Amicus Curiae Brief — Brecht v. Abrahamson

IN THE

Supreme Court of the Gnited States

OCTOBER TERM, 1992

TODD A. BRECHT,

Petitioner,

vs.

GORDON A. ABRAHAMSON, Superintendent,

Dodge Correctional Institution,

Respondent.

On Writ of Certiorari to the

United States Court of Appeals for the Seventh Circuit

- BRIEF AMICUS CURIAE OF THE

CRIMINAL LEGAL FOUNDATION

INSU OF RESPONDENT

KENT S. SCHEIDEGGER

CHARLES L, HOBSON*

2131 J. Street (9581 er

Post Oiltice Bex 1199.

CA 95812

Sacramento,

Telephone: (916) 446-0345

Attorneys Amicus Curiae

Criminal fared Level Foundation

*Attorney of Record

“ss

q

F

é

3

q

:

°

4

> ’

a

\ a Pt gis ¥ © intial ee 2. ayy Z *

=) ee SE +} : *

wn mY x + om oa ~ ¢

_ ad So ghee

S

QUESTION PRESENTED

Should the standard for determining whether constitu-

tional error is harmless continue to be the reasonable

doubt test of Chapman v. California?

(i)

TABLE OF CONTENTS

Ee icceaeescecedvoncescscoccess i

cers cc kcsceeesecccccvccecese iv

ins ccc cegececoeececesess l

Summary of factsandcase ...............5sceeees a

Cine eeddovccccesevceccoeses 2

DC PCERb dd sbwbescessevcecccccececsocces 3

I

Stare decisis does not preclude a re-examination

ED 5 cc ccccsccssccccosccess 3

A. Stare decisis Te 4

B. The limits of stare decisis .... 2... 0.006600, 6

C. Chapman as precedent ..............6555. 9

Il

The reasonable doubt standard of Chapman should be

DUCE CN Se hash nesecceneccececscccece 12

A. Appellate courts and reasonable doubt ..... 12

B. The Chapman standard is too high ......... 13

C. Chapman is inconsistent ................. 19

D. Chapman is unnecessary ................. 20

itl

The standard of Kotteakos v. United States is the

proper standard for replacing Chapman .......... 23

(iii)

