Update on the Cost, Quality, and Availability of Defense
Representation in Federal Death Penalty Cases
Preliminary Report on Phase One of the Research

Jon B. Gould
Lisa Greenman

In June 2008, the Judicial Conference Committee on Defender Services
approved dissemination of the Preliminary Report on Phase One

Table of Contents
I.

Introduction..........................................................................................................................1

II.

Decisions by the Department of Justice to Seek the Death Penalty: Their Implications
for Defense Costs.................................................................................................................3
A. Defendants Subject to Capital Prosecution: Potential Death Penalty Cases,
1989-2007 ................................................................................................................3
B. The Attorney General’s Decision-Making and How It Has Changed .....................4
C. Resolution of Authorized Cases...............................................................................7
D. Geographic Distribution of Authorized Cases.......................................................11

III.

Costs of Defending Federal Capital Cases.........................................................................16
A. Cases Examined .....................................................................................................16
B. Data Examined.......................................................................................................19

IV.

Findings..............................................................................................................................22
A. Total Case Costs ....................................................................................................22
B. Comparison to the Spencer Report ........................................................................24
C. Attorney Costs .......................................................................................................26
D. Expert Costs ...........................................................................................................29
E. Transcript Costs .....................................................................................................30

V.

Explanations for and Predictors of Case Costs ..................................................................31
A. Hypotheses.............................................................................................................31
B. Use of Experts........................................................................................................33
C. Case Characteristics ...............................................................................................35
D. Capital Culture .......................................................................................................36
E. Attorney General Making the Authorization Decision..........................................37
i

F. Geography..............................................................................................................38
G. Other Factors..........................................................................................................39
VI.

Influence of Case Costs on Outcome.................................................................................39

VII.

Conclusion .........................................................................................................................41

Appendix A: Spencer Report Recommendations .........................................................................43
Appendix B: Data Examined ........................................................................................................48
Appendix C: Supplemental Data ..................................................................................................49

ii

Figures and Tables
Figure or Table
Page
Figure 1: Potential Federal Capital Defendants, by Calendar Year......................................... 4
Figure 2: Federal Capital Authorization Decisions, by Authorization Year ........................... 7
Figure 3: Number of Federal Capital Defendants Tried, by Calendar Year............................ 8
Figure 4: Number of Federal Death Sentences, by Calendar Year........................................ 10
Table 1: Federal Death Penalty Prosecutions, 1989-2007 ..................................................... 11
Figure 5: Number of Authorized Capital Defendants by Circuit........................................... 12
Figure 6: Federal Capital Trials by State, 1989-1997............................................................ 14
Figure 7: Federal Capital Trials by State, 1998-2007............................................................ 15
Figure 8: Federal Death-Eligible Cases Studied, 1998-2004................................................. 18
Table 2: Total Cost for Defense Representation in Federal Capital Cases, 1998-2004 ......... 24
Figure 9: Cost Variance in Authorized Capital Cases, Trials vs. Pleas................................. 24
Table 3: Total Case Cost, Spencer Report Compared to Update............................................ 25
Table 4: Attorney Cost, 1998-2004 ....................................................................................... 26
Table 5: Attorney Hours, Spencer Report Compared to Update ........................................... 27
Table 6: Attorney In-Court Hours, Spencer Report Compared to Update ............................ 28
Table 7: Attorney Out-of-Court Hours, Spencer Report Compared to Update ..................... 28
Table 8: Expert Costs, Spencer Report Compared to Update................................................ 30
Table 9: Transcript Cost Per Defendant, 1998-2004 ............................................................. 31
Figure 10: Division of Expert Costs for Trial Cases.............................................................. 34
Figure 11: Division of Expert Costs for Plea Cases .............................................................. 34
Table 10: Factors that Predict Increased Case Cost, Bivariate Relationships ....................... 36
iii

Table 11: Effect of Attorney General's Decision-Making on Case Length ........................... 38
Table 12: Relationship Between Trial Case Cost and Death Sentence ................................. 40
Figure C.1: Number of Federal Capital Defendants Going to Trial/Trials Commencing,
by Calendar Year .................................................................................................................... 49
Figure C.2: Population of Federal Death Row by State and Circuit...................................... 50

iv

I.

Introduction 1
Ten years ago, in May 1998, the Subcommittee on Federal Death Penalty Cases of

the United States Judicial Conference Committee on Defender Services approved the
document entitled “Federal Death Penalty Cases: Recommendations Concerning the Cost and
Quality of Defense Representation,” which has become known as “the Spencer Report,” after
the Subcommittee Chair, Judge James R. Spencer. The Subcommittee’s recommendations
(provided as Appendix A, p. 43) were subsequently adopted by the Judicial Conference of
the United States in September 1998 and have been heavily relied upon by courts and
defense counsel since then.
Much of the Spencer Report remains as relevant in 2008 as it was in 1998 – for
example, its detailed description of the characteristics of federal death penalty prosecutions
and its discussion of the special duties such cases impose on defense counsel. Similarly, the
majority of its recommendations remain sound, as they endorse policies derived from core
values consistently enunciated by the Defender Services Committee: the need for high
standards for appointed counsel, fair funding for the defense that is predicated on sound casebudgeting practices, and an appreciation of the scope and depth of counsel's responsibilities.
The recommendations encourage the Administrative Office of the U.S. Courts to provide
training and technical support for court-appointed counsel, and promote strategic case
management and other approaches intended to contain costs while maintaining high-quality
representation.

1

For the research and writing of this report, the Office of Defender Services of the Administrative Office of the
U.S. Courts thanks Jon Gould, J.D., Ph.D., Director of the Center for Justice, Law, and Society at George
Mason University. Professor Gould was assisted by Lisa Greenman, staff attorney with the Federal Public
Defender Organization for the District of Maryland. Sylvia Fleming, Information Technology Specialist in the
Office of Defender Services, compiled extensive quantitative data from the CJA payment system, and Holly
Stevens, doctoral candidate at George Mason University, provided valuable technical and editorial assistance.

1

On the other hand, case-related data in the Spencer Report, including the cost
information frequently relied upon by defense counsel and the courts in assessing the
reasonableness of individual case budgets, is substantially outdated. The Spencer Report
examined cases from 1989 to 1997, the first years of the modern federal death penalty. 2
Only a limited number of cases had reached resolution in the trial courts by that time, and no
case had proceeded through the appellate and post-conviction stages. There is now an
additional decade of experience in the federal courts to draw upon, during which time there
have been significant changes in the way cases are charged, investigated, and litigated. As of
early 2008, more than 400 federal death penalty prosecutions have been authorized, resulting
in more than 175 federal death penalty trials involving about 230 defendants. Direct appeals
for approximately 60 death sentence cases have been resolved or are pending, and
approximately 30 capital post-conviction proceedings pursuant to 28 U.S.C. § 2255 have
concluded or are underway.
For these reasons, the Administrative Office's Office of Defender Services has
undertaken this re-examination of the cost, quality, and availability of defense representation
in federal death penalty cases since the time of the Spencer Report. Research for this update
is progressing in two phases. In this, Phase One, cost data on federal death penalty cases
between 1998 and 2004 have been collected and analyzed. The preliminary findings that
follow summarize quantitative data on the costs associated with capital representation and
employ statistical techniques to elucidate the factors that affect case cost. Phase Two will

2

Congress revived the federal death penalty in 1988, when it authorized capital punishment for the narrow
category of "drug kingpin" murders. 21 U.S.C. § 848, codifying Pub. L. 100-690, 102 Stat. 4382 (1988),
recodified in 2006 as 18 U.S.C. § 3599. In 1994, Congress passed the Federal Death Penalty Act, which
expanded the number of federal crimes punishable by death from one to approximately 50. 18 U.S.C. § 35913598, codifying Pub L. 103-322, 108 Stat. 1959 (1994).

2

examine these statistical findings through interviews and other qualitative research
techniques to understand the effect these factors may have on the cost, quality, and
availability of defense representation in federal death penalty cases. Phase Two also will reevaluate the Spencer Report's recommendations and, as appropriate, offer new
recommendations.
II.

Decisions by the Department of Justice to Seek the Death Penalty: Their
Implications for Defense Costs
A.
Defendants Subject to Capital Prosecution: Potential Death Penalty
Cases, 1989-2007
Although this study focuses primarily on what happens after the Attorney General has

authorized capital prosecution for a defendant ("authorized" cases), both the total number of
potential capital prosecutions ("death-eligible" cases) and the process by which the
Department of Justice chooses the defendants against whom it will seek the death penalty
(the death penalty "authorization" process) have profound consequences for defense
representation. 3 It is therefore necessary to develop an understanding of patterns and trends.
As previously noted, in 1994 Congress expanded the number of offenses punishable
by death from one to 50. As a result, in the second half of the 1990s there was a sharp jump
in the number of federal defendants potentially subject to the death penalty. According to the
Federal Death Penalty Resource Counsel Project, which monitors federal prosecutions for the
Defender Services program (see note 4, infra), there were 26 death-eligible defendants in
1993, 63 in 1994, and then upwards of 150 in almost every subsequent year.

