Skip navigation

Philadelphia's Crowded, Costly Jails - The Search fo Safe Solutions, PEW, 2011

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Philadelphia’s Crowded,
Costly Jails:
The Search for Safe Solutions

Aerial photo of the Philadelphia Prison System
main campus in Northeast Philadelphia
CREDIT: POLICE AERIAL UNIT

Philadelphia’s Crowded,
Costly Jails:
The Search for Safe Solutions

The Pew Charitable Trusts’ Philadelphia Research Initiative provides timely, impartial
research and analysis on key issues facing Philadelphia for the benefit of the city’s citizens
and leaders. Pew is a nonprofit organization that applies a rigorous, analytical approach
to improve public policy, inform the public and stimulate civic life. For more information,
visit www.pewtrusts.org/philaresearch.

CONTENTS

1.
2.
3.
4.
5.
6.
7.

Foreword

1

Executive Summary

2

The Issue

6

The Philadelphia Prison System: An Overview

10

How Philadelphia’s Jail Population Rose: 1999–2008

13

How Philadelphia’s Jail Population Fell: 2009–2010

19

Decision Points for Admission to Jail

22

What is Being Done to Reduce the Jail Population

29

What More Can be Done to Reduce the Jail Population

32

Notes on Methodology

36

Acknowledgements

38

Foreword >>
FROM THE MOMENT THE PHILADELPHIA RESEARCH INITIATIVE came into existence, taking a hard
look at the Philadelphia Prison System was near the top of our agenda. We knew that the population in
the city’s jails—and the cost to the taxpayers—had been rising year after year with seemingly no end in
sight. We wondered whether it had to be this way.
In the summer of 2009, as we turned our attention to the subject, there were about 9,400 men and
women housed in the prison system, which occupies a vast complex along State Road in Northeast
Philadelphia. On a per capita basis, among the 50 counties and cities with the most inmates, Philadelphia
had the fourth highest inmate population in the country.
Then, something unexpected happened. The inmate population started to decline. And it kept falling
into 2010.
The recent decline tells us something important: The city’s jail population can be controlled when officials
in the criminal justice system work together to make it happen. And if that work is undertaken with caution, creativity and an understanding of the makeup of the jail population, reducing the numbers can be
done while protecting public safety.
Much of the drop in Philadelphia’s inmate population is due to state legislation that has moved one
group of convicted criminals from the city jails to the state prisons. Some of it is due to a reduction in arrests. Measures are being taken to make the criminal justice system more efficient and more innovative
in the ways it deals with incarceration.
This report does a number of things. It analyzes why the city’s jail population rose for most of the last
decade. It studies why the population fell in the past year. It examines what is being done to manage the
population, and it looks at measures in place elsewhere that might be worth considering for Philadelphia.
In doing so, it builds on the work on state prisons and corrections done by our colleagues at the Public
Safety Performance Project of the Pew Center on the States. The Public Safety Performance Project has
helped states such as Texas, Nevada and Kansas advance fiscally sound, data-driven policies and practices
in sentencing and corrections that protect public safety, hold offenders accountable and control corrections costs.
This is not a report about conditions inside the city jails. The focus is on how people come to be in jail,
how long they stay and how they get out.

Philadelphia’s Crowded, Costly Jails: The Search for Safe Solutions is primarily the work of Claire Shubik-Richards, senior associate at the Philadelphia Research Initiative. Our original data analysis, which is
central to this report, was performed by Don Stemen, assistant professor of criminal justice at Loyola
University Chicago.
The Philadelphia Prison System is an integral part of a criminal justice system designed to protect the
lives and property of all Philadelphians. Our purpose in this report is to give policy makers and the public

Larry Eichel
Project Director
The Philadelphia Research Initiative

FOREWORD

the information and perspective necessary to understand the issues connected to managing this system.

1

Executive Summary
Today, the government of the city of Philadelphia spends seven cents out of every tax dollar on holding people
in its jails. That is more than it spends on any other function besides police and human services—and as much
as it spends on the streets and health departments combined. Its spending on jails is nearly as high as that of
Cook County, Illinois, even though Cook County, which includes the city of Chicago, has more than three times
as many residents as Philadelphia. Over the last 10 years, the departmental budget has more than doubled,
reaching $240 million in the current fiscal year. That figure, however, understates the true cost of prison operations, as it does not include employee benefits. With benefits, the number is about $290 million.

This increase in spending has been driven by a rising number

Since mid-2009, however, the population in the prison system

of inmates. As of mid-2008, the most recent date for which

has fallen steadily; the average daily population stood at

comparative numbers are available, Philadelphia had the

8,464 for March 2010, down 13 percent from the peak. And

fourth-highest jail population on a per capita basis among the

some city officials, including Prison Commissioner Louis Giorla

cities and counties with the nation’s 50 largest jail populations.

and Deputy Mayor for Public Safety Everett Gillison, say that

From 1999 through 2008, the Philadelphia Prison System saw
its average daily inmate count climb by 45 percent, peaking

2

without jeopardizing public safety.

largest such increases in the country, and it came at a time

These developments in the inmate population—both the

when jail populations in the nation’s two largest jurisdictions,

long-term rise and short-term fall—generally have not tracked

New York City and Los Angeles County, were declining.

the crime rate; for most of the past decade, as the inmate

PHILADELPHIA PRISON SYSTEM BUDGET

$241

$240

2009

2010 *

250

$224
200

$131

$142

$184

$193

$194

2006

$172

150

$155

$118

0

Source: Philadelphia Prison System; Mayor’s Budget
Note: Does not include employee benefits.
* Estimated FY2010 budget from the Mayor’s proposed FY2011-2015 Five-Year Financial Plan
** Projected FY2011 budget from the Mayor’s proposed FY2011-2015 Five-Year Financial Plan

2008

2007

2004

2003

2002

2001

2000

1999

50

2011 **

100

$234

$208

2005

[IN MILLIONS]

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

at 9,787 for the month of January 2009. This was one of the

they believe that the population can be substantially lowered

Riverside Correctional Facility
CREDIT: ROBERT LECONEY

population numbers were rising, the arrest numbers often
were declining. Rather, the shifts in the population are related
primarily to changes in procedures, legislation and policies involving the police, the courts and the various elements of the
criminal justice system.
To a large extent, the evidence in this report indicates that the
size of the population of the Philadelphia Prison System is
within the power of policy makers to control—without compromising the fight against crime. It suggests that Philadelphia
can have fewer people in jail, save money and be no less safe.
Some of the factors that dictate the size of the inmate population have also contributed to other problems in the city’s
criminal justice system. The same delays that can result in the
guilty going free can keep the accused in custody longer than
is necessary.

the total annual cost of the system and dividing it by the number of inmate-days consumed in a year—does not represent
how much the system saves by reducing the daily population
by one. Because most jail costs are fixed, the savings is only
about $20 for the first day, when clothing is distributed and
medical tests are conducted, and less for every day thereafter,
our analysis shows. The savings are higher for inmates who
need costly medical or mental-health treatment.
Despite its name, the Philadelphia Prison System is actually a
complex of jails; prisons are usually thought of as state and
federal institutions where convicted criminals serve out sentences of substantial length. As such, Philadelphia’s inmate
population consists of three groups: accused individuals being
held pretrial, convicts serving out sentences of less than two
years, and probationers and parolees who have violated the
terms of their community supervision.

Leaders of Philadelphia’s criminal justice system, motivated in
ernment, have been working together for the last few years to
reduce the inmate population, largely under the auspices of the
city’s Criminal Justice Advisory Board. As the latest numbers indicate, they have had a measure of success. They see no contradiction between a lower jail population and safe streets.
The recent population drop has allowed the city to budget
about $15 million less for the prison system for the year beginning July 1, 2010 than for the previous year. More substantial savings will be realized if the population continues to fall.
That would allow for the closure of individual facilities and reductions in the size of the system’s workforce.
It is often said that it costs $95 to keep someone in a Philadelphia jail for a day. But that figure—which is derived by taking

Behind the Long Rise and Recent
Decline of the Jail Population
Our data analysis shows that the rise in the inmate population
from 1999 through 2008 had little to do with the convicted
criminals and mostly to do with the pretrial population:
• During that decade, the percentage of bed-days in the
Philadelphia jails consumed by pretrial inmates on an annual basis rose from 44 percent of the total to 57 percent.
• Most of the increase in pretrial admissions came from
individuals charged with misdemeanors, which range
from disorderly conduct and loitering to simple assault
and some types of theft of up to $2,000. The number of
inmates admitted pretrial on felony charges remained
relatively constant.

EXECUTIVE SUMMARY

part by the pressure to control spending throughout city gov-

3

• In recent years, fewer individuals have been released
without bail than in years past; in cases where bail has
been set, the average amount of the bail has risen.
These two factors, higher bail and more people being ordered to post it, have driven up the number of people
who spend time in jail before the resolution of their
cases. The guidelines laid out for the magistrates who
set bail have not changed, but the magistrates are following the guidelines in only about half of the cases.

doing their time in a state prison at state expense. As a
result, several hundred inmates were moved from the
city to the Pennsylvania Department of Corrections. Others are now going to state prison upon sentencing rather
than staying in the city jails.

• In the past few years, several other factors have contributed to the rising pretrial numbers. The police have
made more arrests for drug possession. And changes in
state law have led to more arrests for drunk driving and
to higher bail being set for individuals charged with carrying a firearm without a license.

are doing so now.

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

• Compared to other urban jurisdictions, Philadelphia
had—and still has—a relatively large group of inmates,
15 percent, who stayed in jail for 120 days or more awaiting trial. This is due primarily to the length of the court
process. Many pretrial stays are less than two weeks.

4

In addition, the Philadelphia Prison System experienced a
steady increase in the number of inmates jailed for violating
the terms of their probation (community supervision in lieu
of incarceration) and parole (community supervision after incarceration):
• The number of individuals admitted for such violations
went from a low of 3,101 in 2000 to 5,900 in 2008, a rise
of 90 percent.
• Those individuals were staying in jail longer. Their average length of stay, which was 49 days in 2000, rose as
high as 73 days in 2007. One reason is that the court
process for hearing a violation can be cumbersome and
lengthy.
The decline in the inmate population started early in 2009 and
accelerated as the year went on. Several factors contributed
to the drop:
• Most important, there was a sharp decrease in the
number of sentenced inmates in the city jails due largely
to a change in state law. In the fall of 2008, the state legislature ended a practice that had given individuals with
sentences of two to five years the option of staying in the
Philadelphia Prison System, at city expense, rather than

• In addition, crime declined in Philadelphia in 2009, and
arrests were down by 11 percent. Admissions declined
by 5 percent. While the jail population and the arrest totals have not always moved in the same direction, they

• A number of changes related to probation appear to
have played a role in reducing the city’s jail population. In
2009, for the first time in a decade, new cases accepted
by the Adult Probation and Parole Department leveled off
and, with it, the number of people incarcerated for violations. At the same time, the department’s reorganization
made it easier for individuals deemed to be at low risk of
committing major crimes to comply with the terms of
their supervision. And the Philadelphia court system inaugurated a specialty court to hear violations of probation
and parole more quickly.

Strategies to Reduce the
Jail Population Safely
In Philadelphia’s criminal justice system, there are six decision
points that determine whether an individual will be sent to jail.
They are arrest, charging by the district attorney’s office, preliminary arraignment, disposition, probation violations and
outstanding warrants.
For policy makers looking to control the jail population, there
is no one decision point that holds the key, nor is there any
single solution. Their challenge is to build on progress already
made while protecting public safety; crime is far and away the
top local concern of city residents, according to polls done by
the Philadelphia Research Initiative in 2009 and 2010.
Our analysis shows that the size of the jail population in
Philadelphia is driven largely by inmates held prior to trial, not
by convicted criminals. So the greatest advances are likely to
be achieved by focusing on the parts of the criminal justice
system that impact that group.

Among the steps being taken to streamline the criminal justice
system—and reduce the jail population—are these:
• District Attorney R. Seth Williams is restructuring his office’s charging unit with the intent of weeding out weak
cases early. Although reducing the jail population is not the
primary goal of this change, it could contribute to that end.
• Through the use of videoconferencing and other measures, the court system is expanding the use of “crash
court,” an expedited-plea process in Municipal Court for
people accused of lower-level misdemeanors and being
held pretrial. One purpose of crash court is to avoid situations in which the indigent, if unable to post even modest
bail, wind up spending more time in jail pretrial than they
would have if convicted.

• Devising programs—as New York City and Santa Cruz
County, California, have—to allow defendants not in custody to deal with missed court dates and probation appointments quickly and effectively without the threat of
swift incarceration.
• Taking some of the savings from reducing the jail population and putting the money into programs aimed at further reducing the population—with the goal of producing
additional savings in the years to come. Federal legislation
to encourage such investment is pending.
• Tracking the performance of the criminal justice system
to give officials a clear sense of the impact of various reforms and potential reforms—on the jail population and
other elements of the system.

• City officials have begun the process of planning what is

This is a time of great ferment in the Philadelphia criminal jus-

known as a day reporting center as an alternative to jail.

tice system. A series of budget crises has placed renewed em-

Exactly who would be eligible to use the center remains

phasis on reducing the jail population, saving money and

to be seen.

making the entire system more efficient. Local criminal justice

As part of this study, we examined measures taken to address

stakeholders, having made some progress on both fronts, are
planning additional reforms. More are likely to come from the

the size of the jail populations in other jurisdictions. From that

Pennsylvania Supreme Court, which has created a panel to

research, a number of policy options emerged. They include

look into the operations of the criminal courts in Philadelphia.

the following:

This report is intended to help those policy makers and inter-

• Expanding the options for diverting troubled, low-level
offenders out of the court system so that their addictions
or mental health problems can be addressed in a more appropriate setting, as happens in Bexar County, Texas.

ested citizens chart an informed course for the future.

• Revitalizing Philadelphia’s often-ignored bail guidelines
with the goal of identifying defendants accused of relatively minor crimes who are at low-risk of failing to appear
in court and allowing as many of them as possible to stay
out of jail pending trial. Guidelines in New York City and
Montgomery County, Maryland, appear to be working.

• Developing new responses to probation violations so
that so-called technical infractions, such as missing a meeting with a probation officer, can be punished without sending the individual back to jail. The state of Georgia has
done this.

EXECUTIVE SUMMARY

• Expanding the range of pretrial services so that more defendants can remain in the community, as has been done
in Washington, D.C.

5

1

The Issue
I don’t want to spend more money on prisons, but I have to.

—Mayor Michael A. Nutter

The mayor of Philadelphia made that statement in November 2009. Ten months earlier, in January 2009, the
population in the Philadelphia Prison System had peaked at a monthly average of 9,787 after rising steadily and
seemingly inexorably for years. But by the time Nutter spoke those words, the numbers had started to drop, and
not by insignificant amounts. And the decline continued into 2010. In March, there was an average of 8,464 people in the system, down 13 percent from the high point.

The decline in the number of inmates caused prison spending,

than with individuals charged with offenses such as drug possession

which had more than doubled in the previous decade, to level off

and drunk driving. And the long-term rise in population had less to

at an estimated $240 million for fiscal year 2010—about $9 million

do with convicted criminals than accused individuals being held pre-

below the figure officials had budgeted. It also allowed city officials

trial.

to do something they had not done in a very long time: allocate
less money for the prison system for the next fiscal year than for the

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

current one.

6

The factors at the heart of the long-term growth in the population
of the Philadelphia Prison System are some of the same ones that
have contributed to other problems in the city’s criminal justice sys-

Neither the long-term rise in the prison population nor the recent

tem, including how long it takes to resolve cases. Prolonged court

fall is an inexplicable phenomenon. Nor is either one directly con-

proceedings can result in dangerous criminals going free without

nected to the crime rate in Philadelphia, although the drop in ar-

ever facing the charges against them. And they can keep nonviolent

rests starting last year is a factor in the recent population decline.

individuals in custody for months, awaiting resolution of their cases.

Rather, the rise and fall of the population are the result of administrative, legislative and policy decisions made at various levels of the
criminal justice system.

None of this is news to the criminal justice stakeholders. Earlier this
year, District Attorney R. Seth Williams said that the criminal justice
system in general—and the city’s prisons in particular—have been

The overriding lesson from these developments is this: to a substan-

burdened with “junk” cases that consume resources better allo-

tial degree, the size of Philadelphia’s prison population is within the

cated to more serious ones. Said Ellen T. Greenlee, the city’s chief

control of public officials.

public defender: “For justice and for economic reasons, we can’t
continue to jail everyone we’re now jailing. We have to find a better

DEFINING THE SYSTEM
Although Philadelphia calls its detention facilities “prisons,” they
are, in fact, “jails,” as the term is commonly used, meaning local institutions for housing individuals awaiting trial or those convicted
of relatively minor crimes. The word “prison” usually is reserved for
state institutions for the confinement of people convicted of more
serious crimes.

