no<>v£”'l’-v.~ —~.

.-.v~»n-'»Q

-

.

v
.

 

.......u'..u..«

 
 

I I'll I I I l I l‘ I I ill.

 

 

 

 

 

 

Jllllllllllllllllllllllllllllllllllllllllll t

31293 01683 899 91 °

This is to certify that the

dissertation entitled

The Nature of Agenda in the United States s“
Supreme Court and Courts of Appeals ’
:

presented by )

3

Mark S . Hurwitz

has been accepted towards fulfillment

u‘-——-f—“f> a.

of the requirements for

V!

Ph.D. Political Science

 

 

 

 

 

 

 

 

 

 

degree in .*

O

l

>

..

i

guru” 5' w ’

agor professor l’

Reg nald‘dé Sheehan t

Date August 10, 1998 ;
l

MSU is an Affirmative Action/Equal Opportunity Institution O~ 12771 f
f

if 7 i
,>

o

l

5

'

LIBRARY r

MlChigBF “gate 'r

. ' t

UfllVera'i‘Ey .

 

the

4.“.

\

- t

.,

 

THE NATURE OF AGENDA IN THE UNITED STATES SUPREME COURT
AND COURTS OF APPEALS

VOLUME I
By

Mark S. Hurwitz

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

1998

 

 

33w.
N
3mm.“ ‘
“mug: - '
“\L,‘ A.
O
9
"s -
~l')- 2-9
. KR“ 9"».
. ¥‘-.
1
31"Vka
c \ a J
:-.. «‘2‘
u‘
D
§
SEQ '
(-I‘ “a.
‘39 5.
‘n.u:v.\q

‘ .
1“: '
: .‘h'x
.P‘K" ‘
‘K 3" .
S “1‘ Q0.- ‘
\‘A ‘_
'\
4—h4
01-; , .
“Sc“-
L‘.'t,, -
“a“ ‘u

ABSTRACT

THE NATURE OF AGENDA IN THE UNITED STATES SUPREME COURT
AND COURTS OF APPEALS

By

Mark S. Hurwitz

While there is a good deal of scholarly research on the agenda in the United States
Supreme Court, much less is known about agenda setting in the United States Courts of
Appeals. Moreover, even less is known about whether the agenda of one of these institutions
influences the agenda of the other and, if so, the nature of any such agenda change. These
questions are especially vexing in light of the disparate jurisdiction of the Supreme Court,
which is principally discretionary, and of the Courts of Appeals, which is generally
mandatory, and considering the Supreme Court’s limited oversight of the circuit courts.
Utilizing both the United States Supreme Court Judicial Database and the United States
Courts of Appeals Data Base, and taking a neoinstitutional approach, I address these issues
systematically by examining the agendas of both of these judicial institutions and, more
importantly, employing time series analyses to identify the extent to which agenda setting
on the Supreme Court inspires agenda change on the Courts of Appeals, and vice versa. I
find that under some conditions the Supreme Court inspires agenda change in the Courts of
Appeals, while under other circumstances it is the Courts of Appeals which take the lead in
setting the agenda of the federal courts. Moreover, additional factors enter into the issue of

agenda setting in these courts as well.

Dedicated to my wife Sheralee,

for her abiding love, encouragement, and support

iii

ACKNOWLEDGMENTS

It was rather simple for me to decide to leave the practice of law in New York City
in order to pursue a Ph.D. in political science. It was much harder, however, to make this
transition work and realize my goals. As with any demanding task such as this, I could not
have succeeded solely on my own, and thus I would like to recognize those people who
represent a significant part of my achievement.

To begin, I appreciatively acknowledge the members of my dissertation committee
at Michigan State University. First among them is Reggie Sheehan. As chair of my
committee, he has guided me through the dissertation process, ensuring I was on the right
track, while always encouraging me to carry through with my ideas, even when they
diverged from his. As a judicial scholar, he has provided a stellar model for me to emulate,
and I hope to be as successful as he has become. Finally, as a friend, he has demonstrated
the importance of balancing professional success with other important aspects of life, such
as our respective families — of course, quality time on the golf course doesn’t hurt, either.
I thank him for all of these things.

I also would like to thank David Rohde who granted me a PIPC fellowship in my first
year of graduate school and has continued to support my studies since then. While there
were many tangible benefits for being a part of the PIPC program, it is the intangible values
I acquired from PIPC for which I am most grateful. In particular, I am indebted to being part

iv

 

z' ‘T’Jfl "' ,5.
I; hos“. y.

'2.
u
o

4 .-'
-~..-..~ .

K‘- .L ._
0' 'i | H
1". L‘ Hobs $.‘\

 

of an academic environment in which graduate students work with each other in a positive
setting. While the other PIPC fellows are responsible for this constructive climate as well,
as director of PIPC it was he who had the foresight to design the program as it is and to
choose these prospective scholars. I am convinced that he is quite correct in his assumption
that graduate students can learn as much from their peers as from their mentors, and PIPC
is substantial evidence in this regard. Accordingly, I thank him for the opportunity to be a
part of this wholly successful and enjoyable program, and I wish PIPC well in the future.
Dave, as you once noted in regard to my achievements, PIPC is batting 1.000, and I am
certain its record will continue to flourish under your guidance.

Two other members of my committee, Jim Granato and Melinda Gann Hall, deserve
special recognition. Jim provided me with the appropriate tools and methods to pursue
rigorous scholarship in political science, and Melinda compelled me always to conceptualize
the theoretical bases of my work. Notwithstanding, I am most appreciative and thankful to
each of them for the ceaseless support, counsel, and fi'iendship they have imparted to me
since I have known them. They both illustrate that genuinely decent people and successful
political scientists are not necessarily mutually exclusive concepts.

I also would like to thank the final member of my committee, Harold Spaeth, for
reading drafts of my dissertation and assisting me with data coding decisions.

A number of other faculty members have been concerned with my success and helped
me in my graduate studies, and they deserve to be acknowledged as well, including Brian
Silver, Darren Davis, Carol Weissert, and Jeff Riedinger. Additionally, I would like to thank
Ada F inifier who, as editor of the American Political Science Review, chose me to be one of

V

 

.
m. |.,;o 'F'.)"
a 1...! ..I.§.

 

’1

‘Yfign'ia .4.
s... a.‘s‘.\ i~
t

- n

o . _. .
f.“ I\ «2" .
«L... .....g\ 5

“"~ . ‘
a 4 '1" ‘1'; m
f““"'--~ t...

..
4..-“
u.‘ "" 5‘."
b‘t-ll...‘ L‘h. ‘
.
I

. I
he 0..

«L. t. s...:01‘: \

 

Y‘ ~'. _
I‘ a.

l :a- . ‘

u. L"\:T.‘h

. .

”no -

. 3 ‘\ .-.

:- «an; ':

‘ ~~ .: .

I. ‘ U: ”‘93 u
inu5 A

11'“ s '

3.: “J: ' -.‘-..
‘t 1 ma

“‘ .
.L -
‘MR ‘7‘ i“?
b ‘ k v
"‘V‘

.1
\‘K‘
‘Il TN
“'i¢\‘v~ ,L
"li ju K
k. ‘
4' o
.. M"
‘\k:\_ ‘49.
"u
L:
A“
‘ My“! m
‘ 6
.l‘
l'r
an \-
- ".3.“
. ““8 ‘. 1
‘ u.
.‘.

her first interns, a position I found to be very gratifying and intellectually stimulating. My
experience at the APSR should be invaluable in my endeavor to move my name from the
front inside cover to the Table of Contents page, a goal I hope to achieve while she remains
the APSR editor.

I again must thank all those scholars who believed in me and my work, supported my
aspirations throughout my graduate studies, and sought to advance my career in a positive
direction, particularly when the circumstances were such that it was difficult or inconvenient
for them to do so. It is my heartfelt promise to each of you, and you know who you are, that
I will do everything possible to become the political scientist you believe I can be, to act as
professionally in my career as you have been in yours, and to leave you with nothing but a
sense of pride in me. Your support and concern are debts I cannot repay, but they are
something I always will remember and of which I will remain deeply appreciative.

I would like to acknowledge a few graduate students as well. In particular, I enjoyed
sharing an office with the PIPC boys, including Bryan Marshall, Brandon Prins, and Roger
Moiles, who provided lots of encouragement and laughs. Carry on, gentlemen. ’
Additionally, I especially want to acknowledge Matt Kleiman, for nothing in particular and
everything in general. He is a great friend who, if he wants to, should become one of the top
scholars in the profession. Besides, having a fellow fan of the New York Rangers, Yankees,
and Knicks around has made living in Michigan all that much better.

I also must thank all those in my family for all they have done for me, particularly
my parents, who believed in me long before I ever did. Their encouragement and love have
served me immeasurably in all aspects of my life. Their greatest success, however, pertains

vi

 

t ‘ i
. O - 0-,... o
“i: ‘ ~
.0 Ads...-
0

R:0;O°; ~zm,.
.Hu..~ \c..l\
A

Q. - $ ..
-J. 'l'lf 171;“

a , .
‘~~- -

n 3‘ 1’ V
*4‘...-\ -1-

 

. A) Mr .
~u5-‘t5‘. i.4\ “
I
' Q. I
an 5 flu“) la“ 3
“N “l3 A 5‘ ~

 

I. .
\ ~l K~
“ “4' " J p. -
. iL l~‘ ‘3‘“
' C
" I. .
t.‘ “-10.,“ '
P‘s .
9 t. x ‘ \‘
a~l~l;.-‘V:.~
"\ .u—o _
‘\.oo\.‘. '4

'0 _ 3‘. .u
' dl A q
A .._I\).'\Uo. 3..

.«l

w
V
I...
«fit
t. “IL 3,.) ‘L
W; M .-
o»
I
n“-
‘W'fl‘vs'-
\ D
u \ ‘ka .
\ \
\‘zy ‘
° 1» y...
Th5. n
“IN? r.
..\

s. \ ,

“ 31‘ t‘
‘4‘ up
‘A

to the family they have raised. Having each come from difficult family backgrounds, my
parents came together to provide all of their children with a set of values and morals second
to none. The fruit of their efforts is found in the lives of their four children and respective
families, and (at this point) their eight grandchildren. I look forward to going to the Marvin
and Patricia Hurwitz family reunions for many years to come. I am very proud of you both.

Last, but certainly not least, I must acknowledge the importance in my life of my wife
Sheralee, my daughter Rachel, and my new son Jacob. All of you continue to provide me
with the love, strength, and incentive to do my best in everything I do. I could not have
wished for a better family, and I truly feel blessed by all of you. Most prominently, however,
is my love, respect, and admiration for you, Sheralee. You have given up so much of
yourself so I could pursue what I needed to do for me, yet I never received anything but
encouragement and support from you. I have no doubt that I would not be where I am now
if not for you, and I am eternally grateful. Besides being the best wife and friend I ever could
have, you are the best lawyer I know. I am so pleased we have not forgotten the meaning of
these words we chose to symbolize our lives almost six years ago (Springsteen 1992):

Now there’s a beautiful river in the valley ahead

There ’neath the oak’s bough soon we will be wed

Should we lose each other in the shadow of the evening trees

I’ll wait for you

And should I fall behind

Wait for me.

My beloved, this dissertation is dedicated to you.

