; Miscarriages of Justice and the Constitution
Learning Center
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Miscarriages of Justice and the Constitution


  • pg 1
									    DRIPPSMACRO                                                                 637
                                                                     8/11/99 4:16 PM

           Miscarriages of Justice and the

                            Donald A. Dripps*

    Accurate fact-finding surely counts as a primary
purpose of criminal procedure. Conventional wisdom holds
that our system is highly successful at preventing
erroneous convictions, but far less successful at preventing
erroneous acquittals or dismissals. This asymmetry is
commonly attributed to the rights of the accused,1
particularly the constitutional safeguards in the Bill of
Rights.2 This essay challenges the conventional wisdom on
three important fronts.
    First, while the common belief that many guilty people
go unwhipped of justice accords with the facts, the
prevailing belief that few innocent people are convicted
seems quite improbable. Our criminal justice system, then,
needs to do better at both convicting the guilty and
acquitting the innocent.
    Second, despite the Warren Court’s criminal procedure
revolution, constitutional law does remarkably little to
prevent unjust convictions.       The criminal procedure
revolution made the Bill of Rights for the most part
applicable to the states through the Fourteenth
Amendment. The Bill of Rights, however, provides quite
inadequate safeguards against unjust conviction. The Fifth

    * James Levee Professor of Law and Criminal Procedure, University of
Minnesota Law School. Thanks to Richard Frase of helpful comments on an
earlier draft. Responsibility for the opinions expressed is solely the author’s.
    1. See, e.g., Ronald J. Allen, The Simpson Affair, Reform of the Criminal
Justice Process, and Magic Bullets, 67 U. Colo. L. Rev. 989, 1009 (1996) (“A lot of
guilty people walk. It is the price we pay for a lot of things, like protecting
innocent people wrongly accused, reducing government intrusions on privacy, and
ensuring that private citizens be treated with a certain degree of dignity and
    2. See, e.g, James J. Tomkovicz, An Adversary System Defense of the Right
to Counsel Against Informants: Truth, Fair Play, and the Massiah Doctrine, 22
U.C. Davis L. Rev. 1, 92 (1988) (“Such social costs are inherent in the adversary
system, in the grant of counsel, and in our commitment to values beyond truth.
We pay a similar price for every ‘criminal’ protection in the Bill of Rights.”).
DRIPPSMACRO                       8/11/99 4:16 PM

638           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

and Sixth Amendment provisions relating to trial
procedure indeed ensure that the trial will be fair, but they
do not speak to the fairness of the antecedent investigation
by the police.       Fourth, Fifth and Sixth Amendment
doctrines regulating police investigation focus on
preventing police abuse of citizens outside the process of
formal prosecution, not on ensuring that the evidence the
police are allowed to gather is collected reliably and
comprehensively.       If constitutional doctrine is to do
anything more to prevent miscarriages of justice, lawyers
and judges must turn to the Due Process Clause, not to the
Bill of Rights.
     Third, while society indeed pays a price for
constitutional rights, constitutional doctrine leaves more
room than is commonly supposed3 for procedural reforms
calculated to prevent the escape of the guilty. In one form
or another, the single most important cause of lost but
deserved convictions is the failure of witnesses to cooperate
with the government. Legislatures, judges, and police
administrators could obviate much of the witness
noncooperation problem by arranging for the prompt
questioning of witnesses before trial. The immediate
accusation and assignment of counsel that would enable
cross-examination of witnesses before trial could coexist
with the existing rules of police interrogation.          The
Constitution, however, even as presently construed, also
leaves room for reforms aimed at questioning the suspect
before trial, after accusation and assignment of counsel.
     In short, our criminal process does not fulfill its
instrumental mission very well, but the Constitution is not
the major cause of the system’s deficiencies. With respect
to reducing the frequency of unjust dismissals or acquittals,
the Constitution does not stand in the way of reforms that
might improve the system’s performance. With respect to
reducing the frequency of unjust convictions, the

    3. For example, Gordon Van Kessel identifies constitutional criminal
procedure as the leading obstacle to reforming criminal justice to facilitate truth-
finding. See Gordon Van Kessel, Adversary Excesses in the American Criminal
Trial, 67 Notre Dame L. Rev. 403, 486 (1992).
DRIPPSMACRO                     8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                639

Constitution even now offers hope—politically speaking,
the only hope—for significant reforms.
     Part I develops the evidence suggesting the system’s
deficiencies. Part II connects constitutional doctrine to
unjust convictions. Focusing on eyewitness identification
evidence, Part II explains why the prevailing doctrinal
focus on the Bill of Rights tolerates unreliable
investigations. Part III turns to the problem of unjust
acquittals, and suggests that Fifth and Sixth Amendment
doctrine permit some promising reforms aimed at
convicting more of the guilty.


A. False Convictions

     The conventional wisdom is reassuring.         Learned
Hand believed the risk of unjust conviction to be “an unreal
dream.”4 A recent survey of judges, prosecutors and
defense lawyers found that the most common estimate of
the frequency of false convictions was 0.5%.5          In a
thoughtful article, Samuel Gross thought it plausible to
assume that 95% of charges are true and that 80% of cases
of mistaken identity are weeded out before trial.6 Even so
zealous a defense lawyer as Allen Dershowitz takes the
view that “[a]lmost all criminal defendants are, in fact,
     Along with others,8 I have long suspected this
comforting view. After all, what other decision-making

    4. United States v. Garsson, 291 F. 646, 649 (S.D.N.Y. 1923).
    5. C. Ronald Huff et al., Convicted but Innocent: Wrongful Conviction and
Public Policy 61 (1996).
    6. Samuel Gross, Loss of Innocence: Eyewitness Identification and Proof of
Guilt, 16 J. Legal Stud. 395, 451 (1987).
    7. Alan Dershowitz, The Best Defense xxi (1982).
    8. See Daniel Givelber, Meaningless Acquittals, Meaningful Convictions: Do
We Reliably Acquit the Innocent?, 49 Rutgers L. Rev. 1317 (1997); Martin Yant,
Presumed Guilty (1991).
DRIPPSMACRO                      8/11/99 4:16 PM

640           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

process run by the government is practically infallible? My
skepticism is now reinforced by disturbing evidence drawn
from DNA test results in criminal investigations.
     DNA tests have cleared several innocent persons who
had been convicted of serious crimes. There is nothing
terribly new about this research angle. More than sixty
years ago Borchard published a classic book documenting
case studies of false convictions.9 Borchard had little
trouble identifying the causes of false convictions:
erroneous eyewitness identifications,10 official misconduct,11
and inadequate defense representation.12 Sadly, the DNA
case studies show the same factors at work, despite the
Warren Court’s criminal procedure revolution.13
     Case studies, however, shed no real light on the
question of how frequently false convictions occur.
Obviously, the case studies mark the low estimate of how
many false convictions occur, but just as obviously in many
cases of false conviction conclusive evidence of innocence
never surfaces. The really startling fact is the frequency
with which DNA tests exonerate suspects under
investigation by the police.
     A Department of Justice study found that, out of a
sample of more than 21,000 cases, DNA testing exonerated
the suspect in 23% of the cases.14 In another 16% of the

     9. Edwin M. Borchard, Convicting the Innocent (1932).
    10. See id. at xiii (“Perhaps the major source of these tragic errors is
identification of the accused by the victim of a crime of violence.”)
    11. Id. at xv (“Yet in only a few of the cases can it be said that no fault,
carelessness, or overzealousness can be charged to the prosecution.”).
    12. See id. at xx (“In the majority of these cases the accused were poor
persons, and in many of the cases their defense was for that reason inadequate.”)
    13. See Peter Neufeld & Barry Scheck, Commentary, in Edward Connors et
al., Convicted by Juries, Exonerated by Science: Case Studies in the Use of DNA
Evidence to Establish Innocence After Trial at xxx (1996), (“Interestingly, in
many respects the reasons for the conviction of the innocent in the DNA cases do
not seem strikingly different than those cited by . . . Borchard . . . . Mistaken
eyewitness identification, coerced confessions, unreliable forensic laboratory
work, law enforcement misconduct, and ineffective representation of counsel,
singly and often in combination, remain the leading causes of wrongful
    14. Connors et al., supra note 13, at 20.
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                     641

cases, DNA tests produced inconclusive results.15 In other
words, in those cases in which there was a conclusive DNA
test result (a sample of many thousands of cases), 27% of
the suspects were exonerated by the test.16
     If the cases selected for DNA testing are
representative of cases at large, then our trial process is
being trusted to negative far more unfounded charges than
commonly supposed. The DNA cases might differ from
other cases in some ways. First, DNA cases tend to be
serious cases, involving a strong community demand for
justice. Second, prosecutors might decline to seek DNA
evidence to gild the lily of a strong case. An inconclusive
result might muddy the waters at trial, and there is the
expense to consider. Third, DNA tests might be done on
nonsuspects to exclude persons whom the defense might
cast in the role of the real culprit.
     None of these explanations seems very powerful. True,
the pressure to obtain a conviction in a high-profile rape or
homicide case might tempt police and prosecutors to look
hard for ways to pin the offense on some luckless innocent.
Serious cases, however, are the focal point of concern about
miscarriages of justice; they are the cases with the most
serious consequences attending false conviction. If “heater”
cases17 have a higher error rate than ordinary cases, we
should still be concerned about the system’s reliability
when the stakes are high.
     Moreover, “heater” cases involve not only more
pressure to convict, but also the devotion of additional

   15. Id.
   16. There are no apparent system factors that might have changed recently to
suggest that the frequency of exonerations might decline. At the moment, in
cases processed by the Minnesota Bureau of Criminal Apprehension, “[a]bout 35
to 40 percent of initial suspects in criminal cases are cleared after DNA analysis.”
Reuben Rosario, Crime Lab Serves Justice Through Science: Unsung Heroes of
Law Enforcement Wear White Coats, St. Paul Pioneer Press, Dec. 1, 1997, at 1B,
   17. I borrow the label from Myron W. Orfield, Jr., Deterrence, Perjury, and
the Heater Factor: An Exclusionary Rule in the Chicago Criminal Courts, 63 U.
Colo. L. Rev. 75 (1992). In Chicago courthouse jargon, a “heater” case is a
prosecution of a notorious crime, characterized by an intense public demand for
DRIPPSMACRO                      8/11/99 4:16 PM

