Maxwell v Bishop Brief for Petitioner

62 pages

Document text

The following text has been generated automatically during the document scanning process. It may contain errors, omissions, or incorrect formatting. Please refer to the PDF document if this text is not usable.

Download

By downloading, publishing, or reusing the Materials outside of the Website, you hereby confirm and attest that you have read and agree to the LDF Archival Collection Terms of Use linked below. We draw your attention to the following:

For those Materials owned by LDF, your use of the Material outside the Site is solely for one or more of the following (the “Permitted Purposes”):

Non-commercial research and private study;
Teaching, scholarship and instruction (including distance learning);
Non-commercial print and non-commercial online use;
Criticism, comment or news reporting; or
For people with visual disabilities (See Section 121 of the Copyright Act).
For those Materials not owned by LDF you understand that:
- It is your responsibility to determine whether the way you plan to use the materials falls within the copyright exception of Fair Use.
- For any use outside of Fair Use, it is your responsibility to seek and obtain all required permissions from any existing rights holders.

LDF hereby grants you the right to use LDF owned Materials for the Permitted Purposes without seeking further permission from LDF.

If you publish or redistribute any Materials (including excerpts and quotes), you must include a credit line and/or proper citation to the source of the Material. However, you may not use the name of LDF in any manner which creates any false association between you and LDF or that incorrectly implies any sponsorship or endorsement of you by LDF.

NOTICE: Uses that do not comply with these Terms or that do not fall under Fair Use (or another exception) may make you liable for copyright infringement.

Read the full LDF Archival Collection Terms of Use Here.

I agree to the terms and conditions for use of this content.

If you would like to use this content in other ways, contact us at

This item is featured in:

Color photo of Elizabeth Bartholet, smiling and seated indoors during her oral history interview.

Elizabeth Bartholet Oral History

Collection
- Brief Collection
Taxonomy
Identifier
- 54155a59-bc9a-ee11-be36-6045bdeb8873

Cite this item

Brief Collection, LDF Court Filings. Maxwell v Bishop Brief for Petitioner, 1968. 54155a59-bc9a-ee11-be36-6045bdeb8873. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/6cc0a752-255a-4da0-b687-60d9b1a46400/maxwell-v-bishop-brief-for-petitioner. Accessed November 07, 2025.

Copied!

APPENDIX A

I n the

(Emtrt af tlji' Mtutefr States
October T erm, 1968

No. 622

W illiam L. M axwell,
Petitioner,

— v.—

O.E. B ishop, Superintendent of
Arkansas State Penitentiary,

Respondent.

ON WRIT OF CERTIORARI TO THE UNITED STATES
COURT OF APPEALS FOR THE EIGHTH CIRCUIT

BRIEF FOR PETITIONER

J ack Greenberg
J ames M. N abrit, III
N orman C. A maker
M ichael M eltsner
J ack H immelstein
E lizabeth D uB ois

10 Columbus Circle
New York, New York 10019

George H oward, Jr.
329V2 Main Street
Pine Bluff, Arkansas 71601

A nthony G. A msterdam
3400 Chestnut Street
Philadelphia, Pennsylvania 19104

Attorneys for Petitioner

I N D E X

Opinions Below ...................................................................... ^

9Jurisdiction ............................................................................... *

Questions Presented .............................................................. ^

Constitutional and Statutory Provisions In vo lved ........ 3

0
Statement ...................................................................................

Summary of Argument .......................................................... ®

A rgument
Introduction: The Fact and Implications o f
Arbitrary Capital Sentencing ................................. H

I. Arkansas’ Practice o f Allowing Capital Trial
Juries Absolute and Arbitrary Power to Elect
Between the Penalties o f Life or Death for the
Crime o f Rape Violates the Rule o f Law Basic
to the Due Process Clause ...... ..................... -....... 24

A. The Power Given Arkansas Juries is E s
sentially Lawless ................................................. 27

B. The Grant of Lawless Power in Capital
Sentencing is Unconstitutional........................ 45

II. Arkansas’ Single-Verdict Procedure for the
Trial of Capital Cases Violates the Constitution 66

Conclusion ....................................................................................

PAGE

A ppendix A
Evidence .and Findings Below Relating to Racial
Discrimination by Arkansas Juries in the Exer
cise o f Their Discretion to Sentence Capitally
for the Crime o f Rape .............................................

A ppendix B
Available Information Relating to the Propor
tion o f Persons Actually Sentenced to Death,
Among Those Convicted o f Capital C rim es....... 24a

A ppendix C
Manner of Submission o f the Death-Penalty
Issue at Petitioner Maxwell’s Trial .................... 35a

PAGE

TABLE OF AUTHORITIES

Cases: PAGE

Alford v. State, 223 Ark. 330, 266 S.W.2d 804 (1954)
28, 30, 69

Allison v. State, 204 Ark. 609, 164 S.W.2d 442 (1942) 31
Amos v. State, 209 Ark. 55, 189 S.W.2d 611 (1945) .... 69
Andrews v. United States, 373 U.S. 334 (1963) ............ 71

Ball v. State, 192 Ark. 858, 95 S.W.2d 632 (1936) ........ 31
Baxstrom v. Herrold, 383 U.S. 107 (1966) .................. 45,53
Behrens v. United States, 312 F.2d 223 (7th Cir. 1962),

aff’d, 375 U.S. 162 (1963) ............................................. 71,72
Bevis v. State, 209 Ark. 624, 192 S.W.2d 113 (1946) .... 68
Bieard v. State, 189 Ark. 217, 72 S.W.2d 530 (1934) .... 68
Black v. State, 215 Ark. 618, 222 S.W.2d 816 (1949) -.3 1 , 68
Blake v. State, 186 Ark. 77, 52 S.W.2d 644 (1 9 3 2 ).......... 30
Bonds v. State, 240 Ark. 908, 403 S.W.2d 52 (1966) ..... 69
Boone v. State, 230 Ark. 821, 327 S.W.2d 87 (1959) ..... 29
Boykin v. Alabama, O.T. 1968, No. 642 ................7,12, 20, 22
Brady v. Maryland, 373 U.S. 83 (1963) ........................ 25, 27
Brown v. Board o f Education, 347 U.S. 483 (1954) ...... 25
Browning v. State, 233 Ark. 607, 346 S.W.2d 210 (1961) 69
Bruton v. United States, 391 U.S. 123 (1968) ..............36, 76
Burgett v. Texas, 389 U.S. 109 (1967) ............................ 75
Burns v. State, 155 Ark. 1, 243 S.W. 963 (1922) ........28, 30

Carson v. State, 206 Ark. 80,173 S.W .2d 122 (1943) —. 31
Chambers v. Florida, 309 U.S. 227 (1940) ................. ...34, 53
Cheney v. State, 205 Ark. 1049, 172 S.W.2d 427 (1943) 31
Clark v. State, 135 Ark. 569, 205 S.W. 975 (1918) ........ 69
Cline v. Frink Dairy Co., 274 U.S. 445 (1927) ................ 47
Cole v. Arkansas, 333 U.S. 196 (1948) .............-.............. 49
Coleman v. United States, 334 F.2d 558 (D.C.Cir. 1964) 72

PAGE

Cook v. State, 225 Ark, 1003, 287 S.W.2d 6 ( 1 9 5 6 , .... 31
Couck v. United StatesjM® ™ 5 1 8 58

■ ■ SCnrtis v. State, 188 Ark. 36, 64 S.W.2d 86 (1933) .......

Diaonv. State, 189 Ark. 812, 75 8 ^ .2 6 242 (1934)....... «

“ X l " u . S: 145 (1968, . . . I ................33

Edens v. State 2 3 5 68

........... 31

Fay v. Noia, 372 U.S. 391 (1963) .... ........................
Fergason v. Georgia, 365 U.S. 570 (19 1 ' 68
Fielder v. State, 206 Ark. 5 1 76 S.W.2d 233 (1943 .
Fi„ d s v. State, 203 Ark. 1046 * ^ » 78
Fradv v. United States, 348 4 .4d 84,
Freedman y. Maryland, 380 U.S. 51 (1965) ...................

Gadaden v. United States, 223 2
Gaines v. State, 208A A 293 S^W. ^ ‘Jg. 0| 69
Gerlach v. State, 217 JGfclOS,229 SI7i. .... g> ^ ^ ^
Giaccao v. Pennsylvania, 1Ana w 2 6 1 (1 9 11 )-- 35
Gilchrist v. State, 100 Ark. 330 140 A W . 261 911)..... ^
Gonzales v. United States, 348 U S . 407 1955 •-
Green r. State, 51 Ark. 189 \ 73
Green v. United States, 365 M . 30 ^
Green v. United States, 313 F.2 \ 7-̂

denied 372 U.S. 951 (1963) .........................................

Griffin v. California, 380 U.S. 609 (1965) ....................... 1
Griffin v. Illinois, 351 U.S. 12 (1956) ...............................

Hadley v. State, 205 Ark. 1027, 172 S.W.2d 237 (1943) 31
Haeue v. C.I.O., 307 U.S. 496 (1939) .............. ------..........
Ham v. State, 179 Ark. 20,13 S.W.2d 805 (1929) ......... ^
Hamilton * A l a b « 1 U.(A 52 <19 1 — ~ 34

r ^ ^ ~ - 3 2 4 S.W.2d 520 (1959,....28, 31

Hi,l v. United States 368 VS.42 4 ^ ^ 31

r r ::v S t ; 2 18A; , * * * * « » < « « *
Holden v. Hardy, 169 U.S. 366 (18!l ) -■-"•••“ 35

" 188 a ! ’ 323, 67 S .W ,d 91 (1934, 31

____P„i o a___ _ 447 P.2d 117, 73 Cal.In re Anderson, Cal.2d
Bptr. 21 (1968) ......................................... ’ 50, si, 53, 56

49
In re Gault, 378 U.S. 1 .................... ............ 77
Irvin v. Dowd, 366 U.S. 717 (1961) ......................

Jackson v. '^ .W .O d ®

Jones v. State, 204 Ark. 61, 161 S.W.2d 173 (1942) ...

" ^ " S ^ c i r ^ : : ' 7 4

PAGE

McDonald v. State, 225 Ark. 38, 279 S.W.2d 44 (1955) 34
McGraw v. State, 184 Ark. 342, 42 S.W.2d 373 (1931).... 68
McGuire v. State, 189 Ark. 503, 74 S.W.2d 235 (1934) 67
Malloy v. Hogan, 378 U.S. 1 (1964) ...............................10, 72
Marks v. State, 192 Ark. 881, 95 S.W.2d 634 (1936) .... 31
Marshall v. United States, 360 U.S. 310 (1959) ........... 74
Matter of Sims and Abrams, 389 F.2d 148 (5th Cir.

1967) ................................................................................... 4a
Maxwell v. Bishop, 385 U.S. 650 (1967) ......................... 6
Maxwell v. State, 236 Ark. 694, 370 S.W.2d 113 (1963) .....1,

4,7a
Maxwell v. Stephens, 229 F. Supp. 205 (E.D.Ark. 1964)

aff’d 348 F.2d 325 (8th Cir. 1965), cert, denied, 382
U.S. 944 (1965) .............................................................. 2,4

Mempa v. Rhay, 389 U.S. 128 (1967) ...........................27,71
Moore v. State, 227 Ark. 544, 299 S.W.2d 838 (1957) .... 69
Moorer v. South Carolina, 368 F.2d 458 (4th Cir.

1966) ................................................................................... 4a
Morgan v. United States, 304 U.S. 1 (1938) ................. 49

N.A.A.C.P. v. Button, 371 U.S. 415 (1963) ........ 9,35,47,59
Nail v. State, 231 Ark. 70, 328 S.W.2d 836 (1959) .....31, 35
Needham v. State, 215 Ark. 935, 224 S.W.2d 785 (1949) 30
Niemotko v. Maryland, 340 U.S. 268 (1951) ................. 5

Osborne v. State, 237 Ark. 5, 371 S.W.2d 518 (1963) .... 69

Palmer v. State, 213 Ark. 956, 214 S.W.2d 372 (1948).... 31
People v. Hines, 61 Cal.2d 164, 390 P.2d 398, 37 Cal.

Rptr. 622 (1964) .......................................................41, 51, 72
Powell v. Alabama, 287 U.S. 45 (1932) ........................... 34
Powell v. State, 149 Ark. 311, 232 S.W. 429 (1921) ..... 67

PAGE

Vll

PAGE

Rand v. State, 232 Ark. 909, 341 S.W.2d 9 (1960) ....... 69
Ray v. State, 194 Ark. 1155, 109 S.W.2d 954 (1937) ..... 28
Rayburn v. State, 200 Ark. 914, 141 S.W.2d 532 (1940) 67
Rhea v. State, 226 Ark. 664, 291 S.W.2d 521 (1956) .... 69
Rideau v. Louisiana, 373 U.S. 723 (1963) ....................... 36
Rinaldi v. Yeager, 384 U.S. 305 (1966) ......................... 53
Rorie v. Statq, 215 Ark. 282, 220 S.W.2d 421 (1949) .... 31
Rosemond v. State, 86 Ark. 160, 110 S.W. 229 (1908) .... 35

Sanders v. United States, 373 U.S. 1 (1963) ................... 6
Scarber v. State, 226 Ark. 503, 291 S.W.2d 241 (1956) ....28,

30
Shuttlesworth v. City of Birmingham, 382 U.S. 87

(1965) .....|......................................................................... 58
Simmons v. United States, 390 U.S. 377 (1968) ......... 73,76
Sims v. Georgia, 385 U.S. 538 (1967) ............................... 76
Skaggs v. State, 234 Ark. 510, 353 S.W.2d 3 (1962) ..... 67
Skinner v. Oklahoma, 316 U.S. 535 (1942) ..... ..9,10, 27, 54,

61, 65,71, 75
Smith v. Cahoon, 283 U.S. 553 (1931) ............................. 47
Smith v. State, 194 Ark. 1041, 110 S.W.2d 24 (1937) .... 31
Smith v. State, 205 Ark. 1075,172 S.W.2d 248 (1943).... 28
Smith v. State, 230 Ark. 634, 324 S.W.2d 341 (1959) ..... 28,

30, 31
Specht v. Patterson, 386 U.S. 605 (1967) ....10, 27,49, 70, 75
Spencer v. Texas, 385 U.S. 554 (1967) ........................ 75, 76
State v. Peyton, 93 Ark. 406, 125 S.W. 416 (1910) ....... 34
Stein v. New York, 346 U.S. 156 (1952) ......................... 34

Thompson v. City of Louisville, 362 U.S. 199 (1960) .... 45
Tigner v. Texas, 310 U.S. 141 (1940)............................... 61
Townsend v. Burke, 334 U.S. 736 (1948) ....................... 77
Turnage v. State, 182 Ark. 74, 30 S.W.2d 865 (1930) .... 31

Vlll

United States v. Behrens, 375 U.S. 162 (1963) .....25,71,72
United States v. Beno, 324 F.2d 582 (2d Cir. 1963) ..... 74
United States v. Curry, 358 F.2d 904 (2d Cir. 1965) .... 78
United States v. Jackson, 390 U.S. 570 (1968) .....10, 73, 77
United States v. Johnson, 315 F.2d 714 (2d Cir. 1963)

cert, denied, 375 U.S. 971 (1964) ...............................71,72
United States v. National Dairy Prods. Corp., 372 U.S.

29 (1963) ........................................................................... 60
United States ex rel. Rucker v. Myers, 311 F.2d 311

(3rd Cir. 1962), cert, denied, 374 U.S. 844 (1963) .... 74
United States ex rel. Scoleri v. Banmiller, 310 F.2d 720

(3rd Cir. 1962), cert, denied, 374 U.S. 828 (1963) .... 74

Walton v. State, 232 Ark. 86, 334 S.W.2d 657 (1960) .... 28
Ward v. State, 236 Ark. 878, 370 S.W.2d 425 (1963) .... 69
Watson v. City of Memphis, 373 U.S. 526 (1963) ....... 25
Weakley v. State, 168 Ark. 1087, 273 S.W. 374 (1925).... 69
Webb v. State, 154 Ark. 67, 242 S.W. 380 (1922) .......28, 30
Wells v. State, 193 Ark. 1092,104 S.W.2d 451 (1937) .... 28
Whitus v. Balkcom, 333 F.2d 496 (5th Cir. 1964) ....... 10, 76
Wilkerson v. State, 209 Ark. 138,189 S.W.2d 800 (1945) 31
Williams v. Georgia, 349 U.S. 375 (1955) ....................... 34
Williams v. New York, 337 U.S. 241 (1949) ................... 70
Williams v. Oklahoma, 358 U.S. 576 (1959) ................... 70
Willis v. State, 220 Ark. 965, 251 S.W.2d 816 (1952)....67, 68
Winters v. New York, 333 U.S. 507 (1948) ..................... 47
Witherspoon v. Illinois, 391 U.S. 510 (1968) —.11, 27, 33, 34,

47,48,49,
61, 70, 78

Wright v. State, 243 Ark. 221, 419 S.W.2d 320 (1967)....67,
68

PAGE

Yick W o v. Hopkins, 118 U.S. 356 (1886) ........... 5, 9, 47, 56,
57, 62

Young v. State, 230 Ark. 737, 324 S.W.2d 524 (1959) .... 31

S ta t u t e s :

Federal:
10 U.S.C. $920 (1964) ......................................................... 21
18 U.S.C. $2031 (1964) ........................... ......................... 21
28 U.S.C. $1254(1) (1964) ................................................. 2
28 U.S.C. $1291 (1964) ......... i.......................................... 2
28 U.S.C. $2241(c)(3)(1964) ............................................. 2
28 U.S.C. $2244(b) (Supp. II, 1966) ............................... 6
28 U.S.C. $2253 (1964) ....................................................... 2
28 U.S.C. $2254 (Supp. II, 1966) ..................................... ' 6

State:
Ala. Code $$14-395, 14-397, 14-398 (Recomp. Vol. 1958) 21
Ark. Acts 1915, No. 187, $ 1 ............................................... 2®
Ark. Stat. Ann. $28-707 (1962 Repl. Vol.) ................ 68
Ark. Stat. Ann. $41-2205 (1964 Repl. Vol.) ................ 35
Ark. Stat. Ann. $41-3401 (1964 Repl. Vol.) ............... 34
Ark. Stat. Ann. $41-2205 (1964 Repl. Vol.) ..................... 35
Ark. Stat. Ann. $41-3403 (1964 Repl. Vol.) ....3,21,24,27
Ark. Stat. Ann. $41-3405 (1964 Repl. Vol.) . 21
Ark. Stat. Ann. $41-3411 (1964 Repl. Vol.) . 21
Ark. Stat. Ann. $43-2108 (1964 Repl. Vol.) . 28
Ark. Stat. Ann. $43-2153 (1964 Repl. Vol.) ......3,21,27
Ark. Stat. Ann. $43-2155 (1964 Repl. Vol.) . 24
Ark. Stat. Ann. $43-2155 (1964 Repl. Vol.) ................... 24

Cal. Penal Code $190.1 (Supp. 1966) ............................... 78
Conn. Gen. Stat. Rev. $53-10 (Supp. 1965) ............... ...... 78

PAGE

D.C. Code Ann. $22-2801 (1961) ....................................... 21

Fla. Stat. Ann. $794.01 (1964 Cum. Supp.) ................... 21

Ga. Code Ann. $$26-1302, 26-1304 (1963 Cum. Supp.) .... 21

Ky. Rev. Stat. Ann. $435,090 (1963) ............................... 21

La. Rev. Stat. Ann. $14:42 (1950) ................................... 21

Md. Ann. Code $$27-463, 27-12 (1967 Cum. Supp.) ....... 21
Miss. Code Ann. $2358 (Recomp. Vol. 1956) ................... 21

N.C. Gen. Stat. $14-21 (Recomp. Vol. 1953) ................... 21
N.Y. Pen. Law $$125.30,125.35 (Cum. Supp. 1968) ....... 78
Nev. Rev. Stat. $$200,363, 200.400 (1967) ..................... 21

Okla. Stat. Ann. Tit. 21, $$1111, 1114, 1115 (1958) ....... 21

Pa. Stat. Ann., tit. 18, $4701 (1963) ................................. 78

S.C. Code Ann. $$16-72, 16-80 (1962) ............................. 21

Tenn. Code Ann. $$39-3702, 39-3703, 39-3704, 39-3706
(1955) ................................................................................. 21

Tex. Code Crim. Pro., Art. 37.07 (1967) ....................... 78
Tex. Pen. Code Ann., Arts. 1183, 1189 (1961) ..... ......... 21

Vernon’s Mo. Stat. Ann. $559,260 (1953) ....................... 21
Va. Code Ann. $$18.1-44, 18.1-16 (Repl. Vol. 1960) ..... 21

PAGE

Other Authorities:

American Law Institute, Model Penal Code, Tent.
Draft No. 9 (May 8, 1959), Comment to $201.6....... 10,70

American Law Institute, Model Penal Code, $210.6
(P.O.D. May 4, 1962) .........................................62, 64, 74, 78

Bedau, A Social Philosopher Looks at the Death Pen
alty, 123 Am. J. Psychiatry 1361 (1967) ................... 16

Bedau, Death Sentences in New Jersey 1907-1960, 19
Rutgers L. Rev. 1 (1964) ...........................................13,30a

Bedau, The Death Penalty in America (1964) .......15,16, 26

Cardozo, Law and Literature (1931) ............................... 35

DiSalle, Comments on Capital Punishment and Clem
ency, 25 Ohio St. L.J. 71 (1964) ..................................... 13

Duffy & Hirshberg, 88 Men and 2 Women (1962) ......... 12

Florida Division of Corrections, Fifth Biennial Report
(July 1, 1964-June 30, 1966) (1966) ....................... 28a, 32a

Garfinkel, Research Note on Inter- and Intra-Racial
Homicides, 26 Social Forces 369 (1949) ................... 16

Handler, Background Evidence in Murder Cases, 51 J.
Crim. L., Crim . & P ol. Sci., 317, 321-327 (1960) ......... 71

H.L.A. Hart, Murder and the Principles of Punish
ment: England and the United States, 52 Nw. U.L.
Rev. 433, 438-439 (1957) ...................................... .........70-71

Hartung, Trends in the Use of Capital Punishment, 284
Annals 8 (1952) ...............................................................16, 26

House of Commons Select Committee on Capital Pun
ishment, Report (H.M.S.O. 1930), para. 177

PAGE

Institute of Judicial Administration, Disparity in Sen
tencing of Convicted Defendants (1954) ................... 36

Johnson, Selective Factors in Capital Punishment, 36
Social Forces 165 (1957) ............................... 13,16, 27a, 30a

Johnson, The Negro and Crime, 271 Annals 93 (1941) .. 16

Kalven & Zeisel, The American Jury (1966) ...............26, 34,
37, 41, 44, 30a

Knowlton, Problems of Jury Discretion in Capital
Cases, 101 U. Pa. L. Rev. 1099 (1953) .......................26, 71

Lawes, Twenty Thousand Years in Sing Sing (1932) .... 12
Lewis, The Sit-In Cases: Great Expectations, 1963

Supreme Court Review 101, 110 ................................... 47

Mattick, The Unexamined Death (1966)........................... 16

New York State Temporary Commission on Revision
of the Penal Law and Criminal Code, Interim Report
(Leg. Doc. 1963, No. 8) ................................................... 70

Packer, The Limits of the Criminal Sanction (1968)
92-94 ....................................................................... ;............ 57

Packer, Making the Punishment Fit the Crime, 77
Harv. L. Rev. 1071 (1964) ............................................... 65

Partington, The Incidence of the Death Penalty For
Rape in Virginia, 22 Wash. & Lee L. Rev. 43 .......26a, 27a

Pennsylvania, Joint Legislative Committee on Capital
Punishment, Report (1961) ........................................... 16

Perkins, Criminal Law (1957) ........................................... 34
President’s Commission on Law Enforcement and Ad

ministration of Justice, Report (The Challenge of
Crime in a Free Society) (1967) ...............................16, 36

PAGE

Royal Commission on Capital Punishment, 1949-1953,
Report (H.M.S.O. 1953) (Cmd. No. 8932), 6, 12-13,
195, 201-207 .........................................................................

Rubin, Disparity and Equality of Sentences—A Con
stitutional Challenge, 40 F.R.D. (1966) ............... 16, 36, 54

State of California, Department of Justice, Division of
Law Enforcement, Bureau of Criminal Statistics,
Report (Crime and Delinquency in California, 1967)
(1968) ................................................................................ 32a

State of Georgia Board of Corrections, Annual Report
(1965), (1966), (1967) .............................................28a, 32a

State of Maryland, Department of Correction, For
tieth Report (1966) ...................................................28a, 32a

Statement by Attorney General Ramsey Clark, Before
the Subcommittee on Criminal Laws and Procedures
of the Senate Judiciary Committee, on S. 1760, To
Abolish the Death Penalty, July 2, 1968, Department
of Justice Release, p. 2 ................................................... 12

Symposium on Capital Punishment, 7 N.Y.L.F. (1961) 23

Tennessee Department of Correction, Departmental
Report (1966) .............................................................29a, 32a

United Nations, Department of Economic and Social
Affairs, Capital Punishment (ST/SOA/SD/9-10)
(1968) .............................................T................................16,22

United States Department of Justice, Bureau of Pris
ons, National Prisoner Statistics................................... 14

No. 42, Executions 1930-1967 (June, 1968) ....14,15, 24a,
28a, 32a, 33a, 34a

No. 41, Executions 1930-1966 (April, 1967)....... 24a, 27a,
28a, 29a, 31a, 32a, 33a

XIV

No 39, Executions 1930-1965 (June, 1966)....... 24a, 27a,
28a, 31a, 32a, 33a

No. 37, Executions 1930-1964 (April, 1964 [sic:
1965]) ........................... —...... 24a, 27a, 28a, 31a, 32a

No. 34, Executions 1930-1963 (May, 1964) 24a, 27a, 31a

PAGE

No 32, Executions 1962 (April, 1963) ...................13,14,
24a, 27a, 31a

No. 28, Executions 1961 (April, 1962) ....... 24a, 27a, 31a

United States Senate, Sub-Committee on Criminal
Laws and Procedures of the Committee on the Judi
ciary, Hearings on S. 1760, to Abolish the Death
Penalty (Unprinted Report of Proceedings, March
20, 1968) ...........................................................................12’ 13

West, Dr. Louis J., “ A Psychiatrist Looks at the Death
Penalty,” Paper Presented at the 122nd Annual
Meeting of the American Psychiatric Association,
Atlantic City, New Jersey, May 11, 1966, p. 2 ........ . 11

W olf Abstract of Analysis of Jury Sentencing in Capi
tal Cases : New Jersey: 1937-1961,19 Rutgers L. Rev.
56 (1964) ........................................................................... 31a

62 Harv. L. Rev. 77-78 Due Process Requirements of
Definiteness in Statutes (1948) ................................... 49> 50

109 U. Pa. L. Rev. 67, 90 (1960) .............................. 47,59,61

109 U. Pa. L. Rev. 69, 81 (I960) ....................................... 52
69 Yale L.J. 1453, 1459 (1960) ........................................... 55

In t h e

CEourt of tli£ T&mUb States
October T erm, 1968

No. 622

W illiam L. Maxwell,
Petitioner,

—v.—

O.E. B is h o p , Superintendent of
Arkansas State Penitentiary,

Respondent.