Mark S. Hurwitz
August 10, 1998

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................ x
LIST OF FIGURES ............................................................................................................ xii
INTRODUCTION ............................................................................................................... 1
CHAPTER 1
THE NATURE OF AGENDA IN THE FEDERAL COURTS: A REVIEW ..................... 3
The Relevance of Agenda Formation and Political Institutions ............................. 5
Agenda Setting in the Courts ................................................................................ 10
Dynamic Changes in the Agendas of the Federal Courts ...................................... 18
CHAPTER 2
THEORETICAL RELATIONSHIPS AND RESEARCH DESIGN AND
METHODOLOGY ........................................................................................................... 40
A Theoretical Basis of the Nature of Agenda in the Federal Courts ..................... 41
Research Design and Methodology ...................................................................... 51
Data ........................................................................................................... 52
Methods .................................................................................................... 54
Validity ..................................................................................................... 57
CHAPTER 3
AN EXPLORATION INTO THE CAUSAL ASSOCIATIONS BETWEEN THE
AGENDAS OF THE SUPREME COURT AND COURTS OF APPEALS .................... 61
Agenda Influences in the Federal Courts in General ............................................ 63
Specific Issues, Agendas, and the Federal Courts ................................................. 67
Data, Methods, and Analysis ................................................................................ 76
Discussion ............................................................................................................. 96
CHAPTER 4

A CLOSER EXAMINATION OF THE RELATIONSHIP BETWEEN THE CIVIL
LIBERTIES AGENDAS OF THE SUPREME COURT AND COURTS OF

APPEALS ........................................................................................................................ 1 55
Issues Within Civil Liberties, Agendas, and Theoretical Suppositions ............... 157
Civil Rights .............................................................................................. 158

viii

First Amendment ...................................................................................... 160

Due Process .............................................................................................. 161
Privacy ...................................................................................................... 162
Data, Methods, and Analysis ................................................................................ 164
Discussion ............................................................................................................ 169
CHAPTER 5
MULTIVARIATE ANALYSES OF THE INFLUENCES ON THE AGENDAS
IN THE SUPREME COURT AND COURTS OF APPEALS ........................................ 202
Influences on Agendas in the Federal Courts ....................................................... 203
Endogenous Influences on Agenda ......................................................... 205
Judicial Influences on Agenda .................................................................. 208
Exogenous Judicial Interventions ............................................................. 214
External Influences on Agenda ................................................................ 216
Data and Methods ................................................................................................. 221
Coding of Data .......................................................................................... 221
Data Transformations for Time-Series Analysis ...................................... 224
Estimation of an Endogenous System ...................................................... 227
Analysis ................................................................................................................ 233
Civil Liberties System .............................................................................. 234
Criminal Procedure System ...................................................................... 236
Economics System ................................................................................... 237
Discussion ............................................................................................................ 238
CONCLUSION ................................................................................................................ 258
APPENDICES
Appendix A — Coding Decision Rules ................................................................. 263
Appendix B - Landmark Decisions ...................................................................... 266
LIST OF REFERENCES ................................................................................................. 268
LIST OF CASES .............................................................................................................. 281

ix

 

‘1 y s‘. 4 .\. 6 q 8 9

I .A I d ulw v . v a 11.; i 1 -I|.. all. Q. .s

«1. A1. ‘\. «\. ~\. ~\. ‘\. \\. I‘ 4

. n.“ . s.» . an 27.. ..._..\ .. gm . A...‘ . ._\ . m1» . Pm

.. y . . .u. A . s . .. . . .. .

..q .u. .1 an . .ut . .1. Pu h... “A hm hm

$ . TA 7 A Tl. .7 i T1 T11 T.... Ti Tl. \rl.

 

Table 3.1
Table 3.2

Table 3.3
Table 3.4
Table 3.5
Table 3.6
Table 3.7
Table 3.8

Table 3.9
Table 3.10

Table 3.11
Table 3.12
Table 3.13
Table 3.14
Table 3.15
Table 3.16
Table 3.17

Table 3.18
Table 3.19
Table 4.1
Table 4.2

Table 4.3
Table 4.4
Table 4.5
Table 4.6

Table 4.7
Table 4.8
Table 4.9
Table 4.10

LIST OF TABLES

Augmented Dickey-Fuller Tests for Unit Roots (Series in Levels) ......... 100
Augmented Dickey-Fuller Tests for Unit Roots (Series in First

Differences) .............................................................................................. 101
Granger Causality Results ........................................................................ 102
Variance Decomposition of Civil Liberties Cases (Differenced Data).... 107
Variance Decomposition of Civil Liberties Cases (Level Data) .............. 108
Variance Decomposition of Economics Cases (Differenced Data) .......... 109
Variance Decomposition of Economics Cases (Level Data) .................... 1 10
Variance Decomposition of Criminal Procedure Cases (Differenced

Data) ......................................................................................................... 1 1 1

Variance Decomposition of Criminal Procedure Cases (Level Data) ...... l 12
Variance Decomposition of Union and Labor Cases (Differenced

Data) ......................................................................................................... 1 13
Variance Decomposition of Union and Labor Cases (Level Data) .......... 114
Variance Decomposition of Attorneys Cases (Differenced Data) ............ 115
Variance Decomposition of Attorneys Cases (Level Data) ...................... 1 16
Variance Decomposition of Federalism Cases (Differenced Data) .......... 117
Variance Decomposition of Federalism Cases (Level Data) .................... 1 18
Variance Decomposition of Interstate Relations Cases (Level Data) ....... 1 19
Variance Decomposition of Federal Taxation Cases (Differenced

Data) ......................................................................................................... 120
Variance Decomposition of Federal Taxation Cases (Level Data) ......... 121
Variance Decomposition of Miscellaneous Cases (Level Data) ............. 122

Augmented Dickey-Fuller Tests for Unit Roots (Series in Levels) ......... 174
Augmented Dickey-Fuller Tests for Unit Roots (Series in

First Differences) ...................................................................................... 175
Granger Causality Results ........................................................................ 176
Variance Decomposition of Civil Rights Cases (Differenced Data) ........ 17 8
Variance Decomposition of Civil Rights Cases (Level Data) .................. 179
Variance Decomposition of First Amendment Cases (Differenced

Data) ......................................................................................................... 180
Variance Decomposition of First Amendment Cases (Level Data) ........... 181
Variance Decomposition of Due Process Cases (Differenced Data) ......... 182
Variance Decomposition of Due Process Cases (Level Data) ................... 183
Variance Decomposition of Privacy Cases (Differenced Data) ............... 184

X

Table 4.1 1
Table 5.1
Table 5.2
Table 5.3
Table 5.4
Table 5.5
Table 5.6
Table 5.7
Table 5.8
Table 5.9
Table 5.10
Table 5.11

Table 5.12

Variance Decomposition of Privacy Cases (Level Data) ......................... 185
Supreme Court Civil Liberties (Burger) - ZSLS Regression Results ...... 246
Supreme Court Civil Liberties (Warren) - ZSLS Regression Results ...... 247
Courts of Appeals Civil Liberties (Burger) - 2SLS Regression

Results ....................................................................................................... 248
Courts of Appeals Civil Liberties (Warren) - ZSLS Regression

Results ....................................................................................................... 249
Supreme Court Criminal Procedure (Burger) - ZSLS Regression

Results ....................................................................................................... 250
Supreme Court Criminal Procedure (Warren) - ZSLS Regression

Results ....................................................................................................... 251
Courts of Appeals Criminal Procedure (Burger) - ZSLS

Regression Results .................................................................................... 252
Courts of Appeals Criminal Procedure (Warren) - ZSLS

Regression Results .................................................................................... 253
Supreme Court Economic Cases (Burger) — 2SLS Regression

Results ....................................................................................................... 254
Supreme Court Economic Cases (Warren) - ZSLS Regression

Results ....................................................................................................... 255
Courts of Appeals Economic Cases (Burger) - ZSLS Regression

Results ....................................................................................................... 256
Courts of Appeals Economic Cases (Warren) - ZSLS Regression

Results ....................................................................................................... 257

xi

Figure 1.1
Figure 1.2
Figure 1.3
Figure 1.4
Figure 1.5
Figure 1.6
Figure 1.7
Figure 1.8
Figure 1.9
Figure 1.10
Figure 1.11
Figure 1.12
Figure 1.13
Figure 1.14
Figure 1.15
Figure 1.16
Figure 1.17
Figure 3.1

Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Figure 3.6
Figure 3.7
Figure 3.8
Figure 3.9
Figure 3.10
Figure 3.11
Figure 3.12
Figure 3.13
Figure 3.14
Figure 3.15
Figure 3.16
Figure 3.17
Figure 3.18

LIST OF FIGURES

Civil Liberties Cases, 1940-1995 ................................................................ 31
Civil Rights Cases, 1940-1995 ................................................................... 31
First Amendment Cases, 1940-1995 .......................................................... 32
Due Process Cases, 1940-1995 ................................................................... 32
Privacy Cases, 1940-1995 ........................................................................... 33
Supreme Court Civil Liberties Cases, 1946-1995 ...................................... 34
Courts of Appeals Civil Liberties Cases, 1940-1988 ................................. 34
Economic Cases, 1940-1995 ....................................................................... 35
Criminal Procedure Cases, 1940-1995 ....................................................... 35
Union and Labor Cases, 1940-1995 ............................................................ 36
Attorneys Cases, 1940-1995 ...................................................................... 36
Federalism Cases, 1940-1995 ..................................................................... 37
Interstate Relations Cases, 1940-1995 ........................................................ 37
Federal Taxation Cases, 1940-1995 ........................................................... 38
Miscellaneous Cases, 1940-1995 ............................................................... 38
Supreme Court All Cases, 1946-1995 ........................................................ 39
Courts of Appeals All Cases, 1940-1988 .................................................... 39
Innovation Accounting of Civil Liberties Cases (Differences) ................ 123
Innovation Accounting of Civil Liberties Cases (Levels) ........................ 124
Variance Decomposition of Civil Liberties Cases (Differences) ............. 125
Variance Decomposition of Civil Liberties Cases (Levels) ..................... 126
Innovation Accounting of Economics Cases (Differences) ...................... 127
Innovation Accounting of Economics Cases (Levels) .............................. 128
Variance Decomposition of Economics Cases (Differences) ................... 129
Variance Decomposition of Economics Cases (Levels) ........................... 130
Innovation Accounting of Criminal Procedure Cases (Differences) ........ 131
Innovation Accounting of Criminal Procedure Cases (Levels) ................ 132
Variance Decomposition of Criminal Procedure Cases (Differences) ..... 133
Variance Decomposition of Criminal Procedure Cases (Levels) ............. 134
Innovation Accounting of Union and Labor Cases (Differences) ............ 135
Innovation Accounting of Union and Labor Cases (Levels) .................... 136
Variance Decomposition of Union and Labor Cases (Differences) ......... 137
Variance Decomposition of Union and Labor Cases (Levels) ................. 138
Innovation Accounting of Attorneys Cases (Differences) ........................ 139
Innovation Accounting of Attorneys Cases (Levels) ................................ 140

xii

Figure 3.19
Figure 3.20
Figure 3.21
Figure 3.22
Figure 3.23
Figure 3.24
Figure 3.25
Figure 3.26
Figure 3.27
Figure 3.28
Figure 3.29
Figure 3.30
Figure 3.31
Figure 3.32
Figure 4.1
Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8
Figure 4.9
Figure 4.10
Figure 4.1 1
Figure 4.12
Figure 4.13
Figure 4.14
Figure 4.15
Figure 4.16
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4

Variance Decomposition of Attorneys Cases (Differences) ..................... 141