This report looks at the size and nature of the inmate population that
drives the city’s spending on its jails. Our research shows that the

way.” Pamela P. Dembe, president judge of the Philadelphia Common Pleas Court, put it another way: “Incarceration should be used
to keep the monsters away.”
Mayor Nutter’s point-person on criminal justice, Deputy Mayor for
Public Safety Everett A. Gillison, agrees and has said that the prison
population could be lowered further, to 6,500, without jeopardizing
public safety. Prisons Commissioner Louis Giorla said that he looks
forward to a day when the facilities he runs will have 6,000 inmates.
Such a decline would result in substantial savings for the city’s taxpayers.

changes in the inmate population over the last 10 years, first up and

What makes the population decrease of the past year particularly

now down, have had less to do with murderers, rapists and robbers

notable is the magnitude and persistence of the population in-

While the jail population rose for most of the past decade, the number of adult arrests fell slightly.
1.2 ADULT ARRESTS IN PHILADELPHIA

66,097

74,444

72,080

2007

66,693

71,420

6,908

7

70 –

2006

7,397

68,158

7,574

65,955

7,916

7,818

69,499

8

80 –

77,281

8,562

2001

8,602
8,289

76,750

9

90 –

2000

9,321

8,892

76,293

9,399

100 –

1999

10

NUMBER OF ARRESTS [IN THOUSANDS]

60 –

6,578

50 –

2009

2008

2005

40 –

2004

5

2003

2010*

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

6

2002

AVERAGE DAILY POPULATION [IN THOUSANDS]

1.1 AVERAGE DAILY POPULATION, PHILADELPHIA
PRISON SYSTEM 1999 – 2009

Source: Philadelphia Prison System

creases throughout the past decade. From 1999 to 2009, the average daily population of the Philadelphia Prison System grew from
6,578 to 9,321, an increase of 45 percent. See Figure 1.1.
The recent decline in the population is only barely reflected in the

Source: Pennsylvania Uniform Crime Reporting System

The National Perspective
How does what is happening in the Philadelphia jails compare to
what is happening in other jurisdictions around the country?

average daily population for 2009 because the big fall-off did not

One of the most telling ways to look at Philadelphia in a national

start until last autumn. For 2009 as a whole, the average daily popu-

context is by jail population per capita. Among the 50 jurisdictions

lation decreased only a small amount, from 9,399 to 9,321. The

in the country with the largest jail populations, Philadelphia had

population for 2010 is on track to be much lower.

the fourth highest rate of incarceration as of mid-2008, the last date

The decade-long population growth that peaked early in 2009 took
place at a time when the number of people arrested in Philadelphia
was trending lower, fluctuating between a high of approximately
77,281 in 2001 and a low of 65,955 in 2003. See Figure 1.2.
As Figures 1.1 and 1.2 indicate, there appears to be no direct relationship between the size of the population in the prison system
and the crime rate, as reflected by total arrests.

for which data from the federal Bureau of Justice Statistics are available. As shown in Figure 1.3, Philadelphia had 5.72 individuals behind bars for every 1,000 residents; the highest figure for any large
jurisdiction was New Orleans at 7.76.
In most places, jails are county institutions, meaning that the areas
they cover—unlike the area covered by the Philadelphia Prison
System—include suburbanites as well as city residents. This makes
precise city-to-city comparisons impossible and, to some degree,

Even with the reduced inmate population this year, the Philadelphia

overstates Philadelphia’s incarceration rate compared to urban coun-

Prison System remains the third-biggest department in the city

ties that also have large numbers of suburbanites.

budget, trailing only human services and police. The city plans to
spend $51 million more this year on its prison system than on fire
protection. It spends about as much on prisons as for the streets
and health departments combined. The growth of the system’s
budget is shown in the figure on page 4.
Actually, the current $240 million figure understates the true cost of
the prison system, since Philadelphia places the healthcare and
pension benefits of all city employees in a separate category. When

Even so, and taking into account the recent decline in the population in Philadelphia jails, the city’s population number on a per
capita basis is among the highest in the country.
Another way to compare Philadelphia to other jurisdictions is in terms
of spending. Many metropolitan jurisdictions have struggled with rising jail spending in recent years, but the growth in Philadelphia in the
past decade has been dramatic. See Figure 1.4.

those benefits are included, the total cost rises to about $290 mil-

Consider that last year the city spent almost as much on incarcera-

lion, slightly more than 7 percent of the city’s general fund budget.

tion as Cook County, which has three-and-a-half times as many resi-

Chapter 1 THE ISSUE

Note: *2010 shows average daily population from January through March.

7

dents as Philadelphia and includes a city, Chicago, which has twice
1.3 JAIL POPULATION PER 1,000 RESIDENTS IN THE
50 LOCAL JURISDICTIONS WITH THE LARGEST
JAIL POPULATIONS

as many people as does Philadelphia. In the 2009 fiscal year,
Philadelphia spent $241 million, Cook County $255 million.
And while Philadelphia has experienced one of the biggest in-

NEW ORLEANS 7.76
BALTIMORE (CITY) 6.28
SHELBY COUNTY, TN 6.27
PHILADELPHIA 5.72
DAVIDSON COUNTY, TN 5.62

creases in jail expenditures over the last decade among the jurisdictions studied, it also experienced one of the biggest increases in
jail population. As shown in Figure 1.5, other jurisdictions had large
increases, including Allegheny County, which includes the city of
Pittsburgh, and Suffolk County, which includes the city of Boston. At

WASHINGTON, DC 5.02

the same time, the nation’s two largest local jurisdictions, New York

JACKSONVILLE, FL 4.54

City and Los Angeles County, have seen their jail numbers decline.

LEE COUNTY, FL 4.36
POLK COUNTY, FL 4.23
GWINNETT COUNTY, GA 4.19
BERNADILLO COUNTY, NM 4.10

What, then, is behind the rising and falling number of inmates in
Philadelphia’s jails? The answer lies in a series of policy and administrative choices:

DENVER COUNTY, CO 4.01

• Who should be detained prior to trial and who should be al-

ORANGE COUNTY, FL 3.99

lowed to remain in the community while his or her case proceeds?

DEKALB COUNTY, GA 3.92
PINELLAS COUNTY, FL 3.91
COBB COUNTY, GA 3.66
FRESNO COUNTY, CA 3.38

• How long does it take to try a case?
• Are other sanctions besides a trip to jail used to punish those
who break the rules governing their probation or parole?

HILLSBOROUGH COUNTY, FL 3.35
SUFFOLK COUNTY, MA 3.29
SACRAMENTO COUNTY, CA 3.29
MILWAUKEE COUNTY, WI 3.18
BROWARD COUNTY, FL 3.14
ESSEX COUNTY, NJ 3.06
EL PASO COUNTY, TX 3.01
KERN COUNTY, CA 2.98
ALAMEDA COUNTY, CA 2.97
MECKLENBERG COUNTY, NC 2.92
MIAMI-DADE COUNTY, FL 2.84
SAN BERNARDINO COUNTY, CA 2.79

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

FULTON COUNTY, GA 2.75

8

DALLAS COUNTY, TX 2.68
TRAVIS COUNTY, TX 2.67

• Which convicted inmates serve out their sentences in the local
jail and which are sent to state prisons?
The answers to these and similar questions determine how many
people are jailed in Philadelphia, how long they stay and, consequently, how much money the city spends on incarceration.
Said Michael Jacobson, New York City’s commissioner of corrections and probation from 1995 to 1998: “Once you look at any of
these [local jail] systems, you see there are people who don’t have
to be incarcerated who are, and there are people who are being
jailed for a certain amount of time when they could just as safely be
staying half or two-thirds that amount of time. The people who work
in these systems know this. They know their populations are not
solely linked to the crime rate. They’re linked to policy decisions.”

ALLEGHENY COUNTY, PA 2.66
SANTA CLARA COUNTY, CA 2.66
MARION COUNTY, IN 2.65
HARRIS COUNTY, TX 2.51
BEXAR COUNTY, TX 2.50
MARICOPA COUNTY, AZ 2.34
PALM BEACH COUNTY, FL 2.28
FRANKLIN COUNTY, OH 2.16
LOS ANGELES COUNTY, CA 2.03
ORANGE COUNTY, CA 2.00

COMPARISON JURISDICTIONS
In deciding which jurisdictions to use for statistical comparison, we
chose places with one of four factors in common with Philadelphia.

TARRANT COUNTY, TX 2.00

1. POPULATION. We looked at New York City and the counties

COOK COUNTY, IL 1.88

containing the four other cities more populous than Philadel-

SAN DIEGO COUNTY, CA 1.78

phia—Los Angeles County, Cook County (Chicago), Harris

RIVERSIDE COUNTY, CA 1.69
NEW YORK CITY 1.66
CLARK COUNTY, NV 1.66
KING COUNTY, WA 1.41
WAYNE COUNTY, MI 1.20

County (Houston) and Maricopa County (Phoenix).
2. CRIME RATE. Philadelphia has a relatively high crime rate. Fulton
County (Atlanta) and Wayne County (Detroit) do, too.
3. DENSE, NORTHEAST LANDSCAPE. Philadelphia has a lot in
common with Baltimore and Suffolk County (Boston)

Note: To calculate these numbers, we used the average daily populations for local
jails for the 12 months ending June 30, 2008, as reported by the U.S. Bureau of
Justice Statistics, and the population estimates from the U.S. Census Bureau for
July 1, 2008. The 2008 jail numbers are the latest available.

4. LEGAL FRAMEWORK. Allegheny County (Pittsburgh) functions
under the same legal system as Philadelphia.

1.4 PERCENTAGE CHANGE IN JAIL EXPENDITURES
1999 – 2009
MARICOPA COUNTY (PHOENIX)
PHILADELPHIA
BALTIMORE CITY

164%

1.5 PERCENTAGE CHANGE IN JAIL POPULATIONS
1998 – 2008
ALLEGHENY COUNTY (PITTSBURGH) 59%

105%

PHILADELPHIA 53%

83%

SUFFOLK COUNTY (BOSTON) 39%

HARRIS COUNTY (HOUSTON)

MARICOPA COUNTY (PHOENIX) 34%

73%

ALLEGHENY COUNTY (PITTSBURGH)

HARRIS COUNTY (HOUSTON) 29%
54%

COOK COUNTY (CHICAGO) 6%
SUFFOLK COUNTY (BOSTON)

48%

BALTIMORE CITY 6%
WAYNE COUNTY (DETROIT)

44%

FULTON COUNTY (ATLANTA)
COOK COUNTY (CHICAGO)
NEW YORK CITY*
0

50

-6% LOS ANGELES COUNTY

42%

-15% WAYNE COUNTY (DETROIT)

40%

-21% NEW YORK CITY
-35% FULTON COUNTY (ATLANTA)

23%
100

150

Source: Allegheny County Prison System, Maryland Department of Corrections,
Cook County Hospital & Health System and Cook County Department of Management and Budget, Fulton County Sheriff’s Department, Harris County Sheriff’s
Department, Maricopa County Sheriff’s Department, New York City Office of Management and Budget, Philadelphia Prison System, Suffolk County Sheriff’s Department, and Wayne County Communications Department.

-40

-30

-20

-10

0

10

20

30

40

50

Source: U.S. Department of Justice, Bureau of Justice Statistics, Jail Inmates at Midyear,
Average Daily Population 1998 and 2008.

Note: For all jurisdictions, inmate healthcare was included in the calculation. For
Baltimore City, Fulton County, Harris County, Philadelphia, and Suffolk County, this
was included in the county jail budget. In Allegheny County, inmate health care cost
was obtained from Allegheny’s Office of Budget and Finance; in Cook County from
Cook County Health and Hospitals System; in Maricopa County from Maricopa’s
Office of Management and Budget; in New York from the New York City Office of
Management and Budget; and in Wayne County from Wayne County Communications.
The following jurisdictions do not include employee benefits in their jail budgets:
Cook County, New York City, Philadelphia, and Wayne County.
Los Angeles County is not included because the County Sheriff’s Department was
unable to provide FY1999 expenditures.
* New York City numbers are based on 2000 expenditures.

Motivated by lawsuits challenging the overcrowding in the Philadel-

population declined from 2009 into 2010.

phia Prison System and by the city’s budget crisis, city officials have

For those who run the criminal justice system, the challenge now is

been working for the last several years to reduce the jail population.

to figure out how to make the current decline a lasting one. Said

In recent months, under the auspices of the city’s Criminal Justice

Prisons Commissioner Giorla, “There have been downturns [in the

Advisory Board, leaders of the criminal justice system have been re-

prison population] before … But this is the first time it’s been done

examining the policy and administrative choices that contributed to

through cooperation and coordinated efforts of the whole criminal

the population increases of the past.

justice system. So we believe we can sustain it.”

Much of this work has focused on ensuring that convicted inmates

The stakes are high. “You could drop [the prison population] by 10,

eligible to serve their sentences in state-run facilities are identified

and if it’s the wrong 10, the city is going to be more dangerous,”

and transferred out of the Philadelphia Prison System—a move that

said Police Commissioner Charles H. Ramsey. “You could also prob-

shifts rather than reduces costs and does not involve moving in-

ably drop the prison population by 1,000 without impacting public

mates out of custody and into the community. To a lesser degree,

safety—but what matters is who those 1,000 people are and what

officials have taken steps to streamline court processes in order to

we’re doing with them.”

shorten the stays of individuals incarcerated for probation and parole violations.

Succeeding will require the consensus of the criminal justice community, the support of elected officials, the buy-in of a public that

These initiatives have required cooperation and coordinated plan-

consistently lists crime as its top local concern, and a willingness

ning among the courts, the prosecutors, the defense attorneys and

to do things in ways that have not been done before in Philadel-

the prisons. As a result of their work and other factors, the city’s jail

phia.

Chapter 1 THE ISSUE

Turning the Corner

9

The Philadelphia Prison
System: An Overview

2

The Philadelphia Prison System consists of six facilities on a sprawling main campus on State Road in Northeast
Philadelphia, as well as several other smaller, privately run facilities throughout the city.
The buildings on State Road were designed to accommodate roughly 6,500 inmates—the size of the 1999 jail
population.1 By adding dormitory space to areas built for common or administrative use, the prison system has
increased its rated capacity to about 8,500.2 Despite the added beds, Philadelphia’s jails, like those in Baltimore,
Houston, Phoenix and Chicago, have struggled to operate within this expanded capacity; in 2009, the jail population in Philadelphia fluctuated between 100 and 112 percent of rated capacity.3

The Curran-Fromhold Correctional Facility is the system’s intake and

Lehigh County remains in place. Other metropolitan jurisdic-

administrative building; it opened in 1995 and houses approximately

tions—including Fulton County (Atlanta) and Harris County

3,000 inmates. There are three other jails for men: the Philadelphia

(Houston)—have similar arrangements.

Industrial Correctional Center (opened in 1986; 1,100 inmates), the
Detention Center (opened in 1964; 1,300 inmates) and the House
of Correction (opened in 1925; 1,500 inmates). The jail for women,
the Riverside Correctional Facility, opened in 2004, houses approximately 750 inmates. The Alternative and Special Detention Divi-

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

sion—a group of buildings and trailers on State Road and a number

10

of small facilities throughout the city—houses up to 1,000 workrelease, weekend-stay and other minimum-security inmates.
To manage overflow population, Philadelphia, like other cities,
employs two main strategies:4
• Philadelphia pays other counties to house inmates in their jails.
In fiscal year 2009, the city had contracts with Lehigh County in
Pennsylvania and Passaic and Monmouth Counties in New Jersey, paying Lehigh $90 per inmate per day, Passaic $88 and Monmouth $105. When the population in the Philadelphia jails
started to fall, the city stopped sending inmates to New Jersey,
saving roughly $4.9 million on an annual basis.5 The contract with

1.

• In some facilities, Philadelphia assigns a third inmate to cells
that are rated for two, a practice known as triple celling.6 At the
Detention Center, where inmates live in open dorms rather than
separate cells, extra bunks have been added to housing areas
that were already considered full; other jurisdictions employ similar measures. These practices are the subject of a civil rights lawsuit against the city.
Although putting additional inmates into existing space is not as
expensive as sending them to other counties, it does result in increased overtime pay for guards—about $1.5 million per year in
Philadelphia, according to local officials. In addition, the litigation
resulting from these conditions has a high price; according to estimates provided by the city, the annual cost of dealing with the current cases and settling prior ones is about $1.2 million.7 All of the
costs associated with overcrowding give city officials a financial incentive to reduce the population in the Philadelphia Prison System
as much as possible, assuming they can protect the public at the
same time.

Talking about jail capacity is less straightforward than it might seem. There is design capacity, which is the number of people the jail was designed to house. This can be
different from rated capacity—the number of people a rating agency has certified the jail to house—especially if, as in Philadelphia, additional beds have been added to
spaces not originally designed as dormitories. And there is operational capacity, which is the number of people the jail can accommodate while keeping a sufficient number of beds unoccupied in order to allow inmate movement and repairs.

2.

Todd D. Minton and William J. Sabol, Jail Inmates at Mid-Year 2008 (U.S. Department of Justice, Bureau of Justice Statistics, 2009).

3.

Ibid.

4.

In addition to these two strategies, starting in 2006 Philadelphia retrofitted the gymnasium area at the old Holmesburg jail to serve as a temporary intake center. Holmesburg, site of a 1973 riot in which two correctional officers were killed, was closed in 1995 as the result of federal litigation related to prison conditions. The temporary intake area was closed in September 2009 when the population decreased to 9,300.

