The Vortex
The Concentrated Racial
Impact of Drug Imprisonment
and the Characteristics of
Punitive Counties
A Justice Policy Institute Report
December 2007

The Justice Policy Institute is a public policy
institute dedicated to ending society’s reliance
on incarceration and promoting effective
solutions to social problems.

	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Contents
2	

I. Introduction	

Incarceration rates for drug offenses have risen dramatically

6	

II. Context	

Who uses drugs? Who is admitted to prison for drug offenses?

10	

Section III.	

Who is most affected by drug admissions at the county level?

13	

Section IV.	What are the spending practices of counties that admit drug
offenders at high rates?

16	Section V.	What are the sociodemographic characteristics of counties that
incarcerate drug offenders at high rates?
    	
21	

	

20    Multiple Variable Analysis

VI. Recommendations	 A call for evidence-based drug enforcement practice

23	Appendix A 	The 198 counties analyzed in this study with overall drug	
admission rate, white drug admission rate, African American 	
drug admission rate, and the ratio of African American to 	
white drug admission rates
27	

Appendix B 	

27	

Appendix C 	Correlations between drug imprisonment rate and social	
structural variables

27	

Appendix D 	

Distribution of social structural variables

U.S. States examined in this study, by U.S. Census Bureau Regions

28	Appendix E 	Ordinary least squares estimates from regression of drug	
imprisonment rates on sociodemographic, budget, index crime 	
rate, and region variables for 198 large-population counties/	
municipalities (2002)
29	Endnotes

I. Introduction: Incarceration rates for drug

offenses have risen dramatically

Over the course of the last 35 years, the rate at
which the U.S. places its citizens in jails and prisons
has risen dramatically. For the first 70 years of the
twentieth century, U.S. incarceration rates remained
relatively stable at a rate of about 100 per 100,000
citizens. Since 1970, the U.S. has experienced a large
and rapid increase in the rate at which people are
housed in federal and state correctional
facilities. Currently, the U.S. incarceraAfrican Americans made
tion rate is 491 per 100,000.1
up 13 percent of the total
The exceptional growth in the prison
U.S. population, but acpopulation has been driven in large
counted for 53 percent of
part by the rate at which individuals
are incarcerated for drug offenses.2 Besentenced drug offenders
tween 1995 and 2003, the number of
in state prisons in 2003.
people in state and federal prisons incarcerated for drug offenses increased

Drug Offenses

Table 1. Though the European Union has 200
million more inhabitants than the United States,
the U.S. incarcerates nearly 10 times as many
people for drug offenses.
U.S. Population (2003)

282,909,885

TOTAL U.S. Prisoners

2,085,620

Federal Prisoners

86,972

State Prisoners

250,900

Jailed Prisoners

170,751*

TOTAL

508,623

European Union Population (2003)

483,297,500

TOTAL EU Prisoners

600,619

Prisoners for Drug Offenses

55,830**

Sources: U.S. Bureau of the Census, American Community Survey; Bureau of Justice Statistics (BJS), Prisoners in 2005, Prisoners in 2003, and
Profile of Jail Inmates, 2002; Council of Europe, SPACE I Survey, 2003
*Estimated using 2002 BJS data.
**Four of the 27 European Union countries did not have data available
by offense type for 2002. Those countries are Austria, Belgium, Czech
Republic, and Poland. Combined, their populations make up approximately 14 percent of the total EU population. This data includes current
EU members.

		

by 21 percent, from 280,182 to 337,872.3 From
1996 to 2002, the number of those in jail for drug
offenses increased by approximately 47 percent, from
111,545 to 164,372.4 This does not include people
imprisoned for other offenses where drugs, the drug
trade, or other drug activities were a feature of the
offense.
The increase in incarceration of drug offenders translates directly to an increase in prison expenditures.
The American Correctional Association estimates
that, in 2005, the average cost of incarcerating one
person for one day was approximately $67.55. The
cost of incarcerating drug offenders in state or federal
prisons amounts to a staggering eight billion dollars
per year.5
There is little evidence to suggest that high rates of
incarceration affect drug use rates or deter drug users.
Researchers have previously found that decreases in
crime in the 1990s were not attributable to an increase in the number of prisons or the increase in the
incarceration rate.6 A Justice Policy Institute (JPI)
study further substantiated these findings by investigating the relationship of incarceration to the rate
of drug use in states. In fact, when observed over a
three-year period, states with high incarceration rates
tended to have higher rates of drug use.7
The growing rate of incarceration for drug offenses is
not borne equally by all members of society. African
Americans are disproportionately incarcerated for
drug offenses in the U.S., though they use and sell
drugs at similar rates to whites.8 As of 2003, twice as
many African Americans as whites were incarcerated
for drug offenses in state prisons in the U.S.9 African
Americans made up 13 percent of the total U.S. population, but accounted for 53 percent of sentenced
drug offenders in state prisons in 2003.10
Over the last several years, JPI has studied drug imprisonment and racial disparities in admission rates
for drug offenses at the state level. While this statelevel information concerning drug offenses and racial

The Vortex: 	The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

	

disparities has been important for consideration in
state policy and laws concerning drug offenses, it has
become apparent that local policies shape the day-today identification of drug users and their entry into
the criminal justice system. It is particularly important to examine the relationship between racial disparities and social policies at the local level.
This report describes the relationship between drug
admission rates and the structural and demographic
characteristics of counties—budgets and spending
for law enforcement, unemployment rates, poverty
rates, and the percentage of the population that is
African American.
In summary, this report finds that:
• While tens of millions of people use illicit drugs,
prison and policing responses to drug behavior
have a concentrated impact on a subset of the population. In 2002, there were 19.5 million illicit drug
users, 1.5 million drug arrests, and 175,000 people
admitted to prison for a drug offense.11 While there
is some variation in reported drug use rates between
different counties and different states, there is much
greater variation between one locality’s propensity to
send people to prison for a drug offense compared
to another’s.
• Whites and African Americans report using and
selling drugs at similar rates, but African Americans go to prison for drug offenses at higher rates
than whites. Survey research shows that whites and
African Americans report illicit drug use and illicit
drug sales at similar rates.12 However, at the local
level, African Americans are admitted to prison for
drug offenses at much higher rates than whites. In
2002, African Americans were admitted to prison for
drug offenses at 10 times the rate of whites in the 198
largest population counties in the country.13 Ninetyseven percent (193 out of 198) of large-population
counties have racial disparities in drug admission rates.
In Appendix A, JPI reports drug admission rates, by
race, for each of the 198 large-population counties
that are the focus of this study.
• Counties that spend more on policing and the
judicial system imprison people for drug offenses
at higher rates than counties that spend less on
law enforcement. Counties that spend a larger proportion of their budgets on policing or the judicial
system imprison more people for drug offenses.
Similarly, those counties that have higher per capita
spending on law enforcement or the judiciary send
more people to prison for drug offenses. These find	

	

ings were statistically significant and stood the test
of multivariate analyses that controlled for other factors, including crime rates, region, poverty, unemployment, the proportion of the population that is
African American, and other spending practices.
•  Counties with higher poverty
rates send people to prison for drug
offenses at higher rates than counNinety-seven percent
ties with lower poverty rates. The
(193 out of 198) of largeoverall drug admission rate for the 10
population counties have
counties with the highest percent of
racial disparities in drug
people living in poverty is six times
admission rates.
higher than for the 10 counties with
the lowest poverty rates. A multivariate analysis controlling for crime rates,
region, unemployment, the proportion of the population that is African American, and spending practices revealed that the correlation between poverty
and drug admission rates is statistically significant.
• Counties with higher unemployment rates imprison people for drug offenses at higher rates
than those counties with lower unemployment
rates. The 10 counties with the highest unemployment rates had drug admission rates that were, on
average, nearly four times that of the counties with
the lowest rates of unemployment. Though these
findings were not statistically significant in the multivariate analysis, they require further discussion and
research in this report.
• Counties with larger proportions of African
Americans in the community sent people to prison
for drug offenses at higher rates. The drug imprisonment rate in the quartile of counties in which African Americans make up the largest percent of the
population has nearly twice the imprisonment rate of
the quartile of counties with the smallest percentage
of African Americans. The positive relationship between drug admissions and the percentage of African
Americans in the community proved to be statistically significant in a multivariate analysis controlling
for crime rates, region, poverty, unemployment, and
spending practices.

Methodology
For this report, we combined data from multiple
sources to calculate county-level rates of admission
to state prisons for drug offenses. We calculated these
rates for the entire population, and separately for the
white and African American subpopulations of each
Justice Policy Institute	



I. Introduction

Incarceration Rates Have Risen

What is a county?
The U.S. Census Bureau states that a county is the primary legal division
of every state except Alaska and Louisiana. A number of geographic entities are not legally designated as counties, but are recognized by the U.S.
Census Bureau as equivalent to counties for data presentation purposes.
These include the boroughs, cities, municipalities, and census areas in
Alaska; parishes in Louisiana; and cities that are independent of any county
in Maryland, Missouri, Nevada, and Virginia. Most of the jurisdictions examined in this report are true “counties.” A few of the jurisdictions examined here are cities or municipalities that have county-like government
structures. These include St. Louis City, MO, and New York City.
Counties traditionally are charged with performing state-mandated duties
such as education, transportation (roads), record keeping, courts, policing,
and jails. Increasingly, counties have been administering programs related
to business development, child welfare, and employment, among others.
The National Association of Counties reports that 35 percent of county
revenue comes from taxes, with the remaining revenue coming from state
and federal sources.

What is a drug admission?
A drug admission is an event in which a person is admitted to a state
prison for a drug offense.
The drug admission rate is the number of drug admissions per 100,000
in the general population in a given year. County-level drug admission
rates reflect the number of state prison admissions that are the result of
sentences that were imposed in each particular county. As such, the drug
admission rate is a measure of action taken by criminal justice institutions
against individuals residing in particular jurisdictions.
In this report, we use the term “drug imprisonment rate” or “rate of admission to prison for drug offenses” as synonyms of the drug admission
rate.
Most research into the social correlates of criminal justice processes uses
the incarceration rate as the outcome variable. The incarceration rate is
the number of individuals who are housed in prison at any given time, for
every 100,000 people in the population. The number of people in prison in
a given year includes individuals who were admitted to prison during that
year or in any previous year.
Given this basic distinction between the prison admission rate and the
incarceration rate, the prison admission rate is a much better measure of
action taken within specific jurisdictions, for a specified time period. The
use of prison admissions data allows us to examine the relationships between jurisdictions’ demographic structure, budgetary decisions, and use
of prison in a temporally-focused manner.

	

county. We then conducted analyses to determine
the characteristics of counties that are associated with
rates of drug admissions. We linked data sources at
the county level using Federal Information Processing Standards (FIPS) State and County Codes.
This report focuses on counties or municipalities
with populations of 250,000 or more. In 2002, the
U.S. population was approximately 288 million.14
Fifty-seven percent of the U.S. population (167
million individuals) lived in the 227 large-population counties or municipalities that had populations
larger than 250,000 in 2002. This report analyzed
data for the 198 large-population counties for which
data was available, representing 147,633,335 million
people, or 51.2 percent of the U.S. population.
The primary source of information for this report is
the most recent data available from the National Corrections Reporting Program (NCRP).15 The NCRP
is the only data source available for examining annual admissions to state prisons by jurisdiction, race,
offense, and other variables. The NCRP provides individual-level data on each admission to state prisons
in a given year. We aggregated this individual-level
data at the county level. In 2002, these data existed
for 38 states. The Department of Justice has gathered
and released this data annually since 1983.
This report focuses on 2002 because it is the year for
which the most recent NCRP data is available. For
the current research, the measure of rates of admission to prison for drug offenses includes only admissions for which a drug behavior was the offense with
the longest associated sentence. The National Corrections Reporting Program data report up to three
offenses associated with each prison admission, and
highlights the offense with the lengthiest sentence.
In a substantial percentage of cases for which a drug
behavior was the offense with the longest associated
sentence, the type of drug offense was listed as “unspecified.” Because of this ambiguous reporting of
data by the NCRP, admission to prison for all types
of drug offenses was combined into one composite
measure per county.
This report also focuses on 2002 because of the availability of data from the U.S. Census Bureau’s 2002
Census of County and Municipal Governments.16
 Other research derived from the 1997 Survey of State and
Federal Inmates indicates that 54.5 percent of inmates had
been convicted for trafficking, 27.1 for possession, and 15.6 for
possession with intent to distribute. An additional 2.8 percent
were convicted for some other offense. King, Ryan S. and
Mauer, Marc (2002), “Distorted Priorities: Drug Offenders in
State Prisons.” Sentencing Project: Washington, DC.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

This rich data source provides detailed county-level
expenditure data in multiple categories. Within the
borders that define the geographic areas of counties
are city and town governments with independent
budgets through which services are provided—above
and beyond the services provided by the county. Detailed budgetary information for these municipalities
was also collected by the U.S. Census Bureau as part
of the 2002 Census of Governments. This combination of county and municipal budget information
within county lines allows for comprehensive countylevel expenditure estimates, regardless of whether the
ultimate source of the expenditure was local, state,
or federal.
This report also uses data from the FBI’s Uniform
Crime Report,17 the U.S. Bureau of the Census,18
and the U.S. Department of Labor’s Bureau of Labor Statistics.19
In order to closely examine relationships between
imprisonment rates and county characteristics, JPI
aggregated individual-level prison admissions at the
county level and linked this data to county-level demographic data from Census Bureau sources. JPI
created four variables representing spending practices: per capita policing expenditures, per capita
judicial expenditures, percent of the county budget
devoted to policing, and percent of the county budget devoted to judicial expenditures. JPI also created
three variables representing key demographic characteristics of counties, including the poverty rate,
unemployment rate, and percent of African Americans in the county. We also examined county-level
crime rates. We conducted a number of bivariate
and multiple variable analyses to examine the characteristics of counties that are associated with drug
admission rates.
In this report, total population refers to people of all
ages, races, and ethnicities. African American refers
to individuals who, regardless of ethnicity, are categorized as either “black alone,” or “black, in combination with one or more other races.” White refers to
individuals who, regardless of ethnicity, are categorized as either “white alone” or “white, in combination with one or more other races.” Because ethnicity
is not taken into account in the data sets utilized in
this report, Latino or Hispanic data is not available.