642           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

resources to the investigation. The police may be more
willing to give up and close the investigation of a less
serious offense, but it is also true that an innocent person
mistakenly suspected of a less serious crime will not be
exonerated by further police work on the case. Anyone who
fits will do. If a brief investigation shows that nobody fits,
the police in a minor case will give up. But if the brief
investigation shows that somebody fits—or even sort of
fits—the investigation will go no further.
     As for the testing of nonsuspects, DNA testing is
expensive and time-consuming. The police ordinarily do
not have tests done on anyone who is not a strong, if not
the prime, suspect. Typically the government has probable
cause before obtaining a tissue sample from the suspect,
although the development of DNA databases may change
the situation. In short, the sample of persons tested is
pretty close to a sample of persons who would, in the
absence of DNA evidence, have gone to trial.
     Prosecutors may not wish to endanger a strong case
with an inconclusive test result,18 but prosecutors must also
consider the possibility that the jury will expect DNA
evidence and hold its absence against the government.19
Moreover, the last nail in the coffin has considerable
utility. Positive test results can persuade the defense to
plead guilty, or persuade the prosecution to reject a
generous plea agreement.
     In other words, at least in serious cases that do not
rest on overwhelming evidence of guilt—the cases the trial
process is for—a significant number of defendants who go
to trial are not guilty. The truth-finding mission of the
trial is more than rhetoric; it is deadly serious business.
     There is a widespread but unmentionable view to the

   18. See Givelber, supra note 8, at 1376 (“The weaker the prosecution’s case,
the more likely that the prosecutor will seek additional evidence from DNA
   19. See Lisa Bouwer Hansen, Stemming the DNA Tide: A Case for Quality
Control Guidelines, 16 Hamline L. Rev. 211, 245 n.226 (“In fact, in some cases,
the jury may expect DNA evidence to be presented and may consider it significant
when it is not.”) (citation omitted).
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                   643

contrary, which might fairly be labeled “legal cynicism.”
On this account, the trial is not meant to test truth at all,
but to ratify the police investigation without the necessity
of substituting police investigation for trial by
constitutional amendment. Because the constitutional
safeguards and nonconstitutional exclusionary rules, such
as those barring proof of a defendant’s record, make it
nearly impossible to convict the guilty with probative
evidence, it becomes necessary for the police to extract
admissions by fraud or fear, induce accomplice testimony,
and prompt eyewitness identifications. These forms of
proof, although of dubious reliability, nonetheless persuade
pliable jurors. Since the rules of evidence keep the reliable
evidence from the jury, it becomes necessary and even
desirable to convict those known by the police to be guilty
on the force of unreliable evidence.
     As we shall see, the Constitution, for the most part,
enables evidentiary reforms that would promote convicting
the genuinely guilty. But even if this were not true, we
should reject legal cynicism because the police simply do
not know who is guilty and who is not. Basing their
guesses on prior records and the whispers of informants,
the police turn out to be quite wrong in a large percentage
of cases. Perhaps the rules of evidence are more rational
than supposed. At any rate, the rules of evidence can be
made more rational than they are. The trial, therefore,
should test the truth, not the good faith of the police.
     There is some goods news about the system in the
research on false convictions. The case studies of false
convictions are virtually devoid of convictions entered on a
plea of guilty.20 Perhaps innocent people often plead guilty,
but the plea terminates the search for exonerating evidence
on their behalf. Still, plenty of defendants plead guilty
while maintaining their innocence to friends and family.

   20. See Gross, supra note 6, at 415 (only two out of 97 false convictions in
sample resulted from guilty pleas); Connor et al., supra note 13, (out of 28 cases
involving falsely convicted defendants later exonerated by DNA testing, only one
(Vasquez) pleaded guilty). But see Huff et al., supra note 5, at 73-74 (“Many
innocent defendants are convicted after entering guilty pleas.”).
DRIPPSMACRO                      8/11/99 4:16 PM

644           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

Because most convictions result from pleas, not from trials,
we can take some comfort from the scarcity of
demonstrably false guilty pleas.
    The frequency of exculpatory DNA test results
suggests that many innocent defendants go to trial. The
case studies of false convictions show how the trial process
can fail to protect the innocent. The major reasons, in
order of importance, are inaccurate identifications, official
misconduct, and ineffective defense counsel.21         False
confessions and bad luck play roles as well. Perhaps
surprisingly, these are the same factors that Borchard
pointed out more than sixty years ago.22 Studies since
Borchard’s have shown the same causes.23 How could the
criminal procedure revolution have done so little to
promote the reliability of the trial process? And is there
anything legal doctrine can do to improve the situation?
Part II takes up these questions, but first we must survey
the other side of the coin. How often does our system of
criminal procedure permit the guilty to escape?

B. Unjust Dismissal and Acquittals

    A great many crimes are not reported, and a great
many reported crimes are not cleared by the police. No
doubt public policy can encourage reporting and facilitate
the identification of offenders. Rape shield laws and
computerized fingerprint databases offer illustrations. My
present focus, however, is on the effectiveness of our
adjudicatory procedures in convicting arrested suspects
who are, in fact, guilty.
    The prevailing view among scholars,24 judges,25 and

   21. See supra notes 9-13; Huff et al., supra note 5, at 66-82.
   22. See supra notes 9-13.
   23. See Jerome Frank & Barbara Frank, Not Guilty 199-249 (1957); Edward
Radin, The Innocents 17-53 (1957) (police and prosecutorial misconduct); id. at
85-104 (identifications); id. at 233-34 (inadequate defense representation).
   24. See, e.g., Givelber, supra note 8, at 1321 (“America’s criminal justice
system creates a significant risk that innocent people will be systematically
convicted. It also creates a distinct and presumably larger risk that the guilty
will be acquitted.”); Gross, supra note 6, at 451 (assuming for purposes of
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                     645

ordinary citizens26 holds that a great many guilty
defendants escape their just deserts. Now this proposition
derives some of its support from the premise that almost all
criminal defendants are guilty. Given this premise, the
figures on case attrition—the frequency with which arrests
and even formal charges do not lead to convictions—offer
obvious proof of widespread unjust acquittals and
dismissals. Roughly speaking, half of all arrests do not end
in conviction.27
     Perhaps, however, case attrition reflects the weeding
out of weak cases, including many cases against innocent
defendants. If the skeptical view defended earlier about
the system’s success at avoiding unjust convictions is well-
taken, however, the syllogism loses some of its force. We do
not know for sure that the vast majority of cases lost by the
prosecution are, in fact, meritorious.
     Nor can many dismissals or acquittals be attributed to
exclusionary rules of evidence that benefit the patently
guilty. A few guilty defendants indeed escape justice in
this way, but any fair view of the empirical evidence
suggests that very small percentages of cases are lost on
Fourth or Fifth Amendment grounds.28 Some time back I

illustration that out of 950 cases against guilty defendants, 140 would be
dismissed and 35 acquitted at trial).
    25. See, e.g., Bivens v. Six Unknown Named Agents, 403 U.S. 388, 416 (1971)
(Burger, C.J., dissenting) (exclusionary rule frees “countless guilty criminals”);
Harold J. Rothwax, Guilty: The Collapse of Criminal Justice in America (1996).
    26. See, e.g., U.S. Dep’t of Justice, Sourcebook of Criminal Justice Statistics
1996 at 118 tbl. 2.9 (1997) (asked how much confidence they had in the criminal
justice system, 20% of respondents said a great deal or quite a lot; 40% said some;
and 35% said very little); id. at 149 tbl. 2.50 (asked whether courts in the
community deal too harshly, about right, or not harshly enough with criminals,
78% of 1996 respondents said not harshly enough).
    27. See, e.g., Yale Kamisar et al., Modern Criminal Procedure 25 (8th ed.
1994) (“Studies on the attrition of felony arrests do indicate, however, that the
combination of pre-filing review and post-filing review is likely to produce a total
rejection rate in the neighborhood of 50% for a ‘typical’ urban jurisdiction.”);
Brian Forst, Criminal Justice System: Measurement of Performance, in 2 Ency. of
Crime & Justice 479 (Sanford Kadsih ed. 1983) (typical rate of conviction
following arrest is 50%); Givelber, supra note 8, at 1136-1338 (in 1990, survey of
75 largest counties in the United States found that 42% of defendants charged
with violent felonies had case dismissed).
    28. On the Fourth Amendment, see Thomas Davies, A Hard Look at What We
DRIPPSMACRO                       8/11/99 4:16 PM

646           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

characterized such losses as “exotic,”29 and since then the
Supreme Court has made the rules more generous to the
police and trial courts have become generally more
conservative, less sympathetic to suppression motions.30
     The process of case attrition, however, bears a
remarkably arbitrary relation to the probability of the
defendant’s guilt. A series of studies in the 1980’s found
that the most common reason for loss of a case is witness
non-cooperation.31 The situation has not changed for the