Variance Decomposition of Attorneys Cases (Levels) ............................. 142
Innovation Accounting of Federalism Cases (Differences) ...................... 143
Innovation Accounting of Federalism Cases (Levels) .............................. 144
Variance Decomposition of Federalism Cases (Differences) ................... 145
Variance Decomposition of Federalism Cases (Levels) ........................... 146
Innovation Accounting of Interstate Relations Cases (Levels) ................. 147
Variance Decomposition of Interstate Relations Cases (Levels) .............. 148
Innovation Accounting of Federal Taxation Cases (Differences) ............. 149
Innovation Accounting of Federal Taxation Cases (Levels) ..................... 150
Variance Decomposition of Federal Taxation Cases (Differences) .......... 151
Variance Decomposition of Federal Taxation Cases (Levels) .................. 152
Innovation Accounting of Miscellaneous Cases (Levels) ......................... 153
Variance Decomposition of Miscellaneous Cases (Levels) ...................... 154
Innovation Accounting of Civil Rights Cases (Differences) .................... 186
Innovation Accounting of Civil Rights Cases (Levels) ............................ 187
Variance Decomposition of Civil Rights Cases (Differences) ................. 188
Variance Decomposition of Civil Rights Cases (Levels) ......................... 189
Innovation Accounting of First Amendment Cases (Differences) ........... 190
Innovation Accounting of First Amendment Cases (Levels) .................... 191
Variance Decomposition of First Amendment Cases (Differences) ......... 192
Variance Decomposition of First Amendment Cases (Levels) ................ 193
Innovation Accounting of Due Process Cases (Differences) .................... 194
Innovation Accounting of Due Process Cases (Levels) ........................... 195
Variance Decomposition of Due Process Cases (Differences) ................. 196
Variance Decomposition of Due Process Cases (Levels) ......................... 197
Innovation Accounting of Privacy Cases (Differences) ........................... 198
Innovation Accounting of Privacy Cases (Levels) ................................... 199
Variance Decomposition of Privacy Rights Cases (Differences) ............. 200
Variance Decomposition of Privacy Cases (Levels) ................................ 201
Aggregate Economic Cases, 1940-1995 ................................................... 244
Judicial Liberalism, 1952-1990 ................................................................ 244
Liberalism of the President and Congress, 1954-1992 ............................. 245
Public Mood, 1953-1992 .......................................................................... 245

xiii

T“;hr*‘ 1“ 1'-

. , '.
ut...~‘c¢\s1 1a --

ae genie of “ t

I
H

o - ‘ ‘
I, :I-“ '. ',,;¢-
"sunny; .4.“

U

‘ l’ I' . .
"K «a. | _. a .
r‘sfid\ ;\ 11LIe§ .“

‘ ‘ 1- .
‘.o.‘ .' l.t .‘
Kauai-l “QR-n.4,.

 

1'1".” “P'H \ln .
.Ls. “1.5.133 U

:1? vs.
U6 u l\ \c

. ,,
g. 'm'tsn:

- .'
ub_

ht~q
3*
\‘ A": - 9
“‘15 3\ 3r 4
o ‘(M ‘

(__
:51

INTRODUCTION

Setting an agenda is an exceedingly important responsibility of any political .
institution. In addition to allocating the resources of those currently before the institution,
the agenda of a political institution defines its priorities and future events, as conflicts are
designated, alternatives are clarified, power is exercised and allocated, and the direction of
public policies is delineated. Perhaps, then, setting its agenda is the most important act any
political institution exercises. This contention is as applicable to the judiciary as it is for the
other branches of government.

For these reasons, setting the judicial agenda conceivably is the most critical stage
of the policy process, as it is the stage at which private and public entities interact with the
government as they seek to rectify their complaints. At the same time, if a case survives on
the judicial agenda, decisions by courts on the issues presented have broad, policy-related
effects beyond the distinct consequence on the particular parties before the judiciary. Thus,
studies on agenda setting in the courts are vital if we are to learn about the nature of these
effects by and on the courts, the government, and society.

Acknowledging this fact, there have been myriad studies on the United States
Supreme Court. Nevertheless, most of this research considers the Supreme Court in
isolation, judging the incidences when the Court is more likely than not to decide a case.
Moreover, these studies usually consider the Supreme Court and only the Supreme Court,

1

DS Tr-ff a. . . u. --

-»u: a

a Q
sax-fl' "F.,:"" '
Una-"Jun. un-

' r
“I“ .394-
bl‘ “HUmT v- \
I“.

 

{Yeh‘ ‘ a
.1, 3 ' -
~ the §_‘n-.m

as other political institutions are ignored. That is, there is little research on the agendas of
courts other than the Supreme Court; and, few studies compare the Supreme Court’s agenda
with that of other courts or political institutions, the repercussion being that studies
comparing the effects between the Supreme Court and other courts have not taken place.
Notwithstanding this perceived bias in the research, the Supreme Court does not exist
totally in isolation. As a result, it is plausible that these other political intuitions in general
and courts in particular might influence the Supreme Court, just as the Supreme Court’s
agenda may affect other political institutions. Specifically, is there any interactive effect
between the agenda of the Supreme Court and that of the United States Courts of Appeals,
which endure just below the Supreme Court in the hierarchical system of federal courts?
Since the Supreme Court has discretionary jurisdiction, while the Courts of Appeals have
mandatory jurisdiction, it seems feasible that the Supreme Court will influence the agenda
in federal appeals courts; moreover, whereas it is constrained to an extent by institutional
rules and structures, in turn the Supreme Court could be influenced by the agenda of the
courts from which the majority of its cases stern. These issues will be addressed in some

detail herein, in an attempt to discover the nature of agenda setting in the federal courts.

CHAPTER 1

THE NATURE OF AGENDA IN THE FEDERAL COURTS: A REVIEW

One of the most important acts anypolitical institution can engage in is setting its
agenda. Indeed, the process of setting an agenda embraces the entire array of an institution’s
possible actions, including allocating resources of those currently before the political
institution, defining its priorities and future events, designating conflicts, clarifying
alternatives, exercising and allocating power, and determining the direction of public policies
(Schattschneider 1960; Kingdon 1984; Riker 1993). As Schattschneider (1960, 68)
provided: "the definition of the alternatives is the supreme instrument of power. " Perhaps,
then, setting its agenda is the most important act any political institution exercises.

The significance of agenda setting is as befitting to the judiciary as it is for the other
political institutions. For an institution which has discretion over its own agenda, as does
the United States Supreme Court, the process of agenda building and setting is far from
random. In fact, the Supreme Court’s agenda is determined by both internal and external
influences and constraints on the institution (Pacelle 1991; Epstein and Segal 1997).
Moreover, once the Court establishes its priorities through its agenda, the effects of those
choices reverberate throughout the judiciary, the legal profession, other political institutions,
even society (Caldeira and Wright 1988). Thus, studies on agenda setting in the courts are

3

vital if we are to learn about the nature of these effects by and on the judiciary, the judicial
process, other political institutions, and society.

One of the institutions potentially affected by the Supreme Court’ s choices on agenda
is the United States Courts of Appeals. The jurisdiction of the Courts of Appeals essentially
being mandatory, they must accept nearly all appeals properly made, a consequence of which
is that the circuit courts have almost no control over their agenda considerations. The agenda
of the appeals courts, therefore, is determined almost entirely by external forces including,
among other things, the Supreme Court. At least, that is one of my arguments posited in this
study. In particular, I argue in part that changes in the Supreme Court’s agenda over time
signal rational litigants, interest groups, and lawyers that the Court is becoming more or less
inclined to review particular types of cases. As a result, the agenda of the Courts of Appeals
will fluctuate in line with the dynamic agenda priorities set by the Supreme Court, as
litigants and their representatives will increase or decrease their docket filings at the Courts
of Appeals accordingly.

Since the jurisdiction of the Supreme Court is almost wholly discretionary, and due
to the importance of this institution atop the judicial hierarchy, public law scholars generally
have found the nature of its agenda more interesting than that of the federal appeals courts;
consequently, there have been scores of studies on the agenda of the Supreme Court, yet
there have been few inquiries on the agenda of the Courts of Appeals. Further and more

significantly, there has been no apparent research comparing the agendas of these federal

 

 

courts, let alone whether the agenda of one has any influence, systematic or otherwise, on
the other.I

This research is an exploration in that direction, as I seek to determine the effects on
the agendas of these institutions. In particular, this study examines the extent to which the
agenda of the Supreme Court inspires changes in the agenda of the Courts of Appeals, and
vice versa, in an attempt to ascertain the exogenous and endogenous nature of the agendas
of these institutions, including both the influences which drive their agenda priorities and the

effect each court has with respect to the other.

THE RELEVANCE OF AGENDA FORMATION AND POLITICAL INSTITUTIONS

Setting a political institution’s agenda is of paramount import. As Hammond (1986,
381-3 82) stated in a non-judicial context: "in some cases . . . whoever controls the agenda
controls the outcome." See e. g., McKelvey (1976; 1981); Plott (1976); Plott and Levine
(1978). Since agenda formation affects or may even determine the output of any political
institution, issues relating to agenda inevitably are linked to the process of policy making.
Since setting an agenda is an integral part of the policy process for any political institution,
application of the policy literature on agendas easily can be extended to an institution such

as the courts, which regularly decides important political and societal issues. Accordingly,

 

IIn fact, the only research of which I am aware which looks into judicial agendas
while taking into account the preferences of, or constraints set by, another political institution
is the preliminary study by Epstein and Segal (1997), who compared the potential actions by
Congress on the Supreme Court’s process of agenda setting.

5

it is appropriate to begin with a discussion of some of the major works on the policy process,
particularly with respect to agenda development.

In his seminal study Agendas, Alternatives, and Public Policies, Kingdon (1984;
1995) suggested that the manner in which agendas are set and alternatives specified is
influenced by both systematic and random factors, particularly when it comes to selecting
public policies on the agenda. With respect to the nonrandom determinants, Kingdon
asserted that for each policy area there are separate streams, each of which affects agendas
and resulting policies. First, there is the political stream, which constitutes electoral and
partisan considerations, institutional and interest group pressures, and public opinion. The
policy stream is composed of an amalgamation of proposals, some more prominent than
others, advanced to solve society’s problems. Finally, such societal ills, how they are
perceived over time by both the masses and political elites, their prominence, and whether
they can be solved by the available alternatives, are found within the problems stream.
These streams generally run independently of each other until they are "coupled" (1984, 1 82)
due to some change arising from the political and/or problem stream.2 When the streams
finally meet, Kingdon argued, policy windows open which yield an occasion for particular
measures and, consequently, a potentially changing agenda

Simply because a policy window opens does not ensure that certain policies will be
enacted, let alone be advanced onto the agenda. First, policy windows "stay open for only

short periods" (174); moreover, they do not always open in a predictable fashion.

 

2As a consequence, the political and problems streams pertain to the process of
setting the agenda, while the policy stream concerns the available alternatives.

6

Accordingly, the participants of the policy process must take expedient advantage of the
policy window when it Opens. These participants, known as policy entrepreneurs, are
"advocates for proposals for the prominence of an idea" (129). As such, policy entrepreneurs
embrace and polish their policy proposals, awaiting the opportunity to project them when a
policy window opens. Once the process is in motion during the short-lived window of
opportunity, the entrepreneurs add their proposals to the debate, aspiring for, and
championing, their adoption. Whether or not policy entrepreneurs are successful depends
upon issues such as bargaining, majority coalition and consensus building, and resource
commitment.

Kingdon’s thesis on agendas and public policy has been exceptionally influential
among policy scholars. More recently, in Agendas and Instability in American Politics,
Baumgartner and Jones (1993) built on Kingdon’s work while they provided a somewhat
different yet valuable perspective concerning the manner in which agendas are determined.
In the heart of their analysis, they contended that the policy process is characterized by both
long-term stability and short-tenn change. That is, protracted periods of incrementalism are
displaced by abrupt shifts in the policy landscape. Baumgartner and Jones referred to this
ebb and flow of stability and change in politics as "punctuated equilibria" (18), suggesting
that inquiry for a general equilibrium in the policy process presumably would be in vain.

One of the prime contrasts between the study by Kingdon and that by Baumgartner
and Jones is the latter supported their theoretical analysis with credible empirical verification
and results. Accordingly, their evaluation of agendas and the policy process was
generalizable across many distinct policy arenas or, as they prefer, policy subsystems.

