The Societal Cost of the

Exclusionary Rule:
An Empirical Assessment
Peter F. Nardulli

A key criticism that has emerged in the debate over the search and seizure exclu-
sionary rule is that it exacts heavy societal costs in the form of lost prosecutions
and that such costs outweigh any demonstrated social benefits. This article ex-
amines the costs of three exclusionary rules using data collected for 7,500 cases in
a nine-county study of criminal courts in three states. It emphmizes motions to
suppressphysical evidence but for comparativepurposes also includes motions to
suppress confessions and identifications. The results show that the various exclu-
sionary rules exact only marginal social costs. Motions to suppress physical evi-
dence are filed in fewer than 5 % of the cases, largely drug and weapons cases,
while serious motions to suppress identifications and confessions are filed in 2%
and 4% of the cases. The success rate of motions to suppress is equally marginal.
Successful motions to suppress physical evidence occur in only 0.69% of the
cases, while successful motions to suppress identifications or confessions occur
much less often. Moreover, not all who successfully suppressed evidence escaped
conviction, especially when only an identification or a confession was sup-
pressed. In all, only 46 cases-less than 0.6% of the cases studied-were lost-
because of the three exclusionary rules combined, most of them involving of-
fenses that would have incurred less than six months’ imprisonment or first of-
fenders. Finally, the impact of unsuccessful motions on subsequent plea bargain-
ing was found to be marginal; only unsuccessful motions to exclude confessions
resulted in any real sentencing concessions.

INTRODUCTION
The exclusionary rule is a legal device by which illegally obtained evidence
is made inadmissible at trial. Under the rule a defendant makes a motion to
Peter F. Nardulli is associate professor in the Institute of Government and Public Affairs and the
Department of Political Science at the University of Illinois, Urbana. B.A. 1969, Northern Illinois Univer-
sity; J.D. 1973, Ph.D. 1975, Northwestern University.
This research was made possible by grants from the National Institute of Justice (Grant No. 81-IJ-
CX-027) and the American Bar Foundation. The author also thanks his collaborators, James Eisenstein
and Roy Flemming, on the larger project from which this study is derived, for permitting him to pursue
this diversion. The overall data collection effort on which this analysis is based required extensive amounts
of field work by the three principal investigators and a staff of research assistants and data collectors. The
taped interviewsproduced over 10,ooO pages of transcripts. In addition, several years of effort have been
invested in data refinement, reduction, and analysis.
The point of view expressed here is the author’s and does not necessarily represent the views of the NIJ,
the ABF, or his collaborators.

0 1983 American Bar Foundation 585

586 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

suppress evidence alleged to have been illegally obtained. If the motion is

granted, the evidence in question is inadmissible for purposes of establishing
guilt. The rule has been in effect in federal courts since the 1914 decision in
Weeks v. United States, I and a number of state courts have also adopted
such a rule. * Exclusionary principles gained additional prominence in 1961
when the Supreme Court ruled in Mapp v. Ohio3that, in the case of physical
evidence, the rule must be applied in state criminal proceeding^.^ Subse-
quently, the exclusionary principle has also been applied to confessions’ and
identifications6 (i.e., lineup or photo identifications) that are illegally ob-
tained. Thus while the exclusion of physical evidence obtained by illegal
searches and seizures is often referred to as “the exclusionary rule,” in fact
there are three distinct exclusionary rules applying to physical evidence, iden-
tifications, and confessions. (For clarity, when I speak of “the exclusionary
rule” I am referring to the search and seizure rule. Otherwise I will refer to
the specific rules or the three “exclusionary rules” combined.)
Despite the fact that the principle had been law for years, the search and
seizure exclusionary rule experienced a renewed surge of criticism in the late
1960s. As Bradley Canon noted in 1974:
In the last few years, . . . the rule’s vitality has declined precipitously, and its
‘chances of reaching adulthood unimpaired can be diagnosed as uncertain at
best. Quite recently judges, lawyers, scholars, and politicians have evidenced
considerable dissatisfaction with the rule, and consequently it is once again
surrounded by controversy. But the new assault is by no means a mere revival
of the pre-Mupp arguments. Now, to paraphrase Holmes, the life of the con-
troversy is not logic but experience.’
In retrospect it is clear that the impetus for these renewed criticisms was the
mounting concern with crime and the political successes of the law and order
movement, contributing to the election of Richard Nixon, the appointment
of conservative justices to the U.S. Supreme Court, and the more conserva-
tive makeup of Congress. The rule’s critics also perceived the new Supreme
Court Justices as being somewhat amenable to abolishing or modifying the
exclusionary rule, at least as it applies to physical evidence.s Moreover,
unlike building new prisons, enhancing law enforcement activities, hiring

1. 232 U S . 383 (1914).

2. Indeed, by 1960the exclusionary rule had been adopted by nearly half the states. See Elkins v. United
States, 364 U S . 206, 224-32 (1960).
3. 367 U S . 643 (1961).
4. For a thorough review of the history and development of the exclusionary rule see Yale Kamisar,
Does (Did) (Should) the Exclusionary Rule Rest on a “Principled Basis” Rather than an “Empirical Pro-
position”? 16 Creighton L. Rev. 565 (1983). See also Steven Schlesinger, Exclusionary Injustice: The
Problem of Illegally Obtained Evidence ch. 2 (New York: Marcel Dekker, 1977).
5 . Mallory v. United States, 354U.S. 449 (1957); Escobedo v. Illinois, 378 U S . 478 (1964); Miranda v.
Arizona, 384 U S . 436 (1966).
6. United States v . Wade, 388 U.S. 218 (1967) (lineups); Gilbert v. California, 388 U.S. 263 (1967)
(identifications).
7. Bradley C . Canon, Is the Exclusionary Rule in Failing Health? Some New Data and a Plea Against a
Precipitous Conclusion, 62 Ky. L.J. 681, 683 (1974).
8. Because illegally obtained identifications and confessions are less credible as evidence, they have re-
mained in the background of the current controversy.
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 587

more prosecutors, etc., abolishing or modifying the exclusionary rule would

not involve financial costs. All these reasons have contributed to the eleva-
tion of the exclusionary rule to a high priority in the law and order move-
ment’s agenda for the criminal justice system. Thus, the recent Attorney
General’s Task Force on Violent Crime called for the creation of a “good
faith’’ exception to the rule.9
THE“RENEWED”EXCLUSIONARY
RULEDEBATE
As Canon notes,I0 the renewed attacks on the exclusionary rule are not
merely extensions of the pre-Mapp debate (i.e., balancing individual rights
against public safety). To understand the parameters of the renewed debate,
we must look to what might be viewed as the twin pillars on which the Mapp
decision was based. According to a widely cited 1970 article by Dallin Oaks:
Two types of justifications have been urged for the exclusionary rule, one nor-
mative and one factud. The normative justification is the evil of government
participation in illegal conduct. The factual justification lies in the assertion
that excluding evidence will reduce violations of the search and seizure rules.
This result is supposed to follow in the short term from deterrence and in the
long term from education.”
Critics of the rule in the post-1970 debate have disavowed the validity of the
first pillar and called for empirical evidence of the second. As Oaks notes:
Although the normative justification that the Supreme Court has referred to
as the “imperative of judicial integrity” continues to appear in the rhetoric of
Supreme Court decisions, it is doubtful that this argument decides cases.
Despite bold pronouncements about not being a “party to lawless invasions, ”
federal courts have not yet been forbidden from entering a valid judgment of
conviction against a defendant who was brought before the court by illegal
means such as kidnapping, arrest without probable cause, or arrest on a war-
rant that was illegal or insufficient. Moreover, it is difficult to accept the pro-
position that the exclusion of improperly obtained evidence is necessary for
“judicial integrity” when no such rule is observed in other common law juris-
dictions such as England and Canada, whose courts are otherwise regarded as
models of judicial decorum and fairness.’*
Following up on the Oaks research, Chief Justice Burger, in his widely dis-
cussed dissent in Bivens v. Six Unknown Federal Narcotics Agents, states:
“Some clear demonstration of the benefits and effectiveness of the exclu-
sionary rule is required to justify it in view of the high price it exacts from so-
ciety-the release of countless guilty criminals.”13
Proponents of the rule have argued that Oaks and other critics have dis-
torted the historical rationale and logic behind the rule. Most recently, in a