5.

In FY2008, Philadelphia paid Monmouth County approximately $4,400,000 and Passaic approximately $500,000 for housing Philadelphia inmates.

6.

The House of Corrections, Philadelphia’s oldest facility, was designed to house one inmate per cell. For decades, two inmates per cell have been the norm in this facility,
and some cells currently have three inmates.

7.

This figure combines settlement costs (indemnities) paid by the prison system for civil rights cases ($925,000) and the average annual cost of hiring outside counsel in
these cases. According to Michael Resnick, chief of staff to the deputy mayor for public safety, the city has paid a combined $820,000 over the last three years for outside
counsel in prison-related civil rights litigation.

Who is in Jail?
There are several ways to answer the question of who is in Philadelphia’s jails. One is through demographics. On a typical day last
year, as shown in Figure 2.1, 66 percent of the inmates in the
Philadelphia Prison System were African American men.

On a typical day last year, 82 percent of Philadelphia’s jail inmates

Inmates also were young: as shown in Figure 2.2, 48 percent of
them were between the ages of 18 and 29. Note that 18 percent of
Philadelphia’s overall population is in this age group.8

or second-degree misdemeanors (crimes with maximum penalties

were charged with or convicted of felonies; in Pennsylvania, felonies
include all crimes in which the maximum sentence is greater than
five years. Sixteen percent were charged with or convicted of firstbetween two and five years). The other 2 percent were charged with
or convicted of summary offenses, such as minor shoplifting and
disorderly conduct, or third-degree misdemeanors, such as prostitu-

A second way to describe the jail population is by how inmates
come to be incarcerated. There are three main pathways to jail:

tion, loitering and persistent disorderly conduct.9 These numbers
are broken down by type of inmate—pretrial, violator or sen-

• Individuals are held prior to trial out of concern that they will
not show up for court if they are allowed to remain at liberty.
This is the pretrial population.

tenced—in Figure 2.4.

• Individuals are convicted of crimes and sentenced to a period
of incarceration. This is the sentenced population.

MENTAL HEALTH, SUBSTANCE ABUSE AND HOMELESSNESS
Another way to examine Philadelphia’s jail population is by the con-

• Individuals already convicted of a crime and sentenced to a
period of supervision in the community are incarcerated as a result of an alleged or proven violation of the conditions of their
community supervision. This is the violator population.

ditions that have contributed to inmates’ criminal behavior. Prior
studies estimate that:
• About 30 percent of inmates in the Philadelphia Prisons experience mental illness;

On a typical day last year, as shown in Figure 2.3, less than a quarter
of the inmates in the Philadelphia jails were convicted criminals
serving sentences. More than half were being held pretrial, and almost all of the rest were violators.

• About 42 percent report having abused drugs and alcohol;10
• And a third expects to be homeless at the time of release.11
For some, said Bruce Herdman, chief of medical operations for the
Philadelphia Prison System, “the prison has by default become the

A third way to look at who is in jail is by the severity of the charges
inmates have faced or will face.

2.1 DAILY JAIL POPULATION BY
RACE AND GENDER

primary source of social, medical and psychiatric care.”

2.2 DAILY JAIL POPULATION BY
AGE

2.3 DAILY JAIL POPULATION BY
PATHWAY

UNDER 18
OVER 50

1%

7%
BLACK MALE
HISPANIC
FEMALE

1%

12%

SENTENCED

18–24

20%

23%

30%

PRETRIAL

WHITE
FEMALE

7%
11%

40–49

57%

3%
BLACK
FEMALE

30–39

24%

18%
25–29

18%
WHITE MALE

PROBATION/ PAROLE
VIOLATORS

HISPANIC MALE

OTHER

2%

Source: Philadelphia Prison System
Note: Daily Population on June 30, 2009. In Figure 2.3, “other” represents individuals held for other jurisdictions.

8.

U.S. Census Bureau, 2008 American Community Survey 1-Year Estimates (2008), Table B01001 (accessed April 12, 2010).

9.

In most other states, a felony is defined as a crime with a maximum sentence of one year or more. So what is a first- or second-degree misdemeanor in Pennsylvania is a
felony in most other jurisdictions. And a misdemeanor in most jurisdictions is classified as a third-degree misdemeanor in Pennsylvania.

10. John S. Goldkamp, E. Rely Vilcicã, Doris Weiland, Wang Ke, Confinement and the Justice Process in Philadelphia: Its Features and Implications for Planning, Crime and
Justice Research Center, Temple University (2006) 72.
11. Philadelphia Prison System, Homelessness Among PPS Inmates (2004), provided by the Philadelphia Prison System (unpublished memoranda on file with the author).

Chapter 2 THE PHILADELPHIA PRISON SYSTEM: AN OVERVIEW

66%

11

A fourth way of looking at the population is in terms of the substan-

THE CHALLENGE OF MULTIPLE HOLDS

tive nature of the most serious or “top” charges lodged against
them.

Some inmates are in jail for more than one reason. They may be held

Thirty percent of all inmates had an offense against a person—such

on bail for a new arrest combined with a probation violation and/or

as assault, armed robbery, harassment, rape or murder—as their

an outstanding bench warrant. Criminal-justice stakeholders refer

top charge. Twenty-eight percent were charged with property

to such inmates as having “multiple holds.” And because there are

offenses including theft, robbery and arson, and 25 percent with

multiple reasons for keeping them in jail, it can require multiple actions to get them out in a timely fashion, even if none of those rea-

drug offenses. These numbers are broken down by type of inmate

sons on its own would merit a prolonged stay in custody.

in Figure 2.5.

2.4 DAILY JAIL POPULATION BY CHARGE SEVERITY
PRETRIAL

PROBATION/PAROLE VIOLATOR

SENTENCED

FELONY

FELONY

FELONY

90%

74%

68%

9%

FIRST- AND
SECOND- DEGREE
MISDEMEANOR

FIRST- AND
SECOND-DEGREE
MISDEMEANOR

FIRST- AND
SECOND-DEGREE
MISDEMEANOR

23%

30%

SUMMARY OFFENSE/
THIRD- DEGREE MISDEMEANOR

SUMMARY OFFENSE/
THIRD- DEGREE MISDEMEANOR

SUMMARY OFFENSE/
THIRD- DEGREE MISDEMEANOR

1%

3%

2%

Source: Philadelphia Prison System
Note: Daily Population on June 30, 2009.

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

Pretrial charge is defined as the top charge the inmate was facing on June 30, 2009. Violator charge is defined as the top charge the inmate was convicted of in the case that resulted
in a sentence of probation or a period on parole.

12

2.5 DAILY JAIL POPULATION BY CHARGE TYPE
PRETRIAL

PROBATION/PAROLE VIOLATOR

PROPERTY

29%

OTHER

SENTENCED

PERSON

PROPERTY

PERSON

PROPERTY

16%

31%

17%

27%

6%
OTHER

PERSON

40%

9%

VEHICLE

3%
VEHICLE
DRUGS

19%

9%

WEAPONS

7%

DRUGS

DRUGS

37%

WEAPONS

OTHER

6%

30%

8%
VEHICLE

WEAPONS

2%

4%

Source: Philadelphia Prison System
Note: Daily Population on June 30, 2009.
Person offenses include all offenses listed under Offenses Involving Danger to the Person in the criminal code, such as assault, neglect of a care-dependant person, rape, and homicide.
Property offenses include all offenses listed under Offenses against Property in the criminal code, such as theft, vandalism, trespass, and burglary.
Drug offenses include all offenses listed under Title 35, Chapter 6 of the Penn. Statutes such as purchase, manufacturing, and sale of a controlled substance.
Vehicle offenses include offenses such as failure to properly insure a vehicle, DUI, and vehicular homicide.
Weapons offenses include offenses such as possession of a firearm by a former convict and carrying a firearm without a license.
Other offenses include all other offenses listed under 18 Pa.C.S.A. or other section of the Penn. Statutes that were not included in other offense categories such as bribery, perjury,
resisting arrest, obstruction of justice, disorderly conduct, public drunkenness, and prostitution.

3

How Philadelphia’s Jail
Population Rose: 1999–2008

To understand what caused the population to increase over the past decade—and what has caused it to go
down in the past year—one needs to look at the population in another way. The size of the jail population is a
function of the number of inmates admitted and the length of time they stay: the number of bed-days consumed.
One inmate who stays one day takes up one bed-day. Another inmate who stays 50 days takes up 50 bed-days
and has a much larger impact on the prison’s budget. In the past decade, as Figure 3.1 shows, the annual number
of bed-days consumed in the Philadelphia Prison System has risen 38 percent.

In 2009, the total number of bed-days was 3,310,991, up from

from the profile of inmates in the prison system on any one day. For

2,400,970 in 1999. This increase was caused by a rise in the number

example, 39 percent of all pretrial admissions last year were for in-

of jail admissions combined with the relatively long lengths of stay ex-

mates charged with misdemeanors, but only 10 percent of the pre-

perienced by many inmates.

trial inmates jailed on a given day were facing such charges.

Over that period, our data analysis shows that admissions to the

The rise in total bed-days was caused by the pretrial and violator

Philadelphia jails grew 27 percent, from 30,599 to 38,890. See Figure

populations and was offset somewhat by a decrease by the sen-

3.2. This increase was due largely to increases in pretrial admissions

tenced population. This is shown in Figure 3.4.

(40 percent) and admissions for probation/parole violations (80 percent).

From 1999 through 2009, the number of bed-days used by pretrial
in 2009, pretrial inmates consumed 57 percent of all bed-days, up

have the longest stays; many pretrial inmates make bail after only a

from 44 percent in 1999. This meant that the Philadelphia Prison Sys-

few days. This is illustrated in Figure 3.3. As a result of these varia-

tem was populated mostly by people awaiting trial, not by convicted

tions, the profile of inmates admitted in a year looks quite different

criminals serving sentences.

2.5

3,310,991

3,365,420

3,230,759

3,139,851

2,889,340

2,764,449

2,853,570

2,700,057

3.0

2,524,522

3.5

3,025,606

4.0

2,400,970

DAYS [IN MILLIONS]

3.1 JAIL BED-DAYS CONSUMED 1999 – 2009

Source: Philadelphia Prison System

2.0
1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Note: Individuals who were admitted
but stay less than one day are counted
as staying half a day regardless of the
amount of time they actually spent in
the facility. Numbers are calculated
based on number of days spent in the
Philadelphia Prison System at time of
release. Because some individuals admitted in 2009 have not yet been released, bed-days in 2009 are
undercounted.

Chapter 3 HOW PHILADELPHIA’S JAIL POPULATION ROSE: 1999–2008

inmates increased from approximately 1,050,000 to nearly 1,900,000;

Length of stay varies by inmate type. Sentenced inmates typically

13

3.3 LENGTH OF JAIL STAY BY PATHWAY

PRETRIAL
PROBATION/
PAROLE VIOLATORS

34,127

32,747

33,256
31,066

30,599

31,516

35,000

31,548

40,000

38,890

38,398

40,858

45,000

33,654

NUMBER OF ADMISSIONS

3.2 ADMISSIONS TO THE PHILADELPHIA PRISON SYSTEM

25,000

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

31–120
DAYS

OVER
120
DAYS

42%

21%

22%

15%

8–30
DAYS

31–120
DAYS

OVER
120
DAYS

13%

43%

29%

16%

0%

20,000

8–30
DAYS

LESS
THAN
7 DAYS

SENTENCED

30,000

LESS
THAN
7 DAYS

LESS
THAN
7 DAYS

8–30
DAYS

31–120
DAYS

OVER
120
DAYS

31%

14%

31%

33%

20

40

60

80

100

PERCENT

Source: Philadelphia Prison System
Source: Philadelphia Prison System

14

3.4 JAIL BED-DAYS CONSUMED BY PATHWAY 1999 – 2009
DAYS [IN MILLIONS]

THE ROLE OF “FREQUENT FLYERS”
Twenty-eight percent of the people admitted to the Philadelphia

2.0

prison return at least once in any given year. Forty percent return
within two years. At the prison, these inmates are known as frequent flyers.
1.5

Our analysis of prison admissions and releases shows that these
chronic offenders are predominantly male, single and unemployed.

PRETRIAL

Most are accused of drug offenses, others of the kind of property
crimes that, while non-violent, can have a real impact on the lives
1.0

of the victims. Many suffer from some form of mental illness. Most

SENTENCED

are readmitted for the same type of crime for which they’ve gone
to jail in the past. And the majority, 75 percent, are readmitted ei-

.5

ther pretrial or for a probation or parole violation.

PROBATION/PAROLE VIOLATORS

“Chronic, low-level, offenders, they’re the biggest pain in our neck,”
said Prisons Commissioner Giorla. “We have one guy who’s been
OTHER

coming in and out of here for 30 years. Usually [he’s picked up on]
retail theft. Each time, they set his bail higher so he’s here a little
2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

0
1999

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

Note: Data reflects number of intakes not number of individuals.

longer. He’s functionally mentally ill. His medical treatment alone is
costing a fortune.”

Source: Philadelphia Prison System

The Rise in the Pretrial Population

3.5 PRETRIAL ADMISSIONS BY CHARGE SEVERITY

number had jumped 40 percent to 26,095. Most of the increase in
pretrial admissions came from individuals who were charged with
first- or second-degree misdemeanors. The number of inmates admitted pretrial charged with felonies rose by a smaller amount. See
Figure 3.5.
Three variables factor into pretrial admissions: the number of peo-

NUMBER OF ADMISSIONS

In 1999, 18,605 inmates were admitted to jail pretrial; by 2009 that
20,000

FELONY
15,000

10,000
FIRST- OR SECOND-DEGREE MISDEMEANOR

ple arrested, the number of people required to post bail and the
amount of bail set. In Philadelphia, all of these factors have been
5,000

SUMMARY/THIRD-DEGREE MISDEMEANOR

ARREST: DRIVING UNDER THE INFLUENCE AND
DRUG POSSESSION. Overall, the number of arrests in
trending downward. In 2009, for instance, there were roughly the

2009

2008

2007

2006

2005

2004

2003

2002

Philadelphia has remained fairly constant over the last decade,

2001

1999

0
2000

working at times to push up pretrial admissions.

Source: Philadelphia Prison System

same number as in 2002. But from 2005 to 2008, the total went up,
due largely to an increase in arrests for driving under the influence
(DUI) and drug possession. See Figure 3.6.
For DUI, legislative changes produced much of the increase in arrests and pretrial admissions. Starting in 2004, Pennsylvania enacted

3.6 ARRESTS FOR DRUG POSSESSION AND DRIVING
UNDER THE INFLUENCE
11,865

12,000

11,450

new laws which, among other things, lowered the minimum blood
alcohol level for DUI from .10 to .08.12 These efforts to get tougher

10,000

on drunk driving led to more arrests: 5,478 in 2008, up from an average of 4,000 per year between 2000 and 2005. More people were

8,000

9,216

10,499
9,836

DRUG POSSESSION

admitted to jail pretrial as a result.

4,000

offenders had bail set. This meant that a lower percentage was released on their own recognizance (ROR), meaning without bail—40
percent in 2009 compared to 46 percent in 2003. See Figure 3.7.

3,939

2009

2008

2007

2005

2,000

Source: Pennsylvania Uniform Crime Reporting System

3.7 BAIL DECISIONS
YEAR

RELEASED
WITHOUT BAIL

BAIL SET

MEAN BAIL

2003

46%

54%

$14,445

2004

44%

56%

$15,335

tution. In 2003, 74 percent of such defendants were released on

2005

40%

59%

$17,402

their own recognizance; in 2009, 60 percent were.

2006

40%

59%

$21,417

Another shift in bail decisions came in cases in which the most serious

2007

45%

54%

$24,940

charge was carrying a firearm without a license.

2008

44%

56%

$21,164

In 2005, Pennsylvania changed the sentencing guidelines to increase

2009

40%

60%

$20,008

The decline in the use of ROR and increase in bail were seen across
the board. But it was particularly pronounced for those accused of
third-degree misdemeanors, such as disorderly conduct and prosti-

the punishment for carrying a firearm without a license. The new
guidelines called for a presumed jail sentence of one to two years;
the previous ones had called for community-based sanctions in
some cases.13 This change appears to have had an impact on bail

Source: Preliminary Arraignment System
Note: Each year less than one percent of defendants are detained with bail
denied at preliminary arraignment. Bail amounts of $1,000,000 or more were
excluded from this calculation; overall, these represented just 0.4 percent of
all bail amounts set. Their inclusion would distort typical bail amounts.

12. For a description of the law change, see Pennsylvania Driver and Vehicle Services, “0.8 DUI Legislation,” Pennsylvania Department of Transportation,
http://www.dmv.state.pa.us/legislation/dui.shtml (accessed April 12, 2010).
13. Adoption of Sentencing Guidelines, 35 Pa. Bull. 1508 (March 5, 2005).

Chapter 3 HOW PHILADELPHIA’S JAIL POPULATION ROSE: 1999–2008

BAIL: CHANGED DECISIONS AND RISING
AMOUNTS. From 2003 to 2009, a higher percentage of accused

3,778

DUI

department intensified its street-patrol efforts to combat streetcorner crime with an emphasis on clearing high-crime locations.

