1999, Western Criminology Review. All Rights Reserved.



An Analysis of Curfew Enforcement and Juvenile Crime in California

 

Mike A. Males

Dan Macallair

 

Citation: Males, Mike and Dan Macallair. 1999. "An Analysis of Curfew Enforcement and Juvenile Crime in California." Western Criminology Review 1 (2). [Online]. Available: http://wcr.sonoma.edu/v1n2/males.html.


ABSTRACT

In recent years cities and localities across the country have expanded the use of youth curfews to address growing public concern about juvenile crime and violence. By reducing the number of youth on the street during certain hours, curfews are assumed to lower the risk factors associated with youth crime. Curfews have been widely cited by policy makers as an effective tool for reducing youth crime. However, no comprehensive analysis of the effects of these laws has been completed. This study analyzes arrest, reported crime, and mortality data from jurisdictions throughout California for the 1980-97 period. There is no support for the hypothesis that jurisdictions with curfews experience lower crime levels, accelerated youth crime reduction, or lower rates of juvenile violent death than jurisdictions without curfews.


Keywords: crime policy, curfew, youth crime, incapacitation, youth violence, violent death

An Analysis of Curfew Enforcement and

Juvenile Crime in California1

National and state leaders have endorsed the implementation and enforcement of stronger "status offense" laws to control youth crime. So-called status offenses apply to youth but not adults, such as running away from home, truancy, underage drinking, incorrigibility, and presence in public during certain hours. The last of these, night time and schoolday curfews, have won the most attention and have been cited for their potential to reduce juvenile crime (Krikorian 1996; Ricardi 1997). Proponents argue that they protect youth and the public from violence and criminality and prevent violators from more serious offenses (Ruefle, Reynolds, and Brantley 1997). Detractors warn that the arrest of youth for acts that would not be crimes if committed by adults violates basic constitutional guarantees, leads to antagonism between non-criminal youth and law enforcement, and is an inefficient way to control crime (Harvard Law Review Association 1997).

A major underlying assumption of curfews is that they reduce risk by removing juveniles from public space. However, there is no systematic study of the effects of curfew. Instead of presenting controlled data, proponents and opponents of curfews have made anecdotal statements to the media such as, "Monrovia, California's curfew adoption was followed by a 32 percent decline in residential burglaries" (Ricardi 1997). However, this assertion and others like it require scrutiny. For example, Monrovia had already experienced a forty percent decline in juvenile burglaries, and only had about a dozen juveniles arrested for burglary for each of several years prior to adopting curfew (Criminal Justice Statistics Center 1978-97). Their decline is also not compared to cities that did not enforce curfews. Thus, it is interesting to note that the following hypothetical but factual statement could be made with equal justification: "In 1992, San Francisco authorities dismantled their previously vigorous curfew enforcement, which had resulted in 1,400 arrests during the previous five years. Only three curfew arrests were made during 1993-97. Crime plummeted. From 1992 to 1997, juvenile murders declined fifty percent, property crimes reported to police declined thirty-six percent, and violent crimes reported to police declined by forty-one percent, the latter of which was the largest crime decrease of any large California city. Therefore, abolishing curfews is a crucial step to reduce youth crime and victimization." 

Surprisingly, given that curfew arrests of California youth rose fourfold between 1989 and 1997 (from 5,400 to 22,400, respectively; see Figure 1), we are not aware of any systematic study of whether California's curfew laws control crime. A search turned up only twenty-five studies of curfews nationwide (three in California) since 1990. These reach mixed, often opposite, conclusions, more often philosophical rather than analytical in nature. Even those that provide statistical analysis are limited to selected time periods, specific types of crimes and areas, and do not compare jurisdictions that enforce curfews to those that do not (Ruefle et al. 1997; Harvard Law Review Association 1997). Several representative research studies follow.

Figure 1

For example, the U.S. Conference of Mayors recently surveyed the nation's 1,010 cities with populations of more than 30,000. The survey asked law enforcement authorities if they could credit curfews for any recent improvements in juvenile crime. Only one-third, or 347, of the cities responded to the survey. Of those, eighty-eight per cent claimed that their curfews reduced youth crime, even though the survey "did not include a statistical analysis of the effect curfews have had on crime" (Wilgoren and Fiore 1997).

A potentially interesting report on "comprehensive, community-based curfew programs" in seven cities was done by the Office of Juvenile Justice and Delinquency Prevention (OJJDP 1996: 1). The report concludes that "information by communities where curfews have been implemented indicates that comprehensive, community-based curfew programs are helping to reduce juvenile delinquency and victimization." However, this study does not present consistent or controlled information necessary to support the conclusion. For some cities, selected juvenile victimizations are cited; for still others, selected juvenile arrests; for others, selected crimes (such as burglaries or vehicle theft) reported to police. No controlled comparisons are made between cities that did and did not enforce curfews. Clearly, for any jurisdiction and year, some crimes and victimizations will increase and others will decrease due to natural fluctuations that are not controlled for. Selective citation of these findings are practically useless for evaluation purposes.