IV

What “error” there was, if any, is harmless

Os 2k EE re ee te

eee e ee &

TABLE OF AUTHORITIES

Cases

Arizona v. Fulminante, 113 L. Ed. 2d 302,

ED cccuswatearcccews secs 17, 26

Brecht v. Abrahamson, 944 F. 2d 1363

Ci ddacstvdandbanseeveesvaseds 2, 23

Bruton v. United States, 391 U. S. 123, 20 L. Ed. 2d 476,

EE cu iddédddcdsbanbiecnecuces 16

Burnet v. Coronado Oil & Gas Co., 285 U. S. 393,

76 L. Ed. 815, 52 S. Ct. 443 (1932) ............... 7

Chapman v. California, 386 U. S. 18, 17 L. Ed. 2d 705,

SPE CE chduedvcsdeastanucestan Passim

Cohens v. Virginia, 6 Wheat. (19 U. S.) 264,

FS bn 606.0 b0-60.dd se beri cecendes 25

Delaware v. Van Arsdall, 475 U. S. 673, 89 L. Ed. 2d 674,

Ss ED ob k on cdeseetnasss ives 17, 29

Doggett v. United States, 120 L. Ed. 2d 520,

PE ED ca Bdedccdeeskqusbcsuess 18

Doyle v. Ohio, 426 U. S. 610, 49 L. Ed. 2d 91,

EE ode Gss onde bh Ghee nn sbaiee'ess 27

Estes v. Texas, 381 U.S. 532, 14 L. Ed. 2d 543,

PEPE cucccdbcccckhehececeschxko4 21

Fahy v. Connecticut, 375 U.S. 85, 11 L. Ed. 2d 171,

i socsncksdanseekeevnenes ees 19

Gideon v. Wainwright, 372 U. S. 335, 9 L. Ed. 2d 799,

Pe CE acute veneers ceeedeliaes ets 20

Griffin v. California, 380 U. S. 609, 14 L. Ed. 2d 106,

oe Seer feiveuneesen 22

Harrington v. California, 395 U. S. 250, 23 L. Ed. 2d 284,

SUNS EE ab ab vevkeucedeucctuds 14, 16

Helvering v. Hallock, 309 U. S. 106, 84 L. Ed. 604,

cece ndhc scant wennae ess 4

Hertz v. Woodman, 218 U. S. 205, 54 L. Ed. 1001,

hai. ot ids de scewes aso 7

Hutchins v. Wainwright, 715 F. 2d 512 (CA11 19 5).. 20

In re Winship, 397 U. S. 358, 25 L. Ed. 2d 368,

SG PEON vo cccccdcocevdscsnescs 19, 23

Jackson v. Virginia, 443 U. S. 307, 61 L. Ed. 2d 560,

FP CED UeeES bbb es ccdcendcaene 12, 20

Johnson v. Zerbst, 304 U. S. 458, 82 L. Ed. 1461,

PE DUPED ccc ccececcesceseceseessces Q

Kotteakos v. United States, 328 U. S. 750, 90 L. Ed. 1557,

66 S. Ct. 1239 (1946) ...... 9, 10, 14, 22, 23, 24, 25, 26

Mapp v. Ohio, 367 U. S. 643, 6 L. Ed. 2d 1081,

es ED 6 ch one sees iexacden nse ess 20

Marks v. United States, 430 U. S. 188, 51 L. Ed. 2d 260,

SPER TED Knvccccondesccanuvepecveces 5

McDonough Power Equipment, Inc. v. Greenwood,

464 U. S. 548, 78 L. Ed. 2d 663,

SP Ee EEE newcnsccccccvnsesessesecs 15

Milton v. Wainwright, 407 U. S. 371, 33 L. Ed. 2d 1,

Pe ED 6. 006.6606 66d eEdceceeresess 14

Miranda v. Arizona, 384 U. S. 436, 16 L. Ed. 2d 694,

PENG ED oc e see ccecccctvnnnecsoces 20

Patterson v. McLean Credit Union, 491 U. S. 164,

105 L. Ed. 2d 132, 109 S. Ct. 2363 (1989) ........ 7,8

Payne v. Arkansas, 356 U. S. 560, 2 L. Ed. 2d 975,

PEED bp vccovecesccocesceceeesess 26

Payne v. Tennessee, 115 L. Ed. 2d 720,

Se Es TGEEEED Kcddctcnsvscvecccess 5,7, 11

Mt

Vil

Planned Parenthood v. Casey, 120 L. Ed. 2d 674,

Ty aa 4, 5, 6, 8,9, 11

Rose v. Clark, 478 U. S. 570, 92 L. Ed. 2d 460,

ee Oe 2 12

Rose v. Lundy, 455 U. S. 609, 71 L. Ed. 2d 379,

Raa 22, 26

Sawyer v. Smith, 111 L. Ed. 2d 193,

On eas 21

Schneble v. Florida, 405 U. S. 427, 31 L. Ed. 2d 340,

92 S. Ct. 1056 (1972)... 0... cece eee eee eee 16, 17

State v. Brecht, 421 N. W. 2d 96 (Wis. 1988) ... 2, 27, 28

Stone v. Powell, 428 U. S. 465, 49 L. Ed. 2d 1067,

Se ED Soveccccasecncecedpeoues’s 21

United States v. Bryan, 339 U. S. 323, 94 L. Ed. 884,

WE Ge MP EEIUED cvceccocciesces eee ie ask 8

United States v. Hasting, 461 U. S. 499, 76 L. Ed. 2d 96,

ee Gnas voceseseosccave’s 12, 13, 28

United States v. Lane, 474 U. S. 438, 88 L. Ed. 2d 814,

Se Py PEP CUI ho ccc ccesesescs 23, 24, 25, 27

United States v. Mechanik, 475 U. S. 66, 89 L. Ed. 2d 50,

eos vos eGo aw ea wes ta 0 x 15

United States v. Lehder-Rivas, 955 F. 2d 1510

in a0t.ous sss sees Rwenneeieas autos 20

United States v. Scott, 437 U. S. 82, 57 L. Ed. 2d 65,

rik nce Javea neue pabe ene ee 8

United States v. Title Ins. & Trust Co., 265 U. S. 472,

68 L. Ed. 1110, 44S. Ct. 621 (1924) .............. 4

United States v. White, 124 F. 2d 181 (CA2 1941) ... 18

Vasquez v. .liilery, 474 U.S. 254, 88 L. Ed. 2d 598,

6 con ceccceennenanneess 4,5,6

Vill

Wainwright v. Sykes, 433 U. S. 72, 53 L. Ed. 2d 594,

Pe SE EOED bcc ccccnncedeswewgneees e's 18

Williams v. Kaiser, 323 U. S. 471, 89 L. Ed. 398,

ee EEE ococcuetGhecandeteatebdeses 21

Wolf v. Colorado, 338 U. S. 25, 93 L. Ed. 1782,

69 S. Ct. 1359 (1949) ...... Lcdicuseveeseeseves 23

Yates v. Evatt, 114 L. Ed. 2d 432,

See EEE ocheciudusvancusves 12, 1748

United States Constitution

a i a ee ees 10

United States Statutes

A A Perry eee ee eee 10, 23

Former 28 U. S. C. § 391, 40 Stat. 1181 ............ 24

Rules of Court

Federal Rule of Criminal Procedure 52(a) ......... 27

Treatises

J. Wigmore, Evidence (Tillers rev. 1983) ........... 14

R. Traynor, The Riddle of Harmless Error

PEE ssa ccebetneewoabun es 13, 14, 15, 19, 22, 25, 27

Miscellaneous

Douglas, Stare Decisis,

49 Cotem. L. Rev. 735 (1949) ..... ccc cccces 4,11

The Federalist No. 78 (A. Hamilton)

I bis o0bRbocccescccccceses 5, 6

Field, Assessing the Harmlessness of Federal

Constitutional Error—A Process in Need of a Rationale,

BaP Wie WO Me ROU. BO CEDFED oc ccc ccvcecccscces 15

Friendly, /s Innocence Irrelevant? Collateral

Altack on Criminal Judgments,

38 U. Chi. L. Rev. 142 (1970) .............., 19, 26

Goldberg, Harmless Error: Constitutional Sneak Thief,

71 J. Crim. L. & Criminology 421 (1980) ......... 16

Mouse, Harmless Constitutional Error:

The Implications of Chapman v. California,

Se DENN. E. TUOV. SEP CTPG) oc ccc scccvcccccees 20

Rehnquist, The Notion of a Living Constitution,

Pe Bs ry PU vc cvvccceceescsneses rae

Note, Harmful Use of Harmless Error in Criminal Cases,

64 Cornell L. Rev. 538 (1979) ............... iS, 21

Saltzburg, The Harm of Harmless Error,

ee ED coca nscncvicerccesses 23

IN THE

Supreme Court of the United States

OCTOBER TERM, 1992

TopbD A. BRECHT,

Petitioner,

vs.

GORDON A. ABRAHAMSON, Superintendent,

Dodge Correctional Institution,

Respondent.

BRIEF AMICUS CURIAE OF THE

CRIMINAL JUSTICE LEGAL FOUNDATION

IN SUPPORT OF RESPONDENT

INTEREST OF AMICUS CURIAE

The Criminal Justice Legal Foundation (CJLF)' is a

nonprofit California corporation organized to participate

in litigation relating to the criminal justice system as it

affects the public interest. CJLF seeks to bring the

constitutional rights of the victims and of society to rapid,

efficient, and reliable determination of guilt and swift

execution of punishment.

The present case involves the standard to be applied

to determine whether constitutional error in a particular

case is harmless. The current reasonable doubt standard

allows too many reversals for unimportant errors. The

resulting loss of valid conditions for unimportant errors is

1. CJLF has received written consent from both parties to file this brief.

contrary to the rights of victims and society which CJLF

was formed to advance.

SUMMARY OF FACTS AND CASE

Petitioner Brecht was convicted of first degree murder

for shooting and killing his brother-in-law, Roger Hart-

man. State v. Brecht, 421 N. W. 2d 96, 98 (Wis. 1988). At

trial, Brecht put on the defense that the shooting was an

accident. Amongst the evidence used to counter this

_ defense was evidence that on three separate -occasions

Brecht had the opportunity to raise his story about the

accidental shooting of his brother-in-law, but did not do

so. See id., at 100-101. Each of these opportunities

happened before Brecht’s Miranda warnings. See ibid.

During Brecht’s cross-examination and on closing

argument, the district attorney mentioned that Brecht had

never said to anyone before the triai that the shooting was

accidental. Brecht v. Abrahamson, 944 F. 2d 1363, 1365

(CA7 1991).

The Wisconsin Court of Appeal found the district

attorney's statemen's to be violations of Doyle v. Ohio.

Ibid. The Wiscoisin Supreme Court reinstated the

conviction, iinding Doyle error, but finding it harmless.

See id., at 1365-1366. The federal district court granted

Brecht’s habeas petition, again finding Doyle error and

concluding that it was not harmless. /d., at 1366. The

Seventh Circuit found Doyle error but held it harmless

under the siandard of Kotteakos v. United States. Id., at

1375-1376.

SUMMARY OF ARGUMENT

Stare decisis does not preclude a re-examination of this

Court’s decision in Chapman v. California. Stare decisis is

a policy designed to foster public respect for the law by

preserving expectations and advancing the notion of a

coherent body of precedent. Chapman is not insulated

from re-examination, because it invokes minimal reliance

interests and is a constitutional decision.

The reasonable doubt standard of Chapman should be

overruled. Reasonable doubt is a test appellate courts are

ill-suited to apply. It interferes with the jury’s findings,

and it requires reviewing courts to make an extensive

examination of the record.

The reasonable doubt standard is also too high.

Properly applied, it is a rule of near automatic reversal.

This can lead courts, including this one, to back away from

the Chapman standard. Such backtracking can only cause

confusion within the courts.

The time for retaining Chapman has passed. While

reasonable doubt may have been necessary when states

opposed federal authority, that is no longer the case.

State courts can be trusted with a less strict standard.

The proper standard for harmless constitutional error

is found in Kotteakos v. United States. The Kotteakos test

provides a proper, congressicnally-empoweredcompromise

between the need to prevent technical reversal and the

need to preserve the integrity of constitutional rules.

Whatever error, if any, that exists in the present case

is harmless. The allegedly harmful statements were no

more than trivial overstatements.

ARGUMENT

I. Stare decisis does not preclude a re-examination of

Chapman v. California.

The Seventh Circuit's decision below demonstrates the

weakness of the reasonable doubt standard of Chapman v.

California, 386 U. S. 18 (1967). While the court below

limited its holding to the collateral enforcement of

prophylactic rules, its reasoning may be carried much

further. If the Chapman standard may be questioned

under this limited circumstance, then a more general

reappraisal of the harmless error standard is in order.

The first step in appraising whether the Chapman

standard should be replaced is the doctrine of stare decisis.

