Federal Courts: Differences Exist In Ordering Fines and Restitution
(Letter Report, 05/06/99, GAO/GGD-99-70).

Pursuant to a congressional request, GAO provided information on federal
offenders who were ordered to pay criminal fines and victim restitution,
focusing on: (1) the percentage of offenders who were ordered to pay
fines or restitution in fiscal year (FY) 1997 and those who were not;
(2) differences across judicial circuits and districts in the percent of
offenders who were ordered to pay fines or restitution and those who
were not; and (3) officials' opinions about possible reasons for those
differences.

GAO noted that: (1) while many factors influenced whether an offender
was ordered to pay a fine or restitution, the judicial circuit or
district where the offender was sentenced was a major factor during
fiscal year (FY) 1997; (2) most of the approximately 48,000 federal
offenders sentenced under United States Sentencing Commission Guidelines
in FY 1997 were not ordered by the courts to pay a fine or restitution;
(3) the percentage of offenders who were ordered to pay fines or
restitution varied greatly across the 12 federal judicial circuits and
94 federal judicial districts; (4) across districts, the percentage of
offenders who were ordered to pay fines ranged from 1 percent to 84
percent, and the percentage of the offenders who were ordered to pay
restitution ranged from 3 percent to 49 percent; (5) an important factor
in determining whether an offender was ordered to pay a fine or
restitution was the type of offense committed; (6) while 6 percent of
offenders sentenced for immigration offenses were ordered to pay a fine,
almost one-third of property offenders were ordered to pay; (7) besides
the type of offense committed, other factors that were associated with
whether an offender was ordered to pay included factors such as sex,
race, education, citizenship, length of sentence, and type of sentence
imposed, such as prison, probation, or an alternative; (8) even after
controlling for all of these factors for four specific types of offenses
in GAO's multivariate statistical analyses, the judicial circuit or
district in which the offender was sentenced continued to be a major
factor in determining whether an offender was ordered to pay a fine or
restitution; (9) some court officials and prosecutors provided
explanations of why differences existed among the districts; (10) some
attributed the differences to the nature and type of offenses committed
or types of offenders sentenced in the districts; (11) some officials
believed that the culture in the judicial district among the prosecutors
and court officials contributed to whether offenders were fined or
ordered to pay restitution; and (12) since the imposition of restitution
for certain offenses became mandatory with the passage of Mandatory
Victims Restitution Act (MVRA), the percentage of offenders, overall,
who were ordered to pay restitution during FY 1997 actually declined
from 26 percent of offenders sentenced for crimes committed before April
24, 1996, to 12 percent of offenders sentenced for crimes committed on
or after April 24, 1996, when MVRA became effective.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-99-70
     TITLE:  Federal Courts: Differences Exist In Ordering Fines and
	     Restitution
      DATE:  05/06/99
   SUBJECT:  Fines (penalties)
	     Law enforcement
	     Restitution
	     Comparative analysis
	     Crimes or offenses
	     Federal courts
	     Criminals
IDENTIFIER:  Crime Victims Fund

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  This text was extracted from a PDF file.        **
** Delineations within the text indicating chapter titles,      **
** headings, and bullets have not been preserved, and in some   **
** cases heading text has been incorrectly merged into          **
** body text in the adjacent column.  Graphic images have       **
** not been reproduced, but figure captions are included.       **
** Tables are included, but column deliniations have not been   **
** preserved.                                                   **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************
FEDERAL COURTS: Differences Exist In Ordering Fines and
Restitution (GAO/GGD-99-70) FEDERAL COURTS Differences Exist In
Ordering Fines and Restitution

United States General Accounting Office

GAO Report to the Chairman, Senate Committee on the Judiciary, and
the

Chairman, Subcommittee on Crime, House Committee on the Judiciary

May 1999 

GAO/GGD-99-70

May 1999   GAO/GGD-99-70

United States General Accounting Office Washington, D. C. 20548

General Government Division

B-279743

Page 1 GAO/GGD-99-70 Federal Courts

GAO May 6, 1999 The Honorable Orrin G. Hatch Chairman, Committee
on the Judiciary United States Senate

The Honorable Bill McCollum Chairman, Subcommittee on Crime
Committee on the Judiciary House of Representatives

As you requested, this report discusses federal offenders who were
ordered to pay criminal fines and victim restitution. The
objectives of this report are to (1) identify the percentage of
offenders who were ordered to pay fines or restitution in fiscal
year 1997 and those who were not, (2) identify differences across
judicial circuits and districts in the percent of offenders who
were ordered to pay fines or restitution and those who were not,
and (3) provide officials' opinions about possible reasons for
those differences. Based on discussions with your staff, we also
documented changes in the rate at which offenders were ordered to
pay restitution before and after the Mandatory Victims Restitution
Act (MVRA) (title II of P. L. 104- 132) was enacted April 24,
1996. This is the second of two reports concerning victim
restitution and criminal fines based on your request. Our first
report 1 concerned how offenders are required to make payments on
fines and restitution that have already been ordered, including
the guidelines available for determining payment schedules and how
payment schedules were established.