Individual cities, including those cited in the OJJDP study, have provided anecdotal and numerical claims regarding the effects of curfew. The city of Los Angeles analysis appears to be the most complete (LAPD 1998). From July 1997 through July 1998, the city issued three semiannual reports on its Enhanced Curfew Enforcement Effort for selected nights and areas. The reports reached contradictory conclusions. One report finds that stronger curfew enforcement "has impacted" and another that curfew enforcement "has not greatly impacted" violent crime, youth arrests, and youth victimization. Interestingly, the six-month period of the most intensive and stepped-up curfew enforcement (4,810 arrests from May to October 1997) produced no effect on crime or victimization. When curfew arrests and enforcement were cut back sharply in the following period, youth crime and victimization declined in the enforcement areas.

Finally, an analysis of Monrovia's daytime curfew by Klein (1998), submitted as a declaration in a lawsuit and cited by Monrovia in the press (Harrahill v. Santoro 1998), concludes that the curfew reduced Part I (FBI felony violent and property) crime in the city by at least seven percent from 1993 to 1996. Klein's conclusion is based on the number of Part I crimes reported to police in 1996 (the most recent year for which statistics are available) compared to 1993 (the last full year before the adoption of curfew in October 1994) for school hours and nonschool hours. In effect, Klein assumes all Part I crimes in Monrovia are committed by juveniles and that the entire reduction evident in 1996 can be explained by the curfew ordinance. However, crime clearance reports by the FBI (1997) show that only one in five reported Part I crimes are committed by (that is, cleared by the arrest of) juveniles, making this measure too imbued with adult "noise" to serve as a reliable indicator of youth crime. As will be shown in the case study of Monrovia (below), Klein's figures appear to be in error, and, when corrected, show a larger decrease in reported crime during non-curfew than during curfew enforcement hours. Overall, Klein's method fails to utilize the data available to reach a more precise conclusion.

In short, even though data exist to make more than guesses about the effects of curfews, there is insufficient research to reach a conclusion about the effects of curfew on crime or youth safety. Claims to date that curfews affect crime and youth behavior represent little more than unsupported assertions. Without long term, large scale, and controlled statistical analyses, it is not possible to reach preliminary conclusions as to whether curfew enforcement reduces, increases, or has no effect on crime.

Top of page

METHODOLOGY

Data on crime by offense type, age of arrestee, year, and county are taken directly from the California Department of Justice's Law Enforcement Information Center (LEIC), annually reported statewide in Crime and Delinquency in California and by county and city in California Criminal Justice Profile through 1995, and statewide and by county in their 1996 and 1997 updates. This study uses the LEIC's definition of youth (age 10-17) and adult (age 18-69) and estimates of population for each group in calculation of crime rates. The categories of youth crime examined are all arrests, felonies, violent felonies, homicides, property felonies, and misdemeanors (Division of Law Enforcement 1978-95; Criminal Justice Statistics Center 1978-1997). "All arrests" and "misdemeanors" do not include arrests for status violations. Population figures are from the California Department of Finance's Demographic Research Unit.

Our assumption is that youth crime levels (expressed as rates per 100,000 youth age 10-17) should be reduced in locales with greater status/curfew law enforcement. California counties are laboratories for the study of this question. Los Angeles (1,166 curfew arrests per 100,000 youth in 1996-97), San Diego (669), Fresno (626), Orange (553), and Santa Clara (514) counties have rates of curfew arrest many times higher than San Francisco (1), Sacramento (3), Alameda (53), and Contra Costa (121) counties. Further, youth violent crime arrest rates in 1996-97 varied sixfold, from 1,666 per 100,000 youth in San Francisco County to 261 per 100,000 in Riverside. Youth property felony rates range from 1,554 in Fresno and 1,499 in San Francisco to 649 in Riverside. San Francisco's rates are elevated because it is the only California county wholly comprised of a city. Relative to adults, the youth felony arrest rate is 1.5 times higher in Santa Clara and 1.2 times higher in Orange, twice the net youth felony rate of Riverside (0.6) and Sacramento (0.7). Rates of status crime arrests varied twenty-fold in 1996-97, from 1,623 per 100,000 youth in Fresno and 1,341 in Los Angeles counties to 77 in Sacramento and 93 in San Francisco.

Whether enforcement of curfews is related to higher or lower levels of youth crime is examined by means of a standard correlation analysis. Correlations determine whether variable A is related to variable B positively (A rises as B rises, A falls as B falls), negatively (A rises as B falls, A falls as B rises), or not at all. The type of correlation test of annual changes in arrest rates used here is known as "differencing." Correlation by the differencing method factors out the artificial trending patterns natural to time series which hamper standard correlations. Instead of comparing the entire time-series A with time-series B, differencing compares the year-to-year changes in A (in this case, rates of curfew enforcement) with corresponding year-to-year changes in B (rates of other crime or violent death).