Refusing to overturn a decision because of stare decisis can

be an act of judicial economy, avoiding a re-examination

of the case by sticking to the narrower grounds of stare

decisis.

But stare decisis is not a shield for the poorly reasoned

or contradictory decision. It “is a principal of policy, and

not a mechanical formula... .” Helvering v. Hallock, 309

U. S. 106, 119 (1940). Therefore, stare decisis should not

be viewed as a means of preserving particular decisions,

but as the method of preserving the policies the doctrine

serves.

A. Stare Decisis.

The fundamental purpose of stare decisis is to preserve

the rule of law. See Planned Parenthood v. Casey, 120

L. Ed. 2d 674, 700, 112 S. Ct. 2791, 2801 (1992) (“Casey”).

This allows people to rely upon the validity of decisions in

conducting their affairs, and it combats the belief that

decisions are the personal whims of judges. See Vasquez

v. Hillery, 474 U.S. 254, 265-266 (1986). By allowing

decisions to stay in place, a coherent body of precedent is

created. This establishes the rule of law.

Thus, the most important role of stare decisis is

preserving expectations. “Stare decisis provides some

moorings so that men may trade and arrange their affairs

with confidence. Stare decisis serves to take the capricious

element out of law and give stability to society.” Douglas,

Stare Decisis, 49 Colum. L. Rev. 735, 736 (1949).

If people did not know what to expect from the courts,

our lives, particularly in the commercial sphere, would

become much more difficult. Contracts could become

voided, the title could change hands, and money could be

lost if precedent is overruled. See, e.g., United States v.

Title Ins. & Trust Co., 265 U. S. 472, 486-487 (1924). Thus

stare decisis is strongest when property and contract rights

are involved. See Payne v. Tennessee, 115 L. Ed. 2d 720,

737, 111 S. Ct. 2597, 2610 (1991).

In criminal cases, different interests are involved.

While reliance is important in substantive criminal law, it

is not protected through stare decisis but rather by limita-

tions on retroactivity. See, e.g., Marks v. United States, 430

U. S. 188, 196-197 (1977). In criminal procedure, reliance

is almost entirely on the side of the prosecution, because

only the prosecution’s victories are subject to appellate or

collateral attack. When a defendant wins an acquittal

based upon an existing procedure, the Double Jeopardy

Clause prevents reversal. Only in rarest situations, such

as restrictions on immunized testimony, will a defendant’s

reliance interest be protected by stare decisis. Criminal

procedure precedent favoring criminal defendants involves

a more general interest in law. See Vasquez v. Hillery,

supra, 474 U. S., at 265-266.

This interest, like reliance, varies in strength with the

type of precedent. It is most important in those extremely

rare cases involving matters of intense national controver-

sy. In these cases, overruling even an obviously wrong

decision may be too expensive to the country.

The Judiciary is the “least dangerous” of the branches

because it has the least powers of compulsion, having “no

influence over either the sword or the purse....” See

The Federalist No. 78, at 465 (A. Hamilton) (Rossiter ed.

1961). Instead, the courts must rely on their persuasion

and prestige ‘o insure popular acquiescence to their

decisions. Thus the Court’s power “lies . . . in its legitima-

cy, a product of substance and perception that shows itself

in the people’s acceptance of the Judiciary as fit to

determine what the Nation’s law means and to declare

what it demands.” Casey, supra, 120 L. Ed. 2d, at 707, 112

S. Ct., at 2814.

In order to preserve its power of persuasion, at times

this Court will preserve some decisions it otherwise would

overrule. This doctrine has two applications. First, when

the number of decisions overturned has reached a satura-

tion point, “a point beyond which frequent overruling

would overtax the country’s belief in the Court’s good

faith.” /d., at 708, 112 S. Ct., at 2815.

The other circumstance involves those rare cases of

exceptional public controversy.

“Where, in the performance of its judicial duties, the

Court decides a case in such a way as to resolve an

intensely divisive controversy... its decision has a

dimension that the resolution of the normal case does

not carry. It is the dimension present whenever the

Court’s interpretation of the Constitution calls the

contending sides of a national controversy to end their

national division by accepting a common mandate

rooted in the Constitution.” Jbid.

In these cases, the need to maintain the public’s belief

in the rule of law can be overwhelming. When it resolves

such controversies, any backtracking by the Court can be

seen as “a breach of faith” with the rest of the public. /d.,

at 709, 112 S. Ct., at 2816. Such a result would cause

“profound and unnecessary damage to the Court’s legiti-

macy, and to the Nation’s commitment to the rule of law.”

Ibid.

Most decisions, however, are not so controversial. In

most cases, the rule of law is meant to assure society that

decisions are based in law, and not in proclivities of

individuals. See Vasquez v. Hillery, supra, 474 U. S., at

265.

B. The Limits of Stare Decisis.

While stare decisis is an important doctrine serving a

useful social policy, it does not have the same force as a

statute or the Constitution. The Judiciary’s role in our

society is as an interpreter of laws. See The Federalist No.

78, at 467 (A. Hamilton) (Rossiter ed. 1961). Therefore,

stare decisis, while respected, cannot deter this or any

other court from its ultimate duty of interpreting the law.

Stare decisis is the servant, not the master, of the law.

This Court has recognized that the doctrine does not

have the force of a rule of law and may be overridden

when appropriate. “Whether it [stare decisis| shall be

followed or departed from is a question entirely within the

discretion of the court... .” Hertz v. Woodman, 218 VU. S.

205, 212 (1910). While it will be followed in most cases,

this is only because usually “it is more important that the

applicable rule of law be settled than it be settled right.”

Burnet v. Coronado Oil & Gas Co., 285 U.S. 393, 406

(1932) (Brandeis, J., dissenting).

Perhaps the most important factor in limiting stare

decisis is the ability of other bodies to overturn this

Court's decisions. This Court is particularly reluctant to

overturn its own statutory interpretations, because Con-

gress “remains free to alter what [this Court has] done.”

Patterson v. McLean Credit Union, 491 U.S. 164, 173

(1989).

Constitutional cases are another matter. Because

““‘correction through legislative action is_ practically

impossible’ ” constitutional cases have much less prece-

dential force than statutory cases. Payne v. Tennessee, 115

L. Ed. 2d 720, 737, 111 S. Ct. 2597, 2610 (1991) (quoting

Burnet, supra, 285 U. S., at 407 (Brandeis, J., dissenting));

see Patterson, supra, 491 U.S., at 172-173. Given the

necessary tension between our democratic ideals and

judicial review under the Constitution, see Rehnquist, The

Notion of a Living Constitution, 54 Tex. L. Rev. 693, 695-

696 (1976), this Court must be ready to re-examine its

constitutional decisions in order to maintain the democrat-

ic nature of our society.

Refusing to re-examine an incorrect opinion that the

public cannot overturn is corrosive to this Court's public

respect. Thus the decision to uphold the incorrectly

decided Lochner v. New York, 198 U.S. 45 (1905) until

1937 helped to damage this Court as a public institution.

See Planned Parenthood v. Casey, 120 L. Ed. 2d 674, 705,

112 S. Ct. 2791, 2813 (1992). “Of course it is embarrass-

ing to confess a blunder; it may prove even more embar-

rassing to adhere to it.” United States v. Bryan, 339 U. S.

323, 346 (1950) (Jackson, J., concurring).

This Court is also more willing to re-examine decisions

that have developed contradictions over time. Thus, this

Court will not allow stare decisis to preserve inconsistent

or difficult to administer decisions. See Patterson, supra,

491 U. S., at 173. Similarly, if the conditions that motivat-

ed a decision change, then there is good reason to over-

rule the prior decision. See Casey, supra, 120 L. Ed. 2d,

at 700, 112 S. Ct., at 2809.

