunk Driving Identity Theft Elderly Abuse Hate Crime
Source: FBI (2010a).
Chapter 2. Measuring Crime and Criminal Behavior 27
Year Crime
Violent Crime
Property Crime Murder Rape Robbery Assault Burglary
Larceny- Theft
Motor Vehicle Theft 1990 5,802.7 729.6 5,073.1 9.4 41.1 256.3 422.9 1,232.2 3,185.1 655. 1991 5,898.4 758.2 5,140.2 9.8 42.3 272.7 433.4 1,252.1 3,229.1 659. 1992 5,661.4 757.7 4,903.7 9.3 42.8 263.7 441.9 1,168.4 3,103.6 631. 1993 5,487.1 747.1 4,740.0 9.5 41.1 256.0 440.5 1,099.7 3,033.9 606. 1994 5,373.8 713.6 4,660.2 9.0 39.3 237.8 427.6 1,042.1 3,026.9 591. 1995 5,275.0 684.5 4,590.5 8.2 37.1 220.9 418.3 987.0 3,043.2 560. 1996 5,087.6 636.6 4,451.0 7.4 36.3 201.9 391.0 945.0 2,980.3 525. 1997 4,927.3 611.0 4,316.3 6.8 35.9 186.2 382.1 918.8 2,891.8 505. 1998 4,620.1 567.6 4,052.5 6.3 34.5 165.5 361.4 863.2 2,729.5 459. 1999 4,266.6 523.0 3,743.6 5.7 32.8 150.1 334.3 770.4 2,550.7 422. 2000 4,124.8 506.5 3,618.3 5.5 32.0 145.0 324.0 728.8 2,477.3 412. 2001 4,162.6 504.5 3,658.1 5.6 31.8 148.5 318.6 741.8 2,485.7 430. 2002 4,125.0 494.4 3,630.6 5.6 33.1 146.1 309.5 747.0 2,450.7 432. 2003 4,067.0 475.8 3,591.2 5.7 32.3 142.5 295.4 741.0 2,416.5 433. 2004 3,977.3 463.2 3,514.1 5.5 32.4 136.7 288.6 730.3 2,362.3 421. 2005 3,900.5 469.0 3,431.5 5.6 31.8 140.8 290.8 726.9 2,287.8 416. 2006 3,838.3 480.6 3,357.7 5.8 31.7 150.6 292.6 735.2 2,221.4 401. 2007 3,748.8 472.0 3,276.8 5.7 30.5 148.4 287.4 726.0 2,186.3 364. 2008 3,669.0 457.5 3,211.5 5.4 29.7 145.7 276.7 732.1 2,164.5 315. 2009 3,465.5 429.4 3,036.1 5.0 28.7 133.0 262.8 716.3 2,060.9 258.
Table 2.2 Crime Rates, 1990–
Source: FBI (2010a).
the Volstead Act in 1933, which effectively removed criminals from the alcohol business. It dropped even further during World War II when most young men were in uniform and overseas, showed a sharp rise when they returned, and then settled into a relatively peaceful period during the 1950s–early 1960s. Murder rates then started a precipitous rise beginning in the late 1960s. The late 1960s–early 1970s was a period of huge changes in American society. Opposition to the Vietnam War combined with the civil rights and feminist movements led to the widespread questioning of many of the fundamental values of American society. When values and norms are questioned, they become weaker in their ability to regulate behavior. Behavioral deregulation led to all kinds of experimentation with alternative lifestyles, including the use of drugs. The emergence of crack cocaine in the early 1980s led to a period of gang wars over territory, just like the gang wars over alcohol did in the 1920s. Crack cocaine is easier to make, conceal, and sell than barrels of beer or bottles of whiskey, so crack dealing is more of an “equal opportunity” enterprise than supplying illegal
28 CRIMINOLOGY: THE ESSENTIALS
alcohol was. Numerous young “gangbangers” took advantage of the opportunity for easy money, sparking a decade-long street war with other like-minded individuals. The decrease in the homicide rate in the early 1990s can be attributed to several factors including a large decrease in the crack market and in gang warfare as territories became consolidated by the strong pushing out the weak. Severe penalties for sale and possession of crack, and the danger from others trafficking in the same market, may have also driven out many dealers. One of the biggest factors in the drop in homicide rates is medical and technological improvements. Cell phones for reporting incidents are everywhere, and emergency medical technicians are alerted and dispatched swiftly. Once hospitalized, victims have the benefit of all that medical experts have learned about treating violent traumas since the Vietnam War. It is estimated that we would be experiencing 5 times the murder rate today if medicine and technology were at the same level as in 1960 (Harris, Thomas, Fisher, & Hirsch, 2002), which is something to remember when comparing U.S. rates to those of less developed coun- tries lacking America’s medical and technological advantages. Thus, many other factors known and unknown have contributed to the fluctuations in the homicide rate we have observed over the course of the 20th century, which makes it always difficult to explain crime trends to everyone’s satisfaction. Problems With the UCR UCR data have serious limitations that restrict their usefulness for criminological research, particularly research seeking to uncover causes of crime. Some of the more serious of these limitations are outlined below.