Know (and Still Need to Learn) About the “Costs” of the Exclusionary Rule: The
NIJ Study and Studies of “lost” Arrests, 1983 A.B.F. Res. J. 611; Peter Nardulli,
The Societal Cost of the Exclusionary Rule: An Empirical Assessment, 1983
A.B.F. Res. J. 585; Report of the Comptroller General, Impact of the Exclusionary
Rule on Federal Criminal Prosecutions (Rep. No. CDG-79-45). On the impact of
Miranda, compare, e.g., Paul Cassell, Miranda’s Social Costs: An Empirical
Reassessment, 90 Nw. U. L. Rev. 387 (1996) with Stephen Schulhofer, Miranda’s
Practical Effect: Substantial Benefits and Vanishingly Small Social Costs, 90 Nw.
U. L. Rev. 500 (1996) and Paul Cassell, All Benefits, No Costs: The Grand Illusion
of Miranda’s Defenders, 90 Nw. U. L. Rev. 1084 (1996) and Stephen Schulhofer,
Bashing Miranda is Unjustified—and Harmful, 20 Harv. J.L. & Pub. Pol’y 347
(1997) and Paul Cassell, Miranda’s “Negligible” Effect on Law Enforcement: Some
Skeptical Observations, 20 Harv. J.L. & Pub. Pol’y 327 (1997). Even Professor
Cassell concedes that approximately eighty percent of interrogated suspects
waive their Miranda rights. See Cassell, An Empirical Reassessment, 90 Nw. U.
L. Rev. at 493-494. It seems highly unlikely that many of the suspects who
invoke their Miranda rights could be induced to make incriminating admissions
by tactics that would survive scrutiny under a reinvigorated voluntariness test of
the sort Professor Cassell envisions. Moreover, the police have an incentive to
work harder at obtaining waivers, and admissions, from suspects in cases that
depend on obtaining incriminating admissions. Thus even if Miranda causes a
few suspects not to talk, there is no reason to think that these suspects would
have given statements that were necessary and sufficient for a conviction in the
absence of Miranda.
   29. Donald Dripps, Beyond the Warren Court and its Conservative Critics:
Toward a Unified Theory of Constitutional Criminal Procedure, 23 U. Mich. J.L.
Reform 591, 626 (1990).
   30. See Thomas Y. Davies, Exclusionary-Rule Exception Flies in the Face of
the Fourth Amendment, Chi. Trib. Feb. 20, 1995, at 19.
   All of the available statistics consistently show that only between one-half
   and two and one-half percent of felony arrests are ‘lost’ in American
   jurisdictions because of unconstitutional searches. The rate is probably
   about twice that range if only drug cases are concerned, but it is virtually
   zero in violent crime arrests. The Supreme Court and even the Reagan
   Justice Department accepted these estimates a decade ago, and these
   ranges would almost certainly be even lower today. Id.
   31. See Floyd Feeney et al, Arrests Without Conviction: How Often They
DRIPPSMACRO                      8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                  647

better, and may have changed for the worse.32
     Sometimes the case is continued so often that
witnesses who have appeared to no avail on several prior
occasions lose interest and give up.         Sometimes the
defendant or his sympathizers have intimidated the
witnesses. Sometimes the witness is reached by bribes
rather than by threats. Sometimes, in domestic violence or
sexual assault cases, there has been a rapprochement
between the witness and the defendant. As that noted
evidence scholar Mr. Dooley put it, “nothing walks itself
into evidence.” The system is troubled little indeed by the
suppression of physical evidence or confessions, but very
greatly vexed by the loss of witnesses.
     Sometimes witnesses will drop out of an apparently
strong but factually false case. There is nothing to prevent
a gang-banger, coincidentally charged with a crime he did
not commit, from intimidating or bribing the mistaken
eyewitnesses. Likewise there is nothing that says that an
innocent defendant may not succeed in the game of
continuing the case until the witnesses does not show up.
But there is likewise no reason to suppose that the
witnesses are less likely to cooperate in the prosecution of
the innocent than in the prosecution of the guilty. The
empirical studies on case attrition do not attribute a
significant share of dismissals to factual innocence.33 No
doubt some lost cases involved innocent people, but there

Occur and Why 196-99 (1983) (summarizing studies); Brian Forst et al, Arrest
Convictability as a Measure of Police Performance 9-10 (1982).
    Particularly striking . . . is the extent to which witness and evidence
    problems were recorded as reasons for arrests being rejected at screening.
    In each of the jurisdictions [studied], at least 50 percent of the arrest
    rejections were attributed to those two problems, and in three jurisdictions
    the percentage was 70 or higher. Id.; Michael Graham, Witness
    Intimidation 1-8 (1985) (reviewing studies).
   32. See generally Witness Intimidation: Showdown in the Streets—
Breakdown in the Courts, Hearings Before the House Committee on the
Judiciary, 103d Cong., 2d Sess. (1994).
   33. See Feeney et al, supra note 31, at 243 (“Most suspects who are arrested
but not convicted are thought by police and prosecutors to be guilty. Many cases
of this kind are dropped for evidentiary reasons but could and probably ought to
be salvaged.”).
DRIPPSMACRO                       8/11/99 4:16 PM

648           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

is, sadly, no reason to suppose that innocent defendants are
more likely to be found in the cases that are dismissed than
in the cases that go forward.
      Even if the witnesses cooperate, the trial may result in
a mistaken acquittal. Kalven and Zeisel found that in 20%
of the acquittals in their sample, the trial judge thought
there should have been a conviction.34 The judges may be
wrong, or basing their opinion on information the jury
didn’t hear, like arrest reports and prior convictions. On
the whole, however, it seems likely that juries acquit at
least some defendants who are not only guilty but whom
the evidence shows to be guilty.
      If we recall, however, that the majority of convictions
result from pleas, we can see that the real hole in the
system is not in the trial, but in the pretrial stage.35 Once
it becomes clear that the government has a strong case, the
factually guilty plead in large numbers. This tendency in
turn suggests that jury trials do not run a terribly high risk
of false acquittal, or far more guilty defendants would opt
for trial. It also suggests that if the government can secure
a strong case that can be used at trial, no matter whether or
how the witnesses testify at the trial, the biggest single hole
in the system would be plugged.
      The obstacles to such a strategy are the hearsay rule
and its constitutional counterpart, the Sixth Amendment
Confrontation Clause. The difficulty of the challenge can
be gauged from the failure of the law-and-order types to
condemn these sweeping exclusionary rules. The right to
cross-examine the government’s witnesses in a criminal
case is rightly thought to be an essential safeguard against
unjust conviction. Thus it would seem that the system lies
between a rock and a hard place. Trial judges can either
require the witnesses to testify at the trial, which many of

   34. See Harry Kalven, Jr. & Hans Zeisel, The American Jury 68 (1966).
   35. Acquittal after trial terminates one percent of state prosecutions and 2.8%
of federal prosecutions. David Luban, Are Criminal Defenders Different?, 91
Mich. L. Rev. 1729, 1729 (1993) (footnotes omitted). By contrast, one study found
that 23% of a sample of more than 1500 cases were dropped because of witness
noncooperation. See Graham, supra note 31, at 4.
DRIPPSMACRO                  8/11/99 4:16 PM

1999]            MISCARRIAGES OF JUSTICE                 649

them will refuse to do, or admit police versions of what
witnesses said at the time of the crime, when the witnesses
themselves may very well deny that those statements were
made or now tell a different version of the facts.
     This apparent collision between the instrumental
functions of adjudication and the constitutional rights of
the accused, however, is as overstated as the claim that
constitutional safeguards make convicting the innocent
next to impossible. My claim here is that the Constitution
can be construed to significantly reduce the risk of unjust
conviction, and that even as presently construed, the
Constitution permits reforms that would significantly
reduce the number of unjust dismissals. Some innovative
but plausible constitutional interpretations could reduce
the number of unjust dismissals still further.


A. Constitutional Law for the Innocent: Diagnosis and

     The Warren Court cases indeed worked a revolution in
the administration of justice in the states. The revolution,
however, took the doctrinal form of incorporating the Bill of
Rights into the Fourteenth Amendment. The “fundamental
fairness” test of due process that governed state
prosecutions from Hurtado v. California36 until, roughly,
Mapp v. Ohio,37 has never been officially repudiated. The
constitutional spotlight, however, shines quite distinctly on
the Fourth, Fifth, and Sixth Amendments.
     From the standpoint of protecting the innocent, this
regime is disappointing. Some items in the Bill of Rights
are key components of due process and critical safeguards
against unjust conviction. Notice, counsel, confrontation,

  36. 110 U.S. 516 (1884).
  37. 367 U.S. 1 (1961).
DRIPPSMACRO                     8/11/99 4:16 PM

650           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

and compulsory process all promote accurate adjudication.
These provisions fall far short, however, of what accurate
adjudication requires.
     Many other provisions in the Bill of Rights protect
persons not charged with any offense against government
overreaching during the investigation of crimes. Fourth
Amendment limits on searches and seizures, and Fifth
Amendment limits on interrogation, do not promote
accurate adjudication, except in the important sense that
they prevent blameless people from official abuse without
any trial at all. Important as that function is, it does not
help the falsely accused defendant avoid unjust conviction.
Indeed, the Fifth Amendment privilege against self-
incrimination can actually frustrate the defense of the
     To go beyond the important but inadequate safeguards
in the Bill of Rights, an innocent defendant must look
either to statutory law or to the Due Process Clause. The
statutory turn can be dismissed pretty quickly.
Legislatures,    state    and   federal,   operate    under
overwhelming tough-on-crime political pressures. These
pressures, in turn, are quite rational. Most citizens face
zero risk of unjust prosecution, but considerable risk of
crime. From their perspective, Blackstone’s ten to one ratio
is empty rhetoric. Honestly rephrased to express today’s
politics, the aphorism would go like this: “Better that
several innocent young black men should be convicted than
that one guilty young black man should be acquitted.” So
long as street crime is the province of young males, and so
long as crime is higher in minority communities, police
suspicion will be aimed at young, disproportionately black,
men. Police errors will fall along with police successes on
the politically marginalized members of the target class.38
     That leaves the Due Process Clause. The Burger and
Rehnquist Courts have not questioned the incorporation
decisions, but have very much narrowed the meaning of the

   38. For a fuller treatment, see Donald A. Dripps, Criminal Procedure,
Footnote Four, and the Theory of Public Choice; Or, Why Don’t Legislatures Give
a Damn About the Rights of the Accused?, 44 Syracuse L. Rev. 1079 (1993).
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                     651

Bill of Rights provisions. In response, one might have
anticipated that litigants would bring due process
challenges to circumvent the prevailing conservative
interpretation of the incorporated amendments.            The
Court, however, has taken the view that the incorporated
amendments define due process in criminal cases, and that
free-standing due process claims are to be rather
skeptically received.
     For example, excessive force claims against the police
practically cry out for due process analysis, for in such
cases the suspect’s claim is that the police have bypassed
the trial and imposed their own brand of unofficial
punishment.39 But the Court has insisted on analyzing
excessive force and even excessive deadly force cases under
the Fourth Amendment,40 preferring to say that a suspect
shot to death by the police has been “seized” rather than
“deprived of life.”41 While the Court’s opinions contain dicta
suggesting the substantive due process might prohibit
police practices that fall outside the scope of “searches and
seizures,”42 neither the Burger nor the Rehnquist Court has
ever reversed a conviction because the police practices were
shocking to the conscience.
     Thus the conservative Court practically has banned
due process analysis from police practices cases, leaving the
field regulated solely by the Fourth, Fifth and Sixth
Amendments. This is a grave disservice to innocent
suspects.     Pretrial procedure can leave the criminal
defendant facing erroneous but now entrenched
identification testimony, without the benefit of exculpatory
physical evidence the police neglected to collect or preserve,
defended by an overworked lawyer with no time to conduct