3

The number of such "death-eligible" prosecutions affects the availability of qualified capital defense attorneys
and the cost of representation because, pursuant to 18 U.S.C. §3005, any defendant charged with an offense
punishable by death must be appointed two counsel, at least one of whom has special qualifications, and who
are paid, pursuant to 18 U.S.C. § 3599, at the increased rate applicable to death penalty cases. These
appointments remain in effect, and the higher rate of compensation applies, unless and until the Department of
Justice notifies the court that it will not seek the death penalty against the defendant.

3

Figure One illustrates this trend, showing the number of prosecutions alleging
offenses punishable by death between 1989 and 2007. These numbers should be viewed as a
good estimate, but not a precise count of each and every such case. 4
Figure 1:
Potential Federal Capital Defendants, 1989- 2007, by Calendar Year

,~

Ifolll

1+0110

~,.

HotSl

~

.

::~
B.

1m 11193.\194 Il1!l519!l111997199819!l!12lDl2lXl1 2lXtl2lXl32OOl200S::lOO81D07

Yellr

The Attorney General's Decision-Making and How It Has Changed

The Department of Justice does not permit a federal prosecutor to seek the death
penalty for a defendant unless specifically authorized to do so by the Attorney General of the

4

The Federal Death Penalty Resource Counsel Project provided the data underlying Figures One through
Seven. Since comparable numbers are not readily ascertainable from information made public by the
Department of Justice, a request for this information will be made in Phase Two of the research. The Federal
Death Penalty Resource Counsel Project, funded by the Administrative Office's Office of Defender Services,
provides training and advice to appointed counsel and the courts. (Spencer Report at 29.)

4

United States. 5 Prior to 1995, the Attorney General would review only those death-eligible
cases in which a United States Attorney sought permission to seek the death penalty. If local
federal prosecutors wished to seek the death penalty, the Attorney General could authorize it
or not. There was, however, no scenario in which a capital prosecution would be sought in
the absence of a local request. Furthermore, if the death penalty was authorized by the
Attorney General, discretion to resolve the case reverted back to the local U.S. Attorney, and
a plea agreement could be negotiated between the parties without further involvement from
the Attorney General.
Beginning in 1995, Attorney General Janet Reno promulgated a formal protocol that
centralized death penalty decision-making and required review of all death-eligible cases by
the Attorney General, regardless of whether the local U.S. Attorney viewed death as the
appropriate punishment. 6 For the first time, it became possible that capital prosecutions
would be ordered in cases in which they were not desired by local prosecutors; however,
following the authorization decision, discretion still returned to the local U.S. Attorney to
accept a plea bargain for a less-than-death sentence. In practice, according to the Federal
Death Penalty Resource Counsel, Attorney General Reno rarely exercised her authority to
overrule a U.S. Attorney's recommendation against the death penalty. When she did, each
such case was ultimately resolved through a negotiated plea agreement resulting in a
sentence less than death. These plea agreements were facilitated by the previously described
policy that returned decision-making authority respecting plea bargains to the local U.S.
Attorney's Office after an authorization decision was rendered by the Attorney General.

5

U.S. Attorney's Manual, Title 9-10.000.

6

Id.

5

Death penalty decision-making changed again in 2001, with the next Attorney
General, John Ashcroft, showing less deference to local prosecutors. The number of capital
prosecutions authorized without the recommendation of the local U.S. Attorney increased
substantially. 7 Attorney General Ashcroft also instituted a formal policy requiring approval
from Washington before an authorized case could be settled by plea agreement. Over time,
proportionally fewer cases have reached a negotiated resolution and a greater proportion of
cases have gone to trial. 8
Figure Two (p. 7) reflects the number of defendants for whom the Attorney General
authorized a capital prosecution each year between 1989 and 2007, a total of 435. The
pattern of prosecutions in Figure Two matches the trend in death-eligible offenses found in
Figure One (p. 4), with cases rising significantly following passage of the 1994 Federal
Death Penalty Act and then remaining at or above that level thereafter. Interestingly, the
number of cases authorized by the Department of Justice has varied from 2002-2007, rising
in 2003 and 2006 and dropping in 2007. 9

7

The Federal Death Penalty Resource Counsel Project has determined that Attorney General Reno required a
death penalty prosecution without a request for permission from the local prosecutors 14 percent of the time (26
of 182 authorization decisions) and that Attorney General Ashcroft did so 30 percent of the time (42 of 139
authorization decisions).

8

The Justice Department's reluctance to approve settlements proposed by local prosecutors has provoked
expressions of concern in Congress and from some judges and former U.S. Attorneys. See, generally,
Oversight of the Death Penalty: Hearing Before the Senate Committee on the Judiciary, 110th Cong. (2007);
John Gleeson, Supervising Federal Capital Punishment: Why the Attorney General Should Defer When U.S.
Attorneys Recommend Against the Death Penalty, 89 Va. L. Rev. 1697, 1699-1700 (2003); Adam Liptak, The
Death Penalty: A Witness for the Prosecution, N.Y. Times, Feb. 15, 2003, at B9.

9

The sharp decrease in 2007 coincides with the resignation of Attorney General Alberto Gonzales. As a point
of clarification, Figure Two does not reflect the total number of authorized federal death penalty cases pending
in the courts each year, but rather the number of times the Department of Justice authorized a capital
prosecution.

6

Figure 2:
U.S. Department of Justice Authorization Decisions, 1989-2007,
by Year of Authorization

D~
l_l~l~lmlml_l_l_l_l_l_~~~~~~~xm

-

"'"

,u

Year

C.

Resolution of Authorized Cases

By the end of 2007, 233 authorized cases were tried. 10 Figure Three (p. 8) shows
the number who went to trial each year between 1989 and 2007. There were a total of 176

10

The vast majority of these 233 defendants proceeded through both guilt and penalty phases, though in some
instances a two-phase capital trial was not completed (e.g., the guilt phase verdict acquitted the defendant of the
capital charge), and in others the case was resolved in another fashion (e.g., after trial commenced, there was a
negotiated guilty plea to a sentence less than death or withdrawal of the death penalty authorization). Cases are
assigned to the calendar year in which trial began.

7

trials for these 233 defendants, reflecting the fact that some trials involved more than one
capital defendant. 11
Figure 3:
Number of Federal Capital Defendants Tried, 1989-2007, by Calendar Year

~

~"
1+021 ...:1

~"

1+-141+-1'

- -

--

~"

..

.

~;

~;

•
•

~

~

~,

1.,990'll9l ,mll193'!i194 1_1_,997 19!1l119!l92lDl2lXl1 =21X13 :11D1200S::!lXI81Dt11

Year

The number of capital trials has risen at a relatively consistent rate between 1989 and
2007, even while the number of death penalty prosecutions authorized by the Attorney
General has varied. Stated differently, regardless of the number of federal capital
prosecutions authorized, fewer authorized defendants each year are resolving their cases
through pretrial guilty pleas, and a greater proportion are going to trial. As explained later,
capital trials are more expensive than cases resolved through plea agreements. Thus, if the

11

A chart showing both the number of capital trials and the number of defendants tried by calendar year is
provided in Appendix C (p. 49).

8

trend identified in Figure Three (p. 8) continues – i.e., fewer capital defendants plead guilty –
costs may continue to rise even if the total number of capital prosecutions stabilizes.
Among the authorized cases that proceeded to trial, only a quarter ended with the
defendant being sentenced to death (defendants almost never plead guilty and accept a death
sentence). Through the end of 2007, 61 of the 233 capital defendants who proceeded to trial
were sentenced to death. 12 Figure Four (p. 10) shows the number of death sentences
obtained each year from 1989 through 2007, and Table One (p. 11) provides summary data.
A chart reflecting the number of defendants presently under sentence of death, and the states
and federal circuits in which their cases were tried, is provided in Appendix C (p. 50).

12

Among the 435 defendants for whom capital prosecution has been authorized, the cases of many of those
authorized later in the time period are still pending. Of the 233 that went to trial, two defendants were
sentenced to death twice, the second death sentence following an appellate reversal and retrial; for purposes of
this report, each such defendant is counted only one time.

9

Figure 4:
Number of Federal Death Sentences, 1989-2007, by Calendar Year

•

--

• • '99011i19l ,mli93'!194 1995199619911!l91l1!l!l!l2lDl:.:Dl1 =2lIln ::llOoI3XIS:Dl8:rxn

Year

10

Table 1:
Federal Death Penalty Prosecutions, 1989-2007
Defendants Authorized
Capital Trials
Capital Defendants Tried
Defendants Sentenced to Death
Death Verdicts as a Percentage
of Capital Defendants Tried

D.

435
176
233
61
26%

Geographic Distribution of Authorized Cases

The geography of the federal death penalty has changed as well. In the first ten years
after the federal death penalty was reinstated, capital prosecutions clustered mainly in states
with active state court death penalty practices, such as Texas, Virginia, Missouri, and
Georgia. Subsequently, however, the Department of Justice has sought to eliminate this
geographic disparity by authorizing capital prosecutions more broadly. 13 Thus, an increasing
proportion of prosecutions has been brought in places with little or no death penalty

13

"[T]he goal of the Department's death penalty review and decision-making process is nationwide consistency
in the fair and even-handed application of federal capital sentencing laws in appropriate cases, irrespective of
geography or local predisposition for or against the death penalty." Statement of Barry Sabin, Dep. Ass’t Att’y
Gen. of the United States. Oversight of the Death Penalty: Hearing before the Senate Committee on the
Judiciary, 110th Cong. (2007).