4,748

4,578

account for a 7 percent increase in arrests from 2005 to 2009.
According to Deputy Police Commissioner William Blackburn, the

5,478

6,000

2006

For drug possession, a change of focus by the police appears to

15

decisions. Prior to the change, about 50 percent of defendants facing that charge were released on their own recognizance. Last year,
only 5 percent were, meaning more of them wound up in jail pretrial.
Bail determination is not meant to be an element of punishment;
while a case is pending, the presumption of innocence remains in
place. The point of pretrial detention and bail is to deal with people
who might not otherwise appear in court. Several factors go into
making the bail determination: whether the defendant has a valid
phone number, has a job, lives with a spouse or children, has a criminal history or has missed court dates in the past.14 The severity of
the offense with which the individual is accused also is taken into
account. In the case of carrying a firearm without a license, it appar-

DOING TIME FOR TRAFFIC VIOLATIONS
You do not expect to get jail time for failing to pay a parking fine.
And that almost never happens unless you are already in jail.
In 2008, our analysis shows that more than 1,500 inmates had their
stays in the Philadelphia Prison System extended—typically by one
or two days—so that they could be transported by the sheriff’s department to traffic court to pay fines. That number was up from less
than 200 inmates in 2003. While other circumstances initially
brought these inmates to jail, they were kept in jail after those issues
were resolved so that they could appear in traffic court.
In a system that uses about three million bed-days a year, this prac-

ently made a big difference.
While the use of bail was increasing, so was the amount of bail
being set. From 2003 to 2009, our analysis shows, the mean bail set
for all offenses rose nearly 39 percent, from $14,445 to $20,008, as
shown in Figure 3.7, although it has fallen in the past few years.
These two factors, higher bail and more people being asked to post

tice, which accounts for perhaps 2,000 bed-days, is not a big item.
But city officials are concerned about the propriety of incarcerating
people for traffic fines. And they know that holding inmates for
extra days and transporting them to traffic court may cost more
than the system recoups in fines paid. They are working to change
this practice.

it, drove up the number of defendants who could not make bail and
consequently ended up spending time in jail pretrial, thereby
adding to the overall jail population in Philadelphia.

In Philadelphia, nearly a quarter of detained pretrial defendants are jailed for more than
60 days, a higher share than in some other jurisdictions.
3.8 PRETRIAL LENGTHS OF STAY

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

100

16

80

60

40

20

0

COOK COUNTY

FULTON COUNTY

LESS THAN
3 DAYS

3 TO 7 DAYS

HARRIS COUNTY

LOS ANGELES
COUNTY

8 TO 30 DAYS

31 TO 60 DAYS

NEW YORK CITY

PHILADELPHIA

61 TO 120 DAYS

GREATER THAN
120 DAYS

Source: Sherriff’s departments of Cook, Fulton, Harris, and Los Angeles counties, New York City Department of Corrections, and the Philadelphia Prison System.
Note: Cook County includes Chicago. Fulton County includes Atlanta. Harris County includes Houston.

14. These are the factors collected and weighed by Philadelphia’s bail guidelines according to Todd Van Gunten, Preliminary Arraignment System (PARS) Technical Lead for
the Philadelphia Department of Technology. They include several of the factors the criminal code enumerates for consideration at preliminary arraignment. See Release
Criteria, 234 PA. CODE §523 (2001).

PRETRIAL LENGTHS OF STAY: VERY SHORT OR
VERY LONG. In Philadelphia, defendants are required to post

STAYING TOO LONG IN JAIL

10 percent of the assigned bail amount. Almost all defendants with

Philadelphia inmates sometimes spend additional time in jail be-

bail set at $500 or less are able to post the money within one day of

cause the orders releasing them were late getting from the court

arraignment.15 For those with bail over $500, the picture is very dif-

to the jail.

ferent. Forty percent of them never post bail. The other 60 percent
typically spend 5 to 15 days in jail while they put together the

In 2001, Philadelphia civil rights lawyers sued the city on behalf of

money required for their release.

defendants who remained in jail past the date they were supposed
to be released. Plaintiffs included a mix of both pretrial and sen-

About half of the pretrial population stays in jail for less than seven

tenced inmates. According to David Rudovksy, one of the lawyers

days. These people are not being detained to protect public safety;

in the case, the problem was that “sometimes the Clerk of Quarter

one way or another, they will be out on the street for most of the

Sessions wasn’t sending the release paperwork from court, and on

time prior to the resolution of their cases. The fact that they are ad-

the prison end, sometimes the paperwork did not reach the file or

mitted to jail in the first place—as well as how long they stay—is pri-

the officer in charge of releases. Simply, there was not a system in

marily a function of how long it takes them to post bail.

place to ensure that an ordered release was, in fact, effectuated.”

For the Philadelphia Prison System, these short stays are particularly

Thanks to procedures enacted as part of the settlement of the litiga-

expensive. Much of the cost of incarcerating someone is incurred at

tion, the staff of the Defenders Association now generates a daily list

intake. At admission, prison officials conduct medical examinations,

of their clients whose releases have been ordered. They then check

issue clothing, catalogue belongings and begin a lengthy paperwork

the files at the prison to make sure all of the releases actually hap-

process. This is true even for inmates who stay one day or

pen.17 Attorney Tom Innes, who does most of the checking, counted

less.16

20 instances in October 2009 when defendants would have remained

Philadelphia also has a comparatively large group of pretrial in-

in jail past their release dates had he not notified the prison.

mates who stay for more than 120 days, about 15 percent of all
pretrial admissions. In other jurisdictions, including the counties that
include Atlanta, Chicago and Houston, 5 percent of the pretrial
population stays this long. See Figure 3.8. The difference is due,
at least in part, to the length of the court process in Philadelphia.
Our analysis of data from the Administrative Office of Pennsylvania
Courts shows that 75 percent of cases in the city remain unresolved
after three months. As a result, criminal justice stakeholders report,
defendants sometime serve more time before their cases are resolved than they would have received in a sentence. Ultimately,

Admissions for Violations Increased
The Philadelphia Prison System also experienced a steady increase
in the number of inmates admitted for violating the terms of community supervision. Such supervision falls into two categories: probation
and parole. Probation is community supervision in lieu of incarceration; parole is supervision after incarceration. Both are supervised by
Philadelphia’s Adult Probation and Parole Department.18

many defendants held pretrial are released on time served.
Philadelphia’s jails for violating probation or parole rose from 3,101 in

jurisdictions, they did not get any longer during the past decade.

2000 to a high of 5,943 in 2009, contributing to the overall growth of

One factor that appears to have helped—even though bail amounts

the prison population during that time. In 2009, violators accounted

were going up—was the opening in 2007 of a bail office at the pris-

for 15 percent of all jail admissions, up from 10 percent in 2000.

ons. Before then, bail could be posted only at the Criminal Justice
Center in Center City. Anyone wanting to post bail for an inmate
had to go to Center City to deliver the bail deposit, then head out
to Northeast Philadelphia to pick up the inmate. Now both tasks
can be accomplished at once. This change has resulted in a notice-

This rise tracks with the overall increase in the number of probation
and parole cases. From 2004 to 2009, new cases accepted by the
Philadelphia probation department rose from 19,065 to 26,318. See
Figure 3.10.

able decrease in the time it takes to post bail. In 2006, it took a de-

Philadelphia’s experience in this regard is similar to the experiences

fendant with bail set at less than $500 an average of five days to

of jurisdictions across the country. From 1984 to 2009, the number

post the 10 percent needed. Today it takes an average of one day.

of Americans under community supervision rose from 1.6 million to

Similar trends are evident for defendants with higher bail rates.

5 million—one in every 45 adults.19

15. Ninety percent of defendants with bail of $500 or less eventually post bail. Some defendants who do not post bail do so as the result of a legal calculation rather than a
lack of money. Many defendants in the Philadelphia prison system have more than one reason for being in custody at a given time (see “The Challenge of Multiple Holds”
on page 12.) For example, a defendant might have bail set for $500 on a new arrest and also be detained on a bench warrant in a previous court case. In such a situation,
the defendant would be detained on the warrant even after posting bail. Individuals in this situation often choose not to post bail, as they would not be released anyway.
16. New York City is piloting a program in two boroughs designed to reduce the number of inmates who spend only a few hours or days in jail while putting together small
amounts of bail. In the program, defendants ordered to post $250 or less are reinterviewed after preliminary arraignment. If they are able to identify people who can post
bail for them soon, they are kept in the police holding cell rather than transported to jail for intake processing.
17. No similar process exists for inmates represented by private counsel.
18. Individuals who, because of sentence length, serve sentences in state facilities are supervised by the Pennsylvania Board of Probation and Parole. For purposes of this report, admissions to the Philadelphia Prison System based on community supervision violations include inmates under the supervision of both the Philadelphia and state
agencies. Of the 5,943 admissions based on violations in 2009, 9 percent (or 535) involved individuals supervised by the state’s probation and parole agency.
19. Pew Center on the States, One in 31: The Long Reach of American Corrections (2009).

Chapter 3 HOW PHILADELPHIA’S JAIL POPULATION ROSE: 1999–2008

As shown in Figure 3.9, the number of individuals admitted to
While pretrial stays in Philadelphia may be long compared to other

17

The idea behind the use of probation is to try to punish and rehabil-

One reason for these significant lengths of stay is the cumbersome

itate offenders without incarcerating them. But a lot of offenders on

and lengthy court process for hearing a violation of probation or pa-

probation end up in jail. How many? It’s hard to say because no one

role. Violation hearings traditionally have been held in front of the

keeps track. Some get there by failing to comply with what is known

judge who sentenced the defendant in the first place. Scheduling a

as a “technical condition” of their supervision. A probationer fails a

hearing before that judge can take time. And when the defendant’s

drug test, and a court hearing on the violation is scheduled—or

alleged violation is for a new crime, a different judge is assigned to

misses a meeting with a probation officer and a warrant is issued.

hear the new arrest. The original judge often waits until after the

Others are re-arrested for new crimes.

new case is concluded before conducting a hearing on the probation violation, delaying final resolution and keeping in jail a defen-

While admissions for probation and parole violations have risen
steadily, the length of stay for this population has gone up as well,

dant who might otherwise have been released.

thereby contributing to the growth in the jail population. The average length of stay, which was as low as 49 days in 2000 and as high
as 73 days in 2007, stood at 68 days in 2008.20 See Figure 3.11.

As new probation and parole cases rose from
2004 to 2008, so did admissions to jail for
probation and parole violations.

DAYS

3.11 PROBATION/PAROLE VIOLATORS LENGTH OF STAY
IN JAIL
80

73.1

60

67.9

67.7

70

57.8

57.7
49.4

50

52.4

60.6

63.2

54.6

40

3.9 PROBATION/PAROLE VIOLATORS ADMITTED TO JAIL

2008

2007

2006

2005

2004

2003

2002

2001

2000

10

1999

20

0

4,364

3,803

3,477

3,331

3,101

4,000

3,389

5,000

4,012

5,111

6,000

5,943

5,900

7,000

3,288

NUMBER OF ADMISSIONS

30

Source: Philadelphia Prison System
Note: Length of stay calculated based on number of days spent in the Philadelphia
Prison System at time of release. Length of stay not calculated for 2009 admissions
because many had not been released at the time of data collection.

18

HOW LONG IT CAN TAKE TO
2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

2,000

RESOLVE A PROBATION VIOLATION
To illustrate the impact of the court process on the jail population,
we asked several attorneys to share with us examples from their case-

Source: Philadelphia Prison System

loads. A criminal defense attorney pulled the following case:
A twenty-five year old man was on probation for a 2006 drug convic-

3.10 PHILADELPHIA’S ADULT PROBATION AND
PAROLE DEPARTMENT NEW CASES

tion. On November 7, 2009, he was a passenger in a car that was

20,000

26,318

22,605

22,205

19,200

25,000

19,065

30,000

26,216

pulled over by the police. An altercation ensued, and the driver and
CASES

the probationer were charged with aggravated assault. The bail commissioner set the probationer’s bail at $2,500 on the new charge. The
man chose not to post the bail since the arrest meant that he would
be held on the probation violation anyway. And he went to jail.
On January 11, 2010, after the case was continued four times, there
was a preliminary hearing on the assault charges, and the charges
were withdrawn. But the man’s hearing on the probation violation—

15,000

for getting arrested—had yet to take place. So he was sent back to
jail to wait to appear before the judge who had sentenced him on
2009

2008

2007

2006

2005

10,000
2004

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

3,000

the drug charges back in 2006.
On February 3, the violation hearing was held before the original sen-

Source: Adult Probation and Parole Department

tencing judge, who ordered him released. Total time in jail: nearly

Note: Data not available prior to 2004.

three months.

20. The median length of stay was 26 days at the low in 2000 and 34 days at the high in 2007. In 2008 it was 30 days.

How Philadelphia’s Jail
Population Fell: 2009–2010

4

After decades of growth, the population of the Philadelphia Prison System started to fall in 2009. See Figure 4.1.
As a result, the system was able to cancel some contracts to send inmates out of county and close a housing unit
at the Alternative and Special Detention Division. By February 2010, officials announced that the system was on
target to end the fiscal year $9 million under budget.
What caused this turnaround? Here are some of the answers.

A Decline in Sentenced Inmates
The largest factor in the lower population was a drop in the number

4.1 AVERAGE DAILY POPULATION, PHILADELPHIA PRISON
SYSTEM JANUARY 2009 – MARCH 2010

of convicted inmates serving their sentences in the Philadelphia

8,000

to longer sentences, the kind that had to be served in state prisons
and not in the Philadelphia jail.
There is no question, however, about what happened more recently.
In 2008, due in large part to the coordinated efforts of the Philadel-

7,500

7,000

phia judges, attorneys and advocates, the state legislature changed
the law regarding where sentenced inmates do their time.21
The change, which took effect in November 2008, meant that two

6,500

6,000

groups of convicted inmates would no longer have the option of
staying in the Philadelphia jails; they would have to go to state

5,000
2009

Source: Philadelphia Prison System

21. An act amending sentencing laws in the Pennsylvania Consolidated Statutes, H.B. 4, Gen. Assem. 2007 Sess. (Pa. 2008).

20 1 0

MAR

FEB

NOV

OCT

JUN

multiple sentences totaling two to five years.

MAR

tences of two to five years. The other consisted of individuals with

FEB

5,500
JAN

prison instead. One group consisted of inmates with single sen-

Chapter 4 HOW PHILADELPHIA’S JAIL POPULATION FELL: 2009–2010

8,464

8,595

8,717

community. Others speculate that judges sentenced more inmates

8,662

lowed some inmates to serve all or part of their sentences in the

JAN

8,500

ing for some sentenced felony offenders; these programs have al-

DEC

drug treatment court, DUI court and the use of electronic monitor-

9,177

SEP

opment of new alternative-to-incarceration programs, including

8,923

9,317

9,268

AUG

9,437

9,370
JUL

9,391

9,345

9,000

MAY

early stages of this decline. Some attribute the trend to the devel-

APR

9,500

Officials in the criminal justice system are not sure what caused the

9,501

that number had shrunk to 5,105.

9,787

10,000

admitted to serve their sentences in Philadelphia custody. By 2009,

9,656

Prison System. In 1999, as shown in Figure 4.2, 6,642 inmates were

19

Fewer Arrests
4.2 SENTENCED ADMISSIONS TO PHILADELPHIA PRISON
SYSTEM

Another factor in the decreasing jail population was the decline in

page 7. This appears to have contributed to several trends: the lev7,655

tions, and an overall 5 percent drop in admissions. That said, the
6,613

7,092

6,905

6,954

a 27 percent decrease in the number of inmates admitted for viola6,869

7,000

8,782

8,409

8,000

than in 2008, a decrease of about 11 percent. See Figure 1.2 on
eling off of pretrial admissions (26,095 in 2008 and 26,346 in 2009),

7,268

9,000

6,642

NUMBER OF ADMISSIONS

arrests. In 2009, the Philadelphia police made 8,000 fewer arrests
10,000

degree to which this drop in arrests caused the population decline
is not clear. As we have seen, a variety of factors play into the numbers, and in years past, arrest rates and the jail population have not
5,105

6,000

always moved in the same direction.

5,000

Changes in Probation and Parole

4,000

From 2000 to 2008, the number of probation and parole violators
admitted to the Philadelphia Prison System grew by an average of 7
2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

3,000

percent a year. In 2009, however, this trend leveled off. And the
number of bed-days consumed by violators fell somewhat.25

Source: Philadelphia Prison System

Two factors appear to have contributed to these trends. One was
that the number of new probation and parole cases remained constant at just over 26,000. See Figure 3.10. The other was that signifi-

According to Philadelphia Deputy District Attorney Sarah V. Hart,

cant changes were taking place at the Philadelphia Adult Probation

one of the prime movers behind the legislation, “We looked at

and Parole Department.

other states, and we were one of only three states where inmates
with sentences of over two years could end up in the county jail. It
didn’t make sense. From the budget perspective, we wanted long-

A GROWING STATE PRISON POPULATION

term inmates with the state, not with the city. The state has better
services to meet the treatment needs for long-term inmates.”