- The UCR data significantly underrepresents the actual number of criminal events in the United States each year. According to a nationwide victim survey, only 47% of victims of violent crime and 40% of victims of property crime indicated that they reported their victimization to the police
Figure 2.3 Murder Rates in the United States, 1900–
Source: Bureau of Justice Statistics, 2010. Key facts at a glance. Available at http://bjs.ojp.usdoj.gov/content/glance/hmrt.cfm
30 CRIMINOLOGY: THE ESSENTIALS
Because NIBRS data provide information about the offender and the victim (victims can identify physical characteristics of perpetrators), it can be used to try to resolve certain criminological issues. One issue is the disproportionately high rate of arrest for African Americans in the United States. The question for criminologists becomes this: Is the disproportion in arrests the result of disproportionately high black involvement in crime or the result of discriminatory arrest patterns of police? This issue was explored by D’Alessio and Stolzenberg (2003) using NIBRS data from 17 states and 335,619 arrests for rape, robbery, and aggravated and simple assault. Their results indicate the odds of arrest for robbery, aggravated assault, and simple assault were significantly greater for white offenders than for black offenders, but there was no significant racial difference in the probability of arrest for rape. In other words, white offenders were more likely to be arrested for violent crimes other than rape than black offenders. The researchers concluded that the disproportionately high black arrest rate is likely attributable to their disproportionately higher involvement in crime. For instance, African Americans committed 5,278 robberies in those states for which only 21.4% were arrested; whites committed 2,620 robberies for which 30.8% were arrested. Similar results based on NIBRS data were found in Pope and Snyder’s (2003) analysis of 102, violent incidents committed by juveniles; that is, white juveniles were significantly more likely to be arrested than black juveniles. y (^) Crime Victimization Survey Data and Their Problems Crime victimization surveys involve asking large numbers of people if they have been criminally victimized within some specified time frame regardless of whether they reported the incident to police. U.S. Census Bureau personnel interview a national representative sample of people age 12 or over on behalf of the Bureau of Justice Statistics (BJS) twice each year. This survey is known as the National Crime Victimization Survey (NCVS), and in 2008, a total of 77,852 people from 42,093 households were interviewed (Rand, 2009). The NCVS requests information on crimes committed against individuals and households, the circumstances of the offense, and personal information about victims (age, sex, race, income, and education level) and offenders (approximate age, sex, race, and victim–offender relationship). Figure 2.4 presents highlights from the 2009 NCVS report. Victimization surveys have their own dark figures as well as other problems that make them almost as suspect as the UCR. Some of these problems include the following:
- Crimes such as drug dealing and all “victimless” crimes such as prostitution and gambling are not revealed in such surveys for obvious reasons. And because murder victims cannot be interviewed, this most serious of crimes is not included.
- Because NCVS only surveys households, crimes committed against commercial establishments such as stores, bars, and factories are not included. This exclusion results in a huge underestimate of crimes such as burglaries, robberies, theft, and vandalism.