   39. See Johnson v. Glick, 481 F.2d 1028 (2d Cir. 1973); Bradley M. Campbell,
Excessive Force Claims: Removing the Double Standard, 53 U. Chi. L. Rev. 1369,
1390-91 (1986) (“The underlying concern when police use excessive force is
precisely that the police have been able to ‘skip the trial’ and proceed directly to
punishing the detainee.”).
   40. Graham v. Connor, 490 U.S. 386 (1989).
   41. Tennessee v. Garner, 471 U.S. 1, 9 (1985).
   42. See United States v. Russell, 411 U.S. 423, 432 (1973); Hampton v. United
States, 425 U.S. 484, 495 n.7 (Powell, J., concurring).
DRIPPSMACRO                  8/11/99 4:16 PM

652           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

a new investigation. This can (and does) happen, without
any unreasonable searches, without any compelled
testimony, and without any denial of counsel.          The
distinction between investigation and adjudication is far
less palpable than current doctrine admits.          When
defendants have challenged clearly adjudicatory procedures
on procedural due process grounds, the modern Court has
continued the basic confusion of substantive and
procedural due process. After some early willingness to
apply the administrative due process balancing test of
Mathews v. Eldridge43 in criminal cases,44 in Medina v.
California45 the Court took a very different approach. The
issue in Medina was whether a state rule casting upon the
defendant the burden of proving incompetence to stand
trial by a preponderance of the evidence violated due
process. The defendant attacked this allocation of the
burden of proof under the Eldridge test.
     Writing for the Court, Justice Kennedy rejected both
the specific claim and the reliance on Eldridge:

   In the field of criminal law, we have defined the category of
   infractions that violate fundamental fairness very narrowly
   based on the recognition that, [b]eyond the specific
   guarantees enumerated in the Bill of Rights, the Due
   Process Clause has limited operation. The Bill of Rights
   speaks in explicit terms to many aspects of criminal
   procedure, and the expansion of those constitutional
   guarantees under the open-ended rubric of the Due Process
   Clause invites undue interference with both considered
   legislative judgments and the careful balance that the
   Constitution strikes between liberty and order.

   The proper analytical approach, and the one that we adopt
   here, is that set forth in Patterson v. New York, which was
   decided one year after Mathews.

  43. 424 U.S. 319 (1976).
  44. See United States v. Raddatz, 447 U.S. 667, 677-79 (1980); Ake v.
Oklahoma, 470 U.S. 68, 77-83 (1985).
  45. 505 U.S. 437 (1992).
DRIPPSMACRO                      8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                    653

   It goes without saying that preventing and dealing with
   crime is much more the business of the States than it is of
   the Federal Government, and that we should not lightly
   construe the Constitution so as to intrude upon the
   administration of justice by the individual States. Among
   other things, it is normally within the power of the State to
   regulate procedures under which its laws are carried out,
   including the burden of producing evidence and the burden
   of persuasion, and its decision in this regard is not subject
   to proscription under the Due Process Clause unless it
   offends some principle of justice so rooted in the traditions
   and conscience of our people as to be ranked as
   fundamental. As Patterson suggests, because the States
   have considerable expertise in matters of criminal
   procedure and the criminal process is grounded in centuries
   of common-law tradition, it is appropriate to exercise
   substantial deference to legislative judgments in this area.
   The analytical approach endorsed in Patterson is thus far
   less intrusive than that approved in Patterson.46

    Note how the Court repeatedly slips back into the
language of “fundamental fairness,” a substantive due-
process concept, to articulate the proper inquiry into
procedural fairness.     Note also how Justice Kennedy
deploys the case for judicial restraint in substantive due
process cases as a reason for taking a lamentably narrow
approach to procedural due process questions.
    In Medina, the Court consulted common-law practice,
found “no settled tradition on the proper allocation of the
burden of proof in a proceeding to determine competency,”47
and therefore rejected petitioner’s historical argument.
But the Court then rather pointedly went on to conduct a
purely instrumental analysis. In the language of the
opinion, “[d]iscerning no historical basis for concluding that
the allocation of the burden of proving incompetence to the
defendant violates due process, we turn to consider
whether the rule transgresses any recognized principle of

  46. Id. at 443-46 (citations and all internal quotations marks omitted).
  47. Id. at 446.
DRIPPSMACRO                  8/11/99 4:16 PM

654           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

‘fundamental fairness’ in operation.”48
     In this portion of the opinion, the Court frankly
consulted instrumental considerations in light of the
special character of the criminal process. The majority
rejected the State’s contention that precedents upholding
shifting the burden of proving the insanity defense to the
defendant compelled rejection of Medina’s claim because
competence goes to the present ability to conduct one’s
defense, not the past ability to understand or control one’s
behavior. Nor was the case analogous to burden allocation
cases involving suppression motions, where the object is
deterring police misconduct. The Court considered, but
rejected, the argument that competence is so vague a
concept that borderline cases must be resolved in favor of
the defendant. In the end, what due process requires is “a
reasonable opportunity to demonstrate that he is not
competent to stand trial.”49 Given that the preponderance
of the evidence standard makes the burden of proof only a
tie-breaker, the State’s procedure satisfied this standard.
     Note that Justice Kennedy supports his position not
with the text of the Fourteenth Amendment, nor with any
inference from its historical purposes. And while the
Medina opinion relies on Patterson, it openly admits that
other cases have taken quite different analytical
approaches. Justice Kennedy’s case for choosing Patterson
relies on federalism and the absence of standards to guide
due process analysis, not on any consistent command from
the precedents. The various attitudes brought together by
the Medina opinion make a good deal of sense so long as
questions about criminal procedure are asked in
substantive due process terms. Substantive due process
has a questionable textual warrant, an unsavory history,
and lies at the center of the violent arguments about Roe v.
Wade.50 So the justices are naturally a little ginger about
substantive due process, and they look to history and
tradition to constrain the exercise of their constitutional

  48. Id. at 448.
  49. Id. at 451.
  50. 410 U.S. 113 (1973).
DRIPPSMACRO                    8/11/99 4:16 PM

1999]            MISCARRIAGES OF JUSTICE                              655

chancellor’s foot.
     But there is another way of asking criminal procedure
questions, in terms of procedural due process. In a
substantive due process analysis, the question is: Does the
state’s procedure violate the unenumerated limits on the
state’s police power, by offending fundamental rights
validated by tradition?     In a procedural due process
analysis, the question is: Does the state’s procedure
subject the individual to an unacceptably high risk of an
erroneous decision?
     Indeed, hidden within the Medina opinion is the ghost
of procedural due process, struggling to be heard. If we
strip away the references to fundamental fairness, what
the Medina Court actually did was to recognize two ways
for a defendant to challenge a state rule on procedural due
process grounds. One approach is historical; if a founding-
era consensus against the state’s procedure can be shown,
the state loses. The other approach is instrumental, i.e.,
whether the procedure is unfair “in operation.” Despite all
the language in Medina about history, it seems pretty clear
that the historical avenue is illusory, leaving in practice
only the instrumental test.
     Suppose a criminal defendant shows a complete
common-law consensus to the effect that a perfectly reliable
type of evidence should be excluded. For example, a child
abuse defendant moves to bar the victim, now six years old,
from testifying. At common law, this would have been the
usual result, on the ground that a child of tender years
cannot understand the oath and its potentially eternal
consequences.51 Would any court do anything but laugh at
the defendant’s due process argument, despite its
impeccable historical pedigree?
     On the other hand, suppose the state excludes reliable
defense evidence that would not have been admissible at
common law. For example, suppose a state procedural rule
has the effect of preventing the defendant herself from

   51. See 2 J. Wigmore, Evidence § 508 (James Chadbourn ed. 1979) (citing
Hale and describing common-law rebuttable presumption of incompetence below
ages fourteen and seven).
DRIPPSMACRO                      8/11/99 4:16 PM

656           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

testifying. At common law, the defendant could not be
sworn as a witness (again on the theory that this made
damnation too likely).52 Could the state justify its rule on
the basis of common-law history? Suppose a state barred
coconspirators from testifying on one another’s behalf.53 Or
prohibited a defendant from cross-examining a witness who
gave testimony favorable to the state after being called by
the defense?54 All of these cases actually reached the
Supreme Court, and in each the Court struck down the
state rules, history notwithstanding.55
     In other words, once the “fundamental fairness”
language gets put in perspective, and once the historical
analysis is understood as pretty much beside the point,
Medina holds that the bottom-line inquiry in a criminal-
side procedural due process case is instrumental. A great
many other cases implicitly follow this instrumental
approach. Procedural due process cases recognize, under
appropriate circumstances, constitutional rights to a
change of venue to avoid local prejudice;56 to the disclosure
of exculpatory evidence during the discovery process;57 to
rules of evidence that exclude unreliable prosecution
evidence58 and that admit reliable exculpatory evidence;59 to
public funds to pay for expert witnesses for indigent

   52. See id. § 575.
   53. At common-law joint defendants were barred from testifying as parties,
although they could give evidence if charged separately. See id. at 832.
   54. For the common-law pedigree of the so-called voucher rule, see IIIA
Wigmore, Evidence ‘ 896 (James Chadbourn ed. 1970).
   55. See Washington v. Texas, 388 U.S. 14 (1967) (striking down bar on co-
conspirator testimony under sixth amendment compulsory process clause);
Chambers v. Mississippi, 410 U.S. 284 (1973) (striking down voucher rule on due
process grounds); Rock v. Arkansas, 483 U.S. 44 (1987) (striking down on
compulsory process grounds ban on testimony by accused who had been subjected
to hypnosis prior to trial).
   56. Rideau v. Louisiana, 373 U.S. 723 (1963).
   57. United States v. Bagley, 473 U.S. 667 (1985); Brady v. Maryland, 373 U.S.
83 (1963).
   58. Foster v. United States, 394 U.S. 440 (1969) (suggestive lineup excluded
on due process grounds).
   59. Crane v. Kentucky, 476 U.S. 683 (1986) (due process and compulsory
process analysis); Chambers v. Mississippi, 410 U.S. 284 (1973).
DRIPPSMACRO                      8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                  657

defendants,60 against knowing use of perjured testimony by
the prosecution;61 to appear at trial in civilian clothes, not
prison issue;62 to limiting closing argument by the
prosecutor;63 and to the reasonable doubt standard of proof
respecting elements of the charged offense.64
     In each instance, the Court’s inquiry has been
instrumental; in each case, history has not been
dispositive; and in each case, the state practices held
inconsistent with due process do not shock the conscience.
Implicit in these due process cases is the appropriate
standard for judging due process challenges to adjudicatory
procedures, which is whether the challenged practice
creates a significant and avoidable risk of unjust
conviction.     This standard resembles the Mathews
standard, but it takes account of the distinctively weighty
individual interest against erroneous “termination” of one’s
liberty from criminal conviction.
     In these criminal-procedure due process cases the
Court does not follow the Betts v. Brady65 case-by-case
approach. Instead, modern due process cases establish
principles that apply in categorical terms to future cases.
For example, under Brady v. Maryland the government
must turn over relevant, exculpatory evidence on request,
in every case. Under Winship, the government must prove
the elements of the charge beyond reasonable doubt in
every case. Sometimes doctrine is stated in such fact-
specific terms that the standard requires case-by-case
adjudication, as with questions about changes of venue to
counter pretrial publicity. In the main, however, the Court
has applied the painfully-learned lessons about regulating
police investigation to the regulation of trial procedure.
Due process, no less than the Fourth or Fifth Amendments,