11

experience, some of them jurisdictions that have chosen not to have a state death penalty. 14
Figure Five illustrates this trend. In the first decade of the federal death penalty, cases were
authorized in each of the 12 federal circuits, although the distribution was uneven. Capital
prosecutions were most numerous in the Fourth Circuit, while the Seventh Circuit had the
fewest authorized cases.
Figure 5:
Number of Authorized Federal Capital Defendants, 1989-1997 and 1998-2007,
by Circuit
Y.~rs

01989-1997
•

1998-2007

,•

~

•

~
~

••=~
~-~
••
•

~

~

10

Z

,.L.L

14

Puerto Rico, where a number of federal capital prosecutions have been authorized, has a constitutional
prohibition against the death penalty. The District of Columbia has rejected by referendum an effort to establish
a local death penalty. Additional non-death penalty states where federal capital prosecutions have been
authorized include Iowa, Michigan, West Virginia, Vermont, and Massachusetts. See also, Rory K. Little,
Good Enough for Government Work? The Tension Between Uniformity and Differing Regional Values in
Administering the Federal Death Penalty, 14 Fed. Sentencing Rep. 7 (2001); Benjamin Weiser & William
Glaberson, Ashcroft Pushes Executions in More Cases in New York, N.Y. Times, Feb. 6, 2003, at A1.

12

A decade later, the Fourth Circuit still had the largest federal capital caseload – about
one-fifth of all federal death penalty cases brought nationwide – but the proportional shares
among the circuits changed significantly. Whereas the Ninth Circuit had only four percent of
federal capital cases between 1989 and 1997, its proportion increased threefold to 13.8
percent of the national docket in 1998-2007. Similarly, the Second Circuit's percentage rose
from 8.5 percent to 12.5 percent of the national capital caseload, while the Fifth and Eleventh
Circuits represented a smaller share of the expanded national docket. In the period between
1998 and 2007, the number of authorized federal capital prosecutions increased in each
circuit except the Tenth.
These dynamics are displayed at the state level in Figures Six and Seven (p. 14 and
15). Although Figure Five (p. 12) includes all authorized cases, both pleas and trials, Figures
Six and Seven portray the geographic distribution of the trial cases, comparing the period of
the present study with the years addressed in the Spencer Report. As these maps illustrate,
the Department of Justice has brought the federal death penalty to a larger group of states in
the past decade, resulting in more cases being tried in jurisdictions that have not frequently
experienced death penalty prosecutions.

13

Figure 6: Number of Federal Capital Defendants by State, 1989-1997

NO.

Deferod~rot5

o

n

..,
Sl>'e

" Oel.

p
A~.R

tA

to

FL

,

liA

IL

KS

LA

, , ,

.,,'

HO

NJ

NN

NY

OK

,

P.\

1-3

~

4-5

!III

6-9

_

10-1 •

_

IS

.

TX

Of "'0'"

VA

"

14

Figure 7: Number of Federal Capital Defendants by State, 1998-2007

No. Defendants

·-,

·-.

10-15

"'ate
.Ocl.

= ..

•

o~'.

.., .., , ". ", ",
.,

"'•

..

<0

,.

.

..

, , "• .-, , ", , "', "'•
"', ", " "' " ", ",
'", ", '", n
" w•
"
"
"
oc

, ",

..

....
~

15

As later sections of this report explain, the changes initiated by Congress and the
Department of Justice have had profound effects on the cost of defending federal capital
cases. The Federal Death Penalty Act vastly increased the number of capital-eligible crimes,
but the impact could not be fully captured at the time of the Spencer Report, two years later.
In addition, in the past decade, the Department of Justice has authorized capital prosecutions
more frequently, and more often without the request of local prosecutors, than at an earlier
time. More of these cases are being brought in jurisdictions with little or no death penalty
experience, and a greater percentage of federal capital prosecutions are proceeding to trial
rather than being settled by plea agreement.
III.

Costs of Defending Federal Capital Cases
A.

Cases Examined

To analyze the cost of defense representation in authorized federal death penalty
cases, a database of such cases was developed. As with the Spencer Report, research focused
on panel attorney appointments – those cases in which a capital defendant is represented by
attorneys appointed pursuant to the Criminal Justice Act (CJA), whose compensation by the
federal courts is recorded in the CJA payment system. All federal death penalty
representations that began in 1998-2004 and concluded in the district court by the end of
2004 were examined, and almost all were included in the database. 15
Excluded from the database were those very few cases in which representation was
provided, in whole or in part, by privately retained counsel, or by counsel whose services

15

This time period immediately follows the years covered by the Spencer Report and represents a span in
which the maximum rate of compensation for defense counsel in panel appointments remained constant at $125
per hour (the same hourly rate that applied during the Spencer Report). In a few cases, payments were made for
services performed after December 31, 2004. Because the amounts were small and comparatively insignificant,
these cases were retained in the research sample.

16

were provided pro bono, without compensation. For purposes of cost analysis, cases in
which representation was provided, in whole or in part, by a federal defender organization
were excluded because the cost of representation would not have been captured by the CJA
payment system. These federal defender cases were identified, coded, and made part of the
database for other purposes, however. Federal defender organizations were involved in a
higher proportion of cases during the period of this study than during the period of the
Spencer Report, and the significance of this increased responsibility for representation will
be explored in Phase Two.
Also excluded from the database were authorized cases that did not proceed as capital
prosecutions through ultimate disposition, either because the death penalty authorization was
withdrawn by the Department of Justice or the notice of intent to seek the death penalty was
dismissed. 16 In both sets of circumstances, the representation would have then gone forward
as a non-capital matter and thus costs could not reasonably be compared with those in cases
that proceeded to conclusion as capital matters. However, as with the Spencer Report, a
companion sample of death-eligible, non-authorized cases was drawn in order to examine the
increased cost of representation when the Department of Justice authorizes a death penalty
prosecution.
All reported data refer to the defense of an individual defendant in a single case, or as
the CJA payment system uses the term, a "representation." When the term "case" is used

16

Authorization might be withdrawn if, for example, new information relevant to guilt or punishment
convinced prosecutors and the Attorney General that death was not an appropriate sanction. Notice of intent to
seek the death penalty might be dismissed upon the order of a judge for a legal reason, such as having been filed
too close to an already established trial date, or because the defendant was diagnosed with mental retardation.
Cases in which the defendant entered a plea of guilty to the death-eligible charge in exchange for the
prosecution agreeing to a life sentence and dismissing the death penalty request were retained in the sample.

17

(e.g., "median case cost"), it refers to the representation of a single defendant who may or
may not have been among multiple defendants joined in an indictment.
Figure Eight illustrates the dataset of federal death-eligible cases examined. For the
period 1998-2004, the study examined 214 cases, of which 119 were authorized as capital
prosecutions. Of these, 95 were CJA panel attorney representations and 24 involved
representation by federal defenders. Thirty-three of the 95 authorized panel cases were
resolved by plea agreement and 62 were tried. The costs from the 95 authorized panel cases
were compared with the costs in 95 death-eligible but not authorized CJA panel attorney
cases. 17

Figure 8
Federal Death-Eligible Cases Studied, 1998-2004
214 cases examined

119 authorized cases

95 CJA panel cases

95 unauthorized CJA panel cases

24 FDO cases

62 trials

33 pleas

17

The non-authorized cases were chosen at random from a list maintained by the Federal Death Penalty
Resource Counsel Project. Because these 95 cases were selected for comparison of costs with the 95 authorized
CJA panel attorney cases, only representations that did not involve a federal defender organization were
included.

18

B.

Data Examined

Phase One analyzes the cost of federal death penalty representations and examines
the factors that may influence case cost. For each of the 95 CJA panel attorney
representations reflected in Figure Eight, more than 40 sets of data were collected and coded.
The list of recorded data is delineated in Appendix B (p. 48). Initially, seven elements of
case cost were coded from data collected from the CJA payment system. These included: 18
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ

Total case cost
Cost of counsel
Cost of experts
Attorney hours
In-court attorney hours
Out-of-court attorney hours
Cost of transcripts
These cost variables ("dependent variables") were matched against factors that may

have influenced case cost ("independent variables"). The latter were collected from a variety
of sources, including the CJA payment system, the federal courts' electronic case
management system (PACER), and the Federal Death Penalty Resource Counsel Project. In
some cases, the attorneys of record were contacted to obtain necessary information.
Some of the independent variables are straightforward and easily understood:
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ

Defendant’s name
Circuit of prosecution
District of prosecution
State of prosecution
Whether the case was resolved by trial or plea
How the case concluded – whether acquittal, dismissal, or verdict of guilt
Defendant’s sentence (if not acquitted or dismissed)
Number of defendants named in the indictment

18

"Total case cost” reflects the combined costs of counsel and experts (or, in CJA terminology, "services other
than counsel"), but does not include transcript costs or certain travel costs that may be billed directly through
the court rather than through CJA vouchers. "Cost of transcripts" reflects payments to the official court reporter
for producing a record of court proceedings.