ON WRIT OF CERTIORARI TO THE UNITED STATES
COURT OF APPEALS FOR THE EIGHTH CIRCUIT

BRIEF FOR PETITIONER

Opinions Below

The opinion of the United States District Court for the
Eastern District of Arkansas denying petitioner’s applica
tion for a writ of habeas corpus is reported at 257 F. Supp.
710. It appears in the Appendix [hereafter cited A ......] at
20-41. The opinion of the United States Court of Appeals
for the Eighth Circuit affirming the district court is re
ported at 398 F.2d 138, and appears at A. 43-74.

Opinions at earlier stages of this proceeding are re
ported. The opinion of the Supreme Court of Arkansas
affirming petitioner’s conviction for the crime of rape an
sentence of death is found sub nom. Maxwell v. State, 26b
Ark 694 370 S.W.2d 113 (1963). Opinions on disposition
of an earlier application for federal habeas corpus are

found sub nom. Maxwell v. Stephens, 229 F. Supp. 205
(E.D. Ark. 1964), aff’d, 348 F.2d 325 (8th Cir. 1965), cert,
denied, 382 U.S. 944 (1965).

Jurisdiction

The jurisdiction of the district court was conferred by
28 U.S.C. §2241(c)(3) (1964). Jurisdiction of the court of
appeals was conferred by 28 U.S.C. §§1291, 2253 (1964).
The jurisdiction of this Court rests upon 28 U.S.C. §1254
(1) (1964).

The judgment of the court of appeals was entered on
July 11, 1968. (A. 75.) The petition for a writ of certio
rari was filed in this Court on October 9, 1968, and was
granted on December 16, 1968 (A. 76), limited to Ques
tions 2 and 3 of the petition.

Questions Presented

I. Whether Arkansas’ practice of allowing capital trial
juries absolute discretion, uncontrolled by standards or
directions of any kind, to impose the death penalty upon
a defendant convicted of the crime of rape violates the
Due Process Clause of the Fourteenth Amendment?

II. Whether Arkansas’ single-verdict procedure, which
requires the jury to determine guilt and punishment simul
taneously in a capital case, and thus requires a capital
defendant to elect between the exercise of his privilege
against self-incrimination and the presentation of evidence
requisite to rational sentencing choice, violates the Fifth
and Fourteenth Amendments?

Constitutional and Statutory Provisions Involved

The case involves the Fifth and Fourteenth Amendments
to the Constitution of the United States.

It also involves A rkansas Statutes A nnotated, §§41-
3403, 43-2153 (1964 Repl. vol.), which provide, respectively,
as follows:

41-3403. Penalty for Rape. — Any person convicted
of the crime of rape shall suffer the punishment of
death [or life imprisonment],

43-2153. Capital cases — Verdict of life imprison
ment. — The jury shall have the right in all cases
where the punishment is now death by law, to render
a verdict of life imprisonment in the State penitentiary
at hard labor.

Statement

Petitioner, William L. Maxwell, a Negro, was tried in
the Circuit Court of Garland County, Arkansas, in 1962
for the rape of a 35 year old, unmarried white woman. (A.
20.) As we shall describe more fully below, his trial was
conducted pursuant to the ordinary Arkansas procedures
for trying a capital case upon a plea of not guilty. The
issues of guilt and punishment were simultaneously tried
and submitted to the jury, which was given no instruc
tions limiting or directing its absolute discretion, in the
event of conviction, to elect between the punishments of
life or death.1 (A. 28-30, 40-41; see pp. 27-32, 66 n. 69, 35a
infra.)

1 Technically, an Arkansas jury chooses life by returning the
“verdict of life imprisonment” authorized by Ark. Stat. A nn.
§43-2153 (1964 Repl. vol.), text, supra. It chooses death by re
turning a guilty verdict without mention of life imprisonment,
upon which the death sentence is imposed as a matter of course
under Ark. Stat. A nn. §41-3403 (1964 Repl. vol.), text, supra.
(See A. 29, 44.)

Maxwell’s jury elected the punishment of death. The
Supreme Court of Arkansas affirmed. Maxwell v. State,
236 Ark. 694, 370 S.W.2d 113 (1963).2 A 1964 federal
habeas corpus proceeding challenging his rape conviction
and death sentence brought no relief. Maxwell v. Stephens,
229 F. Supp. 205 (E.D. Ark. 1964), aff’d, 348 F.2d 325
(8th Cir. 1965) (one judge dissenting), cert, denied, 382
U.S. 944 (1965).

The present habeas corpus proceeding was commenced
by a second federal petition, filed July 21, 1966. (A. 2-12.)
This petition raised, inter alia, three related constitutional
attacks upon petitioner’s death sentence. First, it com
plained of “ the unfettered discretion of the jury to choose
between the sentence of life or death. The jury’s choice
was . . . unregulated by legal principles of general appli
cation, was left to be determined by arbitrary and dis
criminatory considerations, and was in fact arbitrary and
discriminatory in petitioner’s case.” (A. 8.) Second, the
petition challenged Arkansas’ capital trial practice under
which “ the issues of guilt or innocence and of life or death
sentence [are] . . . determined by a jury simultaneously,
after the jury has heard evidence on both issues in the
same proceeding.” (A. 9.) This single-verdict procedure
(as we shall hereafter call it) not merely empowers, but
virtually compels, arbitrary capital sentencing because it
deprives the sentencing jury of information that is requi
site to rational sentencing choice, since “ evidence perti
nent to the question of penalty could not be presented
without prejudicing the jury against the petitioner on the
issue of guilt. . . . Nor could petitioner exercise his con
stitutional right of allocution before the jury which sen
tenced him, without thereby waiving his privilege against
self-incrimination held applicable to the states under the

2 No review of this decision was sought in this Court.

Fourteenth Amendment. . . . ” (A. 9-10.) Finally, petitioner
alleged that the arbitrary capital sentencing practices
which he attacked had in fact resulted in arbitrary appli
cation of the death penalty by Arkansas juries for the
crime of rape: Negroes convicted of the rape of white
women were discriminatorily sentenced to die on account
of race. (A. 5-7.)

The federal district court allowed a thorough evidentiary
hearing on the racial discrimination claim. (A. 17-18.)
That claim has been excluded from the present phase of
the case by this Court’s limited grant of certiorari (A. 76);
and we need not extend this Statement by describing the
evidence presented at the hearing. However, we shall have
occasion to refer to it in the argument portions of this1
brief, under the principle that where a state practice is
challenged as conferring arbitrary and lawless power tend
ing to abuse, in a manner forbidden by the Fourteenth
Amendment, proof that the power has in fact been regularly
abused is entitled to considerable weight.2 For this reason,
we set forth in Appendix A to the brief, pp. la-23a infra,
a detailed description of petitioner’s evidence in the dis
trict court relating to a thorough and extensive empirical
study of capital sentencing by Arkansas juries in rape
cases during the period 1945-1965, together with the find
ings of the district court and of the court of appeals in
reference to this study. [Appendices to the brief will here
after be cited as App., p......a, infra.]

After hearing, the district court rendered its opinion of
August 26, 1966 (A. 20-41), rejecting petitioner’s conten
tion of discrimination (A. 33-40) and upholding the A r
kansas procedures of unfettered jury discretion in capital

3 Louisiana v. United States, 380 U.S. 145 (1 9 6 5 ); Niemotko v.
Maryland, 340 U.S. 268 (1 951); Hague v. C.I.O., 307 U.S. 496
(1 939); Tick Wo v. Hopkins, 118 U.S. 356 (1886).

sentencing (A. 28-33) and the single-verdict capital trial
(A. 40-41). The court declined to stay petitioner’s execu
tion, set for September 2, 1966 (A. 41) and declined to
issue a certificate of probable cause for appeal. Circuit
Judge Matthes of the Eighth. Circuit similarly refused a
stay or a certificate; but petitioner’s execution was stayed
by Mr. Justice White on September 1, 1966; and this
Court subsequently reversed Judge Matthes’ orders and
remanded to the Court of Appeals for hearing of the
appeal. Maxwell v. Bishop, 385 U.S. 650 (1967).

By its opinion of July 11, 1968 (A. 43-74), the court of
appeals rejected on the merits each of petitioner’s consti
tutional claims (A. 47-64 (racial discrimination), 64-68
(unfettered jury discretion), 68-69 (single-verdict proce
dure)). It accordingly affirmed the judgment of the dis
trict court denying petitioner’s application for habeas
corpus relief. (A. 75.)*

Summary of Argument

All informed observers of the institution of capital
punishment in this country have noted its salient char
acteristic: it is unevenly, arbitrarily and discriminatorily
applied. That observation is strikingly borne out on this
record, which demonstrates that Arkansas juries have per

4 Adjudication on the merits was appropriate. State remedies
were exhausted, within 28 U.S.C. §2254 (Supp. II, 1966), since
no procedures are available in the Arkansas courts by which Peti-
tioner can raise his federal constitutional claims. This was alleged
in petitioner’s habeas application (A. 11), and admitted by respon
dent’s response (A. 14). The district court exercised its discretion
under Sanders v. United States, 373 U.S. 1 (1963), and 28 U.S.C.
§2244(b) (Supp. II, 1966), to entertain on the merits each of
petitioner’s present constitutional contentions, although presented
in a second federal habeas corpus petition; and the propriety of its
doing so cannot be questioned. See Maxwell v. Bishop, 385 U.S.
650 (1967).

sistently discriminated on grounds of race in sentencing
men to die. The court of appeals below admitted there
are “ recognizable indicators” that “ the death penalty for
rape may have been discriminatorily applied over the
decades in that large area of states whose statutes pro
vide for it.” (A. 63.) But whether or not racial discrim
ination was here proved or is provable statistically, it
can hardly be denied that the evidence relating to actual
exercise of jury discretion in capital cases “ casts con
siderable doubt upon the quality of justice in those partic
ular cases throughout the system.” 4 * 6 * Extreme arbitrariness
in the selection of the few men sentenced to death and
executed, out of the great number convicted of capital
offenses each year, is patent; and (as we have pointed
out in greater detail in our amicus curiae brief in a com
panion case)6 the very extremity of this arbitrariness may
effectively conceal the workings of racial discrimination
and of every other invidious prejudice forbidden by the
Constitution. At the very least, the record compels this
Court’s strictest scrutiny under the Fourteenth Amend
ment of the regularity and fairness o f the procedures by
which state courts, through their juries, choose the men
who will die.

Petitioner challenges here the practice of uncontrolled
and undirected jury discretion in capital sentencing that
lies at the root of arbitrary and discriminatory imposition
of the death penalty. This is a practice which, even its
defenders must admit, is arbitrary in a legal sense. It
confers upon a group of twelve jurors, selected ad hoc to

6 The phrase is that of petitioner’s expert witness, an eminent
criminologist, testifying in the district court below. See note 19
infra.

* Boykin v. Alabama, O.T. 1968, No. 642. See note 17 infra.

decide a particular case, a power to determine the question
of life or death that is unlike any other power possessed
by a jury, or even by a court, in our legal system. The
life-or-death decision is made utterly without standards
or governing legal principles; it is made without the
limitation of requisite factual findings, or of required
attention to any range or realm of fact, or of required
consideration of any general rule or policy of law; it is
made without any judicial control over the process or
the consequence of the jurors’ determination. The con
scientious and fairminded juror is given not the slightest
idea what he is to do, while the covert discriminator is
given absolute license to practice his biases in the matter
of taking human life. This unfettered jury discretion-
or, rather, naked and arbitrary power, lacking every at
tribute of legal discretion—can be likened only to the
unimaginable procedure of submitting to a jury in a civi
case the unadorned question whether the defendant ought
to be liable to the plaintiff; or, in a criminal case, whether
the defendant has done something for which he should be
punished. Such submissions are not made anywhere m
American law -except in the enormous decision whether
men shall live or die. They violate the rule of law that
is basic to Due Process, and especially critical m regard
to the choice of life or death.

Unfettered jury discretion in capital sentencing exhibits
those vices that have repeatedly been condemned by this
Court under the constitutional principles forbidding in
definiteness in penal legislation. First, a capital defendant
with his life at stake, does not fairly know how to conduct
his defense. He does not know—to take one example
whether the jurors will regard mental disorder as a
mitigating circumstance or an aggravating one; or whether
five jurors will think the circumstance mitigating while
seven vote to kill him for it. As a result, the capital trial

is a grisly game of blind-man’s buff, played for the
defendant’s life. Second, and more important, the con
ferring on the jury of “ a naked and arbitrary power”
(Tick Wo v. Hopkins, 118 U.S. 356, 366 (1886)) to take a
man’s life for any reason or for no reason offends the
principle of legality, of regularity—the principle requiring
a rule of law and not of men—which the Due Process
Clause asserts as a protection against laws that would
otherwise he “ susceptible of sweeping and improper ap
plication” (N.A.A.C.P. v. Button, 371 U.S. 415, 433 (1963),
“ lest unwittingly or otherwise invidious discriminations
are made against groups or types of individuals in viola
tion of the constitutional guarantee of just and equal
laws” (Skitmer v. Oklahoma, 316 U.S. 535, 541 (1942)).
Due Process of Law fundamentally repudiates any such
power, which “ leaves . . . jurors free to decide, without
any legally fixed standards” (Giaccio v. Pennsylvania, 382
U.S. 399 (1966)), whether human life shall or shall not
be forfeit, even as a punishment for crime.

U.
The vices of unfettered jury discretion are compounded

when the jury’s life-or-death decision is made under a
single-verdict procedure. Whereas unfettered discretion
allows the jury arbitrary power, the single-verdict trial
virtually requires that that power be exercised arbitrarily.
This is so because information that is absolutely requisite
to rational sentencing choice cannot be presented to the
jury except at the cost o f an unfair trial on the issue of
guilt or innocence, and of enforced waiver of the defen
dant’s privilege against self-incrimination.

The single-verdict capital trial is federally unconstitu
tional for two reasons. First, it unnecessarily compels
a choice between the defendant’s Fifth and Fourteenth

Amendment privilege against self-incrimination (Malloy
v. Hogan, 378 U.S. 1 (1964)) and his Fourteenth Amend
ment right “ to be heard . . . and to offer evidence of his1
own” (Specht v. Patterson, 386 U.S. 605, 610 (1967))
on the vital question of capital sentencing. As a result,
it unconstitutionally burdens the Privilege (United States
v. Jackson, 390 U.S. 570 (1968); Simmons v. United States,
390 U.S. 377 (1968)) by attaching to its exercise the in
dependently unconstitutional consequence that the capital
sentencing decision is made irrationally (Skirmer v. Okla
homa, 316 U.S. 535 (1942)), because made “upon less than
all of the relevant evidence” (Jackson v. Denno, 378 U.S.
368, 389 n. 16 (1964)). Second, this procedure that forces
the capital defendant to a “choice between a method which
threatens the fairness of the trial of guilt or innocence
and one which detracts from the rationality of the deter
mination of sentence” 7 presents a “ grisly, hard, Hobson’s
choice” 8 * which is fundamentally unfair, in the Due Process
sense, where the wages of the gamble are death.

• • •

Logical presentation requires that our arguments relat
ing to unfettered jury discretion and to the single-verdict
procedure be stated separately. Either argument alone is,
in our view, sufficient to vitiate William L. Maxwell’s sen
tence of death under the Fourteenth Amendment. How
ever, it must be remembered that both of the challenged
procedures were employed at Maxwell’s trial. Their vices,
as we have pointed out, are mutually compounding. Used
together, they deprive Maxwell of his life after a trial ut
terly lacking in the rudimentary fairness and regularity

7 A merican Law Institute, Model P enal Code, Tent. Draft
No. 9 (May 8, 1959), Comment to §201.6, at 64.

‘ Whitus v. Balkcom, 333 F.2d 496, 499 (5th Cir. 1964). Cf.
Pay v. Noia, 372 U.S. 391, 440 (1963).

that Due Process assuredly demands when a state em
powers its jurors “ to answer ‘yes’ or ‘no’ to the question
whether this defendant was fit to live” ( Witherspoon v.
Illinois, 391 U.S. 510, 521 n. 20 (1968)).

ARGUMENT

Introduction: The Fact and Implications of
Arbitrary Capital Sentencing

Knowledgeable observers of the administration of capital
punishment in the United States agree that death is meted
out among persons convicted of capital crimes in a fashion
that is uneven, rationally unsupportable and arbitrary in
the extreme.

“ Of all the uncertain manifestations of justice, capi
tal punishment is the most inequitable. It is primarily
carried out against the destitute, forlorn and forgotten.
. . . Members of racial and cultural minority groups
suffer most. The hundreds of extraneous factors, in
cluding geography, that decide whether a convicted
man will actually live or die, makes capital punishment
a ghastly, brainless lottery.” (Dr. Louis J. West, “ A
Psychiatrist Looks at the Death Penalty,” Paper Pre
sented at the 122nd Annual Meeting of the American
Psychiatric Association, Atlantic City, New Jersey,
May 11,1966, p. 2.)

Arbitrariness in the selection of men to be put to death
takes several forms. First, there is simply the matter of
baseless discrimination among individuals: freakish, whim
sical, erratic difference in the treatment of similar men
guilty o f similar offenses. As the Attorney General o f the
United States put it recently:

“A small and capricious selection of offenders have
been put to death. Most persons convicted of the same
crimes have been imprisoned. Experienced wardens
know many prisoners serving life or less whose crimes
were equally, or more atrocious, than those of men on
death row.” (Statement by Attorney General Ramsey
Clark, Before the Subcommittee on Criminal Laws and
Procedures of the Senate Judiciary Committee, on
S. 1760, To Abolish the Death Penalty, July 2, 1968,
Department of Justice Release, p. 2.)

Those who should surely know—the respected long-time
wardens of Sing-Sing and San Quentin prisons—corrobo
rate the Attorney General. L awes, T wenty T housand
Y ears in S ing S ing (1932) 302, 307-310; D uffy & H irsh-
beeg, 88 Men and 2 W omen (1962) 254-255.®

Second, there is economic and caste discrimination.

“ [T]he death penalty . . . almost always hits the
little man, who is not only poor in material possessions
but in background, education, and mental capacity as
well. Father Daniel McAlister, former Catholic chap
lain at San Quentin, points out that ‘the death penalty
seems to be meant for the poor, uneducated, and legally
impotent offender.’ ” (D uffy & H ieshberg, op. cit.
supra, 256-257.)

8 See also the testimony of Clinton T. Duffy, in United States
Senate, Subcommittee on Criminal Laws and Procedures of the
Committee on the Judiciary, Hearings on S. 1760, To Abolish the
Death Penalty (unprinted report of proceedings, March 20, 1968),
vol. 1, pp. 44 -44A : “ I have often said, and I repeat here, that I
can take you into San Quentin Prison or to Sing Sing, Leaven
worth or Atlanta Prisons and I can pick out many prisoners in
each institution serving life sentences or less, and can prove that
their crimes were just as atrocious, and sometimes more so, than
most of those men on the row.”

Former Ohio Governor Michael DiSalle has made the same
point: “ I want to emphasize that from my own personal
experience those who were sentenced to death and appeared
before me for clemency were mostly people who were with
out funds for a full and adequate defense, friendless, un
educated, and with mentalities that bordered on being de
fective.” (DiSalle, Comments on Capital Punishment and
Clemency, 25 Ohio State L. J. 71, 72 (1964).)10

Third, there is persuasive evidence of that most corrosive
and invidious form of discrimination, racial prejudice, in
the selection of the men who will die. The Federal Bureau
of Prisons maintains reliable statistics on executions in the
United States since 1930. Between that year and 1962, the
year in which petitioner Maxwell was sentenced to die, 446
persons were executed for rape in this country.11 Of these,

10 See also the testimony of Michael DiSalle, in Hearings, note 9
supra, vol. 1, pp. 14-16. The Governor’s observations are supported
by those of scholars who have undertaken to describe the charac
teristics of men on death row in other states. E.g., Bedau, Death
Sentences in New Jersey 1907-1960, 19 R utgers L. Rev. 1 (1964) ;
Johnson, Selective Factors in Capital Punishment, 36 Social
F orces 165 (1957). And see the study of Florida’s death row
population described in the Brief for the N.A.A.C.P. Legal Defense
and Educational Fund, Inc., and the National Office for the Rights
of the Indigent, as Amici Curiae, in Boykin v. Alabama, O.T. 1968,
No. 642, at p. 7 n. 8.

11 The figures below are taken from United States D epartment
of Justice, B ureau of Prisons, National P risoner Statistics,
No. 32, Executions 1962 (April, 1963), which was put in evidence
as petitioner’s Exhibit P-6 at the habeas corpus hearing below.
Table 1 thereof shows the following numbers and percentages of
executions under civil authority in the United States between 1930
and 1962:

Negro
Rape 399 (89.57c)
Murder 1619 (49.1% )
Other 31 (45.6% )
Total 2049 (53.77c)

White Other Total
45 (10.17c)

1640 (49.77c)
37 (54.4% )

1722 (45.2% )

2 (0.47c) 446 (1007c)
39 (1.2%.) 3298 (1007c)
_0 (0.07o) 68 (1007c)
41 (1.170 3812 (1007c)

(Continued on p. 14)

399 were Negroes, 45 were whites, and 2 were Indians. All
were executed in Southern or border States or in the Dis
trict of Columbia. The States of Louisiana, Mississippi,
Oklahoma, Virginia, West Virginia and the District never
executed a white man for rape during these years. Together
they executed 66 Negroes. Arkansas, Delaware, Florida,

(Continued from p. 13)
Table 2 thereof shows the following numbers of executions under
civil authority in the United States between 1930 and 1962, for the
offense of rape, by race and state:

Jurisdiction Negro White Other Total
Federal ............................... 0 2 0 2
Alabama ............................. .... 20 2 0 22
Arkansas .......................... .... 17 1 0 18
Delaware ............................. 3 1 0 4
District of Columbia .... 2 0 0 2
Florida ............................... .... 35 1 0 36
Georgia ............................... .... 58 3 0 61
Kentucky ........................... 9 1 0 10
Louisiana .......................... .... 17 0 0 17
Maryland .......................... .... 18 6 0 24
Mississippi ........................ .... 21 0 0 21
Missouri ............................. 7 1 0 8
North Carolina ............... .... 41 4 2 47
Oklahoma .......................... 4 0 0 4
South Carolina ............... .... 37 5 0 42
Tennessee ........................... .... 22 5 0 27
Texas .................................... .... 66 13 0 79
Virginia ............................. .... 21 0 0 21
West Virginia ................. 1 0 0 1

Total ................................... .... 399 45 2 446

The figures have not changed appreciably since 1963. According
to the latest National Prisoner Statistics Bulletin, United States
Department of J ustice, B ureau of Prisons, National Prisoner
Statistics, N o. 42, Executions 1930-1967 (June, 1968), p. 7, table
1, the following are the numbers and percentages of executions
under civil authority in the United States between 1930 and 1967:

Negro White Other Total
Rape 405 (89 .0% ) 4 8 (1 0 .6 % ) 2 ( 0 .4 % ) 455 (100% )
Murder 1630 (48 .9% ) 1664 (49.9% ) 4 0 (1 .2 % ) 3334 (100% )
Other 3 1 (4 4 .3 % ) 39 (55.7% ) 0 (0 .0 % ) 70 (100% )
Total 2066 (53 .1% ) 1751 (45.4% ) 4 2 (1 .1 % ) 3859 (100% )

(Continued on p. 15)

Kentucky and Missouri each executed one white man for
rape between 1930 and 1962. Together they executed 71
Negroes. Putting aside Texas (which executed 13 whites
and 66 Negroes), sixteen Southern and border States and
the District of Columbia between 1930 and 1962 executed
30 whites and 333 Negroes for rape: a ratio of better than
one to eleven. The nationwide ratio of executions for the
crime of murder was considerably less startling—one Negro
for each one white—but still startling enough, since Negroes
constituted about one-tenth of the Nation’s population dur
ing these years.

Of course, these suspicious figures might be explained, not
by arbitrary and discriminatory administration of the death
penalty, but by some rather extravagant hypotheses about
the Negro crime rate.12 Responsible analysts have rejected
Buch an explanation. With virtual unanimity, commissions
and individuals studying capital punishment have found

(Continued from p. 14)
The following is the breakdown of the 435 men reported under
sentence of death in the country as of December 31, 1967 {id., pp.
22-23, table 1 0 ):

Negro
Nine Northeastern States ........... 33
Twelve North-Central States .... 24
Thirteen Western States ............. 21
Sixteen Southern States ............. 159
Federal ............... :.............................. 1

Total ................................................... 238

White Other Total
29 0 62
31 0 55
68 2 91
66 0 225

1 0 2

195 2 435

12 In fact, the number of crimes committed by Negroes appears
to be three to six times higher than that which the ratio of Negroes
in the population would lead one to expect. See Bedau, T he D eath
Penalty in A merica (1964) 412. Negroes constitute one-tenth or
one-ninth of the population (depending upon the time periods
under consideration). So, instead of the expectable one Negro-
committed crime to every nine white-committed crimes, there are
three to six Negro crimes to every nine white crimes. Far more
crimes numerically are obviously committed by whites than by
Negroes. Yet one Negro murder convict is executed for every
white murder convict executed; and nine Negro rape convicts are
executed for every white rape convict executed. See text, supra.

evidence that the imposition of the death sentence and the
exemse of dispensing power by the courts and the execu
tive follow discriminatory patterns. The death sentence is
disproportionately imposed and carried out on the poor,

e egro, and the members of unpopular groups.” P resi-
dent s Commission on L aw E nforcement and A dminister,
tion oe justice, R epoet (T he Challenoe op Ceime in a
F ree Society) (1967) 143. See also United Nations, De-

aetment of E conomic and Social A ffairs, Capital P un

J oint 1̂ (ST/SO A^ SD/ 9-10) <1968) 32, 98; Pennsylvania,
J oint L egislative Committee on Capital P unishment B e-

mee Ur ° K’
(1964 THE DEATH Penaltt ™ A merica
n /t p t 5 Bedau’ A S°cial Phil0^pher Looks at the

ea Penalty, 123 A m . J. P sychiatry 1361, 1362 (1967) •

t in t* !'n .T ntyr t EqUaUty °f 8entences-A Constitu-
tional Challenge, 40 F.B.D. 55, 66-68 (1966); Johnson, Selec

1957^H0rV n ° T tal Punishment> 36 Social F orces 165
2S4 A J HartU0Df Prends in th° Use of Capital Punishment,
284 A nnals 8, 14-17 (1952); Garfinkel, Research Note on
(19491 ° WT k ra'R™ al Hom™ides, 26 Social F orces 369
(1941) : J0hI1S0n, The Negr° and Crime> 271 A nnals 93

In order to provide a more systematic and rigorous ex
amination of the evidence of racial differentials in capital
sentencing, an extensive empirical study of sentencing pat-
terns in rape cases was undertaken in 1965 by Dr. Marvin
E. Wolfgang, an eminent criminologist, at the request of
counsel for petitioner Maxwell (who also represenTnumer-
ous condemned men in other States). Dr. Wolfgang’s study
covered every case of conviction for rape in 250 counties in
e even States during the twenty-year period 1945-1965. The
data gathered in Arkansas, and Dr. Wolfgang’s analysis of
that data, were the subject of his testimony at the habeas
corpus hearing below. The testimony and the findings of

the lower courts relating to it are described in detail in
Appendix A, pp. la-23a infra. We summarize them briefly
here because of the importance of Dr. Wolfgang’s conclu
sions, which confirm the earlier impressions of racial dis
crimination on the basis of the first fully controlled, exact
ing scientific study of the subject.