7

Nevertheless, both of these perspectives specifically concerned the non-judicial branches of
government, as each pertained primarily to the manner in which agendas are set and public
policy is made via legislative votes and presidential decisions. As a consequence, the query
inevitably becomes, can the theoretical positions and analyses by these policy scholars be
applied to the manner in which agendas are set and policies made in the courts? I contend
that not only are these perspectives applicable to a study on judicial agendas, but that it is
both beneficial and vital that they not be ignored in any such exploration. Indeed, both
judicial and policy scholars have argued that these respective approaches to agendas and
policy can be applied to the courts (see, e. g., Perry 1991b; Barclay and Birkland 1997),
especially since there is little disagreement at this point among judicial scholars and other
political scientists that judges are political actors who seek to impose their policy preferences
on society (Rohde and Spaeth 1976; Rohde 1972; Murphy 1964).3 As argued by policy
scholars: "A political institution deciding on issues of policy fits the parameters of a
traditional policy making institution" (Barclay and Birkland 1997, 9). Since it seems
Obvious that this statement includes the courts, the policy literature is appropriate for any
study on judicial agendas.

How then should these scholarly approaches to the policy process be adopted to this
study on the federal courts? To begin, since the 19305 the Supreme Court’s agenda has

developed from being populated predominantly with economic regulation cases to having

 

3Of course, legal, institutional, and social constraints place limitations Of varying
degrees upon judges in their ability to promote their favored policies to the extent possible
(Brace and Hall 1993; Caldeira and Wright 1988; 1990; Songer 1987), although these checks
are much less apparent or realistic at the Supreme Court than at other courts (Segal 1997).

8

a more extensive emphasis on civil liberties, although this transformation has not been
entirely smooth or without peril. Under these circumstances, the relevance of issues such as
punctuated equilibria and the coupling of policy streams seems axiomatic. With respect to
the judicial process, for instance, policy entrepreneurs could be a Supreme Court justice, the
Solicitor General, the Chair of the Senate Judiciary Committee, litigants and their lawyers,
or interest groups, among others, each of whom vies to advance particular issues onto the
agendas of the courts. Their success depends, in part, on the extent the political and problem
streams provide for an open, relevant policy window (see Pacelle 1991). Or, as Baumgartner
and Jones (1993) suggested, groups seek out certain venues, or institutions, which might
afford them a more favorable hearing (Barclay and Birkland 1997); resort to the courts by
litigants, interest groups, and their representatives seems pertinent in this regard, especially
when the Supreme Court signals that it is becoming more amenable to particular types of
issues.

While scholarly research on agendas and policy in general, such as those by Kingdon
or Baumgartner and Jones, have tended to ignore the courts in their analyses, it does not
follow necessarily that judicial research has neglected either agenda or policy issues as well.
To the contrary, there has been a good deal of notable research on agenda issues in the
courts.4 Nevertheless, more needs to be accomplished. As Schattschneider provided (1960,

68): "[T]he definition of the alternatives is the supreme instrument of power . . . He who

 

4For instance, some judicial research has incorporated various concepts stemming
from the policy literature (see, e. g., Hackman 1966; Barker 1967; Olson 1984; Caldeira and
Wright 1988; Pacelle 1991).

determines what politics is about runs the country, because the definition Of the alternatives
is choice of conflicts, and the choice of conflicts allocates power." Since, as I argue, these
concepts are as applicable for the judiciary as for other political institutions, now it is time

to turn to what we know about judicial agendas.

AGENDA SETTING IN THE COURTS

Public law scholars who have studied agenda formation unquestionably champion
the magnitude of such research. For instance, Caldeira and Wright (1988, 1109) contended
that "the decision to review a case ranks as important as — if not more important than - the
decision on the merits." Pacelle asserted (1991, 23): "At the broadest level, the agenda [of
the Supreme Court] is an integral part of the policymaking process." And, Perry (1991b,
23 5) noted: "Deciding to decide is one of the most important tasks performed by any political
institution. "5 Still, public law scholars must concede that, despite these and other studies,
all important in their own right and representing solid research agendas, analysis of agenda
issues concerning the courts "has not been commensurate with its importance" (Perry 1991 a,
5). Nonetheless, while I believe that more research needs to be performed, this section
reviews much of the past literature on agenda setting in the courts.

Research on issues pertaining to judicial agendas has focused chiefly on the Supreme

Court; and, most of this scholarship has concentrated solely on one (albeit consequential)

 

SSimilar assertions are made by those studying other political institutions. For
instance, Light (1991, 105) insisted: "Selecting alternatives is perhaps the single most
important step in the presidential policy process. . . More than any other choice, the choice
of alternatives determines who gets what when, where, and how."

10

element of agenda formation, the manner in which the Court accepts individual cases.
Accordingly, the earliest studies on the Court’s case selection process were accomplished
by Schubert (1959, 1962), who utilized formal theory and voting patterns on the merits to
argue that justices voted strategically when deciding whether or not to grant writs of
certiorari (hereafter, "cert.") in order to obtain preferred decisions on the merits. In
particular, Schubert asserted that a bloc of justices could force the Court to grant cert. in
order to achieve the bloc’s preferred policy results on the merits. While his game-theoretic
approach was critical in alerting scholars to the study of judicial agenda issues, Schubert
offered no empirical tests of his model.‘5

Other pioneers of this variety of agenda research include Tanenhaus, et. a1. (1963),
who used "cue theory" to explain that cert. more likely would be granted by the Supreme
Court given a cert. petition comprising certain signals. That is, they propounded that
particular cues which were displayed in a petition would alert the justices to their relevance,
thereby increasing the potential of cert. being granted. As they posited (l 18): "The presence
of any one of these cues would warn a justice that a petition deserved scrutiny. If no one
were present, on the other hand, a justice could safely discard a petition without further
expenditure of time and energy." Specifically, Tanenhaus and his colleagues postulated that
the appearance of any of the following cues would lead to a greater incidence of the Supreme
Court granting cert. in particular cases: 1) if review was sought by the federal government;

2) if either conflict among the circuits or dissent among the judges on the court below was

 

6But see Provine (1980), who utilized Justice Burton’s papers to test, and potentially
undermine in part, Schubert’s theory.

11

present; or 3) if either a civil liberties or an economic case was at issue. Scrutinizing the
dockets for the Court’s 1956 through 1958 terms, they charged that all of these cues led to
a greater incidence of case selection, save for the hypothesized cues on economic cases.7
Ensuing research variously has substantiated much of both Schubert’s and
Tanenhaus’ analyses. With respect to Schubert’s claim that cert. votes were associated with
final votes on the merits based on the justices’ policy preferences, Baum (1977) concluded
that cert. decisions were attitudinally-based, as judges are more likely to vote to hear and
decide a case when they disagreed ideologically with the decision in the court below. In that
same vein, Jucewicz and Baum (1990, 133) offered that "the justices followed primarily the
value-based model of case-screening." Brenner and Krol (1989) found evidence that
justices’ votes on cert. and the merits were connected in a number of ways, such that they
were more likely to grant cert. when they disagreed with the decision below or when they
were part of the ideological majority of the Court. Krol and Brenner (1990, 340-342)
subsequently asserted, contrary to a part of their initial findings, that cert. votes likely were
not related to a justice’s expectation of being in the ensuing majority coalition. Nonetheless,
Boucher and Segal (1995, 83 5) then contended that justices do act strategically when voting
on cert. petitions, as they "cast an eye toward the Court’s behavior on the merits."
Additionally, a number of other researchers have verified assorted portions of Schubert’s

examination, though each from a different perspective and to varying degrees of concordance

 

7It should be noted that such a research design clearly is problematic, as it is based
on selection on the dependent variable (see King, Keohane, and Verba 1994). Once the
papers of several retired justices subsequently were released, however, this statistical
dilemma was elucidated to a degree.

12

(see, e. g., Ulrner 1972; Songer 1979; Palmer 1982; Perry 1991a; and Caldeira, Wright, and
Zorn 1996).8

Similarly, several judicial scholars have determined that the presence of certain
factors induced an elevated likelihood of cert. being granted by the Supreme Court, thus in
part ratifying cue theory and the findings by Tanenhaus and his associates. In particular,
researchers have determined that conflict among circuits (U lmer 1983; 1984), review sought
by the solicitor general (Segal 1988), filing of arnicus curiae briefs (Caldeira and Wright
1988) and the presence of particular professional lawyers in a case (McGuire and Caldeira
1993) all ushered a higher probability of the Court granting cert.9

As the above discussion makes clear, the assertions and findings by these innovators
of judicial agenda research generally have been validated by many and diverse researchers. ‘0
Nonetheless, one important study on case selection at the Supreme Court critically
questioned the findings of both Schubert and Tanenhaus. Taking a legalistic approach in
Case Selection in the United States Supreme Court, Provine (1980) argued that the judicial
role restricts the articulation of personal preferences, at least in the cert. process. Utilizing

the Burton papers to test and challenge Schubert’s theory, Provine determined that his

 

8See also Rowland and Todd (1991), who ascertained that the ideology of judges is
associated with their gatekeeping decisions on standing when parties seek access to the
Federal District Courts.

S’Nevertheless, while Caldeira and Wright (1990) found that most of these factors
influenced the cases which made the Court’s ‘discuss list,’ they contended that all were not
instrumental in the decision to grant cert.

I"As Boucher and Segal stated (1995, 825): "Like so many areas in the judicial
subfield, the first work in cert. strategy belongs to Schubert."

13

power-motivated game among strategic voting blocs on cert. decisions did not comport with
empirical reality.l ‘ "The evidence that power politics was not central in case selection, where
such behavior could have been highly effective, suggests that Supreme Court justices deem
outcome-oriented voting inappropriate. It seems likely that a shared conception of the proper
role of a judge prevents the justices from exploiting the possibilities for power-oriented
voting in case selection" (172). With respect to Tanenhaus’ analysis, Provine again
employed Justice Burton’s papers, which enabled her to circumvent Tanenhaus’ problem of
selection on the dependent variable, provided for a greater number of cases to study, and
facilitated an enhanced test of cue theory. ‘2 She insisted that her results illustrated that the
specific cues set forth by Tanenhaus, et. al., necessarily did not lead to a greater incidence
of cert. being granted.

A discussion of Provine’s work on cert. decisions inevitably leads to Perry’s (1991a)
comprehensive study on the cert. process entitled Deciding to Decide: Agenda Setting in the
United States Supreme Court, where Perry also concentrated on the means of individual case
selection. To do so, he interviewed several Supreme Court justices and many of their law
clerks over a number of years, while keeping their identities anonymous, in order to
determine the manner in which the Court decides to accept and deny cases on cert. By

focusing on individual justices and the method of decision making at the cert. stage, Pen'y’s

 

“Justice Burton’s papers were released to the public after his death in 1965;
accordingly, they were not available to Schubert when he conducted his analyses, which
were published in 1959 and 1962.

12As with Schubert, Tanenhaus and his associates did not have access to the Burton
papers when carrying out their study, published in 1963.

14

primary finding was that the cert. process, while more complex than he had anticipated,
displayed a greater prevalence of legal attributes, as opposed to political characteristics
(whether ideological or strategic, such as those posited by the adherents of Schubert and
Tanenhaus), than political scientists had come to believe. This complexity revealed by Perry
is illustrated by the following excerpt from an article in which the authors argue that the
justices generally exhibited strategic behavior during the cert. process: "The essence ”of
Provine’s claim is supported by the recent work of Perry . . . who argues that cert. is
essentially a legal process, not a political one. . . Nevertheless, Perry does not deny the
existence of strategic voting, merely its prevalence" (Boucher and Segal 1995, 825).

Which view of the Supreme Court’s agenda formation is correct, then, the legalistic
one supported by Provine and, to a lesser extent, Perry, or the behavioral and strategic
approaches suggested by Schubert, Tanenhaus, and their followers? Probably both, at least
to a certain extent. As Perry (1991b, 23 8) insisted: "notwithstanding the convincing
argument by Provine that cue theory as described by Tanenhaus was disproved by the Burton
data, the idea of cues has been discarded too quickly. Few cues will be sufficiently important
by themselves to place a case on the discuss list or predict case selection with any degree of
impressiveness. Nonetheless, Tanenhaus’s basic idea of cues still has validity in describing
a decision process, though not as precisely as Tanenhaus envisioned it. "

Thus, while Provine (1980) maintained that the cert. process is essentially a legal
exercise, this is not the sole or even primary manner in which cert. decisions are made. Perry
continued (1991b, 239): "sometimes justices cast cert. votes as this literature [following upon
Schubert and Tanenhaus] describes and sometimes they do not, depending upon their interest

15

in the outcome of a case. . . But the frndings are mixed. . . Findings by various scholars are
consistent enough to know that cert. votes are related to votes on the merits, but results have
not been stable enough, or studies have not measured the same things in such a way as to be
able to say definitively, how the cert. vote and merits are related."