This analysis examines whether year-by-year increases or decreases in the rates of police enforcement of curfews affect the corresponding rates of youth crime. The formula produces a statistic (r) in which a perfect positive correlation is 1.00, a perfect negative correlation is -1.00, and no correlation is 0. If stronger enforcement of curfew laws against youths over the 1980-97 period is negatively correlated with rates of youth crime in a particular jurisdiction (that is, more curfew arrests are accompanied or followed by lower levels of youth crime), it could be argued that curfew reduces crime. A standard significance level (p < .05) is used, which in this case means there is a "statistically significant" odds of one in twenty or less that the similarity (or dissimilarity) between two series is too great to be due to random chance alone. In practice, this means that a few statistically significant findings will themselves turn up by chance. For example, of seventy-two separate correlations, three or four (.05 x 72) would be expected to show significant results by chance. This is especially true in cases in which correlations of aggregate measures (i.e., all crimes) may not be independent, but simply may reflect correlations of certain large sub-measures (i.e., felonies) they include. A few significant findings are not usually considered important, especially if they are in opposite directions.

Since curfew laws by definition are not applied to adults, one might expect them to affect youth but not adult crime rates. For this reason, both raw youth crime and violent death rates and net youth crime and violent death rates (expressed as a ratio to corresponding adult rates) for each year, type of crime or death, and state/county/city jurisdiction are compared, where feasible. Six different analyses are conducted:

(1) Statewide curfew arrest rates and crime rates are examined separately by race/ethnicity for all youths in aggregate and for California's four major groups (white non-Hispanic, Hispanic, Black non-Hispanic, and Asian/other non-Hispanic) for the 1978-97 period, the maximum period for which reliable data exist at this writing. The analysis of six types of crime for all groups in aggregate plus the four racial groups yields thirty separate correlations (six times five) for raw youth crime rates, and thirty for youth crime rates net of adult crime rates.

A statewide comparison of total crime arrests and curfew arrests is also provided. This analysis is intended to examine a possible relationship between raw curfew arrests and overall arrest patterns. It might be argued that statewide statistics fail to capture local variations. Therefore, several local analyses are also performed:

(2) Curfew arrest rates and youth crime rates are compared for the twelve most populous counties for the 1980-97 period; again, the maximum time period for which reliable figures are available. 2 These counties include twenty-two million people, two-thirds of the state population in 1995. They also account for eighty percent of state arrests. This analysis of twelve counties for six types of crime yields seventy-two separate correlations for raw, and net youth crime rates, respectively.

The strength of the arrest measure is that it includes only juvenile offenses. Its weakness is that most crimes reported to police do not result in arrest, rendering "arrests" a problematic measure of "crime." Further, it may be argued, youth arrest rates for other crimes may not be independent of curfew enforcement. More vigorous enforcement of curfews might affect juvenile arrest rates, perhaps increasing them due solely to greater, proactive police attention to and contact with juveniles (Santoro 1998). Therefore, two analyses of another available measure, crimes reported to police, are also included.

(3) We also compare local curfew and reported rates of Part I violent, homicide, property, and arson offenses, and all Part I crimes reported to police for the twelve major counties for the 1980-97 period. Part I crimes are chosen because they are the "index" crimes used by the FBI and are consistently defined and reported over time. The weakness of this measure is that it includes no age detail; in fact, seventy-five to ninety percent of Part I crimes are committed by adults, according to FBI (1997) analysis of crimes cleared by arrest. This introduces the huge "noise" factor of changes in adult offenses blotting out the "signal" of changes in juvenile offenses. Nevertheless, we examine reported offenses here because advocates of curfews have cited this measure. Arson crimes reported to police, though small in number, are analyzed separately since this is the only Part I crime that juveniles commit a large share of (around half, by FBI estimate). This analysis of twelve counties by five different measures of reported crime yields sixty correlations.

(4) Local curfew and Part I crime rates and trends for all twenty-one cities over one hundred thousand population in Los Angeles and Orange counties are compared for the 1990-96 period (see
Appendix C tables for list). The first analysis is correlation of changes in average annual status crime arrest rate changes and average annual index crime rate changes in the 1994-96 period compared to the 1990-93 period (status arrests by city are available only through 1996). Unlike populous counties, smaller cities tend to produce extreme, "outlier" values that can skew averages and correlations. Therefore, a second analysis groups the cities into two natural categories: the thirteen that display low and/or declining rates of curfew law enforcement versus the remaining eight, which show high and rising curfew enforcement. Both average and median changes in index crime rates are reported.