   60. Ake v. Oklahoma, 470 U.S. 68 (1985).
   61. Napue v. Illinois, 360 U.S. 264 (1959); Mooney v. Holohan, 294 U.S. 103
   62. See Estelle v. Williams, 425 U.S. 501, 505 (1976) (holding that defendant
had waived right to appear in street clothes).
   63. Miller v. Pate, 386 U.S. 1 (1967).
   64. In re Winship, 397 U.S. 358 (1970).
   65. 316 U.S. 455 (1942).
DRIPPSMACRO                     8/11/99 4:16 PM

658           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

can be translated into generally applicable rules.

B. Eyewitness Identification Evidence: Diagnosis and
   Prescription Applied

     Erroneous identification evidence remains the single
leading cause of false convictions. The Warren Court
grappled with the problem in the 1960’s, but relied
primarily on the Sixth Amendment right to counsel rather
than on due process.66 The resulting disappointment offers
a good example of how selective incorporation analysis ill
serves the innocent.
     In the first place, the right to counsel approach does
not focus on the underlying problems with the reliability of
the evidence. Giving the suspect a lawyer before a lineup
does . . . . what? The lawyer can testify as a witness about
suggestiveness later on, but any accurate recording of the
session could achieve that much.67 The Fifth Amendment
privilege gives the suspect the right not to answer police
questions, so lawyers can advise clients not to talk during
interrogation. By contrast, the suspect has no Fifth
Amendment right not to appear before the witness,68 and so
the lawyer cannot stop the proceeding. The Wade opinion
never indicated that the police conducting the lineup must
in any way accommodate the objections or suggestions
made by defense counsel.
     Second, the Court’s focus on the Bill of Rights soon
undermined even the modest protection supplied by the
Wade rule. Given the Fifth Amendment privilege, suspects
can be questioned only before the right to counsel attaches.
Therefore it was imperative that the right to counsel attach
late enough in the process for the police to get a crack at
the suspect under the Miranda rules. Fairly supported by
the language of the Sixth Amendment, the Court held that
the “prosecution” does not commence until formal charges

  66. See United States v. Wade, 38 U.S. 218 (1967).
  67. See id. at 254 (White, J., dissenting).
  68. See id. at 221-23.
DRIPPSMACRO                     8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                 659

are filed.69 Thus by delaying accusation until after the
lineup, the police can entirely circumvent the Wade rule.
     Still later, the Court held that even after accusation
the right to counsel does not apply to showing a photo
array to an eyewitness.70 This obviously does not look like
a court hearing; the defendant is not present, the
prosecutor is not present, there is no element of
confrontation between the defendant and the state.
Responding to appearances, the Court held that the photo
array is not a “critical stage” of the proceedings, like a
preliminary hearing or a corporeal identification
     There remains a vestigial due process test, in which
the defense can move to exclude a pretrial identification
obtained under needlessly suggestive circumstances so
strong as to make erroneous identification highly
probable.71 This determination is made on a case-by-case
basis, with no guidance to police for future cases. If the
motion succeeds, the eyewitness may still be allowed to
identify the accused at the trial itself, on the theory that
the in-court identification—rendered under the most
suggestive circumstances imaginable—is independent of
the prior suggested identification.72
     Here we have, in one area of doctrine, the worst of both
selective incorporation and fundamental fairness. The
doctrinal failure is not solely an abstract concern of
Constitution-toting academics. It sends, as the DNA cases
painfully demonstrate, innocent people to prison. It will, in
due course, send an innocent person to execution.
     What can be done about it? To a degree the limitations
on identification testimony are irreducible; perception and
memory just are not very good,73 especially under stress.74

   69. Kirby v. Illinois, 406 U.S. 682 (1972).
   70. United States v. Ash, 413 U.S. 300 (1973).
   71. Stovall v. Denno, 388 U.S. 293, 301-02 (1967); Neil v. Biggers, 409 U.S.
188 (1972); Manson v. Braithwaite, 432 U.S. 98 (1977).
   72. See, e.g., Wayne LaFave & Jerold Israel, Criminal Procedure ‘ 7.4(f) (2d
ed. 1992).
   73. See, e.g., Elizabeth Loftus, Eyewitness Testimony (1979).
   74. See id. at 33.
DRIPPSMACRO                     8/11/99 4:16 PM

660           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

The situation, however, is by no means hopeless. In a
valuable article,75 Gary Wells and Eric Seelau have
attempted to devise practical procedures that would reduce
the frequency of erroneous identifications. They defend
four basic principles.
     First, the lineup or photo spread should be
administered by someone who does not know which of the
subjects is the suspect.76 This is one way, and probably the
only way, to terminate the risk of suggestion. Second, the
eyewitness should be told that the lineup or photo array
may not include any suspects at all.77 Eyewitnesses do a
passable job of identifying the offender when the offender is
in the lineup or array. When the offender is not in the
sample, eyewitnesses have a strong tendency to select the
person who most resembles the offender. Instructing the
witness that a sample may not have a suspect in it reduces
this tendency.
     Third, Wells and Seelau suggest that the lineup or
array include foils who resemble, not the suspect, but the
description of the offender given by the witnesses.78
Putting the suspect’s twin in the lineup would halve the
probability of a correct identification from a perfectly
reliable eyewitness. The objective should be to include foils
who resemble the offender. Thus relative judgment errors
should tend to exculpate, rather than inculpate, the
innocent, without arbitrarily reducing the likelihood that a
guilty suspect will be identified. Finally, Wells and Seelau
recommend obtaining a statement of confidence from the
witness at the time of the identification, so that confidence
at trial is not misinterpreted as confidence at the time of
the identification.79
     If my analysis of procedural due process in criminal
cases is plausible, courts could convert proposals such as

  75. Gary L. Well & Eric P. Seelau, Eyewitness Identification: Psychological
Research and Legal Policy on Lineups, 1 Psychol. Pub. Pol’y & L. 765 (1995).
  76. Id. at 775-778.
  77. Id. at 778-779.
  78. Id. at 779-780.
  79. Id. at 780-781.
DRIPPSMACRO              8/11/99 4:16 PM

1999]         MISCARRIAGES OF JUSTICE                     661

those of Wells and Seelau into Miranda-like rules of
constitutional law. These rules would have Miranda’s
great virtue of clarity. To the extent the police know what
procedures are required, exclusion for violations would be
both rare and appropriate.
    The obvious corollary to such a regime would be a due-
process based ban on any in-court identification evidence.
Evidence rendered under such suggestive circumstances is
unreliable, and a non-suggestive neutral lineup or photo
array easily can be arranged outside the courtroom. Jurors
should be told the reason that they will not hear an in-
court identification, so that they do not entertain
unreasonable doubts about the prosecution’s case.
    Putting these elements together, due process should be
held to require the exclusion of all in-court identification
evidence, and to permit proof at trial of identification when:
DRIPPSMACRO                  8/11/99 4:16 PM

662           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

      (I) The imminent death or incapacity of the witness
          made an immediate identification process
          necessary, and the identification procedure was
          conducted as fairly as the circumstances permitted;

      (II)     The defendant was not in custody, and the
             eyewitness made the identification after being
             shown either

              (1) An open-ended series of photographs (i.e.
                 mugbooks), when the eyewitness is told that
                 the offender may not be included in any of
                 the photographs, and the eyewitness is
                 asked for a confidence statement after any
                 identification of a suspect; or

              (2) A photo array that

              1. Was accurately recorded, by videotape or

              2.    Included photographs of at least five
                   persons other than the defendant;

              3.    Included only photographs that bear a
                   reasonable resemblance to the description
                   of the offender given by the eyewitness;

              4.    Was administered at a time when the
                   police had no suspects, or by persons who
                   were ignorant of the identity of any
DRIPPSMACRO                  8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                 663


              5. Was conducted after the eyewitness was
                 instructed that the witness may be shown
                 multiple arrays, and that the offender
                 might not be in any of them;

              6. In fact involved the presentation of at least
                 two photo arrays;

              7. Included a confidence statement by the
                 eyewitness after any identification; or

     (III) The identification was made after a suspect is in
         custody, at a lineup or photo array, and:

              1. An accurate record of the proceedings was
                 made, by videotape or otherwise;

              2. The proceeding was administered by
                 persons who were ignorant of the suspect’s

              3. The eyewitness was instructed that the
                 witness may be shown multiple lineups or
                 arrays, and that the suspect may not be in
                 any of them; and

              4. At least two lineups or arrays are in fact
                 shown to each eyewitness.
DRIPPSMACRO                 8/11/99 4:16 PM

664           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

              5. That the suspect, after consulting with
                 counsel, have the right to be placed in
                 either the first or second lineup or array
                 shown to the eyewitness.

              6. Each lineup or array shown to the
                 eyewitness consists of at least five persons
                 (including the suspect if the suspect is
                 included in that array or lineup);

              7. The suspect is given the opportunity,
                 through counsel, to strike one of the foils
                 suggested by the government; and to
                 suggest four foils, one of whom must be
                 accepted by the government.

              8. If the eyewitness identifies the suspect as
                 the offender, the eyewitness must be asked
                 to state his or her degree of confidence in
                 the accuracy of the identification.