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

As soon as the change became law, city officials moved aggressively

20

to start transferring an estimated 400 inmates already serving twoto five-year terms in Philadelphia to state facilities.22 Almost all sentenced inmates in this category (except work release prisoners) now
go to state prison at the start of their sentence.23 Although Philadelphia does not keep track of how many inmates go to the state as a
result of the legislation, there is no doubt the change has reduced
the number of sentenced admissions. It also should eventually decrease the average length of stay for the sentenced population in

From 2008 to 2009, even as Philadelphia’s inmate population
peaked and then fell, Pennsylvania’s Prison System was recording
the largest increase in prisoners of any state in the nation in absolute numbers. Only two states, Indiana and West Virginia, had bigger percentage increases than Pennsylvania’s 4 percent. Over the
same period, 26 states had population declines.26
Philadelphia is responsible for a significant share of the increase
being experienced by the state. Last year, the number of new inmates going to the Pennsylvania Department of Corrections from
the Philadelphia courts rose by 6 percent.

the city facilities.24 And ultimately, it should reduce the caseload of

Without enough capacity to handle the growing population, Penn-

the Philadelphia Probation and Parole Department as more proba-

sylvania is sending prisoners to Michigan and Virginia.27 Plans are

tioners and parolees are supervised by the Pennsylvania Board of

in place to build four new state prisons. For fiscal year 2011, Gov-

Probation and Parole.

ernor Edward G. Rendell requested an additional $137 million for
the corrections department.28

22. This estimate, provided by the District Attorney’s Office, includes both inmates transferred under General Transfer Authorization, 61 Pa. Cons. Stat. § 1151 (2009) which
permits the temporary transfer of inmates from county to state facilities and vice versa and under separate petitions based on the law change. See Sentencing Proceeding; Place of Confinement, 42 Pa. Cons. Stat. §9762 (2008).
23. After conviction, inmates sentenced to state time return to the Philadelphia Prison System for a few days prior to being transferred. They are not, however, counted as
sentenced “admissions” in the prison case management system.
24. Data for this report was run in January, 2010, a time at which many inmates sentenced in 2009 were still in custody. Consequently, a length of stay analysis could not be
completed for this group.
25. As discussed in the Notes on Methodology section, bed-day consumption for 2009 is an estimate.
26. Pew Center on the States, Prison Count 2010 (2010).
27. Joseph N. DiStefano and Amy Worden, “2,000 Pa. Prisoners to be Housed Out-of-State,” Philadelphia Inquirer, 22 December 2009, B01.
28. Pennsylvania Governor’s Budget Office, 2010-11 Governor’s Proposed Budget (Harrisburg, 2010), 27.

THE COST OF HOUSING INMATES—NOT A SIMPLE CALCULATION
Officials often say that it costs $95 to house one inmate in jail for
one day, implying that reducing the population by one inmate would

4.3 ESTIMATED DAILY COST FOR ONE INMATE

save that much. But it does not work that way.

Food

The $95 figure is simply the total annual cost of the system divided

Pharmacy

2.29

by the number of inmate-days. Decrease the population by one in-

Lab Work

1.34

Intake diagnostics (one time)

3.82

Clothing (one time)

9.00

mate, and the system must still operate the same facilities and programs.
What does Philadelphia actually save per day by reducing the pop-

Total

ulation by one inmate? It depends on who the inmate is. For every
inmate, the city saves on the cost of that person’s food and clothing.

$ 3.84

$20.29

Note: Information on costs of individual items and services provided
by the Philadelphia Prison System. Analysis of daily cost per inmate
conducted by the Philadelphia Research Initiative.

As Figure 4.3 shows, this runs about $20 for the first day of incarceration—when clothing is issued and medical tests are conducted.
The savings is only about $7 for subsequent days, much more if the
inmate has serious medical or mental health needs.29 (If the inmate

overtime for the additional staff needed for triple-celled areas. And

had been housed in Lehigh County, the daily savings for Philadel-

it would be able to end the practice of sending inmates out of

phia would be $90.)

county.

Should the jail population continue to decrease, there would be

The biggest savings would come through closing an entire housing

more significant savings to be realized. At some point, the prison

area or jail facility. That could happen only if there were a large and

system would be able to reduce triple-celling, saving legal costs and

seemingly lasting drop in the inmate population.

With a nearly flat budget and a shrinking staff—probation and pa-

Based on the results of the experiment, the department began last

role was short 56 probation officers out of 324 approved officer-po-

year to change its reporting requirements and the distribution of its

sitions early in 2010—the department was having trouble managing

caseload.

Working with criminologist Lawrence W. Sherman and statistician
Richard A. Berk, both of the University of Pennsylvania, the department developed a statistical model to determine what factors made

In the new system
• high-risk clients are seen by officers with caseloads of 50 and
have a weekly reporting requirement;31

probationers and parolees likely to commit major crimes within two

• medium-risk clients are seen by officers with caseloads of 120

years after entering the probation or parole system. Using the

cases and must report on a monthly basis; and

model, the department began a controlled experiment in October
2008, giving each probationer or parolee a risk-score based on
those factors including age and first contact with the criminal justice

• low-risk clients are seen by officers with caseloads of 400 and
must report in-person once every six months.

system. The department took a portion of its clients who had scored

“Before now, we’ve dealt with many offenders the same way; now

“low risk” and assigned large numbers of them (400 as opposed to

we’re shifting to dealing with offenders in a more individual, a more

the previous 160) to individual officers. In addition, the reporting

targeted way,” said the department’s director, Robert Malvasutto.

conditions for low-risk probationers and parolees were reduced:

“It’s helped the individual offender; it’s helped the department.”

they would have to make contact with the probation department

Along with these changes within the Adult Probation and Parole

only once every three months, alternating between phone calls and
in-person visits as opposed to one in-person visit per month.
As it turned out, the probationers and parolees involved in the experiment had few new arrests for major

crimes.30

The reduction in re-

Department, the court system has begun to streamline the court
process for violators. These changes, discussed in detail later in this
report, have resolved some probation violations sooner than would
otherwise have been the case. Our own courtroom observations,

porting requirements also gave them fewer opportunities to violate

however, suggest that many of the individuals who have benefited

the terms of their supervision. As a result, there were fewer viola-

from these changes were not in custody when they appeared in

tions. All of this translated into fewer probationers being sent to jail.

court.

29. In 2009, the prison spent around $7.2 million on medication and $14 million in hospital fees.
30. Lindsay Ahlman and Ellen Kurtz, The APPD Randomized Controlled Trial in Low Risk Supervision: The Effect of Low Risk Supervision on Rearrest (First Judicial District of
Pennsylvania, Adult Probation and Parole Department, 2008).
31. On a given day, about half of the offenders supervised by an officer with a high-risk caseload are incarcerated. Consequently, probation officers with high-risk caseloads
typically supervise about 25 offenders in the community.

Chapter 4 HOW PHILADELPHIA’S JAIL POPULATION FELL: 2009–2010

its caseload. So it looked for ways to better utilize its resources.

21

Decision Points for
Admission to Jail

5

At various points in the criminal justice process, decisions are made about who should go to jail and how long
they should spend there. This next section of this report looks at the six main decision points. It describes how
they play out in Philadelphia and in some other jurisdictions.
Understanding all of this is essential to understanding what is being done to try to reduce the jail population in

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

Philadelphia and what else might be done—without jeopardizing public safety.

22

Four of the six main decision points for sending someone to jail are

Charging involves a balancing of priorities and resources. Among

related to pretrial incarceration: arrest, charging by the district attor-

the factors in play are the need to complete the process expedi-

ney’s office, preliminary arraignment and enforcement of outstand-

tiously so that suspects are not held in police lock-ups for excessive

ing warrants. A fifth, case disposition, determines the number of

lengths of time; the need to select the correct charges so that those

sentenced inmates in jail. The sixth, probation or parole, has impact

charges will not be reduced or thrown out later on; and the need to

on the number of inmates jailed for violations.

see that justice is done.

The flow chart on page 23, Figure 5.1, provides a basic map of how

In recent years, getting the work done expeditiously has been the

defendants travel through the Philadelphia criminal justice system

focus of the charging unit in Philadelphia. Under long-time District

and either do or do not enter the Philadelphia Prison System.

Attorney Lynne M. Abraham, who left office in January 2010 after
nearly 19 years in office, charges typically were issued and cases
sent to preliminary arraignment within 24 hours of arrest. Philadel-

Arrest

phia was known for having one of the fastest-charging units in the

If a police officer believes there is probable cause that a person has

country.

committed a crime, the officer can take the suspect into custody or,

The new district attorney, R. Seth Williams, says that the emphasis

for some lower level crimes, give the individual a paper notice to

on speed made the charging unit an “inefficient gatekeeper” in

appear in court at a later

date.32

which weak cases were allowed to go to court. In his view, weeding
out weak cases, if done with care, allows prosecutors to focus on
stronger cases. It also helps manage the jail population. With fewer

Charging by the District
Attorney’s Office

cases being charged, there are fewer people who can be detained
pretrial.

When arrested defendants are brought to a police district for processing, the paperwork is transmitted to the charging unit of the

CHARGING STANDARD. Under Abraham, the district attor-

district attorney’s office. The charging unit then determines whether

ney’s office used a “probable cause” standard to assess what

there is enough evidence to charge the defendant and, if so, what

charges to bring. That meant that charges were lodged if the facts

the charges should

be.33

and circumstances alleged in the police report would lead a reason-

32. Under Pennsylvania criminal procedure, there is a presumption that officers will give paper notices rather than make custodial arrests for all third-degree misdemeanors—
those in which the maximum sentence is less than one year, including prostitution, theft of less than $50 from a motor vehicle and public drunkenness. See Summons and
Arrest Warrant Procedures, 234 PA. CODE §1003(C)(1) (2010). It is the custom in Philadelphia, however, to arrest individuals on these types of charges.
33. Approval of Police Complaints and Arrest Warrant Affidavits by Attorney for the Commonwealth to Local Option, 234 PA. CODE § 507 (2010).

5.1 PHILADELPHIA CRIMINAL JUSTICE SYSTEM FLOW CHART

6

NO PROBABLE CAUSE
NO OUTSTANDING
CRIMINAL JUSTICE MATTER

PROBABLE CAUSE

SUMMONS/CITATION

ARREST

INSUFFICIENT EVIDENCE
TO CHARGE OR
DIVERSION

CHARGING BY THE DISTRICT
ATTORNEY'S OFFICE

OUTSTANDING WARRANT

P R I S O N

2

3

HELD ON BAIL

PRELIMINARY
ARRAIGNMENT

— COURT PROCESS —

4
— CO U RT —

DISPOSITION

— V I OL AT IO N S OF P R OB AT IO N / PA R OL E —

NOT GUILTY/ALTERNATIVE
SENTENCE/TIME SERVED

5
PROBATION OR PAROLE

PROBATION OR PAROLE
TERM CONCLUDED

SENTENCE OF
INCARCERATION

PROBATION OR PAROLE
VIOLATION

PENNSYLVANIA
DEPARTMENT OF
CORRECTIONS

Chapter 5 DECISION POINTS FOR ADMISSION TO JAIL

RELEASED WITH OR WITHOUT BAIL

P H I L A D E L P H I A

R E L E A S E

— P R E - TR I A L —

1

S Y S T E M

POLICE INTERACTION

23

able person to believe that the suspect had committed a specific offense. As a result, most cases were charged with the highest offense

FRONT END V. BACK END

arguable under the facts alleged in the report. Cindy Martelli, who

Philadelphia operates a number of alternative court and sentencing

headed the charging unit for Abraham, said the idea was to look “at

programs. These programs, which include Mental Health Court, DUI

what we can charge if everything on the police report is true” and

Court and Intermediate Punishment, come into play at the “back

not “to determine the quality of the case.”

end” of the criminal justice system, after an offender has accepted

Other jurisdictions, such as San Diego County, California, employ a

a guilty plea or has been convicted.

different standard at charging. “To issue a charge, we use proof be-

Some of these programs, like the newly-created Mental Health

yond a reasonable doubt, meaning that we have to believe we can

Court, are designed to shorten the stays of already-incarcerated in-

get a conviction on this charge before we’ll bring it,” said Terri

mates by helping them make the transition back into the commu-

Wyatt, chief of the case issuance division in the district attorney’s of-

nity. The expansion of such alternative-sentencing options may be

fice there. “If we read the police report, and we can tell we won’t

partially responsible for the decline in the number of sentenced in-

have the evidence to prove a case, we’ll send it back to the police

mates in the Philadelphia Prison System.

and ask for further investigation if appropriate.”

Diversion programs are different. They take offenders out of the
criminal justice system at the “front end,” often before preliminary

PLEA OFFERS. In Philadelphia, an offer from the district attor-

arraignment. The idea is that certain types of offenders can be

ney to the defendant for a plea bargain generally comes several

treated more appropriately outside the traditional criminal justice

weeks or months into a case. Other jurisdictions move more speed-

process and need not be incarcerated.

ily. In San Diego County, a recent study that looked at cases entering criminal court over a three-week span found that 60 percent of
felony cases were concluded within 10 days.34 In New York, plea

• Community Court, in which individuals charged in Center City

bargains are reached at the preliminary arraignment—the post-ar-

with “quality of life crimes” (such as retail theft, minor drug pos-

rest court hearing where bail is set—in 30 percent of all cases, and

session and prostitution) perform community service or get help

another 17 percent are dismissed or diverted at this stage.35 In

obtaining job training, health benefits and other services.

Philadelphia, few cases are disposed of at preliminary arraignment.
And only 25 percent of all cases are concluded in the first 90 days,
with the majority of cases taking 120 days or more, according to our
analysis of court data. These processing times contribute to pretrial
inmates in Philadelphia experiencing long stays in jail.

• Diverting Offenders into Treatment, in which defendants
charged with domestic-violence misdemeanors participate in
counseling classes or substance abuse treatment.
• Drug Treatment Court, in which drug-dependent defendants
charged with first-time drug offenses receive treatment. This

DIVERSION PROGRAMS. For decades, it has been the pracPHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

tice in Philadelphia and other jurisdictions to divert some lower-level

24

program can be used as a diversion program or an alternative
sentence.

offenders out of the formal court process. Once diverted from the
criminal justice process into social-service programs, these defendants are no longer at immediate risk of being incarcerated. In

Preliminary Arraignment

Philadelphia, as elsewhere, the district attorney’s office identifies

For defendants who are arrested and then charged by the district

cases eligible for diversion at the charging stage.

attorney’s office, the question becomes whether they should be put

Although Philadelphia’s diversion programs are generally lauded,
they apply only to certain categories of individuals, with an emphasis
on first-time offenders. Currently, the city operates five programs:

in jail or permitted to remain in the community pending the outcome of their cases. Under Pennsylvania law, a defendant may be
detained prior to conviction only if there is reason to believe he or
she will not appear at subsequent court proceedings. (In some juris-

• Accelerated Rehabilitative Disposition, in which first time, non-

dictions, the law allows public safety to be taken into account—not

violent offenders agree to a period of probation supervision.

in Pennsylvania.) A variety of factors may be considered in making

• Alternative Treatment for Misdemeanants, in which defendants

the determination of how likely a defendant is to appear for court

charged with prostitution, retail theft or drug possession who do

dates, including his or her community ties, criminal history and

not have prior convictions for violent felonies or any felony con-

record of appearing or not appearing in court.36

victions in the past five years are supervised by probation offi-

Much of the information related to community ties and past non-ap-

cers for two years and can be referred to drug and alcohol

pearances is collected by the Pretrial Services Division, an agency of

treatment.

the courts. In the first several hours after arrest, a social worker from

34. “Disposition Trends of Felonies for Sept. 1, 2009 Through Sept. 30, 2009 For All Divisions and Units for All Branches,” provided by the San Diego District Attorney’s Office (unpublished chart on file with the author).
35. New York City Criminal Justice Agency, Annual Report 2009 (2010), 16. “Preliminary arraignment” refers to the hearing in Pennsylvania in which the court rules on probable cause and pretrial detention.
36. Release Criteria, 234 PA. CODE § 523 (2001).

5.2 JURISDICTIONS COMPARED: PRELIMINARY ARRAIGNMENT DECISIONS
NEW YORK CITY

PHILADELPHIA

WASHINGTON, D.C.

RELEASED ON
RECOGNIZANCE OR
WITH SERVICES

85%

RELEASED ON
RECOGNIZANCE

BAIL SET

65%
BAIL SET

34%

RELEASED ON
RECOGNIZANCE

60%

40%
BAIL SET
OR HELD

15%

HELD WITHOUT BAIL

1%
Source: District of Columbia Pretrial Service Agency, New York City Criminal Justice Agency and Philadelphia Preliminary Arraignment System
Note: New York City numbers are for 2008, the most recent data available. Philadelphia and Washington, D.C., numbers are for 2009. The New York chart includes only the 52
percent of cases that were not resolved at preliminary arraignment (by dismissal or plea).

pretrial services interviews the defendant to determine the defen-

dered to pay bail.38 In New York City, the numbers were reversed,

dant’s living situation and employment status. This information is

with 65 percent of defendants released on their own recognizance

entered into the Preliminary Arraignment System (PARS), a comput-

and 34 percent having bail set.39 In Washington, D.C., where the

erized data collection system shared by the police, district attor-

law restricts the use of monetary bail, 85 percent of all defendants

ney’s office and the courts. Using a formula, PARS weights this

were released at preliminary arraignment and the rest held.40 See

information along with information retrieved from the court system’s

Figure 5.2.

current charges. The formula then produces a “score” with a corresponding recommendation that the individual should be released,
detained or have a specified amount of bail set.