	

	

Justice Policy Institute	



II. Context: Who uses drugs? Who is

admitted to prison for drug offenses?
The drug admissions vortex:
annual rates of drug use, arrests,
and prison admissions
In 2002, there were 19.5 million illicit drug users—
approximately 8 percent of the population—in the
United States.20 In the same year, there were approximately 1,538,000 drug arrests,21 or about one arrest for
every 13 drug users nationwide. Nearly half of these
arrests (45.3 percent) were for marijuana, and more
than three-quarters (77 percent) were for possession
of a controlled substance.22 According to the most
recent and most complete data available from the National Corrections Reporting Program,23 there were
175,000 admissions to state prisons for drug offenses in
200224—less than 1 percent of all drug users.
Figure 1: The Drug Admissions Vortex: Annual Rates
of Drug Use, Arrests, and Prison Admissions
19.5 million drug users

1.5 million drug arrests

175,000 admissions to state
prisons for drug offenses in
2002, of which 51 percent
were African Americans

Data for this figure come from SAMHSA (2005), ONDCP (2004), the U.S. Bureau of Justice Statistics, National Corrections Reporting Program (2006), and the U.S. Bureau of the Census, 2004.

	

Despite the growing number of drug arrests over the
years, only a small minority of the large population
of drug offenders in the U.S. are arrested or imprisoned. Laws that are violated by a large percentage
of the population—like drug laws—are particularly
prone to selective enforcement25 and are affected by
the resources available to proactively enforce the laws.
Both the impact of these laws on African Americans
and the relationship of enforcement to admissions
will be discussed at length later in this report.
A growing body of evidence reveals that African
Americans and whites use drugs at similar rates. This
evidence is found in recent data from SAMHSA and
the National Institute on Drug Abuse.
• According to the Monitoring the Future (MTF)
survey conducted by the National Institute on Drug
Abuse,26 African American adolescents have slightly
lower rates of illicit drug use than their white counterparts—whether for illicit drug use generally or
for use of a wide variety of specific drugs, including
crack cocaine. However, African American youth are
still being adjudicated more often for drug offenses
than white youth. In 2002, African American youth,
aged 10 to 17, were brought to court with drug-related cases at a rate of 8.2 per 1,000 compared with
6.0 per 1,000 for white youth.27
• According to the 2002 SAMHSA National Survey
on Drug Use and Health (NSDUH), which samples adolescents as well as adults, rates of current
illicit drug use are only slightly higher for African
Americans than for whites. Eight and a half percent of white Americans were current users of illicit
drugs in 2002, compared to 9.7 percent of African
Americans.28
• In 2002, there were approximately 14 million
white Americans who had used drugs in the previous month, compared to about 2.6 million African
Americans who had done so. In other words, there
were five times as many whites using drugs as Afri-

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

African Americans and whites use and sell drugs at similar rates, yet African
Americans are far more likely to be imprisoned for drug offenses.

Figure 2a. Percentage of reported youth drug use and
sales by race in 2002
50%

Figure 2b. Juveniles detained for drug
offenses per 100,000 by race in 2003
30

45%

White

40%

African American

35%

20

17

30%
25%
20%

17%

15%
10%

10%

13%

13%

10
4

5%
0

0
2002 Drug Use

2002 Drug Sales

Juveniles Detained for Drug
Offenses in 2003

Sources: SAHMSA, 2005 Note: This is data for 12- to 17-year olds; Sickmund, Melissa, Stadky, T. I. and Kang Wei. (2005), “Census of Juveniles in Residential
Placement Databook.”

can Americans.29 However, our analyses indicate that
African Americans are admitted to prison for drug
offenses at nearly 10 times the rate of whites.30

percent of defendants sentenced
for crack cocaine offenses were
African American.31

• SAMHSA reported that in 2002, 24 percent of
crack cocaine users were African American and 72
percent were white or Hispanic, yet more than 80

• Similarly, research indicates
that racial patterns of drug sales
tend to correspond to racial patterns of drug use, and that African
Americans are no more likely to
be involved in drug delivery than
whites. In a report released by the
Office of Juvenile Justice and Delinquency Prevention, the results
of the National Longitudinal Survey of Youth showed that 13 percent of African American youth
reported selling drugs, compared with 17 percent of
white youth.32 However, in 2003, African American
youth were arrested for drug abuse violations at nearly
twice the rate of whites.33

Figure 3. In 2002, African Americans were
admitted to state prisons for drug offenses
at almost 10 times the rate of whites.
300
258.17

250
200
150

White
African American

100
50

27.36

0
2002
The 12 states for which there are no data available in the
2002 NCRP include five states from the Mountain West
(Arizona, Idaho, Montana, New Mexico, and Wyoming),
five states from the Northeast (Delaware, Rhode Island,
Connecticut, Vermont, and Massachusetts), and two
states from the Midwest (Kansas, Indiana).
Data for this figure come from the U.S. Bureau of Justice Statistics, National Corrections Reporting Program (2006), and the U.S.
Bureau of the Census, (2005).

	

…the results of the National
Longitudinal Survey of Youth
showed that 13 percent
of African American youth
reported selling drugs….
However, in 2003, African
American youth were
arrested for drug abuse
violations at nearly twice
the rate of whites.

	

Of the drug users who are admitted
to prison, the vast majority are
people of color.
African Americans and other minority groups are
disproportionately represented among those who
are placed in U.S. prisons for drug offenses, despite
government-sponsored research indicating little
racial variation in drug use34 and drug delivery or
distribution.35
Justice Policy Institute	



II. context

Who Uses Drugs?

If rates of imprisonment were purely a function of individual-level behavior, this relative lack of variation
in rates of drug use or drug sales would suggest that
whites and African Americans are being admitted to
prison for drug offenses at similar rates. However,
there is a large disparity between African American
and white rates of imprisonment for drug offenses.
According to a 2000 JPI report, white Americans
were sent to prison for any offense at a rate of 20 per
100,000 in 1996, compared to a
rate of 279 for African Americans.
Whites experienced a 115 percent
In the 1990s drug admission
increase in rates of admission to
rates varied widely at the
prison for drug behaviors between
state level; however, there
1986 and 1996, while African
Americans experienced a 465 perwas little variation in drug
cent increase.36
use across states.
The reasons behind such differences in admission to prison for
drug behaviors may lie, in part, with several specific
factors related to social policy, law enforcement, and
judiciary systems.
Mandatory minimums were implemented in the
1980s and 1990s with the intention of lowering disparities in sentencing by instructing judges how to
sentence defendants based on the crime. A recent Vera
Institute study indicates that mandatory minimums
have led to an increase in incarceration rates for drug
offenses across states.37 Because African Americans
are more likely than whites to be incarcerated for
drug offenses, the likelihood that they will be incarcerated under a mandatory minimum is also higher.
For example, in Maryland, over the last five years,
500 people were sent to prison on a mandatory minimum; nearly 89 percent were African Americans.38
Mandatory minimums also increase the amount of
time spent in prison for a drug offense.39 Nationally, the average time African Americans served in
prison for a drug offense rose 77 percent from 1994
to 2003, compared to a 28 percent increase for white
drug offenders during the same time period.40
Disparate policing practices that focus attention on
certain communities lead to greater arrest rates for
African Americans. For example, police may focus
their efforts on low-income neighborhoods or racial
or ethnic minority neighborhoods. Police are also
more likely to spot an offense occurring on the street,
but not in a suburban home.41
Disparate treatment before the courts often
stems from generalizations and miscommunications between people of different racial or ethnic
	

backgrounds. In a study examining differences in
sentencing recommendations for African American
and white youth, researchers found that probation
officers viewed crimes committed by youth of color
as caused by personal failure, but viewed crimes committed by white youth as having to do with external
forces.42 Such assumptions and miscommunications
may be further exacerbated by the fact that African
Americans are less likely to have access to effective
counsel. Research has shown that white youth are
twice as likely as African American youth to retain
private counsel. Those youth with private counsel are
less likely to be convicted than youth with either a
public defender or appointed counsel.43
Differences in the availability of drug treatment
for African Americans compared with whites make it
more likely that African Americans will continue to
struggle with drug addiction. In a study of Maryland
drug treatment programs, half of whites successfully
completed the programs, compared with only a third
of African Americans.44
Punitive social spending patterns. Since the 1980s,
states with larger African American populations, on
average, spend less on social welfare programs. These
states with relatively large African American populations also tend to spend more on incarceration. This
state-level relationship between the size of the African American population and punitive public spending patterns has been growing substantially over the
course of the last three decades.45

Though people use drugs at similar
rates across states, rates of
imprisonment vary widely.
The lack of variation in drug use patterns and the
wide variation of drug admission rates at the county
level (which will be discussed later in this report) are
mirrored at the state level. The variation in rates of
reported drug use across the 50 states, however, is
significantly smaller than the jurisdictional variation
in rates of imprisonment for drug offenders that is
described in subsequent sections of this report.46 In
2002, the rate of reported use of any illicit drug in
the last 30 days ranged from 6.1 percent in Iowa,
to 12.2 percent in Alaska. An examination of drug
use rates in smaller substate areas uncovered a similar
range of 5 percent in Utah County, UT, to 13 percent in Northern California.47 The ratio of the highest to lowest levels of drug use at the substate level
was therefore 13:5, or 2.6:1.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

The ratio of highest to lowest state drug imprisonment rates, on the other hand, was much higher, at
27:1. In the mid-1990s, the state of Maine had the
lowest rate of admission to prison for drug behaviors—approximately five admissions per 100,000
people. California had the highest rate of drug admissions, at 134 per 100,000.48
In the 1990s drug admission rates varied widely at
the state level; however, there was little variation in
drug use across states. The percent of drug use varied
from a low of 4.8 percent in West Virginia to a high
of 8.2 in Washington. The rate of admission for drug
offenses varied from 10.57 per 100,000 in West Virginia to 145.9 per 100,000 in California.
Table 2. In 1999, drug admission rates varied
widely across states, though there was very
little variation in drug use rates.
State

1999 Admission Rate
per 100,000

Percent of Illicit
Drug Users in the
Last Month (1999)

Ten States with Highest Admission Rates
California

145.90

7.8

Louisiana

141.47

5.4

Georgia

93.11

5.7

New Jersey

89.90

7.2

Illinois

84.10

6.3

Missouri

79.93

6.1

South Carolina

73.29

5.1

North Carolina

71.20

5.8

New York

71.12

6.6

Tennessee

60.02

5.2

Ten States with Lowest Admission Rates
Ohio

52.47

6.0

Virginia

44.71

4.5

Washington

35.79

8.2

Oklahoma

34.75

5.1

Wisconsin

30.82

6.3

Michigan

27.56

7.1

Pennsylvania

26.70

6.3

Oregon

21.62

7.3

Minnesota

18.39

6.1

West Virginia

10.57

4.8

Source: Schiraldi, V. and Ziedenberg, J. (2003), “Cost and Benefits? The
Impact of Drug Imprisonment in New Jersey.” Washington, DC: Justice
Policy Institute.

	

	

Justice Policy Institute	



section III: Who is most affected by drug

admissions at the county level?
This report focuses on admissions to prison
from 198 counties with a population of 250,000 or
more in 2002. These jurisdictions account for more
than half (51 percent) of the total U.S. population.
The 110,522 individuals who were admitted to state
prison for drug offenses from these counties in 2002
are about 60 percent of the 175,000 drug admissions
reported in that year. In these large-population counties, the overall rate of admission to prison for drug offenses in 2002 was 75 per 100,000 in the population.

Despite the fact that white drug users outnumber
black users by a factor of five there were more than
twice as many African Americans (62,087) as whites
(28,314) admitted to prison for drug offenses from
large-population counties in 2002. The rate of admission to prison for drug offenses is more than
10 times larger for African Americans (262.16 per
100,000) than it is for whites (24.85 per 100,000).

Figure 4. Population Size and Number of Admissions to Prison for
Drug Offenses, by Race, Large-population counties in 2002 (n=198)
African American Drug
Admissions = 62,087

White Drug Admissions = 28,314

White Population = 113,954,520

African American Population
= 23,682,790

Data for this figure come from the Bureau of Justice Statistics, NCRP
(2006), and the U.S. Bureau of the Census (2005).