Individuals convicted of a federal crime can be ordered by the
court to pay a fine or restitution at sentencing. Criminal fines,
which are punitive, are to be paid in most cases to the Department
of Justice's (DOJ) Crime Victims Fund. United States Sentencing
Commission (USSC) Guidelines provide guidance on the minimum and
maximum fine amounts to be imposed by the courts based on the
offense. In establishing the USSC, Congress sought, as one
objective, uniformity in sentencing by narrowing the wide
disparity in sentences imposed for similar criminal offenses
committed by similar offenders. Fines may be waived if the
offender establishes that he or she is unable to pay and is not
likely to become able to pay a fine. MVRA reformed restitution law
and now requires the court to order full restitution in certain
cases to each victim in the full amount of each

1 Fines and Restitution: Improvement Needed in How Offenders'
Payment Schedules Are Determined (GAO/GGD-98-89; June 29, 1998).

B-279743 Page 2 GAO/GGD-99-70 Federal Courts

victim's losses, without regard to the offender's economic
situation. Previously, as with fines, the court could waive
restitution, in most cases, based on the offender's inability to
pay.

While many factors influenced whether an offender was ordered to
pay a fine or restitution, the judicial circuit or district where
the offender was sentenced was a major factor during fiscal year
1997. The large statistical variation among judicial circuits and
districts raises a question, on a broad level, about whether the
objective of uniformity in the imposition of fines and restitution
is being met.

Most of the approximately 48,000 federal offenders sentenced under
USSC Guidelines in fiscal year 1997 were not ordered by the courts
to pay a fine or restitution. About 19 percent were fined by the
courts and about 20 percent were ordered to pay restitution. The
percentage of offenders who were ordered to pay fines or
restitution varied greatly across the 12 federal judicial circuits
and 94 federal judicial districts. Across districts, for example,
the percentage of offenders who were ordered to pay fines ranged
from 1 percent to 84 percent, and the percentage of the offenders
who were ordered to pay restitution ranged from 3 percent to 49
percent. The likelihood of an offender being ordered to pay fines
or restitution could have been three times or more greater in one
federal judicial district than in an adjacent district.

An important factor in determining whether an offender was ordered
to pay a fine or restitution was the type of offense committed.
While 6 percent of offenders sentenced for immigration offenses
were ordered to pay a fine, almost one- third of property
offenders were ordered to pay. Similarly, while 1 percent of drug
offenders were ordered to pay restitution, almost two- thirds of
fraud offenders were ordered to pay.

Besides the type of offense committed, other factors, based on our
statistical analyses, that were associated with whether an
offender was ordered to pay included factors such as sex, race,
education, citizenship, length of sentence, and type of sentence
imposed, such as prison, probation, or an alternative. However,
even after controlling for all of these factors for four specific
types of offenses in our multivariate statistical analyses, the
judicial circuit or district in which the offender was sentenced
continued to be a major factor in determining whether an offender
was ordered to pay a fine or restitution.

Some court officials and prosecutors provided explanations of why
differences existed among the districts. Some attributed the
differences to Results in Brief

B-279743 Page 3 GAO/GGD-99-70 Federal Courts

the nature and type of offenses committed or types of offenders
sentenced in the districts. Some officials believed that the
culture in the judicial district among the prosecutors and court
officials contributed to whether offenders were fined or ordered
to pay restitution. The culture included how prosecutors and court
officials worked together to identify victims and their losses,
among other factors.

Since the imposition of restitution for certain offenses became
mandatory with the passage of MVRA, the percentage of offenders,
overall, who were ordered to pay restitution during fiscal year
1997 actually declined from 26 percent of offenders sentenced for
crimes committed before April 24, 1996, to 12 percent of offenders
sentenced for crimes committed on or after April 24, 1996, when
MVRA became effective. The differences in the likelihood of
offenders being ordered to pay restitution for crimes committed
before and after MVRA became effective varied for specific types
of offenses; for example, the percentage of offenders ordered to
pay restitution increased for robbery offenses but decreased for
larceny and fraud offenses. Some court officials and prosecutors
believed that it was still too early to see the full impact of
MVRA. These officials commented that time is needed to become
familiar with and implement the act. They said that there may also
be mitigating circumstances, such as the recovery of stolen money
in a robbery, to explain why restitution was not ordered in a
particular case.

Congress divided the country into 94 federal judicial districts,
and in each district there is a federal district court. The U. S.
district courts are the federal trial courts the places where
cases are tried, witnesses testify, and juries serve. Congress
placed each of the 94 districts in one of 12 regional circuits,
and each circuit has a court of appeals. If a trial is lost in
district court, the case can be appealed, and the court of appeals
can review the case to see if the district judge applied the law
correctly. Figure 1 depicts a map of the United States that shows
the geographical boundaries of the 94 district courts and the 12
regional circuit courts (including 11 numbered circuits and the
District of Columbia Circuit.) There is also a Federal Circuit
whose court of appeals is based in Washington, D. C., but hears
certain types of cases from all over the country. The Court of
Claims is the trial court from which the appeal arises for the
Federal Circuit. Background

B-279743 Page 4 GAO/GGD-99-70 Federal Courts

Source: Administrative Office of the U. S. Courts data.

Under the law, offenders in federal court may be ordered to pay a
fine or restitution at sentencing. The court is to impose a fine
in all cases, except where the defendant establishes that he or
she is unable to pay and is not

Figure1: Geographical Boundaries of Twelve Regional Circuit Courts
Including the District of Columbia and Ninety Four State District
Courts

B-279743 Page 5 GAO/GGD-99-70 Federal Courts

likely to become able to pay any fine, according to guidelines
issued by the USSC, 2 which interpret federal law.