Dr. Wolfgang’s study began with the collection of data
concerning every case of conviction for the crime of rape
on the docket books of nineteen randomly selected Arkansas
counties, containing 47% of the State’s total population,
for the twenty years 1945-1965. The nineteen counties were
selected by accepted areal sampling methods with the goal
of producing a sample that would be representative of the
State of Arkansas as a whole; and, in the opinion of the
expert statistician whom Dr. Wolfgang employed to per
form the sampling operation, “ inferences drawn from this
sample . . . are valid for the State of Arkansas.” One point
should be stressed. The study, from the outset, concerns
cases of conviction for the capital crime of rape, and what
is studied is the performance of Arkansas juries in select
ing the convicted defendants upon whom they impose the
death penalty. It thus controls completely the possibility,
suggested above, that high frequencies observed in the
sentencing of Negroes to die for the crime of rape might he
explained by the supposition that Negroes commit rape, or
are convicted of rape, more frequently than whites. This
study compares the rate of death sentencing for Negro and
white defendants all of whom have been convicted of rape.

Field researchers dispatched to Arkansas conducted an
exhaustive investigation of each case where a rape convic
tion had been had in the sample counties. They followed a
predetermined pattern for exploring the available sources
of information about each case, beginning with court rec
ords, trial transcripts, witness blotters, file jackets, judicial
opinions, etc., then proceeding to prison and pardon board

records, and finally to newspaper files and interviews with
trial counsel. They had uniform procedures for assigning
priorities to information sources in the event of conflicts;
and they used a uniform schedule, with objectively defined
categories, for recording the data found. At the hearing
below, the State of Arkansas conceded the validity of all of
the data thus gathered and recorded.

The “critical” data for each case were race of defendant,
race of victim, and sentence. Dr. Wolfgang analyzed these
variables and found conclusively that Negro defendants
convicted of the rape of white women were disproportion
ately frequently sentenced to death. Applying tests of
statistical significance that are generally used in the social
sciences (and in other disciplines, such as medical research,
as well), he found that the disproportionate frequency with
which the death sentence was imposed on these Negro de
fendants was so great that there was a less than two per
cent probability of its having occurred by chance. Put
another way, if race were not really related to the capital
sentencing patterns of Arkansas juries, the results observed
in the twenty years between 1945 and 1965 could have oc
curred fortuitously in fewer than two twenty-year periods
since the birth o f Christ. Not surprisingly, the district court
agreed with Dr. Wolfgang in finding that the markedly
over-frequent sentencing to death of Negroes convicted of
rape of white women “could not be due to the operation of
the laws of chance.”

Dr. Wolfgang next proceeded to determine whether any
other ascertainable circumstance in these rape cases could
account for the differential sentencing. The data gathered
by the researchers included not merely race and sentence,
but 28 pages of information about each case: characteris
tics of the defendant (age, family status; occupation; prior
criminal record; etc.) and of the victim (age; family status;

occupation; husband’s occupation if married; reputation
for chastity); nature of the defendant-victim relationship
(prior acquaintance; prior sexual relations, manner
which defendant and victim arrived at the scene of the of
fense); circumstances of the offense (number of offenders
and victims; place; degree of violence or threat employed;
degree of injury received by victim; housebreaking or con
temporaneous offenses committed by defendant; presence
of members of the victim’s family and threats or violence
employed against them; nature of intercourse; involvement
of alcohol; etc.); and circumstances of the trial (plea,
presentation of defenses of insanity or consent; joinder for
trial of other charges against the defendant or co-defen
dants; whether defendant testified; nature of his legal rep
resentation (retained or appointed); etc.). Every one o f
these variables for which sufficient information could be
gathered from the official records and other sources studied
was analyzed with a view to determining whether it might
explain or account for the phenomenon of racially differen
tial sentencing. Dr. Wolfgang concluded that no non-racial
variable of which analysis was possible could account for
the differential observed. His ultimate conclusion was that
Negro defendants who rape white victims have been dis
proportionately sentenced to death, by reason of race, dur
ing the years 1945-1965 in the State of Arkansas.

The district court disagreed with this ultimate conclu
sion but for reasons that the court of appeals appears to
have thought unpersuasive and which will hardly
scrutiny of the record. See Appendix A, PP. 19a-23a infra.
The court of appeals itself rejected petitioners legal con
tention of racial discrimination, for doctrinal reasons that
are not now relevant; but it obviously thought that Dr.
Wolfeang’s factual finding of discrimination was not re
buttable. It expressly found that “ [tjhere are recognizable

indicators . . . that the death penalty for rape may have
been discriminatorily applied over the decades in that large
area of states whose statutes provide for it.”

We have set forth this evidence of arbitrary and dis
criminatory capital sentencing at the outset of our argu
ment for three reasons. First, our specific constitutional
attacks upon the Arkansas death-sentencing procedures by
which petitioner Maxwell was condemned are, in essence:
(1) that the unfettered discretion given Arkansas juries
to select between the penalties of life and death, without
the guidance of standards or control by legal principles of
any sort, allows wholly arbitrary deprivation of human
life, in violation of Due Process, and (2) that Arkansas’
single-verdict practice in capital trials in effect compels the
arbitrary exercise of this arbitrary power because it de
prives the defendant who exercises his privilege against
self-incrimination of the opportunity to present to the
sentencing jury information that is the indispensable pre
requisite of rational sentencing choice. As this Court’s
prior decisions in several differing sorts of cases make
clear, evidence that abuse has in fact occurred has a con
siderable bearing on the issue whether a practice chal
lenged on the ground of lawlessness tending to abuse is sus
ceptible to that challenge. See cases cited in note 3 supra.

Second, there is obviously the most intimate sort of re
lationship between laws maintaining the death penalty,
procedures which allow its imposition arbitrarily, and racial
and caste discrimination in its actual administration. In the
Brief for the N.A.A.C.P. Legal Defense and Educational
Fund, Inc., and the National Office for the Rights of the
Indigent, as Amici Curiae, in the companion case of Boykin
v. Alabama, O.T. 1968, No. 642, we have analyzed one as
pect of that relationship: the point that the “public can
easily bear the rare, random occurrence of a punishment

which, if applied regularly, would make the common gorge
rise.” (Id., at 55.) “ A legislator may not scruple to put a
law on the books whose general, even-handed, non-arbitrary
application the public would abhor-precisely because both
he and the public know that it will not be enforced generally,
even-handedly, non-arbitrarily.” (Id., at 39; see generally
id., at 35-61.)

This is most obviously the case with regard to the death
penalty for rape. Only sixteen American jurisdictions re
tain capital punishment for that offense. Nevada permits
imposition of the penalty exclusively in cases where rape is
committed with substantial bodily harm to the victim. ̂
The remaining fifteen jurisdictions—which allow their
juries absolute discretion to punish any rape with d ea th -
are all Southern or border States.13 14 * The federal jurisdiction
and the District of Columbia also allow the death penalty
for rape in the jury’s unfettered discretion.16 We think the

13 Nev Rev. Stat. §200.363 (1967). See also §200.400 (1967)
(assault with intent to rape, accompanied with acts of violenc
resulting in substantial bodily harm).

14 The following sections punish rape or carnal knowledge unless
otherwise specified. Ala. Code §§14-395, 14-397 14-398 Becomp.
Vol. 1958); Ark. Stat. Ann. §§41-3403, 43-2153 (1964 Repl. Vo . ) ,
see also §41-3405 (administering potion with n 964 Cum
&41-3411 (forcing marriage); Fla. Stat. Ann. §794.01 ( l yo
S »p p 5 G . Cod, Ann. §§26-1302, 26-1304 (196 C u m -S u p P .h
K y Rev. Stat. Ann. §435.090 (1963); La. Rev. Stat. Ann §14.42
(1950) (called aggravated rape but slight force is su c
constitute offense; also includes carnal knowledf e) = ^ th
Code §27-463 (1967 Cum. Supp.) ; see also §27-12 (assault witn
intent to rape); Miss. Code Ann. §2358 (Recomp. Vol 1956),
Vernon s Mo" Stat. Ann. §559.260 (1953) ; N O. Gem Stat §14-21
("Recomn Vol 1953) ; Okla. Stat. Ann. Tit. 21, §§1111, 1114, l i t

1958); S.C. Code Ann. §§16-72, 16-80 (1962) (includes assault
' • Q(.+OTTirit tn rane as well as rape and carnal knowledge) , lenn.
S d , A n” W 9 4 7 0 2 39 3703, 39P3704, 39-3706 (1 9 5 5 ,; Tea r « .
Code A nn, arte. 1183,1189 (1 961); Va. Code Ann. §18.1-44 (Repl.
Yol 1960); see also §18.1-16 (attempted rapel>-

is i s u .s .c . §2031 (1964); 10 U.S.C. §920 (1964); D.C. Code
Ann. §22-280i (1961).

relationship is obvious between this map of the legal inci
dence of capital punishment for rape and the discriminatory
exercise of juries’ discretion in the actual imposition of
death sentences. It is also worth noting that, outside the
United States, rape is punishable by death only in Malawi,
Taiwan, and the Union of South Africa.1'

The mediating links between the allowability on the
statute books of the death penalty for a crime and its actual
use against the few, arbitrarily selected outcasts yearly
chosen to die are provided by the death-sentencing proce
dures challenged in the present case. It is these procedures
by which laws of apparently uniform application are con
verted in practice into instruments of the most vicious dis
crimination. Their rare, arbitrary and discriminatory use
against the poor and the disfavored insulates the laws, in
turn, against fair public scrutiny and reprobation. At the
same time that a capricious, ad hoc selection of the men to
be killed makes sentencing patterns virtually immune
against judicial control under the Equal Protection Clause,16 17 *
the indefinite and arbitrary character of the sentencing
procedures themselves effectively precludes constitutional
control of particular death sentences rendered by individual

16 United Nations, Department of E conomic and Social A f
fairs, Capital P unishment (S T /S O A /S D /9 -1 0 ) (1968), pp. 40,
86.

17 W e make this point at greater length in the Brief for the
N .A A .C .P . Legal Defense and Educational Fund, Inc., and the
National Office for the Rights of the Indigent, as Amici Curiae,
in Boykin v. Alabama, O.T. 1968, No. 642, at 53-55. The opinion
of the district court below presents an obvious instance of judicial
inability to detect racial discrimination where it is concealed under
the additional veil of ad hoc arbitrariness. See A. 37-40. “ Ob
viously, a State can discriminate racially and not get caught at it
if it kills men only sporadically, not too often. By being arbitrary
in selecting the victims of discrimination, a State can get away
with both arbitrariness and discrimination.” Boykin Brief, supra,
at 54.

. . . . . H the arbitrariness and discrimination infecting

^ administration of the
^ r ^ ^ C - ^ a r i n e s s a n d

discrimination. . . fi
, . Q to our third point relating to the signifi- Thm brings ns to our t h P discriminstory capi-

CT °tf n c i n r t l e r f w e l i e r has observed that: '-most
tal sentencing. nerDe alitv is a most im-
dramatically when h e m a ^ m Cofli(oi P m is h .

portent ebsmen i ' ^ The Fourteenth Amend-
merit, 7 Jn.x .Jj. _ > aoK|D Plpment 0f justice under the

ConstHutkm^ aneTthe^Due

upon the individual.
What is in issue in this case is the

regularity and to take human
in proceedings by w 1 procedures
fife. Where the consequences o the use of those P
are marked by what the court of appeals belo ^ ^

edged were applied
penalty for ™pe ” 7 , area of states whose statutes
o « r the decades in a„ extensive, painstaking

S £ o ' 1 — study of the application of those pro-

i!The opinion of the court of appeals below
eaJord in L ily well I . rests part of the
the court cannot de.te,ct ned petitioner to die. See A- 59-64.
particular jury winch condemn P bp impossible be-
Any such work of detection woMd o standards governing
cause the entire absence ^ Arkansas al of the propriety
the jury’s sentencing dec* on preclu P^mination which might
S t S t r : T o t S ^ i c a b l e act of impropriety.

cedures “ casts considerable doubt upon the quality of
justice in those particular cases throughout the system” 19—
surely the procedures which allow these uses and conse
quences call for the most critical and searching scrutiny of
which courts are capable, to assure consistency with Due
Process. Such scrutiny, as we shall now show, finds Arkan
sas’ capital sentencing procedures drastically deficient.

Arkansas’ Practice of Allowing Capital Trial Juries
Absolute and Arbitrary Power to Elect Between the
Penalties o f Life or Death for the Crime of Rape Vio
lates the Rule o f Law Basic to the Due Process Clause.

Reading the formal provisions o f Arkansas statutory
law governing punishment for the crime o f rape, it is
easy to be lulled into a quite misleading frame o f mind.
The statutes say, in effect, that the penalty for rape is
death, except that the jury may instead elect to sentence
a defendant convicted o f rape to life imprisonment. A r k .

S tat. A n n . $§41-3403, 43-2153 (1964 Repl. vol.), p. 3 supra.
The image conveyed is that death is the ordinary and
usual consequence o f a rape conviction, while the sentence
o f life imprisonment is some form o f extraordinary dis
pensation from the true course o f the law.

This image is as dangerous as it is wrong. Its danger
lies not primarily in the sort o f simplistic legal reasoning
which has sometimes been supposed to be applicable to it:
that a dispensing procedure which grants a gratuitous
benefit, rather than imposing a burden, escapes the con-

19 Dr. Marvin E . W olfgang, testifying on cross examination be
low (Tr. 8 1 ) , quoted by the court of appeals at A . 53. Compare
the phraseology of the court of appeals relative to “ suspicion . . .
with regard to southern interracial rape trials as a group over a
long period of time.” (A . 61-62.)

trol of constitutional safeguards designed to protect the
individual from arbitrary and overreaching state actio .
W e do not suppose that this Court would for a momen
countenance any such legal argument." The more insidious
danger of the image is a subtle attitude which it engenders
that the process of decision-making by which ^ ly
is selected as the penalty for crime is not really 7
l l o v Z t The defendant, after all, has been convicted

offense whose punishment is death-although some
few defendants may be exempted from the^ctual neces^ y
of dying This attitude fosters a view of the procedur
tor M e e t in g the men who will live and the men who
t i l l die, from among the total number o f men convicte
^ capital offenses, that is both unreal and irresponsible.

W e hope that there can be no doubt about the facts.
The penalty for rape is not d ea th -in Arkansas or any
where else.7 Only one quarter of the total n u m berof A r
kansas rape convictions analyzed by Dr. W olfgang
Suited in a death sentence.91 The twenty-five per e - t
figure is probably somewhat high even for Arkansas,

20 The Fourteenth Amendment s 9 ?, ^ & state is federally
equality apply not merely to sue ^ processes of dis-
compelled to give its C1̂ lzen®’ choose to give them, however
pensing such benefits as the ^ ate “ y ({on 347 u .S . 483 (1954) ;
gratuitously. Broum v. B o a r d / g 526 (1 9 6 3 ); Cox v. Louisiana,
Watson v. City of M em p h is373 U.S. applied to
379 U S 536, 555-558 (1965). The P ™ P ie TT a 162 (1963 ,
criminal sentencing, ^
and to capital sentencing in ^ i c r d a r \ rkansas-
U.S. 83 (1963). So, even i f t h e the conferring
capital sentencing procedures defendant those procedures
of the benefit of life to a «on^ uf J ^ c S T ^ d Equal Protec-
are nonetheless constrained by the Due Process
tion Clauses, as Brady squarely holds. were

21 See Petitioner’s Exhibit P -4 A p p e n d s B teble ^
fourteen death sentences imposed in a footnote are

22 The fifty-five cases ?ew case! found on the
all of those analyzed by Dr. woiigaug

and it appears far higher than the percentage of rape
convicts who are sentenced to death in other states where
the offense is potentially capital.23 It is also true that
the penalty for first-degree murder is not death—in A r
kansas or anywhere in the United States.24 By far the
greater number of first-degree murder convicts, like rape
convicts, are sentenced to some punishment other than
death.25 The testimony of Attorney General Clark, quoted
at p. 12 supra, was neither heedless nor uninformed:
“ Most persons convicted of the same crimes [for which
“a small and capricious selection of offenders have been
put to death” ] have been imprisoned.”

What is important here is not the respective percentages
of men sentenced to life and to death (we shall recur to
their significance shortly), but rather the point that a
highly selective process of making individuating judgments
is occurring, called forth by a state’s statutes which give
its juries the option between a death sentence and some
thing less. This process begins at the point of a defen-

docket books could not be analyzed, because information relating
to the critical variables was not discoverable. These were ordi
narily non-death cases, since official record-keeping in death cases
tends to be more fulsome.

23 See Appendix B, pp. 24a-34a infra.
24 The death penalty for first-degree murder is no longer manda

tory anywhere in the United States. See B edau, The D eath
! Penalty in A merica (1964) 27-30, 45-52.
j Indeed, there are very few crimes in the United States today
j that carry a mandatory death penalty, and those few are for the
j most part of the obscure sort under which no one is ever charged
i (treason, in several states; perjury in a capital case, etc.). See
I ibid. And see Ka lven & Zeisel, T he A merican J ury (1966) 301,

435; Hartung, Trends in the Use of Capital Punishment, 284
A nnals 8 (1952); Knowlton, Problems of Jury Discretion in
Capital Cases, 101 U. Pa . L. Rev. 1099 (1953).

25 See Appendix B, pp. 24a-34a infra.

dant’ s conviction for a capital crim e; it applies to all
defendants so convicted; and it involves the m aking of
differentiations between them, choosing those ones am ong
their total number whose lives are to be taken.

The question is not w orth debating whether the F o u r
teenth Am endm ent’s basic requirem ents o f regularity ,
fundam ental fairness, and even-handedness operate as

constraints upon such a process of ° T i T s V
they do. W ith ersp oon v. Illinois, 391 U.S. 510 (19 ) .
B ra d y v. M aryland, 373 U.S. 83 (1963). And see Spech t
v. P a tterson , 386 U.S. 605 (1967); M em p a v. R h a y , ^
U.S. 128 (1967); Skinner v. Oklahoma, 316 U.S. 535
As Judge Sobeloff has written in another connection:

“ Under our constitutional system, the payment
which society exacts for transgression of the law does
not include relegating the transgressor to arbitrary
and capricious action.” (Landm an v. P eyto n , 370 F.2d
135, 141 (4th Cir. 1966).)

The issue is whether the selection process used by the
State o f A r k a n s a s -a n d by m ost other A m erican states
which retain capital punishm ent, we m ust a d d -c o m p o r ts
with the relevant Fourteenth A m endm ent constraints o
“ relegates the transgressor to arbitrary and capricious
action” in the manner o f deciding whether he lives or dies.

We therefore pass to an examination of the Arkansas

process.

A. The Power Given Arkansas Juries Is Essentially Lawless.

1 A rk. Stat. A nn . $41-3403 (1964 RepL vol.), P- 3
supra, fixes the punishment for rape as death. Pursuant
to A rk. Stat. A n n . $43-2153 (1964 Repl. vol.), p. 3 su pra .

“The jury shall have the right in all cases where
the punishment is now death by law, to render a
verdict of life imprisonment. . . 21 22 * * * *

This option conferred upon the jury is the exclusive
method of selecting, among convicted capital offenders,
the ones who die. A jury trial may not be waived in a
capital case in Arkansas. A rk . Stat. A nn . $43-2108
(1964 Repl. v o l.) ; Scarber v. State, 226 Ark. 503, 506,
291 S.W.2d 241, 242-243 (1956).

2. The only procedural requisite for valid exercise
of the option is that the jury must be told it has a choice
between life imprisonment and death.27 It is apparently
sufficient, to meet this requirement, that the court hand
the jury alternative life and death verdict forms, identify
them, and tell the jury it may use either. Smith v. State,
230 Ark. 634, 642-643, 324 S.W.2d 341, 346 (1959). The
Arkansas Supreme Court has expressly approved submis
sion of the penalty option to the jury under instructions
which do nothing more than tell the jury it has a choice.
Hays v. State, 230 Ark. 731, 736, 324 S.W.2d 520, 523-524
(1959) (“ ‘ . . . it will be your duty to assess the punish
ment. The punishment for murder in the first degree is

28 The predecessor of this section, first giving juries^ capital
sentencing discretion in Arkansas, was enacted in 1915. Arkansas
Acts 1915, No. 187, §1, at p. 774.

22 Alford v. State, 223 Ark. 330, 266 S.W .2d 804 (1954); Smith
v State, 205 Ark. 1075, 172 S.W.2d 248 (1 943); Crowe v. State,
178 Ark. 1121, 13 S.W.2d 606 (1929); Webb v. State, 154: Ark
67 242 S W 380 (1922). In murder cases, it is also required that
the jury be’ told to determine the degree of the offense; and an
explicit first degree verdict is necessary to support a death sen
tence Walton v. State, 232 Ark. 86, 334 S.W .2d 657 (I9 6 0 ) ,
Jones v. State, 204 Ark. 61, 161 S.W .2d 173 (1 942); Boy v Stole,
194 Ark. 1155, 109 S.W .2d 954 (1937); Wells v S ate 193 Ark.
1092, 104 S.W .2d 451 (1937) ; but see Burns v. State, 155 Ark. 1,
243 S.W . 963 (1922).

death by electrocution or, at the option of the> jury,. im
prisonment in the penitentiary f ° r th® £ ™ ° ^ g W 2 d
life.’ ” ) ; Hodges v. State, 210 Ark. 672, 674, 197 b. VV.z
12 53 (1946) (“ ‘Gentlemen of the jury, if you
r ^ n - r iic l of guilty an , Ox the P— t * * £
vour verdict will be [reciting verdict form]. • • • ?
S I „ return a verdict and fix the pumshment at life
imprisonment, you evill return have

been^ foum^in^which more elaborate instructions were

given.

3. Arkansas Supreme Court decisions
nature of the option given to the jury are ’
informative. They do no more than to recite
propositions that: “ The legislature evidently meant f
the jury to eserc.se its discretion in selecting the pun

.. S ,.,,230 A * .821 i f £

S the Penalty to
or death. The following transpire . , ,t say The law

‘“ By the Court: No, no, ^ im p riso n m en t in the
simply says that the penal y ^ ^ penalty shall be death in
event of a finding of gui y, P it t0 the jurors to

SMTSi o.'

.. .B y the Court
Jury, you have asked ^ c0^ d e r w l in law to be the most
your deliberations which is ^ electrocution, or
severe penalty. Whethe q statutes provide that
whether it is M e d .r in S t o degree shell

i S n S o ^ S e m s i s a matter for each
Of you to decide to your own sa*isfactl” ' other questions!

‘“ By the Court: Gentlemen, are there any otne

“ ‘By Member of rJ buJ f have got to go by the
‘“ By the Court: i am sony,

law.’ ”

ment . . .” (Needham v. State, 215 Ark. 935, 939, 224
S.W.2d 785, 787 (1949); see Webb v. State, 154 Ark. 67,
72-75, 242 S.W. 380, 383-384 (1922)), and that “ this op
tion lies entirely with the jury.” A lfo r d v. State, 223
Ark. 330, 332; 266 S.W.2d 804, 805 (1954)). The only
Arkansas case found which discourses more fully on the
nature of the option is Scarber v. State, 226 Ark. 503,
505, 291 S.W.2d 241, 242 (1956), where the Supreme Court
said that a jury charge requiring that the jurors not
impose the death penalty unless satisfied beyond a rea
sonable doubt that the defendant should receive it “ placed
a greater burden on the state than it was required to
assume under the law.”

4. One point is quite clear however. The jury’s judg
ment is entirely unreviewable.29 * * 32 33 Although the Arkansas
Supreme Court has spoken o f its power to “reduce a
punishment imposed upon the verdict of a jury . . . on
account of . . . excessive punishment,” Blake v. State, 186
Ark. 77, 80-82, 52 S.W.2d 644, 646 (1932), it has repeatedly
made clear that this power may be exercised to reduce
a death sentence only where (as in Blake) the evidence
is insufficient to sustain a jury verdict o f conviction of
the capital charge. E.g., Smith v. State, 230 Ark. 634,
643-644, 324 S.W.2d 341, 3* 46 (1959).80 Otherwise, with
regard to the death penalty, “as we have many times
stated, the matter of assessing punishment is strictly
within the province of the jury, and we have no power

29 “The statute gives the jury, and not the court, the right to
reduce the punishment. . . . ” Burns v. State, 155 Ark. 1, 8, 243
S.W . 963, 967 (1922) (dictum). Accord: Webb v. State, 154 Ark.
67, 72, 242 S.W . 380, 383 (1922) (dictum).