19

The remaining factors are more complex and are discussed in some detail in the
paragraphs following this list. They include:
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ

Number of defense counsel
Number of prosecutors
Number of defendants
Number of homicides alleged in the indictment
Race, gender, and ethnicity of defendant and victims
Involvement of victims in criminal activity
Prosecution’s allegation of "future dangerousness"
Number of offenses alleged in the indictment
Nature (magnitude) of offenses alleged in the indictment
Case length
Features of the Department of Justice’s capital authorization process
Experts utilized by the defense
State history and experience with death penalty litigation
The number of attorneys who staffed each side of a case was noted. The recorded

number reflects all lawyers whose appearance in the case, as reflected on PACER, lasted
longer than 30 days. In all cases, defendants were represented by at least two lawyers
(consistent with the requirements of 18 U.S.C. §3005), and in some cases more lawyers
entered appearances, either to substitute for or to supplement the two statutorily required
counsel. These numbers do not reflect additional counsel, if any, permitted to assist
appointed counsel for limited purposes. The number of prosecutors showed greater
variability, ranging from one to as many as seven lawyers who made a court appearance and
remained in the case for longer than a month.
The numbers of defendants and of charged homicides in each representation were
coded. These data included the number of codefendants joined in the case, regardless of
whether they were charged with a capital crime, as one measure reflecting the size of the
case. Also recorded was the number of homicides attributed in the indictment to the
defendant named in the representation, as an indicator of the scope of criminal responsibility
for that defendant. (The number of homicides alleged as "aggravating factors" in support of
20

a death sentence were not included in this figure, as the data could not be easily verified.)
The race, ethnicity, and gender of victims and defendants, as well as whether victims were
themselves connected to criminal activities, were recorded. Where prosecutors alleged
"future dangerousness" of the defendant as an aggravating factor in support of the death
penalty, these data were noted.
The total number of indicted offenses in each case, whether charged against that
particular capital defendant or a codefendant, was recorded. In most cases, the capital
defendant faced the greatest number of charges, but when he did not, this variable was
another measure of the size and scope of the case. A category was created to note three types
of indictments that are generally complicated and multifaceted, indicating case magnitude:
continuing criminal enterprise (CCE), racketeering (RICO), and terrorism.
Several measures reflecting the length of a case were recorded. The date of the first
appointment of counsel was noted as a "start date." The date on which the government filed
notice of its decision to authorize capital prosecution was recorded, as was the date of
conclusion in the district court, whether by acquittal, dismissal, or sentence. The data also
indicate which of two Attorneys General authorized the capital prosecution (the study period
encompassed part of the tenure of Attorney General Reno and the full tenure of Attorney
General Ashcroft).
Several variables were collected to reflect the use of experts in each representation.
The term "experts" is used here, as in the Spencer Report, to embrace the "services other than
counsel" that are authorized under the Criminal Justice Act. 18 U.S.C. § 3006A(e). Thus, in
addition to professionals traditionally thought of as experts, such as doctors or scientists, the
term includes fact investigators, paralegal assistants, reproduction services, as well as
fingerprint examiners, pathologists, mitigation specialists, jury consultants, etc. Information
21

was drawn from the CJA payment system reflecting notations contained in the form CJA 31
that defense counsel use when itemizing case expenses. The CJA 31 provides 24 potential
categories of services other than counsel that may be used in a representation. 19 The data
include these itemized costs from each case.
Finally, data were collected to reflect the history and experience of states with the
death penalty. These variables are used later to examine the relationship between state
capital practice and case cost in federal capital prosecutions. Data were collected indicating
whether there was a state death penalty statute in the jurisdiction where the federal
prosecution was brought; the number of years following Furman v. Georgia 20 before that
state reinstituted the death penalty; the number of years following Gregg v. Georgia 21 before
the state executed a defendant; 22 the number of defendants on that state's death row as a
percentage of its population; the number of executions in the state post-Gregg as a
percentage of its population; and the size of the state’s death row as related to its crime rate.
IV.

Findings
A.

Total Case Costs

The costs of defending federal death-eligible cases spanned a wide range between
1998 and 2004. As Table Two (p. 24) indicates, the cost of death-eligible, non-authorized

19

The CJA 31 was revised to capture these 24 categories in 1995, and includes a category designated as
"other," for types of services not identified. Prior to that, only twelve categories (including "other") were
available on the form.
20

Furman v. Georgia, 408 U.S. 238 (1972), held death sentencing cruel and unusual under the Eighth
Amendment based on its randomness, effectively bringing capital punishment to a halt while states re-evaluated
their sentencing procedures.

21

Gregg v. Georgia, 428 U.S. 153 (1976), upheld the constitutionality of capital sentencing statutes drafted to
comply with Furman by providing for guided decision-making.
22

In states that have not reinstituted the death penalty, these figures respectively reflect the time between
Furman and Gregg and the present.

22

cases ranged from a high of $681,556 to a low of $1,613. The median cost was $44,809.
Because the median reflects the middle case of a sample, it is generally a better measure of a
“typical” case than an average, or mean cost figure. 23 Nevertheless, Table Two (p. 24) also
provides the mean cost, which at $76,665 indicates that costs varied more extensively in
cases above the median than in cases below the median (or, put another way, the highest cost
cases were further from the median than were the lowest cost cases).
Authorized cases are substantially more expensive than non-authorized cases. The
median amount of $353,185 for authorized cases in Table Two (p. 24) indicates that cases in
which a capital prosecution was authorized cost seven times more than those death-eligible
cases that were not authorized. Authorized cases ranged in cost from a high of $1,788,246 to
a low of $26,526. As this wide range from high to low is present in the non-authorized cases
as well, it probably reflects an inherently broad range of litigation complexity that exists in
all death-eligible cases, authorized and non-authorized. As Table Two (p. 24) shows, the
median cost for authorized cases that were tried ($465,602) is 2.3 times greater than for those
that were pled ($200,933). It is not surprising that trials would generally be more costly than
pleas. Nor is it surprising that, in the context of capital litigation, a case that is resolved with
a guilty plea is, nevertheless, resource intensive, as Figure Nine (p. 24) illustrates. As the
Spencer Report explained, in order to reach disposition in a capital case, defense counsel
must thoroughly investigate and prepare for trial. Figure Nine contains two plea cases that
are "outliers" by virtue of their high cost. If these are removed from the analysis, the range
of plea costs narrows considerably.

23

The median reflects the middle unit in a sample of cases. In a sample of 101 cases, for example, the 51st case
would be the median. The mean, or average, adds costs from all of the sampled cases and divides by the
number of cases.

23

Table 2:
Total Cost for Defense Representation in Federal Capital Cases, 1998-2004
Type of Case

Median

Mean

High

Low

$44,809

$76,665

$681,556

$1,613

$353,185

$491,905

$1,788,246

$26,526

Trials

$465,602

$620,932

$1,788,246

$67,366

Pleas

$200,933

$245,946

$1,174,942

$26,526

Not Authorized
Authorized

Figure 9:
Cost Variance in Authorized Capital Cases, Trials vs. Pleas (1998-2004)
Comparison of Total Cost of Federal Death Penalty Cases
by Type of Proceeding
2,000,000

1,800,000

1,600,000

Cost (dollars)

1,400,000

1,200,000
Plea
1,000,000
Trial
800,000

600,000

400,000

200,000

0

B.

Comparison to the Spencer Report

Case costs have changed substantially since the period covered by the Spencer
Report. As Table Three (p. 25) shows, costs have risen for each category of death-eligible
cases, whether authorized or not, trials or pleas. Unfortunately, it is not possible to offer a

24

direct comparison between current data and those compiled in the Spencer Report. Although
the Spencer Report published mean data on case cost, it excluded from its analysis some
exceptional cases. The Spencer sample did not include certain cases that had atypically high
costs (e.g., the Oklahoma City bombing cases), and included cases that would have been
excluded from the present study because they had no penalty phase trial (for example, the
jury's guilt phase verdict precluded the death penalty, or a guilty plea to a non-death sentence
was entered before verdict).
It can be stated, however, that costs have risen across all categories of cases and that
they have risen more considerably among authorized cases than in non-authorized cases, and
in trials more than in pleas. Indeed, the most significant change in case cost between the
time of the Spencer Report and the current update is that the cost of trials has risen
considerably. Not only did more federal capital cases proceed to trial between 1998 and
2004 than at the time of the Spencer Report (Figure Three, p. 8), but also the cost of those
trials was higher (Table Three). A later section of this report offers preliminary explanations
for these developments, which will be explored in greater depth in Phase Two.
Table 3:
Total Case Cost – Spencer Report Compared to Update
Type of Case

Not Authorized
Authorized
Trials
Pleas

1998-2004
Mean
Median
$76,665
$44,809
$491,905
$353,185
$620,932
$465,602
$245,946
$200,933

Spencer Report
Adjusted Mean Only
$55,773
$218,113
$269,139
$192,333

Some of the increase in total case costs is probably attributable to a rising Consumer
Price Index (CPI) between the time of the Spencer Report and the present update. Over the
25

period of this update, the CPI rose at an annual rate of 2.45 percent. Inflation does not
explain the entire rise in case costs since the time of the Spencer Report, nor does it
distinguish between rates for authorized and non-authorized cases or pleas and trials, but the
CPI does suggest that some portion of rising costs is due to forces outside of litigation.
Although the maximum hourly rate for defense counsel remained fixed between both
periods, expert expenses, attorney travel, and the cost of associates and paralegals were
subject to inflation.
C.