     In emergencies, of course, it makes sense to take
advantage of what may be the last chance to obtain
evidence from an eyewitness. When no suspect is in
custody, it is impossible to involve the suspect or counsel
for the suspect in the process. On the other hand, when no
suspect is in custody the police frequently have no suspect
in mind, so that the risk of suggestion is diminished.
Nonetheless, safeguards of the sort proposed could reduce
the risk of misidentification considerably.
     Once the suspect is in custody, there seems to be no
valid ground to refuse the defense some due process rights
that might be exercised through counsel at identification
proceedings. The participation of counsel here costs the
government nothing except a slight delay. Given actual
due process rights against suggestive procedures, defense
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                    665

counsel has an important role to play during the
identification session. The due process rights that call for
elections by the suspect after consulting with counsel are
whether to appear in the first or the second group; whether
to strike a government-supplied foil, and, if so, which one;
and whether to put forth a group of foils for the defense,
and if so, what foils. If the defense does not supply three
foils the government need not accept any of them. The
process of striking and supplying foils is designed to make
sure that the panel includes foils who resemble the
offender. While that might be done by post hoc review
under a reasonableness standard, it seems far more
desirable to design procedures that can be satisfied in every
case, and that, when satisfied, conclusively establish the
fairness of the proceedings.
     Since counsel here is required by due process, Kirby v.
Illinois80 is no barrier to the proposed regime.81 Note also
that the regime contemplates no subsequent inquiry into
reliability; if the government complies with the specified
procedures, the identification will be admissible, and not
otherwise. Further questions of reliability would be for the
jury. Only when the government invokes the emergency
exception, or proceeds with an array when the suspect is
not in custody, would it run the risk of exclusion under a
general standard rather than a specific rule. If the
involvement of counsel is thought too burdensome, then it

   80. 406 U.S. 682 (1972).
   81. See Morrissey v. Brewer, 408 U.S. 471 (1972) (due process right to counsel
at parole revocation hearings); Gagnon v. Scarpelli, 411 U.S. 778 (1973) (no per se
due process right to counsel at parole hearings, but indicating that due process
might require appointed counsel in exceptional cases). In Lassiter v. Dep’t of
Social Services of Durham County, 452 U.S. 18 (1981), the Court, dividing 5 to 4,
rejected a due process claim for appointed counsel in custody-termination
proceedings. Lassiter, however, like Gagnon, left open the possibility that in
particular cases the appointment of counsel might be constitutionally-required.
See id. at 32-33; Garramone v. Tomo, 94 F.3d 1446 (10th Cir. 1996) (holding that
due process required appointment on facts of instant case). Douglas v. California,
372 U.S. 353 (1963) recognized a right to appointed counsel on appeal, based on
the equal protection clause. Miranda v. Arizona, 384 U.S. 436 (1966) recognized a
right to counsel during police interrogation based on the fifth amendment. There
is thus good authority for the proposition that the constitution may require the
appointment of counsel even when the sixth amendment does not.
DRIPPSMACRO                    8/11/99 4:16 PM

666           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

would make sense to treat post-arrest identification
proceedings under the same rules suggested for pre-arrest
photo arrays under heading (II)(2).
     The costs and burdens of the approach might give
some pause.         Such skeptics should reconsider the
seriousness of their commitment to avoiding erroneous
convictions. Moreover, because the available evidence
seems to show that eyewitnesses do no worse with
photographs than with corporeal lineups,82 we might well
move toward substituting scrupulously fair photo arrays in
favor of suggestive lineups. Administratively, one obvious
possibility would be to have the crime lab administer
identifications; but the critical point is not that the
identification be identified by persons with any particular
competence, only that the composition of the panel, and the
eyewitness making the decision, be as free as possible from
the influence of anyone who knows who the suspect is.
Security is not an issue—at least until a photograph grabs
an officer’s gun and escapes.
     The approach defended here recognizes that a fair trial
can do little to remedy an unfair investigation. As Lloyd
Weinreb pointed out twenty years ago, once the police have
gained physical control of the situation, there is little
justification for leaving further factual investigation wholly
in their hands.83 The constitutional text confines selective
incorporation scrutiny to the trial. Only a turn to due
process can protect the innocent against slipshod or rigged
investigations. To succeed, that turn must think about due
process not in terms of fundamental fairness in isolated
cases, but as rules of procedure that the police can follow in
every case. Cases such as Miranda84 and Belton85 prove
that constitutional rules are possible. The identification
cases make a compelling case for formulating new

  82. See B.L. Cutler et al., Conceptual, Practical, and Empirical Issues
Associated with Eyewitness Identification Test Media, in Adult Eyewitness
Testimony: Current Trends and Developments 163 (D.F. Ross et al. eds. 1994).
  83. Lloyd Weinreb, Denial of Justice (1977).
  84. Miranda v. Arizona, 384 U.S. 436 (1966).
  85. New York v. Belton, 453 U.S. 454 (1981).
DRIPPSMACRO                    8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                 667

constitutional rules based on due process, rather than the
Bill of Rights.


     The contemporary constitutional law of criminal
investigations, crafted by the Burger and Rehnquist
Courts, may fairly be characterized as one long reaction
against the Warren Court landmarks. Ironically, however,
because Mapp, Massiah, and Miranda never did all that
much to impede good policework,86 trimming back the
Fourth and Fifth Amendment rights, and the exclusionary
remedy available to defendants has not really done much to
promote law enforcement.          The major obstacle legal
doctrine poses to convicting more of the guilty is the
hearsay rule coupled with the Confrontation Clause. On
this front, however, the Supreme Court did most of what
needs to be done on behalf of law enforcement almost thirty
years ago.
     In California v. Green,87 the Court, Justice Brennan
alone dissenting, recognized two important limits on the
scope of the Confrontation Clause. First, if the declarant
does testify at the trial, the Constitution permits the use at
trial of any prior statement by the witness as substantive
evidence so long as the declarant is subject to cross-
examination about the pretrial statement at the trial.88
Second, preliminary hearing testimony that was subject to
cross-examination by the defendant at the time it was
given may be admitted so long as the declarant either
testifies at the trial or is unavailable to testify at trial.89
Many years later the Court held, in United States v.
Owens,90 that a witness is subject to cross-examination for

  86.   See supra note 28.
  87.   399 U.S. 149 (1970).
  88.   Id. at 153-164.
  89.   Id. at 165.
  90.   484 U.S. 554 (1988).
DRIPPSMACRO                      8/11/99 4:16 PM

668           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

Confrontation Clause purposes even if the witness has
little recollection of the statement or of the crime. All that
is required is that the witness answer the questions on
cross, without taking a contempt sanction or claiming a
privilege. The declarant is subject to cross even if the
answers to the questions are “I don’t remember” or “you
must be confusing me with my evil twin brother.”
      The criminal justice system has not taken full
advantage of the principles recognized in Green. One
reason for this lies with the Federal Rules of Evidence. In
adopting the rules, Congress rejected a proposal to make
prior inconsistent statements by testifying witnesses
admissible as substantive evidence.91 Congress instead
opted for a requirement that the prior statements be given
before a grand jury or in some other formal “proceeding.”
While only a few states permit no substantive use of prior
inconsistent statements,92 by no means all permit
substantive use of all prior inconsistent statements of
testifying witnesses. Many follow the federal rule, and
others impose somewhat different limits on those prior
statements that may be admitted as substantive evidence.93
      Often enough a witness called by the prosecution at
trial flips and swears to some implausible exculpatory
account. An inculpatory statement made to a police officer
that does not qualify as an excited utterance will then be
admissible only to impeach. If the government does not
have enough other evidence to take the case to the jury, the
case will be lost on a motion for a directed verdict.
      More commonly, a witness will fail to cooperate by
telling the prosecutors that she will not testify. Few
jurisdictions have been willing to use the contempt power
to force reluctant witnesses to the stand, and even then the
witness may, from fear or loyalty, take a contempt sanction

   91. For the text and history of FRE 801(D)(1), see, for example, Christopher
Mueller & Laird Kirkpatrick, 1998 Federal Rules of Evidence 178-90 (1998).
   92. Jack B. Weinstein et al, Evidence: Cases and Materials 529 (9th ed. 1997)
(only 3 states forbid all substantive uses of prior statements).
   93. See id.; Stanley Goldman, Guilt by Intuition: The Insufficiency of Prior
Inconsistent Statements to Convict, 65 N.C. L. Rev. 1, 45-48 (1986).
DRIPPSMACRO                     8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                                669

rather than testify. Also common is the situation in which
the witness indicates a willingness to testify before trial,
but then fails to show up on the appointed date.
     In these cases the first prong of Green offers an
important opportunity. Suppose that the defense had an
opportunity to cross-examine witnesses on the record in the
immediate aftermath of the crime.            If a videotaped
deposition could be taken immediately, witnesses might
give statements in the heat of the event, before their
integrity is tempted by fear, greed, or loyalty.
     Under the Federal Rules of Evidence, such a
deposition is a perfect insurance policy. If the declarant
testifies consistently with the deposition, there is no need
for the deposition. If the declarant refuses to testify or is
unavailable, the deposition comes in as former testimony.
And if the declarant testifies inconsistently with the
deposition, the deposition comes in under FRE 801(d)(1)(b).
     Of course we do not take depositions in the criminal
practice. Illustrative is Rule 15 of the Federal Rules of
Criminal Procedure, which authorizes depositions only in
“exceptional circumstances.”94 The rule contemplates that
parties may depose only their own witnesses, and the
courts have construed “exceptional circumstances”
narrowly. Depositions may be taken only to preserve
testimony, not for discovery.95 The courts view defense
depositions as tactical gambits for unauthorized
discovery,96 and government depositions as reducing the
incentives to present live testimony.97
     There is another unstated barrier to early depositions
in criminal cases.       To be admissible under Green,
deposition testimony needs to be subject to cross-
examination by counsel98 (unless the accused waived

   94. Fed. R. Crim. Pro. 15(a).
   95. See Charles Alan Wright, Federal Practice and Procedure—Criminal 2d
§ 241(1982 & 1998 pocket part).
   96. United States v. Kelley, 36 F.3d 1118, 1124-25 (D.C. Cir. 1994); United
States v. Cutler, 806 F.2d 933, 935-36 (9th Cir. 1986).
   97. United States v. Mann, 590 F.2d 361, 367 (1st Cir. 1978).
   98. Pointer v. Texas, 380 U.S. 400 (1965).
DRIPPSMACRO               8/11/99 4:16 PM