(5) Since curfew advocates claim enforcement reduces juvenile exposure to danger, local curfew enforcement and rates of violent death among youth are compared for the 1980-96 period to the twelve largest counties. Both absolute (raw) and net (compared to adult) youth rates of the four major types of violent death (homicide, suicide, motor vehicle accidents, and other accidents) and all violent deaths in aggregate are examined. This comparison of five types of death for twelve counties yields sixty absolute and sixty net correlations.

(6) Finally, there are two case studies. The city of Monrovia's experience with a daytime, anti-truancy curfew is compared for juvenile arrests and for crimes reported to police on a monthly basis from January 1992 through July 1997. This yields a before and after period evenly bracketing the curfew's adoption in October 1994. The analysis includes a comparison of crime trends during and outside of curfew enforcement periods and comparisons with the crime experiences of eleven neighboring cities. The full study is in
Appendix A.

In the second case study, youth arrest, reported crime, and youth violent death trends in the cities of San Francisco (which went from strong to zero enforcement of curfews in the 1990s) and San Jose (which went from zero to strong curfew enforcement in the 1990s) are examined for the period from 1985 through the latest available (mid-1998). Both absolute and net (of adult) trends are included. This last comparison examines curfew effects under extreme conditions -- a change from vigorous to nil enforcement versus nil to vigorous enforcement. The full study is in
Appendix B.

The detailed tables used in the data analysis are available in Appendix C. These show raw counts for arrests, reported crimes, violent deaths, and populations by year, county, and age group where applicable, along with rates per 100,000 population for each.

Top of page

RESULTS

Statistical analysis provides no support for the proposition that stricter curfew enforcement reduces youth crime or risk of violent fatality either absolutely or relative to adults, by location, by city, or by type of crime. Curfew enforcement generally has no discernible effect on youth crime. In those few instances where a significant effect is identified, it is more likely to be positive (that is, greater curfew enforcement is associated with higher rates of juvenile crime) than negative.

Analysis of Curfew Enforcement and Juvenile Arrest Data

Statewide effects over time and by race/ethnicity. As shown in Table 1, of the thirty correlations of statewide rates of youth crime with race/ethnicity for the 1978-97 period, seven are significantly positive, none are significantly negative, and twenty-three indicate no effect.

Table 1

Correlation of Curfew Enforcement with Absolute Youth Crime Rate,
by Race/ethnicity, California, 1978-97

Offense

All
White
Hispanic
Black
Asian

All crime*

.470*
.698*
.525*
.199
.332

Felonies

.017
.384
.144
.039
.243

Homicide

-.121
-.103
.155
-.114
.355

Violent Crime

.078
.375
.276
-.206
.397

Property crime

.001
.241
.093
.030
.131

Misdemeanors

.540*
.583*
.567*
.184
.236

Drug offenses

.278
.430
-.028
.279
.451

Source: California Criminal Justice Statistics Center, California Criminal Justice Profiles, Statewide, 1978-95 (annual) and 1996, 1997 updates.
Note: "All crime" excludes status offenses.
*r > .40 or r < -.40, p < .05, 16 df.

 

Table 2 shows that of the thirty correlations of net youth crime rates compared to adult rates, four are significantly positive, none are negative, and twenty-six are not significant.

Table 2

Correlation of Curfew Enforcement with Net Youth Crime
Rate,* by Race/ethnicity, California, 1978-97

Stronger curfew

enforcement is

associated with:

Less youth crime
More youth crime
No effect on crime

All youth crime

Hispanic
None
White

(excluding status)

Black

Asian

Youth felonies

None
None
All races

Youth homicide

None
None
All races

Youth violent crime

Asian
None
All races

Youth property crime

None
None
All races

Youth misdemeanors

Hispanic
None
White

Youth drug offenses

All races
Black
 Asian 

Source: Authors' calculations from data in Table 1.

*Relative to corresponding adult crime rates. "All youth crime" excludes status offenses.
 

Greater curfew enforcement is associated with significantly higher absolute rates of misdemeanor arrest for whites, Hispanics, Asians, and all youth in aggregate. Curfew enforcement is also associated with higher rates of violent crime by Asian youth and with higher rates of all types of arrest (subtracting curfew arrests) among white and Asian youth. No significant effect is found on rates of juvenile arrest for property crime, violent crime, homicide, all felonies, or all offenses.

Moreover, there are no significant effects of stricter curfew enforcement on either absolute or net rates of juvenile crime rates in the following year (a test of the hypothesis that stricter curfew enforcement has delayed effects). Only two of sixty comparisons show significant results, and both are positive.

Correlations by county over time and by locale. Of the seventy-two correlations for absolute rates of six types of crime in the twelve largest counties for 1980-97, eight are significantly positive, two are negative, and sixty-two are not significant (see Table 3). A similar pattern emerges when seventy-two correlations are examined for net rates of six types of crime in the twelve largest counties for 1980-97. Seven are significantly positive, four are negative, and sixty-one are not significant (Table 4).