10	

Despite similar patterns of drug use,
African Americans are far more likely
than whites to be admitted for drug
offenses at the county level.
African Americans make up more than half (51 percent) of all admissions to prison for drug offenses.
Despite the fact that white drug users outnumber
black users by a factor of five49 there were more than
twice as many African Americans (62,087) as whites
(28,314) admitted to prison for drug offenses from
large-population counties in 2002.
The rate of admission to prison for drug offenses is more
than 10 times higher for African Americans (262.16 per
100,000) than it is for whites (24.85 per 100,000).

Ninety-seven percent (193 out of 198)
of large-population counties have
racial disparities in drug admission
rates.
Racially disparate rates of admission to prison for
drug offenses are nearly universal among large-population counties in the U.S. Even counties with the
lowest overall rates of admission to prison for drug
offenses display wide racial disparities in those admission rates. Four of the five counties that did not
witness racially disparate drug admissions had very
small percentages of African Americans in their populations (Rockingham, NH = .89 percent; Washington, OR = 1.95 percent; Utah, UT = 0.64 percent;
San Luis Obispo, CA= 2.4 percent), and the fifth
(Clayton, GA = 58 percent) had a sizable majority of
African Americans.
The highest county-level drug admission rate for
whites is 149.5 per 100,000 in Kern, CA. The highest county-level rate for African Americans is nearly
seven times higher—at 1013.9 per 100,000 in the
population in San Francisco, CA.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

III. ???????????????

Who Is Most Affected?

Across counties, there is wide
variation in drug admission rates.
The rate of admission for drug offenses varies substantially across counties. For example, Mecklenburg
County, NC, which contains the city of Charlotte,
has the lowest rate of admission to prison for drug
offenses, at 2.57 per 100,000. Kern County, CA,
which contains Bakersfield, has the highest drug admission rate at 320 per 100,000—more than 124
times higher than Mecklenburg.
At the local level, these rates translate into a striking
number of drug admissions per year—especially for
those jurisdictions with a large population and a high
drug admission rate. The metropolises of Los Angeles County and New York City, for example, each
have drug admission rates of approximately 100 per
100,000. In 2002, 9,768 people from Los Angeles
and 8,161 people from New York City were admitted to a state prison for drug offenses. The practice
of processing and admitting so many individuals to
prison for drug offenses requires large and efficient
policing and judicial machines.

Despite similar rates of drug use
across counties, drug admission
rates vary substantially.
SAMHSA has collected drug use estimates for substate areas in the U.S.,50 making a direct comparison
of drug use and drug admissions possible for only
41 of the counties examined in this study. Most of
the substate areas for which drug use estimates are
available are made up of numerous counties within
states, precluding county-level analysis. A comparison of these 41 counties indicated that the percent of
people who use drugs in various jurisdictions varies
very little, compared to the variation in rates of admission for drug offenses.
The range of illicit drug use in the counties examined
ranged from 12.5 percent of people aged 12 or older
in Multnomah County, OR, to about 5 percent of
people in Utah County, UT. This range appears to
be a good representation of the range for all substate
areas for which drug use estimates are available. Few
substate areas have rates of drug use that are higher
than Multnomah County’s 12.55 percent, and the
highest rate is 13.5 percent for any substate area.51
The sharp contrast in the range of percent of drug users
and the rate of drug admissions can be seen in the difference in the ratio between lowest to highest. Among
	

	

Table 3. Counties with the Highest and Lowest
Rates of Admission to Prison for Drug Offenses
in 2002
Counties with the Highest Drug Admission Rates
Kern, CA

319.86

Atlantic, NJ

256.34

Orleans, LA

249.54

St. Louis City, MO

239.10

Camden, NJ

217.21

Cuyahoga, OH

209.42

Jefferson, LA

185.96

San Bernardino, CA

170.15

Cook, IL

166.25

Alameda, CA

154.93

Counties with the Lowest Drug Admission Rates
Washington, OR

8.03

Cumberland, ME

7.43

Fairfax, VA

6.92

Wake, NC

6.07

Rockingham, NH

5.21

Bucks, PA

3.93

Howard, MD

3.84

Montgomery, MD

3.74

Guilford, NC

3.48

Mecklenburg, NC

2.57

Data for this table come from the U.S. Bureau of Justice Statistics,
National Corrections Reporting Program (2006), and the U.S. Bureau
of the Census (2004).

the 41 large-population counties for which drug use
rates are available, the ratio of the highest to the lowest county-level drug use rate is 12.55 percent to 5.03
percent for a calculated ratio of 2.5. In contrast to this
relatively small range between the counties with the
highest and lowest rates of drug use, the ratio of the
highest to the lowest county-level drug imprisonment
rate is 319.86 to 2.57 for a calculated ratio of 124.5.
Clearly, rates of drug use are not driving the drug imprisonment rate at the county level. Despite similar
percentages of the population using drugs, a number
of counties experience very different admission rates.
• Though Rockingham County, NH, has a larger
percent of illicit drug users, Jefferson Parish, LA,
Justice Policy Institute	

11

Section III.

Who Is Most Affected at County Level?

has a drug admission rate that is 36 times that of
Rockingham.
• About 7 percent of the populations of both Riverside County, CA, and Palm Beach County, FL, reported illicit drug use in the last month. Despite the
similarity in drug use, Riverside County admits people to prison for drug offenses at about eight times
the rate as Palm Beach County.

• Despite similar rates of drug use in Cook County,
IL, and Macomb County, MI, Cook County has a
drug admission rate that is more than seven times
greater than Macomb.

Table 4. There is no relationship between the rates at which people are sent to prison and the
rates at which people use drugs in counties.
Substate/County

State

Drug Admission
Rate 2002

Percent Using Illicit Drugs in
Past Month (2002–2004)

Ten Counties with the Highest Drug Admission Rates and Their Drug Use Rates
Philadelphia

PA

116.67

10.80

Davidson

TN

119.31

8.98

Milwaukee

WI

123.14

9.49

Hamilton

OH

124.74

8.86

Oklahoma

OK

125.21

9.26

Tulsa

OK

128.47

9.99

Jackson

MO

130.00

10.62

Riverside

CA

148.14

7.61

Cook

IL

166.25

8.77

Jefferson

LA

185.96

8.47

Ten Counties with the Lowest Drug Admission Rates and Their Drug Use Rates
Macomb

MI

22.51

8.59

Hidalgo

TX

20.99

5.04

Palm Beach

FL

18.82

7.92

Utah

UT

17.53

5.03

Washtenaw

MI

17.05

9.00

Prince Georges

MD

11.16

7.08

Anne Arundel

MD

10.33

7.19

Cumberland

ME

7.43

10.55

Rockingham

NH

5.21

10.40

Montgomery

MD

3.74

6.42

Data for this table come from SAMHSA (2006), NCRP (2006), and U.S. Bureau of the Census (2004).
* This list is derived from the 41 substate areas in SAMHSA’s drug use report (2006) that correspond to the 198 counties included in this report.

12	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Section IV. What are the spending practices

of counties that admit drug offenders at
high rates?

Levels of funding permit
resource-driven discretion.
Individual criminal behaviors recorded by local police forces are the basic building blocks of the Uniform Crime Reports (UCR), which are the crime
rates that are heralded in the press and analyzed on
an annual basis. Police forces have been significantly
enhancing their resources and capacity to detect and
record crimes since the 1970s. This enhanced capacity was at least partially responsible for violent crime
rates that trended steadily upward in the 1970s and
1980s—despite the fact that victimization rates during the same time period held steady.52
Police forces have varying amounts of discretion as to
whether they will record an incident as an official crime,
depending on the incident. Violent crimes are the least
ambiguous, and the presence of victims necessitates action and reporting by police.53 For violent crimes, the
person engaging in the illegal behavior is the primary
actor in the production of the crime—the agents of the
criminal justice system are essentially reactive.
On the other end of the spectrum, local police forces
have a great deal of discretion when it comes to policing and recording drug offenses, and a jurisdiction’s
prosecutors and judicial branch have discretion in
the application of charges and sentences for those
arrested. Drug offenses are quite common and relatively constant. They rarely have clearly defined victims that necessitate formal reporting procedures by
	

	

the police. Rates of victimless crimes such as drug
offenses tend to have strong correlations with the
number of personnel assigned to police those specific
behaviors.54 In order to observe, arrest, and
make an official crime statistic of a person
engaging in a drug offense, police must be
Policing resources
in proactive pursuit.
determine the size of
Policing resources determine the size of the
the mouth of the drug
mouth of the drug enforcement vortex.
enforcement vortex.
The size of the policing budget—whether
measured in absolute or relative terms—
determines the extent to which a jurisdiction can engage in proactive pursuit of people engaging in drug
behaviors. The size of the judicial budget, in turn,
determines the number of those caught in the vortex
that can be dispatched to prison. The figures that follow illustrate the correlation between spending and
drug admission rates.
•  As shown in Figure 5, the drug imprisonment rate
Counties with greater per capita policing expenditures
have higher rates of drug imprisonment

Figure 5. Rate of Drug Imprisonment by Per Capita Policing Budget
198 Counties with Population Greater than 250,000 (2002)
Imprisonment Rate per 100,000

The vast differences in drug admission rates
across counties—despite similarities in the percent
of drug users—suggest that there are differences in
the way counties construct policies concerning drug
offenses. This section of the report will examine the
extent to which highly variable resources of police
forces and judicial departments across the U.S. have
an impact on their local drug imprisonment rates.

120
100.7

100
80

69.4

60
40

53
34.3

20
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Highest
Quartile

Per capita policing budget (range $65–$510)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

Justice Policy Institute	

13

Section IV.

What Are the Spending Practices?

in the quartile of counties with the highest per capita
policing expenditures (100 per 100,000) is about
three times greater than the drug imprisonment rate
in the quartile of counties with the lowest per capita
policing expenditures (34 per 100,000).

in the quartile of counties with the highest per capita
judicial55 expenditures (108 per 100,000) is more
than three times as large as the drug imprisonment
rate in the quartile of counties with the lowest per
capita policing expenditures (34 per 100,000).

•  Similarly, in Figure 6 the drug imprisonment rate

•  As illustrated in Figures 7 and 8, rates of admission to prison for drug offenses also increase with the
percentage of the budget that jurisdictions devote to
policing and judicial activities.

Counties with greater per capita judicial expenditures
have higher rates of drug imprisonment

107.7

To further substantiate these results, JPI conducted
a multiple variable analysis that controlled for the
crime rate, region of the country, the poverty and
unemployment rates, and the percent of each county’s population that is African American. The results
strongly suggest that the resource-driven discretion
that local police forces have is the engine driving the
wide variation in local drug imprisonment rates. This
relationship is evident in this study’s finding that policing budgets are positively associated with the drug
imprisonment rate—even after controlling for the
crime rate.

Highest
Quartile

“Laws that are widely violated...
especially lend themselves to selective and arbitrary enforcement.”

Figure 6. Rate of Drug Imprisonment by Per Capita Judicial Budget
198 Counties with Population Greater than 250,000 (2002)

Imprisonment Rate per 100,000

120
100
80

70.1

60
40

46.1
34.3

20
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Per capita judicial budget (range $3.25–$238.56)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

— Charles Reich, Yale University Law School

Counties that devote a greater percentage of their budgets
to policing have higher rates of drug imprisonment

Figure 7. Rate of Drug Imprisonment by Percent of County
Budget Devoted to Policing
198 Counties with Population Greater than 250,000 (2002)
94.8

Imprisonment Rate per 100,000

100
90

81.3

80
70
60

63.1
52.6

50
40
30
20
10
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Highest
Quartile

Percent of county budget devoted to policing (Range: 2.24%–11.5%)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

14	

Resources available to police and the
judiciary also encourage selective
enforcement of drug laws.
It has been noted that “Laws that are widely violated ... especially lend themselves to selective and
arbitrary enforcement.”56 While there were approximately 19.5 million drug users in the U.S. in 2002,
police forces across the country carried out about 1.5
million arrests for drug offenses that year. Because of
the large number of drug users at the national and local levels, police forces can selectively target distinct
subpopulations for scrutiny and arrest.
A recent in-depth analysis of drug enforcement patterns in Seattle57 indicates that African Americans
are disproportionately arrested for drug delivery offenses, and that these disproportions are not due to
any extraordinary characteristics of those African
American arrestees, the behaviors they engaged in,
or the communities in which they were arrested. In

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Counties that devote a greater percentage of their budgets
to the judiciary have higher rates of drug imprisonment

Figure 8. Rate of Drug Imprisonment by Percent of County
Budget Devoted to Judicial Expenses
198 Counties with Population Greater than 250,000 (2002)
98.2

Imprisonment Rate per 100,000

100
90

78.8

80
70
60

—Katherine Beckett, Professor at the University of Washington

54

52.8

…these patterns appear to reflect a
racialized conception of who and what
comprises the drug problem in Seattle….
Remedying racial disparity in drug law
enforcement will require a thorough
re-thinking and reorientation of Seattle
Police Department drug law enforcement
practices.