Stare decisis is also weaker in procedural cases. These

decisions do not foster the same reliance interests. In

criminal procedure there is minimal reliance interest, and

nearly all of that is.on the side of the prosecution. See

part I, ante, at 5. Furthermore, changes in procedure

usually will not invoke the same public reaction that

occurs from changes in substantive law. Procedural rules

typically matter only to litigants. As most people are

spared the burden of litigation, changes in procedure are

unimportant to most. No person could or should plan his

life around potential litigation, especially criminal litiga-

tion. Unlike substantive decisions, rules of procedure

have little effect on our daily lives.

The fact that the stare decisis is not a “mechanical

rule” is demonstrated by the hierarchy within the doctrine.

While some decisions are virtually etched in stone, others

warrant more flexibility. For those cases less worthy of

stare decisis, “ ‘the process of trial and error so fruitful in

the physical sciences is appropriate also in the judicial

function.” ” United States v. Scott, 437 U. S. 82, 101 (1978)

(quoting Burnet, supra, 285 U. S., at 406-408 (Brandeis, J.,

dissenting)).

C. Chapman as Precedent.

When deciding whether stare decisis should preserve a

decision from re-examination, this Court “is customarily

informed by a series of prudential and pragmatic consider-

ations designed to test the consistency of overruling a

prior decision with the ideal of the rule of law, and to

gauge the respective costs of reaffirming and overruling a

prior case.” Planned Parenthood v. Casey, 120 L. Ed. 2d

674, 700, 112 S. Ct. 674, 2808 (1992). This involves a look

into the type of decision involved, the reliance interest it

invokes and whether the decision involves a divisive public

controversy. See part 1 A & B, ante, at 4-8. Under all of

these factors, Chapman v. California, 386 U. S. 18 (1967)

invokes-minimal protection from stare decisis.

Chapman invokes minimal reliance interest. The

reasonable doubt standard imposed by Chapman is not

something people will rely upon in their daily affairs.

Both sides in litigation have an interest in keeping the

proceedings free of constitutional error. There is a strong

incentive to avoid constitutional error favoring the other

side. It is the nature of litigation that parties should try to

correct error favoring their opponents at the trial level,

instead of waiting for the appeal. At best, it is a very risky

proposition to forego curing adverse constitutional error

at trial.

Many constitutional rights can be waived by a failure

to press for them at trial. See Johnson v. Zerbst, 304 U. S.

458, 464 (1938). Furthermore, the notion of wily trial

counsel allowing the opponent or the court to commit

error at trial, so that reversal may be obtained upon

appeal, is contrary to the premise behind harmless error

rule. See Kotteakos v. United States, 328 U. S. 750, 759-

760 (1946). Therefore, no litigant should allow adverse

constitutional error to go unchallenged at trial in the

hopes of obtaining a reversal. Any reliance on Chapman

to reverse on the basis of an error which could have been

10

corrected at trial is not legitimate reliance, and it is not

entitled to protection.

Chapman is also a constitutional decision. The

Chapman Court was unclear on its authority for its

decision. The most precise statement comes in the cryptic

passage:

“We have no hesitation in saying that the right of

these petitioners not to be punished for exercising

their Fifth and Fourteenth Amendment right to be

silent—expressly created by the Federal Constitution

itself—is a federal right which, in the absence of

appropriate congressional action, it is our responsibility

to protect by fashioning the necessary rule.” Chapman

v. California, 386 U. S. 18, 21 (1967).

An uncritical appraisal of this passage could lead an

observer to conclude that Chapman is no more than a rule

of court readily changed by Congress. This ignores the

constitutional basis of this Court’s authority in Chapman.

The only authorities for overriding state laws are the

Constitution, the laws enacted under it, and treaties. U. S.

Const. art. VI, § 2. Chapman did not invoke any statute

or treaty. The Chapman rule is either constitutionally

based or it is illegitimate. Cf. Chapman, supra, 386 UV. S.,

at 51 (Harlan, J., dissenting) (Chapman characterized as

a “new constitutional doctrine”).

Chapman's reference to the power of Congress to

legislate in this area is hollow. When Chapman was

decided, Congress had enacted a statute, 28 U.S.C. §

2111, that addressed harmless error. The Chapman Court

recognized that section 2111 did not discriminate between

constitutional and nonconstitutional error. 386 U. S., at

22. In Kotteakos v. United States, 328 U. S. 750, 764-765

(1946), this Court indicated that the predecessor to section

2111 might apply to constitutional error. Kotteakos also

established a standard for determining the harmlessness of

error under section 2111. /d., at 765.

11

Yet Chapman did not rely on section 2111 to establish

a harmless error rule. Its reference to the absence of

“appropriate congressional action” indicates that section

2111 did not give the Chapman Court the authority to

formulate a rule for harmless constitutional error. The

break with section 2111 is completed by the adoption of

the reasonable doubt standard, which is higher than the

one used in Kotteakos. As the Chapman Court ignored

the authority of section 2111 in this case, Chapman must

be a constitutional decision.

The only “congressional action” Chapman would deem

“appropriate” would be an act establishing the same high

standard created in Chapman. It could not mean the

lower standard supplied by section 2111 and Kotteakos. In

order to replace Chapman with a lower standard, either

Chapman will have to be overruled or the Constitution

must be amended. Therefore, like other constitutional

decisions, Chapman should be easier to modify or overrule

under stare decisis. See Payne v. Tennessee, 115 L. Ed. 2d

720, 737, 111 S. Ct. 2597, 2610 (1991).

Some constitutional decisions do warrant more stare

decisis protection than others. These involve particw arly

divisive national controversies that this Court’s decision

has substantially resolved. See Planned Parenthood v.

Casey, 120 L. Ed. 2d 674, 708, 112 S. Ct. 2791, 2815

(1992). Chapman is not such a case. While it is impor-

tant, the harmless error standard does not and is not likely

to divide the country. It involves the sort of technical

issue that the lay public is unlikely to understand well, if

at all. Cf. Douglas, supra, 49 Colum. L. Rev., at 737.

Changing Chapman will not erode the nation’s confidence

in the courts. Cf. Casey, supra, 120 L. Ed. 2d, at 708-709,

112 S. Ct., at 2815-2816.

12

Il. The reasonable doubt standard of Chapman

should be abandoned.

A. Appellate Courts and Reasonable Doubt.

The most evident problem with the Chapman standard

is that reasonable doubt is a test appellate courts are

particularly ill-suited to apply. Asking an appellate court

to find that an error was harmless beyond a reasonable

doubt comes perilously close to second-guessing the jury.

While the reasonable doubt question goes to harmless

error and not guilt, in practice, application of the Chap-

man standard is very similar to a re-examination of the

jury's finding. Determining whether the State has demon-

strated, beyond a reasonable doubt, that the error did not

contribute to defendant’s conviction frequently entails an

examination of the strength of the case under the reason-

able doubt standard.’ See United Siates v. Hasting, 461

U. S. 499, S09 (1983). “Where a reviewing court can find

that the record developed at trial establishes guilt beyond

a reasonable doubt, the interest in fairness has been

satisfied and the judgment should be affirmed.” Rose v.

Clark, 478 U. S. 570, 579 (1986).

Appellate courts are not well-suited to make this type

of finding. A reasonable doubt standard, if properly

applied, imposes an enormous burden on reviewing courts.

Diligent application of the reasonable doubt standard

necessarily entails substantial work. Frequently, detailed

examination of the entire record will be necessary to

ascertain whether there is reasonable doubt. Cf. Jackson

v. Virginia, 443 U. S. 307, 337 (1979) (Stevens, J., dissent-

ing). The reasonable doubt finding in Chapman can be

just as involved as the one for guilt.