In establishing the USSC, Congress sought, as one objective,
uniformity in sentencing by narrowing the wide disparity in
sentences imposed for similar criminal offenses committed by
similar offenders. USSC's Guidelines state that the amount of the
fine should always be sufficient to ensure that the fine, taken
together with other sanctions imposed, is punitive. Except where
the fine is established by specific statute, the fine should be
within a range established by the USSC. The range is based on the
offense level. The base offense level, which is a numerical score,
is established in the Guidelines for each type of crime.
Adjustments can be made to the base offense level for such things
as offense characteristics, offender's role, the victim,
obstruction of justice, and acceptance of responsibility. For
example, the Guideline range for fines for offense level 8 is $1,
000 to $10,000; for offense level 38 the guideline range for fines
is from $25,000 to $250,000. There is also a criminal history
category that helps the judge determine whether an offender should
be sentenced higher or lower within the Guideline range. According
to the USSC Guidelines, in determining the amount of the fine, the
court should consider, among other factors, the need for the
combined sentence to (1) reflect the seriousness of the offense
(including the harm or loss to the victim and the gain to the
defendant), (2) promote respect for the law, (3) provide just
punishment, and (4) afford adequate deterrence.

According to the USSC Guidelines, the court should also consider
any evidence presented on the offender's ability to pay a fine in
light of the offender's earning capacity and financial resources,
as well as the burden that the fine places on the offender and his
or her dependants, any restitution or civil obligations the
offender is required to make, other consequences of conviction
such as civil obligations, whether the offender has been
previously fined for a similar offense, and any other pertinent
equitable considerations. The court may waive the fine or impose a
lesser fine if it has been established that the offender is not
able and is not likely to become able to pay all or part of the
fine, even with the use of a reasonable installment schedule. The
court may also waive the fine or impose a lesser fine if the
imposition of a fine would unduly burden the defendant's
dependents.

2 USSC was created by Congress in the Sentencing Reform Act of
1984, P. L. 98- 473. The act required USSC to develop a system of
sentencing guidelines.

B-279743 Page 6 GAO/GGD-99-70 Federal Courts

Since passage of the Victims of Crime Act of 1984 (P. L. 98- 473),
most fines have been used to support DOJ's Crime Victims Fund. The
fund provides, among other things, grants for victim assistance
programs and compensation to victims.

Restitution is to be paid to the victim of the crime and should
reflect actual losses suffered as a result of the crime. However,
mandatory restitution as part of a federal criminal sentence is a
relatively recent idea. Initially, the courts did not recognize
restitution as a separate term of a criminal sentence. This
changed with the Victim and Witness Protection Act of 1982 (P. L.
97- 291), which was the first statute that broadly addressed
victim restitution in general. Before enactment, imposition of an
order of restitution was completely within the discretion of the
court and could only be ordered as a condition of probation.
Later, in 1992, with the enactment of the Child Support Recovery
Act (P. L. 102- 521), Congress introduced the idea of mandatory
restitution into federal law. The act mandated that courts impose
restitution on defendants convicted of willful failure to pay past
due child support. In the Violence Against Women Act of 1994
(title IV of P. L. 103- 322), Congress identified certain other
types of crimes subject to mandatory restitution, such as sexual
abuse, sexual exploitation and other abuse of children, domestic
violence, and telemarketing.

On April 24, 1996, MVRA was enacted as title II of the
Antiterrorism and Effective Death Penalty Act of 1996 (P. L. 104-
132), reforming restitution and altering the way it is to be
enforced. MVRA now requires the court to order restitution for
each offender who has been convicted or has pled guilty to the
following charges, without regard to the offender's economic
situation:

 a crime of violence (as defined in 18 U. S. C. 16);

 an offense against property under title 18 of the U. S. Code,
including any offense committed by fraud or deceit; or

 an offense relating to tampering with consumer products (18 U. S.
C. 1365). According to MVRA, there must be an identifiable victim
or victims who have suffered a physical injury or monetary loss.
The only other exception to mandatory restitution is for an
offense against property if the court finds that

 the number of identifiable victims is so large that it makes
paying restitution impractical or

B-279743 Page 7 GAO/GGD-99-70 Federal Courts

 determining complex issues of fact related to the cause or amount
of the victims' losses would complicate or prolong the sentencing
process to a degree that the need to provide restitution to any
victim is outweighed by the burden on the sentencing process.

Prior to MVRA, the law provided that the court could consider, in
determining whether to impose an order of restitution, the
financial resources of the offender, the financial needs of the
offender and his or her dependents, and other factors the court
deemed appropriate. Under MVRA, the court cannot waive restitution
based on the offender's economic circumstances. However, the court
can order the offender to make nominal periodic payments if the
offender's economic circumstances do not allow for (1) the payment
of any amount of a restitution order or (2) the payment of the
full amount in the foreseeable future, under any reasonable
schedule of payments.

According to the legislative history, an intent of MVRA was to
establish one set of procedures for the issuance of restitution
orders in federal criminal cases. The history cited fiscal year
1994 USSC statistics that showed that about 20 percent of federal
criminal cases had restitution orders. The history also cited
rates for specific types of crimes, such as about 55 percent of
offenders sentenced for robbery were ordered to pay restitution.
The history cited these figures as indications that more progress
in ordering restitution remained to be made.

The objectives of this report are to (1) identify the percentage
of offenders who were ordered to pay fines or restitution in
fiscal year 1997 and those who were not, (2) identify differences
across judicial circuits and districts in the percent of offenders
who were ordered to pay fines or restitution and those who were
not, and (3) provide officials' opinions about possible reasons
for those differences. We also documented changes in the rate at
which offenders were ordered to pay restitution before and after
MVRA was enacted April 24, 1996.