- Victimization data do not have to meet any stringent legal or evidentiary standards in order to be reported as an offense; if the respondent says he or she was robbed, a robbery will be recorded. UCR data, on the other hand, passes through the legal sieve to determine whether the reported incident was indeed a robbery.
- Other problems involve memory lapses; providing answers the respondent thinks the interviewer wants to hear; forgetting an incident; embellishing an incident; and any number of other misunder- standings, ambiguities, and even downright lies that occur when one person is asking another about his or her life experiences.
Chapter 2. Measuring Crime and Criminal Behavior 31
- Consistent with the above, there are suggestions that just as underreporting plagues UCR data, over- reporting may plague NCVS data (O’Brien, 2001). Whatever the case may be, we find many anoma- lies when comparing the two sources of data. For instance, substantially more crimes appear in police records than NCVS victims claim to have reported to the police. The discrepancy is easily explained for burglary and motor vehicle theft because the NCVS does not include commercial establishments in their reports. It is more difficult to explain discrepancies in violent crime reports, however. One explanation is that the NCVS does not include victims who are under 12 years of age, whereas the UCR does, although it is difficult to believe that children under 12 account for 15 to 20% of the violent victimization known to the police. NCVS researchers are aware of the many problems that arise when asking people to recall victimization and have initiated many interview improvements in their methodology, one of which is the bounding inter- view. This technique involves comparing reported incidents from the same household in the current interview with those reported 6 months prior. When a report appears to be a duplicate, the respondent is reminded of the earlier report and asked if the new report represents the incident previously mentioned or if it is different. Other techniques used to minimize some of the reported problems mention above are available on the NCVS website at www.icpsr.umich.edu/NACJD/NCVS/. Figure 2.5 provides an example of the kinds of questions asked by NCVS workers.
Source: Rand, M. (2009). Note number of households was 119,503,530 in 2007 and 121,141,060 in 2008.: Detail may not add to total because of rounding. Total population age 12 or older was 250,344,870 in 2007 and 252,242,520 in 2008. Total
~Not applicable. *Difference is significant at the 95%-confidence level.
Type of crime
Number of victimizations Ratesa^ Percent change 2007 2008 2007 2008 2007–2003b All crimes 22,879,720 21,312,400 ~ ~ Violent crimesc^ 5,177,130 4,856,510 20.7 19.3 - 6.9% Rape/sexual assault 248,280 203,830 1.0 0.8 - 18. Robbery 597,320 551,830 2.4 2.2 - 8. Assault 4,331,530 4,100,850 17.3 16.3 - 6. Aggravated 858,940 839,940 3.4 3.3 - 2. Simple 3,472,590 3,260,920 13.9 12.9 - 6. Personal theftd^ 194,060 136,710 0.8 0.5 - 30.1% Property crimes 17,508,530 16,319,180 146.5 134.7 - 8.1%* Household burglary 3,215,090 3,188,620 26.9 26.3 - 2. Motor vehicle theft 979,640 795,160 8.2 6.6 - 19.9* Theft 13,313,800 12,335,400 111.4 101.8 - 8.6*
Figure 2.4 Criminal Victimization, 2007–
Chapter 2. Measuring Crime and Criminal Behavior 33
y (^) Areas of Agreement Between the UCR and NCVS
To the extent that two or more data sources tell us the same thing, our confidence in both is increased. The UCR and NCVS agree on the demographics of crime in that they both tell us that males, the young, the poor, and African Americans are more likely to be perpetrators and victims of crime than are females, older per- sons, wealthier persons, and persons of other races. Both sources also agree as to the geographic areas and times of the year and month when crimes are more likely to occur. Over a 3-year period, O’Brien (2001) found that NCVS victims reported that 91.5% of those who robbed them and 87.7% of their aggravated assault assailants were male, as were 91.2 and 84.3%, respectively, of those arrested for those offenses. Likewise, NCVS victims reported that 64.1% of those who robbed them and 40% of their aggravated assault assailants were African American. These percentages fit the UCR arrest statistics for race almost exactly: A total of 62.2% arrested for robbery were African American, as were 40% of those arrested for aggravated assault. Thus, the two data sets agree almost perfectly. Comparisons of UCR and NCVS data have often proven very useful to resolve issues such as these. For instance, some feminist criminologists are of the opinion that women are becoming more “masculinized” as a result of assuming “male” roles in the workforce, and that this is reflected in the increased rates of female arrests for