Table 3

Correlation of Curfew Enforcement with the Absolute Youth Crime Rate,
California's Twelve Largest Counties, 1980-97

Stronger curfew enforcement is associated** with:

More youth crime
Less youth crime
No effect on crime

All youth crime*

0
1
11

Youth felonies

2
0
10

Youth homicide

2
1
9

Youth violent crime

1
0
11

Youth property crime

0
0
12

Youth misdemeanors

3
0
9
Source: California Criminal Justice Statistics Center, California Criminal Justice Profiles, 1980-95 (annual) and 1996, 1997 updates, by county.

*"All youth crime" excludes status offenses.
**r > .40 or r < -.40, p < .05, 16 df.

 

 Table 4

Correlation of Curfew Enforcement with Net Youth Crime Rate,*
California's Twelve Largest Counties, 1980-97

Stronger curfew enforcement is associated** with:

More youth crime
Less youth crime
No effect on crime

All youth crime*

0
2
10

Youth felonies

1
0
11

Youth homicide

2
1
9

Youth violent crime

1
0
11

Youth property crime

1
0
11

Source: California Criminal Justice Statistics Center, California Criminal Justice Profiles, 1980-95 (annual) and 1996, 1997 updates, by county.

*This is relative to corresponding adult crime rates.
"All youth crime" excludes status offenses.
**r > .40 or r < -.40, p < .05, 16 df.

 

The eleven significant findings in Table 4 are more than would be expected in seventy-two correlations by chance alone, which suggests that curfews might be associated with some effect on crime. However, the significant findings are in opposite directions. This effect may be interpretable. For example, police in some areas may enforce curfews more when serious crime is down, freeing up police time to attend to lesser offenses. Police in other areas may use curfews for visible "crackdowns" when serious crime rates are high. The explanation most consistent with these significant findings, in light of the preponderance of non-significant results, is that curfew enforcement is not the cause of increased or decreased crime, but rather a reflection of varying police responses to crime.

Top of page

County-by-county arrest comparisons by level. If strong curfew enforcement reduces youth crime, net youth crime rates relative to adult crime rates should be lower in high curfew-enforcement counties than in low curfew-enforcement counties. The data do not support this expectation (see Table 5). In 1996-97, higher rates of curfew enforcement are not associated with youth crime levels. Note that counties vary by a factor of more than one hundred in their level of curfew enforcement. In particular, the much publicized curfew crackdowns in Fresno, San Diego, and Santa Clara counties are associated with higher levels of youth crime in every category, both in absolute numbers and relative to adult crime. Conversely, low curfew enforcement counties such as Riverside and Sacramento have lower rates of youth crime relative to adult crime. Other counties show mixed results. Overall, there are no significant effects of curfew.

Table 5

Correlations of Net Youth Crime Levels and
Curfew Enforcement Level, by County, 1996-97

County

Ratio, curfew enforcement rate to 12-county average*
Ratio, net youth felony rate to 12-county average*

Alameda

0.13
1.06

Contra Costa

0.30
0.98

Fresno

1.55
1.07

Los Angeles

2.88
0.99

Orange

1.37
1.15

Riverside

0.54
0.60

Sacramento

0.01
0.68

San Bernardino

1.23
0.81

San Diego

1.65
1.09

San Francisco

0.00
1.09

Santa Clara

1.27
1.39

Ventura

1.07
1.10

Source: California Criminal Justice Statistics Center, California Criminal Justice Profiles, 1980-95 (annual) and 1996, 1997 updates, by county.

*"Net youth felony rate" is the rate of arrests for felonies among those age 10-17 divided by the rate of felony arrest for those age 18-69. Ratios express the curfew arrest rate and net youth felony rate for each county divided by the corresponding average for all twelve counties. Approximations of each county's raw curfew enforcement rate and net youth felony arrest rate can be obtained by multiplying each county's ratio by the average rate.

 

Analysis of status crime arrests and index crime by city. Los Angeles and Orange county cities show wide variation in status law enforcement: Burbank, Huntington Beach, and Los Angeles display very high and rising rates of status and curfew arrest while Pasadena and Anaheim have almost none. Los Angeles and Huntington Beach show rapidly increasing rates in the mid-1990s after low rates of enforcement in the early part of the decade while West Covina and Fullerton display declining rates from high levels in the early 1990s. Status arrest rates for youth and total index crime rates for the twenty-one largest cities (those with 100,000 or more people in 1995) are examined for the 1990-96 period.

Comparing the 1994-96 period with the 1990-93 period, the simple correlation of status crime arrest levels with index crime rates yields a non-significant positive result (r = .237, 19 df). This indicates that status crime arrests are more often accompanied by increasing rather than decreasing index crimes. However, since several "outliers" (cities with large changes in status crime arrest rates) affect this simple correlation, the result may be misleading. In view of this, the cities were divided into two natural categories: those with low and/or declining rates of status crime arrests (thirteen cities), and those with high and rising rates of status crime arrests (eight cities). These two sets differed dramatically in status crime levels.