2. In some cases the entire record cannot be used to justify a finding of

harmlessness. Where the error prevents the jury from considering some

evidence such as an erroneous presumption, then the evidence not

considered may not support a harmless error finding. See Yates v. Evatt,

114 L. Ed. 2d 432, 449-450, 111 S.Ct. 1884, 1893-1894 (1991).

13

“This Court is far too busy to be spending countless

hours reviewing trial transcripts in an effort to deter-

mine the likelihood that an error may have affected a

jury’s deliberations. In this case the parties did not

provide us with a printed appendix containing any

portion of the trial testimony or with any of the trial

exhibits that are discussed at some length in the

transcript. I have spent several hours reviewing the

one copy of the trial transcript that has been filed with

the Court. But I have not read all of its 1,013 pages,

and I have read only a few of the 450 pages of the

transcript of the suppression hearing. The task of

organizing ana digesting the testimony is a formidable

one.” United States v. Hasting, 461 U. S. 499, 516-517

(1983) (Sievens, J., concurring).

Finally, the Chapman standard’s placement of the

burden on the state is also contrary to good appellate

procedure. Burdens of proof and presumptions are trial,

not appellate tools. “These presumptions and burdens,

which expedite the trial process of recreating the facts that

lead to the litigation are unsuited to the appellate process

of determining whether an error was harmless.” R.

Traynor, The Riddle of Harmless Error 25-26 (1970). By

placing elements of the trial into an appellate process, the

Chapman standard blurs the line between the appellate

courts and the trial courts, further weakening the jury

process.

B. The Chapman Standard Is Too High.

Setting the proper standard for harmless error can be

tricky. Setting the standard too low can subvert the

Constitution. See Chapman v. California, 386 U. S. 18, 23-

24 (1967). Yet if the standard is too strict, then it can

create contempt for the standard and for the Constitution

it seeks to protect. Chapman errs too much towards this

latter problem.

14

The single greatest problem with Chapman is that the

standard, if diligently applied, requires near automatic

reversal. See R. Traynor, The Riddle of Harmless Error 43-

44 (1970). It is always difficult to overcome the test of

reasonable doubt, a standard “approaching certainty.” See

id., at 44.

The difficulty of passing the Chapman test is further

compounded by the nature of the inquiry being made.

Chapman does not involve the relatively straightforward

question put before the jury over whether the prosecution

proved beyond a reasonable doubt that defendant commit-

ted the charged crime. Chapman is less factual and more

psychological. It does not allow the reviewing court to

weigh the evidence itself, but instead requires it to gauge

the error’s impact on the mind of the average jury mem-

ber. See Harrington v. California, 395 U.S. 250, 254

(1969); Milton v. Wainwright, 407 U.S. 371, 377-378

(1972). It is difficult tobe nearly certain about such a

secondhand conclusion.

The difficulty of finding harmless error under Chap-

man runs contrary to the doctrine of harmless error. The

doctrine of harmless error came from our sense of justice.

The rule of per se reversal that was adopted throughout

much of this country was an extreme interpretation of an

English decision, which was so contrary to prior common

law precedent that Wigmore called it “heresy.” See 1 J.

Wigmore, Evidence § 21, at 888 (Tillers rev. 1983). This

led to great injustices as many cases, particularly criminal

convictions were reversed on technicalities. “All the

sentiment is thrown to weight the scales for the crimi-

nal—that is, not for the mere accused, who may be

assumed innocent, but for the man who upon the record

plainly appears to be the offender that the jury has

pronounced him to be.” /d., at 895. Similar sentiments

led the states and Congress to enact harmless error rules.

See Kotteakos v. United Siates, 328 U. S. 750, 759 (1946).

15

The sense of justice that motivates the harmless error

doctrine will fuel substantial resistance to needless restric-

tions on a court’s ability to find error harmless.

“If courts conscientiously apply the Chapman test, they

will usually find themselves compelled to balk at

declaring a belief that there is no reasonable possibility

that an error influenced the jury. The alternative is to

give the test lip service while tacitly discounting it.

The very severity of the test invites such discounting to

preclude well nigh automatic reversal.” Traynor,

supra, at 44 (emphasis in original); see also Field,

Assessing the Harmlessness of Federal Constitutional

Error—A Process in Need of a Rationale, 125 U. Pa. L.

Rev. 15, 61, n. 126 (1976).

Trials are costly for the parties, the jurors, and society.

See McDonough Power Equipment, Inc. v. Greenwood, 464

U.S. 548, 553 (1984). Retrials are even more costly.

Retrial of a reversed conviction requires parties, witnesses,

jurors, and courts “to repeat a trial that has once taken

place.” United States v. Mechanik, 475 U. S. 66, 72 (1986).

This can be particularly burdensome to the victims who

“may be asked to relive their disturbing experiences.”

Ibid. The passage of time makes retrial even costlier. As

memories fade and witnesses disappear, conviction may be

impossible on retrial. Even if there is a conviction, the

delay impedes society's interests in retribution and deter-

rence. /bid. “The doctrine of harmless error conserves

judicial resources by obviating retrial for trivial mistakes.”

Note, Harmful Use of Harmless Error in Criminal Cases, 64

Cornell L. Rev. 538, 565 (1979). As Chapman will offend

a court’s sense of economy as well as its sense of justice,

courts will be even more inclined to evade strict applica-

tion of the reasonable doubt standard.

The difficulty courts have with applying the Chapman

standard is demonstrated by this Court’s harmless error

cases. The Chapman Court criticized the California

harmless error rule because it relied too much upon other

16

“ ‘overwhelming evidence’ ” of guilt. Chapman, supra, 386

U. S., at 23. Instead of looking at whether defendant was

guilty in spite of the error, the Chapman test required

courts to decide whether the error contributed to the

comviction. /d., at 26. But in Harrington v. California,

supra, this Court began to abandon this central tenet of

Chapman.

Harrington was one white man who was tried with

three black men for attempted robbery and first degree

murder. 395 U.S., at 252. At trial, the confessions of

Harrington’s two nontestifyingcodefendants were admitted

in violation of Bruton v. United States, 391 U.S. 123

(1968). Harrington, supra, 395 U. S., at 252. Neither of

these confessions implicated Harrington by name, but each

referred to the fourth conspirator as the “ ‘white guy.’ ”

Id., at 253. The confessions were found harmless because

they were cumulative and “the case against Harrington

was so overwhelming... .” /d., at 254.

Although the majority claimed it was not departing

from Chapman, ibid., the dissent claimed that Chapman

had been overruled. See id., at 255 (Brennan, J., dissent-

ing). While Chapman may not have been dead, it was

certainly maimed. it is difficult to see how two confes-

sions stating that one white man and three black men

committed the crime did not, beyond a reasonable doubt,

contribute to the conviction of the one white man being

iried along with three black men. See id., at 257

(Brennan, J., dissenting); Goldberg, Harmless Error:

Constitutional Sneak Thief, 71 J. Crim. L. & Criminology

421, 427 (1980). This is not to say that Harrington reached

the wrong result. The evidence against Harrington was

overwhelming and consequentially the error was trivial in

the context of the trial. See Harrington, supra, 395 U. S.,

at 252-254. But the result cannot be reached without a

move away from Chapman.

In Schneble v. Florida, 405 U.S. 427 (1972), the

departure from Chapman continued. The strength of the

17

case without the error was the primary issue under

harmless error analysis. “In some cases the properly

admitted evidence of guilt is so overwhelming, and the

prejudicial effect [of the error] is so insignificant by

comparison” that the harmlessness of the error “is clear

beyond a reasonable doubt.” /d., at 430. While the

Schneble Court recognized the standard enunciated in

Chapman, it continued this Court’s move away from it.