violent crimes. Darrell Steffensmeier and his colleagues (Steffensmeier, Zhong, Ackerman, Schwartz, & Agha, 2006) used a comparison of data trends reported in the UCR and the NCVS from 1980 to 2003 to explore the issue of whether the violent crime gap between males and females is closing. They found that both sources reported little or no changes in the gender ratio for violent crimes such as murder, rape, and robbery, but that the UCR reports indicated a sharp rise in assaults by females. Does this mean that women became more violent over the period examined, or does the increase reflect the behavior of the police more than the behavior of women? The authors conclude that net-widening policy shifts have esca- lated the arrest proneness of females for “criminal assault” (e.g., policing physical attacks/threats of mar- ginal seriousness) rather than that women have become more violent. In other words, UCR increases in female arrests for assault are explained by changes in police policy, something that could not have been determined without examining both data sources. The addition of the NCVS and NIBRS to the nation’s crime databases thus has great utility for settling some major quarrels among criminologists of different ideological persuasions, although not to the satisfaction of everyone. Note from Table 2.1 that this trend was still in evidence when comparing UCR arrests for aggravated assault and simple (labeled “other” in the table) assault from 1999 to 2009. Female aggravated assault decreased by 2.5%, but female arrests for simple assault increased by 15.9%. The overall female increase in violent crime arrests of 1.4% is entirely a function of the huge increase in female robbery arrests over the period.
y (^) Self-Report Crime Surveys and Their Problems
Self-report surveys of offending provide a way for criminologists to collect data without having to rely on government sources. Questionnaires used in these surveys typically provide a list of offenses and request subjects to check each offense they recall having committed and how often, and sometimes whether they have ever been arrested, and if so, how many times. Self-report surveys have relied primarily on college and high school students for subjects, although prison inmates and probationers/parolees have also been surveyed. Several studies have addressed the issue of the accuracy and honesty of self-reported offenses in various ways, and the results have generally been encouraging, at least for uncovering the extent of minor offenses.
34 CRIMINOLOGY: THE ESSENTIALS
On average, known delinquents and criminals disclose almost 4 times as many offenses as the non-delin- quents. Cartier, Farabee, and Prendergast’s (2006) study of methamphetamine use and self-reported crime and recidivism among 614 California parolees, for instance, found that meth use was significantly related to self-reported crime and claim a high level of agreement between self-reported crime and actual crimes committed. Had differences such as these not been found, the validity of the self-report procedure would have been in doubt. The greatest strength of self-report research is that researchers can correlate a variety of characteristics of respondents with their admitted offenses that go beyond the demographics of age, race, and gender. For instance, they can attempt to measure various constructs thought to be associated with offending, such as impulsiveness, empathy, and sensation seeking, as well as their peer associations and their attitudes. The evidence indicates that self-reported crime measures provide largely accurate information about some illegal act that occurred at some time in the respondent’s life. However, there are a number of reasons why self-report crime surveys also provide a distorted picture of criminal involvement.
- The great majority of self-report studies survey “convenience” samples of high school and college students, populations in which we don’t expect to find many seriously criminally involved individu- als. Most self-report studies thus eliminate the very people we are most interested in gathering information about. One strength of the self-report method, however, is that is appears to capture the extent of illegal drug usage among high school and college students, something that neither the UCR nor the NCVS attempts to do.
- Self-report studies typically uncover only fairly trivial antisocial acts such as fighting, stealing items worth less than $5, smoking, and truancy. Almost everyone has committed one or more of these acts. These are hardly acts that help us to understand the nature of serious crime. A connected problem is that some researchers lump respondents who report one delinquent act together with adjudicated delinquents who break the law in many different ways and at many different times.