670           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

counsel), and the defense needs to know the charges before
conducting an effective cross. The defendant, shielded by
the Fifth Amendment privilege, is deemed ineligible for the
rudimentary discovery allowed to civil litigants. And
because police interrogation cannot succeed in the presence
of defense counsel advising the suspect to invoke the
privilege, the idea of providing counsel at the moment of
arrest is unthinkable.        Thus the gap between the
investigation and the judicial process has never been closed
as Green would permit. In short, the fear that extending
the right to counsel to the moment of arrest would abolish
police interrogation has prevented taking usable evidence
from the witnesses until long after the crime.
     It does not have to be that way, especially in this era of
sophisticated technology. Imagine that police, summoned
to the scene of an armed robbery of a convenience store,
arrest a suspect a few blocks from the store based on a
general description from the clerk and the customers.
Present practice is for the police to escort the suspect back
to the store, where the witnesses will identify the suspect
under the most suggestive possible conditions. The police
will then administer Miranda warnings and drive the
suspect to the station for booking and a bindover hearing in
court. The descriptions given by the witnesses to the police
will not be admissible at any later trial, and the
identifications will be admissible only if the witnesses
     Meanwhile defense counsel will not enter the picture
until the bindover hearing at the earliest, and in many
cases not until a subsequent arraignment. During this
window the suspect might confess, but it is at least as
likely that the witnesses will be reached or lose interest.
     Now suppose that a different approach were taken, an
approach that aims at taking statements usable in
evidence at the earliest possible moment. At the scene, the
police might videotape statements from the eyewitnesses,
giving descriptions of the offender and any details of the
offense. These recorded statements may be admissible as
excited utterances, and they might also prove useful for
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                    671

impeachment or for refreshing recollection at trial.
Moreover, Green leaves legislatures free to authorize the
receipt of such prior inconsistent statements as substantive
     Suppose further that as soon as charges are filed and
counsel is retained or assigned for the defense that
videotaped depositions of the witnesses are taken. This
could be done immediately after arraignment,99 days,
rather than months, after the crime. Once those
depositions are in hand, reaching the witnesses later will
not help the defendant.           If the witness becomes
unavailable, the deposition testimony will be admissible
under the former testimony exception. If the witness
testifies inconsistently with the deposition testimony, the
deposition testimony will be received not just to impeach,
but as substantive evidence as well. If the witness testifies
consistently with the deposition testimony, the deposition
testimony is unnecessary. No matter what the witness
does at trial, the inculpatory story will be told.
     If that were the case, the percentage in witness
intimidation would decline once the depositions were given.
Bribery or witness intimidation would not help the
defendant, but it could land the perpetrators in the
penitentiary.     Organized gangs may still perceive an
incentive to retaliate against witnesses to intimidate other
witnesses in the future. Even in this situation, however,
knowledge that the pending charge cannot now be defeated
will reduce, albeit not eliminate, the incentives for
intimidation and corruption.100 The more rational the gang,

   99. Pretrial procedure varies a good deal among jurisdictions.                By
“arraignment,” I mean the first occasion at which the accused enters a plea to the
charge after it is read in court. Since entry of the plea requires that counsel be
retained, appointed, or waived, and can occur only after the charge is filed, the
prerequisites of admissible pretrial testimony under Green are present at this
  100. See Graham, supra note 31, at 280 n.1:
   No one is asserting that preservation of testimony will stop all intimidation
   of witnesses. Live testimony accusing the defendant of a crime will be
   more persuasive to a jury than hearing a prior statement of the witness
   read by the attorney for the government. Moreover, members of organized
   crime may believe it is advantageous in the long run to make examples of
DRIPPSMACRO                      8/11/99 4:16 PM

672           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

the more likely its leaders are to realize that fixed costs are
     Fears of unreciprocated defense discovery are
misplaced, at least so long as both sides can depose any
witness as of right. Obviously the defendant could not be
deposed because of the privilege against self-incrimination,
but defense witnesses can be deposed as easily as
government witnesses. Guilty defendants typically know
the identities of the witnesses against them from first hand
knowledge, so that depositions would do little to encourage,
and much to forestall, witness intimidation. Discovery as
such is a good thing, and there should be powerful reasons
indeed to justify unnecessary surprises at trial. Indeed,
many prosecutor’s offices maintain an open file policy to
convince defense lawyers to accept plea offers, a practice
that would not exist if defense discovery had pernicious
     As for the fear of trial by deposition, the proper focus
for this concern is the hearsay rule, which governs
admissibility of deposition testimony, not the rules of
procedure, which only determine whether a deposition can
be taken. If the unavailability requirement is enforced, the
government will still be under pressure to produce the
witnesses to testify at the trial.
     There is, however, a strong argument that video-taped
depositions justify a much more casual attitude toward
unavailability. Both Green and Ohio v. Roberts101 implied

    anyone who testifies even after the damage has been done and the
    defendant convicted. Assuming some murders will occur, the question
    remains whether the murders that do occur will be significantly fewer than
    would have occurred without preservation of testimony.             Currently,
    murders do more than reduce impact or set an example—they also
    eliminate damaging testimony that might put the perpetrator in jail. One
    trusts that a preservation proceeding will create a distinction that amounts
    to a significant difference.
Videotape of course reduces—and may well eliminate—the gap between the
impact on juries of preserved and live testimony.
  101. 448 U.S. 56 (1980). I say “implied” because the Court found that the
declarant was in fact unavailable; the opinion lays down a since-sometimes-
followed test that demands a showing of unavailability of the declarant plus a
showing that the hearsay statement is of a reliable sort. Strictly speaking,
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                   673

that former testimony from an available declarant who
does not testify would violate the Confrontation Clause, but
both cases involved a cold transcript, not a videotape.
Since Roberts the availability requirement comes and goes
as the cases seem to require.102
     Given videotaping, a strong argument can be made to
abolish the unavailability requirement in the former
testimony exception. In the first, place videotaping enables
the jury to observe demeanor.103         Empirical evidence
overwhelming indicates that ordinary people cannot detect
deception by observing demeanor.104 The preference for live
testimony, however, does not depend entirely on the ability
of jurors to observe symptoms of deceit. Live testimony
may improve attentiveness and retention.           Empirical
evidence seems to show that videotaped testimony is at
least as good as live testimony on this front.105 The
enthusiastic reception of videotaped depositions in civil
cases is instructive, if not conclusive.
     In the second place, the opponent of the deposition

however, the Court did not hold in Roberts, and hasn’t held anywhere else, that a
conviction obtained with a cross-examined statement from an available witness
must be reversed.
  102. Compare id. (confrontation clause requires production of available
declarants) with United States v. Inadi, 475 U.S. 387 (1986) (confrontation clause
does not require showing of unavailability of declarants of coconspirator
statements) and Idaho v. Wright, 497 U.S. 805 (1990) (applying Roberts
unavailability requirement to statements offered under catchall exception) and
White v. Illinois, 502 U.S. 346 (1992) (declining to impose Roberts unavailability
requirement to excited utterances or statements made for purposes of medical
  103. A point noted by many courts with respect to depositions in civil cases.
See Michael Henke & Craig Margolis, The Taking and Use of Video Depositions:
An Update, 17 Rev. Litig. 1, 14 n.57 (1998).
  104. See Jeremy A. Blumenthal, A Wipe of the Hands, A Lick of the Lips: The
Validity of Demeanor Evidence in Assessing Witness Credibility, 72 Neb. L. Rev.
1157 (1993); Olin Guy Wellborn III, Demeanor, 76 Cornell L. Rev. 1075 (1991).
  105. See Albert W. Alschuler, Implementing the Criminal Defendant’s Right to
Trial: Alternatives to the Plea Bargaining System, 50 U. Chi. L. Rev. 931, 1023-
1024 (1983); Gerald Miller et al., The Effects of Videotape Testimony in Jury
Trials: Studies on Juror Decision Making, Information Retention, and Emotional
Arousal, 1975 BYU L. Rev. 331 (1975); Daine M. Hartmus, Videotrials, 23 Ohio
N.U. L. Rev. 1, 6-11 (1996).
DRIPPSMACRO                       8/11/99 4:16 PM

674           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

testimony can subpoena the readily available witness.106
Indeed in criminal cases it would be unconstitutional to
deny the subpoena power to the defense. If live testimony
is critical to the defense, and the declarant is readily
available, nothing prevents the defense from taking
advantage of the Compulsory Process Clause.
     As a general matter the government quite properly
bears the burden of producing the witnesses, because the
government is in a better practical position to monitor
them.107 If, however, prompt videotaped depositions were
taken, a strong argument could be made that, for all
practical purposes, the government had produced the
witnesses. True, live testimony would add somewhat to the
reliability of the inquiry, but the addition would be far less
than what live testimony adds over a cold transcript. A
defendant who insists on that marginal difference might
well be asked to bear the burden of summoning the
     Under current law, trial judges frequently bend over
backwards to admit hearsay statements that were not
recorded, were not sworn, and were not cross-examined
when witnesses refuse to testify or give incredible stories
on the stand.108 If depositions were routinely taken in
criminal cases immediately after arraignment, and if those

  106. The Supreme Court has placed some importance on this point. See United
States v. Inadi, 475 U.S. 387, 396-398 (1986); Dutton v. Evans, 400 U.S. 74, 88
n.19 (1970) (plurality opinion).
  107. If a witness is not produced by the prosecution on its own initiative under
the confrontation clause, the defendant can always produce the witness at his
own request under the compulsory process clause. The question under the
confrontation clause is therefore not whether the witness should be produced at
all, but whether the witness is better produced on the prosecution’s initiative or
at the defendant’s request.
Peter Westen, Confrontation and Compulsory Process: A Unified Theory of
Evidence for Criminal Cases, 91 Harv. L. Rev. 567, 622 (1978).
  108. For example, in Williamson v. United States, 512 U.S. 594 (1994), the trial
judge admitted a statement to police from an intimidated witness, in which the
declarant, who had himself been caught with a large quantity of illegal drugs,
described the defendant as the mastermind behind the offense. The Supreme
Court reversed, but the dynamics of the situation cannot be cured by isolated
Supreme Court decisions. Trial judges forced to choose between hearsay and
silence will be inclined to opt for the hearsay.
DRIPPSMACRO                       8/11/99 4:16 PM