80 Where the evidence will not support conviction on the capital
charge but will support conviction of a lesser included offense,
the Arkansas Supreme Court reduces the penalty to that provided
by law for the lesser offense. This is not, of course, review of the
sentence but of the conviction.

to change the fixed punishment unless the proof fails
to sustain the charge for which the defendant is con
victed.” Nail v. State, 231 Ark. 70, 85, 328 S.W.2d 836,
845 (1959). Accord: Smith v. State, 230 Ark. 634, 324
S.W.2d 341 (1959); Young v. State, 230 Ark. 737, 324
S.W.2d 524 (1959); Hildreth v. State, 215 Ark. 808, 223
S.W.2d 757 (1949); Rorie v. State, 215 Ark. 282, 220
S.W.2d 421 (1949); Palmer v. State, 213 Ark. 956, 214
S.W.2d 372 (1948); Allison v. State, 204 Ark. 609, 164
S.W.2d 442 (1942); Turnage v. State, 182 Ark. 74, 30
S.W.2d 865 (1930).81 * * * See also Hays v. State, 230 Ark.
731, 324 S.W.2d 520 (1959); Hodges v. State, 210 Ark.
672, 197 S.W.2d 52 (1946).32 This powerlessness of the
Arkansas Supreme Court to review a death sentence stands
in sharp contrast to its freely exercised power to reduce
jury-fixed sentences in non-capital cases.88

81 The only death case in which the Arkansas Supreme Court
has ever assumed to reduce the penalty without upsetting the con
viction is Davis v. State, 155 Ark. 245, 244 S.W . 750 (1922). In
Davis, the Supreme Court disbelieved the complaining witness in
a rape case and responded by granting clemency to the defendant.
Its power to do so was later repudiated, and Davis expressly over
ruled, in the Allison case, text supra, 204 Ark., at 614; 164 S.W.2d,
at 445.

32 It is in the context of this line of decisions that one must
read the phrase in Ezell v. State, 217 Ark. 94, 102, 229 S.W.2d
32, 36 (1950), declining to reduce a death sentence fixed by a
jury “ if it be conceded that we have such power.” See also Black
v. State, 215 Ark. 618, 625, 222 S.W .2d 816, 820 (1949).

33 Carson v. State, 206 Ark. 80, 173 S.W.2d 122 (1943) ; Hadley
v State, 205 Ark. 1027, 172 S .W .2d 237 (1943) ; Marks v. State,
192 Ark. 881, 95 S.W.2d 634 (1936); Ball v. State, 192 Ark. 858,
95 S.W .2d 632 (1936); Hudspeth v. State, 188 Ark. 323, 67 S.W .
2d 191 (1934). Even where the court refuses to reduce a jury-
fixed non-capital sentence, it does not talk of lack of power; it
finds no “abuse of discretion” by the jury, Cook v. State, 225 Ark.
1003, 287 S.W .2d 6 (1956); Wilkerson v. State, 209 Ark. 138, 189
S.W .2d 800 (1945); Cheney v. State, 205 Ark. 1049, 172 S.W .2d
427 (1943), and concedes that “we may reduce extreme penalties
when not supported by the evidence,” Smith v. State, 194 Ark.
1041, 1045, 110 S.W.2d 24, 26 (1937).

5. Describing the Arkansas death-sentencing practice,
used at petitioner Maxwell’s trial, the district court below
wrote:

“ It may be conceded that the Arkansas statutes
dealing with rape and dealing with capital punishment
do not purport to set up any standards by which the
jury is to exercise its discretion in determining whether
it should exercise the power conferred upon it by
section 43-2153, and it will be assumed that no such
standards are to be found in the reported decisions
of the Supreme Court of Arkansas. Nor did the Cir
cuit Court in its charge to the jury attempt to lay
down any principles which should be applied in deter
mining whether petitioner, if convicted, should be
punished by life imprisonment rather than by death.”
(257 F. Supp. at 716; A. 30.)34 *- 36

34 Although the transcript of petitioner’s state trial has not been
included in the materials certified to this Court by the Court of
Appeals for the Eighth Circuit, it was before the District Court
in petitioner’s earlier federal habeas corpus proceeding, see 229
F. Supp. 205, and hence by agreement available to the district
court in the present proceeding (A . 17). The district court refers
to it in the quoted passage, and in stating, for example, that peti
tioner did not testify at his trial (A . 41). The Court of Appeals
sent for the transcript while the appeal was under submission,
and its opinion refers explicitly to it (A . 59, n. 5). For the in
formation of the Court, we set forth in Appendix C the manner
in which the sentencing option was submitted to petitioner’s trial
jury.

36 In a footnote to this passage, the district court adds: “It does
not appear that counsel for petitioner requested any instructions
on the subject.” (257 F. Supp. at 716, n. 6 ; A . 30, n. 6.) The
court of appeals also noted that “the defense in Maxwell’s rape
trial requested no instructional standards.” (398 F.2d, at 149;
A. 65.) But neither court found that petitioner had thereby com
mitted the sort of intentional bypassing required to forfeit fed
eral claims under Fay v. Noia, 372 U.S. 391 (1963); and no such
finding could be made. Petitioner’s constitutional complaint is
that there are no standards fixed by Arkansas law to govern the
jury’s death-penalty decision, and in the absence of such standards,

We note the salient characteristics of the death-sen
tencing process just described.

First, the objective of the process is to differentiate
among individuals, in order to select from the total num
ber of men convicted of rape those who will live and those
who will die. The purpose for which this differentiation
is made is the most arcane, the most intractable judgment
known to the criminal law. “It should be understood that
much more is involved here than a simple determination of
sentence. For the State . . . empowered the jury in this
case to answer ‘yes’ or ‘no’ to the question whether this de
fendant was fit to live.” Witherspoon v. Illinois, 391 U.S.
510, 521 n. 20 (1968).

The ends thought to be served by capital punishment—
hence, the considerations which bear on the fitness of pre
scribing it in a particular case— are variable, and hardly
a matter controlled by consensus or enlightened by the
commonly shared values of our society. Arkansas’ legisla
ture has not identified the ends which it thinks worthy of
consideration in support of its general decision to retain
the death sentence for the offense of rape; nor has it pro
scribed consideration of whatever other ends may occur
to individual men sitting in judgment in rape cases. The
sentencer is informed only by his personal intuition and
particularistic experience concerning the goals in whose
interest he is empowered to kill other men. Sentencing in
non-capital cases—however complex the judgments it may
require—at least rests upon the common assumption that
the offender is to be salvaged if he is salvageable con
sistently with community protection. Capital sentencing

counsel could not have conceived what to request the court to
charge on the issue. Surely, counsel was not required to make up
his own standards, in order to supply the deficiency of Arkansas’
legislation.

rejects that assumption, in the case of some men but not
others, without explaining who or why.

Second, the consequences to the individual who is sen
tenced to die are far less recondite than the objectives
which may, or may not, underlie the decision to kill him.
“Whatever the differences on which this decision hinges,
they remain demeaningly trivial compared to the stakes.”
K alven & Z eisel, T he A merican J ury (1966) 448-449.
The enormity of the stakes is altogether obvious.18

Third, the range of cases and convicted defendants sub
ject to the selection process, and among whom it operates
to differentiate, is extraordinarily broad. The crime of
rape in Arkansas (as in other states where it is capital)
encompasses every variety of consummated sexual assault,
whether on a child or a mature woman, whether the victim
is brutally injured or physically unharmed, whether the
assailant is a prowling stranger or the victim’s social com
panion.87 The range of capital murder cases is similarly

3« W e think we need not elaborate the point before this Court,
which has so frequently recognized it. Witherspoon v. Illinois, 391
U.S. 510 521 n. 20 (1968); Hamilton v. Alabama, 368 U.S. 52
(1961); Williams v. Georgia, 349 U.S. 375, 391 (1955); Stein v.
New York, 346 U.S. 156, 196 (1952); Chambers v. Florida, 309
U.S. 227, 240 (1 940); Powell v. Alabama, 287 U.S. 45 (1932).

37 The Arkansas statute in effect from 1842 until 1967 defined
rape in terms of Blackstone’s venerable formulation as “the carnal
knowledge of a female, forcibly, and against her will.” Ark. Stat.
Ann. § 41-3401 (1964 Re.pl. vol.). But, in this formulation, ‘forc
ibly” means nothing more than the degree of force necessary to
achieve intercourse, and “against her will” signifies only “without
her consent.” This is the doctrine generally followed by states
which retain the common-law formulation of rape, see P erkins,
Criminal L aw (1957) 110-112, 119-127; and it is the rule an
nounced by the Arkansas cases. McDonald v. State 225 Ark. 38,
279 S.W .2d 44 (1955) ; Fields v. State, 203 Ark. 1046, 159 S.W.2d
745 (1942) ; Davis v. State, 155 Ark. 245, 244 S.W . 750 (1922) ;
State v. Peyton, 93 Ark. 406, 125 S.W . 416 (1910); Harvey v.
State, 54 Ark. 425, 14 S.W . 645 (1890).

In 1967, Arkansas enacted a new statute dividing rape into
degrees but not materially affecting the nature of the charge which

U
K

■
■■

■
..a

---

broad.88 It is quite inconceivable that the legislature which
authorized the death penalty in all of these cases thought
that it should be used in very many of them. And, indeed,
it is used in practice in only a relatively small number.89
The prescription of the death penalty, with its discre
tionary incidence, is therefore a striking example of pur
posive overbreadth—that style of legislation which sweeps
far more broadly than its intended target, leaving to ad hoc
judgment in administration the job of deciding what that
target shall be.37 * * * * 38 39 40

Fourth, in the case of the death penalty, the instrument
of administration is the lay jury, selected to try a partic-

constitutes the capital degree- Arkansas Acts 1967, No. 362, p.
830, codified as Ark. Stat. Ann. §41-3401 (1968 Cum. Supp.).
First degree rape, which is capital, consists of sexual intercourse
with a female “by forcible compulsion,” or where the female is
“incapable of consent by reason of being physically helpless, or
mentally incapacitated,” or is less than eleven years old. Second
and third degree rape, which are non-capital, involve intercourse,
whether or not consensual, with females under the ages of 14 and
16 respectively.

38 Arkansas has the common form of first degree murder statute,
A rk . Stat. A nn . § 41-2205 (1964 Repl. vol.), encompassing murders
by poison or lying in wait, felony-murders (arson, rape, robbery,
burglary or larceny), and deliberate and premeditated murders.
The State follows the widely accepted notion that “ premeditation
and deliberation to do murder may be formulated in the assailant’s
mind upon the instant. It does not have to exist in the mind an
appreciable length of time.” Nail v. State, 231 Ark. 70, 75, 328
S.W.2d 836, 839 (1959). See House v. State, 230 Ark. 622, 324
S.W . 112 (1959); Jenkins v. State, 222 Ark. 511, 261 S.W.2d 784
(1953); Gilchrist v. State, 100 Ark. 330, 140 S.W . 261 (1911) ;
Rosemond v. State, 86 Ark. 160, 110 S.W . 229 (1908); Green v.
State, 51 Ark. 189, 10 S.W . 266 (1889). Consequently, as Mr.
Justice Cardozo pointed out many years ago, the line between
second and first degree murder is paper thin; and virtually all
murders are potential first degree cases if the jury takes that view
of them. Cardozo, L aw and L iterature (1931) 97-101.

39 See Appendix B, pp. 24a-34a infra.
40 This Court has pointed out that such legislation is inherently

“susceptible of sweeping and improper application.” N.A.A.C.P.
v. Button, 371 U.S. 415, 433 (1963).

ular case. In pointing out this circumstance, we imply no
general criticism of trial by jury. Recognition of the unique
virtues of jury trial, see Duncan v. Louisiana, 391 U.S. 145
(1968), does not preclude the simultaneous recognition
that jury trial also has certain peculiar dangers, requiring
especial safeguards under the Constitution. See Jackson
v. Denno, 378 U.S. 368 (1964); Rideau v. Louisiana, 373
U.S. 723 (1963); Bruton v. United States, 391 U.S. 123
(1968). And it can hardly be doubted that the jury system,
for all of its merits, is that form of judicial process which
is least capable of developing uniform and consistent rules
of decision.41

Comparison of the jury with a sentencing judge is in
structive in this regard—and none the less because sen
tencing by judges in this country has itself fallen far short
of the marks of regularity and even-handedness.42 The
judge, at least, is a professional sentencer. The very fact
that he sentences a considerable number of offenders pro
motes some consistency in their sentencing, even if it be
only the consistency of his habits. In the process, hopefully,
he gains some generalizable experience, develops bases for
rational comparative judgments; in any event, minimally,
he evolves a “ feel” for how the cases “ line up.” As a pro
fessional, he is somewhat better guarded against unex
amined visceral urges and inflammable emotions than are

41 The point has been succinctly stated, with particular reference
to jury sentencing, by the cognizant Task Force of the National
Crime Commission. “ [T]he transitory nature of jury service vir
tually precludes rational sentencing.” P resident’s Commission on
Law E nforcement and A dministration of J ustice, Task F orce
Report: T he Courts (1967) 26.

42 See e g Rubin, Disparity and Equality of Sentences— A Con
stitutional Challenge, 40 F.R.D. 55 (1966); Institute of J udicial
A dministration, D isparity in Sentencing of Convicted Defen
dants (1954).

lay jurors.43 He shares with other judges and sometimes
with corrections personnel a sense of common enterprise
and responsibility in regard to sentencing offenders that
can be, and often is, a regularizing influence. He talks with
other judges; lawyers talk to him about his sentencing
practices and theirs; he has contact with corrections peo
ple ; and he can be affected, as well, by more formal controls
and guides:—occasional instances of appellate sentencing
review or expressed disapprobation, sentencing conferences
and councils, etc. Notwithstanding all of these moderating
forces, as we have said, judicial sentencing often is ex
tremely erratic. How much more erratic and uneven then
must jurors natively tend to be, who are subject to not a
single one of the controls working on the judge’ A unique
array of twelve untrained individuals assembles; hears one
case; sentences one man; then disperses. We repeat that
we would not in the least disparage jurors’ functioning m
this fashion to decide issues—whether factual, judgmental
or moral—adequately framed by uniform rules of law.
But the prospect that, without rules, the jurors will them
selves supply uniformity or regularity, is hopeless.

Fifth, in making its selection of the men to die, out of all
those convicted, the jury is ordinarily deprived by A r
kansas’ single-verdict procedure of information that is nec
essary to any sort of rational selective judgment. We de
velop this point independently in Part II, pp. 66-78 infra;
but it must be considered, together with the points made

43 “The judge very often perceives the stimulus that moves the
jury but does not yield to it. Indeed it is interesting how often
the judge describes with sensitivity a factor which he then excludes
from his own considerations. Somehow the combination of official
role, tradition, discipline, and repeated experience with the task
make of the judge one kind of decider. The perennial amateur,
layman jury cannot be so quickly domesticated to official role and
tradition; it remains accessible to stimuli which the‘ Judge wi
exclude.” K alven & Zeisel, T he A merican Jury (1966) 497-498.

in paragraphs First through Fourth above, as composing
the background and concrete setting within which Arkan
sas juries exercise the “ option” or “ discretion” allowed
them in capital sentencing. Having sketched the back
ground, we now come to that “ discretion,” which is the
nub of our constitutional complaint.

Sixth, the power of life-or-death decision given to the
jury is absolutely lawless. It is a raw, arbitrary power to
kill or to let live, unguided by principle, undirected by
concern for specified relevant facts, uncontrolled by any
general rules of law, unleashed of any requirement that it
be exercised pursuant to valid reasons or even to agreed-
upon reasons (since the twelve jurors may vote to kill for
twelve disparate reasons), and uncontrollable, unreviewable
by any other power in the legal system. It is, simply, the
power to take away a convicted man’s life for any reason
(good or bad, rational or irrational, generally applicable
or trotted out for the occasion) or for no reason at all—
on a whim, a caprice—or because the defendant did not
take the witness stand; because he took the stand, slander
ously claimed consent, and was disbelieved; or because of
the color of his Bkin.

Let us examine more closely the nature of this extraor
dinary “ discretion.”

(1) Its exercise is not required to rest upon any
prerequisite findings of fact. Such findings, of course,
are required to support virtually every other judgment
of a jury in our legal system. “ I f you find that the
defendant did strike the plaintiff with his car . . . ”—
thus runs the archetypal jury charge. The requirement
of factual findings serves both to guide the conscien
tious juror and to provide a basis for judicial control
by review of the unconscientious one. But a death-
sentencing jury need find no facts, either of a specific

sort (for example, that the rape resulted in substantial
bodily harm to the victim ;44 or that the defendant has
been convicted, or has committed, other rape of-
enses4S), or of a more general sort (for example, that
the defendant is a likely rape recidivist, or that he is
incapable of rehabilitation).

(2) Nor is there any legal prescription, to guide
the jury or enable review of its judgment, of any
preclusive factual findings: circumstances that ex
clude the death penalty (for example, the defendant’s
youth; or conduct by the victim which led the defend
ant on).

(3) Nor is the jury required to consider, or to take
account of, any specified facts or concerns.

(4) Nor is there any enumeration, for the jurors,
of specified facts or concerns which they may consider.
The jury’s attention is not directed to any range or
realm of available factual or judgmental considera
tions.

(5) Nor is there enumeration and proscription of
impermissible considerations. The jury is not told
that it may not sentence the defendant to death for
his unregenerate bad taste in taking the stand and
perjuriously claiming consent; or for not taking the
stand.

(6) There is no prescription of guiding criteria,
legal norms, standards or principles for judgment. In

«« The State of Nevada requires such a finding as the precondi
tion of the imposition of a death sentence in a rape prosecution.
See note 13 supra.

« This is the sort of finding required to support the imposition
of the harsher penalty allowable under common state legislation
dealing with recidivists.

some areas of law, where legal doctrine can do no bet
ter, jurors are permitted to employ such general stan
dards as “ reasonable care” or “ the conduct of a rea
sonable man.” These standards are imprecise, but they
are standards; they serve to tell the jury what the law
has determined is the test or benchmark by which the
defendant is to be judged. The defendant is liable to
the plaintiff if he did not act with the degree of care
which a reasonable man would have employed; the jury
may decide (within limits) what that degree of care is,
but they are told that they must not find against the
defendant if he exercised it. No equivalent guidance
is provided to the death-sentencing jury. It may__
indeed, it must—make up its own governing principles,
which may or may not he those that other juries apply
to other convicted men. In fact, not even this much
rational deliberation is required, for it is illusory to
speak about the death-sentencing “ jury” and “ its” gov
erning principles. The jurors are not directed or re
quired to discuss or to agree upon any common set of
principles; and the court’s instructions (unlike even
such formulations as the “ reasonable man” ) give them
nothing to focus any such discussion.

“ The precise point which prompts the penalty
in the mind of any one juror is not known to us
and may not even be known to him. Yet this dark
ignorance must be compounded twelve times and
deepened even further by the recognition that any
particular factor may influence any two jurors in
precisely the opposite manner.

“ . . . Such factors as the grotesque nature of
the crime, the certainty of guilt, or the arrogant
behavior of the defendant may conceivably have
assured the death penalty. . . . Yet who can say

that these very factors might not have demon
strated to a particular juror that a defendant, al
though legally sane, acted under the demands of
some inner compulsion and should not die? . . •
(.People v. Hines, 61 Cal.2d 164, 169, 390 P.2d 398,
402, 37 Cal. Rptr. 622, 626 (1964).)

(7) There is no prescription even of the ultimate
legal goals and purposes to be considered in the formu
lation of criteria to guide the death-sentencing deci
sion. Anglo-American law addresses its vaguest sort
of directions to courts in such matters as child custody
cases, where the “best interests of the child” or some
equivalent formulation is the touchstone, or in eco
nomic regulation, where “ unreasonable” restraints of
trade are forbidden. Death-sentencing jurors have not
even the direction and delimitation, the concentration
of attention upon specified general objectives, which
these vague formulations comport. Rather, the power
of life-or-death decision is wholly unguided and un
constrained, unlike any other decision made by a law
ful tribunal within our legal traditions.41 It can only
be likened to the power that would be conferred by a
practice—inconceivable in any American court—of
submitting to a jury in a civil case the naked ques

« gee K alvin & Zeisel, T he A merican J ury (1966) 435:
“The discretion which the [death-sentencing] jury in the

United States is asked to exercise is, it should be emphasized,
striking: there is neither rule nor standard to guide it hor
this reason, comparison of judge and jury decision must here
depart from the standard pattern of analysis which discussed
disagreement in terms of why the jury differred from the
judge. W e have viewed the latter [m all sorts of other legal
judgments] as a kind of baseline representing the law, and
we have tried to trace the nuances of jury judgment as it
deviated from the legal norm of the judge. For the death
penalty, however, the judge is not ‘the law but merely an
other decider. In no meaningful sense can it be said that t
judge’s decision is more representative of the law than is the
jury’s.”

tion: “ Should the defendant be liable to the plaintiff!”
— or, in a criminal matter: “Has the defendant done
something for which he should be punished!”

“We wouldn’t turn it over to a jury, the deter
mining of whether the father or the mother or
whether the grandmother or a sister-in-law got
the child, according to the absolute whim or ca
price, or, as you put it, the discretion of the jury.
We wouldn’t turn over to the whim of a jury the
determination o f whether a fox terrier belonged
to the husband or the wife in a separation. We
wouldn’t let a jury determine that with absolute
discretion. Any issue in the whole legal system
that you can think of, rights, property rights, per
sonal rights, are guided by precedents, by stan
dards, and to leave to a jury the absolute discre
tion to determine whether a person lives or dies,
without any guidance, or any compass or standard,
principles or anything else, is foreign to the whole
basic tradition of the Anglo-Saxon common law.
That is the hurdle you have to face and that the
Court has to meet in passing on this question.” 47 * * * * * * * * * * * *

47 W e quote this passage because it summarizes better than any
thing written, on the subject, and better than anything we could
say, the nature and the vice of standardless discretion in capital
sentencing by a jury. The passage is a transcription of a statement
by Chief Justice Traynor during a colloquy with the Assistant
Attorney General of California in the course of the argument in
In re A nderson,------- C al.2d --------- , 447 P.2d 117, 73 Cal. Rptr. 21
(1968), a case discussed at pp. 45-56 infra. See Transcript of
Proceedings in the Supreme Court of California, In re Anderson,
Crim. No. 11,572, March 28, 1968, pp. 107-108. However, because
questions asked from the bench may convey a misleading impres
sion of a judge’s views, and because we would not wish to run the
slightest risk of misrepresenting those of Chief Justice Traynor
— even while using his words exclusively for the purpose of ex
pressing our own thoughts— we feel obliged to set forth here two
subsequent, related passages in the colloquy:

“Chief Justice Traynor: I wonder if you were making
this argument, Mr. Harris. I don’t want to put words into

(8) There is no review of the jury’s decision, and
no judicial safeguard of any sort against invidious dis
criminations and other abuses. It is therefore not sur-

your mouth. You look at this problem of standards right in
the face and it’s awfully easy to talk about ships without
rudders or compass, boats without oars and so forth, and about
assigning a fox terrier to one person or another without
standards, but in this area you just can’t find workable stan
dards. Any kind of formula that you put up would simply
be magic words, at best. W e haven’t heard of any precise
standards that wouldn’t be like mouthing formulas to the
jury that it should take into consideration such and such
factors and so forth. You can’t get anything as precise as a
rudder or an oar and so forth. That being the case, the ques
tion is whether you are going to have death or not. The State
could say that every killing or certain killings, just auto
matically brought on the death penalty. The present system
is a dispensation that the State has made. It is futile to think
of standards. I don’t know whether you make that argument
or not.

“ Chief Justice Traynor: Let me see if I can restate your
argument about discretion, that there are many instances, as
you pointed out, where matters are left to the discretion of
the trial judge, and that is because the appellate courts, in
their wisdom and experience, don’t know what precise stan
dards should be set down, because they are not— they haven’t
had enough experience, haven’t had enough cases to lead them,
haven’t had enough specific items. So when the appellate court
doesn’t know the answer, it turns the matter over to the dis
cretion of the trial judge, just as in many instances when we
don’t know the answer to a real tough question, we turn it
over to a jury.

“Now, here is the toughest question of all that human beings
have to face, as to whether a man should die or not. A man,
say, who has thrown gasoline into a tavern and caused, say,
seven or eight people to be burned; a man who rapes a girl
and savagely cuts her u p ; who does all of the other heinous
things that some of these people have been convicted of.

“Now, the question is this: Should that person die or not!
Where are you going to get the guidance that is going to tell
you whether a person should die or not!

“Maybe the answer is that because you have no answer, a
civilized society wouldn’t have the death penalty. But you
might retaliate to that, that is really a question for the legis
lature, but on this question of standards it is impossible to

prising that the performance of jurors in the exercise
of the lawless discretion given them is not merely ir
regular48 and arbitrary in particular cases,49 hut ex-

articulate a standard. Of all the questions that could be pro
pounded to mankind, this is one which would defy a Solomon.’

(Id., at pp. 113-114, 120-121.) W e might add, concerning this
last quoted passage, that again in our judgment it goes directly to
the heart of the matter. W e quite agree that common experience
furnishes no easy answers relating to the standards that should
govern capital sentencing. A ll the more reason, we have suggested,
why the death-sentencing decision cannot properly be left to the
unguided decisions of individual juries. (See pp. 33-34 supra.)
W e deny that it would be impossible for a legislature to provide
standards governing capital sentencing, if the legislature gave the
matter proper attention and had rational ultimate goals in author
izing capital punishment as the available penalty in any case.
(See note 67 infra.) However, if the nature of the death penalty,
or of a legislature’s reasons for ordaining it, is such that no reg
ular, rational, even-handed policies of general applicability can he
formulated to govern its administration— with the necessary result
that individuals must be condemned to die irregularly, irrationally,
unevenly: in short, arbitrarily and without due process of law—
then it is our view precisely that the Due Process Clause forbids a
civilized society to use this sort of penalty. I f a sanctions use
cannot be made consistent with due process, the Constitution of the
United States makes quite clear that the sanction, not due process,
must be abandoned.

48 See K alven & Zeisel, T he A merican J ury (1966) 437-449.
Examining jury penalty decisions in 111 death cases, and the
presiding judges’ evaluations of them, Kalven and Zeisel find that
jurors and judges agree on a sentence of imprisonment in 76 cases,
agree on death in 14, and disagree in 21. Significantly, therefore,
where death is the outcome, there is more often disagreement than
agreement. Patterns in jury sentencing emerge, in the sense that
death-sentence cases are frequently characterized by the same sorts
of aggravating circumstances. But these same factors appear in
many cases where the death sentence is not imposed. “Many of the
murder cases in which the judge and jury disagree on the death
penaly appear no less heinous than those in which they agree. Id.,
at 439. “ The leniency categories have a plausible ring. But the
brute fact is that each time one of the factors listed was persuasive
to one of the deciders, it was unpersuasive to the other. Either the
judge or the jury was willing, despite the presence of the leniency-
disposing factor, to have the defendant executed. Id., at 444.