- Even though most people are forthright in revealing minor antisocial behaviors, most people do not have a serious criminal history, and those who do have a distinct tendency to underreport their crimes (Hindelang, Hirschi, & Weis, 1981). As the number of crimes people commit increases, so does the proportion of offenses they withhold reporting, with those arrested for the most serious offenses having the greatest probability of denial (Farrington, 1982).
- Males tend to report their antisocial activities less honestly than females and African Americans less honestly than other racial groups (Cernkovich, Giordano, & Rudolph, 2000; Kim, Fendrich, & Wislar, 2000). This evidence renders suspect any statements about gender or racial differences regarding antisocial behavior that are based on self-report data. When it comes to relying on self-report data to assess the nature and extent of serious crime, it does us well to remember the gambler’s dictum: “Never trust an animal that talks.” We should not end on a pessimistic note about self-reports, however. Many of the major multi-million dollar longitudinal studies going on today have built-in safeguards against researchers naïvely taking sub- jects at their word. A number of studies verify self-report accounts with police records and other social agencies, a practice which further helps us to gain a grasp on the reliability of self-report studies. For instance, a large longitudinal (following the same people across the lifespan) cohort study (one that studies a set of individuals who share a common characteristic, such as being born in the same month in the same geographic area) showed that individuals from the lowest social class category reported 3.21 times more
36 CRIMINOLOGY: THE ESSENTIALS
the number of pending cases from 2005 to 2009. In 2009, investigations resulted in 156 convictions of corporate criminals, and numerous other cases are pending trials and plea agreements, although this is likely a drop in the bucket compared to the actual number of crimes committed by corporate criminals. During 2009, the FBI secured $6.1 billion in restitution orders and $5.4 million in fines from corporate criminals (www.fbi.gov/stats-services/publications/financial-crimes-report-2009).
y (^) The Dark Figure of Crime The dark (or hidden) figure of crime is that portion of the total crimes committed each year that never comes to light. Figure 2.7 presents three diagrams that show the different dark figures for the three major measures of criminal behavior. (The dark figures are represented by the dark shading in each diagram.) Each diagram shows the degree to which crimes of various levels of seriousness are most likely to be detected by each measure (“victimless” crimes excluded). In the top diagram displaying UCR data, you can see that very few trivial offenses are reported in official statistics, and most of those that are will be dis- missed as unfounded by the police. For official statistics, then, the dark figures are highly concentrated at the nonserious end of the crime seriousness spectrum. The middle diagram reveals that the dark figures for victimization data are primarily concentrated in the nonserious end of the spectrum also, although to a lesser degree than in the case of official data. The failure of victimization data to pick up these minor offenses is largely due to survey subjects not remembering all incidences of victimization. In the bottom diagram, we see that most of the dark figures in the case of self-reports are concentrated in the upper end of the seriousness continuum rather than the lower end. This is partly due to the fact that (a) nearly all self-report surveys exclude the most persistent serious offenders from their subject pools, and (b) many of the most serious offenders who remain in self-report subject pools do not reveal the full extent of their criminal histories.
y (^) What Can We Conclude About the Three Main Measures of Crime in America? All three main measures of crime in America are imperfect measures, and which one of them is “best” depends on what we want to know. UCR data are still probably the best single source of data for studying serious crimes, and indeed, the only one for studying murder rates and circumstances. For studying less serious but much more common crimes, either victimization or self-report survey data are best. If the interest is in drug offenses, self-reports are the preferable data source. Because all three data sources converge on some very important points about crime, they enable us to proceed with at least some confidence in our endeavors to understand the whys of crime. The basic demographics of crime constitute the raw social facts that are the building blocks of our criminological theories. If street crime is concentrated among the lower socioeconomic classes and in the poorest neighborhoods, we can begin to ask such questions as, does poverty “cause” crime, or does some other variable(s) cause both? Is social disorganization in a neighborhood independent of the people living in it, or completely dependent on the people living in it? Why do females always and everywhere commit
Chapter 2. Measuring Crime and Criminal Behavior 37
Figure 2.7 Differing Proportions of Reported/Unreported Crimes for the Three
Major Measures of Crime
*Light shading = proportion of crimes reported. Dark shading = proportion not reported.