Attorney Costs

Defense counsel's time, both in- and out-of-court, and the use of experts on behalf of
the defense influence case cost. Tables Four through Eight present these findings from the
current study, and (in all but Table Four) compare these data with those presented in the
Spencer Report. (It is not possible to compare total attorney cost from the most recent cases
to those at the time of the Spencer Report because the earlier study did not produce these data
as a portion of total costs.) Table Four provides data on attorney cost in the 1998-2004
sample. Similar to total case cost, the median expense for an attorney’s time was
significantly greater – 6.5 times more – in authorized cases than in death-eligible, nonauthorized prosecutions. Attorney cost in authorized cases was 2.8 times greater in trials
than for pleas.
Table 4:
Attorney Cost – 1998-2004
Type of Case

Median

Mean

Not Authorized

$42,148

$62,336

Authorized

$273,901

$363,776

Trials

$352,530

$462,037

Pleas

$122,772

$176,464

26

These same trends are reflected in Table Five, which presents the total attorney
hours. In death-eligible cases between 1998 and 2004, defense attorneys spent 4.6 times
more hours on authorized than non-authorized cases (comparing medians), and 2.6 times
more time on trials than on pleas for authorized cases. The fact that the ratio between
authorized and non-authorized cases is greater in Table Four (p. 26) than in Table Five
probably reflects the fact that, in most instances, attorneys are paid a lower hourly rate in
death-eligible cases that are not authorized as capital prosecutions than in those cases that are
authorized. In addition, attorney cost includes certain travel and other expenses that do not
correlate with hours, and these are presumably higher in authorized cases than nonauthorized matters because of the more extensive nature of capital as opposed to non-capital
defense.
Table 5:
Attorney Hours – Spencer Report Compared to Update
Case Type

Not Authorized
Authorized
Trials
Pleas

1998-2004
Mean
Median
637
436
2,815
2,014
3,557
2,746
1,403
1,028

Spencer Report
Adjusted Mean Only
429
1,464
1,889
1,262

As previously mentioned, it is difficult to make a direct comparison to the Spencer
data, since the information there reflects adjusted mean costs. Still, it is instructive to note
that attorney hours have risen over time for each type of case, but much more significantly in
authorized cases that go to trial. Whatever forces drive case cost – which this report
addresses in a later section – they appear to influence attorney hours more extensively in
authorized capital trials than in other representations.
27

Tables Six and Seven compare attorneys’ time in- and out-of-court when representing
death-eligible defendants. Attorneys spent more time in- and out-of-court in authorized cases
than in non-authorized matters and more in trials than pleas. Moreover, attorney time has
risen in all forms of representation from the time of the Spencer Report (even accounting for
the hybrid quality of the Spencer data). The data clearly show that the ratio of attorney time
spent in-court to out-of-court is greater for trials than pleas. In capital trials, defense counsel
spent a median 353 hours in-court and 2,373 hours out-of-court. In pleas, they spent a
median 42 hours in-court and 992 hours out-of-court.
Table 6:
Attorney In-Court Hours – Spencer Report Compared to Update
Case Type

Not Authorized
Authorized
Trials
Pleas

1998-2004
Mean
Median
106
34
401
306
537
353
142
42

Spencer Report
Adjusted Mean Only
38
231
409
61

Table 7:
Attorney Out-of-Court Hours Spencer Report Compared to Update
Case Type

Not Authorized
Authorized
Trials
Pleas

1998-2004
Mean
Median
531
350
2,414
1,645
3,019
2,373
1,261
992

Spencer Report
Adjusted Mean Only
391
1,233
1,480
1,201

28

D.

Expert Costs

The use of experts has a substantial influence on case cost. As Table Eight (p. 30)
indicates, experts were utilized in both authorized and non-authorized cases. There is a
significant difference, however, in the prevalence, and hence cost, of expert assistance
between authorized and non-authorized cases. Whereas the median for expert costs was
$5,275 in non-authorized cases, it was $83,029 in authorized cases. Further, experts were
utilized more extensively in capital trials, where the median cost for experts was $101,592,
than in pleas, where the median was $42,049.
Although the comparison between the Spencer Report and the present update is
imperfect, Table Eight (p. 30) suggests expert costs have risen substantially in capital trials.
As discussed in Section V, this trend likely reflects, among other developments, the
requirements for effective defense representation set forth by the U.S. Supreme Court in
recent years in cases such as Wiggins v. Smith, 539 U.S. 510 (2003). 24 In addition, there has
been a geographic shift in where federal capital prosecutions are authorized, and an increase
in the complexity of prosecutions, which also may influence the use and cost of experts.

24

Wiggins discussed defense counsel's responsibility to conduct a full social history investigation of a capital
defendant and endorsed the American Bar Association's standards as guidance as to what constitutes reasonable
performance of counsel in a death penalty case.

29

Table 8:
Expert Costs – Spencer Report Compared to Update
Case Type

Not Authorized
Authorized
Trials
Pleas

E.

1998-2004
Mean
Median
$14,330
$5,275
$128,129
$83,09
$158,895
$101,592
$69,482
$42,049

Spencer Report
Adjusted Mean Only
$10,094
$51,889
$53,143
$51,028

Transcript Costs

Finally, another contributor to the cost of defense representation is the expense of
acquiring the transcript of district court proceedings. 25 These figures are not included in the
data in Table Two (p. 24) that report total case cost; however their expense, especially in
capital trials, is noteworthy, with a median cost per capital trial defendant of $10,269. Table
Nine (p. 31) presents data on transcript costs per representation, which show a sizeable
difference between costs expended for transcripts in capital trials when compared to other
proceedings. It should be noted that transcript costs are incurred separately by each
defendant in a case (as well as by the court and by the prosecution). That the mean expenses
for non-authorized cases ($4,144) and capital pleas ($1,337) are substantially higher than the
median costs in these cases ($210 and $82, respectively) indicates that there are some
unusual cases in these categories that have substantially higher transcript costs than the other

25

Although transcripts of all pretrial and trial proceedings must be produced for an appeal, they are also
typically ordered for counsel's use during the pendency of a case in the trial court, as verbatim records are relied
upon for pleadings, examinations, and arguments. Because some transcript costs reflected in the CJA payment
system for the cases included in this analysis may have been incurred in connection with an appeal, while others
were generated during the course of the trial representation, they were not included in the "total case costs."
Regardless of when in the process they are incurred, the costs of transcript production are an expense associated
with capital trials.

30

more “typical” cases. By contrast, whereas some transcripts in capital trials cost more than
the median case, the relative similarity between mean and median costs ($16,487 versus
$10,269) shows that most capital trials carry a considerable transcript expense.

Table 9:
Transcript Cost Per Defendant, 1998-2004
Case Type

Median

Mean

$210

$4,144

$5,223

$11,274

Trials

$10,269

$16,487

Pleas

$82

$1,337

Not Authorized
Authorized

V.

Explanations for and Predictors of Case Costs
A.

Hypotheses

Numerous factors likely account for the increase in overall case costs in the ten years
following the Spencer Report. The significance of some, like the changing geography of
authorized capital prosecutions, can be established by analyzing the data presently on hand,
and others remain hypotheses that will be explored in Phase Two. Although a fuller
discussion will be made in the final report, several factors are presently clear.
ƒ

Inflation. Some of the increase in case costs is likely related to inflation. The rate of
inflation from 1998 to 2004 averaged approximately 2.45 percent annually, which
over six years could compound to approximately 15 percent. Although the CJA
hourly rate of $125 remained constant between the time of the Spencer Report and
this update, all costs other than legal fees, including those of experts, paralegals,
associate counsel, travel, copying, technology, etc., were subject to inflation.

ƒ

Differences in cases studied. The sample of cases in the Spencer Report was
constructed differently from the present sample. For example, the Spencer Report
excluded certain high cost cases, like the two representations arising out of the
Oklahoma City bombing, and included certain cases that did not proceed through a
penalty phase trial, either because the government withdrew its request for the death
penalty or because the defendants were acquitted of the capital charge. (See Spencer
31

Report at 8, n. 13.) For those reasons, costs reported in the Spencer study may have
appeared lower than they would have using the current methodology.
ƒ

Expanded areas of litigation. Defense practice has continued to evolve and become
more sophisticated since the first decade of the federal death penalty. For example,
both the prosecution and the defense now make greater use of scientific evidence and
experts and mount more extensive challenges to such evidence. The issue of the
defendant's future dangerousness within the Federal Bureau of Prisons is one such
area. Such changes in practice will be investigated through qualitative methods in
Phase Two.

ƒ

Pretrial litigation of mental health issues. As a result of changes in case law and new
Federal Rule of Criminal Procedure 12.2, governing mental health evaluations of the
defendant, both the prosecution and defense now engage in more extensive pre-trial
litigation over mental health issues. In addition to more time being spent on these
issues, the hourly rate paid to mental health experts appears to have risen significantly
since the time of the Spencer Report.

ƒ

Higher expectations for performance of defense counsel in death penalty cases. In
recent years, standards of practice for providing constitutionally adequate defense
representation have been clarified, particularly with respect to the duties of counsel
vis-a-vis investigating and presenting evidence in mitigation of sentence. See
Wiggins v. Smith, 539 U.S. 510 (2003) and the standards it endorsed, the American
Bar Association's Guidelines for the Appointment and Performance of Counsel in
Death Penalty Cases (1989) (Rev. ed. 2003).