49 See pp. 11-13 supra.

hibits grossly unconstitutional discriminations60 61 which
the courts have been unable to control or correct. We
put aside, for present purpose, the question whether a
state is federally obligated to provide at least some
minimal form of judicial review of a jury’s determina
tions;51 or whether, if its highest court freely reviews
jury-fixed sentences in non-capital cases (as does Ar
kansas’), it may deny all power of review in capital
cases alone.62 The point we make here is that a total
absence of standards to govern the jury’s decision
making function—a defect that would defeat any
meaningful judicial review which is allowed—is the
more baneful, if possible, where there is no judicial
review. The jury makes the one and only judicial de
cision that a capital defendant is allowed on the ques
tion whether he lives or dies; and makes that decision
entirely arbitrarily.

B. The Grant of Lawless Power in Capital Sentencing is
V nconstitutional.

We submit that the practice just described violates the
rule of law basic to the Due Process Clause. The same
submission was recently made to the Supreme Court of
California, which rejected it by a vote of four Justices to
three. In re Anderson,------ Cal.2d------ , 447 P.2d 117, 73
Cal. Rptr. 21 (1968). Justice Tobriner, joined by Chief
Justice Traynor and Justice Peters, dissented in an opinion
that states our position with incomparable lucidity. It was
the view of the dissenters that the California statutes grant
ing unfettered discretion to juries in capital sentencing

“ violate the Fourteenth Amendment of the Constitu
tion of the United States because they provide no

60 See pp. 13-20 supra.
61 Cf. Thompson v. City of Louisville, 362 U.S. 199 (1960).
62 Cf. Baxstrom v. Herrold, 383 U.S. 107 (1966).

standards or tests whatsoever to enable judge or jury
to decide why one convicted capital defendant should
die and another should live. The California penalty
trial leaves this vital decision to the unguided whim
and caprice of the trier of fact; this irrational process,
the antithesis of due process, has no place in the con
stitutional structure o f American law.” (Tobriner, J.,
joined by Traynor, C.J., and Peters, J., concurring
and dissenting in In re Anderson, supra, 73 Cal. Rptr.
at 36.)63

We commend the reasoning of the Anderson dissenting
opinion to this Court. After examining the nature of the
power which standardless capital-sentencing legislation
confers upon the jury and concluding that it requires the
jury “ to perform a sui generis function which subjects the
convicted capital defendant to a power of arbitrary deci
sion’’ (Anderson Dissent, p. 40 (original emphasis)), the
opinion states the constitutional premise against which
such a power must be judged:

“ The constitutional imperative that laws infring
ing upon life and liberty be framed in terms of rea
sonably ascertainable standards is central to our ad
ministration of criminal justice.” (Anderson Dissent,
p. 42.)

Surely, this proposition is beyond dispute. For whatever
else “due process of law” may encompass, it has always
been thought to impose some demand of fundamental pro
cedural regularity in decision-making, some insistence upon
the rule of law, some adherence to the principle established
by Magna Carta that the life and liberty of the subject

63 The dissenting opinion is at 73 Cal. Rptr. 36-59. W e shall
hereafter cite it as Anderson Dissent, p. . . . , referring to the
pages in 73 Cal. Rptr.

should not be taken but by the law of the land. This Court
has long condemned the sort of vagueness in criminal
statutes that “ licenses the jury to create its own standard
in each case,” Herndon v. Lowry, 301 U.S. 242, 263 (1937).
See, e.g., Smith v. Cahoon, 283 U.S. 553 (1931); Cline v.
Frink Dairy Co., 274 U.S. 445 (1927); Winters v. New
York, 333 U.S. 507 (1948). Statutes which authorize ad
hoc adjudications unconstrained by legal principles of gen
eral application thereby confer a “naked and arbitrary
power” (Tick Wo v. Hopkins, 118 U.S. 356, 366 (1886)),
which is at war with Due Process. The vice of such statutes
is not alone their failure to give fair warning of prohibited
conduct, but the breadth of room they leave for jury ar
bitrariness and the influence of impermissible considera
tions, N.A.A.C.P. v. Button, 371 U.S. 415, 432-433 (1963);
Freedman v. Maryland, 380 U.S. 51, 56 (1965); Lewis,
The Sit-In Cases: Great Expectations, 1963 Supreme
Court R eview 101, 110; Note, 109 U. P a. L. R ev. 67, 90
(1960), including racial considerations, see Louisiana v.
United States, 380 U.S. 145 (1965); Dombrowski v. Pfister,
380 U.S. 479 (1965); Cox v. Louisiana, 379 U.S. 536 (1965),
and every other insidious urging of caprice or prejudice.

The Anderson dissent points out that this basic princi
ple of legality, of the rule of law, is no less applicable to
a court’s sentencing function than to its guilt-determining
function. It notes this Court’s recent explicit recognition
in Witherspoon v. Illinois, 391 U.S. 510, 521 n. 20 (1968),
that while sentencing choice—and, in particular, the choice
of life or death—may be :

“different in kind from a finding that the defendant
committed a specified criminal offense, . . . this does
not mean that basic requirements of procedural fair
ness can be ignored simply because the determination
involved in this case differs in some respects from the

traditional assessment of whether the defendant en
gaged in a proscribed course of conduct.”

Witherspoon, say the Anderson dissenters:

“ therefore squarely stands for the proposition that the
requirements of the Fourteenth Amendment—those
that go to the very fairness and integrity o f the
penalty-determining process—apply to the procedure
pursuant to which the state, be it by judge or jury,
takes the life of a capital offender.” (Anderson Dis
sent, p. 40.)

And other recent decisions of the Court point ineluctably
to the same conclusion. See cases cited at p. 27 supra.

The question, then, is whether an unlimited grant of
arbitrary power to make the life-or-death sentencing choice
falls afoul of the Fourteenth Amendment’s requirements
that “go to the very fairness and integrity of the penalty
determining process.” The Anderson dissenters conclude,
and we here submit, that it does, for several reasons.

First, the totally undefined issue that is presented for
the jury’s decision makes it impossible for the defendant,
whose life is at stake, advisedly and intelligently to pre
pare and present his defense on the sentencing question.
We shall see in Part II of this brief, pp. 66-78 infra, that
the single-verdict trial procedure in capital cases im
measurably exacerbates the unfair litigation posture in
which the defendant is put, by requiring him to fashion
his case in a manner that sets at loggerheads his interests
in a fair trial of the guilt question and a sound penalty
determination. But, even without this additional handicap,
the capital defendant who confronts a jury armed with
unconfined and unguided power to sentence him to life or
death is denied the least semblance of a fair trial on the
matter of penalty. This Court has long recognized the

“ principle of procedural due process . . . that notice of
the specific charge, and a chance to be heard in a trial of
the issues raised by that charge, if desired, are among the
constitutional rights of every accused in a criminal pro
ceeding in all courts, state or federal.” Cole v. Arkansas,
333 U.S. 196, 201 (1948). See In re Gault, 387 U.S. 1, 33-34
(1967). Even in non-criminal matters, the Court has found
a Due Process right to adequate notice of the issues posed
for adjudication in a proceeding affecting individual in
terests. E.g., Holden v. Hardy, 169 U.S. 366, 389 (1898);
Morgan v. United States, 304 U.S. 1 (1938); Gonzales v.
United States, 348 U.S. 407 (1955). A fortiori, such no
tice and an “ opportunity to be heard” are required in the
sentencing phase of a criminal case, Specht v. Patterson,
386 U.S. 605, 610 (1967), particularly where the jury is
“ empowered . . . to answer ‘yes’ or <no’ to the question
whether this defendant [is] . . . fit to live,” Witherspoon
v. Illinois, 391 U.S. 510, 521 n. 20 (1968). But without
rules of decision “ sufficiently definite to guide the lawyer
in contesting their applicability to his client . . . , the right
to a hearing would be of little value, because there would
be no adequate standards toward which argument and
evidence could be directed.” Note, 62 Harv. L. Rev. 77, 78
(1948).64

64 The Anderson Dissent, p. 44, n. 9 quotes the following passage
from this Harvard Note, which identifies the two due process con
cerns that are at the core of the present case:

“ ‘ [D]ue process requires the adjudication of an individual’s
rights and duties to be governed by rules of sufficient objec
tivity to guard against an arbitrary or ad hominem result.
It follows that such rules, when embodied in a statute, must
be definite enough to enable the judge to make rulings of law
and charges to the jury which are so closely referable to the
statute as to assure consistency of application.

“ ‘Due process requires, in addition, that the rules be suffi
ciently definite to guide the lawyer in contesting their ap
plicability to his client. Without this latter requirement, the
right to a hearing would be of little value, because there would

It would be fair to describe the defendant’s situation in
a capital trial as a deadly guessing game. He cannot know
what facts or positions will save him or cost him his life.56
The legislature has not told him, or told the jury, when
capital punishment is to be imposed or why it is to be im
posed; and the jurors are free to take their own views.
Views on capital punishment being what they are in our
society, and the justifications for the death penalty being
rather obscure matters of personal taste in the absence of
some legislative specification, the jurors’ attitudes will
range wildly. A defense of consent which may not con
vince the jury to acquit the defendant may nevertheless
lead it to the conclusion that—because the prosecutrix was
less than firm in her resistance—the defendant should not
be executed. Or the jury may think the degree of resistance
by the prosecutrix irrelevant. Particularly where the de
fendant is Negro and the prosecutrix white, the defense
of consent may cause the jury to execute the defendant for
the unpardonable offense of adding slander to rape. Given
its grisly risks, how is counsel to know what he stands
to gain by making the defense T 65

be no adequate standards toward which arguments and evi
dence could be directed. Both of these requirements would
seem to be satisfied by the same degree of definiteness; a statute
which is sufficiently definite to guide the judge should also
be sufficiently definite to guide the lawyer in litigation, be
cause there the function of each is the same— to test applica
tion of rules to particular situations.’ (Note, Due Process
Requirements of Definiteness in Statutes (1948) 62 Harv. L.
Rev. 77-78.)”

65 “W e must assume that in establishing the alternative punish
ments of death and life imprisonment rather than setting down
one mandatory punishment, the Legislature concluded that not all
capital offenders should be punished by the extreme penalty. . . .
Yet the current administration of the death penalty, by failing to
provide criteria sufficiently ascertainable to guide courts and juries
in making that distinction, prevents a convicted capital defendant
from knowing how to show that he falls within the class of capital
offenders for whom the law contemplates the lesser penalty of life
imprisonment.” Anderson Dissent, p. 44.

If the defendant was under the influence of alcohol or
drugs, the jury may take the view that that is a mitigating
incident—or an aggravating one. Five jurors may have
sympathy for a defendant who displays a severe mental
or emotional disorder not amounting to legal insanity;
while seven, who would reserve the death penalty for just
such perverted savages, vote to kill him. for his disorder.
See People v. Hines, quoted at pp. 40-41, supra. The re
sult is not a litigation in any ordinary sense but a flailing
contest or a gambling escapade in which the wages of a
misplaced guess are death. This is simply not a Due
Process trial. See Anderson Dissent, pp. 43-44.

Second, “ the complete absence of standards in the ad
ministration of the death penalty deprives the convicted
capital defendant of any way to protect himself against an
arbitrary imposition of the death penalty. Since no limita
tions bound the exercise of the discretion of the trier of
fact, the defendant can neither challenge the evidence in
troduced . . . on the ground of insufficiency nor seek review
on the ground of erroneous application of the death penalty
to him.” Anderson Dissent, p. 44. Standardless death-
sentencing thus effectively strips the convicted capital de
fendant of all of the safeguards of the Constitution in con
nection with the life-or-death penalty decision. Jurors may
inflict death upon him for reasons or by processes forbidden
by the clearest constitutional commands, and yet get away
with it. The inability of the district and circuit courts
below to come to grips with petitioner Maxwell’s contention
of racial discrimination is a striking example of how con
stitutional protections founder in the sea of limitless dis
cretion that envelops and conceals the jury’s decision
making process. The district court avoided Maxwell’s com
pelling statistical showing of racially discriminatory sen
tencing patterns on the part of Arkansas juries by as
suming that the factors which might be affecting the jurors

in their selection of the men sentenced to die were so subtle
and complex that statistics could not take account of them
all. (See A. 39-40.) Thus, although the court itself found
as a fact that Negroes convicted of the rape of white vic
tims were disproportionately frequently given death sen
tences, and although it was uncontested that no non-racial
factor of which analysis was possible on the basis of an
exhaustive empirical study could explain the racial death-
sentencing differential, the court was able to imagine fac
tors that must have escaped identification or measurement
and that could account for the differential.56 Perhaps. Such
is the range of freedom in the jurors’ sentencing decision
that almost any speculation is possible.

The court of appeals took a somewhat more hard-headed
view of the evidence and did not really deny that petitioner
had proved a state-wide practice of racial discrimination.
It rejected his Equal Protection contention principally on
the ground that the particular jury which sentenced Max
well to die had not been shown to have a discriminatory
motivation. (A. 59-64.) Doubtless this last conclusion is
factually correct. Since, under Arkansas law, Maxwell’s
jurors could have had almost anything in mind—or nothing
in mind—to support their death verdict, it is quite impos
sible to say.

It has been pointed out that one of the purposes of the
Due Process requirement of definiteness in penal legisla
tion is to delimit “ to what extent the administration of
public order can assume a form which, first, makes pos
sible the deprivation sub silentio of the rights of particular
citizens and, second, makes virtually inefficacious the fed
eral judicial machinery established for the vindication of
those rights.” Note, 109 U . P a . L. R ev . 69, 81 (1 96 0 ). The 55

55 Viz., the “issue of consent” discussed at A. 40.

function of courts to protect against the deprivation of
specific constitutional guarantees—such as the guarantee
against racial discrimination—“ requires at the least a con
tinued surveillance to assure the regularity of legislated
and administered modes of applying public compulsion to
all particular men. It requires, as a corollary, the mainte
nance of a state of institutions in which , that surveillance
is kept effective.” Id., at 89. If ever a form of penal enact
ment flouted these requirements, and thereby established a
regime of administration in which nothing was certain but
that denials of federal rights would go undetected and un-
correctible, standardless death-sentencing legislation is that
form.

Third, the arbitrary power which this legislation confers
to differentiate among convicted rape defendants by sen
tencing some to life and some to death for any or no reason
violates a constitutional concern in regard to which “ the
equal protection and due process clauses of the Four
teenth Amendment overlap.” Anderson Dissent, p. 47.
This is the concern for even-handed treatment which al
lows “no invidious discriminations between persons and
different groups of persons,” Griffin v. Illinois, 351 U.S. 12,
17 (1956), and ordains that all shall “ ‘stand on an equality
before the bar of justice in every American court,’ ” ibid.,
quoting Chambers v. Florida, 309 U.S. 227, 241 (1940). A
corollary of these propositions is that differences in the
treatment of individuals by the courts must have some ra
tional basis: that is, there cannot be differentiations with
out differences, nor differentiations along lines of differ
ence that have no rational relation to the purpose for dif
ferentiating. E.g., Rinaldi v. Yeager, 384 U.S. 305 (1966);
Baxstrom v. Herrold, 383 U.S. 107 (1966). Particularly
where “legislation . . . involves one of the basic civil rights
of man . . . , strict scrutiny of the classification which a
State makes . . . is essential, lest unwittingly or otherwise,

invidious discriminations are made against groups or types
of individuals in violation of the constitutional guaranty
of just and equal laws.” Skinner v. Oklahoma, 316 U.S.
535, 541 (1942).

The Skinner case held that a state could not, by legisla
tive classification, draw the line of sexual sterilization be
tween thieves and embezzlers. “ When the law lays an un
equal hand on those who have committed intrinsically the
same quality of offense and sterilizes one and not the other,
it has made . . . [a constitutionally forbidden invidious
discrimination].” Id., at 541. We take it that, under
Skinner, a state could not constitutionally draw the line of
death between thieves and embezzlers either; nor could it
permit its juries to draw so irrational a line in the exercise
of a general death-sentencing discretion. I f this is so, it
is difficult to conceive how Arkansas may give its juries
arbitrary power—which necessarily includes the compe
tence to draw lines as unsubstantial as the line between
thieves and embezzlers, or more unsubstantial still—with
out falling afoul of the same constitutional interdiction.

This is not to say, of course, that a state may not consti
tutionally provide for differential sentencing of offenders
convicted of the identical offense. There is no inconsist
ency whatever between the principle of equality and a
procedure for individualizing sentencing—even death sen
tencing. “Equality and individualization go together; they
complement each other. They are of one piece.” 67 This is
so because any system of individualization that efficiently
serves the penal interests of the state must necessarily be
based upon some rational scheme for differentiating among
offenders; and if the bases of differentiation are rational,

57 Rubin, Disparity and Equality of Sentences— A Constitutional
Challenge, 40 F.R.D. 55, 69 (1966).

the Fourteenth Amendment is satisfied. The trouble with
standardless death-sentencing legislation is that it author
izes irrational differentiation.57 68 69 By

“providing no classificatory criteria whatsoever, [it]
permits the trier of fact arbitrarily to determine
whether [a] particular convicted capital defendant
should suffer death rather than life imprisonment.
Such absence of classification renders the current pro
cedure insufficient to satisfy even the minimum require
ment of the equal protection clause that differential
treatment of persons prima facie similarly situated
must bear some rational relation to a discernible legis
lative purpose. In failing to provide any rational basis
upon which to justify an imposition of the death pen
alty on those particular capital defendants sentenced
to death rather than life imprisonment, [standardless
death-sentencing statutes make an invidious discrimi
nation]. . . . ‘It is clearly unconstitutional to enable a
public official [let alone an ad hoc group of 12 individ
uals]6’ to * * * engage in invidious discrimination
among persons or groups * * * by use of a statute pro
viding a system of broad discretionary * * * power.’
(Cox v. State of Louisiana, . . . 379 U.S. 536, 557

68 “Recognition of the problem created by disparate sentences
does not, however, compel the conclusion that all offenders who
commit the same offense ought to receive the same sentence— the
theory of ‘uniform sentencing.’ There may be differentiating per
sonal and social factors which require ‘individualization’— different
sentences for offenders who have engaged in the same anti-social
conduct. But if a system of individualization is to satisfy objec
tions against disparity, the disparity inherent in individualization
must result from the articulated application of uniform criteria
designed to effectuate agreed-upon goals.” Note, 69 Yale L.J.
1453, 1459 (1960).

69 This bracketed insertion is by the author of the Anderson
dissent. The others in the passage are ours. Emphasis throughout
is found in the original.

[1965] . . . ; Yick Wo v. Hopkins, . . . 118 U.S. 356, 366-
368 [1886] . . . .) Thus, the trier of fact’s absolute
power to impose the death penalty on a convicted cap
ital defendant violates the basic interdictment of the
equal protection clause that no person shall hold ‘a
naked and arbitrary power’ to make invidious dis
criminations against another. (Yick Wo v. Hopkins
supra, 118 U.S. at pp. 366-368 . . . ) ’ ’ (Anderson Dissent
pp. 49-50.)

Fourth, and most fundamentally, the process of the stan
dardless death-sentencing decision subjects the defendant
to an exercise of legally unregulated power which is anti
thetical to the rule of law expressed by Due Process.

The epitome of the vice of absence o f procedural safe
guards inherent in vague and standardless statutes
thus permeates the administration of the death pen
alty: the defendant sentenced to death cannot even
show that an ad hoc group of 12 jurors . . . exercising
absolute power over his life abused this discretion.
[The statutes giving the jury such power] . . . thus
deny due process of law because they provide no
standards by which an abuse of discretion by the trier
of fact can be curbed or even subjected to review. The
power of the trier of fact to decree the death sentence
is as broad and arbitrary as it is absolute and un
touchable.” (Anderson Dissent, p. 44.)

In this aspect, what is constitutionally wrong with stan
dardless death-sentencing is that it collides violently with
a principle so basic that it hardly ever is or needs to be
explicitly asserted, because our entire system of law as
sumes it. It is the principle of legality. Standardless
death-sentencing is its opposite: arbitrariness. In order to
protect against the regime of arbitrariness, the fabric of

our legal structure—laws, Constitution, and courts—has
been established. The function of the structure, to main
tain the rule of law, has as its one most essential objective
to deny arbitrariness entrance to the system. For “ the
very idea that one man may be compelled to hold his life,
or the means of living, or any material right essential to
the enjoyment of life, at the mere will of another, seems
to be intolerable in any country where freedom prevails, as
being the essence of slavery itself.” Yick Wo v. Hopkins,
118 U.S. 356, 370 (1886).

Under the principle of legality, it could scarcely be con
tended that an Arkansas statute would be valid which pro
vided: “ Whoever is found condemnable in the discretion
of the jury shall be guilty of an offense.” Yet we submit

' “ See Packer, The Limits of the Criminal Sanction (1968)
92-94:

“Let us suppose for a moment that a legislature, contem
plating the momentous problem of drafting a new criminal
code, decided to solve its problem by a single grand enactment:
whoever does anything bad shall be punished as justice may
require. When a court is called upon to decide whether John
Jones, who hit his neighbor over the head with a hammer
during the course of an argument, has violated the statute,
is the court making law or is it applying law? The question
is absurd, of course. In a sense, the court is doing both. But
the significant question, which is concealed by categorical
treatment of law-making and law-application, is: how great
a degree of freedom is the court allocated in deciding whether
to treat John Jones’s conduct as criminal? The extremity of
the example makes the issue clear. . . . It hardly needs argu
ment that the hypothetical criminal enactment just proposed
— whoever does anything bad shall be punished as justice may
require— does violence to the principle of legality. A s the ex
ample shows, it is not enough that the law formally being
applied is law in existence at the time the conduct complained
of occurred. Devices are needed to ensure that the amount of
discretion entrusted to those who enforce the law does not
exceed tolerable limits. But the working out of the devices
and the decision about what limits are tolerable are functions
that fall to the courts. It is, of course, no accident that they
fall to the courts; neither, however, is it the result of any

5 8

that this suppositious statute stands no differently in light
of the relevant concerns of Due Process than the unregu
lated sentencing practice under which Arkansas juries con
demn some but not other men to die. The statute and the
sentencing practice have an indistinguishable vice: both
permit adjudications that take away a man’s liberty or his
life on an entirely ad hoc basis, without reference to rules
of decision that have ever been applied, or will ever be ap
plied, to any other man. When selection among individ
uals for the purpose of killing some of them is made in
this fashion—according not to laws of general application
but simply to the caprices of moment-to-moment, arbitrary
opinions11—more is wrong than that the men selected to die
are treated unequally as compared with the rationally un
differentiated men selected to live, or that the men selected
to die are treated irrationally with regard to any purpose
that might be advanced for the selection process or its
lethal consequences. What is wrong, more essentially, is
that the men selected to die are treated lawlessly. “ Cer
tainly one of the basic purposes of the Due Process Clause

omnicompetent lawgiver’s deliberate plan. It is, very simply,
an institutional necessity. . . .

“ . . . The devices worked out by the courts to keep the prin
ciple of legality in good repair comprise a cluster of doctrines
that give the criminal law much of its distinctive content. For
our purposes it is enough to identify and describe two of these
doctrines, which fairly represent the values involved. They are
the void-for-vagueness doctrine and the doctrine requiring
strict construction of penal statutes. . . .

“ Under the vagueness doctrine in its starkest form, the court
says to the legislature: you have given so much discretion in
picking and choosing among the various kinds of conduct to
which this statute may be applied that we will not let it be
applied at all. That is unquestionably the response that an
American court would give to the prosecution of John Jones
under the hypothetical bad conduct statute discussed above.”

81 Cf. Mr. Justice Black, concurring, in Cox v. Louisiana, 379
U.S. 536, 579 (1965), approved in Shuttlesworth v. City of Birming
ham, 382 U.S. 87, 90 (1965).

has always been to protect a person against having the
Government impose burdens upon him except in accordance
with the valid laws of the land.” Giaccio v. Pennsylvania,
382 U.S. 399, 403 (1966).

It is true, of course, that the “whoever-is-found-con-
demnable” statute deals with the definition of crime, while
Arkansas’ sentencing practice regulates the degree of pun
ishment for crime. Questions of “ fair notice” that are im
plicated in crime-defining provisions are not necessarily in
volved in provisions relating to penalty alone. But, as we
have said above, the Due Process requirement of definite
ness in penal laws is not simply a command of notice; it is
also a command of regularity. N.A.A.C.P. v. Button, 371
U.S. 415, 433 (1963); Note, 109 U. P a . L. R ev. 67, 90
(1960). We think it obvious that “ whoever-is-found-con-
demnable” is constitutionally bad not principally because a
man does not know how to behave consistently with it, but
because—however he behaves—he may be arbitrarily and
capriciously taken by the heels. The fault of that sort of
statute is that it “ injects into the governmental wheel so
much free play that in the practical course of its operation
it is likely to function erratically—responsive to whim or
discrimination unrelated to any specific determination of
need by the responsible policy-making organs of society.
. . . ” Ibid. The precise vice inheres in wholly unregulated
jury discretion to sentence a convicted man to life or death.
He too may be dealt with arbitrarily, his life extinguished
for any reason or for none at all.

Giaccio v. Pennsylvania, supra, supports, if it does not
compel, the conclusion that Arkansas’ standardless grant
of discretion to its juries in capital sentencing is unconsti
tutional. What was at issue there, as here, was a state
practice governing disposition. No “ fair notice” problem
was involved—except, of course, the problem present in at

least an equal degree in a capital case tried to a jury with
limitless sentencing power62—that it was impossible at the
trial to know what issues were being tried. But this proce
dural deficiency, however important, was not the crux of
Giaccio. That decision turned squarely on the proposition
that the Fourteenth Amendment forbade Pennsylvania to
leave its “ jurors free to decide, without any legally fixed
standards,” 382 U.S. at 402-403, whether to impose upon
a defendant a rather small item of costs.63 It is not evident
why, in the infinitely more significant matter of sentencing
men to death, Arkansas juries can be permitted the same
lawless and standardless freedom.

Nor does footnote 8 in the Giaccio opinion, relied upon
by both courts below,64 blunt the implications of the Giaccio
holding for our present purposes. In that footnote, the
Court said that it intended to cast no doubt upon the
constitutionality of leaving to juries finding defendants
guilty of a crime the power to fix punishment “ within
legally prescribed limits.” The problem in this case is
precisely that there were no “ legally prescribed limits,” in
any realistic sense, to the jury’s discretion. The Giaccio
footnote speaks for jury sentencing generally, not capital
sentencing. But, once again:

“ It should be understood that much more is involved
here than a simple determination of sentence. For the
State . . . empowered the jury in this case to answer
‘yes’ or ‘no’ to the question whether this defendant

62 See pp. 48-51 supra.
63 It should be noted that no First Amendment rights or other

preferred federal guarantees demanding the special protection
afforded by a heightened requirement of statutory specificity, see
United States v. National Dairy Prods. Corp., 372 U.S. 29, 36
(1963), were involved in Giaccio.