9 . The Attorney General’s Task Force on Violent Crime: Final Report, Recommendation, 40,55-56
(Washington, D.C.: Department of Justice, 1981).
10. Supra note 7.
1 1 . Dallin Oaks, Studying the Exclusionary Rule in Search and Seizure, 37 U. Chi. L. Rev. 665, 668
(1970).
12. Id. at 669.
13. 403 U S . 388,416 (1971).
588 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

thoroughly documented argument, Yale Kamisar argues that deterrence is

not mentioned at all in the early decisions that led to the exclusionary rule;
the real purpose of the rule, he contends, is the normative 0ne.I4 Discussing
the Weeks case, which established the federal exclusionary rule, Kamisar
notes:
Nowhere in Weeks is the exclusionary rule called a “remedy” and nowhere in
the opinion is there any discussion, or even mention, of the effectiveness of the
exclusionary rule versus the effectiveness of tort remedies, internal self-disci-
pline or other alternatives. Although few would suspect this from the argu-
ments of the rule’s opponents or from recent majority opinions of the Su-
preme Court, the Weeks opinion contains no language that expressly justifies
the rule by reference to a supposed deterrent effect on police officials.
Indeed, no less formidable a critic of the exclusionary rule than Chief Jus-
tice Burger recognized, before ascending to his present post, that the Weeks
case “rest[s] on the Court’s unwillingness to give even tacit approval to official
defiance of constitutional provisions by admitting evidence secured in viola-
tion of the Constitution. The idea of deterrence may be lurking between the
lines of the opinion but is not expressed.”
Nor, so far as I have been able to tell, is the idea of deterrence expressed for
the next thirty-five years.1J
Thus, the thrust of Kamisar’s argument is that empirical data on the ef-
fects of the rule should be irrelevant: the normative justification is compell-
ing and constitutionally sufficient to support the rule.I6
Nevertheless, in its 1974 decision in United States v. Calandra,” the Bur-
ger Court adopted the view that the principal rationale for the exclusionary
rule was its value in deterring the police from illegal behavior. In a recent
symposium Canon aptly summarized what has occurred in the reformulation
of the debate: “By portraying the rule as a pragmatic social policy rather
than a basic constitutional principle, its critics have shifted the scope of the
debate from arguments about constitutional law and judicial integrity (where
they lost) to arguments about the empirical data.”’*
In the mid-l970s, several empirical studies attempted to assess the deter-
rent effects of the rule on police behavior.” The research design and method-

14. Kamisar, supra note 4.

15. Id. at 598 (footnotes omitted).
16. See also Lane V. Sunderland, The Exclusionary Rule: A Requirement of Constitutional Principle,
69 J. Crim. L., Criminology, &Police Sci. 141 (1978); Thomas S. Schrock & Robert C. Welsh, Up from
Calandra: The Exclusionary Rule as a Constitutional Requirement, 59 Minn. L. Rev. 251 (1974); Critique,
On the Limitations of Empirical Evaluations of the Exclusionary Rule: A Critique of the Spiotto Research
and United States v. Calandra, 69 Nw. U.L. Rev. 740, 776-90 (1974).
17. 414 U.S. 338 (1974).
18. Bradley C. Canon, Ideology and Reality in the Debate over the Exclusionary Rule: A Conservative
Argument for Its Retention, 23 S. Tex. L.J. 559, 563 (1982); see also Critique, supra note 16, at 776-77.
19. Oaks, supru note 11; Michael Ban, The Impact of Mupp v. Ohio on Police Behavior, delivered at
the annual meeting of the Midwest Political Science Association, Chicago, 1973; id., Local Courts v. the
Supreme Court: The Impact of Mapp v. Ohio, delivered at the annual meeting of the American Political
Science Association, New Orleans, 1973; James E. Spiotto, Search and Seizure: An Empirical Study of the
Exclusionary Rule and Its Alternatives, 2 J. Legal Stud. 243 (1973); Canon, supra note 7; Bradley C.
Canon, Testing the Effectiveness of Civil Liberties Policies at the State and Local Levels: The Case of the
Exclusionary Rule, 5 Am. Politics Q. 57 (1977); Effect of Mapp v. Ohio on Police Search-and-Seizure
Practices in Narcotics Cases, 4 Colum. J . L. & SOC.Probs. 87 (1968).
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 589

ology of those studies was criticized, however,20and the empirical evidence

on deterrence remains ambiguous.2’Nonetheless, in calling for abolition of
the rule its critics have argued that more than ambiguous evidence on deter-
rence is needed to justify the rule. They have argued that the social costs it
imposes are so heavy as to outweigh any possible benefits. Thus, this “new”
argument against the rule has assumed the form of a cost-benefit analysis,22
with most of the discussion asserting that the rule fails to provide the touted
benefit-deterrence of police misconduct.
For a long time the costs of the rule-”the release of countless guilty crimi-
nals,” in Burger’s word~~~-were simply assumed and largely unexamined,
though several defenders of the rule questioned whether the costs were as
great as the critics claimed.24Then in 1977, a study of the effects of the rule in
federal prosecutions was conducted by the General Accounting Office at the
request of Congress.25That study found that federal prosecutors dropped
very few cases because of the exclusionary rule. In fact, it found that only
0.4% of all cases declined for prosecution by federal prosecutors were de-
clined primarily because of illegal search problems.26Moreover, it found that
only 10.5% of all defendants who were prosecuted filed search and seizure
motions to suppress27and concluded that in only 1.3% of the 2,804 cases
studied was evidence excluded;2seven then convictions were obtained in half
the cases in which evidence was excluded.29
More recently, in 1982 the National Institute of Justice (NU) conducted an
in-house study of the effects of the exclusionary rule using various California
data for the years 1976-79.30That study concluded that the exclusionary rule
exerted a “major impact on . . . state prosecutions,” especially drug cases.3’
Using official California data, the NU study reported that in the four years
studied, prosecutors had declined to file complaints in 86,033 felony arrests,

20. Bradley C . Canon, The Exclusionary Rule: Have Critics Proven That It Doesn’t Deter Police? 62
Judicature 398 (1979); Critique, supra note 16, at 744-64.
21. See United States v. Janis, 428 US. 433,450-52 n.22 (1976) (after reviewing the various empirical
studies of the rule’s deterrent effects and criticisms of those studies, Justice Blackmun, writing for the
Court, concluded: “The final conclusion is clear. No empirical researcher, proponent or opponent of the
rule, has yet been able to establish with any assurance whether the rule has a deterrent effect”); Stone v.
Powell, 428 U.S. 465, 492 n.32 (1976) (“the evidence [on deterrence] derived from recent empirical
research is still inconclusive”). See also Canon, supra note 18, at 560-72; Donald L. Horowitz, The Courts
and Social Policy 220-54 (Washington, D.C.: Brookings Institution, 1977).
22. E.g., Richard A. Posner, Excessive Sanctions for Governmental Misconduct in Criminal Cases, 57
Wash. L. Rev. 635 (1982). But see Arval Morris, The Exclusionary Rule, Deterrence and Posner’s
Economic Analysis of Law, 57 Wash. L. Rev. 647 (1982).
23. See supra text accompanying note 13.
24. See, e.g., Yale Kamisar, The Exclusionary Rule in Historical Perspective: The Struggle to Make the
Fourth Amendment more than ‘an Empty Blessing,’ 62 Judicature 337, 341.
25. Report of the Comptroller General of the United States, Impact of the Exclusionary Rule on
Federal Criminal Prosecutions Rep. No. CDG-7945 (19 Apr. 1979).
26. Id. at 14.
21. Id. at 8.
28. Id. at 11.
29. Id. at 13.
30. National Institute of Justice, Criminal Justice Research Report-The Effects of the Exclusionary
Rule: A Study in California (mimeo.; Washington, D.C.: Department of Justice, National Institute of
Justice, 1982).
31. Id. at 2.
590 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983:585