Review of the literature thus far demonstrates that much of the scholarship on j udicial
agendas concentrates on the Supreme Court; moreover, the core of this focus almost entirely
is on the cert. process of individual case selection, with no definitive portrayal thereof.
Nevertheless, as Perry (1991b, 244) admonished: "One way to become broader and more
theoretical, and to tie in more directly with other political science research, is to expand the
horizon from case selection to the broader concept of agenda setting." One public law
scholar who took this advice to heart in his research on the Supreme Court was Pacelle
(1991). Incorporating a distinct theoretical and methodological approach fiom the
preponderance of agenda research which had transpired before him, in The Transformation
of the Supreme Court’s Agenda Pacelle (23) focused on the Court’s agenda "as an
institutional entity rather than the summation of individual case selection."l3

Pacelle’s study encompassed a 50 year period, from the time of the constitutional
crisis of the late 19303 through the initial terms of the Rehnquist Court in the 19803. He

discovered that the alterations of the Court’s docket over time embodied a dynamic,

 

13Interestingly, the few studies which have viewed judicial agendas in broad,
institutional terms concerned either lower federal or state courts (see, e. g., Grossman and
Sarat 1975; Atkins and Glick 1976). With respect to these works, Pacelle (1991, 11)
maintained: "These studies are different in emphasis but support the propositions that
underlying social, political, and economic activities affect the complexion of the agenda and
that the agenda has an independent empirical meaning."

16

nonrandom process, such transformations being steered by landmark decisions which were
espoused when relevant policy windows opened (see Kingdon 1984). The result was an
incremental yet systematic and orderly increase in the number of civil liberties cases over
time which replaced the economic cases that earlier had predominated the Supreme Court’s
docket. Nevertheless, Pacelle argued that inertia stemming from institutional constraints
functioned to impede both wholesale agenda change and the pace of dynamic alteration of
the Court’s agenda.

Understanding agenda issues is vital, therefore, if we are to learn fully about the
nature and priorities of political institutions, including the judiciary. As Pacelle (1991, 1)
succinctly professed: "Changes in agenda priorities and judicial doctrine are a function of
institutional transformations and may reflect broader-scale changes occurring in the external
political, social, and economic environment." Yet, as I believe I have demonstrated, most of
the research on judicial agendas concerns only the Supreme Court, as there is a relative
dearth of research on the agendas of other federal courts.“ Furthermore, most of the
research on the Supreme Court’s agenda is limited to individual case selection, as few beside
Pacelle (1991) have inquired into "the broader concept of agenda setting" (Perry 1991b, 244)
or other external influences on judicial agendas. Whatever the variety of research, there are
few if any studies making comparisons, systematic or otherwise, of the Supreme Court’s
agenda to the agenda any other court, whether federal or state. Indeed, the only comparative

study of any sort of which I am aware is the preliminary one by Epstein and Segal (1997)

 

l“Howard’s (1981) study on the Courts of Appeals is a notable exception, although
lack of systematic analysis, to a certain extent, limits his implications.

17

which explored the Court’s ostensible strategic behavior on agenda setting in reaction to
potential congressional action.ls Just as studies across various courts are necessary to
determine a general theory of judicial decision making (see Hall and Brace 1995), it
behooves us to make similar comparisons in studies on issues other than decision making,
such as the manner in which judicial agendas are set. As Epstein and Segal counseled (1997,
21): "we encourage more studies on the external context of Court agenda setting."
Accordingly, it remains to be seen if, in fact, the federal judicial system "articulates as a
system" (Frankfurter and Landis 1927, 3), at least with respect to the agenda-setting nature
of the federal courts. Therefore, it is my intent that this study fill part of this void by
systematically comparing and contrasting the agendas of the Supreme Court and the Courts

of Appeals over time.

DYNAMIC CHANGES IN THE AGENDAS OF THE FEDERAL COURTS
In the previous sections I discussed the manner in which scholarly research has
viewed agendas, both in general and with specific regard to the judiciary. In this section I
address the procedures by which the Supreme Court and Courts of Appeals set their
respective agendas and the manner in which these agendas have changed over the years.
The dialogue on agendas and public policy characterized a political environment in
which issues appeared on the agenda when appropriate opportunities arose and relevant

entrepreneurs took advantage of such prospects. Nevertheless, this review did not consider

 

lSlt remains to be seen whether Epstein and Segal (1997) expand upon or follow
through on their study of external constraints on the Court’s agenda.

18

the fact that particular institutions may have additional political constraints which neither
Kingdon (1984) or Baumgartner and Jones (1993) alluded to concerning the formation of the
institution’s agenda. In particular, they did not address the practical limitations on the
federal courts’ ability to form their agendas (Epstein and Segal 1997).

Some of these restraints on agenda formation stem from the Constitution. For
instance, as provided in Article 111, Section 2 thereof,’6 no issue can be addressed on the
Supreme Court’s agenda unless it is presented to the Court as a legitimate case or
controversy. '7 Moreover, there are standards which potentially limit the development of new
issues in a case before the Court, although issue fluidity may be commonplace and tend to
limit this legal principle (McGuire and Palmer 1995). The Constitution firrther abridges the
ability of the Supreme Court to set its own agenda by confining its jurisdiction to federal
questions and cases brought under diversity of citizenship. These various constitutional
specifications and restrictions preclude the Supreme Court fiom enjoying complete and
unfettered discretion over formation of its agenda.

Likewise, a distinction exists between the Supreme Court’s original and appellate

jurisdiction. Generally speaking, the Constitution provides Congress with the ability to

 

l‘SAll references to the Constitution in this section are made to Article 111, Section 2,
unless provided specifically otherwise.

17Generally speaking, the criteria for determinating whether real case or controversy
exists include: the parties must have standing; the case must be ripe; the case must neither
be moot nor a political question; available remedies must be exhausted; the decision being
appealed from must be final; and, advisory opinions from the court are precluded (Baum
1995; Fine 1997). Nevertheless, courts usually can evade most of these technical
requirements if they so desire (Segal and Spaeth 1993).

19

institute the Supreme Court’s appellate jurisdiction through legislation, while it precludes
the legislature from modifying the Court’s original jurisdiction (Baum 1995). See also
Marbury v. Madison, 1 Cranch 37 (1803); Ex parte McCardle, 7 Wallace 506 (1869). As
a consequence, Congress has the potential to narrow further the Supreme Court’s ability to
compose its own agenda, at least with respect to its appellate jurisdiction. Since the vast
majority of cases before the Supreme Court derive from its appellate jurisdiction, not its
original jurisdiction,l8 Congress’ power to adjust the Court’s authority to hear cases, while
not absolute, is quite extensive (Baum 1995).

While Congress’ broad authority to abridge the appellate jurisdiction of the Supreme
Court hypothetically might hamper the Court in its ability to construct its own agenda, in
reality Congress’ revisions of the Court’s appellate jurisdiction over the years instead have
afforded the Court with an enhanced ability to fashion its agenda. This has been
accomplished by a series of laws which have diminished the category of cases the Supreme
Court must hear (i. e., the Court’s mandatory jurisdiction) while expanding the Court’s power
to select the cases it wishes to hear (i. e., its discretionary jurisdiction). Most of the
legislation reducing the Court’s mandatory jurisdiction has been grounded primarily upon
perceived workload concerns of the Court. As well, each law was passed with the Court’s

consent, and many times in light of strong judicial lobbying.19

 

18The few cases decided under the Supreme Court’s original jurisdiction usually have
arisen as disputes between two states (Baum 1995; Carp and Stidham 1991).

1""In May 1988, Chief Justice Rehnquist wrote that the elimination of the Supreme
Court’s mandatory jurisdiction was ‘the primary legislative goal of the Court’" (Carp and
(continued...)

20

The Judiciary Act of 1925 was the first major legislative effort specifically designed
to lessen the Court’s workload by amending its appellate jurisdiction, as this law eliminated
many of the issues which previously arose under the Court’s mandatory jurisdiction. More
importantly, the Judges Bill (as it is commonly known) created the writ of certiorari and
required that the majority of cases decided by the Court derive from such writs. The
establishment of writs of cert. essentially granted the Court with discretionary jurisdiction,
a power it did not have ordinarily prior to enactment of the Judges Bill. That is, instead of
being compelled to decide all cases properly brought under writs of appeal, the Supreme
Court now could grant cert, and thus decide cases, on a discretionary basis.2o

Subsequent to the Judges Bill, a number of laws was passed by Congress which
continued to increase the Operation of the discretionary cert. process while abating the
availability of mandatory writs of appeal. This legislative trend culminated with passage of
the Judicial Improvements and Access to Justice Act of 1988, which for all practical
purposes eliminated writs of appeal, leaving the cert. process as the principal means by
which a case could reach the Supreme Court (see Baum 1995; Carp and Stidham 1991;
Pacelle 1991; Perry 1991a). Ergo, from 1925 through 1988, Congress fundamentally

transformed the Supreme Court’s jurisdiction "from primarily mandatory to almost entirely

 

‘9(...continued)
Stidham 1991, 43; quoting fiom, The Third Branch 1988).

2°As Justice Murphy elucidated: "Writs of certiorari are matters of grace." Wade v.
Mayo, 334 US. 672, 680 (1948). Of course, the procedure by which the Supreme Court now
grants cert. is the rule of four, whereby the approval of a minimum of four justices is
necessary for the Court to decide a case on its merits; if less than four justices want to hear
a case, then cert. is denied, leaving to stand the decision below.

21

discretionary" (Jucewicz and Baum 1990, 123). The integral consequence of these changes
in the Supreme Court’s discretionary jurisdiction is a heightened, indeed nearly absolute,
capacity to determine its own agenda, circumscribed only by the constitutional requirements
that cases and controversies, brought under the auspices of either a federal question or
diversity, be decided by the Court.”

To the contrary, the jurisdiction of the Courts of Appeals, and thus the manner in
which the agenda of these courts is defined, is very disparate from that of the Supreme Court.
Article 111, Section 1 of the Constitution provides Congress with authority to establish federal
courts inferior to the Supreme Court. Although the Constitution does not mandate Congress
to do so, in 1891 the modern Courts of Appeals were established with enactment of the
Circuit Courts of Appeal Act. As a result of this legislation, nine circuits courts were
formulated. The Court Of Appeals for the District of Columbia Circuit was incorporated into
this system in 1893; finally, the Courts of Appeals for the Tenth and Eleventh Circuits were
added in 1929 and 1981 when existing circuits were divided (Howard 1981; Carp and
Stidham 1991; Stidham, Carp, Songer and Surratt, 1992). Accordingly, a dozen separate,

geographically-based federal Courts of Appeals exist today.22

 

2'The Supreme Court cannot initiate cases; like most other courts, it must await cases
to be brought to them. Notwithstanding this technical, potentially limiting requirement, this
likely does not handicap the Court’s potential to shape its own agenda, simply due to the fact
that in recent years the Court has received nearly 7,000 petitions per term, of which it has
decided approximately 100 by full, written opinions (Baum 1995; Carp and Stidham 1991).

22Pursuant to the Judiciary Act of 1789, the First Congress created the initial federal
circuit courts. These courts remained active in various forms until they were abolished in
1911, after becoming obsolete with passage of the 1891 law (Carp and Stidham 1991).