See A. 32, 67.

was fit to live.” (Witherspoon v. Illinois, 391 U.S.
510, 521 n. 20 (1968).)

Given the imprecision of the sentencing art, even when
performed by judges, see Tigner v. Texas, 310 U.S. 141,
148-149 (1940), it may well be that juries can constitu
tionally be given some discretion in selecting a smaller
or larger fine, a longer or shorter term of years, partic
ularly where the range of choice is relatively circumscribed
and the effect of the choice is somewhat qualified by parole
statutes and the continued availability of post-conviction
process to rectify after-discovered mistakes made at the
trial. But the degree of arbitrariness allowed a State is
not so liberal where grave and irremediable punitive ex
actions are at stake, see Skinner v. Oklahoma, 316 U.S.
535 (1942); and none is graver or more irremediable than
the sentence of death by electrocution.

Nor did the Court in Giaccio, adverting to the problem
of jury sentencing, have before it what has been presented
here: evidence showing that in fact the capital sentencing
discretion given juries has been exercised on the ground
of race explicitly forbidden by the Constitution. See pp.
13-20 supra. We have not the slightest hesitation in con
tending that the arbitrary and standardless discretion
afforded Arkansas juries constitutes per se a flagrant vio
lation of the Fourteenth Amendment. But, in addition,
this discretion has “ in operation displayed the latitude it
[allows] . . . for discontrol, irrationality and irregularity.”
Note, 109 U. Pa. L. Rev. 67, 108 (1960). There can be no
better demonstration o f the potential, the inevitable
tendency, of this sort of law than what has happened in
its administration in a State where race has spelled the
difference between life and death. See cases cited in note 3
supra.

Of course, petitioner does not contend that the Due
Process Clause forbids entirely the exercise of discretion
in sentencing, even by a jury and even in a capital case.
Ways may be found to delimit and to guide discretion,
narrow its scope and subject it to review; and these may
bring a grant of discretion within constitutionally tolerable
limits. Whether the approach taken by a State such as
Nevada, which makes certain reviewable findings of fact
the indispensable condition of imposing capital punish
ment,'15 or that of the Model Penal Code, which both estab
lishes prerequisite findings and enumerates aggravating
and mitigating circumstances,” or that of the states which
provide plenary review of capital jury sentencing by trial
and/or appellate courts, would be constitutional, is not
the question presented. Here, Arkanses has taken none
of these available steps, used none of these available de
vices, to hold its jurors back from the exercise of “ a naked
and arbitrary power,” Tick Wo v. Hopkins, 118 U.S. 356,
366 (1886).

One point involved in this totally unconfined regime of
discretion deserves emphasis. It is often said, and rightly,
that it is a vexing and difficult job to select from among
all capital offenders those upon whom death will actually
be imposed. See note 47 supra. But by making the death
penalty non-mandatory, the Arkansas legislature has as
sumed that such a selection can be made, and has required
it to be made. Selection might be achieved in either of
two ways: by leaving the matter to ad hoc judgment, or
by formulating generally applicable principles of dif
ferentiation. A legislature may adopt the ad hoc approach
for one or another reason. It may have no generally appli-

” See note 13 supra.
” See A merican Law Institute, Model Penal Code, §210.6

(P.O.D. May 4, 1962), pp. 128-132.

c a U e p r in c ip le s in v ie w , a n d in t e n d t h a t th e s e le c t io n b e

“ a d e a r b i t r a r i l y . In t h a t c a s e , i t i s o b v io u s ly h e e d le s s of
D u e P r o c e s s .

Or a legislature may find the articulation of general prin
ciples too burdensome, at the same time that it opes an
expects its jurors to act rationally-to develop death-
sentencing patterns that are regular and non-arbitra y.
Its expectation may or may not be fulfilled: from the sum
of ad hoc life-or-death decisions, a pattern may or may
emerge. I f it does not, the legislature’s product—whatever
its intendment—is again heedless of Due Process.

But if a rational pattern does emerge in the ad hoc selec
tive decisions of particular juries, it will be because there
exist rational general principles of differentiation to which
the juries have more or less conformed. Such rational
general principles, if they exist, must be e^Press^ le’
finable, describable in legislation. It is hardly to be sup
posed that what a random, shuttling corps of legally un
trained jurors can appreciate and act upon is not suscepti
ble of formulation by a legislative body. I f there are sorts
or classes of cases in which the death penalty is not gen
erally imposed, these can be identified and described; and
the use of the penalty in such cases disallowed. I f there are
rather limited sorts of cases in which the penalty generally
is imposed, these can be enumerated, circumscribed, and
the penalty limited to them.” I f these things cannot be

^As^suggested in the text at pp. 38-45 supra, there are a con-
■aprahle number of ways in which a legislature might choose to

d e S the death-sentencing discretion of jurors^ It m g ht condi-

da , SofoTidant’s likelihood of recidivism, his reformability, etc.),
to the defendanti i k e ™ a ° tion to ’enumerated aggravating
It might direct} • „Jht direct the jury’s attention to enumerated

r l k e t i d i n g of specifed .ire™ -

p u re lV 'lrb itra ^ ^ f4** ‘“ f ' of ,he de>«> P -a lty i8
bejone beca J o( lh‘ % £ £ + £ . “ j ^

m akin g^n T h op^n or ite.tWeeD rational decision-

For even if patterns of rational sentencing choice emertre
wrtlnn ; system, of wholly unfettered d i s c r L n T Z s "
th n f h • a r a c t e n s t l c a n d c o n s e q u e n c e o f s u c h a s y s t e m
h a t c h o ic e s a r e a l lo w e d a n d w i l l b e m a d e , in p a r t i c u l a r

c a s e s , m u t t e r d i s r e g a r d o f th e p a t t e r n Art ? • •

bo a „ d m a y s e n t e n c e a d e f e n d a n t to d ie b e c a m e " s to o

y o u n g , 01 to o o ld , o r to o s ic k , o r to o h e a l t h v n r + u
Hons or too black. With human life at etake ’̂ a systemttat'
permits this result also is heedless of Due Process.

Concededly, the goals of sentencing are comDlex ■
designing devices for achieving them th« f i Pl ’ d m
some tolerance under the C o Z i ^ l L i L ™ ' W
petitioner ManweU’s sentence was submitted

~ ass “ 32 * ~
or aspects o f petitioner's conduct in relation to " e h p„“

It might fix governing rules criteria CT lderati°ns.
lesser generality. It might identifv I P clpl®s> of greater or
ment which should inform the de^lnn purP°?es of capital punish-
might use a c o m b in X n oil ihtlo a X ™ s°f *“ b'riterta. Or it
Penal Code. See note 66 supra. PP M ^es’ as does the Model

6 5

poses. They were not required or invited to consider the
extent of physical harm to the prosecutrix, the moral
heinousness of the defendant’s acts, his susceptibility or
lack of susceptibility to reformation, or even the appropri
ateness of the deterrent effect of killing this defendant (as
distinguished from any other) “pour decourager les
autres. Cf. Packer, Making the Punishment Fit the
Crime, 77 Habv. L. Rev. 1071, 1077 (1964). They were per
mitted to choose between life and death for any reason
rational or irrational, or for no reason at all— on an im
pulse, a surge of hatred, a vague distaste for petitioner or

s color. In making the determination to impose the
death sentence, they acted wilfully and unreviewably
without guidance and without controls. Nothing therefore
assured that there would be the slightest thread of connec
tion between the sentence they exacted and any reason
able justification for exacting it. Cf. Skinner v. Oklahoma,
supra. To concede the complexity and interrelation of
sentencing goals, see Packer, supra, is no reason to sus
tain a procedure which ignores them all. It is futile to put
forward justification for a death so inflicted; there is no
assurance that the infliction responds to the justification
or will conform to it in operation. Inevitably, under Ar
kansas’ death-sentencing practice, capital punishment in
the few, arbitrarily selected cases where it is applied is
both irrational and lawless. Thus it denies Due Process
to the men condemned to die.

II.

Arkansas’ Single-Verdict Procedure for the Trial of
Capital Cases Violates the Constitution.

Arkansas- practice of submitting simultaneously to the
tnal j „ ry the two i8suea o f guilt aild punishmenl y
.ta case compounds the vice o f law,egs . dig “ P

* ‘ r US“ '1’- ^ maki" g ‘ l virtaal|y impossible for the
Unde. A r k " 0186 " diSCretio” “ rational fashion."
Unde. Arkansas procedure, the jury hears evidence si-

resolves°both ° ” T ” ” °f ^ S" d p“ ” ishmeut, and
cedure there 68 ° SmgIe sitling- UndCT UA pro-eclure there ls no separate hearing on penalty anart
ton. the criminal trial itself, and L o p Z u n i t y for
locution or for the presentation of evidence in mitiga-

tion of sentence after the finding of guilt but prior to the
life-death sentencing choice.69

n..d™f ttr k tiT c e“u’rtbS ”u r k i , °M th“* " n o ™
But the court d i d n o t , * „ ? ? « * * « *™1 ”, ( A 68.)
intentional bypassing r e a S S U? ted. to the sort of
against federal habeas corDus See COIlstltu^ onal claims as

(A. Joh4 l j :kansas procedure is thus described by the district court

testify in his own behalf I f hP ° r may not

a S n s T h i T l n l S a m S S E S

iio n rf gufiht\ndlitbhrati0nt-0f thf6 jUry reJate to bo^hThe

theri? iS n° P°St-the defendant should receive ” 7 punishment which

The effect of this method of proceeding is obvious, and
devastatingly prejudicial to the accused. He is whipsawed
between his rights of allocution and to present evidence
to support a rational sentencing decision, and his privilege
against self-incrimination. I f he wishes personally to
address the jurors with respect to their decision whether
he shall live or die, he can do so only at the price of
taking the stand and thereby surrendering his privilege.70
He is subject not only to incriminating cross-examination
but also to impeachment71—a process which, in Arkansas,
involves the introduction by the prosecution of an uncom
monly prejudicial lot of ordure.72 If he exercises the

70 See note 69 supra.

71 When a criminal defendant “voluntarily takes the stand as a
witness in his own behalf he is subject to the same cross-examination
to which any other witness might be subjected.” Dillon v. State,
222 Ark. 435, 442, 261 S.W.2d 269, 273 (1953). For the applica
tion of the rule in capital trials, see Wright v. State, 243 Ark. 221,
419 S.W.2d 320 (1967); McGuire v. State, 189 Ark. 503, 74 S.W.2d
235 (1934). “ ‘This court had adopted the rule that witnesses,
including the accused, may be impeached on cross examination by
drawing out the fact that they have committed other crimes and
immoralities.’ ” Skaggs v. State, 234 Ark. 510, 511, 353 S.W .2d 3,
4 (1962) (conviction reversed on other grounds). Accord: Ray
burn v. State, 200 Ark. 914, 141 S.W .2d 532 (1940) (capital case).

72 As the Skaggs case, note 71 supra, indicates, the Arkansas law
relating to impeachment permits cross examination regarding not
merely prior convictions, but prior criminal acts not resulting in
convictions, and other “immoralities.” Wright v. State, 243 Ark.
221, 419 S.W.2d 320 (1967) (rape defendant asked whether several
persons had not told him to quit hanging around their places of
business because he made indecent proposals to women); Edens v.
State, 235 Ark. 178, 359 S.W .2d 432 (1962) (defendant in false
pretenses case asked whether he had not defrauded each of several
named persons); Willis v. State, 220 Ark. 965, 251 S.W .2d 816
(1952) (homicide defendant asked whether he did not once get
drunk and have a wreck and kill a boy) ; Powell v. State, 149 Ark.
311, 232 S.W. 429 (1921) (defendant in carnal abuse case asked
whether he had not cohabited with his wife before they were mar
ried ).

The sorts of misconduct which may thus be brought out are ap
parently limitless. They need have no real relation to the trait of

privilege, on the other hand, he risks an uninformed, arbi
trary, and uncompassionate death verdict. Should he wish
to present background and character evidence to inform
the jury’s sentencing choice, he may do so only at the cost

truthfulness. See, in addition to the cases cited above, the following
exemplary cases involving capital trials: Black v. State, 215 Ark.
618, 222 S.W.2d 816 (1949) (convictions for “ certain misdemean
ors,” including drunk driving, elicited); Edwards v. State, 208
Ark. 231, 185 S.W .2d 556 (1945) (conviction and fine on a liquor
charge elicited); Bieard v. State, 189 Ark. 217, 72 S.W.2d 530
(1934) (general reputation of the defendant three or four years
prior to the offense adduced); Curtis v. State, 188 Ark. 36, 64 S.W.
2d 86 (1933) (involvement in illicit relationship with a named
woman elicited). Although the doctrine is that misconduct which
is too remote in time may not be elicited, at least if it does not
involve a conviction, Dixon v. State, 189 Ark. 812, 75 S.W.2d 242
(1934) (20-year old homicide), a 12- to 14-year old drunk-driving
accident is not too remote, Willis v. State, supra; nor is a 19-year
old juvenile conviction for auto theft, Dillon v. State, 222 Ark. 435,
261 S.W .2d 269 (1953), nor a 20-year old fine on a liquor charge,
Edwards v. State, supra. Impeachment relating to general miscon
duct or specific bad acts is limited to cross examination, but prior
convictions may be proved almnde. Holcomb v. State, 218 Ark. 608,
238 S.W.2d 505 (1951). To the extent that Ark. Stat. Ann. §28-
707 (1962 Repl. Vol.) may seem more restrictive than the rules
described in this footnote, “ [t] he cases do not appear to have held
to the strict language of the statute.” Id. at 612, 238 S .W .2d at 507.

The Arkansas court has shown no disposition to restrict the exhi
bition of prior misconduct for the purposes of impeachment even
where it is of such a nature as to be devastatingly prejudicial with
regard to the crime charged. See the following cases regarding
capital trials: Wright v. State, supra (rape defendant asked about
instances when he had been told to quit hanging around places
because he made indecent proposals to women); Bevis v. State, 209
Ark. 624, 192 S.W .2d 113 (1946) (homicide defendant asked if he
did not shoot his first w ife ); Gaines v. State, 208 Ark. 293, 186
S.W.2d 154 (1945) (homicide defendant asked whether he had not
previously shot his brother-in-law and a neighbor, nearly killing
them ); Fielder v. State, 206 Ark. 511, 176 S.W.2d 233 (1943)
(homicide defendant asked whether he did not shoot two other men
to death) ; Ham v. State, 179 Ark. 20, 13 S.W.2d 805 (1929) (homi
cide defendant asked whether he had not killed a man a few days
ago) ; McGraw v. State, 184 Ark. 342, 42 S.W.2d 373 (1931) (defen
dant, charged with wife’s homicide, asked whether he had not
previously killed his wife’s step-father).

of opening the question of character generally prior to
the determination of guilt or innocence, thereby risking
the receipt of bad-character evidence13 ordinarily ex
cludable because highly prejudicial on the guilt question.14
Or he may avoid that risk of prejudice by confining the
evidence at trial to matters relevant to guilt, letting the
jury sentence him to life or death in ignorance of his
character.

A procedure of this sort in unconstitutional, both be
cause it results in a fundamentally unfair trial and be
cause it infringes the several federal constitutional rights
which it sets at loggerheads. To appreciate why this is

73 Arkansas follows the ordinary rules permitting the cross exam
ination of a defense character witness in the “have you heard” form
which permits the eliciting of every prejudicial occurrence, real or
imagined, in the defendant’s history. See Amos v. State, 209 Ark.
55, 189 S .W .2d 611 (1945); Clark v. State, 135 Ark. 569, 205 S.W.
975 (1918). The prosecution may also elicit the defendant’s bad
reputation in regards to traits broader than those which the defen
dant’s character witness bolstered. See Weakley v. State, 168 Ark.
1087, 273 S.W . 374 (1925).

74 In Arkansas, as elsewhere, the prosecution is forbidden to open
the question of the defendant’s character, or to develop bad-
character evidence in his case-in-chief. See, e.g., Long v. State, 240
Ark. 687, 401 S.W.2d 578 (1966); Browning v. State, 233 Ark. 607,
346 S.W.2d 210 (1961); Rand v. State, 232 Ark. 909, 341 S.W.2d
9 (1960). Unless the defendant testifies or opens the character
question, evidence of prior crimes is inadmissible. Bonds v. State,
240 Ark. 908, 403 S.W.2d 52 (1 966); Rhea v. State, 226 Ark. 664,
291 S.W.2d 521 (1956). Of course, Arkansas has the usual excep
tions relating to proof of intent, motive, modus operandi, etc., e.g.,
Osborne v. State, 237 Ark. 5, 371 S.W .2d 518 (1963); but these
rules were exhaustively considered in Alford v. State, 223 Ark. 330,
266 S.W.2d 804 (1954), and there delimited in such a way as to
make prior crime evidence inadmissible in the ordinary rape case,
Under Alford, prior-crime evidence would also be inadmissible—
again, unless the defendant takes the stand or opens the character
question— in the ordinary murder case. Moore v. State, 227 Ark.
544, 299 S.W.2d 838 (1957). Compare Ward v. State, 236 Ark. 878,
370 S.W .2d 425 (1963) (sexual perversion) ; Gerlach v. State, 217
Ark. 102, 229 S.W.2d 37 (1950) (assault with intent to rape).

so, one must begin by recognizing what this Court has
several times said, that much evidence which is not rele
vant to the issue of guilt of the charge for which the
capital accused is on trial—evidence which, indeed, is
prejudicial and inadmissible on the issue of guilt—is highly
relevant to a non-arbitrary decision on the question of
punishment. “ [MJodern concepts individualizing punish
ment have made it all the more necessary that a sentenc
ing judge not be denied an opportunity to obtain pertinent
information by a requirement of rigid adherence to re
strictive rules of evidence properly applicable to the trial.”
Williams v. New York, 337 U.S. 241, 247 (1949); see also
Williams v. Oklahoma, 358 U.S. 576, 585 (1959); Wither
spoon v. Illinois, 391 U.S. 510, 521 n. 20 (1968). A fortiori,
a jury engaged in the task of determining whether a defen
dant shall live or die needs much information that cannot
and should not be put before it within the confines of tradi
tional and proper limitations on the proof allowable as
going to guilt or innocence. It is fair to say that the over
whelming weight of considered contemporary judgment
concurs in the conclusion that, whether discretionary death
sentencing be done by a judge or jury, it is the imperative
condition of rational sentencing choice that the sentencer
consider more information about the individual defendant
than is likely or permissibly forthcoming on trial of the
guilt issue. E.g., H ouse of Commons Select Committee
on Capital P unishment, R eport (H.M.S.O. 1930), para.
177; R oyal Commission on Capital P unishment, 1949-
1953, R eport (H.M.S.O. 1953) (Cmd. No. 8932), 6, 12-13,
195, 201, 207; A merican Law I nstitute, Model P enal
Code, Tent. Draft. No. 9 (May 8, 1959), Comment to §201.6
at 74-76; New Y ork State T emporary Commission on R e
vision of the P enal L aw and Criminal Code, Interim
R eport (Leg. Doc. 1963, No. 8) (February 1, 1963), 15-16;
H.L.A. Hart, Murder and the Principles of Punishment:

England and the United States, 52 Nw. U.L.Rev. 433, 438-
439 (1957); Knowlton, Problems of Jury Discretion in
Capital Cases, 101 U.Pa. L.Rev. 1099, 1109, 1135-1136
(1953); Handler, Background Evidence in Murder Cases,
51 J. Crim.L., Crim. & Pol. Sci. 317, 321-327 (1960).

The single-verdict procedure therefore confronted peti
tioner, on trial for his life, with a gruesome Hobson’s
choice:

Petitioner had a crucial interest—amounting, indeed, to
an independent federal constitutional right, see Skinner v.
Oklahoma, 316 U.S. 535 (1942)—that his sentence be ra
tionally determined. The Constitution guaranteed him,
also, certain procedural rights in this sentencing process:
inter alia, “an opportunity to be heard . . . and to offer evi
dence of his own.” Specht v. Patterson, 386 U.S. 605, 610
(1967); Mempa v. Rhay, 389 U.S. 128 (1967). As the basis
for a rational sentencing determination, he would want to
present to the sentencing jurors evidence of his history,
his character, his motivation, and the events leading up
to his commission of the crime of which he was guilty (if
he was guilty). The common-law gave him a right of allo
cution which is an effective vehicle for this purpose, as well
as for a personal appeal to the jurors, where capital
sentencing is discretionary. This Court has already recog
nized that allocution may in some circumstances rise to the
dignity of a due process command.76

w ffflt v. United States, 368 U.S. 424, 428-429 (1 962); see
Andrews v. United States, 373 U.S. 334, 336-337 (1963). Acting
on these opinions, several circuit courts have a
constitutional right. Green v. United States, 313 F.2d 6 (1st Cir.
1963), cert, denied 372 U.S. 951 (1963) ; United Statesx. Johnson,
315 F.2d 714 (2d Cir. 1963), cert, denied, 375 U.S. 971 (1964) ,
Behrens v. United States, 312 F.2d 223 (7th Cir. 1962), o ff d, 375
U.S. 162 (1963). . „ . ,

The Hill case holds that, absent “ aggravating circumstances
(368 U.S. at 429), failure of a sentencing judge to invite tHe

But to exercise his right of allocution before verdict on
the guilt issue, petitioner was required to forego his con
stitutional privilege against self-incrimination. Malloy v.
Hogan, 378 U.S. 1 (1964); Griffin v. California, 380 U.S.
609 (1965). He was required to take the stand and be sub
jected to cross-examination that could incriminate him.76

defendant to exercise his right of allocution does not violate due
process. But Hill was not a capital case, nor a case in which
counsel was denied the right to present pertinent information to
the sentencer, nor “a case where the defendant was affirmatively
denied an opportunity to speak during the hearing at which his
sentence was imposed . . . [nor where the sentencer] was either
misinformed or uninformed as to any relevant circumstances . . .
[nor where it was even claimed] that the defendant would have
had anything at all to say if he had been formally invited to
speak.” Ibid. In an Arkansas capital case, the following “aggravat
ing circumstances” are presented: (A ) the circumstances that the
right of allocution has had unique historical significance in capital
cases, see Couch v. United States, 235 F.2d 519, 521 (D.C. Cir.
1956) (opinion of Judge F a h y ) ; Coleman v. United States, 334
F.2d 558, 566 (D.C. Cir. 1964) (opinion of Judges Burger and
McGowan); (B ) the circumstance that in capital cases matters
which may affect the sensitive discretion of the jury in its life-
death choice are traditionally viewed with a stricter eye to possi
bilities of prejudice than are other matters in the criminal process,
see People v. Hines, 61 Cal. 2d 164, 390 P.2d 398, 37 Cal. Rptr. 622
(1964); (C) the circumstance that Arkansas’ single verdict pro
cedure “affirmatively” denies a defendant his opportunity to address
the jury on sentence, within the meaning of Hill, supra; (D ) the
circumstance that, for the reasons set out above in text, a jury
making a capital sentencing choice on no other information than the
trial of the guilt issue allows is invariably “either misinformed or
uninformed” within the meaning of H ill; (E ) the circumstance
that, under the Arkansas procedure, not only is the defendant de
nied the right to make a personal statement without giving up his
constitutional privilege against self-incrimination, but he is simi
larly denied the right to have his counsel supply evidence on the
sentencing issue without incurring the risks of prejudice on the
guilt issue, see the Johnson and Behrens cases cited, supra; also
Gadsden v. United States, 223 F.2d 627 (D.C. Cir. 1955); Jenkins
v. United States, 249 F.2d 105 (D.C. Cir. 1957).

76 Even apart from cross-examination, allocution before verdict
of guilt destroys the privilege, for much of the value of the defen
dant’s personal statement to his sentencer derives from its spon-

He declined to surrender the Privilege, and did not address
the jury. (A. 41.) The jurors who sentenced him to die
therefore had neither heard his case for mercy, nor even
the sound of his voice.

United States v. Jackson, 390 U.S. 570 (1968), holds that
the exercise of the Fifth Amendment privilege in criminal
trials may not be penalized or needlessly burdened. Simul
taneous submission to the jury of the guilt and death
penalty issues results in just such a needless burden. The
burden is the obvious and onerous one that the defendant
must go to his death like a dumb animal, forbidden to plead
his case and incapable even of impressing the human qual
ity of his voice on his sentencing jury. It is a needless
burden because the State has ample means to avoid it by,
for example, a bifurcated jury trial, judge sentencing, or
the elimination of the death penalty. Consistently with
Jackson, Simmons v. United States, 390 U.S. 377, 394
(1968), points the way to condemnation of the “ undeniable
tension” between constitutional rights presented here. The
question in Simmons was whether a defendant might be
obliged either to give up what he believed to be a valid
Fourth Amendment claim or “ in legal effect, to waive his
Fifth Amendment privilege against self-incrimination” ;
and this Court there held it “ intolerable that one constitu
tional right should have to be surrendered in order to
assert another.” Ibid.

But the single-verdict practice which petitioner challenges
presents grave problems in addition to this conflict between
the right of allocution and the privilege. If the defendant * 570

taneity, see Green v. United States, 365 U.S. 301, 304 (1961)
(opinion of Mr. Justice Frankfurter). This same spontaneity—
unguided by the questions of counsel— leaves the defendant im
permissibly unprotected as he appears before a jury which has
not yet decided on his guilt. Cf. Ferguson v. Georgia, 365 U.S.
570 (1961).

seeks to present to the jury evidence of his background
and character, apart from his own statement, the prosecu
tion may counter with evidence of the defendant’s bad char
acter, including evidence of unrelated crimes. The prohibi
tion which ordinarily keeps this sort of evidence from the
trial jury sitting to determine the issue of guilt is “ one of
the most fundamental notions known to our law,” United
States v. Beno, 324 F.2d 582, (2d Cir. 1963), arising “ out
of the fundamental demand for justice and fairness which
lies at the basis of our jurisprudence,” Lovely v. United
States, 169 F.2d 386, 389 (4th Cir. 1948); see Marshall v.
United States, 360 U.S. 310 (1959). Allowing the trial jury
access to unfavorable background information, however
pertinent to the issue of punishment, and however clearly
limited by jury instructions to that use, may itself amount
to a denial of due process of law. Compare United States
ex rel. Scoleri v. Banmiller, 310 F.2d 720 (3d Cir. 1962),
cert, denied, 374 U.S. 828 (1963), with United States ex rel.
Rucker v. Myers, 311 F.2d 311 (3d Cir. 1962), cert, denied,
374 U.S. 844 (1963). In any event, the possibility that the
background information will be strongly prejudicial forces
a defendant to a “ choice between a method which threatens
the fairness of the trail of guilt or innocence and one which
detracts from the rationality of the determination of the
sentence.” A merican L aw’ I nstitute, Model P enal Code,
supra at 64.