To identify the number of offenders who were ordered to pay fines
or restitution and those who were not, we used USSC data for
fiscal year 1997. USSC maintains a computerized data collection
system, which forms the basis for its clearinghouse of federal
sentencing information. USSC requests that each probation office
in each judicial district submit the following documents on every
defendant sentenced under the guidelines:

 indictments,

 presentence report, Objectives, Scope, and

Methodology

B-279743 Page 8 GAO/GGD-99-70 Federal Courts

 report on the sentencing hearing,

 written plea agreement, and

 judgment of conviction. Data from these documents are extracted
and coded for input into USSC's databases. We did not
independently verify the accuracy of the USSC data. However, USSC
takes several steps to insure the reliability and completeness of
the data system.

To identify the differences across the judicial circuits and
districts in the percentage of offenders who are ordered to pay
fines or restitution and those who are not, we used USSC's data
for fiscal year 1997 to identify the percentage of offenders who
received a fine or restitution by court district. Because we found
that type of offense was strongly related to whether fines or
restitution was ordered, we also analyzed the differences among
judicial circuits and districts in fines and restitution ordered
by selected types of offenses. To determine the percentage of
offenders ordered to pay fines, we selected larceny, fraud, and
drug trafficking crimes because of the large number of offenders
sentenced under USSC guidelines in federal courts for these
crimes. To determine the percentage of offenders ordered to pay
restitution, we selected robbery, larceny, and fraud crimes
because (1) there were a large number of offenders sentenced in
federal courts and (2) these crimes involve an act upon another
person, so we believed there was the reasonable expectation of an
identifiable victim.

To identify possible reasons for sentencing differences among
judicial circuits and districts, we analyzed the overall percent
of fines and restitution ordered by circuits and districts and for
the specific types of crimes we chose. In performing our analysis,
we first considered all federal offenders and how the likelihood
of being ordered to pay fines or restitution was affected by

 selected demographic characteristics of the offenders (sex, race,
citizenship, education, and number of dependents);

 the type of offense they committed (whether it involved property,
drugs, firearms, fraud, immigration, a violent or other offense);

 characteristics of the offender's sentence (whether it occurred
before or after MVRA was enacted, whether the offender was
sentenced to prison, probation, or an alternative sentence
imposed, and whether there was more than a single count of
conviction); and

 the circuit and district in which the sentencing occurred.

B-279743 Page 9 GAO/GGD-99-70 Federal Courts

We found that a number of factors were strongly associated with
whether fines or restitution were ordered and that the type of
offense was one of the most important. Therefore, we conducted a
series of multivariate statistical analyses for specific types of
offenses to estimate the effects of the other factors on fine and
restitution orders, independent of one another. We used logistic
regression techniques to estimate, simultaneously, the effects of
these different factors on the odds of an offender being ordered
to pay restitution among larceny, robbery, and fraud offenses and
then to estimate the effects of these factors on the odds of an
offender being ordered to pay fines among larceny, drug
trafficking, and fraud offenses. 3 In one series of regression
models, we employed circuit as one of our independent variables to
determine how much variation across circuits in the odds on
ordering fines and restitution persisted for these types of
offenders after other factors were controlled. In a second series
of regressions, we restricted our attention to offenders in the 10
largest districts, based on the number of offenders sentenced
during fiscal year 1997 for each of the 4 types of offenses
selected, to get a sense of the extent of variation across
districts. (It was not always the same 10 districts for each type
of crime.) There were not sufficient numbers of offenders ordered
to pay fines for robbery offenses or restitution for drug
trafficking offenses in the districts for us, in our opinion, to
reliably perform the multivariate statistical analyses in those
cases. We also believed it would be too time consuming to perform
the multivariate statistical analyses for offenders sentenced for
all types of federal offenses separately in all 94 districts.

We interviewed officials from the DOJ's Executive Office of the U.
S. Attorneys, the Administrative Offices of the U. S. Courts
(AOUSC), USSC, and selected Federal District Courts to obtain
their opinions of why the differences existed in the percentage of
offenders ordered to pay fines and

3 The statistical terms probability, odds, and odds ratio are used
in this report. Probability, expressed as a proportion or
percentage, is the ratio of the number of outcomes that will
produce a specific event to the total number of possible outcomes.
For example, the probability of an offender being ordered to pay a
fine was 0. 19 (19 out of every 100 offenders, or 19 percent, were
so ordered.) However, probability is limited by the bounds of 0
and 100. For comparison purposes, odds are better than probability
because they are not restricted by these boundaries. Odds is the
ratio of the probability of an event occurring to the probability
of it not occurring. For example, the odds of an offender being
ordered to pay a fine was 0. 23 (19 of 100 ordered to pay divided
by the 81 of 100 who were not). More simply, odds of 0.23 means
that 23 were ordered to pay a fine for every 100 who were not.

We compare odds by taking odds ratios. For example, the odds ratio
is determined by dividing the odds of an offender being ordered to
pay restitution in a circuit by the odds of an offender being
ordered to pay restitution in the referent circuit. The referent
circuit is the circuit in which an offender has the lowest odds of
being ordered to pay restitution before controlling for other
factors that might affect whether restitution is ordered. Odd
ratios are, perhaps, most easily understood in terms of times as
likely. For instance, an odds ratio of 4. 5 would be interpreted
as 4.5 times as likely as in the referent circuit.