of which 4,130-or 4.8%-were declined because of illegal search prob-

l e m ~ .The
~ * study also concluded, after analysis of very small samples drawn
from two local prosecutor offices, that about 30% of drug cases were reject-
ed by prosecutors because of illegal searches.” Finally, the study concluded
that “individuals released because of search and seizure problems were those
with serious criminal records who appeared to continue to be involved in
crime after their relea~e.”’~ These findings must be viewed cautiously,
however, as the NU study has been criticized for overstating the effect of the
rule. Specifically, it has been argued that the NU data show that only 0.8% of
all arrests are rejected by prosecutors because of the rule;35that official
statewide statistics show that California prosecutors reject only 2.4%, not
30%, of drug arrests because of illegal and that fewer than 0.3%
of nondrug arrests are rejected because of illegal searche~.~’
These empirical assessments of the rule’s effects have obvious potential
importance for the present debate. Indeed, the NU statistics were recently
cited in Justice White’s concurring opinion in Illinois v. Gates in support of a
purported need to adopt a “reasonable good faith exception” to the exclu-
sionary rule.” The following questions, as well as others, must be addressed
if we are to arrive at a meaningful assessment of the costs of the exclusionary
rule. How many cases are lost? What types of cases? Who are the defend-
ants? What, if any, are the effects of unsuccessful motions on plea bargain-
ing?
THEPRESENT
STUDY
In an attempt to answer these questions, this article draws on data col-
lected in a comprehensive study of criminal courts to evaluate the impact or
costs of the exclusionary rule on felony prosecutions in a more complete, de-
tailed, and systematic manner. The larger study that is the source of the data
reported here is an intensive, collaborative study of criminal courts in nine
medium-sized counties (with populations ranging from 100,OOO to 1,0oO,OOO)
in Illinois, Michigan, and Pennsylvania. 3 9 The Illinois counties were
DuPage, Peoria, and St. Clair; the Michigan counties were Oakland, Kala-
mazoo, and Saginaw; the Pennsylvania counties were Montgomery, Dau-
phin, and Erie.
Medium-sized counties were chosen for the larger study because they were
large enough to produce interesting variance in cases and case-processing
32. Id. at 10 and table 1.
33. Id. at 2, 13.
34. Id. at 2.
35. Thomas Y.Davies, A Hard Look at What We Know (and Still Need to Learn) About the “Costs”
of the Exclusionary Rule: The NIJ Study and Other Studies of “Lost” Arrests, 1983 A.B.F. Res. J . 61 1,
infra, at 633.
36. Id. at 639.
37. Id. at 645-46.
38. 103 S. Ct. 2317, 2342 n.13 (White, J., concurring) (decided June 8, 1983). But see Davies, supra
note 35, at 617.
39. The results of the larger study of these nine county criminal courts are presented in Peter F. Nar-
dulli, James Eisenstein, & Roy B. Flemming, Sentencing as a Sociopolitical Process, Final Report to the
National Institute of Justice, Grant No. 81-1J-CXM)27 (June 1982).
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 59 1

procedures, yet not so large that their size would prevent the research team
from gaining a thorough understanding of how the county court system op-
erated. The nine counties were selected to maximize variance on a variety of
political, social, and economic dimensions that could plausibly affect crimi-
nal court operations. One area of concern was socioeconomic welfare;
another was the political views of the county’s populace. The researchers felt
that these were important because counties with deep social cleavages may
operate differently from more homogeneous and prosperous counties. Also,
more politically conservative counties could be expected to operate different-
ly from more liberal counties. To fulfill these criteria, the researchers selected
for each state one economically declining county (St. Clair, Saginaw, Erie),
one geographically isolated or autonomous county (Peoria, Kalamazoo,
Dauphin), and one suburban ring county (DuPage, Oakland, Montgomery).
Table 1 reports data on economic and political characteristics of the nine
counties. As expected, the ring counties are the most prosperous-with per
capita incomes hovering at about $lO,OOO in 1979. The declining counties are
far less prosperous; per capita incomes stood at somewhat over $6,500. Polit-
ically, DuPage and Dauphin counties appear to be the most conservative,
followed by Peoria and Montgomery counties. The Michigan counties ap-
pear to be fairly moderate, while St. Clair and Erie counties are moderately
liberal.

TABLE 1
Environmental and Structural Characteristics of the Nine Counties
Average
Vote for Con-
Public servative Presi-
Assistance idential FBI’s Uniform
Per Capita Recipients Candidate@)in Overall Crime Rate for
Income, per 100,OOO “Ideological” Political Offenses Against
1979 Populationa Electionsb Category Personsc
Illinois:
DuPage (ring) . . . . . . . . . . . $10,495 713 68% Conservative 205
Peoria (autonomous) . . . . . 8,388 4,689 55 Conservative 860
St. Clair (declining) . . . . . . 6,550 12,409 43 Moderately 792
liberal
Michigan:
Oakland(ring) . . . . . . . . . . 10,675 3,202 53 Moderate 442
Kalamazoo (autonomous) . 7,776 5,838 53 Moderate 582
Saginaw (declining) . . . . . . 7,263 9,778 51 Moderate 333
Pennsylvania:
Moderately
Montgomery (ring). . . . . . . 9,764 1,569 56 conservative 20 1
Dauphin (autonomous) . . . 7,581 5,165 60 Conservative 5 20
Erie(dec1ining). . . . . . . . . . 6,680 5,361 46 Moderately 243
liberal

aFebruary 1980.
b1980,1972, 1%8, 1964.
c1971-80 average; rate per 100,00 population.
592 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

The nine counties also showed some important differences in crime rates.
According to the FBI reports on violent personal crime rates (per 100,OOOpop-
ulation) for the ten years preceding this study (1971-80), Peoria and St. Clair
counties have the highest rates; Kalamazoo and Dauphin counties are far
lower. While two Michigan counties (Oakland and Saginaw) have fairly low
personal offense rates, two of the ring counties (DuPage and Montgomery)
and Erie have the lowest.
It is obvious that these counties differ markedly even though they are all
midsized American communities. Indeed, they were selected because of their
differences. But no claim is made that the criminal courts of these counties
are in any way a representative sample. They are not. However, their diversi-
ty will help undercut the types of biases that often creep into findings based
on only one or two locales.
In these nine counties the research team studied many aspects of criminal
court operations (e.g., bail, motions filed, plea bargaining, sentencing, case
disposition, delay) and collected extensive case data on almost 7,500 felony
defendant^.^^ The number of defendants for the nine counties ranged from
1,162 in St. Clair County to 594 in Erie County. These cases represent
roughly a year’s cases in each county; the nine sets of cases, on balance, re-
flect about nine years of dispositions. Most were disposed of during 1979 and
1980. In most counties all cases for a given time span were included in the
sample. However, in some counties systemic samples were In addi-
tion, 300 interviews were conducted with the judges, prosecutors, and de-
fense attorneys who handled the cases. The taped interviews yielded detailed
insights and information on the operating and structural characteristics of
the nine court systems. The interviewees also completed attitudinal, back-
ground, and personality surveys.
The Exclusionary Rule: An Empirical Assessment
Although the larger study was not specifically designed to examine the im-
pact of motions to suppress incriminating evidence, data on such motions (as
well as others) were collected for every case sampled. Because the scope of
the study is broader than most that have examined the exclusionary rule, the
perspective is a good deal more inclusive. This is useful in conducting the
data analyses as well as in interpreting the results. Moreover, while the pre-
sent controversy concerns only motions to suppress physical evidence, the as-
sessment here also includes motions to suppress identifications and confes-
sions. Inclusion of data on these other types of motions provides some inter-
esting contrasts and a more comprehensive and balanced view.
40. The defendant-incident was the unit of analysis for this study. If more than one charge arose from
the same incident, the other charges would simply be listed as “second offense charged” or “third offense
charged.” A defendant charged with a subsequent offense from a separate incident would show up as a
separate case.
41. A more detailed description of sampling procedures can be found in Nardulli et al., supra note 39,
at ch. 2. Note that a sampling problem in Dauphin County, Pennsylvania, led us to weight several hundred
diversion cases (largely DWI cases) to assure their proper representation in the sample. Thus, while we
sampled only 766 defendants in Dauphin County, the weighted figure is 1,058. Correspondingly, while on-
ly 7,475 total cases were sampled, the weighted number is 7,767.
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 593