In this aspect, the practice of simultaneous submission of
guilt and sentence to a jury is much akin to New York’s
former practice of simultaneously submitting to a jury the
two issues of the voluntary nature of a confession and the
guilt of the accused. The New York practice wTas, of course,
struck down in Jackson v. Denno, 378 U.S. 368 (1964),
where the Court recognized that joint jury trial of the two
issues prevented either from being “ fairly and reliably

determined.” 378 U.S. at 389; see id. at 386-391. One fault
of the practice was that:

“ . . . an accused may well be deterred from testifying
on the voluntariness issue when the jury is present
because of his vulnerability to impeachment by proof
of prior convictions and broad cross-examination. . . .
Where this occurs the determination of voluntariness
is made upon less than all of the relevant evidence.
(378 U.S. at 389 n. 16).77 *

77 To the extent that petitioner’s argument on this branch of
his case draws sustenance from the radiations of the rule of Jack-
son v. Denno, text, supra, his argument is not weakened by Spencer
v. Texas, 385 U.S. 554 (1967), cited by the court of appeals.
The Spencer decision dealt with the Texas procedure for determin
ing the punishment decreed under its recidivist statutes, pursuant
to which a jury trying a criminal defendant was informed by a
reading of the indictment that the defendant had been convicted
of prior crimes, and evidence was introduced in support of these
allegations during the course of trial. The procedure was attacked
on the ground that it unduly prejudiced the defendant on the
issue of guilt, in violation of the Due Process Clause. In rejecting
this attack, the Court held, inter alia, that Jackson v. Denno did
not require a two-stage jury trial for invocation of a state s ha
bitual offender statute and that “the emphasis [in Jackson] was
on protection of a specific constitutional right and the Jackson
procedure was designed as a specific remedy to ensure that an in
voluntary confession was not in fact relied upon by the jury, 385
U.S., at 565. . , ,

But no more “specific constitutional right can be imagined than
that which prohibits the state’s taking of petitioner’s life pursuant
to a sentencing process which is irrational because sentencing
choice is made without information essential to the choice, Skinner
v. Oklahoma, supra; Specht v. Patterson, sxipra. And though the
Court in Spencer read Jackson as not broadly requiring a two-
stage jury proceeding whenever evidence of prior crimes which
might possibly prejudice the jury’s determination of the guilt
issue is presented at the trial (because of the justifications gen
erally accepted for permitting such evidence— see 385 U.S., at
560-61), Jackson’s holding that “a general jury verdict [is] not
a reliable vehicle for determining the issue of voluntariness [of a
confession] because jurors might have difficulty in separating the
issue of voluntariness from that of guilt or innocence, 385 U.S. at
565, is unimpaired. See Burgett v. Texas, 389 U.S. 109 (1967):

And see Bruton v. United States, 391 U.S. 123 (1968).

It is in precisely the same manner that single-verdict
capital sentencing tends either to make trials of guilt
unfair—by forcing the defense to present evidence poten
tially helpful on the punishment issue and prejudicial on
the issue of guilt— or to produce the unfair result that
men are sentenced to death “upon less than all of the rele
vant evidence” if the defense declines to take that risk.78

Simmons v. United States, text, supra. That, by analogy, applies
with equal, if not greater force to the jury’s determination of the
issues of guilt and sentence in a capital case. The “specific remedy”
for reliably determining these issues enjoined on the courts by
Jackson, reinforced by Sims v. Georgia, 385 U.S. 538 (1967), and
endorsed by Spencer is that of separate consideration of the issues.

Moreover, there are crucial differences between the situation pre
sented here and that presented by the practice challenged in
Spencer. First, the sole claim urged by the petitioners in Spencer
was that the introduction of prior crime evidence before jury
verdict on the guilt issue unduly prejudiced the defendant on that
issue (385 U.S. at 559, 567) and was therefore unconstitutional
per se. Petitioner’s submission here differs. The effect of the single
verdict sentencing procedure which he challenges is to confront a
capital defendant with the grim specter of having to sacrifice one
or another of his precious constitutional rights, either allocution
or self-incrimination. By whatever choice he makes, he is preju
diced on either the issue of guilt or sentence; and he must decide—-
with his life at stake—-which kind of prejudice to invite. Cf.
Whit us v. Balkcom, note 78, infra. Also, while in Spencer the
juries were given limiting instructions to guide their considera
tion of the prior crime evidence on the guilt issue, the jury’s dis
cretion in sentencing choice under Arkansas law is not limited or
regulated in any way. See pp. 24-45, supra. Hence there is more
than “the possibility of some collateral prejudice” (385 U.S. at 564)
which the Court in Spencer thought not of sufficient gravity to
condemn the procedure there attacked.

Finally, the Arkansas procedure has been shown on this record
to have resulted in a pattern of racially discriminatory death sen
tencing. This goes far toward satisfying the concern of the Spencer
court over the lack of a convincing showing of prejudice.

78 Such a “ ‘grisly,’ hard, Hobson’s choice” is itself so unfair as
to violate due process. See Whitus v. Balkcom, 333 F.2d 496, 499
(5th Cir. 1964).

This latter alternative was the course of Maxwell’s trial,
and its result. Not only, in such a case, is the jury em
powered to act arbitrarily, see pp. 24-65, supra; it is
virtually compelled to do so for want of information upon
which nonarbitrary choice can be based. In short, although
the Due Process Clause guaranteed Maxwell a fair trial
on the issue of punishment, e.g., Townsend v. Burke, 334
U.S. 736 (1948), as well as on the issue of guilt, e.g., Irvin
v. Dowd, 366 U.S. 717 (1961), the single-verdict procedure
employed in his case required him to purchase the second
of these at the cost of the first. Cf. Fay v. Noia, 372 U.S.
391, 440 (1963).

As the question in United States v. Jackson, 390 U.S.
570 (1968), was whether the provision of the federal kid
naping statute reserving the infliction of the death sentence
to the exclusive province of the jury “needlessly en
courages” guilty pleas and jury waivers and therefore
“needlessly chill[s] the exercise of basic constitutional
rights,” 390 U.S., at 582, 583, so the question here is
whether the simultaneous trial of guilt and punishment
needlessly encourages the waiver of the right to remain
silent or needlessly chills the right to put in evidence rele
vant to rational sentencing and the right of allocution.
“ The question is not whether the chilling effect is ‘inci
dental’ rather than intentional; the question is whether
that effect is unnecessary and therefore excessive.” 390
U.S., at 582. We submit the answer to the question is
clear, in light of the ready availability of alternative modes
of procedure not involving the same destructive collision
of the defendant’s rights— such as the split-verdict proce
dure now in use in a number of jurisdictions19 and uni- 79

79 A man tried on a capital charge in California, Connecticut,
New York, Pennsylvania and Texas receives first a hearing on in
nocence or guilt; then, after a finding of guilt, a separate hearing

formly recommended by modern commentators, see Frady
v. United States, 348 F.2d 84, 91 n. 1 (D.C. Cir, 1965)
(McGowan, J . ) ; cf. United States v. Gurry, 358 F.2d 904,
914 (2d Cir. 1965).80

on the issue of punishment. Cal. Penal Code, §190.1 (Supp. 1966);
Conn. Gen. Stat. Rev. §53-10 (Supp. 1965); N .Y. Pen. Law
§§125.30,125.35 (Cum. Supp. 1968); Pa. Stat. Ann., tit. 18, §4701
(1963) ; Tex. Code Crim. Pro., Art. 37.07 (1967). See also §210.6
of the Model Penal Code, note 66, supra. In view of the availability
of this alternative “split-verdict” mode of procedure which does
not entail the deprivation of petitioner’s rights, the constitutionally
unfair single-verdict procedure used in petitioner’s case can obvi-
ously not be defended on the ground of necessity. Petitioner, of
course, does not contend that the State is constitutionally compelled
to have a bifurcated trial. The bifurcated trial is only one of the
alternatives available to the State which do not entail the needless
burden on capital defendants’ constitutional rights here complained
of.

80 Cf Witherspoon v. Illinois, 391 U.S. 510, 520 n. 18 (1968),
noting that if a defendant could establish that a jury which was
representative with respect to penalty was not representative with
respect to guilt,

“the question would then arise whether the State’s interest in
submitting the penalty issue to a jury capable of imposing
capital punishment may be vindicated at the expense of the
defendant’s interest in a completely fair determination of guilt
or innocence— given the possibility of accommodating both in
terests by means of a bifurcated trial, using one jury to decide
guilt and another to fix punishment.”

CONCLUSION

Petitioner’s trial jury was permitted lawless and arbi
trary power to decide whether he should live or die, in
violation of the rule of law that is fundamental to the
Fourteenth Amendment. Arkansas’ single-verdict capital
trial procedure assured that this arbitrary power would
be used arbitrarily, by depriving the jury of information
requisite to rational sentencing choice, as a consequence ot
petitioner’s exercise of his Fifth and Fourteenth Amend
ment privilege against self-incrimination. The use of these
two procedures at his trial deprived petitioner of his life
without due process of law. The judgment below should
be reversed.

Respectfully submitted,

Jack Greenbebg
James M. Nabeit, III
Norm an C. A maker
Michael Meltsner
J ack H immelstein
E lizabeth DuB ois

10 Columbus Circle
New York, New York 10019

George H oward, J r.
329V2 Main Street
Pine Bluff, Arkansas 71601

A nthony G. A msterdam
3400 Chestnut Street
Philadelphia, Pennsylvania 19104

Attorneys for Petitioner

APPENDIX A

Evidence and Findings Below Relating to Racial Dis
crimination by Arkansas Juries in the Exercise of Tlieir
Discretion to Sentence Capitally for the Crime of Rape.

A. The Evidence of the Wolfgang Study.

Petitioner’s second federal habeas corpus petition, giving
rise to the proceedings now before the Court, alleged that
new evidence had become available with respect to his
claim of racial discrimination in capital sentencing, pre
viously urged at his trial, on his direct appeal, and in his
initial federal habeas application. It averred, specifically,
that a systematic study of Arkansas rape convictions dur
ing a twenty-year period had been:

“ conducted in the summer of 1965, as part of a study
of the application of the death penalty for rape in
eleven southern states. This comprehensive study re
quired the work of 28 law students throughout the
summer, the expenditure of more than $35,000 and
numerous hours of consultative time by expert crimi
nal lawyers, criminologists and statisticians. Petitioner,
who is an indigent, could not have himself at any time
during the prior proceedings in his cause conducted
such a study.” (A. 6, quoted by the court of appeals,
at A. 48.)

At a pre-trial conference, the district court was advised
that petitioner intended to present at an evidentiary hear
ing the results of this comprehensive study. Its pre-trial
conference order reflected that petitioner’s evidence would
consist in part of “ the testimony of Dr. Marvin E. W olf
gang, a criminologist and statistician on the faculty of the
University of Pennsylvania, and . . . certain studies and

a report made by Professor Wolfgang,” which in turn were
based upon “ [b]asic data . . . gathered by law student
field workers from various sources and . . . recorded on
individual case schedules.” (A. 17-18.) Accordingly, the
order provided for procedures to facilitate the establish
ment of “ the validity and accuracy of the individual case
schedules.” (A. 18.)la

“ It was agreed that counsel for Maxwell will make
those schedules available for the inspection of counsel
for Respondent not later than August 10 and will also
furnish the names and addresses of the field workers
who assembled the original data in Arkansas. Not
later than August 15 counsel for Respondent will ad
vise opposing counsel and the Court as to whether,
to what extent, and on what grounds he questions any
individual case schedule.

“ Subject to objections on the ground of relevancy
and materiality, and subject to challenges to individual
case schedules, Professor Wolfgang will he permitted
to testify as an expert witness and to introduce his

A p p en d ix A

la The “ individual case schedules” referred to are the completed
forms, for each case of conviction of rape, of the printed schedule
captioned “ Capital Punishment Survey” admitted as Petitioners
Exhibit P-2 (Tr. 57). [Tr. — references in this Appendix are to
the original transcript of the district court proceedings.] The use
of this printed schedule in the process of data-gathering was ex
plained by Dr. Wolfgang at Tr. 22-25. Instructions given the
field researchers in use of the schedule are included in the record
as an exhibit, Petitioner’s Exhibit P-3 (Tr. 25-27, 57) but, in
view of the respondent’s concession that the facts gathered by
the researchers were accurate, see text infra, no effort was made
in the testimony to demonstrate the steps taken in gathering the
data to assure reliability. See Tr. 25-27. Also, in hght of the
court’s pre-trial conference order, text, immediately infra, the
completed “ individual case schedules” were not introduced in
evidence.

report as a summary exhibit reflecting and illustrating
his opinions. Again subject to objections or challenges
to individual schedules there will be no occasion for
Petitioner to introduce the schedules in evidence or
prove the sources of the information reflected thereon
or therein, or to call the individual field workers as
witnesses.” (A. 18.)

When the case came on for hearing, counsel for peti
tioner announced that no objections had been filed by the
respondent to any of the individual case schedules, so that
“ all of the facts in the schedules are treated as though
they are true, and Dr. Wolfgang’s testimony is to be treated
as though based not on schedules, but on facts which are
established of record . . . As I understand it, the basic facts
on which Dr. Wolfgang’s testimony and his analysis are
made are treated as established for the purpose o f this
case.” (Tr. 8.) Counsel for respondent and the court
agreed with this statement (Tr. 8-9), the court settling that:

“ The basic facts—that is, the age of the victim, the
race, and so on, of the individual defendants, or the
alleged victims—the basic evidentiary facts, as the
Court understands it, stand admitted, and that Dr.
Wolfgang in testifying, or anybody elso who testifies
about these basic figures, will not be faced with an
objection as to the authenticity of his basic data.”
(Tr. 9 .)2a

On this understanding, Dr. Marvin E. Wolfgang was
called as a witness for petitioner. In its written opinion,
the district court termed him a “well qualified sociologist

2a The Court of Appeals accepted this procedure without ques
tion. (A. 48.)

A p p en d ix A

and criminologist on the faculty of the University of Penn
sylvania” and noted that his “qualifications to testify as
an expert are not questioned and are established” (257
F. Supp. at 717-718; A. 33).3a (Similarly, the Court of
Appeals was later to find that Dr. Wolfgang “ obviously
is a man of scholastic achievement and of experience in
his field,” whose “ ‘qualifications as a criminologist have
[concededly] never been questioned by the respondent.’ ”
(398 F.2d at 141; A. 49.)) Dr. Wolfgang’s testimony occu
pies some ninety pages of the transcript of the hearing
(Tr. 10-99); in addition, “ a written report prepared by
him, together with certain other relevant documentary ma
terial, was received in evidence without objection” (257 F.
Supp. at 717-718; A. 33-34). The written report referred
to, Petitioner’s Exhibit P-4, was received as substantive
evidence (Tr. 57), and will be relied upon together with
Dr. Wolfgang’s testimony in the summary of evidence that
follows.

The district court’s opinion fairly summarizes the “ back
ground facts of the Wolfgang study” :4a

“In early 1965 Dr. Wolfgang was engaged by the
NAACP Legal Defense and Educational Fund, Inc., to

A p p en d ix A

3a Interrogation of Dr. Wolfgang establishing his qualifications is
at Tr. 10-14. In addition, Petitioner’s Exhibit P-1, a curriculum
vitae of Dr. Wolfgang, was received in evidence to establish his
qualifications, Tr. 14-16. Dr. Wolfgang is one of the foremost
criminologists in the country.

4a The general scope of the study, which gathered data concern
ing every case of conviction for rape during a 20-year period in
250 counties in eleven States, is described more fully in the- affidavit
of Norman C. Amaker, Esq., attached as Exhibit 2 to the petition
for habeas corpus. For other descriptions, see the Memorandum
and Order, dated July 18, 1966, appended to the opinion in Moorer
v. South Carolina, 368 F.2d 458 (4th Cir. 1966) ; and the opinion
in Matter of Sims and Abrams, 389 F.2d 148 (5th Cir. 1967).

make a study of rape convictions in a number of south
ern States, including Arkansas, to prove or disprove
the thesis that in those States the death penalty for
rape is disproportionately imposed upon Negro men
convicted of raping white women. Dr. Wolfgang was
apprised of the fact that the results of his study mig i
well be used in litigation such as the instant case.

“ As far as Arkansas is concerned, Dr. Wolfgang
caused Mr. John Monroe, a qualified statistician, to
select a representative sample of Arkansas counties
with reference to which the study would be made. Ihe
sample drawn by Mr. Monroe, who testified at the
hearing, consisted of 19 counties in the State.

“ During the summer of 1965 law students interested
in civil rights problems were sent into Arkansas to
gather basic data with respect to all rape convictions
in the sample counties for a period beginning January
1, 1945, and extending to the time of the investigation.
Data obtained as to individual cases were recorded on
individual case schedules. When the work was com
pleted, the individual schedules were turned over to
Dr. Wolfgang for evaluation.

“ The investigation brought to light 55 rape convic
tions during the study period involving 34 Negro men
and 21 white men. The offenses fell into three cate
gories, namely: rapes of white women by Negro men;
rapes of Negro women by Negro men; and rapes ot
white women by white men. No convictions of white
men for raping Negro women were found.” (257 F.
Supp. at 718; A. 34-35. See also the opinion of the
Court of Appeals, 398 F.2d at 141-142; A. 49.)

The design of the investigation was described by Dr.
Wolfgang as a function of its objectives “ to collect the

A p p en d ix A

appropriate kind of data necessary to provide some kind
of empirical study, either in support of, or in rejection of,
the underlying assumption” (Tr. 17)—i.e., that there is
racially differential imposition of the death penalty for
rape in the States studied (Tr. 16-17)— and “ to give the
empirical data the appropriate kind of statistical analysis
that would satisfy scientific requirements” (Tr. 17). The
basic research methodology involved these several stages:
(1) identification of the cases to be studied; (2) collection
of data concerning the critical variables (race of defen
dant, race of victim, sentence imposed) in each case, and
statistical analysis of the relationship between these vari
ables; (3) collection of data concerning other variables
(“control” variables) in each case, and statistical analysis
of the relationship between each such variable and the
critical variables (race and sentence) to determine whether
the operation of the control variables could explain or
account for whatever relationship might be observed be
tween the critical variables; (4) reporting of results of
the analysis. It is convenient to summarize the evidence
presented to the district court under these four heads,
with respect to the Arkansas study. Such a summary can
only imperfectly portray the character and range of the
Wolfgang study. We respectfully invite the Court’s atten
tion to the whole record of the hearing below.

1. Identification of the Cases to Be Studied.

Data were gathered concerning every case of conviction
for rape during a 20-year period (January 1, 1945 to the
summer of 1965) in a representative sample of Arkansas
counties (Tr. 21). Two points should be noted here.

First, because the study begins with cases of conviction
for rape, it addresses itself at the outset to the possibility

A p p en d ix A A p p en d ix A

suggested by the Supreme Court of Arkansas on the direct
appeal in petitioner’s case, Maxwell v. State, 236 Ark. 694,
370 S.W.2d 113 (1963), that any showing that Negroes are
more frequently sentenced to death for rape than whites
might be accounted for by the supposition that Negroes
commit rape, or are convicted of rape, more frequently
than whites. What is compared in this study is the rate
of capital sentencing of Negro and white defendants all of
whom have been convicted of rape.

Second, in order to give a valid basis for generalization
about the performance of Arkansas juries, every case of
conviction for rape in a randomly selected sample of Ar
kansas counties was included in the study (Tr. 62-63). The
county sampling procedure was employed because resources
available for the field study did not permit the gathering
of data in every county in the State (Tr. 21, 107-111), and
because it is “ unnecessary to collect every individual case,
so long as the sample is presumed to be a valid represen
tation—a valid representative one” (Tr. 21). At Dr. W olf
gang’s request, a random sample (Tr. 128) of Arkansas
75 counties -was drawn by Mr. John Monroe, a “qualified
statistician” (257 F. Supp. at 718; A. 34), with seventeen
years experience in sampling and surveys (398 F.2d at 144;
A. 53-54).6a Testifying below, Mr. Monroe described in
detail the sampling process used (Tr. 107-141) to draw
counties “ in such a manner that the sample counties within
each state would provide a representative sampling for that
state so that inferences could be drawn for each state in the
sample and for the region as a whole” (Tr. 107). Nineteen
counties in the State (Tr. 28, 118; 122-123; Petitioner’ s Ex-

6a Mr. Monroe’s qualifications appear at Tr. 104-106. His biog
raphy, in summary form, was admitted as Petitioner s Exhibit P-10
(Tr. 144-145).

hibit P-5, appendices C, D ; Petitioner’s Exhibit P-7) con
taining more than 47 per cent of the total population of
Arkansas (Petitioner’s Exhibit P-4, p. 1; Tr. 130; 398 F.2d
at 144; A. 54) were drawn by a “ theoretically unbiased
random method (Tr. 118). Mr. Monroe testified that “ a
sample is the procedure of drawing a part of a whole, and
if this sample is drawn properly according to the law of
chance, or with known probability, by examining a small
part of this whole, and using the appropriate statistical
methods, one can make valid inferences about the whole
population from examining a small part” (Tr. 116). He con
cluded that his own sample of Arkansas counties “ is a very
reliable sample under the restrictions that we were confined
to, the number of counties that could be investigated during
the time allotted. In other words, for the size of the sample,
the 19 counties, it was a very reliable and highly acceptable
sample insofar as sampling statistics are concerned (Tr.
118; see also Tr. 130, 132). “ I would say that, as far as
the sample is concerned, the inferences drawn from this
sample, as described, are valid for the State of Arkansas
(Tr. 135). See 398 F.2d at 144; A. 54.

(These conclusions were not questioned by the courts
below, although, as we shall see, both courts were con
cerned over the circumstances that Mr. Monroe’s areal
sampling methods resulted in the selection of counties that
lie principally in the southern and eastern portions of the
State. This circumstance was apparently not thought to
impugn the sample’s factual representativeness—to the
contrary, as the record shows and the district court found
(257 F. Supp. at 720; A. 38), the sampling method was “ ac
ceptable statistically”—but it was given importance by the
legal theory of both courts that petitioner was required to
show that Garland County, not the State of Arkansas as a

A p p en d ix A

whole, applied the death penalty for rape discriminatorily.
Notwithstanding this legal conception, neither court below
contested the uncontradicted factual assertions of Mr. Mon
roe, as an expert statistician, that conclusions drawn from
data gathered in his sample counties would be valid for the
State of Arkansas. See 398 F.2d at 144; A. 53-55.)

2. Data Concerning the Critical Variables (Race
and Sentence) and Statistical Analysis of the
Relationship Between Them.

For each individual case of conviction of rape, data were
gathered as to race of defendant, race of victim, and sen
tence imposed (Tr. 28-30).Sa Using approved statistical
techniques, analysis was performed to determine the re
lationship among these variables (Petitioner’s Exhibit P-4,
pp. 2-4). Briefly, the analysis involved these steps: (a)
erection of a scientifically testable “ null hypothesis” “ as
serting there is no difference in the distribution . . . of the
sentence of death or life imprisonment imposed on Negro
or white defendants” (Tr. 30-31; see also Tr. 31-32); (b)
calculation of a “ theoretical or expected frequency” (Tr.
33) which represents the number of Negro defendants and
the number of white defendants (or, more specifically, the
number of Negro defendants convicted of rape of white
victims, and of all other defendants) who would be ex
pected to be sentenced to death if the null hypothesis (that
sentence is not related to race) were valid (Tr. 32-33); (c)

6tt The sources from which these data, and other data relating to
the individual cases of rape convictions studied, were obtained is
described in detail in Petitioner’s Exhibit P-3, pp. 1-6. See note
8a, infra. Because the accuracy of all the basic data was con
ceded by the respondent below, see text supra at pp. la-3a, methods
of data collection and data sources were not developed at the
hearing, and Exhibit P-3 was put in merely for the information
of the court.

A p p en d ix A

10a

comparison of this “ theoretical or expected frequency” with
the frequency of death sentences actually observed in the
collected data for each racial combination of defendants
and victims; and (d) determination whether the discrep
ancy between the expected and observed frequencies is suf
ficiently great that, under generally accepted statistical
standards, that discrepancy can be said to be a product
of the real phenomena tested, rather than of the operation
of chance within the testing process, sampling, etc. (Tr.
33-37). “ If that difference reaches a sufficiently high pro
portion, sufficiently high number, then the assertion can be
made, using again the traditional cut-off point,711 that the
difference is significant and could not have occurred by
chance” (Tr. 34). See 398 F.2d at 142; A. 50. The result
of this analysis, then, is the determination whether there
is a relationship or “ association” between Negro defend
ants convicted of rape of white victims and the death sen
tence imposed by Arkansas juries; and if so, whether that
relationship or association is “ significant” in the statistical
sense that the possibility of its occurrence by chance is so
slight as properly to be discounted. (See Petitioner’s Ex
hibit P-4, pp. 2-4.) (As we shall see infra, such a relation
ship, showing disproportionately frequent death sentencing
of Negroes convicted of rape of white victims, was in
fact established by the data.)

7a Dr. Wolfgang explained in considerable detail the procedures
by which relations among items of observed data are tested statis
tically for reliability, “not only in sociology and social sciences,
but other disciplines as well, . . . such as medical research” (Tr.
36). The basic procedure used in the present study— the chi-square
method of statistical analysis and the traditional measure of statis
tical “significance” which treats as real observed relationships that
could not have occurred more than five times out of one hundred
by chance (expressed in the formula P < .05)— is described at
fr . 33-37, with explication of these matters by reference to the
familiar example of head-or-tail coin tossing.

A p p en d ix A

11a

3. Data Concerning “Control” Variables.

Data gathering did not stop, however, with the facts of
race and sentence. As explained by Dr. Wolfgang, data
were collected on numerous other circumstances attending
each case of conviction for rape that “were felt to be rele
vant to the imposition of the type of sentence” (Tr. 40).
These data were sought by the exhaustive inquiries that
occupy 28 pages of small type on the data-gathering form
that is Petitioner’s Exhibit P-2—inquiries concerning the
defendant (age; family status; occupation; prior criminal
record; etc.), the victim (age; family status; occupation;
husband’s occupation if married; reputation for chastity;
etc.), defendant-victim relationship (prior acquaintance if
any; prior sexual relations if any; manner in which defen
dant and victim arrived at the scene of the offense), cir
cumstances of the offense (number of offenders and vic
tims; place of the offense; degree of violence or threat
employed; degree of injury inflicted on victim if any;
housebreaking or other contemporaneous offenses com
mitted by defendant; presence vel non at the time of the
offense of members of the victim’s family or others, and
threats or violence employed, or injury inflicted if any,
upon them; nature of intercourse; involvement of alcohol
or drugs; etc.), circumstances of the trial (plea; presenta
tion vel non of defenses of consent or insanity; joinder of
defendant’s rape trial with trial on other charges or trial
of other defendants; defendant’s representation by counsel
(retained or appointed) at various stages of trial and sen
tencing; etc.), and circumstances of post-trial proceedings
if any. See 398 F.2d at 142; A. 50-51.