Not only do New York and Washington release a higher percentage
of defendants into the community pretrial, they also have fewer of
them missing court dates. In New York, in 2009, 16 percent of released defendants failed to show up for their court appearances.41

Once the guideline recommendation has been calculated, the de-

In Washington, it was 12 percent.42 Unlike New York City and Wash-

fendant, who is being held in the police district where the arrest

ington, Philadelphia does not keep track of its failure-to-appear rate

took place, is linked by video to a hearing room in the basement of

for defendants released at preliminary arraignment. But our analysis

the Criminal Justice Center.37 At that point, a Municipal Court mag-

shows that the percentage in Philadelphia was about 30 percent.43

istrate determines whether the defendant will be released or have
bail set—and if so, how much. The magistrate takes account of the
guideline recommendation, arguments from the lawyers and the
magistrate’s own judgment.

Why are Philadelphia defendants less likely to make their court appearances even though more of them have had to post bail? The
answer lies largely in how the two main tools designed to ensure
that defendants released into the community appear in court—the

In Philadelphia, 40 percent of the defendants arraigned in 2009

bail guidelines and pretrial supervision—have been implemented

were released on their own recognizance while 60 percent were or-

and maintained.

37. The use of videoconferencing at preliminary arraignment has grown in acceptance around the country. However, the chief judge in Cook County, Illinois, eliminated the
practice in 2008 after reviewing a study that compared bail decisions before and after its introduction. The study found a significant increase in bail amounts for the same
charges under videoconferencing. One contributing factor was the inability of the defendant to consult privately with the public defender prior to the preliminary arraignment. In addition, the authors concluded that the removal of the human presence in the room possibly “encourages harsher responses than would occur if the judge were
faced with a live individual.” Shari Diamond, et al. “Efficiency and Cost: The Impact of Video-conferenced Hearings on Bail Decisions,” Journal of Criminal Law & Criminology, 100 (forthcoming). See also Matthew Walberg, “Video bond court to end; Northwestern study found it set bail 65% higher,” Chicago Tribune, 12 December 2008, 29.
38. Less than one percent were held without bail.
39. New York City Criminal Justice Agency, Annual Report 2009 (2010), 16.
40. Spurgeon Kennedy, Director of Research for the District of Columbia Pretrial Service Agency, “Memoranda regarding council member request for information,” March 26,
2010, provided by the District of Columbia Pretrial Service Agency (unpublished memoranda on file with the author). For publicly available data on this topic, see District
of Columbia Pretrial Service Agency, Leading the Field: FY 2008 Annual Report (2008). Use of pretrial detention also has consequences for the sentenced population. Defendants who are incarcerated pretrial are significantly more likely to be incarcerated after conviction, even controlling for other predictors of sentence severity. See Cassia
C. Spohn, How Do Judges Decide (Thousand Oaks, California: SAGE Publications, Inc., 2008), 101.
41. New York City Criminal Justice Agency, Annual Report 2009 (2010), 26.
42. Clifford T. Keenan (deputy director of the District of Columbia Pretrial Services Agency), interview with the author March 24, 2009.
43. In calculating the percentage of pretrial defendants in the community who miss court dates, we linked two data sources. From the courts, we looked at all cases in which a
bench warrant was issued. We removed all cases in which it was obvious that the warrant was issued for another purpose. In many cases, the bench warrant did not specify a cause. So we may have included some cases with warrants for reasons other than missed court dates. All of these cases were then linked with prison data. If the defendant did not appear in the prison data or was listed as having posted bail within the first few months after admission, the defendant was counted as being in the
community. Stakeholders report that some warrants are issued when the defendant is not absent intentionally. The defendant may be in custody, in the hospital or otherwise unable to get to court. Or the defendant may have left the courtroom temporarily and missed hearing the case called.

Chapter 5 DECISION POINTS FOR ADMISSION TO JAIL

database on criminal history, history of appearing in court and the

25

SPECIAL RELEASE
Since the 1990s, the city has had a “special release” procedure in
place. It mandates review of all cases in which an individual has been

liminary arraignment often focuses on the severity of charges and
whether these charges will be reduced at a later date, rather than
the totality of the factors in the guidelines. And once again, the
amount of bail set varies widely from magistrate to magistrate.47

sent to jail even though the bail guidelines recommended that he

Many magistrates, judges and lawyers believe that the guidelines are

or she be released on his or her own recognizance.

out of date. It has been more than a decade since research was con-

The city’s director of prison management looks at these cases on a
regular basis and sends some of them back to a magistrate. Twice
a week, a magistrate re-hears bail decisions. This process, from preliminary arraignment to re-hearing, can take five to 10 days. Each
week, about 40 cases are reheard. Most result in the immediate release of the defendant.

ducted on how well the guidelines predict whether an individual will
appear for court dates.48 One problem with the guidelines, said Municipal Court President Judge Marsha H. Neifield, is that “there are
now offenses that weren’t even on the radar screen when the guidelines were developed, such as gun straw-purchase [purchasing a gun
for someone else] and violations of protection from abuse orders.”

PRETRIAL SUPERVISION AND SERVICES. Currently,
BAIL GUIDELINES. The bail guidelines were adopted in

Philadelphia’s magistrates have four main choices at preliminary ar-

Philadelphia in the early 1980s. Officials viewed them as a tool to
help manage jail overcrowding and to standardize bail decisionmaking.

fendant without conditions pending trial or order the defendant to

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

The factors considered in the guidelines were based on empirical
research about what might predict a defendant’s appearance or
nonappearance in court including strength of community ties and
employment status. The idea was to help the court make sure that
all defendants who could remain in the community and make their
court dates were allowed to do so, thereby reducing the pretrial jail
population without endangering the public.

26

At the time the guidelines were introduced, researchers and criminal justice officials expected that magistrates would follow the recommendations 70 to 75 percent of the time. The expectation was
that the other 25 to 30 percent of cases would involve factors that
the guidelines could not take into account. In 1981 and 1982, when
the guidelines were tested, magistrate decisions matched the recommendations in 76 percent of all cases. When the guidelines were
updated in 1997, magistrate decisions matched recommendations
65 percent of the time.44
Today, according to the most recent analysis of bail decisions, magistrates follow the guideline recommendations in only 50 percent of
cases. When magistrates do not follow the recommendations, they
usually set higher bail.45 Most criminal justice stakeholders, including some who participate in preliminary arraignments, appear to
know little about what factors the guidelines consider or how those
factors are weighted.46 Attorneys report that the argument at pre-

raignment: set bail, hold the defendant without bail, release the decheck-in periodically with pretrial services.49 Defendants ordered to
check-in attend an orientation about the court process and then call
into an interactive voice-response system once or twice a week.
For some defendants who cannot make bail and thus are jailed, a
judge may reduce bail at a later court hearing. Or the judge may
order that the defendant be released on an electronic monitor or
with intensive supervision and restricted movement. Pennsylvania
criminal procedure limits the amount of time anyone can be held
pretrial to 180 days.50 Frequently, electronic monitoring and intensive supervision are used for defendants who have already been
held for 180 days.
In other jurisdictions, judges have additional choices. They can
order defendants with higher risk profiles to begin substance-abuse
treatment, receive mental health services and engage in frequent
face-to-face contact with pretrial staff. All of this occurs in the community rather than in jail. In Harris County, Texas, which includes the
city of Houston, one option is for defendants charged with DUI to
be given a device that prevents them from starting their vehicles
without passing a breathalyzer test.51 In Washington, D.C., some
pretrial defendants are ordered to wear electronic monitors from
the outset; in Philadelphia, such monitors are used primarily for sentenced offenders. The more options that are available, the more
ways there are to manage accused offenders safely without putting
them in jail or increasing the risk of their not appearing in court.

44. John S. Goldkamp and E. Rely Vilcicã, “Judicial Discretion and the Unfinished Agenda of American Bail Reform,” Studies in Law, Politics and Society 47 (2009) 133-142.
45. John S. Goldkamp et al., Confinement, 72. Our request for data on guideline scores was denied by the PARS steering committee. Consequently, we were unable to conduct our own analysis of bail guidelines recommendations and magistrate decisions. The numbers cited rely on the 2006 analysis conducted by Goldkamp, a professor in
the department of criminal justice at Temple University.
46. In conducting research for this report, we asked for copies of the bail guidelines from the District Attorney’s Office, the court system and the Defenders Association.
No one was able to provide a copy.
47. John S. Goldkamp, et al., Confinement, 72.
48. The 1994 Philadelphia guidelines manual states that the guidelines will be used to “collect data on a periodic basis showing the use of the guidelines and their results
(failures to appear and re-arrests) so that use of the guidelines, the impact of conditions of release and the performance of defendants can be reviewed and modifications
can be made if necessary.” John S. Goldkamp and M. Kay Harris, “Pretrial Release Guidelines,” vol. 3 (draft operational manual, Crime and Justice Research Center 1994).
Little research on guideline effectiveness or revision has been done since those words were written 16 years ago.
49. In addition, pretrial services calls all released defendants—both those released on their own recognizance and those released after posting bail—to remind them of court
dates.
50. Prompt Trial, 234 PA. CODE § 600 (2000).
51.. Harris County Pretrial Services, 2008 Annual Report (2008), 5-6. In 2008, approximately 660 Harris County defendants were required to install ignition interlock devices
while on pretrial supervision.

Disposition
Every case eventually concludes in a verdict, a plea agreement or
dismissal of charges. If the defendant is found guilty, the court can
send him or her to the Philadelphia Prison System if the sentence is
less than two years. Those sentenced to more than two years go to
the Pennsylvania Department of Corrections. The court can also
sentence the defendant to a period of probation or communitybased programming such as substance-abuse or behavior-health
treatment. If the individual has been held in jail pending resolution
of the case, the court can sentence that person to time served and
thus immediate release.
As discussed in the last section, this is the decision point that has

WHEN LANDING IN JAIL MAKES YOU HOMELESS AND
WHEN BEING HOMELESS KEEPS YOU IN JAIL
A 2004 survey of inmates in the Philadelphia Prison System found
that a third of prisoners expected to be homeless at the time of release.52 Of that group, a quarter reported that they were not homeless when they entered jail but had become homeless as the result
of being incarcerated—through losing a job, falling behind in rent
or developing strained relationships with family members and roommates.
According to Byron Cotter, director of alternative sentencing for
the Defenders Association, many inmates find themselves caught
in another bind: incarceration makes them homeless, and being

been a prime focus of the recent changes that have helped drive

homeless makes them ineligible for programs that could get them

down the Philadelphia jail population.

released. Said Cotter, “We could get a lot more inmates released
both pretrial and after sentencing if they had appropriate housing
available; inmates don’t qualify for alternative programs like electronic monitoring if they’re returning to a shelter or to the street.

Defendants who are sentenced to periods of probation or released

The fact is, the city can’t implement alternatives to incarceration

on parole typically sign agreements committing to comply with a

without supportive housing.”

series of conditions. These conditions may include regular check-ins

This situation, which adds to the jail population, is not unique to

with a probation officer, passing drug tests, attending therapy ses-

Philadelphia. In Chicago and New York, the local jails and human

sions and refraining from criminal activity.

services agencies have partnered with a nonprofit group, the Cor-

If there comes a time when the probation officer decides that the
offender has violated his or her agreement, the officer can file a
“violation”—a request that the court revoke the offender’s probationary status and put him or her in jail. If the offender’s violation
is for missing an appointment, the department may also issue a
warrant for the arrest of the offender.

poration for Supportive Housing, to identify and provide housing,

The types of people jailed for probation violations range from violent
criminals who have committed new crimes to individuals originally
arrested for low-level offenses who have failed to show up for a meeting with a probation officer and are picked up on a warrant. Whatever
the circumstances, violators can be held in jail for up to 12 days before seeing a magistrate who has the authority to release them.

advocacy and mental health services to individuals who cycle in and
out of jails and shelters. Early evaluations of these programs show
a significant decrease in shelter and jail stays for participating individuals.53 Philadelphia’s Office of Health and Opportunity has begun
discussion with the group to create a similar program for Philadelphia’s homeless.

Probation officials say that in the past there has been “no hard and
fast rule” for when a violation is filed or a warrant issued but add

Every weekday, a magistrate rules on the detention of between 60
and 100 recently-admitted inmates alleged to have violated the
conditions of their supervision. At those hearings, 20 to 25 percent
are released.

that all have always required supervisor approval. They say that

Rev. Ernest McNear, founder of the Kingdom Care Offender Reentry

ment is overwhelmed. In 2001, the city spent $1 on probation and

Network, a support program for offenders headquartered in South

parole for every $100 spent on jail; today, it spends 70 cents.54 During

Philadelphia, says that the decision to jail someone for a probation

this time the number of probation officers declined while the number

violation often seems arbitrary. “Much of what we see is someone

of cases grew. Today, probation officers have less time and few tools

who smoked drugs and doesn’t want to report so he wouldn’t get a

to deal with individuals who do not comply with their conditions of

hot urine [a positive drug test], or someone who is tired and forget-

probation and parole. After admonishing an offender verbally or

ful about reporting after a year or two,” McNear said. “Depending

sending a letter reminding a missing client of a skipped appointment,

on the officer, that guy might get a violation even though he’s been

there is little an officer can do other than file a violation or issue a

living safely in the community for years. Or he might not.”

warrant that may send the individual to jail.

some discretion is needed in order to manage clients on an individual basis.
One thing is clear. In Philadelphia, the probation and parole depart-

52. Philadelphia Prison System, Homelessness Among PPS Inmates (2004), provided by the Philadelphia Prison System (unpublished memoranda on file with the author).
53. Cooperation for Supportive Housing, Frequent Users of Public Services: Ending the Institutional Circuit (2009).
54. “First Judicial District Budget Submissions FY02-FY09,” provided by the First Judicial District Office of Financial Services (on file with the author).

Chapter 5 DECISION POINTS FOR ADMISSION TO JAIL

Probation or Parole

27

Outstanding Warrants
For some inmates in the Philadelphia jails, incarceration does not

In all, 1,207 individuals participated, many with multiple outstand-

begin with an arrest, conviction or violation of probation. Rather, it

ing warrants or violations of probation.56 According to Joseph A.

begins with a bench warrant.

Lanzalotti, deputy court administrator, the churches were “packed

Here’s how it usually works: An individual accused of a crime misses

door to door with people spilling onto the street.”

a court appearance, or a probationer misses a meeting with a pro-

The program has not been repeated since, although the court al-

bation officer. A warrant is filed calling for the immediate arrest of

lows individuals to clear-up outstanding warrants by coming to the

the individual. For some time, nothing happens. Then, the individ-

Criminal Justice Center voluntarily. Lanzalotti estimates that about

ual has an interaction with police, perhaps for a traffic violation or

90 people do so every day without being incarcerated. More might

something more serious. The warrant pops up on the police officer’s

come except for this: in a survey of Fugitive Safe Surrender partici-

computer screen, and the individual is taken into custody. The

pants, 37 percent said they had been afraid that they would be sent

missed court date or meeting may have been in connection with a

to jail if they returned to court.

felony or a minor offense; it may have occurred a month ago or
three years ago.
As of April 9, 2010, there were nearly 48,000 active bench warrants

WHO’S IN JAIL FOR PROBATION VIOLATIONS

in the Philadelphia court system for missed court dates. Of those
48,000, roughly 17,000 have been issued since January 1, 2005.

One day in November 2009, we observed hearings at the Curran-

The rest are older, some of them decades old.

Fromhold Correctional Facility for individuals jailed for alleged violations of probation or parole. At each hearing, a magistrate

To deal with the problem of missed appearances, a number of juris-

determined if the inmate should be released prior to a judge ruling

dictions have developed programs to follow-up with absent defen-

on the violation.

dants and get them back into court or probation quickly, without
bench warrants and the threat of jail time.

Some of the probationers released from jail that day included the

forming them of their missed dates and instructing them to go to

following:

calling for up to 29 days after the missed court date; half of all defendants return to court within 30 days without penalty.

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

for individuals re-arrested for serious crimes. But many were not.