50
40
30
20
10
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Highest
Quartile

Percent of county budget devoted to judicial expenses (Range: 0.07%–4.17%)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

other words, although African Americans in Seattle
were not selling drugs at a higher rate than whites,
they were targeted more frequently for drug arrests.
Given the racial disparities in drug enforcement
practice highlighted in this in-depth Seattle study, it
is not surprising that the drug imprisonment rate in
King County, WA, was 23 times higher for African
Americans (465 per 100,000) than it was for whites
(20 per 100,000) in 2002.
Wide racial disparities in drug imprisonment exist in
virtually every large-population county in the U.S.
and policing budgets have been found to be closely
associated with the drug imprisonment rate. Given
these facts, the conclusion of the Seattle drug analysis
that drug law enforcement practices must be reevaluated can be applied to large-population counties
throughout the country.

	

	

Justice Policy Institute	

15

Section V. What are the sociodemographic

characteristics of counties that incarcerate
drug offenders at high rates?
The wide racial disparity in drug imprisonment
rates does not reflect the fact that whites and African
Americans use illegal drugs at roughly the same rate59
and that whites and African Americans are engaged in
selling a wide variety of drugs at similar rates.60 Similarly, the significant intercounty variation in the drug
admission rate does not reflect the reality of minimal
variation in rates of drug use across state and local
jurisdictions.61 Since relatively constant patterns of
individual-level drug use do not appear to be driving the widely varying racial and cross-jurisdiction
drug admission rates, it is necessary to examine the
sociodemographic characteristics of places that may
be associated with these disparities in prison admission rates.

Sociodemographic structure
As described earlier in this report, a majority of
Americans live in jurisdictions with populations of
250,000 or more. These large-population counties
are not monolithic in their sociodemographic structure. There is significant variation in
poverty rates, unemployment rates,
and the racial/ethnic composition of
Results of our analysis
these large population centers. Previous
indicate that the drug
criminal justice research has explored
imprisonment rate is
the relationship between criminal justice outcomes and the poverty rate,62
related to the strength of
unemployment rate,63 and the percentlocal labor markets and
age of the population that is African
the ways that our comAmerican.64
munities are racially and
The current research generally indicates
economically stratified.
that wider bands of disadvantaged social strata within counties are associated
with punitive practices with regard to
policing, prosecuting, and ultimately imprisoning
individuals who have engaged in drug behaviors.
The drug imprisonment rate can be seen as a measure of the “ease with which a society imposes pun16	

ishment.”65 Social and behavioral research generally
indicates that punishment is easier to dispense upon
individuals with whom one feels little commonality.66 For those who create and implement criminal
justice policies at the federal, state, or local levels,
this fact often translates into prison populations that
vastly underrepresent people from privileged social
positions, and overrepresent the segments of society
that superficially appear to have little in common
with them.
This report is the first to examine the relationships
between these sociodemographic structures and the
specific annual rate at which people are admitted to
prison for drug offenses. Results of our analysis indicate that the drug imprisonment rate is related to the
strength of local labor markets and the ways that our
communities are racially and economically stratified.
On average, counties with higher unemployment
rates, higher poverty rates, and larger percentages of
African American citizens tend to have higher rates
of admission to prison for drug offenses.

Poverty rate
Prisons in the U.S. are disproportionately populated
by individuals who were living in poverty prior to
their imprisonment. Half of those in prison in the
early 1990s had annual incomes of less than $10,000
in the year prior to the arrest that led to their imprisonment. One-fifth had pre-arrest annual incomes
that were less than $3,000.67
As shown in Figure 10, the drug imprisonment rate
in the quartile of counties with the highest poverty
rates (106 per 100,000) is approximately four times
larger than the drug imprisonment rate in the quartile of counties with the lowest poverty rates (27 per
100,000). These findings were further substantiated
in a multiple variable analysis that also revealed a significant positive relationship.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

What Are the Sociodemographic Characteristics?

A close examination of the 10 counties that have the
highest percent of poverty and the 10 counties that
have the lowest of the 198 in the study reveals a striking difference in the rates of drug admission. The
overall drug admission rate for the 10 counties with
the highest percent of poverty is six times higher than
that for the 10 counties with the lowest percent of
poverty. Whites are admitted for drug offenses in high
poverty communities at three times the rate of whites
in low poverty counties. African Americans in high
poverty counties are admitted at nearly twice the rate
of African Americans in low poverty communities.
This study is not the first to find that jurisdictions with
higher poverty rates tend to be more punitive. Nagel’s
examination of state-level variation in incarceration
rates,68 and Colvin’s study69 of county-level variation,
both found that the rate of poverty in a given jurisdiction is related positively to the rate of incarceration.
Beckett and Western also found that state poverty rates
were positively associated with their respective incarceration rates, suggesting that “poor populations are
subject to greater surveillance.”70 Our results, in addition to these studies, indicate that the rate of poverty
in a state or county is significantly indicative of that
jurisdiction’s willingness to incarcerate its citizens.

Counties with higher percentages of people living in poverty
have higher rates of drug admissions

Figure 9. Rate of Drug Imprisonment by Poverty Rate Quartiles
198 Counties with Population Greater than 250,000 (2002)
120
Imprisonment Rate per 100,000

V. ????????????

76

80
58.4

60
40
27.2

20
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Highest
Quartile

(3.5–7.6450)

(7.6451–10.1000)

(10.1001–12.7000)

(12.7001–33.00)

Poverty Rate (Range: 3.5%–33%)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

Figure 10. The ten counties with the highest levels of poverty
have a higher average drug admission rate than the ten counties
with the lowest levels of poverty.*

Unemployment rate

300
Imprisonment Rate per 100,000

The unemployed are disproportionately represented
among individuals admitted to prison. Approximately one-third of inmates in state prisons were
unemployed immediately prior to the arrest that
led to their incarceration.71 This does not necessarily
mean that the unemployed are more likely to commit
crimes that lead to imprisonment. The unemployed,
who by definition are actively seeking work, are in
competition for local jobs and have the potential to
diminish the relative prestige and power of those in
more privileged positions in the labor market. The
unemployment rate is an indicator of economic distress and slack labor markets in local communities.72
Higher unemployment rates may be associated with
higher levels of anxiety and perceived economic insecurity,73 which may in turn be translated into support
for punitive criminal justice processes74 and therefore
higher drug imprisonment rates.

105.7

100

250
200

All

256.58

White
African American
149.94

150

117.82

100
50

37.56
18.47

11.9

0
Lowest Poverty

Highest Poverty

Data for this figure come from the U.S. Bureau of the Census (2002, 2004, 2005) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

*These comparisons are for illustrative purposes and have not been examined for statistical significance.
The 10 counties with the highest levels of poverty are New York City, Caddo Parish (LA),
Philadelphia, Fresno, St. Louis, Tulare (CA), El Paso (TX), Orleans Parish (LA), Cameron
(TX), and Hidalgo (TX). The 10 counties with the lowest levels of poverty are Waukesha
(WI), Morris (NJ), Howard (MD), Somerset (NJ), Dakota (MN), St. Charles (MO), Rockingham (NH), Anoka (MN), Collin (TX), and Chester (PA).

A closer examination of the relationship between
unemployment and drug admission rates for the 10
counties that have the highest rate of unemployment
and the 10 that have the lowest reveals a similar pattern. Counties with the highest rates of unemploy	

	

Justice Policy Institute	

17

section V.

What Are the Sociodemographic Characteristics?

Counties with larger unemployment rates have higher
rates of drug imprisonment

Figure 11. Rate of Drug Imprisonment by Unemployment Rate
198 Counties with Population Greater than 250,000 (2002)

Imprisonment Rate per 100,000

120
102.6

100
80

64.1

72.3

40

33.7

Racial distributions

20
0
Lowest
Quartile

2nd
Quartile

3rd
Quartile

Highest
Quartile

Figure 12. The ten counties that have the highest rates of unemployment, on average, also have higher drug admission rates
than the ten counties with the lowest rates of unemployment*
450
Imprisonment Rate per 100,000

The relationship between the unemployment rate
and drug imprisonment is of borderline significance
in our multiple variable analysis. This suggests that
other county-level sociodemographic and enforcement characteristics are potentially related to unemployment rates, and are more closely related to the
drug imprisonment rate.

60

Percent of population that is unemployed (Range: 3.2%–12%)
Data for this figure come from the U.S. Bureau of the Census (2004) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

400
350
300
250

427.15
All
White
African American

200

167.67

150
105.64

100
50

29.56

18.27

45.3

0
Lowest Unemployment

Highest Unemployment

Data for this figure come from the U.S. Bureau of the Census (2004, 2005), U.S. Bureau of
Justice Statistics, NCRP (2006), and the U.S. Department of Labor (2007).

*These comparisons are for illustrative purposes and have not been examined for statistical significance.
The 10 counties with the highest unemployment rates are Santa Clara (CA), San Joaquin
(CA), Cameron (TX), Monterey (CA), Clark (WA), Stanislaus (CA), Kern (CA), Hidalgo (TX),
Fresno, and Tulare (CA). The 10 counties with the lowest unemployment rates are Lancaster (NE), Cumberland (ME), Chesterfield (VA), Prince William (VA), Dane (WI), Fairfax
(VA), Howard (MD), Montgomery (MD), Polk (IA), and Knox (TN).

18	

ment have an overall drug admission rate that is
nearly four times the rate of counties with the lowest
rates of unemployment. A similar pattern holds true
for both whites and African Americans.

The high rate of imprisonment of African Americans
for drug offenses clearly has an important adverse
impact on communities. A growing body of public
health research describes a wide range of objectively
identified social ills that are the result of extremely
high rates of imprisonment in local communities
of color,75 including reduced employment rates,76
reduced family income and stability,77 high rates of
homelessness,78 reduced number of citizens who are
eligible to vote,79 increased foster care placements and
the associated risk of psychological and educational
problems for children,80 and reduced health and well
being among women in the community.81
Our results indicate that the proportional size of the
African American population in a community has a
clear relationship with the rate of drug admissions.
The drug imprisonment rate in the quartile of counties in which African Americans make up the largest percentage of the population (93.6 per 100,000)
has nearly twice the drug imprisonment rate as the
quartile of counties with the smallest percentage of
African Americans (49.6 per 100,000).
A closer examination of the relationship between
racial distributions and drug admission rates for
the 10 counties that have the highest percentage of
African Americans and the 10 that have the lowest
reveals a similar pattern. The overall drug admission rate is six times higher for counties with high
percentages of African Americans, relative to those
with the lowest percentage. In counties with higher
percentages of African Americans, whites are admitted for drug offenses at four times the rate of counties with lower percentages of African Americans.
African Americans are admitted at more than twice
the rate in counties with higher percentages of African Americans.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

A growing body of public
health research describes a
wide range of objectively
identified social ills that are
the result of extremely high
rates of imprisonment in local
communities of color…

Counties with larger percentages of African Americans
have higher rates of drug imprisonment.

Figure 13. Rate of Drug Imprisonment by Percent of Population
that is African American
198 Counties with Population Greater than 250,000 (2002)
100
Imprisonment Rate per 100,000

The results of this study indicate that the severely disproportionate impact of drug imprisonment on African Americans is related directly to the level of African
American representation in local counties, and therefore to the persistent ways in which our local communities are stratified by race. This relationship between
African American representation and drug imprisonment persists even after controlling for region, crime,
and important economic and labor market indicators
in our multiple variable analysis. Other research82 has
indicated that the relationship between the size of the
African American population and punitive criminal
justice outcomes is continuing to grow over time.

93.6

90
80

72.1

72.4

2nd
Quartile

3rd
Quartile

70
60
50

49.6

40
30
20
10
0
Lowest
Quartile

Highest
Quartile

Percent of population that is African American (Range: 0.64%–68.0%)
Data for this figure come from the U.S. Bureau of the Census (2005) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

Figure 14. The overall drug admission rate for the ten counties
with the highest average percentage of African Americans is
almost ten times that of the ten counties with the lowest
average percentage.*
175.38

Imprisonment Rate per 100,000

180
160
140
120
100

All
White
112.42

African American

80

69.02

60

44.03

40
20

19.22

11.64

0
Lowest Percentage

Highest Percentage

Data for this figure come from the U.S. Bureau of the Census (2004, 2005) and the
U.S. Bureau of Justice Statistics, NCRP (2006).

* These comparisons are for illustrative purposes and have not been examined for
statistical significance.
The 10 counties with the highest percentages of African Americans are Fulton (GA),
Philadelphia, Caddo Parish (LA), Richland (SC), Shelby (TN), St. Louis, Dekalb (GA),
Clayton (GA), Prince Georges (MD), and Orleans Parish (LA). The 10 counties with the
lowest percentages of African Americans are Utah County (UT), Hidalgo (TX), Cameron
(TX), Rockingham (NH), Larimer (CO), Waukesha (WI), Clackamas (OR), Boulder (CO),
Jefferson (CO), and Lane (OR).

	

	

Justice Policy Institute	

19

What Are the Sociodemographic Characteristics?

section V.