One important caveat should be made. Data used to assess the impact of
the exclusionary rule on felony prosecutions are from motions to suppress at
the trial court level only. In some jurisdictions this approach would be trou-
blesome because prosecutors and/or lower courts (preliminary hearings)
may screen out many problem cases, including some with illegal searche~.~’
But the interviews indicate that there appears to be little screening by prose-
cutors or lower courts in the nine counties. In the Illinois counties the prose-
cutors screened cases, but only in St. Clair did they refuse to initiate prosecu-
tions in more than a handful of cases. In DuPage and Peoria counties screen-
ing simply meant advising the police on appropriate charges. Moreover, ex-
amination of a sample of preliminary hearings in Illinois lower courts com-
piled by the research team shows that motions to suppress occurred in fewer
than 0.4% of the cases.
A similar sample was not available for the Michigan counties, but inter-
views there indicated that screening by prosecutors was not extensive. In
Michigan the lower court proceedings were held in district courts spread
throughout the county and presided over by locally elected judges, who are
expected to be much more concerned with community safety than with legal
niceties. Thus, many preliminary hearings were waived, and most defense at-
torneys said they preferred to make legal motions at the circuit level.
In Pennsylvania, state law empowered local police to initiate felony prose-
cutions. Hence, there was no prosecutorial screening. Here, too, lower court
proceedings were held in district courts spread throughout the county and
staffed by elected magistrates. (Pennsylvania magistrates did not even have
to be lawyers.) Thus there was a real incentive for Pennsylvania defense at-
torneys to postpone legal motions until a case went to the trial court. In sum,
while it is possible that a small number of illegally obtained evidence cases
were screened out by prosecutors or in preliminary hearings in lower courts,
it is unlikely that they would have any substantial effect on the analysis that
follows.
The Incidence of Motions to Suppress Evidence
Drawing on the data for 7,767 weighted cases defendant^),^^ table 2 re-
ports how often motions were filed to suppress physical evidence, an identifi-
cation, or a confession in all nine counties combined.44All are rare events.
The most often filed, the motion to suppress a confession, is filed in fewer
than 7% of all cases. The other two are each filed in fewer than 5% of all
cases.
The picture presented in table 2 needs some clarification. At first glance it
might appear that these motions cumulatively affect about 17% of all cases.
In fact, however, in a good portion of the cases involving a motion to sup-

42. E.g., in California, prosecutors screen out a substantial number of arrests by declining to file a
complaint. See National Institute of Justice, supra note 30.
43. See supra note 41.
44. Whenever there was some ambiguity or doubt as to whether a motion was actually filed, we coded
the motion variable as missing. This was a relatively rare occurrence, but it could happen when part of a
file was found to be missing or when coders had reason to doubt that the file was intact.
594 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

press, more than one motion is made (table 3). In all, motions to suppress
were filed in about 11 Yo of all cases for which we have good data. Just over
half of all cases with motions to suppress (416 of 810) had more than one mo-
tion, with motions to suppress both a confession and an identification ac-
counting for the bulk of the multimotion cases (68%).
Even this figure of 1 1 To is a somewhat inflated estimate of the incidence of
motions if we distinguish serious motions to suppress from pro forma filings.
A county-by-county breakdown of motions (table 4)reveals why. Some im-
portant differences between counties emerge, especially with respect to mo-
tions to suppress identifications and confessions. The incidence in St. Clair
County is four to five times that of the county with the next highest inci-
dence. Indeed, St. Clair accounts for 80% of all motions to suppress identifi-
cations (301 of 375) in the merged samples and 71% of all motions to sup-
press confessions (364 of 5 14). The St. Clair County public defender’s office
has a policy of automatically filing motions to suppress an identification and
a confession along with the standard discovery motion-even in cases with
no identification or confession. The St. Clair public defender motions were
not viewed as real and undoubtedly were not given much attention; virtually
none were granted, as will be seen.
, Thus, a better estimate of how often motions to suppress are filed can be
obtained by excluding the St. Clair public defender cases. With these cases
removed, the total incidence of cases with one or more motions to exclude is
7.6%. Motions to suppress identifications are filed in about 2% of all cases,
while motions to suppress confessions occur in 3.6% of all cases,
To obtain an idea of the characteristics of the cases in which some type of
motion to suppress evidence is made, the incidence of motions was examined
by type of offense and defendant’s criminal record. Table 5 shows that mo-
tions to suppress physical evidence are filed at a higher rate in cases involving
drugs (13%) and unlawful possession of weapons (12%) but are rare in cases
involving personal offenses (4%) and property offenses (30’/0).Motions to
suppress identifications and confessions, however, are filed at a higher rate in
cases involving offenses against persons and property.
Table 6 presents data on defendants’ prior criminal records, classified as
first offenders, moderate, or While there is a slight tendency for
motions to suppress to be filed at a higher rate in cases involving defendants
with more serious criminal records, the differences are quite minor, as the
table These data show that the observed relationships remain fairly
stable.

45. The criminal record variable is a composite measure incorporating data on number of arrests, con-
victions, and incarcerations. The variables were combined and weighted using factor analysis. Nardulli et
al., supra note 39, at 2-24, report the results of the factor analysis. The distinction between defendants
with moderate and serious criminal records is based on this composite measure. Data on first offenders
were removed, and the remaining defendants were divided into two roughly equal groupings based on
scores on the composite criminal record variable. Defendants with higher weighted rankings of arrests,
convictions, and incarcerations were assigned to the “serious” category, the remainder to the “moderate”
category .
46. The figures in parentheses in tables 5 and 6 are the proportions with the St. Clair public defender
cases removed.
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 595

TABLE 2
Frequency of Motions to Exclude Incriminating Evidence, All Counties
(Weighted N=7,767)
Percent of All Cases
Physical Identif-
Evidence cation Confession
% (N) % (M % (M
Motionfiledintrialcourt.. . . . . . . . 4.6 (356) 4.8 (375) 6.6 (514)
Nomotionfiledin trialcourt.. . . . . 92.4 (7,177) 92.2 (7,158) 90.4 (7,019)
Uncertaina .................... -3.0 (234) 3.0 (234) 3.0 (234)
T o t a l . . ..................... 100 100 100
?Some ambiguities or missing documents in the case file made it unclear whether a motion was filed

TABLE 3
Motions to Suppress Incriminating Evidence, Including Multirnotion
Cases
No. of VO of Adjusted Qlo
Cases Cases of Cases=
Nomotion filed.. . . . . . . . . . . . . . . . . 6,714 86.4 89.2

All motions to suppress. . . . . . . . . . . . 1,053 13.6 10.8

Motion to suppress:
Evidence ..................... 244 3.1 3.2
Identification . . . . . . . . . . . . . . . . . . 39 0.5 0.5
Confession .................... 133 1.7 1.8
Evidenceand identification . . . . . . 13 0.2 0.2
Evidence and confession . . . . . . . . . 58 0.7 0.8
Identification and confession . . . . . 284 3.7 3.8
Evidence, identification, and con-
fession ..................... 39 0.5 0.5
Uncertainb .................... 243 3.1 ....
- -
Total .................... 7,767 100.0 100.0
aExcludes missing data. See supra note 27a.
b e e note a to table 2.

TABLE 4
Motions to Suppress by County ( N = 7,533)
% of Cases Involving Motion
No. of To Suppress To Suppress To Suppress Of Any Type
Cases Evidence Identification Confession to Suppress
Illinois:
DuPage . . . . . . . . . . 880 6 0.3 1 7.2
Peoria . . . . . . . . . . . 1,036 3 0.5 3 5.4
St.Clair . . . . . . . . . . 1,157 6 26.2 31 34.9
Michigan:
Oakland.. ........ 797 3 0.8 1 4.8
Kalamazoo . . . . . . . 7 18 2.3 0.3 4 5.7
Saginaw., . . . . . . . . 635 5 1.1 2 6.9
Pennsylvania:
Montgomery . . . . . . 681 14 6.6 6 17.2
Dauphin . . . . . . . . . . 1,047 2 0.2 0.2 1.7
Erie.. . . . . . . . . . . . . 582 2.4 0.3 2 4.7
596 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983:585