B-279743 Page 10 GAO/GGD-99-70 Federal Courts

restitution and their views about the effect of mandatory victim
restitution. We limited our discussions to district courts in the
largest judicial districts, based on number of offenders sentenced
for each of the four types of offenses. We selected these
districts first because they were among the districts that met the
criteria for our multivariate statistical analyses. We then
narrowed our selection to seven districts in two parts of the
country where at least one district was adjacent to another and
there appeared to be variation in the percent of offenders ordered
to pay fines or restitution. Those seven districts were the
Eastern District of Pennsylvania, the District of New Jersey, the
Southern District of New York, the Eastern District of New York,
the Northern District of California, the Central District of
California, and the Southern District of California. At each
district, we interviewed the Chief Judge, the U. S. Attorney, and
the Chief Probation Officer or a representative selected by that
official. However, we did not review court case files at each
district. A review of court case files would have been time
consuming and might not have provided the reason why a fine or
restitution was or was not ordered in a particular case.

We did our work from August 1998 through January 1999 in
accordance with generally accepted government auditing standards.
We requested comments from DOJ, USSC, and AOUSC. We obtained
written comments from DOJ and USSC. AOUSC provided technical
comments. These comments are summarized at the end of this letter
and are contained in appendixes III and IV. All three agencies
provided technical corrections and suggestions.

According to our analysis of USSC data, of the approximately
48,000 federal offenders sentenced in fiscal year 1997 under USSC
guidelines, about 9,000 (19 percent) were ordered to pay fines and
9,600 (20 percent) were ordered to pay restitution. About 2
percent were ordered to pay both fines and restitution. The total
amount of fines and restitution ordered was over $1.6 billion
dollars.

The percent of federal offenders sentenced that were ordered to
pay fines and restitution varied substantially across the 12
federal circuits. Figure 2 shows the percent of offenders ordered
to pay fines ranged from 7 percent in the D. C. Circuit to 42
percent in the Seventh Circuit, which includes Illinois, Indiana,
and Wisconsin. Figure 3 shows the percent of offenders who were
ordered to pay restitution ranged from 15 percent in the Fifth
Circuit, which includes Texas, Louisiana, and Mississippi, to 32
percent in the Seventh Circuit. Table I. 2 in appendix I shows the
percent of offenders ordered to pay fines or restitution, by
judicial circuit. Numbers and Percent

of Federal Offenders Ordered to Pay Fines or Restitution

Percent of Federal Offenders Ordered to Pay Fines or Restitution
by Judicial Circuit or District

B-279743 Page 11 GAO/GGD-99-70 Federal Courts

Source: GAO analysis of fiscal year 1997 USSC data.

Figure2: Percent of Offenders Ordered to Pay Fines, by Circuit

B-279743 Page 12 GAO/GGD-99-70 Federal Courts

Source: GAO analysis of fiscal year 1997 USSC data.

The 94 federal judicial districts also differed greatly in the
percent of offenders ordered to pay fines, ranging from a low of 1
percent of the offenders sentenced in the Eastern District of
Oklahoma to a high of 84 percent in the Southern District of
Illinois. The rate at which offenders were ordered to pay
restitution also varied from 3 percent in the Southern District of
California to 49 percent in the Western District of Wisconsin. The
rate at which offenders were ordered to pay fines or restitution
might be three times or more greater in one district than in an
adjacent district. According to court officials, there should be
less variation in the types of offenses committed among adjacent
districts than there might be among districts located in different
parts of the country. For example, while 45 percent of the
offenders in the district of New Jersey were ordered to pay fines,
12 percent of the offenders in the Eastern District of New York,
which includes Staten Island and Long Island, were ordered to pay.
Similarly, while 22 percent of the offenders in the District of
New Jersey were ordered to pay restitution, 9 percent of the
offenders in the Eastern District of New York were ordered to pay.
Table I. 3 in appendix 1 shows

Figure3: Percent of Offenders Ordered to Pay Restitution, by
Circuit

B-279743 Page 13 GAO/GGD-99-70 Federal Courts

the variation by district in offenders ordered to pay fines or
restitution for all 94 judicial districts.

While many factors influence whether an offender was ordered to
pay a fine or restitution, the judicial circuit or district where
the offender was sentenced was a major factor. Our initial
statistical analysis of all 12 judicial circuits and 94 districts
showed major variation among the judicial circuits and districts
in the likelihood of an offender being ordered to pay a fine or
restitution. However, there was also a strong association between
the type of offense committed and the likelihood of an offender
being ordered to pay a fine or restitution. We then performed a
multivariate statistical analysis for offenders sentenced for four
types of offenses robbery, larceny, fraud, and drug trafficking
and controlled for such things, among others, as offender
characteristics, type of crime committed, length of sentence, and
type of sentence imposed such as prison, probation, or an
alternative. We selected these four types of offenses because they
were well- represented in the number of offenders sentenced in the
districts. We performed our multivariate statistical analysis for
the 10 largest districts in number of offenders sentenced under
the Guidelines for each of the 4 types of offenses during fiscal
year 1997.

Our multivariate statistical analyses for the four types of
offenses showed that the major variation among judicial circuits
and districts, although less, persisted and the likelihood of an
offender being ordered to pay fines or restitution for the same
type of offense could still be many times greater in one judicial
circuit or district than in another. We asked court officials and
prosecutors in seven districts for possible explanations of why
these differences might exist. We received some reasons related
both to the nature of the crimes and the types of offenders
sentenced in particular districts and to the culture of the courts
and the prosecutor's office in the districts such as how well
court officials and prosecutors work together to identify victims
and their losses.