Multiple Variable Analysis
The findings reported in the previous sections describe relationships between single, county-level variables and the rate
of admission to prison for drug offenses. In order to be more confident in these findings, we must control for key factors
that may be mediating or interfering with the relationships between county-level social and budget structures and the
drug imprisonment rate. In other words, it is possible to isolate each variable of interest to examine its relationship with
drug admissions without concern that other variables in the model are affecting the relationship. This analysis controls
for the index crime rate and the geographic region of the counties, as well as all three sociodemographic variables and
the spending variables discussed in this report. It is necessary to control for the factors for the following reasons:
Crime rate: Counties with higher rates of criminal behavior are likely to have higher imprisonment rates and, thus, are also
likely to have greater policing and judicial resources. In this way, the county-level crime rate may be affecting the relationship between the drug imprisonment rate and policing and judicial resources.83
Geographic region: Geographic regions experience vastly different drug admission rates, as well as poverty rates,
unemployment rates, and relative size of the African American population. Failing to control for regional effects may lead
to incorrect interpretation of relationships between these variables and criminal justice outcomes84 such as the drug
imprisonment rate.85
Sociodemographic characteristics: Because poverty, unemployment, and the percentage of African Americans in a
community may be independently related, it is necessary to control for each of those variables.
Spending practices: Counties that spend more on policing are more likely to spend more on the judiciary as well. By including each of these aspects of spending in our multivariate model, it is possible to isolate the impact of each on the drug
imprisonment rate.
For this analysis, JPI created two models. The first model includes the three sociodemographic structure indicators
(unemployment, poverty, and the percentage of African Americans in the county), per capita policing and judicial expenditures, the index crime rate, and a variable for region. The second model differs only in that the per capita budget items
are replaced with indicators of the percentage of the budget devoted to policing and judicial expenses. By establishing
these models, JPI was able to determine the relationships between each independent variable and the drug admission
rate, independent of the effects of the other variables in the model. The first model accounts for about 44 percent of the
county-level variation in the drug imprisonment rate, while the
second model accounts for about 42 percent of the variation.
Model 1

Model 2

Independent Variables
Unemployment Rate

Unemployment Rate

Poverty Rate

Poverty Rate

Percent African American

Percent African American

Per Capita Policing Budget

--

Per Capita Judicial Budget

--

--

Percent of Budget Policing

--

Percent of Budget Judicial

Controlled Variables
Index Crime Rate

Index Crime Rate

Region

Region

Dependent Variable
Drug Admission Rate

20	

Drug Admission Rate

In both models, the poverty rate and the percentage of African Americans in the population persist as significant, positive
predictors of the drug imprisonment rate, even after controlling for crime rate, region, and other variables in the model.
According to our models, however, the unemployment rate
does not display an independent relationship with the drug
imprisonment rate.
Whether viewed in absolute (per capita expenditures) or
relative (percent of budget) terms, policing and judicial expenditures are statistically significant, positive predictors of
admissions to prison for drug offenses, even after controlling for the potentially confounding effects of the crime rate,
region variables, and the sociodemographic variables in the
model. (Refer to Appendix E for regression coefficients and
standard errors for both models.)

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

VI. Recommendations: A call for evidence-based

drug enforcement practice

The relationships between social structure and
drug imprisonment, as well as disparities in drug
imprisonment rates between African Americans
and whites that are nearly universal across counties, strongly suggest a need for a more evidencebased approach to drug enforcement practices in the
United States.
Evidence-based drug enforcement practices means
that rates of drug arrests and imprisonment would
have a direct empirical link to rates of specific drug
behaviors within county borders, and within groups
defined by race/ethnicity, labor market status, or
other relevant stratification factors. To facilitate this
rational and fair approach to drug enforcement, federal and state legislatures must provide funding for
county-level probability surveys and ethnographic
research to reliably determine the rates of specific
types of drug behaviors across categories of race/
ethnicity. State and local officials must then be held
accountable to ensure that rates of drug arrests and
imprisonment be no more disparate across counties
or social stratification categories than rates of relevant
drug behaviors.
With detailed data on drug use rates and patterns
within counties, local police and prosecutors could
be held accountable for disproportional disparities in
drug imprisonment rates between African Americans
and whites, or across other sociodemographic categories. In order to reduce racial disproportions in the
drug imprisonment rate, local police and prosecutors
would need to reduce their emphasis on policing and
prosecuting drug behaviors in largely African American neighborhoods.
If federal, state, or local governments will not fund
the administration of local drug use surveys, research and advocacy organizations should fund this
evidence-generating research and form community
oversight boards to compare the results to publicly
available information on drug imprisonment rates
by race and other relevant social categorizations.

	

	

This data would help hold jurisdictions accountable
for disparities in arrests and imprisonment outcomes
that occur within their borders.

Evidence-based drug enforcement:
Toward de-escalation of the drug war
The drug war is primarily being waged against African American citizens of our local jurisdictions,
despite solid evidence that they are no more likely
than their white counterparts to be engaged in drug
use86 or drug delivery behaviors.87 If evidence-based
drug enforcement were to take place so that whites
were punished for their drug
behaviors at similar rates as
African Americans, the push
“The white majority can ‘afford’
for de-escalation of drug enthe costs associated with
forcement through legislative
mass incarceration because
means would be led by whites
the incarcerated mass is
and others in more privileged
positions who presently supdisproportionately non-white.”
port or are ambivalent about
—David Cole, Georgetown University
current drug control policy.
Law School
As David Cole noted, “The
white majority can ‘afford’ the
costs associated with mass incarceration because the
incarcerated mass is disproportionately nonwhite.”88
If drug laws were enforced among whites as they are
among African Americans, those who are currently
privileged by the status quo would no longer be able
to “afford” punitive drug laws and drug enforcement
practices. If these laws and practices were to become
“unaffordable” to privileged subpopulations through
equitable hyper-enforcement, they would quickly
become a thing of the past.
Alternatively, the more appropriate reduction of
the drug enforcement effort against African Americans—so that it is proportional to their rates of
drug use and delivery behaviors in the community—would bring African American drug imprisJustice Policy Institute	

21

Vi. Recommendations

Evidence-Based Drug Enforcement

onment rates into line with those of whites. This
equalization would, in effect, be a rational de facto
de-escalation of the drug war that is currently being
waged in a way that disproportionately and unfairly
targets African Americans.

Careful consideration of public
safety funding
The social construction of the drug imprisonment
rate in a county begins with the purposeful perception of drug offenses by the police and subsequent
arrests. Additional steps toward the drug imprisonment rate include decisions made by local prosecutors about the type of crimes that individuals should
be charged with once they have been arrested. These
innumerable social acts and decisions are carried out
within the localized social contexts of racial and ethnic distributions, poverty rates, and resources available for policing and judicial functions.
Research presented in this report clearly indicates
that county-level public expenditures on policing
and the judicial branch are strongly related to the
rates at which local individuals are admitted to state
prison for drug offenses. This report also indicates
that counties with higher poverty rates, higher unemployment rates, and larger percentages of African
Americans are generally the most likely to send their
fellow citizens to state prison for drug offenses. Local
decision makers must carefully consider whether punitive and exclusionary patterns of public spending
on drug enforcement is in the best interest of their
diverse constituencies and communities.

10 times higher than Lancaster County’s rate (20 per
100,000).
The sociodemographic contexts in which these local
decisions are made are important as well. Rates of
poverty and unemployment are considerably higher
in Orleans Parish than they are in Lancaster County.
In Orleans and other counties with high drug imprisonment rates, public policing dollars may be
better spent on a wide variety of social, health, and
educational programs that reduce poverty and create
jobs for local citizens. Since the 1980s, there has been
a growing trend at the federal, state, and local levels
to reduce expenditures on social welfare programs in
favor of increased spending on penal institutions89
and law enforcement. Reversing this trend is critical
to reducing the drug imprisonment rate, and its severely disproportionate impact on African American
communities throughout the United States.
While this report carefully examines the role of law
enforcement funding, it should not be read as an
indictment of the behavior of individual police officers. Individual police officers are performing difficult jobs, and most individuals who are imprisoned
for drug behaviors reach that outcome because they
broke established drug laws. However, millions of
additional individuals break those same laws every
day, far beyond the scrutiny and interest of police
forces across the country. The results of the current
research cast a bright light on the drug imprisonment
outcomes of selective drug policing—which is made
possible by the wide range of policing resources that
are available across jurisdictions in the U.S.

In order to illustrate the importance of these public
safety decisions and their consequences, we have isolated two counties for comparison. Orleans Parish,
LA, and Lancaster County, NE, have similar index
crime rates, at 6753 and 6207, respectively. Twice as
much is spent on policing, however, in Orleans Parish. Policing accounts for approximately 6 percent of
all public expenditures in Orleans Parish, compared
to about 3 percent in Lancaster County. Expressed in
per capita terms, $222 is spent on policing for every
citizen of Orleans Parish, relative to $113 in Lancaster County.
These disparate funding levels, which reflect public
safety spending decisions made by local decision
makers, have a dramatic impact on the extent to
which individuals are admitted to prison for drug offenses from each locale. The drug imprisonment rate
in Orleans Parish (250 per 100,000) is more than
22	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Appendix A. The 198 counties analyzed in this study with overall drug admission rate, white drug
admission rate, African American drug admission rate, and the ratio of African American to white
drug admission rates.

County

	

State

Drug
Admission
Rate

White
Drug Admission
Rate

African American
Drug Admission
Rate

Ratio of African American
to white drug
admission rates

MECKLENBURG

NC

2.57

2.24

3.70

2

GUILFORD

NC

3.48

.35

9.05

26

MONTGOMERY

MD

3.74

.62

13.71

22

HOWARD

MD

3.84

1.02

18.99

19

BUCKS

PA

3.93

1.93

56.09

29

ROCKINGHAM

NH

5.21

4.97

.00

0

WAKE

NC

6.07

.78

26.42

34

FAIRFAX

VA

6.92

3.31

38.95

12

CUMBERLAND

ME

7.43

4.21

45.16

11

WASHINGTON

OR

8.03

6.59

.00

0

CLACKAMAS

OR

9.38

8.35

45.75

5

ANNE ARUNDEL

MD

10.33

3.37

49.68

15

BOULDER

CO

11.10

4.93

27.23

6

PRINCE GEORGES

MD

11.16

3.21

13.87

4

HILLSBOROUGH

NH

11.21

9.86

63.87

6

WESTMORELAND

PA

12.49

4.75

311.42

66

LUZERNE

PA

12.71

6.86

271.29

40

MORRIS

NJ

13.37

4.66

229.72

49

WAUKESHA

WI

14.57

11.69

276.50

24

FORT BEND

TX

15.52

1.91

34.99

18

MONTGOMERY

PA

15.79

5.10

135.80

27

SEMINOLE

FL

16.77

5.17

111.01

21

ANOKA

MN

16.79

15.33

66.76

4

GLOUCESTER

NJ

16.79

5.17

117.03

23

MARION

FL

16.88

4.65

105.64

23

WASHTENAW

MI

17.05

2.64

111.90

42

UTAH

UT

17.53

16.30

.00

0

BALTIMORE

MD

18.04

6.33

57.43

9

BERGEN

NJ

18.32

8.06

153.09

19

CUMBERLAND

NC

18.79

3.33

40.79

12

PALM BEACH

FL

18.82

5.01

88.82

18

BURLINGTON

NJ

19.18

4.83

86.31

18

COLLIN

TX

19.23

11.57

79.18

7

FORSYTH

NC

19.37

.44

72.05

164

DAKOTA

MN

19.51

14.85

139.00

9

ERIE

NY

19.79

3.41

103.66

30

LANCASTER

NE

20.19

14.23

179.04

13

INGHAM

MI

20.26

4.26

133.11

31

NORTHAMPTON

PA

20.49

14.64

175.01

12

HIDALGO

TX

20.99

1.17

42.27

36

LANCASTER

PA

21.52

14.54

188.81

13

WESTCHESTER

NY

21.55

2.70

100.78

37

GWINNETT

GA

21.67

14.15

66.31

5

DU PAGE

IL

21.85

11.69

199.61

17

MACOMB

MI

22.51

11.94

276.14

23

OAKLAND

MI

22.53

7.22

143.92

20

PASCO

FL

22.63

12.32

344.99

28

MANATEE

FL

23.17

7.15

183.43

26

	