The Outcome of Motions to Suppress Evidence

Table 7 reports the rulings for each type of motion in the pooled samples.
Successful motions to suppress are rare, although the rate of success varies
with the type of motion. If we exclude the cases for which we could find no
ruling on the motion (between 14 and 7%, depending on the category), we
find success rates of 16.9% for motions to exclude physical evidence, 1.7%
for motions to suppress an identification, and 2.5% for motions to exclude a
confession. The combined success rate for identifications and confessions
without the St. Clairpublic defender cases is about 5 % . Clearly, motions to
suppress physical evidence have an appreciably higher success rate than the
other two.
The success rates in table 7 were calculated using the number of motions
filed as the denominator. Another way of examining success rate-one that
yields a more accurate picture of the overall effect of exclusionary rules on
the dispositional process-would be to calculate suppressions as a percentage
of all cases. These percentages are reported in table 8. Successful search and
seizure motions occurred in only 0.7% of all cases (i.e., 7 in I ,OOO), while suc-
cessful motions to suppress confessions or identifications occurred at even
lower rates.
Tables 9, 10, and 11 report the success rate of motions for the three vari-
ables for which the rate of motion filings was reported: county, offense type,
and criminal record. The most relevant observation in table 9 is that the
counties with the highest rates of motion filings have extremely low rates of
success. Motions to suppress physical evidence were filed at the highest rate
in Montgomery County (about 14%), but only 3% of these motions were
granted, the lowest rate by far in the nine counties. Moreover, only about
0.3% of all cases in Montgomery involved a successful motion to exclude
physical evidence. Thus, the percentage of all cases in which motions to sup-
press physical evidence were granted appears much more level across coun-
ties than does the incidence of motions, ranging from 0.3 to 1.1Vo. A similar
picture appears with regard to motions to suppress identifications and con-
fessions. In St. Clair, which had many pro forma motions to suppress identi-
fications and confessions, no motions to suppress identifications were
granted, and only 0.8% (3 of 364) of the motions to suppress confessions
were granted.
Table 10 shows that the highest success rates for motions to suppress
physical evidence occur in cases involving drugs (25%) and weapons (33%)
offenses. The motion success rate for physical evidence for the other offenses
is about 10%. However, such motions were filed in only 13.1% of all drug
cases and 11.9% of all weapons cases; thus the success rate of such motions
for all drug cases was only 2.9'70, for all weapons cases, 3.4%. Successful
motions occur in only a fraction of all cases in each other offense category,
none approaching 0.5% (0.3% in both offenses against persons and proper-
ty, as well as miscellaneous offense^).^' Motions to suppress identifications
47. None of the motions granted in offenses against persons involved exceptionally serious cases such
as murder, rape, armed robbery, or even unarmed robbery. The motions granted were in indecent ex-
posure, simple battery, and aggravated assault cases.
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 597

TABLE 5
Motions to Suppress by Type of Offense
% of Cases Involving
Motion Motion Motion
to Suppress to Suppress to Suppress
Evidence Identificationa Confessiona
Offense against person . . . . . . . . . . . 4 6 (3) 8 (5)
Property offense. . . . . . . . . . . . . . . . 3 6 (2) 9 (4)
Drug offense. . . . . . . . . . . . . . . . . . . 13 2 (1) 2 (1)
Unlawful possession of a weapon . . 12 1 (0) 5 (4)
Other . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 (2) 3 (2)
N . . .................... 7,533 7,533 (7,019) 7,533 (7,019)
apercentages in parentheses calculated with the St. Clair public defender cases removed

TABLE 6
Motions to Suppress by Defendant’s Criminal Record
% of Cases Involving
Motion Motion Motion
to Suppress to Suppress to Suppress
Evidence Identificationa Confessiona
First offender . . . . . . . . . . . . . . . . . . 5 3 (1) 6 (3)
Relatively moderate record . . . . . . . 4 6 (2) 6 (3)
Serious record. . . . . . . . . . . . . . . . . . 6 6 (3) 9 (5)
N ...................... 5,671 5,672 (5,285) 5,670 (5,283)
aPercentages in parentheses calculated with the St. Clair public defender cases removed

TABLE I
Motion Rulings
Cases Involving Motions to Suppress
Evidence Identificationa Confessiona
Motion denied:
Percent . . . . . . 71.6 91.5 (79.5) 89.1 (79.9)
N ........... 255 343 (1 12) 458 (203)
Motion granted:
Percent . . . . . . 14.6 1.6 (4.2) 2.3 (4.3)
N.. ......... 52 6 (6) 12 (1 1)
No ruling:
Percent . . . . . . 13.8 6.9 (16.3) 8.6 (15.8)
N . . ......... 49 26 (23) 44 (40)
- - - - _ _ _
TotalNo. 356 375 (141) 514 (254)
aFuures in parentheses calculated with the St. Clair public defender cases removed
598 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

TABLE 8
Successful Motions as a Percentage of All Cases
Cases Involving Motions to Suppress
Physical
Evidence Identification Confession
% N % N % N
Percentage of successful motions,
usingallcasesasabase.. . . , . . . . . . . . . . 0.69 7,484 0.08 7,507b 0.16 7,489
aExcludes 49 cases for which no outcome was ascertained.
bExcludes 26 cases for which no outcome was ascertained.
CExcludes 44 cases for which no outcome was ascertained.

TABLE 9
Success of Motions by County
Cases Involv- Successful Motions Total Successful
ing Motions As a?, of No. of Motions as 070
NO. % No. All Motions Cases of All Cases
Motion to Suppress Physical Evidence
DuPage (ring) . . . . . . . . 52 5.8 6 12.8 875 0.7
Peoria (autonomous) . . 36 3.5 11 39.3 1,028 1.1
St. Clair (declining) . . . 69 6.0 10 14.7 1,156 0.9
Oakland (ring) . . . . . . . 27 3.0 I 26.9 796 0.9
Kalamazoo (autono-
mous) . . . . . . . . . . . . 17 2.4 5 29.4 718 0.7
Saginaw (declining) . . . 32 4.6 3 12.0 628 0.5
Montgomery (ring). . . . 93 13.6 2 3.0 655 0.3
Dauphin (autonomous) 16 1.5 4 25 .O 1,047 0.4
Erie (declining) . . . . . . . 14 2.4 4 38.8 581 0.7

Motion to Suppress Identification

DuPage (ring). . . . . . . . 3 0.3 0 0 880 0
Peoria (autonomous) . . 5 0.5 1 20.0 1,035 0.1
St. Clair (declining) . . . 303 26.1 0 0 1,155 0
Oakland (ring) . . . . . . . 6 0.6 0 0 797 0
Kalamazoo (autono-
mous) . . . . . . . . . . . , 2 0.3 0 0 718 0
Saginaw (declining) . . . 7 1.0 2 50.0 63 1 0.3
Montgomery (ring). . . . 45 6.5 3 11.5 66 1 0.4
Dauphin (autonomous) 2 0.2 0 0 1,049 0
Erie (declining) . . . . . . . 2 0.3 0 0 581 0

Motions to Suppress Confessions

DuPage (ring) . . . . . . . . 13 1.4 0 0 878 0
Peoria (autonomous) . . 29 2.8 3 13.0 1,030 0.3
St. Clair (declining) . . . 364 31.3 3 0.8 1,154 0.3
Oakland (ring) . . . . . . . 11 1.2 1 11.1 796 0.1
Kalamazoo (autono-
mous) . . . . . . . . . . . . 28 3.9 3 12.5 714 0.4
Saginaw (declining) . . . 12 1.7 1 14.3 630 0.2
Montgomery (ring). . . . 44 6.4 0 0 660 0
Dauphin (autonomous) 2 0.2 0 0 1,046 0
Erie (declining) . . . . . . . 11 1.9 1 9.1 58 1 0.2
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 599

are most successful in drug cases (1 1-1Yo) and miscellaneous cases (1 1.a%),
but these occur in only 0.2% of all such cases. Most successful motions to
suppress confessions are filed in weapons cases. (Neither of these observa-
tions changes if the St. Clair cases are removed.) Again, only a minuscule
fraction of all cases in each category is affected by successful motions, except
for weapons cases, where 1.3% involve successful motions to suppress con-
fessions.
Table 11 shows that the likelihood that a motion to suppress physical evi-
dence and identifications will be successful decreases with the severity of the
defendant's record. For example, motions suppressing physical evidence
were granted in 0.9% of all cases involving first offenders but in only 0.3 and
0.4% of cases involving defendants with moderate or serious records. The
opposite is true with respect to confessions. But even there, only 0.3% of
cases involving defendants with serious criminal records had confessions
suppressed, compared to 0.1070of cases in which defendants were either first
offenders or had moderate records.
Successful Motions and Conviction
It is obvious from these data that for most types of offenses and offenders
the filing of motions to suppress various types of incriminating evidence is

TABLE 10
Success of Motions by Types of Offenses
Cases Involv- Successful Moiions Total Successful
ing Motions As % of No. of Motions as '70
No. '70 No. All Motions Cases of All Cases
Motion to Suppress Physical Evidence
0ffenses against
persons.. . . . . . . . . . . . 61 3.7 5 9.1 1,682 0.3
Property offenses. . . . . . . 114 3.1 12 12.1 3,613 0.3
Dmgoffenses.. . . . . . . . . 1 1 1 13.1 24 25.0 849 2.8
Weapons offenses . . . . . . 28 11.9 8 33.3 231 3.4
Miscellaneousoffenses . . 39 3.4 3 10.0 1.184 0.3