Initially, we identified a number of potential reasons for the
differences between the judicial circuits and districts in the
likelihood of offenders being ordered to pay fines or restitution.
Judicial circuits and districts varied substantially in the types
of offenders who were sentenced. For example, the percentages of
offenders sentenced for drug trafficking, robbery, and fraud
varied by judicial circuit and district. Judicial circuits and
districts also varied according to the demographic characteristics
of offenders who were sentenced, such as the number of women,
minorities, or noncitizens. There were also differences in the
characteristics of offenders' sentences, such as whether they were
sentenced to prison or Differences Among the

Judicial Circuits and Districts in Ordering Fines or Restitution

B-279743 Page 14 GAO/GGD-99-70 Federal Courts

probation and the length of the sentence or amount of time the
offender would be in custody.

Our preliminary analysis showed that many of these characteristics
affected whether offenders were ordered to pay fines or
restitution. For example, the type of offense had a very
pronounced effect. While 6 percent of immigration offenders were
ordered to pay a fine, almost one- third of property offenders
were likewise ordered. Similarly, while 1 percent of drug
trafficking offenders were ordered to pay restitution, almost
twothirds of fraud offenders were so ordered. Similar disparities
in the likelihood of an offender being ordered to pay fines or
restitution existed among those sentenced for such offenses as
drug trafficking, fraud, and violence. Figures 4 and 5 show the
percentage of offenders who were ordered to pay fines or
restitution, by type of offense.

Source: GAO analysis of fiscal year 1997 USSC data.

Figure4: Percentages of Offenders Ordered to Pay Fines in Fiscal
Year 1997, by Type of Crime

B-279743 Page 15 GAO/GGD-99-70 Federal Courts

Source: GAO analysis of fiscal year 1997 USSC data.

Other factors related to the offender and the type of sentence
also produced variations in the likelihood of an offender being
ordered to pay a fine or restitution in our initial statistical
analysis for all 12 judicial circuits and 94 judicial districts.
For example, overall, females were more likely than males to be
ordered to pay restitution and blacks less likely than whites.
Also, overall, citizens were six times more likely to have
restitution orders than noncitizens. Similar race differences
existed in the likelihood of offenders being ordered to pay fines.
Additionally, offenders who were sentenced to probation were much
more likely to be ordered to pay a fine or restitution than those
sent to prison. Better- educated offenders also were more likely
to be ordered to pay a fine or restitution. Our comparison for
these factors is presented in detail in table I. 1 of appendix 1.

Given the effect of these characteristics, we did multivariate
statistical analyses that reestimated the differences in fines and
restitution across circuits and districts. In these analyses, we
controlled for those characteristics. These analyses took into
account differences in offenders' characteristics across these
judicial locations and the effects of these characteristics on
fine and restitution orders. We looked separately at fine orders
among larceny, drug, and fraud offenders and separately at
restitution orders for larceny, robbery, and fraud offenders. The
statistical

Figure5: Percentages of Offenders Ordered to Pay Restitution in
Fiscal Year 1997, by Type of Crime

B-279743 Page 16 GAO/GGD-99-70 Federal Courts

analyses are presented in full in appendix I. Figures 6 through 11
show the differences among the 10 largest districts in number of
offenders sentenced during fiscal year 1997 for fraud, larceny,
robbery, and drug trafficking.

Source: GAO analysis of fiscal year 1997 USSC data. Source: GAO
analysis of fiscal year 1997 USSC data.

Figure6: Percentage of Offenders Sentenced to Pay Fines for
Larceny Offenses in Selected Districts

Figure7: Percentage of Offenders Sentenced to Pay Restitution for
Larceny Offenses in Selected Districts

B-279743 Page 17 GAO/GGD-99-70 Federal Courts

Source: GAO analysis of fiscal year 1997 USSC data. Source: GAO
analysis of fiscal year 1997 USSC data.

Figure8: Percentage of Offenders Sentenced to Pay Fines for Fraud
Offenses in Selected Districts

Figure9: Percentage of Offenders Sentenced to Pay Restitution for
Fraud Offenses in Selected Districts

B-279743 Page 18 GAO/GGD-99-70 Federal Courts

Source: GAO analysis of fiscal year 1997 USSC data. Source: GAO
analysis of fiscal year 1997 USSC data.

Figure10: Percentage of Offenders Sentenced to Pay Fines for Drug
Offenses in Selected Districts

Figure11: Percentage of Offenders Sentenced to Pay Restitution for
Robbery Offenses in Selected Districts

B-279743 Page 19 GAO/GGD-99-70 Federal Courts

Even after controlling for the characteristics mentioned above, we
found that pronounced differences existed across judicial circuits
and districts for fine and restitution orders. For example, we
found that the likelihood of being ordered to pay restitution in
some districts among offenders convicted of robbery was five or
more times as high as in other districts. Drug trafficking
offenders were 10 or more times as likely to be fined in some
districts as in others. We found similar results for the other
types of offenders and for both fines and restitution, with
offenders in some districts many more times as likely to be
ordered to pay as in others.

Court officials and prosecutors we interviewed offered several
possible explanations why restitution might not have been ordered
in all cases. Some officials noted that there is rarely an
identifiable victim in drug offenses, making an order to pay
restitution unlikely. 4 These officials also identified bank
robbers as being among the poorest of offenders who usually lack
the ability to pay a fine, especially if restitution to the bank
is ordered. Offenders who commit immigration offenses most likely
are aliens who can be deported after being sentenced and serving
any time that may be ordered; collection of a fine might be
unlikely after deportation.