However, these dramatic differences in status crime levels have no consistent effects on local youth crime. Cities with high and rising status enforcement show a slightly larger (-20.6 percent) average decrease in index crime rates compared to cities with low and/or declining status enforcement (-18.8 percent) but a slightly smaller (-19.5 percent vs. -21.3 percent) median decrease in index crime rates (use of the median eliminates the skewness caused by one or two extreme values). Neither of these differences, of course, approaches significance.

Status arrest and index crime changes for the twenty-one cities are shown in Appendix C. Of interest is the fact that all twenty-one cities show declines in index crime rates no matter how vigorous status crime enforcement. The generality of the crime decline casts doubt on claims that any particular city's law enforcement strategy, whether greater curfew enforcement or some other approach, "caused" the crime decrease.

Analysis of Curfew Enforcement and Crimes Reported to Police

If curfews reduce crime, reductions in reported crime (particularly crimes such as arson and robbery, which are most likely to be committed by youth) should occur in areas with strong curfew enforcement. As noted, FBI's clearance data indicate that around four-fifths of reported Part I crime is committed by adults, making this measure problematic. However, crime reported to police is the measure most often cited by curfew advocates. Table 6 reports the average of correlations for curfew arrest rates and rates of Part I crime reported to police for the 1980-97 period.

We find no relationship between stronger curfew enforcement and crimes reported to police. In fact, the results are uniformly random for major crime categories: in four counties curfew arrest is associated with more index crime, two show less, and fifty-four indicate no effect. Unexpectedly, greater curfew enforcement is significantly and positively associated with arson statwide. As noted earlier, arson is a crime most likely to be committed by juveniles. However, no consistent effect is found across the major counties.

Table 6

Correlation of Curfew Enforcement with Part I Crimes
Reported to Police by County, 1980-97

County correlations

Stronger curfew enforcement is associated** with:

Average of county correlations
More reported crime
Less reported crime
No effect on reported crime

All Part I offenses

-.033
0
0
12

Homicide

-.045
2
1
9

Violent

-.040
0
1
11

Property

-.022
1
0
11

Arson

.140
1
0
11

Source: California Criminal Justice Statistics Center, California Criminal Justice Profiles, 1980-95 (annual) and 1996, 1997 updates, by county.

**p < .05.

 

Top of page

Analysis of Curfew Enforcement and Juvenile Violent Deaths

Finally, the President of the United States has suggested that stronger curfew enforcement keeps youth "out of harm's way," enhancing their safety. While a full analysis of injuries (fatal and nonfatal) by location (public or private) of infliction would be ideal, consistent data are not available on a long-term basis to fully investigate this issue. Violent deaths by age and county, available for the 1980-96 period, are used as an indicator of the most serious injuries.

Violent deaths can vary by county for a number of reasons (i.e., higher rates of driving produce more motor vehicle deaths), which should affect both juvenile and adult death rates. In fact, juvenile and adult violent death rates often are highly correlated with each other, by county, and by time period. Therefore, since curfews apply to youth and not to adults, we would expect their most obvious effects to be on juvenile death rates net of (compared to) those of adults. Table 7 shows the correlations of curfew rates with net youth violent death rates. Using absolute rates produces an almost identical result.

Consistent with prior analyses, we find no effect of stricter curfew enforcement on absolute or net violent death rates among youth. In fact, no county showed a significant reduction in any type of youthful violent death associated with stronger curfew enforcement; three had more juvenile deaths; and fifty-seven showed no difference. Three significant findings in sixty correlations, even in the same direction, do not necessarily indicate a pattern.

One would expect curfew enforcement to reduce juvenile motor vehicle deaths the most, since this is the only type of violent death which always takes place in public; to have the least effect on suicide, which nearly always takes place in private; and to exert intermediate effects on deaths from homicide and from non-traffic accidents,which can take place either in public or private. However, these insignificant results do not support any of these expectations.

Table 7

Correlation of Curfew Enforcement with
Juvenile Violent Deaths, by County, 1980-96

County correlations

Stronger curfew enforcement is associated** with:

Average of county correlations
More youth deaths
Fewer youth deaths
No effect on youth deaths

All violent deaths

-.069
0
0
12

Homicide deaths

-.073
1
0
11

Motor vehicle deaths

.115
0
0
12

Other accidental deaths

-.122
2
0
10

Suicide deaths

-.143
0
0
12

Source: California Bureau of Vital Statistics, Microcomputer Injury Surveillance System (MISS), 1985-96 (annual); National Center for Health Statistics, US Mortality Detail File, 1980-95 (annual) .


**p < .05.

 

Curfew Enforcement Case Studies: Monrovia and San Jose/San Francisco

Our case studies of the effects of curfews in Monrovia and San Jose, California, likewise find no effect attributable to curfews in the relevant crime measures either on an absolute basis (within cities) or compared to neighboring cities that did not enforce curfews. These case studies can be found in full in Appendixes
A and B, respectively. The case studies are summarized below.