In New York City, individuals who miss court dates receive calls inthe courthouse to reschedule. The nonprofit agency involved keeps

28

On this day, there were 125 cases before the magistrate. Many were

• A man in his twenties who had been convicted of possession of
narcotics and sentenced to probation three years earlier. In the last
year, he had stopped reporting to his probation officer. When he’d

In Santa Cruz County, California, the probation department works

gone to the Criminal Justice Center to check the date of a traffic

with a community-based organization in an area where there is a

court hearing he had coming up, the clerk there had run his name

high concentration of residents with violations of probation and

though the court’s database and found that he had a warrant for vi-

missed court dates. Before warrants are issued, community mem-

olating the conditions of his probation. He had been held in jail for

bers knock on the doors of such residents and help reconnect them

more than a week.

to the court or their probation officers. According to Barbara Lee of

• A man in his forties who had violated the terms of his probation

the Santa Cruz probation department, the program was initiated

by failing to take the required psychotropic drugs. While in jail, he

because putting many violators in jail seemed inappropriate. “They

had begun to take the drugs again. The prison gave him a three-

haven’t committed a crime, they aren’t dangerous, they’ve just been

day supply of the medication upon his release.

irresponsible,” she said. Since the program’s inception, Santa Cruz

• A professional truck driver who had been stopped while driving

has cut the number of warrants filed by 51 percent.55

his truck. The officer who made the stop said that a records check

Philadelphia has no similar program to actively seek out absent de-

had revealed an outstanding warrant against the man for failing to

fendants and probationers. For several days in 2008, however, the

report to his probation officer. The man told the magistrate that he

city participated in the national Fugitive Safe Surrender program.

had recently taken a new job and had missed one appointment with

Administered by the U.S. Marshals Service and the Philadelphia
courts, the program gave individuals with outstanding warrants the
chance to go to several churches and mosques and have their warrants lifted.

his probation officer while working with the officer to get his reporting time changed from daytime to evening. Before releasing the
man, the magistrate read the probation department’s case file. The
file described the man as an exemplary probationer who had missed
only the one appointment.

55. “Bench Warrants 2005-2009” (March 2010), provided by the Santa Cruz County Probation Department (on file with the author). According to Santa Cruz Probation Director Scott MacDonald, the program is primarily used for missed probation meetings though it can be used for missed court dates. The program, operated by the nonprofit
Friends Outside, costs the county approximately $50,000 annually.
56. “Statistics and Survey from Fugitive Safe Surrender” (March 15, 2008), provided by the First Judicial District of Pennsylvania (on file with the author).

What is Being Done to
Reduce the Jail Population

6

A lot is changing in the Philadelphia criminal justice system, much of it aimed at reducing the jail population
while maintaining public safety. Many of these reforms affect the decision points that determine whether people
go to jail and how long they stay.

Changing the Charging Standard

Expediting Plea Agreements

District Attorney Williams has overhauled the charging unit, crafting

“Crash court” is an expedited plea process run through Municipal

new procedures and giving it a bigger and more experienced staff

Court to deal with individuals being held pretrial on certain lower-

than in the past. “The reinvigorated unit is in the process of chang-

level misdemeanor charges. Cases are selected by the city’s director

ing the charging standards from `probable cause’ to `reasonable

of prison management and the public defender’s office. Under an

doubt,’” Williams said. “I want people to know that only the right

accelerated time frame, the public defender and district attorney

cases are being charged; that when we charge a case, we can con-

negotiate pleas. If a defendant agrees to the plea offer, the case is

vict beyond a reasonable doubt.”

placed on the crash-court docket. On a typical week, about 50

the court process by encouraging more defendants to accept plea
bargains earlier and by keeping more low-level cases out of the

cases are heard in this manner. According to the public defenders,
cases referred to crash court are completed within two to three
weeks after preliminary arraignment.

courts. He also hopes that it will allow his office to focus its re-

Crash court has been in existence for several decades, but now the

sources on the cases that remain. In April 2010, he announced that

district attorney’s office has broadened the range of eligible cases

his office would no longer charge low-level marijuana possession

to include some domestic violence cases and individuals with war-

cases as misdemeanors; instead, it would treat them as summary of-

rants. In addition, the court, the prison system and the attorneys are

fenses—cases that receive citations and do not go through the pre-

working to speed up the process through videoconferencing; previ-

liminary arraignment process.57

ously, defendants were transported to the Criminal Justice Center

These changes have the potential to shrink the jail population and
save money, although Williams said that is not his primary concern.
“My role is to seek justice in individual cases,” he said. “We should

twice a week for crash court. Thanks to videoconferencing, the court
is now able to operate on a daily basis, thereby getting eligible inmates out sooner.

be more efficient in assessing our cases, determining to whom pleas

Said chief public defender Greenlee: “The changes to crash court

should be offered and for whom diversionary programs are most ap-

are making a big difference. The defendants are getting the [plea]

propriate. Such a strategy will allow us to focus more energy and re-

offers earlier. It’s taken a lot of effort, and all of the different stake-

sources at targeting violent criminals.”

holders working together. But it’s paying off.”

The challenge, according to other prosecutors and defense attorneys, will be how to make these determinations when, at preliminary
arraignment, the charging unit often has little more than a barebones police report. And the risk is that an individual released without being charged will go on to commit a serious crime.
57. Nancy Philips, Dylan Purcell, and Craig McCoy, "Philadelphia to Ease Marijuana
Penalty," Philadelphia Inquirer, April 5, 2010, A01.

Chapter 6 WHAT IS BEING DONE TO REDUCE THE JAIL POPULATION

Williams and his team hope that this approach will help speed up

29

Day Reporting Center

Sentencing and Reentry

Philadelphia is making plans for a day reporting center—a place

In 2009, Philadelphia opened the Mayor’s Office of Reentry Serv-

where nonviolent individuals who otherwise would be in jail can

ices, since renamed the Mayor’s Office of Reintegration Service for

report daily and be supervised while remaining in the community.

Ex-Offenders. The office provides services to help convicted in-

Eligibility to participate would be limited, and the proposed center

mates leaving state or local incarceration reintegrate into the com-

would provide participants with a range of supportive services.

munity; the focus is on employment assistance and mentoring.

Plans for the Philadelphia center are taking shape. Policy makers are

Advocates for reentry services argue that these services can reduce

deciding whether it would be used by the probation department for

the populations of the city jail and state prison in two ways. First,

violators, by judges for alternative sentencing, or by the courts and

successful reintegration reduces the number of people who commit

pretrial services for individuals who otherwise would be in jail await-

new crimes. Second, judges and the state parole board are more

ing trial.

likely to release convicted inmates early if inmates can be released

Said Deputy Mayor for Public Safety Gillison: “What do we need

to an established program.

to reduce the jail population? We need a place for people to go if

The city’s new Mental Health Court also has the potential to shorten

we’re not putting them in jail. That’s what this will be.”

the jail-stays of sentenced inmates. Opened in 2009, it provides
behavioral health services and judicial supervision to convicted in-

DAY REPORTING
The term “day reporting center” has different meanings in different

mates with histories of mental illness. Individuals, identified while
in jail, can have their periods of incarceration shortened by agreeing
to participate.

communities.
Generally speaking, a day reporting center is a place where individuals who would otherwise be in jail go for supervision and social
services. Hours of reporting vary. Most centers require that participants stay at the facility for the better part of the day, but some
allow participants with jobs or other obligations merely to check in.
In the last 20 years, day reporting centers have been established
from western Massachusetts to central California, from Chicago to
Orlando. They have been used for individuals who have violated
conditions of probation, for individuals awaiting trial who would otherwise be incarcerated and for people who would otherwise be
serving sentences.
PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

Some centers focus on ensuring that people show up and keep off

30

the streets. Others concentrate on providing social services.
There are few studies on the impact of the centers. Research done
in 1998 and 1999 on the center in Cook County, Illinois, which is designed for pretrial defendants, found that less than 1 percent of the
individuals being supervised failed to appear in court and that 5 percent were arrested while in the program.58
One of the obstacles to establishing a day reporting center, of
course, is finding a neighborhood that accepts it.

Streamlining the Court Process
for Violators
The city is also updating court practices regarding individuals on
probation or parole. Since March 2009, the courts have launched
three programs designed to reduce the time it takes to get a violation of probation heard by a judge:
• Accelerated Violation of Probation Program. Individuals alleged to have violated a “technical” term of their probation—
by missing a meeting or failing a drug test—now can have their
cases heard on an expedited basis by a judge specializing in
such hearings. To participate, they need the consent of the
judge who originally put them on probation. Our observation of
the program indicates that many of the 440 cases it handled in
2009 involved probationers who were not in jail at the time of
their hearings; no statistics on this matter are available. The program affects the jail population to the degree that it deals with
violators who are in custody.
• Non-Sitting Judges Cases. A problem for some probationers
awaiting violation hearings is that the judge who sentenced
them originally now sits in civil court and has trouble making
time for a violation hearing. Since last year, however, more than
1,000 of these hearings have been reassigned to the judge
specializing in violation hearings.
• Advanced Review and Consolidation Program. This covers probationers accused of new crimes. They typically face two legal
problems: the new accusation and a probation violation stemming
from the new arrest. Defendants who accept a plea agreement on
the new charges can have those charges and the probation violation heard together by the violations judge. Many of these cases
also involve people not in jail at the time of the hearing. In 2009,
1,635 cases were resolved through the program.

58. Christine Martin, Arthur J. Lirigio, and David E. Olson, “An Examination of Rearrest and Reincarceration Among Discharged Day Reporting Center Clients,” Federal
Probation, 2003: 26.

Cautiously rolled out last year, these new programs are gaining acceptance. The experience of other jurisdictions indicates that the
programs, if expanded to include more of the people who are in
detention, have the potential to reduce the jail population. For example, in Hillsborough County, Florida, which includes Tampa, it
used to take 37 days on average for a violation of probation to

THE IMPORTANCE OF GOOD PAPERWORK
Whom to admit, whom to detain, whom to let out and when—the
Philadelphia Prison System gets its instructions from the Clerk of
Quarter Sessions.

make it through the court system. Now, the county has a specialty

For 18 years, Vivian T. Miller, an elected official, had the responsi-

court that hears all violations and new cases related to the viola-

bility for recording court orders and transmitting them to the prison.

tions. All outstanding matters are consolidated, and cases get re-

Her office lacked the necessary technology and quality controls.

solved in less than half the time. The county’s former detention
department commander, David Parrish, said the change has been
a substantial contributor to a decline in the inmate population by
1,000 from 2005 to 2009.

Judges, lawyers and prison officials complained that the clerk’s office often recorded and transmitted inaccurate, late or confusing
orders. These mistakes led to some high-profile cases of inmates
getting out too early while others were held too long.
In March, Miller announced her retirement, and on April 1, Mayor
Nutter and President Judge Dembe announced that the court sys-

Other Changes to Court Processes

tem would take charge of the work. “We are excited to bring the

In Philadelphia, it often takes six months or more for a case to move

Dembe said.

functions of the clerk’s office to 21st century standards,” Judge

though the criminal justice system. Judges, prosecutors and defenders agree that continuances are at the heart of this problem.
Once a case is approaching trial, the most frequent cause of continuances is unfinished discovery, said D. Webster Keogh, administrative judge for the trial division. Discovery is the requirement that
prosecution and defense provide each other certain evidence prior
to trial.
To address this issue, the Court of Common Pleas in January 2010
opened Discovery Court, a courtroom dedicated to working out discovery issues prior to trial. All cases headed to trial must first stop in
discovery court. The goal, said Keogh, is to “reduce disposition impediments and get the parties to address the case in a meaningful
way earlier on” and thereby “reduce continuances, witness appearances, police overtime and busy work for the court and move the
process along.”
Although it is too soon to know, this new court could impact the jail
population by reducing the time spent in jail by inmates awaiting
not in custody. Why fewer warrants? Research indicates that the
longer a case lasts, the more likely defendants are to miss court
dates.59 So, speeding up the court process may result in fewer warrants and fewer people in jail as a result.
While Discovery Court may reduce how long some defendants stay
in jail, the impact of another recent court practice is open to debate. On April 17, 2010, the rules governing Philadelphia courts
were changed to increase the time between the preliminary arraignment, when bail is set, and the preliminary hearing, when the court
determines whether the prosecution has grounds to proceed with
the charges.60 Under the old system, proponents of the change say,
prosecutors often did not have time to notify witnesses; as a result,
hearings were postponed, keeping jailed defendants behind bars
pretrial. Others counter that the new system means it will take
longer for some cases to be dismissed, resulting in longer pretrial
stays for some.

59. Goldkamp, et al., Confinement, 66.
60. Amendment to Rule 1003 (Procedure in Non-Summary Municipal Court Cases) 40 PA Bulletin 2012 (2010).

Chapter 6 WHAT IS BEING DONE TO REDUCE THE JAIL POPULATION

trial and by reducing the number of bench warrants issued for those

31

7

What More Can Be Done to
Reduce the Jail Population

The population in the Philadelphia Prison System has fallen, thanks largely to a sharp decline in the number of sentenced inmates there. But the overall population continues to be driven primarily by inmates sent to jail prior to
trial, many of whom only stay for a matter of days. Unless innovations are made at the decision points that influence
pretrial detention, Philadelphia’s gains in controlling its jail population may be hard to maintain or build upon.

Some work is already going on in this area, including the changes in
the district attorney’s charging unit and the planning for a day reporting center.
The final section of this report outlines additional options for
change, based on programs already in place in other jurisdictions. If
any of these options are adopted, they would have to be tailored to
Philadelphia’s needs. Unlike the transfer of inmates to the state system, some of these changes would be slow in producing measuraPHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

ble results. Several are items the city has considered and worked on

32

in the past.
The key to building on recent gains is not just creating individual
new programs; it is designing a coordinated plan that allows for ongoing monitoring and adjustment. Said Joanne Fuller, former director of the Department of Community Justice in Multnomah County,
Oregon, and a noted criminal-justice reform leader: “System reform
is never done—it is an ongoing process where each reform leads to
better understanding both of what is working well and what is not.

Expanding Options for Diverting
Cases away from the Courts
In interviews for this report, multiple stakeholders in the Philadelphia criminal justice system expressed a desire to see the system
expand its programs to divert various types of cases away from the
criminal justice system and thus the possibility of individuals going
to jail. In particular, they cited the need to create alternative responses for nonviolent, chronic offenders who have underlying issues that contribute to their criminal behavior but that are not
addressed through the traditional court process.
“Some of these people go through intake three or four times during
a year,” said Chip Junod, the city’s director for prison management.
“Every time they come through, the city has to pay for initial cost
of intake. It’s easier not to put them in than trying to get them out
after the fact. What we need is something other than jail for these
people.”

So each reform exposes the need for the next reform. It becomes

It was this type of thinking that led Milwaukee County, Wisconsin,

not a thing you do but a way of thinking about your work and the

to expand its diversion programs. “There are low risk offenders in

work of whole systems.”

need of treatment,” said John Chisholm, the district attorney there.

None of this is easy. Most of the measures discussed here have
up-front costs. And all must be tested against the need to maintain
public safety and get defendants who are not in jail to appear in
court.

“I said to the public, ‘Let me create responsive diversion programs
for the categories of offenses and offenders who annoy us but don’t
scare us.’ In the process, we’ll reduce violent crime, and we’ll put
less people in jail.”
None of Philadelphia’s five diversion programs, described earlier,
is targeted at chronic offenders or defendants whose mental health
contributes to their criminal conduct. And while the city has an array
of alternative sentences and specialty courts for these types of offenders, they come into play only after a defendant has entered the

Philadelphia Detention Center
CREDIT: ROBERT LECONEY

Revitalizing the Bail Guidelines
criminal justice system. Without alternatives diverting offenders

As discussed earlier, the guidelines for determining whether a defen-

from the beginning, the city will continue to send chronic, low-level

dant should be released at preliminary arraignment or have bail set—

offenders to jail over and over again, particularly pretrial.

and how much that bail should be—are frequently ignored in
Philadelphia. Bail is being set in an increasing number of cases, bail
amounts have risen and the city has a relatively high rate of defen-

• In Milwaukee County, a network of community-based agencies

dants who miss court dates. Other cities, including New York and

provides services to clients diverted from court. These include

Washington, set bail in a lower percentage of cases and yet report

drug and alcohol counseling, general equivalency diploma and

lower levels of missed court dates. For Philadelphia, the challenge is

work-placement programming for men between the ages of 17

to transform the existing guidelines into a useful and effective deci-

and 19, healthcare for the homeless, life skills and esteem-build-

sion-making tool.

ing for women, and anger management education.

What does such a decision-making tool look like? In New York City

• In Allegheny County, Pennsylvania, at preliminary arraignment,

and Montgomery County, Maryland, courts rely consistently on the

workers from the county’s Office of Behavioral Health screen de-

recommendations generated by bail guidelines—and a relatively

fendants suspected of having behavioral-health problems. When

high percentage of released defendants show up for their court

services are needed, the behavioral-health workers design appro-

dates. The guidelines in those places have three things in common:

priate service-plans. With the consent of all parties, the defendants are released into services without further court involvement.

• Magistrates and judges get full write-ups of all of the factors
that go into a guideline recommendation, such as the defen-

• In Bexar County, Texas, which includes San Antonio, individuals

dant’s living and employment situation. In Philadelphia, the

suspected of minor crimes and thought to be mentally ill are

magistrates get only a guideline score.61

taken by police officers directly to a 60-bed facility known as the
Crisis Care Center. If these individuals turn out to be mentally ill
and accept treatment, they are not at risk of going to jail. “Treatment works,” said Leon Evans, who conceived the program and

• The guidelines are used as the basis for ongoing research into
the predictors of whether a defendant will appear in court. No
such research is being done in Philadelphia.

heads the agency that runs it. “When they’re lucid, these people

• In New York City and Montgomery County, the guideline rec-

have no propensity to commit crimes.”

ommendations do not suggest specific bail amounts, as do

61. New York judges are also given information on factors not considered in their guideline recommendation. When New York revised its guideline tools in 2003, the research
that led to the revision indicated that several elements previously included—someone meeting the defendant at preliminary arraignment as well as the absence of any prior
bench warrants or unresolved cases—were no longer predictive of appearance in court. The Criminal Justice Agency, which compiles the pretrial reports, took these factors
out of the calculation. But at the request of the judges, it continues to include them in its write-ups to the court.