While these explanations offered some reasons why fines or
restitution might not be ordered in all cases, they do not explain
why when controlling for offender characteristics, length of
sentence, and type of sentence such as probation, prison, or an
alternative offenders have a far greater likelihood of being
ordered to pay a fine or restitution for the same type of offense
in one district than in another. We asked court officials and
prosecutors in seven of these districts what some possible
explanations might be for differences between the districts. These
7 districts were among the 10 largest in number of offenders
sentenced for each of the 4 types of offenses. The 10 largest
districts were not always the same 10 districts for each of the 4
types of offenses.

While some of these officials acknowledged that they did not know
the reasons, other officials offered an explanation that the
culture of the court and prosecutor's office within a district
could be a factor. The culture included factors such as how the
court views an offender's ability to pay fines, how well the civil
and criminal attorneys work together in the prosecutor's office,
or how well court officials and prosecutors cooperate

- 4 MVRA expanded discretionary restitution by creating community
restitution for victimless drug offenses (18 U. S. C. 3663c)
allowing for the possibility of restitution, even when an
identifiable victim is not involved. This provision was effective
for offenses on or after November 1, 1997, the date the revised
USSC Sentencing Guidelines went into effect.

B-279743 Page 20 GAO/GGD-99-70 Federal Courts

with each other in identifying victims and the amount of loss for
restitution.

Some officials who had worked in more than one district explained
that the culture was very different among the districts in which
they had worked. These officials explained that, when trying
cases, the diligence with which prosecutors identified victims and
their losses was greater in one district than in the other. In one
district, officials said there was a greater reluctance to pursue
restitution as part of the sentence because there was a strong
belief among prosecutors that the offenders would not, in most
cases, be able to pay it; whereas in another district, every
opportunity was considered by the prosecutors in the prospect of
an offender paying a fine or restitution as part of the sentence.
Some criminal prosecutors noted that in the district to which they
had transferred, civil attorneys sat in the same area as criminal
prosecutors; and the civil attorneys, who are more focused on
monetary issues, assisted criminal prosecutors in developing the
financial aspects, such as restitution, in criminal cases.

MVRA requires the court to order full restitution in certain
cases, including most federal offenses involving a crime of
violence or a crime against property. Prior to MVRA, the court
could waive restitution, in most cases, based on the offender's
ability to pay. MVRA's amendments are to be, to the extent
constitutionally permissible, effective for sentencing proceedings
in cases in which the defendant is convicted on or after the date
of enactment, which was April 24, 1996. However, because of an ex
post facto issue, DOJ has issued guidelines that any provisions of
MVRA for determining whether to impose restitution or the amount
of restitution would be applied only prospectively to offenses
committed on or after April 24, 1996. In general, the ex post
facto clause of the U. S. Constitution has been interpreted to
prohibit the application of a law that increases the primary
penalty for conduct after its commission.

Our statistical analysis showed that the percentage of offenders
ordered to pay restitution, overall, declined after MVRA went into
effect, while the results of the multivariate statistical analysis
for the four types of offenses were mixed on whether the
percentage of offenders changed after MVRA became effective.
Overall, for offenders sentenced during fiscal year 1997, 26
percent of offenders who were sentenced under the USSC Guidelines
for crimes committed before MVRA went into effect were ordered to
pay restitution, compared with 12 percent who were sentenced for
crimes The Effect of MVRA on

Restitution Orders

B-279743 Page 21 GAO/GGD-99-70 Federal Courts

committed after MVRA went into effect 5 (See table I. 1 in app.
I.) Our multivariate statistical analyses showed inconsistent
results across types of offenses. (See table I. 5 in app. I.) For
all districts, offenders who committed fraud offenses after MVRA's
effective date were about 80 percent as likely to be ordered to
pay restitution as those who committed offenses before restitution
became mandatory. Larceny offenders who were sentenced for crimes
committed after MVRA went into effect were about half as likely to
be ordered to pay restitution as those sentenced for crimes
committed before MVRA went into effect. Robbers who were sentenced
for crimes committed after MVRA went into effect were about one-
third more likely to be ordered to pay restitution than robbers
sentenced for crimes committed before MVRA went into effect.

In discussing our results, some court officials and prosecutors
said that it was still too early to assess the full impact of
MVRA. Some officials commented that time is needed to become
familiar with and implement MVRA, especially on the part of the
Assistant U. S. Attorneys who prosecute cases covered by MVRA.
Prosecutors in one district acknowledged that they were not yet
fully implementing the law. Most prosecutors, however, commented
that they did not remove an offender's responsibility for
restitution when drafting plea agreements. 6 The judges we
interviewed, in general, commented that they were implementing the
law and did not have constitutional questions with MVRA.

Although we selected larceny, fraud, and robbery because of the
likelihood of a victim being due restitution, a substantial
percentage of offenders about one- to two- thirds of offenders
sentenced were still not ordered to pay restitution, even if their
crimes were committed after MVRA was passed. Court officials and
prosecutors provided some reasons why restitution might not have
been ordered in these cases. In some cases, stolen money or assets
might have been recovered. In other cases, an offender might have
paid the restitution prior to sentencing, removing the need for a
restitution order. Another reason cited by officials was that the
offense might have been an attempted fraud or attempted robbery,
and the offender was arrested prior to obtaining any money from
the victim. Some officials also cited an exception to MVRA in
ordering mandatory restitution, such as in cases where the number
of victims is so large that it

5 For our analysis, for a crime to be considered subject to MVRA,
it had to be identified in the database as having been committed
on or after April 24, 1996. Overall, 55 percent of offenders in
our database committed crimes after MVRA became effective.