“No-Crime” Crimes(unsubstantiated crimes)
Seriousness
Victimization Survey Data
Self-Reported Data
Seriousness
Crime Prevalence
(a) Low Low
(b)
(c)
LowLow High
High
High
Crimes Reported to Police
High
Crime Prevalence
Seriousness
Crime Prevalence LowLow High
High
Chapter 2. Measuring Crime and Criminal Behavior 39 S U M M A R Y
- Crime and criminal behavior are measured in several ways in the United States. The oldest measure is the FBI’s Uniform Crime Reports (UCR), which is a tabulation of all crimes reported to the police in most of the jurisdictions in the United States in the previous year. The UCR is divided into two parts: Part I records the eight Index crimes (murder, rape, robbery, aggravated assault, burglary, larceny/theft, and arson), and Part II records arrests made for all other crimes.
- UCR data seriously underestimate the extent of crime because the UCR only records reported crimes, ignores drug offenses, and only reports the most serious crime in a multiple-crime event. The problems with the UCR led to the implementation of the National Incident-Based Reporting System (NIBRS).
- The second major source of crime statistics is the National Crime Victimization Survey (NCVS). This survey consists of many thousands of interviews of householders throughout the United States, asking them about their crime victimization (if any) during the previous 6 months. The NCVS also has prob- lems because it leaves out crimes against commercial establishments and relies exclusively on the memory and the word of interviewees.
- The third source of crime data is self-report data collected by criminologists themselves. The advantage of self-report data is that they are derived “from the horse’s mouth,” and typically the questionnaires used ask about “victimless” offenses not covered in either the UCR or NCVS. The major problems with self- report data is that it does not capture serious criminal behavior and is subject to dishonesty in the form of underreporting, especially underreporting by those most seriously involved in criminal activity.
- The UCR, NCVS, and self-report data come to different conclusions on a variety of points, but they agree about where, when, and among whom crime is most prevalent, and the fact that crime has fallen dra- matically in the United States over the past decade. Taken together, then, we have a fairly reliable picture of the correlates of crime from which to develop our theories about explanatory mechanisms.
D I S C U S S I O N Q U E S T I O N S
- Go to the website www.fbi.gov/wanted/topten/fugitives/fugitives.htm-1 for the FBI’s ten most wanted fugitives, and research the background and crimes of one of the people listed there. Then write a one- to two-page summary and report to the class.
- Do you think it wise to make “authoritative” statements or formulate theories of criminal behavior, espe- cially serious criminal behavior, based on self-report data?
- Can you think of other problems possibly associated with asking people about their delinquent or criminal behavior or their victimization other than those discussed in the chapter?
- If you were the American “crime czar,” what would you do to get the various law enforcement agencies to fully implement NIBRS—and no, you can’t just order them to do so.
U S E F U L W E B S I T E S
Bureau of Justice Statistics. www.ojp.usdoj.gov/bjs/ National Archive of Criminal Justice Data. www.icpsr.umich.edu/NACJD/
40 CRIMINOLOGY: THE ESSENTIALS
National Crime Victimization Survey Resource Guide. www.icpsr.umich.edu/NACJD/NCVS/ National Incident-Based Reporting System Resource Guide. www.icpsr.umich.edu/NACJD/NIBRS/ Uniform Crime Reports. www.fbi.gov/ucr/ucr.htm
C H A P T E R T E R M S
Cleared offenses Crime rate Dark figure of crime Hierarchy rule
National Incident-Based Reporting System (NIBRS) National Crime Victimization Survey (NCVS) Part I offenses (or index crimes)
Part II offenses Self-report surveys Uniform Crime Reports (UCR)