1999]              MISCARRIAGES OF JUSTICE                                   675

depositions were routinely admissible at trial even without
a showing of unavailability, witnesses would know that
they could perform their civic duty promptly, need only do
so once, and that they would reduce their exposure to
illegal pressures by doing so. In some cases, this would
mean that videotaped cross-examined testimony would
substitute for live cross-examined testimony. In many
other cases, it would mean that videotaped cross-examined
testimony would substitute for police testimony about
“excited utterances,” or for grand jury transcripts now
admitted under rule 801(d)(1), or under the catch-all
exception, or for no evidence whatsoever.
     The right to physically confront adverse witnesses,
recognized in Coy v. Iowa109 and Maryland v. Craig,110 does
not stand in the way of videotaped depositions in criminal
cases. To begin with, all hearsay from nontestifying
declarants comes into evidence without physical
confrontation. Compared to an excited utterance or a
coconspirator statement, cross-examined testimony on
videotape is far more probative. Justice Scalia’s concern
with physical confrontation is grounded on the fear that
lying comes more easily in the absence of the person falsely
accused.111 Assuming this to be true, it is still the case that
cross-examination is a far stronger check on deception than
physical confrontation. Statements cross-examined by the
defense, with notice of the charge, should not be kept from
the jury for want of physical confrontation anymore than
business records or dying declarations. Indeed, the circuit
courts of appeal have admitted grand jury testimony from
unavailable declarants under the catch-all exception when
the testimony bears special indicia of reliability.112 If
statements given without cross-examination or physical

  109. 487 U.S. 1012 (1988).
  110. 497 U.S. 836 (1990).
  111. See Coy, 487 U.S. at 1017-18; Craig, 497 U.S. at 867-68 (Scalia, J.,
  112. See, e.g., United States v. Carlson, 547 U.S. 1346 (1976); United States v.
Doerr, 886 F.2d 944 (7th Cir. 1989); Curro v. United States, 4 F.3d 436 (6th Cir.
DRIPPSMACRO                    8/11/99 4:16 PM

676           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

confrontation can be received, it seems fair to argue that
cross-examined statements should be received without
physical confrontation.
     Even if a different view were taken of the right to
physical confrontation, depositions could be taken at a
location proximate to the lock-up so that the defendant
could be physically present during pretrial testimony.
Bulky police officers with reputations for overzealousness
might be assigned for security purposes. Alternatively, it
might be possible to take testimony, at least from especially
vital witnesses, at the arraignment itself, converting
arraignment into a preliminary hearing. The security and
administrative burdens might seem out of proportion to the
gain for the truth-finding function, but it would be possible
to arrange for physical confrontation, as well as cross-
examination, of pretrial testimony.
     In sum, the biggest obstacle to convicting more of the
guilty is witness noncooperation, and the best way to
attack the noncooperation problem is to take admissible
statements from the witnesses as quickly as possible.
Professor Graham recognized this opportunity more than a
decade ago.113 Professor Graham, however, favored a
special preservation proceeding, with the defendant
represented by a sort of guardian ad litem, over criminal-
side depositions. Depositions, he argued, take too long and
require the (often intimidating) presence of the defendant.
These concerns, although cogent, do not strike me as
insurmountable. Construed as I suggest, the Confrontation
Clause permits use of statements cross-examined by
defense counsel in the defendant’s absence. If the Clause is
construed     otherwise,    the    physical     confrontation
requirement would apply to any preservation proceeding,
whatever its form.
     That leaves the time factor, but depositions could be
taken as soon as charges are filed. A special preservation
proceeding could not satisfy constitutional requirements
until that time anyway, because meaningful cross-

 113. See Graham, supra note 31, at 252-80.
DRIPPSMACRO                8/11/99 4:16 PM

1999]             MISCARRIAGES OF JUSTICE                677

examination depends on notice of the charge. Professor
Graham’s approach would be an improvement on present
law (and illustrates how much room the Constitution leave
for legislatures interested in addressing the major cause of
lost convictions). I tend to think, however, that cross-
examination by the defendant’s own lawyer is important
enough to justify whatever irreducible delay might
separate depositions testimony from the filing of charges.
     Professor Graham makes a quite important point that
deserves to be carried over into any system of early pretrial
testimony in criminal cases. He notes that when testimony
is taken before defense counsel has an opportunity to
investigate the case, cross-examination questions must be
asked in ignorance of the probable answer.114 His proposal
to permit the defense, but not the government, to introduce
pretrial testimony elicited on cross makes excellent sense,
and would help to strengthen any system of pretrial
testimony against Confrontation Clause attack.
     The Constitution permits prompt pretrial testimony so
long as the defendant gets to cross-examine the witnesses
through counsel with knowledge of the charge. Our
peculiar interrogation regime depends on keeping lawyers
out of the process as long as possible; but the key to
overcoming the witness noncooperation problem is to get
defense lawyers into the case as early as possible. Even
given our present interrogation rules, however, the
government could benefit from the option of securing
admissible witness accounts in favor of approaching the
suspect for a Miranda waiver.
     If our interrogation rules were changed, however, this
election would be unnecessary. If there were some system
for questioning the suspect even after the appointment of
counsel, the defense would gain discovery, but would be
subject to some reciprocal obligations. Nor would there be
any loss of evidence from the suspect, except to the extent
that police questioning methods are more effective than
questioning in court.

 114. See id. at 270-71.
DRIPPSMACRO                       8/11/99 4:16 PM

678           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

     Suppose that state legislation or police regulations
provided that the right to counsel attaches at the moment
of arrest, but that the failure of the suspect to give a
deposition after arrest, while represented by counsel, could
be considered by the jury.115 Suppose further that this
policy made prior convictions inadmissible to impeach
deposition testimony given by the criminal defendant.116
     Griffin v. California117 might be distinguished on at
least two grounds. First, Griffin involved an inference of
guilt from the failure to testify at the trial. With live
testimony before the jury, excesses by the prosecution on
cross and misstatements by the accused cannot be
retracted or corrected. By contrast, if the defendant’s
deposition involved honest mistakes or prosecutorial
browbeating or trickery, these problems could be brought
out to the jury at a later trial, or, in appropriate cases, by a
motion to exclude or redact the deposition.
     Second, absent possible impeachment by prior
convictions, there is far less reason for the accused to
decline to speak. Accordingly, the inference of guilt from
silence is much more rational in a system such as I suggest,
and Griffin, in the end, rests on an equation between
irrational inferences of guilt and constitutionally-forbidden
“compulsion.”      Practically speaking, wholly reliable
evidence, like a well-conducted DNA test or a surveillance
film of the defendant committing the crime, “compels” the
defendant to take the stand and testify to duress, insanity,
consent or what not. Such rational inferences are quite
properly not deemed compulsion, and a counseled suspect’s
refusal to make a statement immune to impeachment with
prior convictions really does support a rational inference of
     In this system, defense lawyers would be welcomed by

  115. Similar proposals have been around a long time. See Donald Dripps,
Foreword: Against Police Interrogation—and the Privilege Against Self-
Incrimination, 78 J. Crim. L. & Criminology 699, 730 n.120 (1988) (citing
proposals of Pound, Kauper, Frankel and Schaefer).
  116. Again, this is not a novel idea. See, e.g., Van Kessel, supra note 3, at 540.
  117. 380 U.S. 609 (1965).
DRIPPSMACRO              8/11/99 4:16 PM

1999]         MISCARRIAGES OF JUSTICE                     679

the prosecution, for defense counsel enables the taking of
depositions from both the witnesses and the accused.
Presently, weeks separate the preliminary hearing from
the crime. Given present interrogation law, depositions
could be taken after arraignment, within days of the crime.
But if counsel could be appointed for the accused at the
time of arrest, prosecutors could issue an information and
take cross-examined statements from the witnesses within
hours of the crime.
    Even as presently construed in Green and Roberts, the
Constitution leaves room for legislatures and courts to
authorize prompt post-arraignment depositions in criminal
cases. This would enable a promising attack on the single
biggest cause of unjust dismissals.           Relaxing the
unavailability requirement for videotaped testimony under
the Confrontation Clause would enable a more effective
response, and authorizing questioning of the suspect after
the right to counsel attaches would permit even quicker,
and so more effective, questioning of the witnesses.


      The conventional wisdom is almost completely wrong.
With the important if qualified exception of Gideon v.
Wainwright, the criminal procedure revolution has done
little to protect innocent people from unjust conviction,
although it has done a great deal to protect innocent people
from arbitrary searches and oppressive questioning.
Modern studies agree with Borchard, Radin, and the
Franks about the causes of unjust convictions.             The
disturbing evidence of DNA exonerations during
investigations should give some urgency to the obvious
need for constitutional doctrine to do more to protect the
innocent. The most plausible doctrinal approach relies on
due process, not on the Bill of Rights.
      As for convicting the guilty, the critics of Miranda may
be right, in the wrong way, when they point to Miranda as
a loophole for the guilty.            Miranda means that
interrogation depends on excluding defense counsel from
DRIPPSMACRO               8/11/99 4:16 PM

680           BUFFALO CRIMINAL LAW REVIEW [Vol. 2:637]

the critical investigative phase of the case. Participation of
defense counsel, however, is necessary to both any effective
system of depositions in criminal cases and to any system
of in-court questioning of the accused. The inability to take
depositions opens the door for defendants to intimidate
witnesses and for witnesses to protect defendants or give
up on the process. Of course, any system of interrogation
that depends on excluding counsel would have the same
effect, so Miranda’s critics—who liked Escobedo even less
than Miranda—have no real cause to complain.
     The Confrontation Clause does not doom us to current
case attrition figures. Even in an interrogation regime
dominated by Miranda, prompt depositions in criminal
cases make sense. We should not forget, however, that
Miranda itself invited experiments with other forms of
regulated interrogation.       Generalizations in criminal
justice are always hazardous, but this one might be
ventured with some confidence: We need to make the
investigation less inquisitorial and more adversarial, and
we need to make the trial process less adversarial and
more inquisitorial. Each of the proposals defended here—
due-process rules to govern identification proceedings,
prompt depositions in criminal cases, and permitting an
inference of guilt from a defendant’s refusal to give
deposition testimony—are all plausible steps toward such a

To top