Motion to Suppress Identifications

Offenses against
persons.. . . . . . . . . . . . 101 6.1 1 1.1 1,682 0.1
Property offenses. . . . . . . 225 6.1 1 0.5 3,613 0.0
Drugoffenses.. . . . . . . . . 20 2.3 2 11.1 849 0.2
Weaponsoffenses . . . . . . 3 1.3 0 0 237 0.0
Miscellaneous offenses . . 23 2.0 2 11.8 1,184 0.2

Motion to Suppress Confessions

Offenses against
persons . . . . . . . . . . . . . 137 8.2 3 2.4 1,682 0.2
Propertyoffenses. . . . . . . 310 8.4 5 1.7 3,673 0.1
Drugoffenses.. . . . . . . . . 21 2.4 1 5.0 849 0.1
Weapons offenses . . . . . . 12 5.1 3 30.0 237 1.3
Miscellaneousoffenses . . 31 2.6 0 0 1,184 0
600 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983:585

rare; successful motions are even rarer. We cannot, however, stop our assess-
ment of the role and impact of the exclusionary rule with the outcome of mo-
tions because even a successful motion to suppress does not necessarily trans-
late into an acquittal or dismissal.
Table 12 reports dispositional data for cases in which a motion to suppress
was made, by outcome of the motion. As shown there, only 22% of the cases
in which physical evidence was suppressed resulted in conviction, while 83 ‘Yo
of the cases in which an identification was suppressed and 58% of the cases in
which a confession was suppressed ended in conviction. The effect of the
successful suppression motions on the conviction rate can be evaluated by
comparing these outcomes to a conviction rate of 88% (n= 6,227) in cases in
which no motions to suppress were filed. Thus, table 12 shows that unsuc-
cessful motions had little impact on conviction rates (compared with the 88%
overall figure), but the impact of a successful motion varied with the type of
motion. Successful motions to suppress the physical evidence have the great-
est impact, because they tend to occur in cases involving drugs and unau-
thorized use of weapons. Without the drugs or the weapon there is usually
not much of a case. Correspondingly, fewer than 22% of defendants who
succeed in having physical evidence suppressed are convicted; while 89% of
those who unsuccessfully filed such motions were convicted. In contrast, the
successful suppression of an identification had virtually no impact on convic-
tion rates, although the number of cases with successful motions is small (6).
Obviously other incriminating evidence was available in these cases. The im-
pact of successful motions to suppress confessions is somewhat greater, but
the conviction rate is still almost 60% (58.3%). Here again, however, the
number of cases is quite small (1 2).
Table 12 also reports the rate at which convictions are “lost” because of
each type of successful motion to suppress. The highest “lost conviction”
rate is, of course, for cases in which physical evidence has been excluded-
but the 40 “lost” cases are barely one-half of 1070 of all cases (0.56%). The
overall lost conviction rates for the other types of motions are not even one-
tenth of 1 %; only one case, or 0.014%of all cases, is lost because of suppres-
sion of an identification, and only 5 cases, or 0.071070 of all cases, are lost be-
cause of suppression of a confession. Moreover, if we combine the effects of
the three exclusionary rules, the 46 lost convictions are only 0.65% (fewer
than 7 in 1 ,OOO) of the 7,035 cases for which we have good data on both mo-
tions and trial court dispositions. But even this modest figure is a somewhat
inflated estimate of the rules’ impact. Just as some cases result in conviction
because of other evidence even after a successful motion to suppress, so some
cases with successful motions that did not result in convictions would have
been lost because of other evidentiary deficiencies. Because about 12% of
cases involving no motions to suppress did not result in convictions, this
figure can be used to adjust the 0.65%estimate for normal attrition; thus, we
arrive at a final estimate of the impact of the exclusionary rules on conviction
rates of 0.57% (6 in 1,OOO). In other words, just over one-half of 1070 of the
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 60 1

TABLE 11
Success of Motions by Defendant’s Criminal Record
Cases Involv- Successful Motions Total Successful
ing Motions As To of No. of Motions as To
No. % No. All Motions Cases of All Cases

Motion to Suppress Physical Evidence

First offenders . . . . . . . . 135 4.7 25 22.1 2.921 0.9
Defendants with rela-
tively moderate crirni-
nal records . . . . . . . . . 58 3.8 5 10.2 1,509 0.3
Defendants with serious
prior record. . . . . . . . . 75 5.3 6 9.2 1,398 0.4

Motion to Suppress Identifications

First offenders . . . . . . . . 98 3.3 3 3.3 2,921 0.1
Defendants with rela-
tively moderate crirni-
nal records . . . . . . . . . 88 5.8 2 2.4 1,509 0.1
Defendants with serious
prior record. . . . . . . . . 86 6.1 0 0.0 1,398 0.0

Motion to Suppress Confessions

First offenders . . . . . . . . 1.59 5.5 3 2. I 2,921 0.1
Defendants with rela-
tively moderate crimi-
nal records . . . . . . . . . 96 6.4 1 1.2 1,509 0.1
Defendants with serious
prior record. . . . . . . . . 121 8.7 4 3.6 1,398 0.3

TABLE 12
Convictions and Motions to Suppress ( N = 7,035)

No.
Convictions Lost
Lost Due to Convictions
No. of No. of 070 Successful as % of
Cases Convictions Convicted Motions All Cases

Motion to Suppress Physical Evidence

Filed but denied . . . . . . 255 227 89 -_ -_
Granted . . . . . . . . . . . . 51 11 21.6 40 0.56

Motion to Suppress Identification

Filed but denied . . . . . . 343 290 84.5 __ __
Granted . . . . . . . . . . . . 6 5 83.3 1 0.014

Motion to Suppress Confession

Filed but denied . . . . . . 458 401 87.6 _- --
Granted . . . . . . . . . . . . 12 7 58.3 5 0.071
602 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

felony cases handled by these court systems were lost due to the operation of
the three exclusionary rules combined.
Characteristics of “Lost Cases’’
A judgment whether the rate at which cases are lost is “too high” may de-
pend on the characteristics of the cases lost because of these rules. Table 13
shows the most serious offenses involved in the 46 lost cases. Fewer than 20%
of these (8) are serious cases: unarmed robbery, arson, and burglary. How-
ever, almost 60% involve victimless crimes: 37% involve the unlawful illegal
possession of drugs and 22% involve possession of a weapon. Another 24%
involve indecent exposure, theft, or some miscellaneous offense.
In terms of the seriousness of the offenses, as defined by the average sen-
tence given to defendants convicted of equivalent offenses in the same coun-
ty, analysis shows that most (65%) of the defendants who were not convicted
following granted motions to suppress would have received sentences of less
than six months. Of the 40 defendants who were not convicted because of
suppression of physical evidence, over 20% (9) would have received less than
one month in jail, an additional 33% (1 5 ) would have received less than two
months. Only one could have expected more than one year.
Criminal record data were missing for 17 of the 46 defendants whose cases
were not convicted due to the exclusionary rules. However, of the 29 defend-
ants for whom we had such information, 18 (62%) were first offenders, 4 had
moderate records, and 7 had serious criminal records. The cases of all 18 of
the first offenders involved motions to suppress physical evidence.

The Impact of Unsuccessful Motions

We should not stop our assessment of the exclusionary rule with the im-
pact of successful motions on convictions. There is a final question whether
unsuccessful motions to suppress affect the dispositional process. For exam-
ple, might cases in which unsuccessful motions are made be more likely to re-
sult in adversary trials and appeals? Moreover, many of the social forces op-
erating in criminal courts that mute the impact of exclusionary principles on
convictions may have other consequences, such as concessions in plea negoti-
ations. Therefore, I will examine the disposition of cases in which an unsuc-
cessful motion to suppress was made. Also, I will look at the impact of these
motions on the types of concessions (charge and sentencing reductions) that
may be made in plea negotiations.
Unsuccessful Motions and Disposition Mode
Table 14 compares the disposition of cases that had unsuccessful motions
to suppress with those that had no such motion. The first row shows that a
significantly higher percentage of defendants whose motions to suppress in-
criminating evidence were denied demanded a trial than did defendants who
did not file such motions. While the difference is statistically significant, it is
not large-8.4 percentage points. Of the 905 defendants with unsuccessful
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 603

motions, 135 went to trial and 108 were convicted. What this means, of
course, is that defendants who thought they had a legitimate argument to ex-
clude incriminating evidence exercised their right to a trial more often than
did other defendants, in order to preserve their right to appeal. These poten-
tial appeals represent about 1.5% (108/7,035) of the trial court dispositions.
However, other research suggests that most such convictions wilI be affirmed
on
Unlike the mode of disposition, there was no statistically significant differ-
ence in conviction rates for the two groups. In other respects most cases in-
volving denied motions were also handled similarly to other cases. About 8%
of those making unsuccessful motions were dismissed, as opposed to 10.3%
of other cases. Slightly over 77% of the cases with denied motions resulted in
some type of guilty plea, while 83.3% of other cases resulted in such a plea.