Justice Policy Institute	

23

Appendix A

County

198 Counties Analyzed

State

Drug
Admission
Rate

White
Drug Admission
Rate

African American
Drug Admission
Rate

Ratio of African American
to white drug
admission rates

HAMILTON

TN

23.92

5.00

95.46

19

SOMERSET

NJ

24.20

12.76

146.63

11

CAMERON

TX

24.32

2.31

35.97

16

HENNEPIN

MN

24.42

7.36

161.10

22

VIRGINIA BEACH

VA

24.43

11.28

76.28

7

CHESTER

PA

24.44

6.34

276.58

44

SNOHOMISH

WA

24.61

22.48

138.01

6

LARIMER

CO

24.94

17.66

65.47

4

NASSAU

NY

25.80

4.69

166.98

36

DANE

WI

25.95

4.46

433.76

97

WILLIAMSON

TX

26.56

12.98

88.54

7

LANE

OR

27.24

22.77

41.39

2

BRAZORIA

TX

27.60

8.90

161.97

18

ROCKLAND

NY

27.76

5.07

143.34

28

DELAWARE

PA

27.83

7.90

130.43

17

MAHONING

OH

28.03

10.49

115.05

11

ALLEGHENY

PA

28.11

6.41

170.02

27

ONONDAGA

NY

28.21

2.26

224.90

99

HENRICO

VA

28.33

11.80

73.49

6

PRINCE WILLIAM

VA

28.54

11.62

90.18

8

BREVARD

FL

28.65

9.76

202.98

21

SUFFOLK

NY

28.66

6.27

227.22

36

KNOX

TN

28.77

11.71

197.32

17

DUTCHESS

NY

29.19

5.20

208.84

40

MONTGOMERY

TX

29.53

13.24

304.38

23

ARAPAHOE

CO

29.60

13.71

145.44

11

SPOKANE

WA

30.18

25.27

226.76

9

ERIE

PA

30.67

11.85

276.41

23

OCEAN

NJ

31.09

15.28

430.79

28

CLARK

WA

33.49

33.93

53.06

2

CHESTERFIELD

VA

33.56

17.66

96.00

5

LEE

FL

33.85

9.90

319.47

32

DENTON

TX

34.80

27.50

95.66

3

SARASOTA

FL

35.33

13.08

485.56

37

MONROE

NY

36.43

5.95

175.44

29

DEKALB

GA

36.48

20.64

49.73

2

SALT LAKE

UT

36.77

33.32

251.64

8

JEFFERSON

CO

36.86

26.27

155.60

6

KENT

MI

36.91

9.31

276.91

30

MARION

OR

37.52

28.64

87.89

3

MULTNOMAH

OR

38.22

20.26

185.63

9

COLLIER

FL

38.31

20.27

316.22

16

DOUGLAS

NE

39.13

26.83

127.53

5

NUECES

TX

40.36

7.85

171.67

22

EL PASO

TX

40.86

3.21

80.51

25

ST. LOUIS

MO

41.74

18.15

135.57

7

LEHIGH

PA

42.52

25.84

343.89

13

DUVAL

FL

42.80

9.79

123.44

13

MIDDLESEX

NJ

42.81

14.71

268.73

18

ST. CHARLES

MO

43.56

35.83

266.08

7

FRANKLIN

OH

44.17

12.00

172.44

14

EL PASO

CO

44.32

23.28

174.04

7

GENESEE

MI

44.40

7.85

180.23

23

24	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

County

	

State

Drug
Admission
Rate

White
Drug Admission
Rate

African American
Drug Admission
Rate

Ratio of African American
to white drug
admission rates

WILL

IL

45.73

12.24

277.14

23

WAYNE

MI

46.30

7.86

97.18

12

BEXAR

TX

46.46

8.07

156.05

19

ADAMS

CO

46.51

17.00

402.40

24

SONOMA

CA

46.54

24.46

320.87

13

MADISON

AL

48.62

17.57

148.93

8

CONTRA COSTA

CA

49.68

25.85

218.78

8

TRAVIS

TX

49.95

9.63

302.63

31

VENTURA

CA

50.13

22.41

241.28

11

LUCAS

OH

50.27

17.19

177.33

10

KING

WA

51.15

20.03

464.64

23

ORANGE

FL

51.88

17.55

173.32

10

MOBILE

AL

52.23

24.79

104.68

4

YORK

PA

52.93

25.99

617.98

24

RAMSEY

MN

53.47

27.94

303.70

11

PINELLAS

FL

53.52

23.34

328.79

14

PIERCE

WA

53.53

38.42

216.24

6

TARRANT

TX

55.91

27.06

186.40

7

SAN LUIS OBISPO

CA

56.04

36.00

32.95

1

SHELBY

TN

57.31

8.79

100.75

11

MONMOUTH

NJ

58.27

19.05

456.11

24

PLACER

CA

58.53

50.68

217.92

4

GALVESTON

TX

62.02

15.99

299.50

19

MONTGOMERY

OH

63.30

19.76

224.68

11

SANTA CRUZ

CA

63.83

29.96

434.89

15

DALLAS

TX

63.84

19.95

175.42

9

BERKS

PA

64.90

53.76

280.94

5

ORANGE

NY

67.55

13.34

461.56

35

COBB

GA

69.38

33.40

211.19

6

VOLUSIA

FL

73.13

23.32

520.01

22

JEFFERSON

KY

73.77

22.12

279.47

13

HILLSBOROUGH

FL

74.58

29.40

298.11

10

JEFFERSON

AL

75.63

30.63

142.51

5

RICHLAND

SC

75.97

8.87

153.23

17

FULTON

GA

76.69

9.70

161.89

17

SAN MATEO

CA

76.79

26.89

946.32

35

BUTLER

OH

77.52

34.16

732.22

21

BROWARD

FL

77.76

23.95

248.81

10

HARRIS

TX

78.04

14.63

279.94

19

SPARTANBURG

SC

78.28

30.05

247.48

8

ESCAMBIA

FL

78.38

24.55

257.11

10

FAYETTE

KY

78.90

27.00

390.03

14

MONTEREY

CA

82.24

21.74

537.87

25

MERCER

NJ

85.13

10.35

344.45

33

STARK

OH

85.20

32.76

668.96

20

KANE

IL

85.32

12.13

809.76

67

POLK

FL

85.62

62.48

225.98

4

SUMMIT

OH

87.30

31.62

424.80

13

LORAIN

OH

88.08

32.36

594.71

18

GREENVILLE

SC

89.18

25.03

360.55

14

ALBANY

NY

91.50

10.78

625.80

58

SANTA BARBARA

CA

92.78

35.50

508.19

14

WINNEBAGO

IL

93.41

22.47

592.92

26

	

Justice Policy Institute	

25

Appendix A

198 Counties Analyzed

County

State

Drug
Admission
Rate

White
Drug Admission
Rate

African American
Drug Admission
Rate

Ratio of African American
to white drug
admission rates

ANCHORAGE

AK

94.80

60.65

319.82

5

CHARLESTON

SC

95.40

10.82

255.09

24

SANTA CLARA

CA

97.95

40.26

594.36

15

WASHOE

NV

98.53

58.48

956.80

16

LOS ANGELES

CA

99.61

24.15

416.16

17

NEW YORK CITY TOTAL

NY

100.95

32.09

174.17

5

PULASKI

AR

106.21

66.00

188.02

3

SAN DIEGO

CA

107.27

41.72

603.57

14

STANISLAUS

CA

110.27

71.59

438.98

6

UNION

NJ

112.10

23.60

380.98

16

DAUPHIN

PA

112.68

23.29

500.69

21

ORANGE

CA

115.40

66.75

463.89

7

TULARE

CA

116.30

44.60

583.37

13

PHILADELPHIA

PA

116.67

85.30

162.71

2

EAST BATON ROUGE

LA

118.20

44.33

223.92

5

DAVIDSON

TN

119.31

26.09

356.93

14

CLAYTON

GA

120.68

138.19

115.98

1

SAN JOAQUIN

CA

122.09

46.63

576.93

12

MILWAUKEE

WI

123.14

26.56

390.09

15

SAN FRANCISCO

CA

123.42

35.83

1013.89

28

PASSAIC

NJ

124.03

28.33

532.36

19

HAMILTON

OH

124.74

32.83

403.46

12

OKLAHOMA

OK

125.21

79.05

292.16

4

SOLANO

CA

127.96

66.75

386.30

6

TULSA

OK

128.47

80.35

417.31

5

FRESNO

CA

128.92

41.29

491.10

12

JACKSON

MO

130.00

49.08

382.90

8

HUDSON

NJ

130.18

24.38

601.30

25

POLK

IA

136.64

113.85

563.57

5

SACRAMENTO

CA

138.84

67.38

512.42

8

ESSEX

NJ

140.30

10.61

291.63

27

CADDO

LA

140.56

45.10

254.80

6

DENVER

CO

147.39

32.75

638.04

19

RIVERSIDE

CA

148.14

71.47

404.97

6

ALAMEDA

CA

154.93

23.11

797.49

35

COOK

IL

166.25

9.64

558.69

58

SAN BERNARDINO

CA

170.15

80.54

407.76

5

JEFFERSON

LA

185.96

65.00

559.84

9

CUYAHOGA

OH

209.42

51.84

597.69

12

CAMDEN

NJ

217.21

76.68

601.69

8

ST. LOUIS

MO

239.10

43.04

409.24

10

ORLEANS

LA

249.54

77.48

331.63

4

ATLANTIC

NJ

256.34

67.06

960.49

14

KERN

CA

319.86

149.48

917.57

6

26	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Appendix B. Distribution of Social Structural Variables

Univariate Description of Study Variables for Study Sample of 198 Counties with Population > 250,000
Range

Mean

Std. Dev.

25th Percentile

50th Percentile

75th Percentile

Drug Imprisonment Rate

2.57 -319.86

61.87

53.02

16.46

35.74

61.36

Unemployment Rate

3.20 – 12.0

5.72

1.45

24.13

44.24

87.50

Poverty Rate

3.50 – 33.00

10.71

4.62

7.68

10.10

12.70

Percent African American

0.64 – 68.11

14.41

13.04

4.60

10.52

20.49

Per Capita Policing Budget

64.78 – 510.82

199.58

74.15

153.34

188.94

227.20

Per Capita Judicial Budget

3.25 – 238.36

62.73

39.08

36.66

51.90

78.95

Policing as Percent of Budget

2.24 – 11.50

5.09

1.49

4.11

4.96

5.85

Judicial as Percent of Budget

0.07 – 4.17

1.55

0.75

1.05

1.43

1.99

4417.72

1915.38

3030.57

4144.71

5621.79

Index Crime Rate

365.78 – 15077.15

Appendix C. Correlations Between Drug Imprisonment Rate and Social Structural Variables
Drug Imprisonment Rate
Unemployment Rate

.341 (.000)

Poverty Rate

.441 (.000)

Percent African American

.329 (.000)

Per Capita Policing Budget

.425 (.000)

Per Capita Judicial Budget

.463 (.000)

Policing as Percent of Budget

.211 (.003)

Judicial as Percent of Budget

.358 (.000)

Index Crime Rate

.339 (.000)

Appendix D. U.S. States Examined in this Study, by U.S. Census Bureau Regions

	

Region

States in Region

West

AK, CA, CO, HI, OR, UT, NV, WA

Midwest

IL, IA, MI, MN, MO, NE, ND, OH, SD, WI

South

AL, AR, FL, GA, KY, LA, MD, MS, NC, OK, SC, TX, TN, VA, WV

Northeast

ME, NH, NJ, NY, PA

	

Justice Policy Institute	

27

Appendix E

Ordinary Least Squares Estimates

Appendix E. Ordinary Least Squares Estimates from Regression of Drug Imprisonment
Rates on Sociodemographic, Budget, Index Crime Rate, and Region Variables for 198
Large-Population Counties/Municipalities (2002)

Variable

Model 1

Model 2

Coefficient
(Std. Error)

Coefficient
(Std. Error)

Unemployment Rate

-.609
(3.061)

-.636
(3.101)

Poverty Rate

.3.776**
(1.062)

3.987**
(1.077)

Percent African American

.939**
(.314)

1.114**
(.310)

Per Capita Policing Budget

.148**
(.045)

__

Per Capita Judicial Budget

.278**
(.095)

__

Percent of Budget Policing

__

6.803**
(2.024)

Percent of Budget Judicial

__

12.858**
(4.363)

Index Crime Rate

.0005
(.002)

.002
(.002)

Northeast

19.407*
(9.405)

31.889**
(9.256)

Midwest

16.586
(9.559)

26.795**
(9.311)

West

28.886*
(11.515)

47.416**
(10.695)