22

In addition to the power to create these inferior courts, the Constitution necessarily
provides Congress with the capacity to assign the appellate jurisdiction to these courts.
Through a variety of laws, Congress accordingly has defined the jurisdiction of the Courts
of Appeals to be predominately mandatory.23 Thus, subject to the constitutional
requirements that federal courts decide only cases or controversies characterized by a federal
question or diversity, the Courts of Appeals must accept and hear each and every properly
appealed case.

The difference between the Supreme Court and the Courts of Appeals with respect
to jurisdiction, therefore, is stark. One consequence of this distinction has been an incredible
expansion of the number of cases commenced in, and decided by, the Courts of Appeals.
The number of cases filed in the Courts of Appeals increased from about 15,000 in 1973 to
over 50,000 in 1993 (Carp and Stidham 1996). Songer ( 1991, 36) similarly reported that
there has been a "sixty-seven-fold increase [in docketed cases] since 1895 and a ninefold
increase just in the twenty-seven years from 1960 to 1987." In contrast, the Supreme Court
issued a mere 84 written opinions in its October 1993 term. It accordingly has been argued
that the Courts of Appeals are effectively courts of last resort (Howard 1981).

Furthermore, with respect to agenda formation, this glaring distinction in jurisdiction
between these federal courts signifies that the Supreme Court has the ability to, and does,

construct its own agenda. While the justices may be constrained in ways political actors in

 

23The mandatory jurisdiction of the Courts of Appeals consists of two broad
categories of appeals: 1) ordinary civil and criminal appeals from the federal district courts,
see 28 US. C. §§ 1291, 1292; and 2) administrative appeals, whether Of rule making or
adjudicative character, see A.P.A. §§ 552b, 702 (Carp and Stidham 1991).

23

the other branches of government are not (Epstein and Segal 1997), for all practical purposes
the Supreme Court has considerable discretion to set its own agenda. The federal appeals
courts, however, remain in the position the Supreme Court was in before enactment of the
Judges Bill in 1925, where "agenda reflected the cases and issues litigants considered
important" (Pacelle 1991, 62). In other words, circuit court judges have no authority over
the mode by which the agenda of their courts is demarcated, as those decisions instead are
made by those who seek access to the appeals courts, i. e., the parties. Thus, the Supreme
Court justices appear to possess the "supreme instrument of power" (Schattschneider 1960,
68) to determine the Court’s agenda, whereas the various judges sitting on the Courts of
Appeals must accept all cases properly brought before them.

The next subject, then, is to examine the agendas of these courts and observe how
they have been modified over the years. That is, which issues has the Supreme Court
emphasized, which has it ignored, and has this pattern changed over time? Similarly, what
types of cases have parties considered to be meaningful enough to petition the Courts of
Appeals? I employ the term "issues" here because a court’s agenda is composed essentially
of the disparate issues which appear in the cases on its docket. The more significant an issue
is to those who affect the agenda, then the greater the number of times that issue will appear
on the court’s agenda relative to other issues. Since absolute numbers provide no relative
or meaningful measure of agenda change, I will note the percentage of the various issues on
these courts’ agendas per year and the manner in which such emphases have changed over

time.

24

The following types of issues comprise the vast majority of cases which shape the
agendas of the federal courts: civil liberties (which specifically includes civil rights, first
amendment, due process, and privacy); economic regulation; criminal procedure; union and
labor; attorneys; federalism; interstate relations; federal taxation; and miscellaneous.
Accordingly, these are the issues I will examine throughout this study as I attempt to
discover the various influences of agenda changes in both the Supreme Court and the Courts
of Appeals. A review of the following graphs demonstrates, in a summary fashion, the
changing accentuations by the agenda-setters on each of these courts.“

To begin, I will look at civil liberties, a topic which has garnered much attention
among political scientists, judicial and legal scholars, the media, and the public. Figure 1.1
demonstrates the dynamics of this issue in both the Supreme Court and Courts of Appeals.
The Supreme Court decided a relatively stable percentage of civil liberties cases from the
mid-19408 through the 19505, while there was quite an increasing emphasis on these cases
in the 19608 which peaked in 1971 and steadily decreased since then. The mean percentage
of civil liberties cases decided by the Supreme Court during this entire time period (from
1946 through 1995) was just under 30 percent. Parties apparently did not value civil liberties
cases with the same significance as Supreme Court justices, since the percentage of these
cases was less in the appeals court than in the Supreme Court for every year in the series.

Nevertheless, there was an analogous, steadily increasing trend of civil liberties cases in the

 

2"These issues are defined by the coding methods utilized by the United States
Supreme Court Judicial Database. The manner in which I operationalized the variables in
order to formulate these graphs is set forth in Chapter 2 and Appendix A.

25

 

 

4

cm courts;
in die Suprem:
fresh 1083
Since
individual lib.
these issues r
m civil ri $5.15

Ill 1h: SUPR-

mm 0f abet
185mm" Vel}

in the Super:
“1"th bfigx
the man PC:

Cases in the
less 111m In

and Speech
”a“ 011151
in 3,, al‘pe
respeqb‘e‘

Iy‘
auflbuzh ‘

circuit courts; yet, it appeared to defy the downward inclination in the most recent history
in the Supreme Court. The mean percentage of these cases the Courts of Appeals from 1940
through 1988 barely rose above ten percent.

Since civil liberties cases are made up of a variety of issues dealing with personal or
individual liberties, I need to examine each of these issues separately. The manner in which
these issues have varied in these courts over the time period is represented in Figures 1.2
(civil rights), 1.3 (first amendment), 1.4 (due process), and 1.5 (privacy). The general trends
in civil rights cases, particularly the upward tendencies but also the recent decreasing trend
in the Supreme Court, closely followed those for civil liberties cases in both courts, with a
mean of about 15 percent and seven percent in the Supreme Court and Courts of Appeals,
respectively (see Figure 1.2). First Amendment cases demonstrated two increasing slopes
in the Supreme Court, the first which peaked in the early years of the Burger Court, the other
which began a few years later and crested at the dawn of Chief Justice Rehnquist’s tenure;
the mean percentage was slightly less than ten percent. In contrast, the percentage of speech
cases in the Courts of Appeals remained relatively low and stable; mean percentage here was
less than two percent (see Figure 1.3).

Due process cases, on the other hand, followed a dissimilar pattern from the rights
and speech cases just described, as due process cases seemed to have peaked in the early
years of this study in the Supreme Court, while a very slight increasing trend was noticeable
in the appeals courts. Averages here for each court were about five percent and two percent,
respectively (see Figure 1.4). Finally, privacy cases were not all that popular in either court,
although some increasing movements and wider variances were apparent for both courts

26

 

  
  
  

heughout the 19795
at only one percent a:
funhennore. ’2
ii ricsin order to ill-
rights cases made up
foiiowed by first 3!th
circuit courts apparen:
not in terms OIICIJIlM
The next issue

35% the Supreme
”“835 that the Cour
Coun' throughout the
Supreme C (Jun, )6
X81‘Ettlieless. while
dei‘line Rm flinch s
midem in the mom
REVisjng 91‘.

Fears. Figure 1.9

I" ,1 A

ncd
' in:
I‘ll]

throughout the 1970s. The mean percentage of privacy cases in the Supreme Court leveled
at only one percent and half of that for the Courts of Appeals.

Furthermore, Figures 1.6 and 1.7 amalgarnate all the issues which compose civil
liberties in order to illustrate collectively their dynamics. These graphs demonstrate that civil
rights cases made up the largest portion of the category of civil liberties in both courts,
followed by first amendment issues; neither the Supreme Court justices nor parties in the
circuit courts apparently placed too much emphasis on due process or privacy cases, at least
not in terms of relative emphasis.25

The next issue concerns economic regulation, the types of cases which Pacelle (1991)
asserts the Supreme Court systematically replaced with civil liberties cases. Figure 1.8
reveals that the Courts of Appeals decided more of these types of cases than the Supreme
Court throughout the series, as the mean values for economic cases is about 25 percent in the
Supreme Court, yet it is a solid 20 percentage points higher in the appeals courts.
Nevertheless, while each court displayed a decreasing emphasis of these cases, the pace of
decline was much steeper in the Courts of Appeals, although a slight upward inclination is
evident in the most recent years of this series.

Revising emphases are visible for criminal procedure cases in both courts over the
years. Figure 1.9 illustrates that for criminal procedure cases in the Supreme Court the
variance remained fairly stable about the mean of 20 percent or so. A very different

propensity was noticeable in the Courts of Appeals, as the percentage of these cases steadily

 

25In this regard, for instance, Roe v. Wade, 410 US. 113 (1973), is among the most
salient cases, privacy or otherwise, ever decided by the Supreme Court.

27

 

    
  

sedated which CUlf‘.‘
cr'rriral procedure cc.
horn er. as the empf
here was nearly 30 IX"
Next I look a:
mange of these cc»
pinged to nearly not}:
Ol‘hese cases for the cr
on union and labor ca:
means for these b(M
0m.t‘I‘ezils. respecriu
AllOme}'S C35:

illVOl'i'Cd m the mailC‘

C01,,

WE MAP-Deals 3

Stable

around 1116 mt

Siemfid 10 be an in-
\1’

“itch ‘ ‘.
us; Is jlkel ,
.\ reSpQ

escalated, which culminated in 1969 with about a quarter of its docket being composed of
criminal procedure cases; the good times in the appeals courts for these cases did not last,
however, as the emphasis on criminal procedure tumbled after this peak. The mean value
here was nearly 30 percent over the series.

Next I look at union and labor cases. In the Supreme Court, a fairly stationary
percentage of these cases was decided until the early 19808, when the percentage drastically
plunged to nearly nothing. In the Courts of Appeals, however, the relatively constant share
of these cases for the entire series demonstrates that litigants did not minimize their emphasis
on union and labor cases as Supreme Court justices apparently did (see Figure 1.10). The
means for these series were about five and eight percent for the Supreme Court and Courts
of Appeals, respectively.

Attorneys cases at times are important to the legal bar and to the specific attorneys
involved in the matter, but these cases are not too prominent in either the Supreme Court or
Courts of Appeals, as Figure 1.1 1 illustrates. Nevertheless, while the variance seemed fairly
stable around the mean of about a half percent attorneys cases in the appeals courts, there
seemed to be an increasing trend in the final 15 to 20 years of the Supreme Court series,
which is likely responsible for this mean to be about one percent for these cases.

Figure 1.12 denotes the dynamics for federalism cases. Specifically, the Supreme
Court never moved too far from its mean percentage of about four percent and, save for a few
boom years in the late 19508, the Courts of Appeals remained fairly close to its mean of less
than one percent of these cases. Similarly, neither court seemed to emphasize interstate
relations cases, although the Supreme Court reserved about one percent of its docket for

28

 

    
  
  

mess cases met the } ‘
any year in the appea.

In contrast. b
stressed cases concer'
changed met time. F -,
ofippeals early on. r.
precipitously. with the
however. exhibited \ e:
four percent in the hig‘.

mod prior to the mi.

for all the years ther

conned :5 miscellanr

1? s
a, hat notable in eir‘

M as only about or

unsolidated lash.

lifjin

Co , -
L111 “ihlle Fiuul»
C 8

ii“ ‘.
Manon earlier i
n

1
\

LDU

4- 31

Wish.) Id: lite r
”- 1e .

these cases over the years; however, the percentage of these cases rarely rose to that level in
any year in the appeals courts (see Figure 1.13).

In contrast, both Supreme Court justices and parties in the appeals courts generally
stressed cases concerning federal taxation, although the significance of these cases has
changed over time. Figure 1.14 illuminates that tax cases were very important in the Courts
of Appeals early on, making up about 20 percent of its docket, but this magnitude dropped
precipitously, with the average for this series being about nine percent. The Supreme Court,
however, exhibited very different behavior. While the mean for federal tax cases was about
four percent in the high Court, it decided a percentage of cases above the mean for the entire
period prior to the mid-19608, while its emphasis on these cases remained below the mean
for all the years thereafter. Finally, those cases not appropriate for categorization are
codified as miscellaneous cases; that is, besides defying classification, they probably are not
all that notable in either substance or numbers. Figure 1.15 corroborates this intuition in
part, as only about one percent of each court’s docket is made up of miscellaneous cases,
although there are a few years in the Courts of Appeals where the percentages appeared to
be nontrivial.