“Thus, unless there is a reasonable possibility that the

improperly admitted evidence contributed to the convic-

tion, reversal is not required.” /d., at 432, compare with

Chapman, supra, 386 U. S., at 26 (placing burden of proof

on prosecution). Once again, this Court reached the

correct result; the Bruton error was unimportant in the

context of the rest of the case. See Schneble, supra, 405

U. S., at 430-431. But once again, this Court departed

from Chapman to reach the proper result.

The practice established in Harrington and Schneble,

looking at the strength of the prosecution’s case without

the error, is now the standard analysis. See, e.g., Delaware

v. Van Arsdall, 475 U.S. 673, 684 (1986); Arizona v.

Fulminante, 113 L. Ed. 2d 302, 323-325, 111 S. Ct. 1246,

1258-1260 (1991). If this Court has difficulty reconciling

Chapman with the principle of harmless error, then other

courts must have similar qualms.

In Yates v. Evatt, 114 L. Ed- 2d 432, 111 S. Ct. 1884

(1991) this Court made some effort to get closer to the

original jurisprudence of Chapman. The Yates Court

disavowed any notion drawn from this Court’s prior cases

that harmless error may be found in erroneous presump-

tion cases by examining the entire record. /d., at 449-450,

111 S. Ct., at 1893-1894. The Yates Court recognized that

Chapman's emphasis was on the error’s effect on the jury,

not on the strength of the case. Where an erroneous

presumption kept a jury from considering certain evidence,

that evidence cannot be used as a basis for harmless error.

Id., at 450-451, 111 S. Ct., at 1894-1895.

18

The Yates Court still gave the strength of the case

great importance. “To say that an error did not contrib-

ute to the verdict is, rather, to find that error unimportant

in relation to everything else the jury considered on the

issue in question as revealed in the record.” /d., at 449,

111 S. Ct, at 1893. The inference from Yates is that

where the jury did consider all evidence, then the entire

record may be used to overcome error. See id., at 450,

111 S.Ct, at 1894. Thus, sufficiently overwhelming

evidence should be able to overcome error even after

Yates, provided that the jury considered it.

The problem courts have with Chapman does not stem

from an unwillingness to enforce the Constitution, but

from the difficulty of applying the exacting reasonable

doubt standard to the very difficult question of the error’s

impact on jury deliberations. This Court has replaced this

question with the more factual, and thus less difficult

inquest into the strength of the case against the defendant.

While this test may“work its own justice, it is not an

acceptable solution. Courts should not say one thing and

do another. Chapman should be replaced with a lower

and more workable standard, one that more accurately

reflects what courts are really doing. Cf. Wainwright v.

Sykes, 433 U. S. 72, 94-95 (1977) (Stevens, J., concurring)

(Fay v. Noia standard should be replaced because courts

were not really using it).

“No judge in so extended a trial can avoid on occasion

rulings that on reflection he will see to have been wrong;

but, unless they cut off some really subs‘antial aspect of

the truth, or let in too distracting issues, they are not

important.” United States v. White, 124 F. 2d 181, 186

(CA2 1941) (Hand, J.).

Learned Hand’s admonition is even more important

now. In the last few decades constitutional law has

become immensely more complicated. See Doggett v.

United States, 120 L. Ed. 2d 520, 539, 112 S. Ct. 2686, 2699

(1992) (Thomas, J , dissenting). Instead of the few, simple

19

rules actually adopted in 1791, the Bill of Rights has been

expanded by decision into a “detailed Code of Criminal

Procedure, to which a new chapter is added every year.”

Friendly, Is Innocence Irrelevant? Collateral Attack on

Criminal Judgments, 38 U. Chi. L. Rev. 142, 155 (1970)

(footnote omitted). Chapman might have been appropri-

ate long ago when constitutional law was less complicated,

but today, where constitutional error is much more

pervasive, a less harsh standard is necessary.

C. Chapman Is Inconsistent.

The inconsistencies generated by Chapman come from

two main sources. First, the harshness of reasonable

doubt standard makes courts unwilling to apply it literally.

Therefore, subsequent cases have moved away from the

harsh test of Chapman. See part II B, ante, at 15-19.

The second source of conflict comes from the Chap-

man Court’s purported application of Fahy v. Connecticut,

375 U. S. 85 (1963). The Fahy Court found that constitu-

tional error prejudiced defendant because there was “a

reasonable possibility that the evidence complained of

might have contributed to the conviction.” Id., at 86-87

(emphasis added). The Chapman Court claimed to be

adopting this standard when it enunciated its reasonable

doubt standard, 386 U.S., at 24, but it is difficult to

reconcile these two standards.

The reasonable doubt standard is the highest one in

the law, “approaching certainty.” See R. Traynor, The

Riddle of Harmless Error 44 (1970). Fahy’s “reasonable

possibility” standard is more equivocal than Chapman.

Even when worded in the negative terms of Chapman

“whether there was no reasonable possibility that the error

contributed to the jury’s verdict” the Fahy standard is less

stringent. .

Reasonable doubt protects each individual from

criminal conviction unless the trier of fact is convinced of

the guilt “with utmost certainty.” In re Winship, 397 U. S.

20

358, 364 (1970). This is a “subjective state of near

certitude.” Jackson v. Virginia, 443 U. S. 307, 315 (1979).

When one deals in “possibility,” certainty is not necessarily

certain.

The difference between Fahy and Chapman would be

academic had Chapman formally abandoned Fahy. As the

Chapman Court did not, a court could claim to be apply-

ing the Chapman standard while applying the less rigorous

Fahy test. This has been done by the Eleventh Circuit. In

Hutchins v. Wainwright, 715 F. 2d 512, 517 (CA11 1983),

the court equated the Fahy test with the Chapman stan-

dard. As Chapman “purported merely to apply the

standard previously established in Fahy” the Hutchins

opinion was deemed acceptable. United States v. Lehder-

Rivas, 955 F. 2d 1510, 1517, n. 3 (CA11 1992). Until the

reasonable doubt standard of Chapman and its improper

reliance upon Fahy are removed, contradictions like this

will be possible and legal.

D. Chapman Is Unnecessary.

* Chapman v. California, 386 U.S. 18 (1967) was

decided in the middle of this Court's criminal procedure

revolution. Landmark cases such as Gideon v. Wainwright,

372 U. S. 335 (1963), Mapp v. Ohio, 367 U. S. 643 (1961),

and Miranda v. Arizona, 384 U. S. 436 (1966) had only

recently been decided. These and other revolutionary

changes in criminal procedure that occurred during the

1960’s caused some resentment in the state courts. Their

criminal cases were now subject to a much higher level of

*federal scrutiny than before. Harmless error could have

been a way for state courts to avoid this Court's decisions.

By forcing states to adopt a high standard for harmless

constitutjonal error, Chapman helped ensure that state

courts did not subvert the Constitution. Cf. Mouse,

Harmless Constitutional Error: The Implications of Chap-

man v. California, 53 Minn. L. Rev. 519, 530, n. 76 (1969)

(Chapman standard is stricter than most state tests).

a

21

Chapman was decided when federal authority was still

bitterly resisted. Only a few years earlier a state court of

general jurisdiction answered a federal constitutional

objection by saying “that the ‘case [was] not being tried

under the Federal Constitution.’” Estes v. Texas, 381

U. S. 532, 536 (1965). Furthermore, the “massive resis-

tance” of some states to the federal authority to enforce

civil rights was not yet dead. Therefore, the Chapman

Court may have had good reason to distrust some state

courts.