TABLE 13
Types of Offenses “Lost” Due to Exclusionary Rule”
% of All “Lost”
Type of Offense No. Cases
Unarmed robbery. ....................... lb 2.2
Aggravated assault ....................... 1 2.2
Arson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2.2
Burglary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5c 10.9
Deviant sexual act (largely indecent exposure) . 4 8.7
Theft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 8.7
Unlawful possession of a
weapon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1od 21.7
Possession of controlled
substance ............................ 17 37.0
Miscellaneous ........................... 3d 6.5
46 100.0

aExcept for some of the footnoted offenses these “lost” cases were due to motions to
su press physical evidence.
gDue to a motion to suppress an identification.
Three cases were due to a suppressed confession.
done case was due to a suppressed confession.

TABLE 14
Unsuccessful Motions and Trial Court Dispositions
Defendants Did Not Defendants Unsuccess-
File a fully Filed a
Motion to Suppress Motion to Suppress
Percentage of defendants demand-
ing a trial ..................... 6.5% (6,146) 14.9%*** (905)
Percentage of defendants
convicted ..................... 88.0% (6,227) 89.0% (919)

***Significance beyond ,001 level.

48. See Thomas Y . Davies, Affirmed: A Study of Criminal Appeals and Decision-making Norms in a
California Court of Appeal, 1982 A.B.F. Res. J . 543. In the year and jurisdiction studied, there were
approximately 10,ooO convictions, id. at 577, but only 544 criminal appeals were decided on the merits, id.
at 556. Of those, 151 raised an illegal search issue but only 8 were successful, id. at 616 table 9.
604 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

Unsuccessful Motions and Plea Bargaining

It is no secret that negotiated guilty pleas are the most common form of
trial court dispositions, and the most significant impact of motions to sup-
press may be on these negotiation^.^^ Because an unsuccessful motion might
be grounds for an appeal, it might become a bargaining tool for a defense at-
torney. Prosecutors might offer a charge or sentence reduction in exchange
for a plea, thereby cutting off most avenues for appeal. This cements the
conviction and eliminates further work on the case. Bargaining concessions
by the prosecutor may also be the state’s way of making “more appropriate”
amends to a defendant whose rights have been violated by the state. That is,
while the criminal court actors handling the case may suspect that a constitu-
tional violation has occurred, there may be scant willingness to free the de-
fendant on such a “technicality.” In such cases the violation may be “worked
out” in the negotiations.
We can examine this possibility by analyzing the impact of unsuccessful
motions on three aspects of plea bargaining. First, we look at whether the
plea is “on the nose” or to a reduced charge. Then we examine cases involv-
ing reduced charges to see whether these motions affect the size of the con-
cession. Finally, we look at the impact on negotiated sentences of unsuccess-
ful motions to suppress.
Multiple regression analysis was used to control for two important case at-
tributes: the seriousness of the offense and the severity of the defendant’s
record,” the variables shown by research to be the most important ones to
use in predicting the outcome of plea negotiations, especially sentencing. s 1
After these control variables were entered into the analysis, we allowed a set
of three “dummy” variables depicting the making of an unsuccessful motion
to suppress of any of the three types to enter. The variables used are reported
in table 15.
Table 16 reports the impact of the unsuccessful motion variables on each
dependent variable. The results are spotty. First, unsuccessful motions to
suppress physical evidence have quite a strong impact on the decision to re-
duce the original charge. Cases in which such motions have been filed and
denied have about a 20% greater chance of obtaining a charge reduction
than do other cases ( B = .21). However, the magnitude of the charge reduc-
tions does not seem to be significant. Not only does the EVMOTN-UN variable

49. This analysis cannot measure the full impact of the exclusionary rules on plea bargaining because
one defense tactic may be not to file the motion in exchange for plea concessions. Obviously, this cannot
be measured quantitatively with this data set. However, while this hidden impact may be a reality, one
must question the bargaining leverage one obtains from a motion that has such a low success rate when
made. Prosecutors are well aware of the potential for success, and it is unlikely that they would be buf-
faloed into many concessions, especially for marginal cases. If thecase for suppression is good, it is unlike-
ly that the defense attorney will pass up a chance for a quick dismissal.
50. The offense seriousness measure is essentially the county-specific mean sentence of the offense in
our county samples. The criminal record measure is a composite variable derived from a factor analysis of
the number of prior arrests, prior convictions, county jail commitments, and penitentiary commitments.
More details on the derivation of these measures can be found in Nardulli et al., supra note 39, at 2-23,
n.25.
51. Id., ch. 6.
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 605

not have a significant impact on GPBREAK, which measures the relative size of
the charging concession (see table 15), it also has no impact on the sentence
given. That is, the reduced charge did not lead to a reduced sentence, even if
we control for the original charge instead of the reduced charge.
The variable representing unsuccessful motions to suppress identifications

TABLE 15
Variables Used in Plea Negotiation Analysis

Symbol Scale/Interpretation
Dependent variables:
Was the guilty plea “on the nose” NOSEPLEA 1 = o n the nose
or to a reduced charge 0 = reduced
Relative magnitude of the GPBREAK [(seriousness of original charge) -
charge concession (seriousness of charge pled to)]/
(seriousness of original charge)
Sentence JAILTIME Minimum months of confinement,
probation coded 0
Control variables:
Seriousness of most serious OFFSER County-specific average sentence for
offense convicted on the offense (months)
Severity of defendant’s criminal CRIMRCD Composite variable reflecting prior
record arrests, convictions, and sentences
‘‘Dummy ” mot ion variables:
Unsuccessful motion to suppress EVMOTN-UN 1 = unsuccessful motion
physical evidence 0 = n o motion
Unsuccessful motion to suppress IDMOTN-UN 1 =unsuccessful motion
identification 0 = no motion
Unsuccessful motion to suppress CONFMOTN-UN 1 =unsuccessful motion
confession 0 = n o motion

TABLE 16
Summary of Regression Analysis for Motion Variables and the Plea Process (OFFSER
and CRIMRCD-Controlled)

EVMOTN-UN IDMOTN-UN CONFMOTN-UN

NOSEPLEA:
B coefficient . . . . . . . . . . . . . . . . . - .21 .03 - .03
Fvalue ...................... 31.4 .88 I .02
Level of significance. . . . . . . . . . . beyond .001 N.S. N.S.
N of cases . . . . . . . . . . . . . . . . 5.561 5,561 5,567
GPBREAK:
Bcoefficient . . . . . . . . . . . . . . . . . .00 .03 .06
Fvalue ...................... .08 .86 1.4
Level of significance. . . . . . . . . . . N.S. N.S. beyond .01 level
Nof cases ................... 2,628 2,628 2,628
JAILMIN:
Bcoefficient . . . . . . . . . . . . . . . . . - 1.4 1.54 - 8.6
Fvalue ...................... .14 .ll 4.4
Level of significance. . . . . . . . . . . N.S. N.S. .05
Nof cases . . . . . . . . . . . . . . . . . . . 4,998 4,998 4,998
606 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 1983585

(IDMOTN-UN) was the weakest. It had no impact on any of the dependent vari-
ables measuring plea concessions. Unsuccessful motions to suppress confes-
sions did, however, have two measurable effects. First, while defendants in
such cases had only about a 3% greater probability of getting a charge reduc-
tion-an effect too small to be statistically significant-the magnitude of the
changes that did occur was statistically significant. That is, the impact of
CONFMOTN-UN on GPBREAK was strong enough not to have occurred by
chance (i.e., real charge concessions appear to have been given). On average,
cases with unsuccessful motions to suppress confessions had the most serious
offense charged reduced by about 6%. This is not a large drop (the average
offense seriousness score is only about 8 months). A more significant effect
occurs on sentencing. The defendants who had an unsuccessful motion to
suppress a confession received a sentence that is, on average, 8 months less
than that received by other defendants with a similar criminal record con-
victed of a similar offense. Further analyses revealed that the impact of an
unsuccessful motion to suppress a confession on sentence increased as the
seriousness of the offense increased. For example, a theft case with an aver-
age sentence of 4 months would get a sentencing concession of 4 months due
to an unsuccessful motion to suppress a confession. A burglary case with an
average sentence of 12 months would get a concession of 8 months, while an
armed robbery with an average of 40 months would get a concession of about
22 months.