ƒ

Geographic shift. As discussed earlier in this report, the Department of Justice has
changed its charging practices and authorized more capital prosecutions in areas that
historically have not embraced or have been resistant to the death penalty. Disparities
between local practice and federal policy may well raise the cost of litigation. A lack
of capital experience might require a greater investment of time to learn a new area,
for example. In addition, a higher proportion of cases being tried in jurisdictions
known for more extensive advocacy by both prosecution and defense, as discussed
below, would increase cost.

ƒ

Changing nature of capital prosecutions. Preliminary data suggest that a greater
proportion of the 1998 – 2004 cases involve highly complex issues at the guilt and
penalty phases than at the time of the Spencer Report, requiring more extensive
defense work. This hypothesis is supported by the increased number of in-court
attorney hours, indicating that substantially more court time is being required for
cases to reach resolution. This hypothesis, which will be delved into further in Phase
Two of the project, suggests that a greater proportion of the overall number of
prosecutions contains features that increase litigation costs. For example, many cases
in the current sample include all or some of the following features:
-

Multiple codefendants

32

-

-

Multiple homicides (in the indictment and/or as aggravating factors alleged in
support of the death penalty)
Complex offenses (such as CCE, RICO, and terrorism allegations that involve
numerous instances of conduct over an extended period of time and in
multiple locations, including outside the United States)
Foreign national and/or non-English speaking defendants, family members, or
victims
Voluminous discovery

B.

Use of Experts

-

-

Phase One connects several of the factors above to the cost of defending federal
capital cases. As Table Eight (p. 30) illustrates, expert costs were higher during the period
studied than during the period covered by the Spencer Report, but were the same percentage
of overall case costs. As Figures Nine and Ten (p. 34) demonstrate, almost 60 percent of
expert expenses were attributable to two types of experts, investigators and mitigation
specialists. (Payments reflected in the "investigator" category on the CJA 31 voucher may
reflect mitigation as well as guilt phase investigation.) Expenses for mitigation specialists
could not be broken out for the Spencer Report analysis – the CJA payment system did not
identify such costs at the time – but as the courts have come to authorize additional work for
this important part of capital representations, mitigation specialists have become a significant
portion (25 percent) of expert expenses. So too, investigative expenses represented almost
one-third of total expert costs, likely reflecting the additional work necessary to investigate
an increasingly complex set of capital authorizations, many involving multiple jurisdictions,
or defendants or witnesses who reside in other countries.
Interestingly, while the total cost of expert expenses is different in capital trials than
pleas, the proportions are similar. Regardless of whether it was resolved by trial or plea, a
capital case required the same type of preparation, including the use of experts. Table Eight
(p. 30) indicates that total expert expenses were almost 2.5 times greater in trials than pleas,
33

but Figures Nine and Ten demonstrate that in both types of authorized cases, investigation
and mitigation experts accounted for about 30 percent of expert costs.
Figure 10:
Division of Expert Costs for Trial Cases
8%
0.8%

Expert Type
Investigator - 32%

32%

Translator - 2%
Mental Health Experts - 13%

25%

Scientific Evidence Evaluation - 1.2%
Pathology & Medical - 4%
Computer - 0.2%
Legal & Paralegal Services - 8%
Jury Consultant - 6%

2%

Mitigation Specialist - 25%

6%
Duplication Services - 0.8%

13%

Other - 8%

8%
0.2%

4%

1.2%

Figure 11:
Division of Expert Costs for Plea Cases
Expert Type

10%

Investigator - 30%

2%
Translator - 1.1%

30%

Mental Health Experts - 10%
Scientific Evidence Evaluation - 2%
Pathology & Medical - 0.4%
Computer - 7%

27%
Legal & Paralegal Services - 10%

1.1%
Jury Consultant - 1.3%
Mitigation Specialist - 27%

10%
Duplication Services - 2%

2%

1.3%
10%

7%

Other - 10%

0.4%

34

C.

Case Characteristics

Quantitative research links higher costs with the presence of several factors that
reflect a representation's complexity. By testing these variables in bivariate, or one-to-one,
relationships, the following factors predicted increased case cost, whether in capital trials or
pleas:
ƒ
ƒ
ƒ
ƒ
ƒ
ƒ

Number of offenses
Number of defendants
Number of victims
Offense complexity
Number of defense counsel
Number of prosecutors
Most of these influences make intuitive sense and are shown in Table Ten (p. 36). As

the number of offenses, defendants, and victims rise in a case, so do the number of facts and
circumstances that counsel must investigate. In turn, they must devote additional attorney
and expert time to a case. Similarly, while every death penalty case is complex to defend,
certain offenses – in particular CCE, RICO, and terrorism – present exceptionally
multifaceted and wide-ranging fact patterns that require additional investigators, attorney
time, and expert consultation.
It may also seem apparent that the number of defense counsel is related to case cost:
generally, the greater the number of defense attorneys involved, the higher the cost. But as
data from both the Spencer Report and this update show, the number of prosecutors is an
even stronger predictor of case cost, at least in trials. Not only can the number of prosecutors
be an effective signal for the complexity of a case – presumably the Department of Justice
assigns additional prosecutors to those cases that are most involved and will require the
greatest effort – but defense teams also require additional attorney time and expert assistance
to keep pace with the prosecution’s resources.

35

Table Ten provides additional details on the relationships between these factors and
case cost, displaying all variables tested that had a statistically significant (i.e., reliable)
relationship to case cost, whether at trial or in a plea. It is apparent that case length is
positively correlated with case cost, meaning that the longer a case lasts the more expensive
it is. 26
Table 10:
Factors that Predict Increased Case Cost – Bivariate Relationships
Factor
Case Length
Case Length
Number of Defendants
Number of Defendants
Number of Victims
Number of Victims
Number of Offenses
Number of Offenses
Number of Prosecutors
Number of Defense Counsel
Offense Type
State Capital Culture

D.

Case Cost
Influenced
Trial
Plea
Trial
Plea
Trial
Plea
Trial
Plea
Trial
Trial
Trial
Trial

Bivariate
Correlation
.487
.428
.244
.748
.466
.557
.530
.414
.612
.480
.356
-.370

Statistical
Significance
.000
.015
.058
.000
.000
.001
.000
.019
.000
.000
.000
.007

Capital Culture

Table Ten also indicates that a state’s “capital culture” – its history and experience
with the death penalty – was negatively associated with trial costs. Federal capital trials are
more expensive when brought in districts where there is no state death penalty, or where the
state has neither sentenced to death nor executed many capital defendants. The specific
factors underlying this phenomenon will be explored in Phase Two.

26

In Table 10, the "Bivariate Correlation" column reflects the strength of the relationship between the variables
in the first two columns. A negative number reflects a relationship that reduces cost. The last column,
"Statistical Significance," indicates the degree of statistical significance the relationship has. Thus, "case
length" has a strong positive correlation with the cost of both trial and plea cases, and increases cost.

36

E.

The Attorney General Making the Authorization Decision

Case length is an important factor in determining cost, and case lengths varied
considerably depending on the Attorney General who authorized the capital prosecution. As
Table Eleven (p. 38) illustrates, trial cases authorized by Attorney General Janet Reno were
completed in a median 688 days, compared to 836 days in cases authorized by Attorney
General John Ashcroft. 27 Capital cases involving guilty pleas were also shorter when they
were authorized by Attorney General Reno, taking a median 578 days, compared to 744 days
under Attorney General Ashcroft. Put another way, the time required to prepare and
complete capital trials was 22 percent shorter, and capital pleas were 29 percent faster, when
begun under Attorney General Reno than Attorney General Ashcroft. Moreover, authorized
cases in this sample were much more likely to be resolved by a guilty plea when begun by
Attorney General Reno (27 trials, 31 pleas) than Attorney General Ashcroft (48 trials, 12
pleas). Since capital cases resolved by guilty pleas are less expensive to litigate than capital
trials (see p. 25), and cases of shorter duration are generally less expensive than those that
last longer (see p. 27), capital litigation was generally less costly for cases authorized by
Attorney General Reno than Attorney General Ashcroft.
Some of these differences between Attorneys General are explained by the amount of
time it took each to decide to authorize a capital prosecution, because the authorization
process was faster in the Reno Justice Department than it was under Attorney General
Ashcroft. As Table Eleven (p. 38) indicates, the Department of Justice took a median 178
days to authorize capital prosecutions under Attorney General Reno. With Attorney General

27

These statistics relate only to those death-eligible cases that were ultimately authorized for capital
prosecution. The sample analyzed here does not include case length data for death-eligible cases that were not
authorized.

37

Ashcroft, the comparable number of days was two-thirds greater, at 297. These distinctions
in case processing will be examined in Phase Two. 28
Table 11:
Effect of Attorney General's Decision-Making on Capital Case Length
and on Plea vs. Trial
(Includes cases litigated by federal defenders as well as panel attorneys)

Attorney General Number of
Median Case Number of
Who Authorized Capital Trials Length
Capital Pleas

Median Case
Length

Median Days
To Authorize

Janet Reno

27 cases

688 days

31 cases

578 days

178 days

John Ashcroft

48 cases

836 days

12 cases

744 days

297 days

F.