The final graphs for this chapter exhibit the agendas for both of these courts in a
consolidated fashion. That is, Figure 1.16 illustrates the changing docket for the Supreme
Court, while Figure 1.17 is its counterpart for the Courts of Appeals.26 In addition to the

discussion earlier in this chapter, what these graphs should make abundantly clear is that the

 

2“Due to the manner in which the United States Courts of Appeals Data Base was
coded, slight double counting inevitably occurred. See Appendix A.

29

agendas of both the Supreme Court and the Courts of Appeals are dynamic structures worthy
of study. In particular, for the benefit of both judicial scholars specifically and political
scientists in general, it is meaningful to determine the influences on the agenda of each court
along with the influences of each court’s agenda on the other. My purpose in the ensuing

chapters will be to ascertain the variety and extent of these effects.

30

Percent Cases on Docket

Percent Cases on Docket

 

 

 

40-
301
1 . I‘ .
A l
201 l ‘ l “\ I
41" '1 ' '4
II \I ‘1 I" court:
101 j‘ ‘ I‘II ~l’ Supreme Court
0 j g I Courts of Appeals

1940 1946 1952 1958 1964 1970 1976 1982 1988 1994
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1 . 1
CIVIL LIBERTIES CASES, 1940-1995

 

 

 

50

40‘

30:

20-

,“ ‘ Court:
\ t \” _
10' ”'\"I’ \l ‘ Supreme Court
‘ 9
I ' I I
O " A I ’ ’ ‘I ______ Courts oprpeals

 

1940 1946 1952 1958 1964 1970 1976 1982 1988 1994
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.2
CIVIL RIGHTS CASES, 1940-1995

31

1“;
I
0

‘0 \ .v
.r I m
l
8 11 m 1 m .
~0¥OOQ CO 80000 sCQULQQ

8 m
.- 1|

H:

190
30

 

Ill} ‘1]11111111111I. .
0

1

“F.“
.01000 :0 00000 .20050Q

 

 

 

 

 

 

 

30

20-
as
x
8
o
c 101
O .
0, Court.
0 —
8
0 VJ SupremeCourt
E 1‘ r‘ I" \ ,

|

g A i"\ I, ‘ z u, \"1 ""1"! —-'
0- ° A!--..‘..A.‘:.l.‘f-...\;:'::.‘!.!-----..--L': .............. C°Urt3°fAPP°a'3

1940 1949 1952 1959 1994 1970 1979 1992 1999 1994

1949 1949 1955 1991 1997 1973 1979 1995 1991
FIGURE1.3
FIRST AMENDMENT CASES, 1940-1995

30

20.
3 '4
3‘
5 10' Court:
m _
8
‘5 SupremeCourt
o .
E ‘ \ I

l - s s

& 0‘ I‘ I’ I A‘ Y L Courtsoprpeals

 

 

1940 1946 1952 19- 1964 1970 1976 1982 1988 1994
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.4
DUE PROCESS CASES, 1940-1995

32

Percent Cases on Docket

20

 

15-
10-

51

 

1940

1943

IA
1946

 

1949

     

1952 1958

Court:

Supreme Court

Courts of Appeals

 

 

 

       

1964 1970 1976 1992 199 19911
1961 1997 1973 1979 1995 1991

 

FIGURE 1.5

PRIVACY CASES, 1940-1995

33

 

Percent Cases on Docket

Percent Cases on Docket

 

 

 

50
4O .
1
30 1
4
20 q Civil Liberties:
. - Privacy
10 . - Due Process
- First Amendment
0 .Civil Rights

1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.6

 

q
1940 1946 1952 1958 1964 1970 1976 1982 1988 1994

SUPREME COURT CIVIL LIBERTIES CASES, 1946-1995

 

40-

104

 

0
1940 1946 1952 1958

     

1964 1970 1976 1982 1988

  
     

Civil Liberties:
- Privacy
- Due Process

- First Amendment

 

- Civil Rights
1994

1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.7

COURTS OF APPEALS CIVIL LIBERTIES CASES, 1940-1988

 

 

 

I Q A v ‘ I,
I. 9‘ \ \ a“. r I w‘v
i ll .9. . .r w. ..
l . 1 l
0 4i I l. | I til I. I I II I IIIIiII I I IIIIIIII p iJII Ii IIII I IIIIIII I. .I IIJIIU IJI l. ‘11."
"a. E. K X m4 n. . 49,» AW» 4% 04 3» . T; «rm 0 ..-.

«.1000 CO 00000 uCGUSOQ uoxooo C0 00000 aCUULDnK

 

Percent Cases on Docket

Percent Cases on Docket

70

 

101

0

 

 

 

70

-V'U-"n-"'--"'--"-'-""'V'Vi'fiffiVVV1iii—U'1'

1958 1964 1970 1976 1982 1988

1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.8
ECONOMIC CASES, 1940-1995

 

60-

501

20-

10-

0

 

 

1940

 

 

1946
1943

1 952
1 949

1958 1964
1955 1961

1 970
1 967 1973

1976 1 982
1 979

1 988
1 985

1994
1991

FIGURE 1.9
CRIMINAL PROCEDURE CASES, 1940-1995

35

 

Court:

1994

Supreme Court

Courts of Appeals

Court:

Supreme Court

Courts of Appeals

 

Percent Cases on Docket

Percent Cases on Docket

20

 

 

 

 

 

 

 

15-
’ 1. - 1 A

a '| l| I \l' I 1 \ ‘ l
101 I". I I I 1 l r \ I

. ,0 ‘ 1 \l | r} '1 I | 'I

I l \ I l 1 , ‘ H I I I

" t l \ l I , v‘ 1’ I "

I“: 1 I , l ,' '

. 'I . I I ‘ Court:

7 ‘I I \
5| \' —

J ' Supreme Court
0 ‘ Courts of Appeals
1940 1946 1952 1958 1964 1970 1976 1982 1988 1994

1943 1949 1955 1961 1967 1973 1979 1985 1991
FIGURE 1.10
UNION AND LABOR CASES, 1940-1995
10
84
5.
4-
Court:
2- Supreme Court

‘I (“ {‘1‘ --9
0" It A I ‘2‘ Courts oprpeals
1940 1946 1952 1958 970 1976 1988 1994

1943 1949 1955 1961 1967 1973 1979981985

FIGURE 1.1 1
ATTORNEYS CASES, 1940-1995

36

1991

Percent Cases on Docket

Percent Cases on Docket

20

 

 

 

 

15-

.l
10-

1

Court:
5‘ —
Supreme Court

: . A

4 . ,n
O A ' " \ ’ \ 2 \ ‘1‘," | Courts oprpeals

1940 1946 1'2 1958 1964 1970 1976 1982 1988 1994
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1 .12
FEDERALISM CASES, 1940-1995

 

10

8-

4:

Court:

 

2 ' Supreme Court

Courts of Appeals

  

 

 

1
1940
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.13
INTERSTATE RELATIONS CASES, 1940-1995

37

 

 

1 ‘I Iilfllll. 1'
AilU F.V

II

«Ow-000 CO 00000 «£00501

Willi- -

0 8
1|

. I} I...

 

II
6

I 4. I xiii :4

noxooo to nonmo QCDEOQ

 

Percent Cases on Docket

Percent Cases on Docket

30

 

 

3.
’\'I
’01
11
I is
I 11' i
I I
It :\' |' :\
101}: 1 l\’|
1
"I‘ \n
'\ 1
LI
0 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

 

 

Court:

Supreme Court

1940 1946 1952 1958 1964 1970 1976 V1982 1988 1994

1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.14
FEDERAL TAXATION CASES, 1940-1995

10

 

84

 

 

1946 .
1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1 .1 5
MISCELLANEOUS CASES, 1940-1995

38

 

Courts of Appeals

Court:

Supreme Court

Courts of Appeals

1.1 .9 ,1 1.1 11

«oxooo co Dunno acookea

I .‘e
o

n
H.-

11

«0&000 CO 00000 21.001401

Percent Cases on Docket

Percent Cases on Docket

 

 

1943

 

1940 1946 1952 1958 1964 1970 1976 1982 1988 1994

Types of Cases:
- Miscellaneous
- Federal Taxation
Interstate Relations
1:] Federalism
EAttcmeys
- Union and Labor
.Criminal Procedure
- Economic
- Civil Liberties
1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.16
SUPREME COURT ALL CASES, 1946-1995

 

 

Types of Cases:
- Miscellaneous
- Federal Taxation
Interstate Relations
CI Federalism
Attorneys
- Union and Labor
- Criminal Procedure

- Economic

.Civil Liberties

 

1940 1946 1952 1958 1964 1970 1976 1982 1988 1994

1943

1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 1.17
COURTS OF APPEALS ALL CASES, 1940-1988

39

CHAPTER 2

THEORETICAL RELATIONSHIPS AND RESEARCH DESIGN AND
METHODOLOGY

In this chapter I intend to present a set of theoretical arguments which explain the
relationships between the changing agendas of the Supreme Court and Courts of Appeals
over time; additionally, I will describe the methods by which I foresee addressing and testing
these conditions in the ensuing chapters. Both theory and methodology, I believe, not only
are important but are indispensable to this (or any) study, because a full portrayal cannot be
accomplished if either of these approaches is shunned or excluded. That is, a strong
theoretical foundation in the social sciences is essential in order to explain and generalize
about the substantive specifics of human events. Nevertheless, empirical implications of any
theory are critical as well (Shepsle 1995; Green and Shapiro 1994). This is particularly true,
and applicable, for studying political phenomena (Riker 1990). As the behavioral
desideraturn provides: "It is possible to understand, explain, and ultimately predict political
phenomena; the most accurate way of accomplishing this feat is by using methods tested by
science - empiricism . . . Thus, research should be theory oriented and theory directed”
(Brown 1981, 58) [emphasis in original]. As outlined in the following pages, I will

incorporate such a theory-oriented and -directed approach to this study.

40

 

ATH‘

{/7
{,3
f

mflfiw'
cohesive iii:
is a a“ '
to he stud}
mom] chC
goajfiritl‘iit
Supremf Ci
eyz' .iisi'l til:
i): parties i
"32:13 31 1h
are mom-at:
their behai'i

Apr
microecono
rational rho

irhaiior. '

Inplemcmei

indix‘idUals'

A THEORETICAL BASIS OF THE NATURE OF AGENDA IN THE FEDERAL COURTS

Since the manner in which the agendas of the Supreme Court and Courts of Appeals
are set is so different, any systematic changes in these agendas must be tied together by some
cohesive theory. The theory I utilize in this study is rational choice. Rational choice theory
has a good deal in common with behavioralism, in that each employs a scientific approach
to the study of politics. Nonetheless, there is a key difference between behavioralism and
rational choice, to wit: rational choice theory has as its base the theoretical assumption that
goal-oriented behavior drives political actors (Rohde and Shepsle 1978). With respect to the
Supreme Court’s agenda, for instance, the actors involved are the justices themselves, who
establish the Court’s docket by use Of its discretionary jurisdiction. On the other hand, it is
the parties and their representatives (whether lawyers or interest groups) which form the
agenda at the Courts of Appeals. Accordingly, this study assumes that these political actors
are motivated by their various goals, and such an underlying assumption can help explain
their behavior.