SUMMARY AND POLICY IMPLICATIONS

Given the results of the empirical analysis it seems clear that the exclusion-
ary rules-as applied to physical evidence, identifications, and confessions-
have a truly marginal effect on the criminal court system. Although no claim
is made that these jurisdictions are “representative,” their diversity, the size
of the data base, and the minuscule effect of the rules make it unlikely that a
more representative sample would produce a very different picture. Exclud-
ing pro forma filings, the three types of motions combined are filed in only
about 7.6% of the cases; motions to suppress physical evidence are filed in
fewer than 5 % , largely drug and weapons cases. The success rate of these
motions is equally marginal-about 17% of motions to suppress physical
evidence are granted (representing 0.69% of all cases), about 5% for identifi-
cations (representing 0.08% of all cases), and 5% for confessions (represent-
ing 0.16% of all cases).
Not all individuals who successfully suppressed evidence escaped convic-
tion, especially when only an identification or a confession was suppressed.
Fewer than 0.6% of the cases in the nine-county sample were lost due to the
three exclusionary rules combined. Moreover, most of the lost convictions
involved first offenders or offenses that would have incurred only a few
months’ imprisonment. First offenders who filed motions to suppress physi-
cal evidence were twice as successful as defendants with criminal records; de-
fendants charged with victimless crimes were much more successful than
No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 607

others. The impact of the rule on plea bargaining was equally marginal. Only
unsuccessful motions to exclude confessions resulted in any real sentencing
concessions.
Some critics may contend that these results are deflated estimates of the ex-
clusionary rules’ impact because they do not incorporate the screening phase
of the process, the prior discussion of screening notwithstanding. While one
can always point to “hidden effects,” these arguments should not be viewed
as compelling nor can they undercut the clear implications of the findings
presented here. Indeed, the findings reinforce earlier conclusions about the
likely “noneffect” of prior screening (in these courts) on this analysis. Why
would prosecutors in charge of screening (who tend to be entry-level prosecu-
tors) or lower court judges (who are usually elected from local districts) take
responsibility for dismissing cases on the basis of legal arguments that have
only remote chances of being accepted by trial court judges? The reality of
the matter is that there is little incentive for them to do so, especially in more
serious cases. Thus, it is unlikely that this analysis has missed the rules’ im-
pact on many serious cases.
Even if one were to reject these contentions and insist that estimates for
pretrial court screening be incorporated in this analysis, the findings would
not be substantially different. For example, the NU study cited earlier showed
that only about 0.78% of all arrests were dismissed because of search and sei-
zure problems. If this estimate were simply added to our estimate of 0.57%
(0.78 + 0.57 = 1 . 3 9 , the policy implications would be unchanged. The costs
of keeping the rules are not “the release of countless guilty criminals”; only a
relatively few marginal offenders are released. If these minuscule costs are
compared with the benefits (the ambiguous deterrence value of the rules), I
can only conclude that there should be no change in the status quo. Critics
appropriately bear the burden of justifying a change. They have not done so.
The case for retention of the present rules is even stronger if one accepts
Kamisar’s resurrection of the moral justification for excluding illegally ob-
tained evidence.
Viewed from a larger perspective, one must also ask, Is the status quo
good enough? Is the right battle being fought? What good is a legal principle
that, in Canon’s words, “has no significant impact on society”?s2Does the
Fourth Amendment deserve only symbolic support? Do the marginal effects
of the rule mean that search violations are overlooked by trial court judges?
One critic of the rule, Steven Schlesinger, contends:
At present, judges participate in a system in which no discipline or punishment
is given to an officer whose misbehavior gives rise to exclusion; the judiciary
has become part of a system in which misbehavior by law enforcement offi-
cials is rampant, apparently not decreasing significantly, and often tolerated
or encouraged by superiors.”

52. Canon, supra note 18, at 578.

53. Schlesinger, supra note 4.
608 AMERICAN BAR FOUNDATION RESEARCH JOURNAL 198x585

Although these charges are not empirically documented (and are not easily
empirically testable), if true they constitute a national scandal of far greater
proportion than the release of a fraction of a percent of defendants due to the
exclusionary rules.
All of us should be alarmed by this situation, whether we are concerned
with civil liberties or the intrusion of government into private lives. A
“trade” of the exclusionary rules for empirically proven devices to curtail
Fourth Amendment abuses would not be a bad one, depending on the actual
devices and the method of adoption. It is a “deal” that proponents of the
rules should consider.
Oddly enough, the seeds of such a device can be found in the work of one
of the rules’ harshest critics, Steven S c h l e ~ i n g e rHe
. ~ ~proposes two alterna-
tives to the exclusionary rules. Neither one is necessarily inconsistent with the
preservation of the rules, and neither is acceptable as presented by Schles-
inger. Nonetheless, both merit further study.
The first part of Schlesinger’s plan is an independent review board to
which judges could refer suspected instances of abuse. The board, which
would have disciplinary powers, would be composed of citizens, judges, and
law enforcement officials. Penalties would range from fines to severance
from the force. With some modifications this might be an effective deterrent.
One necessary change would be the composition of the board. Since the
board’s primary function would be the determination of constitutional viola-
tions based on legal criteria, it should be staffed by lawyers selected from the
local bar association. Citizen and law enforcement representation is unneces-
sary for such deliberations because the only question would be whether there
is a constitutional violation. Penalties should be assessed on the basis of a
predetermined schedule providing for minimum fines and leading to sever-
ance from the department. Provisions for greater penalties could be made
for grievous violations. Cases could be referred to the board by either the
defense attorney or the judge. A minimum fee would be paid to the attorney
preparing the complainant’s case but only in successful cases.

The second part of Schlesinger’s plan calls for a revised civil remedy:
I suggest a statutory civil action with provision for the award of monetary
compensation, the amount depending on the gravity of the misbehavior (but
with a minimum amount to be awarded in the event that any violation is
found). Such an action would allow innocent victims to recover a basic com-
pensatory amount plus counsel’s fees without any showing of specific damage
to the victim or flagrant violation by the officer. The only proof necessary
would be a showing that the victim’s Fourth Amendment rights had been vio-
lated; the greater the invasion of Fourth Amendment rights, the higher the
compensation. . . .
There are now in existence state common law causes of action under which
victims may recover damages, but civil actions are rarely instituted (and even
more rarely won) because it is necessary to show either substantial harm to the

54. Id.. ch. 4.

No. 3 THE SOCIETAL COST OF THE EXCLUSIONARY RULE 609

victim (for compensatory damages) or outrageous official misconduct (for

punitive damages). Winning the proposed civil suit would be easier because no
such showing would be necessar~.~’
While this remedy has merit, it would also require several modifications.
First, Schlesinger would permit only innocent defendants to sue because, he
argues, guilty defendants have no right to privacy: “If criminal activity is
predominantly a public concern, then when the police, either legally or il-
legally, find evidence of such activity, it is not an invasion of the individual’s
privacy to use what the police have found against him in a criminal proceed-
ing. ” s6
This is, of course, a novel idea, unsupported by legal precedent; it is open
to question whether such a remedy could withstand an equal protection at-
tack based on the Fourteenth Amendment. Moreover, I suggest that to make
the remedy even more effective, the police department should be the party
liable for the damages, thus discouraging police departments from condon-
ing illegal behavior on the part of their officers.
This two-pronged approach, which reaches both individual and institu-
tional sources of violations, might become an effective deterrent to Fourth
Amendment violations. But no action to modify or replace the exclusionary
rules should be taken until the effectiveness of these, or similar remedies, has
been empirically demonstrated. If “misbehavior by law enforcement of-
ficials is rampant,” then these actions should be undertaken by legislative
bodies independent of any judicial action on the exclusionary rules.
However, as an inducement to experimenting with and empirical testing of
more effective remedies, perhaps the Supreme Court should make it known
that there are a variety of ways in which the intent of the Fourth Amendment
can be fulfilled. States with empirically proven plans to discipline police mis-
conduct could be excluded from the exclusionary rule, but only after a
demonstration that the state’s alternative actually deterred Fourth Amend-
ment abuse.

5 5 . Id. at 77.
56. Id. at 48.