Chapter 7 WHAT MORE CAN BE DONE TO REDUCE THE JAIL POPULATION

Here are some examples of what other jurisdictions are doing:

33

Philadelphia’s guidelines. Rather, they simply sort defendants

gamut from individuals getting arrested for new and violent crimes

into three categories: low risk of failure to appear (recom-

to those who missed single meetings with probation officers.

mended release), moderate risk of failure to appear (no recommendation), and high risk of failure to appear (bail or detention
recommended).

Recognizing that probation officers have growing caseloads and
limited options, community corrections departments around the
country are working to standardize the circumstances under which

A transformed bail-guideline tool—one that is shown to be effective

violations and warrants can be filed. These departments are estab-

and presents useful information in a clear format—could change

lishing clear policies about when to issue formal violations; they are

current practice. It would have the potential of increasing the per-

developing administrative sanctions that allow probation officers to

centage of defendants who show up for court and reducing the

respond to infractions without using incarceration or involving the

number of pretrial admissions to jail.

court. Sanctions include requiring probationers to make more frequent check-ins, do weekend community service, accept electronic

Creating a Broader Range of
Pretrial Services

monitoring or go to day reporting centers.
In Georgia, the state probation department began employing such
sanctions in select counties in 2004. An initial study of the practice

At preliminary arraignment, Philadelphia’s magistrates have limited

found that probationers in those counties spent far fewer days in jail

choices: release defendants on their own recognizance, require

than probationers in other counties.62 The program has been ex-

them to check-in weekly with a voice-response system, set bail or

panded to include the entire state.

order them held without bail. This could change if the city opens a
day reporting center. But even with this addition, Philadelphia’s pretrial options would remain relatively thin.

Getting People to Show Up

Jurisdictions including Washington, D.C., and Montgomery County,

Every day in the Philadelphia Prison System, a magistrate hears

Maryland, tie pretrial services to their bail guidelines. Defendants

dozens of cases of people jailed on warrants, usually for missed

who score in the moderate-risk category—meaning that they pose

court dates or missed probation appointments. As discussed earlier,

some risk of failing to appear for court dates—are linked to services

jurisdictions such as New York City and Santa Cruz County, Califor-

designed to counterbalance the factors that make them flight risks.

nia, have developed programs that reach out to defendants who

These services typically include drug counseling and housing sup-

miss court dates. These programs give defendants the chance to get

port. Intensive supervision, including electronic monitoring or as-

back into court promptly without punishment, saving jail bed-days.

signment to a day reporting center, is reserved for the members of

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

the group considered least likely to appear in court.

34

Said Nola Joyce, chief administrative officer of the Philadelphia Po-

Reinvesting for Safety and Savings

lice Department and a former police official in the District of Colum-

Jurisdictions that have experienced savings from reduced jail or

bia, “When you put people back into the street without doing an

prison populations are hungry for more. And they are eager to use

effective sort and matching on risk and need, I don’t feel safer be-

these savings to fund other priorities. Philadelphia is no exception.

cause all you’re doing is putting people back on the street. But
what happened in the District made more sense. D.C. Pretrial Services worked to match individuals with programs and appropriate
levels of supervision. You knew that the [pretrial defendants] who
needed it were getting the support they required.”

But some programs that could reduce the population further have
up-front costs. Under a concept known as “justice reinvestment,”
some of the savings generated by reduced incarceration-levels is
put back into such programs. In Texas, for example, state officials
put a portion of the funds no longer needed for prison construction

A broader selection of pretrial services would allow more members

into community alternatives to imprisonment. The result has been a

of the moderate-risk population to remain in the community safely

continued decline in incarceration levels and an estimated savings

pretrial while potentially increasing court appearance rates.

of $230 million over two years.63
A national Justice Reinvestment Act is currently under consideration

Developing New Responses to
Probation Violations

in Congress.64 If enacted, the act would provide federal grants to

In 2009, there were 5,943 admissions to the Philadelphia Prison Sys-

and agree to reinvest “averted prison or jail costs” into programs

tem for violations of probation or parole. These violations ran the

that improve public safety.

local and state jurisdictions that use data analysis to help craft policy
options that safely “manage the growth of corrections spending”

62. Applied Research Services, Inc., An Evaluation of Georgia's Probation Options Management Act (Atlanta: Applied Research Services, Inc., 2007).
63. Council of State Governments Justice Center, Justice Reinvestment State Brief: Texas (New York, NY: Council of State Governments Justice Center, 2007).
64. Criminal Justice Reinvestment Act, H.R. 4080, 111th Cong. (2009).

Tracking Performance
What impact are various measures having on the jail population?

establish key benchmarks, identify when a course correction is

Which decision points in the criminal justice system are functioning

needed and figure out what the new goals should be.65 In many

as intended and which are not working as well?

places, one of those goals is managing the use of incarceration.

How many sentenced inmates are going to state prisons? What is

“We need to set milestones and track our performance,” said

the effect of the revamping of the district attorney’s charging unit?

Deputy District Attorney Hart. “We need to be able to see how

Are new court practices changing the pretrial lengths of stay for in-

the system is performing and be able to systematically identify the

mates? Has the number of violations and warrants in probation and

types of cases that are falling through the cracks.”

parole cases risen or fallen since the probation department changed
its supervision model?

Without careful monitoring, programs designed to reduce jail populations such as day reporting centers and electronic monitoring can

On a monthly basis, Philadelphia’s Criminal Justice Advisory Board

end up serving defendants who would be in the community anyway.

reviews aggregate daily population numbers from the prison. And

These programs ultimately cost money instead of saving it and have

the Court of Common Pleas provides numbers on hearings and dis-

little or no impact on how many people are in custody. Absent reli-

positions. But these statistics are not sufficient to let criminal justice

able and current data, the public is unlikely to support innovations

stakeholders track the performance of the system in general or their

that claim to reduce jail populations without compromising public

reforms in particular.

safety.

An increasing number of other jurisdictions are adopting perform-

Philadelphia Industrial Correction Center
CREDIT: ROBERT LECONEY

65. Performance indicators can be implemented without integrating data systems. Stakeholders identify a few, easily collectable statistics for each decision point or innovation. The agencies responsible for those points or reforms then report out from existing data sources.

Chapter 7 WHAT MORE CAN BE DONE TO REDUCE THE JAIL POPULATION

ance indicators—statistics that help describe changes over time,

35

PPS Data
Data on the population of the prison were obtained from the casemanagement system maintained by PPS. Data included information
for all persons admitted between January 1, 1999 and December
31, 2009. A series of individual-level variables extracted from the
data included demographic information (such as gender, race, and
marital status), arrest charges, admission date, admission reason, release date, release reason, case number, charge disposition and

36

METHODOLOGY

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

sentence and sentence date (if applicable).

Notes on Methodology
This report relies on five data sources to
examine population trends in the Philadel-

PPS data were used to examine the number of admissions, the daily
population on June 30 of each year, the lengths of stay for persons
released from PPS and the number of bed-days consumed. For
each of these, the analyses focused on the individual’s pathway to
jail or hold status (pretrial, violator, sentenced, or other) and the
most serious holding offense for each pathway.
Determining hold status, most serious holding offense and length of
stay in each hold status is a complicated process. Individuals can be
held in jail for multiple reasons at the same time, and their status
can change over time as charges are dropped or adjudicated. For
example, an individual can be admitted pretrial on a new arrest and
as a probation violator at the same time. For each unique intake
into PPS—each time an individual physically enters PPS from outside the jail—the case management system assigns a unique intake
number.
To examine admissions, our analysis counts unique intake numbers—the same method used by the prison system. The analysis allowed individuals to have only one status per unique intake

phia Prison System: admissions data main-

(admission). To determine this status, we ranked serving a sentence

tained by the Philadelphia Prison System

first, being held on a probation or parole violation second, being

(PPS), arraignment data maintained by the

held pretrial third and being held for another jurisdiction fourth. In-

Preliminary Arraignment System (PARS),

dividuals with multiple holds for the same intake number or whose

case-processing data maintained by the

hold status changed while incarcerated (without a new intake) were

Administrative Office of the Pennsylvania
Courts (AOPC), crime data maintained by
the Pennsylvania Uniform Crime Reporting

categorized as having only one status for the purposes of counting
admissions. Thus, an individual who had a status of “sentenced” at
any point during an intake was categorized as “sentenced” for
counting admissions. Individuals could have multiple admissions on

System and jail population data collected

the same case—if, for instance, they posted bail and were released

by the Bureau of Justice Statistics.

pretrial but later came back sentenced.

AOPC Data
After ranking the hold status, the analyses further ranked admissions

Data on the processing of cases through the court system were ob-

by most serious offense as determined by the offense gravity score

tained from the system maintained by AOPC. Data included infor-

contained in the Pennsylvania Sentencing Guidelines. When it was

mation for all persons with criminal cases filed against them in

impossible to match the specific offense to a specific score, the low-

Municipal or Common Pleas Court between October 1, 2006 and

est offense applicable gravity score was assigned to the offense.

December 31, 2009. This time period was chosen because the

To examine the daily population, the lengths of stay and bed-days
consumed, our analysis categorized individuals by a different

Philadelphia court system began reporting data into the AOPC system starting in September 2006.

method. For daily population, we looked at hold status on June 30.

The AOPC data were developed from several data sets, containing

For length of stay and bed-days consumed, we looked at an individ-

information on bail, confinement, charges, dispositions, sentences,

ual’s entire admissions period and calculated how long the individ-

case events and warrants. Data from these sets were merged using

ual was incarcerated on a given hold status. Because individuals

common case number identifiers, which allowed for the creation of

may have multiple hold statuses at the same time, the previously-

a single database and the analysis of linked demographic and case

described ranking of hold statuses was used to determine the pri-

information.

mary hold status at any given point.

AOPC data were used primarily to examine bail outcomes and war-

In counting length of stay and bed-days consumed, an inmate’s stay

rants issued for individuals arraigned between 2006 and 2009. The

was assigned to the year in which the defendant was released, re-

analysis focused on the individuals who posted bail and the time it

gardless of when he or she was admitted. Because data was pro-

took them to post bail. The analysis also considered the cases with

vided to us in January 2010, many inmates admitted in 2009

warrants issued and the warrants issued for individuals released pre-

(particularly sentenced and violator inmates) had not been released

trial. This required merging AOPC data with PARS and PPS data to

by then. Consequently, length of stay and bed-day numbers for

determine which individuals were released ROR and which individu-

2009 were estimated.

als with bail set were released pretrial. Data from AOPC were first
merged with PARS data using the OTN assigned to each charge;

PARS Data

the data were then merged with PPS data using the docket number
assigned at arraignment.

Data on preliminary arraignment decisions were obtained from the
case management system maintained by PARS. Data included infor-

Crime and Arrest Data

mation for all persons who had preliminary arraignments between
January 1, 2003 and December 31, 2009. A series of individual-level

Longitudinal data describing crime and arrest trends were obtained

variables extracted from the data included demographic informa-

from public data sets through the Pennsylvania Uniform Crime Re-

tion, most serious charge, offense tracking number (OTN), arrest

porting System.

date, bail disposition and bail amount.

National Jail Data

who had preliminary arraignments between 2003 and 2009. The

Data describing jail populations in other jurisdictions were obtained

analysis focused on the decision to release individuals on their own

from public data sets maintained by the Bureau of Justice Statistics,

recognizance (ROR) pretrial or to set bail and, if bail was set, the

a division of the U.S. Department of Justice. Additional data on cor-

amount. The analysis examined bail decisions for unique OTNs or

rections spending were obtained directly through interviews and

arrest incidents.

surveys of corrections administrators in other jurisdictions.
METHODOLOGY

PARS data were used to examine the bail decisions for individuals

37

ACKNOWLEDGEMENTS

PHILADELPHIA’S CROWDED, COSTLY JAILS: THE SEARCH FOR SAFE SOLUTIONS

38

This report would not have been possible
without the cooperation of Commissioner

We gratefully acknowledge the judges and administrative staff of
the Philadelphia Court system who gave us their time and insights.
They include Common Pleas Court President Judge Pamela P.
Dembe; Judge D. Webster Keogh, administrative judge of the trial
division; Judge Sheila Woods-Skipper, supervising judge of the
criminal section of the trial division; Municipal Court President
Judge Marsha H. Neifield; and David C. Lawrence, Joseph A. Lanzolotti, Kathleen M. Rapone, Kevin Cross, Keith Smith and their colleagues in court administration. In addition, we would like to thank
all of the members of the Preliminary Arraignment System working
group for providing us data on preliminary arraignments and Todd
Van Gunten, the system’s technical lead, for making sure that we
understood the data.
Additional officials throughout Philadelphia’s criminal justice system
have shared with us their perspective: Director Robert Malvestuto,
Ellen Kurtz, Kathleen M. Intenzo and the staff of the Adult Probation
and Parole Department; William Babcock from Community Court,
Chief Defender Ellen T. Greenlee and her staff at the Defender As-

Louis Giorla and his staff at the Philadel-

sociation, including Byron C. Cotter, Charles A. Cunningham, Abi-

phia Prison System. In addition to provid-

gail Horn, Elizabeth Hurt, John M. Konchack, Thomas J. Innes III,

ing much of the raw data used for the

Joanna McClinton, Stuart Schuman and Adrienne Winney; District

analysis in this report, the staff at the

Attorney R. Seth Williams and his staff, including John Delaney,

prison worked with us to make sure we

Christopher P. Diviny, Robert Foster, Sarah V. Hart, Deborah Harley,

understood the nuances of the system. In

Kristen Heine, Dawn M. Holtz, John Andrew Jenemann, Thomas

particular, we would like to acknowledge
James DiNubile, Roseanne Duzinski,
Robert Eskind, Robert Fitzmartin, Marco

Lipscomb and Catherine Thurston; Deputy Mayor for Public Safety
Everett A. Gillison and his staff, including Chip Junod and Michael
Resnick; Police Commissioner Charles H. Ramsey and members of
the Philadelphia Police Department, including Kevin Bethel, Francis

Giannetta, Bruce Herdman, Cheryl Morri-

Healy, and Nola Joyce; former District Attorney Lynne M. Abraham;

son, Richard Petrelli, Christopher Thomas

Councilman Bill Green; former Prison Commissioner Leon A. King II;

and Robert Tomaszewski.

and former candidate for district attorney Michael Untermeyer.

About the Authors
This report was enriched by our conversations with several service-

Claire Shubik-Richards, senior associate of the Philadelphia Re-

providers and advocates: Rev. Ernest McNear and Greg Thomas from

search Initiative at The Pew Charitable Trusts, was the primary au-

the Kingdom Care Reentry Network; Angus R. Love and Su Ming Yeh

thor and researcher on this report. Data analysis was performed by

from the Institutional Law Project; Amy Augustine and Deborah

Don Stemen, assistant professor of criminal justice at Loyola Univer-

Schlater from the Philadelphia Health Management Corporation;

sity Chicago. The report was overseen and edited by Philadelphia

Cameron Holmes, Betty-Ann Soiefer Izenman and Ann Schwartzman

Research Initiative project director Larry Eichel.

from the Prison Society; and David Rudovsky, civil rights attorney and
professor at the University of Pennsylvania Law School.
A welcome orientation to the Philadelphia criminal justice system,
its data and its history was provided by Temple University professor
John S. Goldkamp.
In addition to these individuals listed by name, many Philadelphia
attorneys, clerks, judges, inmates, scholars and magistrates spoke
with us about their experiences and ideas. We are truly grateful for
their contribution.
We would also like to thank the officials throughout the country who
took time to talk with us about their jails and jail reforms, particularly
Scott MacDonald and Barbara Lee in Santa Cruz County, California;
Jerome McElroy in New York City; Sharon Trexler in Montgomery
County, Maryland; David Parrish in Hillsborough County, Florida;
and Leon Evans in Bexar County, Texas.
And we are thankful for the support and guidance we received from
Adam Gelb, Jake Horowitz and Richard Jerome, our colleagues at the
Public Safety Performance Project of the Pew Center on the States.
Their broad expertise in the field of corrections provided us with wise
counsel in conceiving, executing and completing this report.
Other Pew colleagues contributed to this effort as well. They include
Cindy Jobbins, Donald Kimelman, Jennifer Lowes and Emily
Pennsylvania Law School, joined the Pew team as an intern to assist
on this report.

ACKNOWLEDGEMENTS

Cheramie Walz. Kristen-Elise Brooks, a student at the University of

39

www.pewtrusts.org/philaresearch

THE PEW CHARITA BLE TRUSTS
2005 MARKET STREET, SUITE 1700
PHILADELPHIA, PA 19103-7077

www.pewtrusts.org