Intercept

-51.425

-79.504

R2

.436

.421

Note: Coefficients are unstandardized.
* p < .05; ** p < .01

28	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

Endnotes
1 U.S. Department of Justice, Bureau of Justice Statistics (2005).
“Key Facts at a Glance: Incarceration Rate 1980-2005.” http://
www.ojp.usdoj.gov/bjs/glance/tables/incrttab.htm accessed on
January 11, 2007.
2 Mauer, M. (1999), Race to Incarcerate. New York: The New
Press. Schiraldi, V., Holman, B., Beatty, P. (2000), “Poor
Prescription: The Cost of Imprisoning Drug Offenders in the
United States.” Washington, D.C.: Justice Policy Institute.
3 Harrison, P.M. and Beck, A. J. (2006), “Prisoners in 2005.”
Washington, D.C.: U.S. Department of Justice. Beck, A.J. and
Mumola, C.J. (1999), “Prisoners in 1998.” Washington, D.C.:
U.S. Department of Justice.
4 Harlow, C.W. (1998), “Profile of Jail Inmates 1996.”
Washington, D.C.: U.S. Department of Justice. James, D.J.
(2004), “Profile of Jail Inmates 2002.” Washington, D.C.: U.S.
Department of Justice.
5 American Correctional Association (2006), “2006 Directory
of Adult and Juvenile Correctional Departments, Institutions,
Agencies and Probation and Parole Authorities, 67th Edition.”
Alexandria, VA: ACA. Harrison, P.M. and Beck, A. J. (2006),
“Prisoners in 2005.” Washington, D.C.: U.S. Department of
Justice.
6 Spelling, W. (2000), “The Limited Importance of Prison
Expansion.” In The American Crime Drop, eds. Blumstein,
Alfred and Wallman, Joel. Boston: Cambridge University Press.
Rosenfeld, R. (2000), “Patterns in Adult Homicide: 1980–
1995.” In The American Crime Drop, eds. Blumstein, Alfred and
Wallman, Joel. Boston: Cambridge University Press. Clear, T.
R. (2003), “The Problem with ‘Addition by Subtraction’: The
Prison-Crime Relationship in Low Income Communities,”
In Invisible Punishment: The Collateral Consequences of Mass
Imprisonment. Eds. Mauer and Chesney-Lind. New York: New
Press.
7 Schiraldi, V., Holman, B., Beatty, P. (2000), “Poor Prescription:
The Cost of Imprisoning Drug Offenders in the United States.”
Washington, D.C.: Justice Policy Institute.
8 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes.
9 Harrison, P.M. and Beck, A. J. (2006), “Prisoners in 2005.”
Washington, D.C.: U.S. Department of Justice.
10 U.S. Bureau of the Census (2000). Selected Population
Group: Black or African American alone or in combination
with one or more other races. Web page: http://factfinder.
census.gov/servlet/SAFFIteratedFacts?_event=&geo_
id=01000U.S.&_geoContext=01000U.S.&_street=&_
county=&_cityTown=&_state=&_zip=&_lang=en&_
sse=on&ActiveGeoDiv=&_useEV=&pctxt=fph&pgsl=010&
_submenuId=factsheet_2&ds_name=DEC_2000_SAFF&_ci_
nbr=005&qr_name=DEC_2000_SAFF_R1010&reg=DEC_
2000_SAFF_R1010%3A005&_keyword=&_industry=
referenced April 9, 2007. Harrison, P.M. and Beck, A.
J. (2006), “Prisoners in 2005.” Washington, D.C.: U.S.
Department of Justice.
11 SAMHSA (2005). Results from the 2002 National Survey
on Drug Use and Health: Detailed Tables. Prevalence
Estimates, Standard Errors, and Sample Sizes. Office of
National Drug Control Policy (2004). National Drug Control
Strategy, Data Supplement. Table 34. Web page: http://www.
whitehousedrugpolicy.gov/publications/policy/ndcs04/
data_suppl_2004.pdf accessed on January 28, 2007. U.S.
Department of Justice, Bureau of Justice Statistics. National
Corrections Reporting Program, 2002: [United States]
[Computer file].
12 SAMHSA (2005). Results from the 2002 National Survey
	

	

on Drug Use and Health: Detailed Tables. Prevalence
Estimates, Standard Errors, and Sample Sizes. Snyder, H.N.
and Sickmund, M. (March 2006), Juvenile Offenders and
Victims: 2006 National Report. Office of Juvenile Justice and
Delinquency Prevention.
13 This data is from U.S. Department of Justice, Bureau of
Justice Statistics. National Corrections Reporting Program
National Corrections Reporting Program, 2002: [United
States] [Computer file].
14 U.S. Bureau of the Census (2003). Statistical Abstracts of the
United States. Mini Historical Statistics. Web page: http://
www.census.gov/statab/hist/HS-01.pdf referenced on January
20, 2007.
15 U.S. Department of Justice, Bureau of Justice Statistics.
National Corrections Reporting Program, 2002: [United
States] [Computer file]. Conducted by U.S. Department of
Commerce, Bureau of the Census. ICPSR ed. Ann Arbor, MI:
Inter-university Consortium for Political and Social Research
[producer and distributor], 2006.
16 U.S. Bureau of the Census (2005). 2002 Census of
Governments, Volume 4, Number 5. Compendium of
Government Finances: 2002. GC02(4)-5. U.S. Government
Printing Office, Washington DC. Raw data obtained from
the following web page: http://www.census.gov/govs/www/
02censustechdoc.html. Accessed on January 20, 2007.
17 Federal Bureau of Investigation (2002). Uniform Crime
Report. Washington, D.C.: U.S. Department of Justice.
18 U.S. Bureau of the Census (2005). Annual Estimates of
the Population by Sex, Race Alone, or in Combination and
Hispanic or Latino Origin for Counties: April 1, 2000 to July
1, 2005. Web page: http://www.census.gov/popest/counties/
asrh/CC-EST2005-RACE5.html accessed on January 20,
2007. U.S. Bureau of the Census (2002). Small Area Income
and Poverty Estimates. 2002 State and County Files and
Description. Web page: http://www.census.gov/hhes/www/
saipe/stcty/sc02ftpdoc.html accessed on January 20, 2007.
U.S. Bureau of the Census (2004). County level population
estimates for 2002 obtained from the following web page:
http://www.census.gov/popest/archives/2000s/vintage_
2002/#county, accessed on January 21, 2007. U.S. Bureau of
the Census (2005). Annual Estimates of the Population by
Sex, Race Alone or in Combination and Hispanic or Latino
Origin for Counties: April 1, 2000 to July 1, 2005. Web page:
http://www.census.gov/popest/counties/asrh/CC-EST2005RACE5.html accessed on January 20, 2007.
19 U.S. Department of Labor, Bureau of Labor Statistics (2007).
Local Area Unemployment Characteristics. Web page: http://
www.bls.gov/lau/ accessed on January 28, 2007.
20 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes.
21 Office of National Drug Control Policy (2004). National
Drug Control Strategy, Data Supplement. Table 34. Web page:
http://www.whitehousedrugpolicy.gov/publications/policy/
ndcs04/data_suppl_2004.pdf accessed on January 28, 2007.
22 Office of National Drug Control Policy (2004). National
Drug Control Strategy, Data Supplement. Table 34. Web page:
http://www.whitehousedrugpolicy.gov/publications/policy/
ndcs04/data_suppl_2004.pdf accessed on January 28, 2007.
23 U.S. Department of Justice, Bureau of Justice Statistics.
National Corrections Reporting Program, 2002: [United
States] [Computer file].
24 This is an underestimate of total drug admissions in the U.S.
in 2002, as the NCRP does not contain data for 12 states.
The 12 states for which there are no data available in the 2002
NCRP include five states from the Mountain West (Arizona,
Idaho, Montana, New Mexico, and Wyoming), five states

Justice Policy Institute	

29

endnotes

from the Northeast (Delaware, Rhode Island, Connecticut,
Vermont, and Massachusetts), and two states from the
Midwest (Kansas, Indiana).
25 Reich, C. (1973). “The Law and the Corporate State.” In
Chambliss, W., (ed.), Sociological Readings in the Conflict
Perspective. pp. 445–454. Reading, MA: Addison-Wesley
Publishing Company.
26 Johnston, L., O’Malley, P., and Bachman, J. (2003),
“Monitoring the Future National Results on Adolescent Drug
Use: Overview of Key Findings 2002.” NIH Publication No.
03-5374. Bethesda: MD: National Institute on Drug Abuse.
27 Snyder, H.N. and Sickmund, M. (March 2006), “Juvenile
Offenders and Victims: 2006 National Report.” Office of
Juvenile Justice and Delinquency Prevention.
28 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes.
29 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes.
30 U.S. Department of Justice, Bureau of Justice Statistics.
National Corrections Reporting Program, 2002: [United
States] [Computer file]. Conducted by U.S. Department of
Commerce, Bureau of the Census. ICPSR ed. Ann Arbor, MI:
Inter-university Consortium for Political and Social Research
[producer and distributor], 2006. U.S. Bureau of the Census,
Annual Population Estimates 2000 to 2006, http://www.
census.gov/popest/states/NST-ann-est.html accessed on May
14, 2007.
31 SAMHSA (2005). Results from the 2002 National Survey
on Drug Use and Health: Detailed Tables. Prevalence
Estimates, Standard Errors, and Sample Sizes, U.S. Sentencing
Commission (2006). 2006 Sourcebook on Federal Sentencing
Statistics, Table 34. Washington, DC: U.S. Sentencing
Commission.
32 Snyder, Howard N. and Sickmund, Melissa (2006), “Juvenile
Offenders and Victims: 2006 National Report.” Washington,
D.C.: Office of Juvenile Justice and Delinquency Prevention.
D.C.
33 Snyder, Howard N. and Sickmund, Melissa (2006), “Juvenile
Offenders and Victims: 2006 National Report.” Washington,
D.C.: Office of Juvenile Justice and Delinquency Prevention.
D.C.
34 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes. Johnston, L., O’Malley,
P., and Bachman, J. (2003), “Monitoring the Future National
Results on Adolescent Drug Use: Overview of Key Findings
2002.” NIH Publication No. 03-5374. Bethesda: MD:
National Institute on Drug Abuse.
35 Riley, J. (1997), “Crack, Powder Cocaine, and Heroin: Drug
Purchase and Use Patterns in Six U.S. Cities.” Washington
D.C.: National Institute of Justice and the Office of
National Drug Control Policy; Pennell, S., Ellett, J., Rienick,
C., and Grimes, J. (1999), “Meth Matters: Report on
Methamphetamine Users in Five Western Cities.” Washington,
D.C.: Office of Justice Programs, National Institute of Justice.
36 Schiraldi, V., Holman, B., Beatty, P. (2000), “Poor
Prescription: The Cost of Imprisoning Drug Offenders in the
United States.” Washington, D.C.: Justice Policy Institute.
37 Stemen, D., Rengifo, A., and Wilson, J. (2005), “Of
Fragmentation and Ferment: The Impact of State Sentencing
Policies on Incarceration Rates, 1975–2002.” Vera Institute of
Justice.
38 Justice Policy Institute (2007), “Maryland’s Mandatory
Minimum Drug Sentencing Laws: Their Impact on
Incarceration, State Resources, and Communities of Color.”

30	

Washington, D.C.: Justice Policy Institute.
39 Justice Policy Institute (2007), “Maryland’s Mandatory
Minimum Drug Sentencing Laws: Their Impact on
Incarceration, State Resources, and Communities of Color.”
Washington, D.C.: Justice Policy Institute.
40 U.S. Department of Justice, Bureau of Justice Statistics
(1998), “Compendium of Federal Justice Statistics, 1994,
Table 6.11, at 85.” Washington, D.C.: Bureau of Justice
Statistics; U.S. Department of Justice, Bureau of Justice
Statistics (2004), “Compendium of Federal Justice Statistics,
2003, Table 7.16 at 112.” Washington, D.C.: Bureau of Justice
Statistics.
41 Hoytt, E.H., Schiraldi, V., Smith, B.V., and Ziedenberg, J.
(2002), “Pathways to Juvenile Detention Reform: Reducing
Racial Disparities in Juvenile Detention.” Baltimore, MD:
Annie E. Casey Foundation.
42 Bridges, G.S. and Steen, S. (1998), “Racial Disparities in
Official Assessments of Juvenile Offenders; Attributional
Stereotypes as Mediating Mechanisms.” American Sociological
Review, Volume 63.
43 Hoytt, E.H., Schiraldi, V., Smith, B.V., and Ziedenberg, J.
(2002), “Pathways to Juvenile Detention Reform: Reducing
Racial Disparities in Juvenile Detention.” Baltimore, MD:
Annie E. Casey Foundation.
44 Schiraldi, V. and Ziedenberg, J. (2003), “Race and
Incarceration in Maryland.” Washington, D.C.: Justice Policy
Institute.
45 Beckett, K., Western, B., (2001), “Governing Social
Marginality: Welfare, Incarceration, and the Transformation
of State Policy.” Chapter 4 in Garland, D. (ed.), Mass
Imprisonment: Social Causes and Consequences. Thousand Oaks,
CA: Sage Publications.
46 Wright, D. (2003), “State Estimates of Substance Use
from the 2001 National Household Survey on Drug Abuse:
Volume I. Findings.” (DHHS Publication No. SMA 03-3775,
NHSDA Series H-19). Rockville, MD: Substance Abuse and
Mental Health Services Administration, Office of Applied
Studies.
47 SAMHSA, Office of Applied Studies (2006). Substate
Estimates From the 2002–2004 National Surveys on Drug
Use and Health. Rockville, MD: Substance Abuse and Mental
Health Services Administration.
48 Schiraldi, V., Holman, B., Beatty, P. (2000), “Poor
Prescription: The Cost of Imprisoning Drug Offenders in the
United States.” Washington, D.C.: Justice Policy Institute.
49 Schiraldi, V., Holman, B., Beatty, P. (2000), “Poor
Prescription: The Cost of Imprisoning Drug Offenders in the
United States.” Washington, D.C.: Justice Policy Institute.
50 SAMHSA, Office of Applied Studies (2006). Substate
Estimates from the 2002–2004 National Surveys on Drug
Use and Health. Rockville, MD: Substance Abuse and Mental
Health Services Administration.
51 SAMHSA, Office of Applied Studies (2006). Substate
Estimates from the 2002–2004 National Surveys on Drug
Use and Health. Rockville, MD: Substance Abuse and Mental
Health Services Administration.
52 Jencks (1991), “Is Violent Crime Increasing?” The American
Prospect. (Winter): 98—109; O’Brien, R. (1996), “Police
Productivity and Crime Rates: 1973–1992.” Criminology.
34(2): 183–207.
53 O’Brien, R. (1996), “Police Productivity and Crime Rates:
1973–1992.” Criminology. 34(2): 183–207.
54 Black, D. (1970), “Production of Crime Rates,” American
Sociological Review. 34: 733–748.
55 Judicial expenses refer to administration of the courts and
activities associated with courts including prosecutorial and
defendant programs, law libraries, probate functions, and juries.