Fortunately, times have changed. These problems are

inconceivable today. The supremacy of federal law is

universally acknowledged and is routinely applied in the

State courts. See Stone v. Powell, 428 U. S. 465, 493, n. 35

(1976); Sawyer v. Smith, 111 L. Ed. 2d 193, 210-211, 110

S. Ct. 2822, 2831 (1990). State courts may interpret the

Constitution in ways later disapproved, but they no longer

refuse to follow it.

While the times have changed, Chapman has not. Its

overly restrictive standards still needlessly prevent harm-

less error findings. State courts are now as worthy to trust

with the Constitution as any federal court. They “are no

less under duty to observe the United States Constitution

than is this Court.” Williams v. Kaiser, 323 U. S. 471, 482

(1945) (Frankfurter, J., dissenting). Chapman should not

be retained out of any unwarranted fear of state courts.

The two other possible reasons for retaining the

Chapman standard do noi withstand scrutiny. The high

Chapman standard could be justified on the ground that

constitutional rights are more important than lesser rights,

and thus warrant a higher standard for harmless error.

This confuses the type of error with whether the error

actually was harmless. Nonconstitutional error can be just

as prejudicial as constitutional violations. See Note,

Harmful Use of Harmless Error in Constitutional Cases, 64

Cornell L. Rev. 538, 551 (1979). “Fundamental” and

“constitutional” ceased to be synonymous many, many

22

years ago. See Rose v. Lundy, 455 U. S. 509, 543, n. 8

(1982) (Stevens, J., dissenting).

Kotteakos and Chapman amply illustrate that constitu-

tional rules are not inherently more “important” than

nonconstitutional rules. The error in Kotteakos was in the

instruction on conspiracy. It went to the very heart of the

case. The trial court flatly and wrongly told the jury that

there was one large conspiracy, when in reality there were

many separate, smaller conspiracies. Kotteakos v. United

States, 328 U. S. 750, 767-768 (1946). Such an error raises

a grave possibility of conviction of an innocent person,

because the overt act of one conspirator may be imputed

to another. /d., at 771.

Chapman, in contrast, involves a rule with minimal

relation to accuracy of the verdict. The fact of defend-

ant’s refusal to testify is perfectly obvious to the jury.

Griffin v. California, 380 U. S. 609, 621 (1965) (Stewart, J.

dissenting, joined by White, J.) The inference drawn by

the former California rule is logically vaiid and had been

sanctioned by a host of authorities. /d., at 622. The

Griffin rule is based on an external policy unrelated to the

accuracy of the trial, and it is an extreme extrapolation of

that policy. To say that comment on failure to testify is a

higher order of error than grossly wrong instruction on the

nature of the crime would be patently absurd. Yet

Chapman assigns a higher harmless error standard without

regard to the importance of the error.

Furthermore, the high Chapman standard can work to

denigrate constitutional rights, to avoid the strictness of

the Chapman test. See part II C, anie, at 15. “Reversal

for error, regardless of its effect on the judgment, encour-

ages litigants to abuse the judicial process and bestirs the

public to ridicule it.” R. Traynor, The Riddle of Harmiess

Error 50 (1970).

The other reason given to support the Chapman test is

that the reasonable doubt standard is necessary to pre-

serve defendant's right to a finding of guilt beyond a

23

reasonable doubt under Jn re Winship, 397 U.S. 358

(1970). See Saltzburg, The Harm of Harmless Error, 59

Va. L. Rev. 988, 991 (1973). This rationale cannot save

Chapman. Winship was decided after Chapman, thus the

right announced six years later in Winship was an unlikely

motivation for the Chapman Court. Furthermore, the

Chapman standard actually interferes with jury decisions.

The high standard frequently requires a detailed examina-

tion of the trial record which creates an ever present

danger of appellate second-guessing of the jury’s verdict.

See part II A, ante, at 12-13.

Chapman’s reasonable doubt standard may have once

served a useful purpose. It no longer does. It is impracti-

cal to administer, too difficult to meet, and inconsistent

with other precedent. The Seventh Circuit has refused to

extend the Chapman standard to violations of prophylactic

rules on habeas corpus. Brecht v. Abrahamson, 944 F. 2d

1363, 1375 (CA7 1991). Amicus submits that it is better

to overrule an unprincipled decision than simply to refuse

to extend it. “Wisdom too often never comes, and so one

ought not to reject it merely because it comes too late.”

Wolf v. Colorado, 338 U. S. 25, 47 (1949) (Rutledge, J.,

dissenting).

Ill. The standard-of Kotteakos v. United States is the

proper standard for replacing Chapman.

Congress has mandated the proper standard for

determining harmless error in 28 U. S. C. § 2111.’ This

statute and the cases interpreting it, Kotteakos v. United

States, 328 U. S. 750 (1946) and United States v. Lane, 474

U. S. 438 (1986), provide a flexible, practical standard that

3. “On the hearing of any appeal or writ of certiorari in any case, the court

shall give judgment after an examination of the record without regard to

errors or defects which do not affect the substantial nghts of the parties.”

24

strikes a proper balance between preserving the rule of

law and preventing unnecessary retrials.

The Kotteakos Court interpreted the predecessor to

section 2111, former 28 U. S. C. § 391, 40 Stat. 1181. The

Court set out to strike a proper balance, disregarding

unimportant error while maintaining the integrity of the

law. Kotteakos started by recognizing the purpose behind

harmless error. Before the harmless error rule took hold,

the frequency of reversal was so great “that criminal trial

became a game for sowing reversible error in the re-

cord... .” 328 U. S., at 759. Harmless error was meant

to end the game playing and close the loopholes. See id.,

at 759-760.

These principles spurred Congress to action, and this

Court kept them in mind when interpreting ‘the statute.

It recognized that hard and fast rules could not work well

in an area heavily dependent upon the facts of each case.

Id., at 762. The Court also understood appellate courts

could not determine guilt or innocence. /d., at 763. But

courts still had to assess the impact of the error on the

jury's decision. /d., at 764. The case by case nature of

harmless error also mitigated against any systematic use of

presumptions. The reviewing court had to look to the

particular error in the setting of the particular case before

it. See id., at 765-766.

From these principles the following standard emerged.

“The inquiry cannot be merely whether there was

enough to support the result, apart from the phase

affected by the error. It is rather, even so, whether the

error itself had substantial influence. If so, or if one is

left in grave doubt, the conviction cannot stand.” /d.,

at 765 (emphasis added). 5 Te

This standard has several advantages over any competi-

tors. It is a lower standard than the reasonable doubt

standard of Chapman. See Lane, supra, 474 U. S., at 460-

461 (Brennan, J., concurring in part and dissenting in

part). Therefore, it will avoid many of the problems

25

associated with the Chapman standard It will be less

burdensome to apply than Chapman, as reviewing courts

will not need to make the detailed inquiry of the record

necessary to remove all reasonable doubt. See part II A,

ante, at 12-13. Applying the less strict Kotteakos standard

will promote the social good of fewer reversals for unim-

portant errors and make courts more willing to apply this

Court’s harmless error standard. See part II C, ante, at

13-19. Thus, applying Kotteakos instead of Chapman may

actually lead to a greater respect for constitutional rights.

Cf. R. Traynor, The Riddle of Harmless Error 50 (1970).

Another crucial advantage of Kotteakos is that it comes

with congressional authority. Kotteakos only interpreted

the predecessor to section 2111. The difference between

section 2111 and former section 391 does not undermine

the Kotteakos test. If anything, the new statute provides

a less restrictive definition of harmless error. Section 2111

removes the qualifier “technical” from the statute, provid-

ing possibly greater scope for harmless error. See Lane,

supra, 474 U. S., at 448. In any event, Kotteakos supplies

the current standard for nonconstitutional harmless error

in federal courts. See id., at 448-449.