The district court aptly characterized these factors as
“ Generally speaking, and subject to certain exceptions, . . .
variables . . . which reasonably might be supposed to either

A p p en d ix A

12a

aggravate or mitigate a given rape” (257 F. Supp. at 718
n. 8; A. 35, n. 8). Their exhaustive scope appears upon the
face of Petitioner’s Exhibit P-2, and from Dr. Wolfgang’s
testimony: “The principle underlying the construction of
the schedule [Petitioner’s Exhibit P-2] was the inclusion of
all data that could be objectively collected and transcribed
from original source documents that were available to the
investigators—the field investigators—such as appeal tran
scripts, prison records, pardon board records, and so forth,
and whatever was generally available was included. In this
sense, it was a large eclectic approach that was used for the
purpose of assuring ourselves that we had all available data
on these cases” (Tr. 96-97; see also Tr. 65-70). Dr. W olf
gang conceded that some data potentially pertinent to sen
tencing choice were not collected—for example, strength of
the prosecution’s case in each individual rape trial—but ex
plained that this was because such items were not informa
tion “ that we could objectively collect” (Tr. 97). See 398
F.2d at 142; A. 51.

The pertinency of these data to the study was that some
of the many circumstances investigated, “ rather than race
alone, may play a more important role in the dispropor
tionate sentencing to death of Negro defendants convicted
of raping white victims” (Tr. 40).

“ These factors, not race, it could be argued, may be
determining the sentencing disposition; and Negroes
may be receiving death sentences with disproportionate
frequency only because these factors are dispropor
tionately frequent in the case of Negro defendants.
For example, Negro rape defendants as a group, it
may be contended, may employ greater violence or do
greater physical harm to their victims than do white
rape defendants; they may more frequently be repre

A p p en d ix A

13a

sented at their trials by appointed rather than retained
counsel, and they may more frequently commit con
temporaneous offenses, or have a previous criminal rec
ord, etc.” (Dr. Wolfgang’s written report, Petitioner’s
Exhibit P-4, p. 5.)

In order to determine whether the control variables ex
plained or accounted for the racial disproportion in death
sentencing, analysis had to be made of the relationship
between each such factor for which data were available
and sentence on the one hand, race on the other. Dr. W olf
gang explained that no variable could account for the sig
nificant association between Negro defendants with white
victims and the death sentence unless that variable “ was
significantly associated with the sentence of death or life”
(Tr. 41), and unless it cdso was significantly associated
with Negro defendants convicted for rape of white victims
(Tr. 41-42).

A variable, even though associated with such Negro de
fendants (i.e., found disproportionately frequently in their
cases), could not furnish a non-racial explanation for their
over-frequent sentence to death unless it was itself affect
ing the incidence of the death sentence (as evidenced by
its significant association with the death sentence) (see,
e.g., Tr. 45-46); while a variable which was not associated
with Negro defendants convicted of rape of white victims
could also not explain the frequency with which they, as
a class, were sentenced to death (e.g., Tr. 49-52). See gen
erally Petitioner’s Exhibit P-4, pp. 6-7.

4. Results and Conclusions.

Based on his study of the data gathered for the twenty
years 1945-1965 in the State of Arkansas, Dr. Wolfgang con-

A p p en d ix A

14a

eluded categorically that “ compared to all other rape de
fendants, Negroes convicted of raping white victims were
disproportionately sentenced to death.” (Dr. Wolfgang’s
written report, Petitioner’s Exhibit P-4, p. 8, para. 3 (origi
nal emphasis).) “We found a significant association be
tween Negro defendants having raped white victims and
the disproportionate imposition of the death penalty in com
parison with other rape convictions.” (Tr. 52; see also Tr.
37-39.) Indeed, the disparity of sentencing between Negroes
with white victims and all other racial combinations of con
victed defendants and victims was such that it could have
occurred less than twice in one hundred times by chance
(Tr. 37-38)—i.e., if race were not really related to capital
sentencing in Arkansas, the results observed in this twenty-
year study could have occurred fortuitously in two (or
less) twenty-year periods since the birth of Christ. Thus,
the Wolfgang study convincingly documents the dis
crimination which previously available data—collected less
systematically or in a form permitting less rigorous scien
tific analysis—also suggests: for example, the Federal
Bureau of Prisons’ National Prisoner Statistics for execu
tions during the period 1930-1962 (Petitioner’s Exhibit
P-6, Tr. 99-101), which disclose that more than nine times
as many Negroes as whites were put to death for rape dur
ing this period in the United States. See also pp. 13-16,
supra, of this brief.

A considerable part of Dr. Wolfgang’s testimony was ad
dressed to the question whether this disproportion could be
explained away or accounted for by the operation of other,
non-racial (“ control” ) variables. He testified that after the
Arkansas data were collected, he considered and subjected
to analysis every such variable or factor about which suf
ficient information was available to support scientific study
(Tr. 56, 64-65, 78-80, 97). With respect to a substantial

A p p en d ix A

15a

number of the variables investigated by the field research
ers, their exhaustive exploration81 failed to provide enough

Sa By reason of the court’s pre-trial order and respondent’s con
cession under the procedures fixed by that order that the responses
recorded by the field researchers on the individual case schedules
were accurate (see pp. la-3a, s u p r a ) , petitioner dad not present in
any systematic fashion below testimony relating to the data-gather-
incr procedures. The concession, of course, included the accuracy
of the response “unknown” wherever that appeared on a schedule,
;lnd__as counsel for petitioner pointed out in the district court,
without disagreement from respondent or the court— the response
“unknown” “ means that research, using the State s records and
using all of the resources that we have poured into this case, is
unable to make any better case than this” (Tr. loo-156)

The nature of the. research effort involved is indicated by reti-
tioner’s Exhibit P-3 (Tr. 25-27, 57), the instructions to the field
researchers. Those instructions include the following, at pp. 4 -b :

“Whether the work is done by a single researcher or divided
among more than one, the course of investigation of any spe
cific case will ordinarily involve the following steps:

“ (1) Inspection of the county court docket books for en
tries relating to the case.

“ (2) Inspection of all other records relating to the case
available at the county court: file jackets, transcripts, witness
blotters, letter files, pre-sentence reports.

“ (3) Inspection of appellate court records in any case where
appeal was taken. Appellate court records include the. docket
of the appellate court, its file jacket, record on appeal (it
maintained on file in the appellate court), court opinion or
opinions if any, and appellate court clerk’s letter file.

“ (4) Inspection of prison records of the defendant if he
was incarcerated in a prison which maintains records.

“ (5) Inspection of pardon board records in any case where
the defendant submitted any application for executive clem
ency.

“ (6) If the foregoing sources fail to provide sufficient in
formation to complete the schedule on the case, interview of
defense counsel in the case.

“ (7) If the foregoing sources fail to provide sufficient in
formation to complete the schedule on the case, inspection of
local and area newspaper files for items pertaining to the case.

“Three general directives should be kept in mind:
“ (A ) We are concerned with the sentencing decision, in

each case, of a particular official body at a particular time

A p p en d ix A

16a

A p p en d ix A

information for study. (E.g., victim’s reputation for
chastity, Tr. 79.) Notwithstanding respondent’s pre-trial

(i.e., the trial judge or jury ; the pardon board). Every such
body acts— can act— only on the facts known to it at the time
it acts. For this reason, the “facts” of a case called for by
the schedule, mean, so far as possible, the facts perceived by
the sentencing body. Facts which we know to have been
known to the sentencing body are preferred facts, and sources
which disclose them are preferred sources. (A trial transcript,
where it exists, is therefore the most desirable source of facts.)
Other sources are of decreasing value as the likelihood de
creases that the facts which they disclose were known to the
sentencing body. (A newspaper story which purports to re
port trial testimony, therefore, is to be preferred to one which
purports to report the facts of the offense on the basis of
other sources of information.)

“ (B) After this survey is completed, its results will be made
the basis for allegations of fact in legal proceedings. I f the
allegations are controverted, it will be necessary to prove them,
and the proof will have to be made within the confines of
ordinary evidentiary rules, including the hearsay principle,
best evidence rule, etc. For this reason, sources of facts which
are judicially admissible evidence to prove the facts which
they disclose are preferred sources. Official records are most
desirable in this dimension; then the testimony of witnesses
having knowledge of the facts (for example, defense counsel),
finally, secondary written sources (for example, newspapers).
Wherever an official record or document may contain perti
nent information, inspect it yourself if you can; don’t take
somebody’s word for what is in it.

“ (C) Many of the facts you need to know will have been
contested in the judicial and post-judicial proceedings lead
ing to a defendant’s sentence and its execution. W e have no
method for resolving factual disputes or, ordinarily, for know
ing how the triers of fact resolved them. As an invariable
rule, then, the facts should be reported in the light most
favorable to the prosecution, and most unfavorable to the
defendant, in every case. I f a trial transcript exists, and if
it contains the testimony of the complaining witness and of
the defendant, resolve all conflicts of testimony in favor of
the complaining witness and report the facts as they might
reasonably have been found by a jury which credited the
complaining witness, drew all rational inferences from her

17a

concession of the accuracy of the field researchers’ re
sponses on the individual case schedules, including the
response “ unknown” where that appeared (see note 8a
supra), counsel for respondent attempted to suggest in
cross-examination of Dr. Wolfgang that these gaps in in
formation impugned the underlying data-gathering process.
Dr. Wolfgang replied:

“the absence of information, I would be unwilling to as
sert is due to lack of any effort. Very diligent efforts
were made by the field investigators to collect the in
formation—from court clerks, from police records,
from prisons, from other sources available in the com
munity—and they were instructed to follow down each
piece of information, each source of information to its
fullest extent, so that I have no reason to doubt that
the effort was made to collect the data” (Tr. 80).

His testimony as a whole makes it clear that—although, as
he put it: “ Information is always limited” (Tr. 72)—he
was confident that he had enough of it to support his con
clusions. (See particularly Tr. 76-79.)

He was able to subject twenty-two “quite relevant vari
ables” (Tr. 78)—in addition to race of defendant, race of
victim, and sentence—to analysis. (See Petitioner’s Ex
hibit P-4, Appendix A ; Tr. 29, 52.) Most of these were not
significantly associated with sentence, and so Dr. W olf-

testimony most strongly against the defendant, discredited
the defendant, and refused to draw any disputable inferences
in bis favor. Treat all other sources in a similar fashion. In
interviews with defense counsel, try to impress upon counsel
that you have to have the facts as they might have appeared
in the worst light for his client. In reading newspaper items
which give conflicting versions of the facts, adopt the version
most unfavorable to the defendant.”

A p p en d ix A

18a

gang could assert categorically that they did not account
for or explain the disproportionately frequent death sen
tencing of Negroes with white victims (Tr. 42-46, 53-54).
These variables included the defendant’s age, whether he
was married, whether he had dependent children, whether
he had a prior criminal record; the victim’s age, whether
she had dependent children; whether the defendant and
victim were strangers or acquaintances prior to the offense;
place where the offense occurred (indoors or outdoors),
whether the defendant committed an unauthorized entry in
making his way to that place; whether the defendant dis
played a weapon in connection with the offense; degree of
seriousness of injury to the victim; and the defendant’s
plea (guilty or not guilty), type of counsel (retained or
appointed), and duration of trial (Tr. 47, 53; Petitioner’s
Exhibit P-4, p. 9, para. 5; Petitioner’s Exhibit P-5). Two
variables were shown to bear significant association with
sentence: death sentences were more frequent in the cases
of defendants who had a prior record of imprisonment, and
in the cases of defendants who committed other offenses
contemporaneously with the rape. But because these vari
ables were not associated with race,9* Dr. Wolfgang con
cluded that they also could not account for the fact that
Negroes convicted of rape of white victims were dispropor
tionately often sentenced to death (Tr. 47-52, 54; Peti
tioner’s Exhibit P-4, pp. 8-9, para. 4; Petitioner’s Exhibit
P-5). Other variables appeared so frequently or so in-

9a Statistical analysis of the association between these variables
and race of the defendant disclosed no significant association.
When defendant-victim racial combinations were, considered, the
numbers of cases for which information was available became too
small for statistical treatment, but on the basis of trend of as
sociation, Dr. Wolfgang concluded that here too there was no
association of significance.

A p p en d ix A

19a

frequently in the total population of cases studied that
statistical analysis of them was “unnecessary and impos
sible” : the fact that they appeared to characterize all cases
(or no cases), irrespective of sentence or of racial com
binations of defendant and victim, pointed to the conclu
sion that they were not available explanations for the re
lationship observed between death sentences and Negroes
with white victims. These variables included the victim’s
reputation for chastity and prior criminal record; whether
the defendant and victim had had sexual relations prior
to the occasion of the rape; the degree of force employed by
the defendant; whether the victim was made pregnant by
the rape • and whether the defendant interposed a defense
of insanity at trial (Tr. 54-55, 94-95, Petitioner’s Exhibit
P-5). Summarizing, Dr. Wolfgang found that no variable
of which analysis was possible could account for the ob
served disproportionate frequency of sentencing to death
of Negroes convicted of rape of white victims (Tr. 56-57).
His ultimate conclusion was:

“ On the basis of the foregoing findings, it appears
that Negro defendants who rape white victims have
been disproportionately sentenced to death, by reason
of their race, during the years 1945-1965 in the State
of Arkansas.” (Dr. Wolfgang’s written report, Peti
tioner’s Exhibit P-4, pp. 13-14 (emphasis added).)

B. The Opinion of the District Court.

Although respondent presented no evidence of any sort
in rebuttal, the district court disagreed with Dr. W olf
gang’s conclusions. It accepted his finding that the differ
ential sentencing to death of Negroes with white victims
“could not be due to the operation of the laws of chance”

A p p en d ix A

20a

(257 F. Supp. at 718; A. 35); but supposed, again without
any sort of evidentiary presentation by the State, that it
might be due to some factor respecting which statistical
analysis had not been possible, such as the issue of consent
in rape cases (257 F. Supp. at 720-721; A. 40). The Court
remarked that the “ variables which Dr. Wolfgang con
sidered are objective . . . broad and in instances . . . im
precise” ; that in many of the individual rape cases studied
“ the field workers were unable to obtain from available
sources information which might have been quite perti
nent” ; and that “ Dr. Wolfgang’s statistics really reveal
very little about the details” of comparative individual cases
of rape. (257 F. Supp. at 720; A. 39.) While recognizing
that “ the sample drawn by Mr. Monroe seems to have been
drawn in a manner which is acceptable statistically” (257
F. Supp. at 720; A. 38), the court noted that the counties
randomly chosen had turned out not to be evenly geographi
cally dispersed, and not to include many counties of sparse
Negro population (ibid.). Garland County, which was not
itself included in the sample, is a county of sparse Negro
population located in a portion of the State in which the
sample counties fell less frequently than elsewhere. For
those reasons, the district court declined to conclude that
“ the Garland County jury which tried petitioner was moti
vated by racial discrimination when it failed to assess a
punishment of life imprisonment” (257 F. Supp. at 719;
A. 37)— (a subjective proposition, parenthetically, which
petitioner’s counsel had explicitly disavowed any intention
to undertake to prove). With regard to the State of
Arkansas generally, the district court thought that the
“ cases studied, and the number of death sentences imposed
are simply too few in number to afford convincing proof”
of racial discrimination (257 F. Supp. at 720; A. 38).

A p p en d ix A

21a

Placing some reliance on the language in Dr. W olfgang’s
written report to the effect that the report was “ prelimi
nary” and “ tentative” (257 F. Supp. at 720; A. 39), the
court concluded:

“ On the meager material before it the Court is sim
ply not prepared to convict Arkansas juries of uncon
stitutional racial discrimination in rape cases. As a
matter of fact, the Court doubts that such discrimina
tion, which is a highly subjective matter, can be de
tected accurately by a statistical analysis such as was
undertaken here. Statistics are elusive things at best,
and it is a truism that almost anything can be proved
by them.” (257 F. Supp. at 720; A. 39-40; see also 257
F. Supp. at 719-720; A. 37-38.)

C. The Opinion of the Court of Appeals.
The court of appeals agreed with the district court in

rejecting petitioner’s contention that his death sentence
must be set aside by reason of racial discrimination in capi
tal sentencing—but for somewhat different reasons. “Like
the trial court, . . . although perhaps not for each and all
of the reasons it advanced, we feel that the [statistical]
argument does not have validity and pertinent application
to Maxwell’s case.” (398 F.2d at 146; A. 59.) “What we are
concerned with here is Maxwell’s case and only Maxwell’s
case.” (398 F.2d at 147; A. 60.)

The court of appeals, in general, appears to concede the
validity of Dr. Wolfgang’s methodology, and even of his
conclusion that a pattern of racial discrimination has been
shown. (398 F.2d at 146-148; A. 59-64.) It finds that Dr.
Wolfgang testified his report was “preliminary” only “ in
the sense that the other states for which data was being col

A p p en d ix A

22a

lected would be included in the final report.” (398 F.2d at
143; A. 52-53.) And the court accepts that there are “ rec
ognizable indicators” “ that the death penalty for rape may
have been discriminatorily applied over the decades in that
large area of states whose statutes provide for it.” 10a
Nevertheless the court holds, in effect, that this statistical
showing of discrimination is irrelevant because it does “not
show that the petit jury which tried and convicted Maxwell
acted in his case with racial discrimination” (398 F.2d at
147; A. 60) or that Maxwell himself “ was the victim of dis
crimination based on race” (398 F.2d at 148; A. 64.)

“ [W]hatever . . . suspicion it may arose [sic] with
respect to southern interracial rape trials as a group
over a long period of time, and whatever it may dis
close with respect to other localities, we feel that the
statistical argument does nothing to destroy the in
tegrity of Maxwell’s trial.” (398 F.2d at 147; A. 61-62.)

This is so because the Arkansas statistics “ do not relate
specifically to Garland County where this particular offense
was committed and where Maxwell was tried and con
victed.” (398 F.2dat 146; A. 59.)

“ [W ]hile it is true that it is in a sense the state
which prosecutes, nevertheless the county has a char-

10a (398 F.2d at 148; A. 63.) Despite the generality of the court’s
language, this finding must relate specifically to the State of A r
kansas. This is so because no proof was offered on this record of
the. results of the Wolfgang study in other States. The only evi
dence of record with regard to areas outside Arkansas was Peti
tioner’s Exhibit P-6, the National Prisoner Statistics. These show
raw numbers of executions for rape and murder, by race, for all
the States. (See pp. 13-15 n. 11, supra.) They do no more than
show that Dr. Wolfgang’s conclusions for Arkansas are conform
able to the national pattern. Thus, if there is south-wide discrimi
nation, there is most assuredly discrimination in Arkansas.

A p p en d ix A

23a

acter, too. . . . Yet the Garland County statistics [pro
duced in the earlier habeas corpus proceeding] . . .
afford no local support to the petitioner’s statistical
argument. The evidence produced at the prior hearing
and at this one discloses only Maxwell as a recipient of
the death penalty in Garland County for rape.

“ . . . [W ]e are not yet ready to nullify this peti
tioner’s Garland County trial on the basis of results
generally, but elsewhere, throughout the South.” (398
F.2d at 147; A. 61.)

We think it apparent, on a fair reading of the court of
appeals’ opinion, that that court found unavoidable Dr.
Wolfgang’s conclusion concerning racial discrimination in
capital sentencing by Arkansas juries in rape cases. At
the very least, the court of appeals explicitly admitted
that there are “ recognizable indicators” and “ ground for
suspicion that the death penalty for rape may have been
discriminatorily applied over the decades in that large
area of states whose statutes provide for it.” (398 F.2d at
148; A. 63.)

A p p en d ix A

24a

APPENDIX B

Available Information Relating to the Proportion of
Persons Actually Sentenced to Death, Among Those
Convicted of Capital Crimes.

It is exceptionally difficult to obtain information which
will permit construction of a reliable estimate concerning
the proportion of persons actually sentenced to death,lb
among all those convicted of capital crimes. In order to
calculate this proportion, two comparable figures are re
quired : the number of persons convicted of a given capital
offense, and the number of persons sentenced to death
for it. The figures must be comparable in the sense that
they relate to the same jurisdiction, the same period of
time, and the same offense.

Figures for the number of death sentences imposed
during recent years, for each of the several capital offenses,
in each of the capital-punishment jurisdictions, may be
estimated—albeit somewhat inexactly—from data reported
in the National Prisoner Statistics Bulletins on Executions.
The relevant tables are cited in the footnote,Jb which also

,b By “persons actually sentenced to death,” we mean persons
upon whom the sentencing authority, judge or jury, imposes a
death sentence. W e do not mean to take account of later judicial
decisions vacating the sentence, of executive clemency, or of any
other factor that may result in the initial death sentence not being
carried out.

2b The National Prisoner Statistics Bulletins on Executions are
published annually. The latest is United States Department of
J ustice, B ureau of P risons, National P risoner Statistics, No.
42, Executions 1930-1967 (June, 1968). In the columns below we
list the numbers and titles of the issues relied upon in this A p
pendix, the abbreviation which will hereafter be used to designate
each, and the pages and tables of each that contain pertinent in
formation.
Number, title, date Abbreviation Page and table
No. 28, Executions NPS (1961) p. unnumbered,
1961 (April, 1962) table 5

25a

sets forth the citation form that will be used hereafter
to describe each bulletin. There are several problems with
the tables as sources of information relating to the num
ber of persons sentenced to death. First, the figure which
they report represents the number of persons received
by the respective state penitentiaries from the courts,
each year, under sentence of death. This may be fewer
than the actual number of death sentences imposed in
some states, where the practice is to hold a death-sentenced
man in a local facility pending post-verdict motions and/or
appeals. Second, for this or other reasons, a few death
sentences imposed each year are not reported to the NPS
until the following year. The latest NPS Bulletin gives
revised figures for prisoners received from court under
sentence of death for each year since 1961 (NPS (1967),
p. 12, table 4 ); but these figures are not broken down by
jurisdiction and offense. Figures broken down by juris
diction and offense must be taken from the earlier annual
reports, and have not been revised.

But these problems are inconsequential compared to
those of ascertaining comparable figures relating to the

A p p en d ix B

Number, title, date Abbreviation Page and table
No. 32, Executions NPS (1962) p. unnumbered,
1962 (April, 1963) table 5
No. 34, Executions
1930-1963 (May, 1964)

NPS (1963) p. 14, table 5

No. 37, Executions
1930-1964 (April,
1964 [sic: 1965])

NPS (1964) p. 14, table 4

No. 39, Executions
1930-1965 (June, 1966)

NPS (1965) p. 14, table 4

No. 41, Executions
1930-1966 (April, 1967)

NPS (1966) p. 13, table 4

No. 42, Executions
1930-1967 (June, 1968)

NPS (1967) pp. 12-17, tables 4-7

26a

number of convictions for capital offenses. Conviction
figures are almost nowhere published. The rare published
figures suffer from assorted woes that virtually destroy
their usefulness. Some states report judicial statistics
for fiscal years, making comparison with the calendar-
year NPS reports difficult. Most states report conviction
figures for categories of cases (“murder” ; “ sexual o f
fenses” ) that include, but are not entirely composed of,
capital crimes. Many states report not conviction figures,
but figures concerning the number of commitments to the
state penitentiary under conviction and sentence for given
offenses. These figures omit large numbers of convicted
persons: namely, those sentenced to imprisonment in local
facilities, or to probation, or whose sentences are sus
pended. Secondary sources reporting conviction figures
suffer from the same defects and often, in addition, they
report totals for a span of years that includes both a
period of mandatory capital sentencing and a succeeding
period of discretionary capital sentencing within a juris
diction.

Confronted with these problems, we adopt the approach
of setting out below all of the information we can find
relating to the question of what proportion among all
convicted capital offenders are actually sentenced to death.
The information comes from a variety of states, for a
variety of periods, and has a variety of problems. We
report each item separately, and explain its problems as
we see them.

A. Rape.

1. Partington appears to say that, between 1908 and
1963, there were 2798 offenders committed to the Virginia
State Penitentiary upon convictions for the capital crimes

A p p en d ix B

27a

of rape, attempted rape, statutory rape, and attempted
statutory rape. There were 68 death sentences in this
group, including two imposed upon defendants who had
been convicted of capital robbery as well as the sexual
offense. For rape alone, there were 1565 commitments,
including 41 death sentences. The principal problem with
these figures appears to be that the commitment figures
are doubtless considerably lower than the number of
capital convictions, since some capital offenders would
have received jail terms, probation, or suspended sen
tences. Partington, The Incidence of the Death Penalty
for Rape in Virginia, 22 W ash. & Lee L. Rev. 43, 43-44,
71-73 (1965).

2. Johnson reports that in North Carolina, between
July 1, 1938 and December 31, 1953, there were 382 convic
tions for rape, as compared with 52 death sentences (ad
missions to death row) upon conviction for rape. The
conviction figure seems unduly low, even considering the
availability to the jury of verdicts for lesser offenses.
Johnson, Selective Factors in Capital Punishment, 36
Social F orces 165, 166 (1957).

3. In the federal courts, where rape is capitally punish
able under both the federal code and the District of Colum
bia Code, there appear to have been a total of 111 rape
convictions between July 1, 1961 and June 30, 1966, ex
cluding the fiscal year 1964-1965, for which no figure is
reported. For the calendar years 1961 through 1966, no
death sentences were imposed for rape by a federal court.
Conviction figures are taken from table D5 in the Annual
Reports of the Director of the Administrative Office of
the United States Courts for the years 1962-1966; death-
sentence figures from NPS (1961) through NPS (1966).

A p p en d ix B

Copyright notice

This collection and the tools to navigate it (the “Collection”) are available to the public for general educational and research purposes, as well as to preserve and contextualize the history of the content and materials it contains (the “Materials”). Like other archival collections, such as those found in libraries, LDF owns the physical source Materials that have been digitized for the Collection; however, LDF does not own the underlying copyright or other rights in all items and there are limits on how you can use the Materials. By accessing and using the Material, you acknowledge your agreement to the Terms. If you do not agree, please do not use the Materials.

Additional info

To the extent that LDF includes information about the Materials’ origins or ownership or provides summaries or transcripts of original source Materials, LDF does not warrant or guarantee the accuracy of such information, transcripts or summaries, and shall not be responsible for any inaccuracies.

Maxwell v Bishop Brief for Petitioner

This item is featured in:

Collection

Taxonomy

Identifier

Cite this item

Copyright notice

Additional info

You may also be interested in