Monrovia. In 1994, Monrovia imposed the nation's first truancy curfew, which made it illegal for youth to be in public during daytime school hours. City police subsequently made more than eight hundred arrests under the ordinance. Initial claims of large declines in crime, widely cited in the press and by politicians, were later admitted by police to have resulted from mistakes in data entry (Johnson 1998). We compared the pre-curfew period of the early 1990s to the post-curfew period. The results of our analysis of Monrovia's curfew effects using monthly police reports and California Criminal Justice Profiles of crime in eleven neighboring cities, are as follows:

San Jose/San Francisco. This is a case study of "curfew extremes" in neighboring large cities. San Francisco vigorously enforced its curfew, with more than 1,400 arrests in the 1987-90 period. It then cut back arrests in the early 1990s and nearly abandoned the curfew altogether (two arrests in 1995-97). San Jose went in the opposite direction: compared to just nineteen curfew arrests in the entire 1987-94 period, police stepped up curfew enforcement to arrest more than 1,600 youth in 1995-97. It would be hard to find two cities whose curfew enforcement policies contrast more sharply. Sixty correlations between curfew arrest rates and rates of youth and adult arrest, crime reported to police, and youth and adult violent death were conducted for the years 1985 through mid-1998. The results show that:

Top of page

DISCUSSION

In recent years public and law enforcement officials have frequently touted curfew laws as essential elements in reducing crime and protecting youth. Despite widespread endorsement of curfews, virtually no substantive analysis prior to this study has tested the hypothesis that tougher curfew enforcement reduces juvenile crime. Through an analysis of official data, this research compares the relative crime and death rates of jurisdictions with greater curfew enforcement and jurisdictions with lesser curfew enforcement. In addition, the study examines the effects of curfew enforcement on specific types of crime and the impact of curfew enforcement on juvenile crime rates relative to adults.

The consistent result of this analysis is that curfew enforcement (even the strongest) has no effect on crime, youth crime, or youth safety no matter what the time period, jurisdiction, or type of crime measure studied. This is a surprise. Intuitively, the incapacitation of a large number of youth from the public should affect the likelihood of that group to commit crimes and to be harmed in public places. We had expected this discussion to revolve around whether curfew enforcement was an efficient expenditure of police resources compared to competing demands (i.e., more policing of "hot spots" for crime, as Boston and New York City have emphasized) in terms of the most "bang for the buck." Instead, it appears that juvenile curfews in California have no discernible crime reduction or safety benefits whatsoever. The consistency of results of the different kinds of analysis of curfew laws undertaken in this study points to the ineffectiveness of these measures in reducing crime and improving safety.

California counties display a number of interesting extremes. In 1997, for example, the city of Los Angeles arrested 11,500 youth for curfew violations, 150 times more than in 1990. Supporters cite the forty percent decline in the rate of serious youth crime from 1990 to 1997. Yet adult crime declined at the same rate, in almost identical fashion, for each type of crime. The bottom line is that the Los Angeles rate of youth felonies and other crimes relative to adults was the same in 1997 as it was a decade earlier. Further, the decline in serious crime in the 1990s in the curfew-crackdown cities of Los Angeles, San Diego, and San Jose was no greater, and in most cases was less, than in zero-curfew San Francisco. Cities and counties with zero or near-zero enforcement of curfews and other juvenile status laws are just as safe (or unsafe) from crime, and have no more juvenile violent deaths, than jurisdictions that have the most vigorous enforcement.

Finally, of much greater significance in crime control is the fact that rates of serious crime among youth are strongly correlated with those of adults around them, both by local area and over time. Significant positive correlations (that is, youth and adult crime rates rise and fall together) are found between rates of youth and adult violent, property, felony, and homicide arrests for the twelve largest counties and for the state as a whole. Youth and adult felony rates were correlated for all four major racial groups, as were violent and property crime arrests for all racial groups except whites, homicide rates for whites and Hispanics, and misdemeanor rates for Asians. Where adults display a high rate of violent crime arrests, so do youth.

Law enforcement authorities have stated that they enforce curfew laws evenhandedly. For most major counties, this appears to be true. Arrest rates of white (non-Hispanic) youth are reasonably similar to those of Hispanics, blacks, and Asians. However, four large counties display discrepant racial/ethnic statistics. In Ventura County, curfew arrests of Hispanic and black youth are 8.4 times and 7.4 times higher, respectively, than those of white youth. In Fresno and Santa Clara counties, Hispanic youth are five times more likely to be arrested for curfew violations, and black youth three times more likely, than white non-Hispanic youth. Los Angeles authorities arrest Hispanic and black youth for curfew violations at rates 2.5 times that of white and Asian youth.