Some will object to Kotteakos on the ground that it was

never meant to apply to constitutional error. See Lane,

supra, 474 U.S., at 460-461 (Brennan, J., concurring in

part and dissenting in part). Kotteakos did not deal with

a constitutional error, but involved a variance between the

proof at trial and the charge in the indictment. See 328

U.S., at 755. Thus any statements in Kotteakos on

constitutional error are no more than dicta. See Cohens

v. Virginia, 6 Wheat. (19 U. S.) 264, 399-400 (1821). --

What the Kotteakos Court did say about constitutional

error was that if error had little effect on the jury “the

judgment should stand except perhaps where the departure

is from a constitutional norm or a specific command of

Congress.” Kotteakos, supra, 328 U. S., at 764-765 (em-

phasis added) (footnote omitted). The use of the equivo-

26

cal “perhaps” indicates Kotteakos could be used for

harmless constitutional error. Footnote 19, which came

after the word “norm” in the passage quoted above is no

more restrictive. It states that cases have been reversed

for “forced confessions... although on other evidence

guilt might be taken to be clear.” /d., at 765, n. 19. This

dealt with just one category of error that, until Arizona v.

Fulminante, 113 L. Ed. 2d 302, 331-332, 111 S. Ct. 1246,

1265-1266 (1991), had not been considered to be subject

to harmless error analysis.‘ See Payne v. Arkansas, 356

U. S. 560, 567-568 (1958).

Kotteakos did not need to dwell on constitutional error

because at that time constitutional law was much less

complicated. Today, however, the Constitution has

become a detailed code of criminal procedure. See

Friendly, Js Innocence Irrelevant? Collateral Attack on

Criminal Judgments, 38 U. Chi. L. Rev. 142, 155-156

(1970). Many of these new rights now protected are far

from fundamental. See Rose v. Lundy, 455 U. S. 509, 543,

n. 8 (1982) (Stevens, J., dissenting). Constitutional law

needs the Kotteakos standard much more today than it did

in 1949. See part II D, ante, at 21.

The other reason not to adopt Kotteakos, that the

Constitution deserves more protection, is dealt with in

part II, ante, at 20-23, which dismisses this argument as a

reason for retaining the Chapman standard. This argu-

ment also denigrates the protection offered by Kotteakos.

As the Court below noted, it is strict enough to protect

federal statutes and rules in federal trials. See Pet. App.

12d. This Court has thought enough of Kotteakos to apply

4. ..nother possible objection, that adopting the Kotteakos standard would

make all constitutional rights subject to harmless error, is meritless. This

Court only needs to adopt the Kotteakos standard for those constitutional

errors that can be harmless. Those rights concerning “the framework

within which the trial proceeds,” Arizona v. Fulminante, 113 L. Ed. 2d

W2, 331, 111 S. Ct. 1246, 1965 (1991) need not be subject to harmless

error.

27

it to the harmless error rule of the Federal Rule of

Criminal Procedure 52(a). See Lane, supra, 474 U. S., at

449. Federal rules and statutes have not collapsed

because of Kotteakos, and neither will the Constitution.

Courts today will not use harmless error to avoid the

Constitution, see part II D, ante, at 20-23, and they should

be given a chance to apply this superior standard.

Adopting the Kotteakos rule has other incidental

benefits. It will lessen the number of standards for

harmless error, a desirable simplification of the law. See

Traynor, supra, at 48. It is also a standard federal courts

are used to applying, thus lessening the costs of transition

from Chapman. But the primary reason for adopting

Kotteakos is respect. Respect for the statute enacted by

Congress, respect for the courts that apply the complex

creature. our constitutional law has become, and respect

for trials that are fundamentally fair.

IV. What “error” there was, if any, is harmless.

The present case demonstrates the difficulty of apply-

ing the exacting standard of Chapman to alleged constitu-

tional error. At trial, three people testified that they had

conversations with petitioner and he never mentiqned

shooting Roger Hartman to any of them. See State v.

Brecht, 421 N. W. 2d 96, 100-101 (Wis. 1988). All of these -

statements occurred before any Miranda warnings. See

ibid. Admission of all three was clearly constitutional.

See id., at 102. Comment on Brecht’s failure to mention

the accident on each of these three occasions would have

been entirely proper.

The alleged violations of Doyle v. Ohio, 426 U. S. 610

(1976) occurred when the prosection made several broad

references to the fact that Brecht had never told his story

until the trial. The prosecutor did not point out that

Brecht had been silent after being given his Miranda

warnings at arraignment. See Brecht, supra, 421 N. W. 2d,

28

at 103. The inference that petitioner did not talk after

Miranda warnings should not constitute a Doyle violation.”

In Vovie, the only silence referred to by the prosecu-

tion was that which occurred after defendant's arrest and

Mirande warnings. See Doyle, 426 U. §., at 612-614. In

this case, the prosecution stated that’ Brecht had never

told his story until trial, he was not punishing petitioner

for heeding the Miranda warnings, but for his failure to

mention the shooting to people he talked to before the

Miranda warnings. It is unlikely that the prosecution even

cared about the post-Miranda silence. The testimony of

the three witnesses was much more damaging than

Brecht’s silence after his initial appearance. Had the

prosecutor simply changed the wording of the offending

Statements to refer only to the testimony of the three

witnesses, the statements would have lost none of their

relevance and been completely insulated from Doyle

attack. The present case involves a poor use of Doyle. It

was meant to protect Miranda's guarantee of silence, not

to correct a prosecutor's syntax.

The fact that courts have found that this violated Dovle

shows how technical constitutional criminal procedure has

become. The district attorney was cognizant of Doyle

error and took steps to avoid it. This must be why he

introduced evidence that Brecht had not been given his

Miranda warnings after the arrest. See Brecht, supra, 421

N. W. 2d, at 100-101. Yet the district attorney was

tripped up because of subsequent slight over-statements

constituting a few lines in a 900-page manuscript. Law as

technical as this needs a less demanding harmless error

standard.

5. The comment in this case is strongly analogous to that in United States

v. Hasting, 461 U. S. 499 (1983), where Justice Stevens found no Griffin

error. /d., at 512 (concurring in the judgment). The majority did not

consider that question properly before the Court, but noted that Justice

Stevens “may well be correct.” /d., at 506, n. 4.

29

Even if this was Doyle error, it was harmless, even

under Chapman. The fact that two of the four courts that

reviewed this case did not find it harmless speaks volumes

about the problems with Chapman. Whatever prejudice

that came from the inference of post-Miranda silence was

overwhelmed by the evidence of Brecht’s failure to bring

up his story during three separate conversations. Thus,

the alleged Doyle error was not prejudicial and therefore

harmless beyond a reasonable doubt.

Harmless error is supposed to “promote|} public

respect for the criminal process by focusing on the

underlying fairness of the trial rather than on the virtually

inevitable presence of immaterial error.” Delaware v. Van

. Arsdall, 475 U. S. 673, 681 (1986). In the present case,

“wo courts applying Chapman have allowed the prosecu-

tion to be tripped up in, at best, a hypertechnical violation

of Doyle. If Chapman can allow this to happen, then it

makes a mockery of harmless error.

CONCLUSION

The judgment of the Court of Appeals for the Seventh

Circuit should be affirmed.

September, 1992

Respectfully submitted,

KENT S. SCHEIDEGGER

CHARLES L. HOBSON*

Attorneys for Amicus Curiae

Criminal Jusiic Justice Legal Foundation

*Attorney of Record