6 This is consistent with MVRA because the mandatory restitution
provisions of the act apply to plea agreements.

B-279743 Page 22 GAO/GGD-99-70 Federal Courts

makes paying restitution impracticable. One district had a number
of telemarketing schemes in which large numbers of victims were
defrauded of small amounts. It was not practical to identify all
victims and obtain restitution for them.

In response to our questions, Chief Probation Officers in the
Northern District of California, which includes San Francisco, and
the Central District of California, which includes Los Angeles,
provided information from local databases on bank robbery cases
where restitution had not been ordered. These databases were
different from the USSC database, and the information covered
different time periods. Information provided by the Chief
Probation Officer in the Northern District of California showed
that restitution was ordered in all bank robbery cases where there
was monetary loss. However, restitution was not ordered in bank
robbery cases where there was no loss or where the stolen money
was recovered. In the Central District of California, the primary
reason given when restitution had not been ordered as part of the
sentence was, as in the Northern District, that there was no loss
or that the stolen money had been recovered. We noted that, in
some cases, probation officers recommended restitution, which
would have been based on an actual loss, as part of the sentence;
but the judge did not follow the recommendation and did not order
restitution as part of the sentence. In other cases, the probation
officer did not recommend restitution, and the judge did not order
it. According to the Chief Probation Officer, the length of the
sentences over 80 years in prison in 2 of the cases might have
been a factor considered by the probation officer in not
recommending to the judge that restitution be part of the
sentence.

Although offender characteristics, type of offense, and the nature
of the sentence all played a role, the judicial circuit or
district where an offender was sentenced was a major factor in
determining the likelihood of an offender being ordered to pay a
fine or restitution during fiscal year 1997. This major variation
among judicial circuits and districts occurred overall for all
federal offenders sentenced under sentencing guidelines during
that year; and, although occurring less, this variation persisted
when we performed multivariate statistical analysis for federal
offenders sentenced under sentencing guidelines for four types of
offenses. The large statistical variation among circuits and
districts raises a question, on a broad level, about whether the
goal of uniformity in the imposition of fines and restitution is
being met. Offenders could be much more likely in some
jurisdictions than in others to be ordered to pay a fine or
restitution for the same type of crime. Conclusions

B-279743 Page 23 GAO/GGD-99-70 Federal Courts

Statistics for fiscal year 1997 are mixed on whether offenders
were more likely to be ordered to pay restitution after MVRA,
which was intended to eliminate much of the discretion judges
previously had in waiving restitution for certain types of crime.
Substantial percentages of offenders sentenced for crimes such as
fraud, robbery, and larceny which, by their nature, indicate a
need for restitution to a victim are not being sentenced to pay
restitution by the courts. However, it may be too early to see the
full impact of the results of the legislation; and there also may
be mitigating circumstances, such as recovery of stolen money, in
explaining why restitution was not ordered in a particular case.

We requested comments from DOJ, USSC, and AOUSC. DOJ and USSC
provided written comments on a draft of this report (see apps. III
and IV). AOUSC provided technical comments. All three agencies
provided technical corrections and suggestions.

DOJ generally agreed with the findings in the draft and advised us
of the steps that have been taken to help ensure that MVRA is
properly implemented. These steps include memoranda providing
guidance on MVRA and additional training efforts. DOJ's comments
recognize that the imposition of restitution is critical to their
law enforcement efforts and that, while a number of steps have
been taken, more remains to be done to increase the number of
cases in which restitution is imposed.

USSC raised several issues. First, USSC noted that training
efforts are planned or under way within the offices of the federal
courts to provide guidance and training on the correct application
of MVRA. Second, USSC also raised a general concern that the scope
and conclusions of the report rely perhaps too heavily on use of
the 1997 data. Although USSC acknowledges that 1997 data are the
only data currently available to study MVRA, they believe that
replicating the study using future data might mitigate the
idiosyncrasies in any given year's caseload. While the scope
suggested by USSC might be beneficial, USSC also does not provide
any evidence that expanding the scope would change our overall
conclusion that the judicial district or circuit where an offender
was sentenced was a major factor in determining the likelihood of
an offender being ordered to pay a fine or restitution. We
incorporated other suggestions by USSC as appropriate.

We have incorporated technical comments and suggestions from the
three agencies in the final report, as appropriate. Agency
Comments and

Our Evaluation

B-279743 Page 24 GAO/GGD-99-70 Federal Courts

We are sending copies of this report to Patrick J. Leahy, the
Ranking Minority Member of your committee, and Robert C. Scott,
the Ranking Minority Member of your subcommittee; Charles E.
Grassley, the Chairman, and Robert G. Torricelli, the Ranking
Minority Member, of the Subcommittee on Administrative Oversight
and the Courts, Senate Committee on the Judiciary; Henry J. Hyde,
the Chairman, and John Conyers, Jr., the Ranking Minority Member,
of the House Committee on the Judiciary; Leonidas Ralph Mecham,
Director, AOUSC; Timothy B. McGrath, Interim Staff Director, U. S.
Sentencing Commission; the Honorable Janet Reno, Attorney General;
and other interested parties. Copies will be made available to
others upon request.

Major contributors to this report are listed in appendix V. If you
have any questions about this report, please call me on (202) 512-
8777.

Richard M. Stana Associate Director Administration of Justice
Issues

Page 25 GAO/GGD-99-70 Federal Courts

*** End of document. ***