This research suggests that the solutions to juvenile crime often championed by law enforcement agencies and public officials must be closely examined rather than evaluated with media-popular anecdotes and selective statistics. Based on the current evidence, a crime reduction strategy founded solely on law enforcement intervention has little effect, suggesting that solutions are more complex and multifaceted (Lundman 1993). Thus, curfews appear to be another "panacea" for crime problems -- a simplistic solution that on closer inspection has no demonstrable effect (Finckenauer 1982). Future policy and research should focus on the potential crime reduction effects of prevention strategies that provide a comprehensive array of services, opportunities, and interventions. While this approach is likely to require a substantial infusion of public resources, the long term benefits appear more promising than panacea approaches such as curfew and status policies.


Appendix A

Appendix B

Appendix C


Endnotes

1. The authors would like to thank Elliott Currie, Pat Jackson, Randy Shelden, Meda Chesney-Lind, Vince Schiraldi, and Tom Gitchoff for their helpful comments. Special thanks to Jodi Schwartz, Jill Herschman, and Christine Gralow for their assistance.

2. The counties examined are Alameda, Contra Costa, Fresno, Los Angeles, Orange, Riverside, Sacramento, San Bernardino, San Diego, San Francisco, Santa Clara, and Ventura.


References

Center for Health Statistics. 1997. Microcomputer Injury Surveillance System 1996. Statewide, by county, and zipcode. Sacramento: Department of Health Services.

Criminal Justice Statistics Center. 1978-1998. California Criminal Justice Profile. Statewide. Counties of Alameda, Contra Costa, Fresno, Los Angeles, Orange, Riverside, Sacramento, San Bernardino, San Diego, San Francisco, Santa Clara, and Ventura (annual). Cities of San Francisco and San Jose (printouts). Sacramento, CA: California Department of Justice.

Demographic Research Unit. 1998. California Statistical Abstract 1997, and California Demographics updates. Sacramento: California Department of Finance.

Division of Law Enforcement. 1978-1996. Crime and Delinquency in California, 1978-1995. Sacramento, CA: California Department of Justice.

Federal Bureau of Investigation. 1997. Uniform Crime Reports for the United States 1996. Washington, DC: US Department of Justice, Tables 28, 38.

Finckenauer, James O. 1982. Scared Straight! and the Panacea Phenomenon. Englewood Cliffs, NJ: Prentice-Hall.

Harvard Law Review Association. 1997. "Curfews May Be Ineffective and Discriminatory." Pp. 193-198 in Juvenile Crime: Opposing Viewpoints, edited by D. Bender and B. Leone. San Diego, CA: Greenhaven Press.

Klein, Stephen P. 1998. February 9 . Declaration. Harrahill v. Santoro.

Krikorian, G. 1996. August 11. "Violent Youth Crime Arrests Decline." Los Angeles Times, A15.

Lait M. 1998. February 10. "Report Questions Teen Curfews." Los Angeles Times, A15.

Lundman, R. L. 1993. Prevention and Control of Juvenile Delinquency. New York: Oxford Press.

Monrovia Police Department. 1992-1997. Monthly report summary, arrests, incidents report to police. Monthly Report. Monrovia, CA. Available from January 1992 through July 1997.

National Center for Health Statistics. 1998. U.S. Mortality Detail File 1985-95. Rockville, MD: Department of Health and Human Services.

Office of Juvenile Justice and Delinquency Prevention (OJJDP). 1996. "Curfew: An Answer to Juvenile Delinquency and Victimization?" Juvenile Justice Bulletin, Summary.

Ricardi, N. 1997. June 5. "Small-Town Success." Los Angeles Times, B14-15.

Ruefle, Reynolds, and Brantley. 1997. "Curfews Can Be Effective and Constitutional." Pp. 187-192 in Juvenile Crime: Opposing Viewpoints, edited by D. Bender and B. Leone. San Diego, CA: Greenhaven Press.

Safe City/Safe Campus Task Force. 1995. "Safe City/Safe Campus Program." City of Monrovia. California (memo).

Santoro J. J. 1998. August. 2 "Curfews Help Make Cities Safer Places." Orange County Register Commentary, 3.

U.S. Bureau of the Census. 1992. Census of Population 1990, California. Washington, DC: U.S. Department of Commerce.

Wilgoren, J. and F. Fiore. 1997. December 2. "Curfews Cited For Drop In Juvenile Crime." Los Angeles Times A1, A26, A27.

Cases Cited

Harrahill et al. v. Santoro.. 1997. Superior Court of the State of California, County of Los Angeles. Case No. BC 170089.

Johnson, R. 1998. January 8. Deposition Testimony. Harrahill v. Santoro. Superior Court of the State of California, County of Los Angeles, Case Number BC 170089, Volume II, 94-105.

Monrovia Police Department. 1998. Crime during school hours. Exhibit 29, Harrahill v. Santoro.
 

Top of page

Return to Current Issue

WCR Home