Geography

As discussed in Section D (p. 11), the Department of Justice has "nationalized"
federal death penalty prosecutions, increasingly authorizing federal capital cases in states that
have limited or no local death penalty experience. In the past decade, a greater proportion of
prosecutions have been both authorized and tried in cities such as New York, San Francisco,
Los Angeles, Washington, D.C., and Boston -- places known for especially vigorous and
hence more costly advocacy on both sides of the bar. 29
The Spencer Report found no regional cost variation, but noted its sample was too
small for meaningful analysis. Preliminary review of the current sample, which includes a
larger number of cases, suggests that there are significant regional variations in the cost of
capital trials. At this time, it is not known precisely what underlies these regional

28

In 2007, the Judicial Conference adopted a policy designed to facilitate prompt resolution of whether the
government will seek the death penalty. The policy urges judges to set reasonable deadlines for various stages
of the authorization process. Section 6.04, "Scheduling of Federal Death Penalty Case Authorization to Control
Costs," Guidelines for the Administration of the Criminal Justice Act and Related Statutes, Volume VII, Guide
to Judiciary Policies and Procedures.
29

Such differences in “local legal culture” have been well documented in socio-legal research, describing a
process by which common experiences of litigation become shared norms for the local legal market. See, e.g.,
Thomas W. Church, Jr. Examining Local Legal Culture, 10 Law and Social Inquiry 449 (1985).

38

differences; whether, for example, judges decline to fund defense requests for services,
defense counsel do not request or expend resources, or these cases are less complex. Phase
Two will try to illuminate the reasons for any regional variation in the cost of capital defense.
G.

Other Factors

Finally, it is important to note those factors that did not predict case cost. None of a
defendant’s recorded demographics – his race, gender, or ethnicity – affected the cost of
federal capital defense representation. The same was true for victims’ demographics (race,
gender, ethnicity) and their connection to criminal activity. 30 Finally, the prosecution’s
allegation of future dangerousness did not predict case cost.
VI.

Influence of Case Costs on Outcomes
This update has focused on the cost of defending capital prosecutions and the factors

that predict or explain case costs. These issues are vitally important, both to understand
which case characteristics affect the cost of defending federal capital matters and also to
recognize how and why these relationships have changed since the release of the Spencer
Report. During the process of analyzing data for this report, however, another issue revealed
itself: the relationship between case cost and the likelihood of receiving a death verdict.
During the period of this study, defendants who received the least amount of attorney and
expert time, and whose defense representation thus cost the least, faced a higher probability
of receiving a death sentence. Specifically, as Table Twelve (p. 40) shows, individuals
whose defense received less than $320,000 in combined attorney and expert assistance – the

30

Interestingly, the interplay between a defendant's and a victim's demographics influenced case costs, but the
small number of cases analyzed was insufficient to achieve statistical significance. In panel appointments,
median total case cost for white defendants tried for capital offenses against white female victims was
$631,382. By contrast, median total case costs for non-white defendants tried for capital offenses against white
female victims were $248,477. Larger sample size or qualitative research could help to elucidate the nature of
this relationship.

39

lowest one-third of federal capital trials – had a 44 percent chance of being sentenced to
death at trial. Individuals whose total representation costs were above that amount – the
remaining two-thirds of defendants – had a 19 percent chance of being sentenced to death.
Defendants in the low-cost group thus were more than twice as likely to be sentenced to
death.
Table 12:
Relationship Between Trial Case Cost and Death Sentence:
Lowest One-Third of Case Cost vs. Remaining 2/3
(Chi-square = .036)

Total Trial Cost

Sentenced to Death

Other Verdicts

Total Cases

Lowest Cost Cases

44 percent

56 percent

18

19 percent

81 percent

43

45

16

61

(under $320,000)
Remaining 2/3 of
Cases
Total Cases

The findings are similar when examining the influence of attorney costs and expert
expenses separately on trial verdicts, although because of the smaller numbers involved, the
results in both instances approach but do not reach the level of statistical significance.
Similarly, these relationships remain when controlling for the type of offense charged.
Examining those cases that did not involve RICO, CCE, or terrorism charges, 31 a defendant
had a 50 percent chance of being sentenced to death if the representation cost less than
$320,000. By contrast, a defendant had a 35 percent chance of receiving a death sentence if

31

Cost is not a direct correlate of outcome for indictments that allege CCE, RICO, or terrorism offenses,
largely because the total case cost in each of these representations is greater than $320,000.

40

the representation costs were more than $320,000. Phase Two will explore these findings in
greater detail. 32
VII.

Conclusion
Phase One of this update has examined the cost of defending federal capital cases

and identified several factors that affect cost. Although federal capital cases are now more
expensive to defend than at the time of the Spencer Report, the largest increase has been in
those federal capital cases that proceed to trial. Several factors influence and predict the
overall cost of defending federal capital cases, including the national inflation rate. Cases
with complex fact patterns, and those that involve several defendants, victims, and
prosecutors, are likely to be more expensive. But perhaps most predictive is the decision of
the Department of Justice to authorize capital prosecutions in particular states.
Phase Two will go beyond the available quantitative data to offer more detailed
observations about the sources of case cost while also considering issues of availability and
quality of counsel. Among such issues are continuity of counsel; counsel’s experience and
training; the availability of resources, including assistance from the Federal Death Penalty
Resource Counsel Project; the significance of federal defender involvement; case budgeting;
and the lessons learned from post-conviction proceedings. As with the Spencer Report,
judges and defense counsel will be interviewed and assistance from the Department of
Justice will be sought to further illuminate these issues. The final report will update the

32

In considering the cost of defense representation in cases where the possible outcome is death, it is
interesting to note the cost of defense representation in complex white collar prosecutions. The criminal
defense of Enron Chief Executive Officer Jeffrey Skilling reportedly cost $70 million, and that of former
HealthSouth CEO Richard Scrushy about $21 million. Kenneth Lay's defense in the Enron prosecution was
reportedly $25 million, and that of Cendant Corporation executive E. Kirk Shelton was $24 million. In 2007
litigation in the Southern District of New York, predictions of the cost of representation for individual former
partners at KPMG ranged from $10 million to $44 million. Beth Bar, Defense Cost Estimates Offered in KPMG
Case, New York Law Journal, July 13, 2007. http://www.law.com/jsp/article.jsp?id=1184231199215

41

findings and conclusions of the Spencer Report and also will examine the Report’s
recommendations.

42

APPENDIX A: SPENCER REPORT’S RECOMMENDATIONS
1. Qualifications for Appointment.
a. Quality of Counsel. Courts should ensure that all attorneys appointed in federal death
penalty cases are well qualified, by virtue of their prior defense experience, training and
commitment, to serve as counsel in this highly specialized and demanding type of litigation.
High quality legal representation is essential to assure fair and final verdicts, as well as costeffective case management.
b. Qualifications of Counsel. As required by statute, at the outset of every capital case,
courts should appoint two counsel, at least one of whom is experienced in and
knowledgeable about the defense of death penalty cases. Ordinarily, "learned counsel" should
have distinguished prior experience in the trial, appeal, or post-conviction review of federal
death penalty cases, or distinguished prior experience in state death penalty trials, appeals, or
post-conviction review that, in combination with co-counsel, will assure high quality
representation.
c. Special Considerations in the Appointment of Counsel on Appeal. Ordinarily, the
attorneys appointed to represent a death-sentenced federal appellant should include at least
one attorney who did not represent the appellant at trial. In appointing appellate counsel,
courts should, among other relevant factors, consider:
i. the attorney's experience in federal criminal appeals and capital appeals;
ii. the general qualifications identified in paragraph 1(a), above; and
iii. the attorney's willingness, unless relieved, to serve as counsel in any postconviction proceedings that may follow the appeal.
d. Special Considerations in the Appointment of Counsel in Post-Conviction Proceedings. In
appointing post-conviction counsel in a case where the defendant is sentenced to death,
courts should consider the attorney's experience in federal post-conviction proceedings and in
capital post-conviction proceedings, as well as the general qualifications set forth in
paragraph 1(a).
e. Hourly Rate of Compensation for Counsel. The rate of compensation for counsel in a
capital case should be maintained at a level sufficient to assure the appointment of attorneys
who are appropriately qualified to undertake such representation.