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

56 Reich, C. (1973), “The Law and The Corporate State.”
In Chambliss, W. (ed.) Sociological Readings in the Conflict
Perspective. Reading, MA: Addison-Wesley Publishing
Company.
57 Beckett, K (2004), “Race and Drug Law Enforcement in
Seattle.” http://www.defender.org/files/Beckett-20040503.pdf
accessed on January 15, 2007.
58 Beckett, K (2004), “Race and Drug Law Enforcement in
Seattle.” http://www.defender.org/files/Beckett-20040503.pdf
accessed on January 15, 2007, p. 87.
59 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes; Johnston, L., O’Malley,
P., and Bachman, J. (2003), “Monitoring the Future National
Results on Adolescent Drug Use: Overview of Key Findings
2002.” NIH Publication No. 03-5374. Bethesda: MD:
National Institute on Drug Abuse.
60 Riley, J. (1997), “Crack, Powder Cocaine, and Heroin:
Drug Purchase and Use Patterns in Six U.S. Cities.”
Washington D.C.: National Institute of Justice and the
Office of National Drug Control Policy; Pennell, S., Ellett,
J., Rienick, C., Grimes, J. (1999), “Meth Matters: Report on
Methamphetamine Users in Five Western Cities.” Washington,
D.C.: Office of Justice Programs, National Institute of Justice;
Beckett, K. (2004), “Race and Drug Law Enforcement in
Seattle.” http://www.defender.org/files/Beckett-20040503.pdf
accessed on January 15, 2007.
61 SAMHSA, Office of Applied Studies (2006). Substate
Estimates from the 2002–2004 National Surveys on Drug
Use and Health. Rockville, MD: Substance Abuse and Mental
Health Services Administration.
62 Reiman, J. (1998), The Rich Get Richer and the Poor Get
Prison: Ideology, Class, and Criminal Justice. Boston: Allyn and
Bacon; McGarrell, E. (1993), “Institutional Theory and the
Stability of a Conflict Model of the Incarceration Rate.” Justice
Quarterly. 10(1): 7-27; Colvin, M. (1990), “Labor Markets,
Industrial Monopolization, Welfare and Imprisonment.”
Sociological Quarterly. 31: 441-457; Beckett, K., and
Western, B., (2001), “Governing Social Marginality: Welfare,
Incarceration, and the Transformation of State Policy.” In
Garland, D. (Ed.) Mass Imprisonment: Social Causes and
Consequences. Thousand Oaks, CA: Sage Publications.
63 Reiman, J. (1998). The Rich Get Richer and the Poor
Get Prison: Ideology, Class, and Criminal Justice. Boston:
Allyn and Bacon; Western, B., and Beckett, K. (1999).
“How Unregulated Is the U.S. Labor Market? The Penal
System as a Labor Market Institution,” American Journal of
Sociology. 104(4): 1030–1060; Box, S., and Hale, C. (1985),
“Unemployment, Imprisonment and Prison Overcrowding,”
Contemporary Crises. 9: 209-228; Inverarity, J., and McCarthy,
D. (1988), “Punishment and Social Structure Revisited:
Unemployment and Imprisonment in the U.S: 1948–1984.”
Sociological Quarterly. 29: 263–279; Inverarity, J., and Grattet,
R. (1989), “Institutional Responses to Unemployment: A
Comparison of U.S. Trends, 1948–1985.” Contemporary
Crises. 13: 351–370.
64 Beckett, K., and Western, B., (2001), “Governing Social
Marginality: Welfare, Incarceration, and the Transformation
of State Policy.” In Garland, D. (ed.) Mass Imprisonment:
Social Causes and Consequences. Thousand Oaks, CA: Sage
Publications; McGarrell, 1993. Nagel, J. (1982), “The
Relationship Between Crime and Incarceration Among the
American States.” Policy Studies Review. 2(2): 193–200;
Joubert, P., Picou, S., and McIntosh, A. (1981), “U.S.
Social Structure, Crime, and Punishment.” Criminology. 19:
344–359.
65 Mauer, M. (2001), “Drug Policy and the Criminal Justice

	

	

System.” Washington: D.C.: The Sentencing Project.
66 Mauer, M. (2001), “Drug Policy and the Criminal Justice
System.” Washington: D.C.: The Sentencing Project.
67 Reiman, J. (1998), The Rich Get Richer and the Poor Get
Prison: Ideology, Class, and Criminal Justice. Boston: Allyn and
Bacon.
68 Nagel, W. (1977), “On Behalf of a Moratorium on Prison
Construction.” Crime and Delinquency. 1977: 154-172.
69 Colvin, M. (1990), “Labor Markets, Industrial
Monopolization, Welfare and Imprisonment.” Sociological
Quarterly. 31: 441-457.
70 Beckett, K., and Western, B., (2001), “Governing Social
Marginality: Welfare, Incarceration, and the Transformation
of State Policy.” In Garland, D. (ed.) Mass Imprisonment:
Social Causes and Consequences. Thousand Oaks, CA: Sage
Publications., p. 44.
71 Reiman, J. (1998), The Rich Get Richer and the Poor Get
Prison: Ideology, Class, and Criminal Justice. Boston: Allyn and
Bacon.
72 Feser, E., Sweeney, S. (2003), “Out-Migration, Depopulation,
and the Geography of U.S. Economic Distress.” International
Regional Science Review. 26(1): 38–67.
73 Burchell, B. (1992), “Towards a Social-Psychology of the
Labor-Market – or Why We Need to Understand the Labor
Market Before We Can Understand Unemployment.” Journal
of Occupational and Organizational Psychology. 65: 345–354.
74 D’Alessio, S., Stolzenberg, L. (2002), “A Multilevel Analysis
of the Relationship Between Labor Surplus and Pretrial
Incarceration.” Social Problems. 49(2): 178-193; Hochstetler,
A., Shover, N. (1997), “Street Crime, Labor Surplus, and
Criminal Punishment, 1980–1990.” Social Problems. 44(3):
358–368; Chiricos, T., Delone, M. (1992), “Labor Surplus
and Imprisonment: A Review and Assessment of Theory and
Evidence.” Social Problems. 39: 421–426.
75 Hagan, J., and Dinovitzer, R. (1999), “Collateral
Consequences of Imprisonment for Children, Communities,
and Prisoners.” In Tonry M., and Petersilia J. (eds), Crime and
Justice: A Review of Research. Chicago: University of Chicago
Press.
76 Western, B., Kling, J., and Weiman, D. (2001), “The
Labor Market Consequences of Incarceration.” Crime and
Delinquency. 47: 410–427.
77 Wilson, W.J. (1987), The Truly Disadvantaged: The Inner
City, the Underclass, and Public Policy. Chicago: University of
Chicago Press.
78 Richie, B.E. (2001), “Challenges Incarcerated Women Face
as They Return to Their Communities: Findings from Life
History Interviews.” Crime and Delinquency. 47: 368–369.
79 Fellner, J., Mauer, M. (1998), “Losing the Vote: The Impact
of Felony Disenfranchisement Laws in the United States.”
Washington, D.C.: Human Rights Watch and the Sentencing
Project.
80 Phillips, S., and Bloom, B. (1998), “In Whose Best Interests?
The Impact of Changing Public Policy on Relatives Caring
for Children with Incarcerated Parents.” Child Welfare. 77:
531–541; Gabel, S. (2003), “Behavioral Problems in Sons
of Incarcerated or Otherwise Absent Fathers.” In Harris, O.,
and Miller, R. (eds.), Impacts of Incarceration on the African
American Family. New Brunswick, NJ: Transaction Publishers.
81 Freudenberg, N. (2001), “Jails, Prisons, and the Health
of Urban Populations: A Review of the Impact of the
Correctional System on Community Health.” Journal of Urban
Health. 78: 214–235; Freudenberg, N. (2002), “Adverse
Effects of U.S. Jail and Prison Policies on the Health and WellBeing of Women of Color.” American Journal of Public Health.
92(12): 1895–1899.
82 Beckett, K., and Western, B., (2001), “Governing Social

Justice Policy Institute	

31

endnotes

Marginality: Welfare, Incarceration, and the Transformation
of State Policy.” In Garland, D. (ed.) Mass Imprisonment:
Social Causes and Consequences. Thousand Oaks, CA: Sage
Publications, p. 44.
83 The information used in this control comes from the FBI’s
county-level index crime rate.
84 Carroll, L., Doubet, M. (1983), “U.S. Social Structure and
Imprisonment: A Comment.” Criminology. 21: 449–456.
85 The information used for this control comes from the Census
Bureau’s four-category region indicator for each jurisdiction.
86 SAMHSA (2005). Results from the 2002 National Survey on
Drug Use and Health: Detailed Tables. Prevalence Estimates,
Standard Errors, and Sample Sizes; Johnston, L., O’Malley,
P., and Bachman, J. (2003), “Monitoring the Future National
Results on Adolescent Drug Use: Overview of Key Findings
2002.” NIH Publication No. 03-5374. Bethesda: MD:
National Institute on Drug Abuse.
87 Riley, J. (1997), “Crack, Powder Cocaine, and Heroin:
Drug Purchase and Use Patterns in Six U.S. Cities.”
Washington D.C.: National Institute of Justice and the
Office of National Drug Control Policy; Pennell, S., Ellett,
J., Rienick, C., Grimes, J. (1999), “Meth Matters: Report on
Methamphetamine Users in Five Western Cities.” Washington,
D.C.: Office of Justice Programs, National Institute of Justice.
88 Cole, D. (1999), No Equal Justice: Race and Class in the
American Criminal Justice System. New York: The New Press,
1999, p. 8.
89 Beckett, K., and Western, B., (2001), “Governing Social
Marginality: Welfare, Incarceration, and the Transformation
of State Policy.” In Garland, D. (ed.) Mass Imprisonment:
Social Causes and Consequences. Thousand Oaks, CA: Sage
Publications, p. 44.

Beatty, Phillip, Amanda Petteruti, and Jason Ziedenberg (2007). The Vortex: The Concentrated Racial
Impact of Drug Imprisonment and the Characteristics of Punitive Counties. Washington, DC: Justice
Policy Institute.

32	

The Vortex: The Concentrated Racial Impact of Drug Imprisonment and the Characteristics of Punitive Counties

About the Authors
Phillip Beatty is a Sociologist with a PhD from American University. As a Consultant to the Justice Policy Institute since 1999, his data analyses have served as the basis
for a number of JPI reports related to drug policy and prison admission rates. Beatty
authored and researched several reports concerning drug policy, including Poor Prescription: The Cost of Imprisoning Drug Offenders in the United States, Cost and Benefits? The
Impact of Drug Imprisonment in New Jersey, and Drugs and Disparity: The Racial Impact
of Illinois’ Practice of Transferring Young Drug Offenders to Adult Court.
Amanda Petteruti is a researcher and policy analyst with approximately seven years of
combined experience in education and criminal justice policy. Early in her career, she
organized a writing program for youth at the National Campaign to Stop Violence and
provided general support to the National Juvenile Defender Center. Prior to joining the
staff of the Justice Policy Institute, she conducted research on issues pertaining to urban
education at the Council of the Great City Schools. Petteruti has earned a Master of
Arts in education policy and leadership from the University of Maryland College Park
and a Bachelor of Arts in sociology from Bates College. Petteruti has contributed to
several reports related to education policy and co-authored Education and Public Safety,
Employment, Wages and Public Safety, Housing and Public Safety, and Drug Treatment
and Public Safety.
Jason Ziedenberg is the co-founder and Executive Director of the Justice Policy
Institute, one of the nation’s leading prison reform think tank. His research and policy
work on juvenile and criminal justice policy is frequently used by nonprofits, foundations, think tanks, law enforcement, community organizations, government, and the
media. He has written over a dozen policy briefs and reports that have sought to
reduce the inappropriate use of prison for drug offenders, including Maryland’s Mandatory Minimum Drug Sentencing Laws: Their Impact on Incarceration, State Resources
and Communities of Color; Proposition 36: Five Years Later; Efficacy and Impact: The
Criminal Justice Response to Marijuana Policy in the United States; Treatment or Incarceration: National and State Findings on the Efficacy and Cost Savings of Drug Treatment
Versus Imprisonment, and Cost and Benefits? The Impact of Drug Imprisonment in New
Jersey. He is the recipient of two Media Advocacy Awards from the National Council
on Crime and Delinquency for exceptional research and communications work in support of prison reform. Ziedenberg has served on the California Governor’s Juvenile
Justice Reform Working Group and the Mayor of Washington DC’s transition team
on corrections. He has represented JPI’s research and analysis before the U.S. Congress,
state legislators, city and county councils, and various national and state commissions
considering juvenile and criminal justice reform. Ziedenberg has a Master in Science
from the Columbia University School of Journalism in New York City, and a Bachelor
of Arts from the University of Toronto.

Acknowledgements
This report would not have been possible without the generous support of the Drug
Policy Alliance, the Butler Family Fund, the Open Society Institute, and the Public
Welfare Foundation. The report was edited by Bonita Sennott and designed by Lynn
Riley Design. JPI staff includes Jason Ziedenberg, Amanda Petteruti, Nastassia Walsh,
Laura Jones, LaWanda Johnson, and Debra Glapion.

	

	

Justice Policy Institute	

33

www.justicepolicy.org