43

2. Consultation with Federal Defender Organizations or the Administrative Office.
a. Notification of Statutory Obligation to Consult. The Administrative Office of the U.S.
Courts (Administrative Office) and federal defender organizations should take appropriate
action to ensure that their availability to provide statutorily mandated consultation regarding
the appointment of counsel in every federal death penalty case is well known to the courts.
(See 18 U.S.C. § 3005.)
b. Consultation by Courts in Selecting Counsel. In each case involving an offense punishable
by death, courts should, as required by 18 U.S.C. § 3005, consider the recommendation of
the district's Federal Public Defender (FPD) (unless the defender organization has a conflict)
about the lawyers to be appointed. In districts not served by a Federal Public Defender
Organization, 18 U.S.C. § 3005 requires consultation with the Administrative Office.
Although not required to do so by statute, courts served by a Community Defender
Organization should seek the advice of that office.
c. Consultation by Federal Defender Organizations and the Administrative Office in
Recommending Counsel. In discharging their responsibility to recommend defense counsel,
FDOs and the Administrative Office should consult with Federal Death Penalty Resource
Counsel in order to identify attorneys who are well qualified, by virtue of their prior defense
experience, training and commitment, to serve as lead and second counsel.
3. Appointment of More Than Two Lawyers.
Number of Counsel. Courts should not appoint more than two lawyers to provide
representation to a defendant in a federal death penalty case unless exceptional circumstances
and good cause are shown. Appointed counsel may, however, with prior court authorization,
use the services of attorneys who work in association with them, provided that the
employment of such additional counsel (at a reduced hourly rate) diminishes the total cost of
representation or is required to meet time limits.
4. Appointment of the Federal Defender Organization (FDO).
a. FDO as Lead Counsel. Courts should consider appointing the district's FDO as lead
counsel in a federal death penalty case only if the following conditions are present:
i. the FDO has one or more lawyers with experience in the trial and/or appeal of
capital cases who are qualified to serve as "learned counsel"; and
ii. the FDO has sufficient resources so that workload can be adjusted without unduly
disrupting the operation of the office, and the lawyer(s) assigned to the death penalty
case can devote adequate time to its defense, recognizing that the case may require all
of their available time; and
iii. the FDO has or is likely to obtain sufficient funds to provide for the expert,
investigative and other services reasonably believed to be necessary for the defense of
the death penalty case.
44

b. FDO as Second Counsel. Courts should consider appointing the district's FDO as second
counsel in a federal death penalty case only if the following conditions are present:
i. the FDO has sufficient resources so that workload can be adjusted without unduly
disrupting the operation of the office, and the lawyer(s) assigned to the death penalty
case can devote adequate time to its defense, recognizing that the case may require all
of their available time; and
ii. the FDO has or is likely to obtain sufficient funds to provide for the expert,
investigative and other services reasonably believed to be necessary for the defense of
the death penalty case.
5. The Death Penalty Authorization Process.
a. Streamlining the Authorization Process. The Department of Justice should consider
adopting a "fast track" review of cases involving death-eligible defendants where there is a
high probability that the death penalty will not be sought.
b. Court Monitoring of the Authorization Process. Courts should exercise their supervisory
powers to ensure that the death penalty authorization process proceeds expeditiously.
6. Federal Death Penalty Resource Counsel.
a. Information from Resource Counsel. In all federal death penalty cases, defense counsel
should obtain the services of Federal Death Penalty Resource Counsel in order to obtain the
benefit of model pleadings and other information that will save time, conserve resources and
enhance representation. The judiciary should allocate resources sufficient to permit the full
value of these services to be provided in every case.
b. Technology and Information Sharing. The Administrative Office should explore the use of
computer-based technology to facilitate the efficient and cost-effective sharing of
information between Resource Counsel and defense counsel in federal death penalty cases.
7. Experts.
a. Salaried Positions for Penalty Phase Investigators. The federal defender program should
consider establishing salaried positions within FDOs for persons trained to gather and
analyze information relevant to the penalty phase of a capital case. FDOs should explore the
possibility that, in addition to providing services in death penalty cases to which their FDO is
appointed, it might be feasible for these investigators to render assistance to panel attorneys
and to other FDOs.
b. Negotiating Reduced Rates. Counsel should seek to contain costs by negotiating reduced
hourly rates and/or total fees with experts and other service providers.

45

c. Directory of Experts. A directory of experts willing to provide the assistance most
frequently needed in federal death penalty cases, and their hourly rates of billing, should be
developed and made available to counsel.
8. Training.
Federal Death Penalty Training Programs. The Administrative Office should continue to
offer and expand training programs designed specifically for defense counsel in federal death
penalty cases.
9. Case Budgeting.
a. Consultation with Prosecution. Upon learning that a defendant is charged with an offense
punishable by death, courts should promptly consult with the prosecution to determine the
likelihood that the death penalty will be sought in the case and to find out when that decision
will be made.
b. Prior to Death Penalty Authorization. Ordinarily, the court should require defense counsel
to submit a litigation budget encompassing all services (counsel, expert, investigative and
other) likely to be required through the time that the Department of Justice (DOJ) determines
whether or not to authorize the death penalty.
c. After Death Penalty Authorization. As soon as practicable after the death penalty has been
authorized by DOJ, defense counsel should be required to submit a further budget for
services likely to be needed through the trial of the guilt and penalty phases of the case. In its
discretion, the court may determine that defense counsel should prepare budgets for shorter
intervals of time.
d. Advice from Administrative Office and Resource Counsel. In preparing and reviewing
case budgets, defense counsel and the courts should seek advice from the Administrative
Office and Federal Death Penalty Resource Counsel, as may be appropriate.
e. Confidentiality of Case Budgets. Case budgets should be submitted ex parte and should be
filed and maintained under seal.
f. Modification of Approved Budget. An approved budget should guide counsel's use of time
and resources by indicating the services for which compensation is authorized. Case budgets
should be re-evaluated when justified by changed or unexpected circumstances, and should
be modified by the court where good cause is shown.
g. Payment of Interim Vouchers. Courts should require counsel to submit vouchers on a
monthly basis, and should promptly review, certify and process those vouchers for payment.
h. Budgets In Excess of $250,000. If the total amount proposed by defense counsel to be
budgeted for a case exceeds $250,000, the court should, prior to approval, submit such
budget for review and recommendation to the Administrative Office.

46

i. Death Penalty Not Authorized. As soon as practicable after DOJ declines to authorize the
death penalty, the court should review the number of appointed counsel and the hourly rate of
compensation needed for the duration of the proceeding pursuant to CJA Guideline
6.02.B(2).
j. Judicial Conference Guidelines. The Judicial Conference should promulgate guidelines on
case budgeting for use by the courts and counsel.
k. Judicial Training for Death Penalty Cases. The Federal Judicial Center should work in
cooperation with the Administrative Office to provide training for judges in the management
of federal death penalty cases and, in particular, in the review of case budgets.
10. Case Management.
a. Non-Lawyer Staff. Where it will be cost-effective, courts should consider authorizing
payment for services to assist counsel in organizing and analyzing documents and other case
materials.
b. Multi-defendant Cases.
i. Early Decision Regarding Severance. Courts should consider making an early
decision on severance of non-capital from capital codefendants.
ii. Regularly Scheduled Status Hearings. Status hearings should be held frequently,
and a schedule for such hearings should be agreed upon in advance by all parties and
the court.
iii. "Coordinating Counsel." In a multi-defendant case (in particular a multi-defendant
case in which more than one individual is eligible for the death penalty), and with the
consent of co-counsel, courts should consider designating counsel for one defendant
as "coordinating counsel."
iv. Shared Resources. Counsel for codefendants should be encouraged to share
resources to the extent that doing so does not impinge on confidentiality protections
or pose an unnecessary risk of creating a conflict of interest.
v. Voucher Review. In large multi-defendant cases, after approving a case budget, the
court should consider assigning a magistrate judge to review individual vouchers. The
court should meet with defense counsel at regular intervals to review spending in
light of the case budget and to identify and discuss future needs.
11. Availability of Cost Data
The Administrative Office should improve its ability to collect and analyze information about
case budgets and the cost of capital cases.

47

APPENDIX B: DATA EXAMINED

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

Defendant name
Circuit
District
Type of Counsel
Case proceeds to conclusion as death authorized
Case resolved by trial or plea
If tried, defendant guilty of the capital count
Sentence
Number of defense counsel
Number of prosecutors
Number of defendants
Number of indicted offenses
Category of offense
Exceptional offenses
Date defense counsel appointed
Date of authorization by DOJ
Attorney General who authorized
Date case ends (sentencing or acquittal)
Number of days from opening to authorization
Number of days from opening to closing of case
Number of victims
Categories of experts used
Expenses for each category of expert used
In-court hours by defense counsel
Out-of-court hours by defense counsel
Total hours by defense counsel
Total cost for experts and defense counsel
Total cost for experts
Total cost for defense counsel
Cost of transcripts
First and last dates of counsel payments
Victim’s connection to criminal activity
Government alleges future dangerousness against defendant
Race of Victim
Race of Defendant
Gender of Victim
Gender of Defendant
Whether state had the death penalty at the time of indictment
Number of years post-Furman that the state (re-)instituted the death penalty
Number of years post-Gregg that state executed its first prisoner
Number of prisoners executed by the state post-Gregg on a per capita basis
State’s current death row population on a per capita basis
State’s current murder rate

48

APPENDIX C: SUPPLEMENTAL DATA
Figure C.1
Number of Federal Capital Defendants Going to Trial/Trials Commencing,
by Calendar Year

o Numb'r 01 Otttndants
• Numb., oITriall

••

"~

,.•

~

,E

Z

•

••<
C•
:

15

~

•
."
•E

,

Z

,

49

Figure C.2
Current Population of Federal Death Row by State and Circuit

"
, t---------,
"

, ",
•,
<
0

•

c

•
·,
••
,c•
•

,
,
0
0

0

•

•,E
z

;
;

, , ,

,
, , , , ,
,

1IIIIIIr

,

-

-

-

-

-

,
---

,

, ,

C-[ ] I

, ,

, ,

-n-[I= -ttl=

-

-

-

-

Current as of June 1, 2008.
Total = 53 individuals presently on death row.

50