Application of rational choice theory to politics has its roots in studies on
microeconomic behavior (Riker 1990; Aldrich 1994). Indeed, the initial application of
rational choice to politics generally concerned an analogous area - issues of mass political
behavior. That is, in a manner similar to microeconomic theory, rational choice was
implemented by assuming that individual political behavior could be explained by

individuals’ maximizing their goals, such as when a citizen votes for a party which she

41

tellers wl
i nilt n T
he}; . rlt
homer. .3

elites like“

‘.
rites ens-a;
5 l

Mipoliti
eiilffi Who
midst the:

Alt

believes will optimize her benefits (see, e. g., Downs 1957).1 However, as Aldrich (1994)
adeptly propounded, political behavior is unlike economic behavior. Even within politics,
however, Aldrich asserted that mass behavior is distinct from elite behavior; in particular,
elites likely care a great deal more about the political outcomes than the masses, and elites
at times are able to affect or change the rules under which those outcomes are produced, an
ability almost entirely lacking in mass actors. According to Aldrich, then, rational choice
theory is best utilized in political science when the behavior to be studied concerns political
elites engaging in salient, high-stakes endeavors. Moreover, when these elite political actors
make political decisions within dynamic institutions which feature endogeneity (i. e., political
elites who make or even change the rules of the institution under which their decisions are
made), then rational choice proves to be exceedingly appropriate.2

Aldrich’s description of the applicability of rational choice theory assuredly
incorporates the political actors in my study. For instance, Supreme Court justices are

political elites who engage in high-stakes behavior as they pursue their judicial functions

 

1Of course, Downs (1957) was not solely interested in the behavior of individuals in the
electorate; he also was concerned with a government’ 3 maximization of political support and,
most importantly, the "mutual interdependence" between the government and the electorate
in pursuit of their respective utilities (72).

2A few examples of studies utilizing rational choice in which the research concerned
political elites making high stakes decisions in endogenous institutions include: Shepsle’s
(1978) work on committee assignments; Schlesinger’s (1966; 1991) research on political
ambition and political parties; the study by Rohde and Spaeth (1976) on Supreme Court
decision making; Rohde’s (1991) examination of parties and leaders in Congress; Krehbiel’s
(1991) informational theory of legislative organization; and, the rationale for political parties
in America by Aldrich (1995). This list, while certainly not exhaustive, is provided merely
to demonstrate situations Specified by Aldrich (1994) to be most appropriate for rational
choice theory.

42

 

pursuant to
endogeneitj
lla its OM
is able to e
{Caldeira a
docket but;
hi the COU
mil since i
illCTCbt' all;
lhfl' deClCit
“filth aliei
mi“ Nb:
Nigerian
interest grc
toms (\lci
mom-S {0!
As
brim-mt
Studying in
of the ins:
Wining“

‘ 1.;
Wing lo

pursuant to the Constitution. Moreover, the Court is a dynamic institution with features of
endogeneity. With complete oversight of its agenda, authority which it gained essentially
via its own lobbying efforts before Congress (Carp and Stidham 1991), the Supreme Court
is able to exploit the judicial process based upon the ideologies of the individual justices
(Caldeira and Wright 1990). This afi’ects not only the types of cases which reach the Court’s
docket but also the eventual decisions on the merits. In other words, different cases accepted
by the Court indicate that distinctive decisions eventually will be conveyed on the merits;
and, since it is the justices themselves who choose the cases to be heard, their own actions
thereby affect their decisions. Similarly, litigants engage in salient political behavior when
they decide whether or not to appeal particular cases to the Courts of Appeals, a process
which affects not only the parties themselves but others in the legal profession when the
judges publish their opinions in these cases. This operation of high-stakes behavior within
endogenous political institutions seems especially clear when professional attorneys and/or
interest groups pursue particular cases (yet not others) before specific judges, circuits, or
courts (McGuire and Caldeira 1993). Accordingly, rational choice theory seems remarkably
apropos for my study.

As stated, rational choice initially was employed in research on mass political
behavior. However, numerous rational choice scholars have moved away merely from
studying individuals; instead, individual behavior has been incorporated within the context
of the institutions housing these political actors and their actions. Terrned "new
institutionalism," this is the study of "political outcomes . . . [which] result from actors’
seeking to realize their goals, choosing within and possibly shaping a given set of

43

 

institutional arrangements, and so choosing within a given historical context" (Aldrich 1995,
6; citing Rohde and Shepsle 1978). On the judicial landscape, new institutionalism, or
neoinstitutionalism, has been defined as a "perspective [which] bridges the gap between the
traditional jurisprudence, attitudinal theory, and institutional analysis. Neoinstitutionalism
[sic] proposes that institutional arrangements, which consist of internal and external rules as
well as organization structures, determine the aggregation of individual preferences within
any given decision-making body" (Brace and Hall 1993, 916). Moreover, these institutional
rules and structures continue to affect political outcomes once decisions are made regarding
structure and prior outcomes. This final point, perhaps, is the key to employing a
neoinstitutional perspective - not only are the institutions themselves of maj or consequence
to individual political behavior, but it is the endogenous nature of the political actors and
their utility-maximizing behavior within these institutions which neoinstitutionalism seeks
to incorporate.

Furthermore, while these institutional rules and structures may be viewed as
constraints on individual behavior, a neoinstitutional perspective realizes that these apparent
restrictions in fact may be opportunities for the political actors within the institution; and,
even if institutional arrangements indeed are confining, in reality they may not be binding
(Aldrich 1994). As Aldrich (227) advanced: "The ‘fundarnental equation’ of the new
institutionalism is that political outcomes are the result of the interplay of actors as they seek
to realize their goals, of the institutional settings in which they act and which they may help
to shape, and of the historical context in which their decisions are set." Thus, according to
neoinstitutionalism, political outcomes are the result of individual, goal-oriented behavior

44

as determined within the limitations and opportunities afforded by political institutions in
which the political actors function.

As a theoretical foundation, then, this study incorporates a neoinstitutional
perspective with respect to the agendas of the Supreme Court and Courts Of Appeals. By
focusing on these courts and their respective agendas in this theoretical manner, hypotheses
which flow from the assumptions of rational choice and neoinstitutionalism can be posited.
Once proposed, these hypotheses can be empirically analyzed.

What, then, are some of these specific assumptions? Under neoinstitutionalism, the
initial assumptions concern basic rational choice suppositions on individual behavior. Thus,
I assume that the individuals involved will seek to further their goals to the extent possible.
At the Supreme Court, this suggests that the individual justices are political actors who
attempt to impose their own policy preferences on society and who seek to influence other
political institutions and actors. At the Courts of Appeals, individual litigants, interest
groups, and their lawyers are assumed to act rationally by utility maximization when
deciding to appeal cases, and it fiirther assumes that they desire to win cases, influence others
in the political process, or both.

The next branch ofneoinstitutionalism concerns the institutional arrangements which
help shape the aggregation of individual preferences. With respect to the Supreme Court,
the assumption is that, while individual justices are potentially restricted by institutional

rules and structures which affect individual preferences, the Supreme Court systematically

4S

determines its agenda over time.3 Thus, constitutional stipulations such as the cases and
controversies requirement, Congress’ ability to set and restrict the Court’s appellate
jurisdiction, the rule of four, and the fact that the Court must await cases to be brought to the
institution, all help to constrain the justices’ ability to set the Court’s agenda. Nevertheless,
these same arrangements can be seen as opportunities for the justices to fashion the Court’s
agenda to their own liking, as the only true and genuine institutional constraints include the
rule of four and any self-imposed restraints relating to judicial role which some individual
justices may adhere to. The institutional structures, on the other hand, are more binding with
respect to the agenda at the Courts of Appeals. Here, the assumption is that parties, interest
groups, and counsel are pressured when making decisions to appeal, such constraints usually
stemming from budgetary or strategic considerations.

By applying this neoinstitutional perspective and focusing on the agendas of the
federal courts in this manner, I am able to make several arguments on the relationship
between the Supreme Court and Courts of Appeals. The first of these arguments is that the
choices the Supreme Court makes concerning its agenda ripple throughout the judiciary.
Thus, agenda priorities generated at the Supreme Court soon will be found in the dockets of
the Courts of Appeals. In particular, when the Supreme Court changes its preferences for
various types of cases over time, signals are sent to parties and their representatives in the

courts below that the Supreme Court is becoming more or less agreeable to certain kinds of

 

3On a broader level with respect to agenda, it is assumed that "the Supreme Court is a
political, unitary actor and the variation in its behavior and policies over time can be
explained systematically" (Pacelle 1991, 13).

46

issues. Ir.

W
by patties
manna
the Supra
lines. T]
Hues xiii;
l0l refrain
"3.51) by
to Oillfii ca
10 lls agen

An
0i lesser p,
"tiled. 1
mt wi
Opposite d,

On the mill

issues. In time, the behavior of litigants is conditioned by the Supreme Court’s actions, as
parties will react accordingly by appealing cases more in line with the Supreme Court’s
changing agenda priorities. AS a result, since the agenda of the Courts of Appeals is
determined by litigious behavior, over time the agenda Of the Courts of Appeals should
fluctuate in response to that of the Supreme Court.

What are some of these signals which are sent by the Supreme Court and acted upon
by parties and their representatives in the Courts of Appeals? Essentially, they are encased
in a variety of dynamic behaviors stemming from the Court’s agenda management. First,
the Supreme Court might accept either a greater or lesser number of cases on particular
issues. The enhanced (or lessened) emphasis provided by the Court for certain types of
issues will make it more likely that litigants will recognize such trends and thereby appeal
(or refrain from appealing) those classes of law suits in the federal courts below. Thus,
Simply by changing its priorities as expressed in the percentage of cases it hears compared

to Other cases, the Supreme Court beckons litigants to its changing preferences with respect
to its agenda.

Another manner in which the Supreme Court signals parties is by issuing a greater

01' lesser percentage of liberal decisions each year. That is, not only is it the types of cases
decided by the Court which matter to litigants but also the direction in which those cases are
decided. In the long term, this signifies that a liberal trend in Supreme Court decision
making will alert parties to appeal cases more in line with such liberal doctrine, while an
0Pposite development in the lower courts Should be apparent when conservative inclinations
0n the high Court are operation. Thus, parties will reflect upon decision-making trends in

47

he Supra"
juéicia' dt
contempt“
lil‘iit‘h ii i
com ati
lexeiopin,
clearings
musinid
decisionSI
docket dll
tiring;
met

hi“: and ‘lie

the Supreme Court and appeal more of those cases which are conducive to that type of
judicial doctrine.4 In the short term, however, a different outcome may transpire. As a
contemporary Court attempts to reverse the decisional trends set by its predecessors with
which it disagrees, the Court may accept more "liberal" issues but decide them in a
conservative manner if it is becoming more conservative, and vice versa if the Court is
developing a more liberal demeanor. When such mixed signals from the Supreme Court
(hearing more liberal cases but deciding them in a conservative direction) are relayed, parties
must decide which trend is the best one to follow. Generally, the directionality of the Court’ 5
decisions will be more persuasive to appellants’ utility, so long as the types of cases on the
dOCKet drops off and falls in line with directionality at a slight lag. Under these
Circumstances, parties Should follow the changing liberalism more than the types of issues
a‘?v<>e13ted by the Court. If, on the other hand, the issues and directionality remain polar
beYond the short term, it is much more likely that parties will follow neither trend during the
time period, whether short or long term.
The Supreme Court ushers in a new direction by issuing landmark decisions (Pacelle
1 99 1). In this way, it is not necessarily the percentage or direction of the cases but, instead,

Which particular cases are so salient that they advise others of the Court’s new agenda

\

4Interestingly, while appellants may pursue more particular types of cases based on the
trel'lds in the Supreme Court, they may be less discriminating about bringing cases in line
with the trend while being a good deal more discerning concerning cases which do not
Conform to the trend. The result is that litigants on the endorsed Side of the trend may appeal
more "bad" cases, while the stronger cases will stem from the disfavored pool, since parties
there must be much more strategic about which cases to appeal (Moe 1985). The result is
a decision-making trend in the Courts of Appeals which is opposite of that from the Supreme
Court (see Sheehan, Hurwitz, and Reddick 1996).

48

mrne.l
p.115 llllC'
Court in Ct
ieizg 3?, ‘3

Pers
jHMit
fihwfi
Wynn“
fireman oi
uxmmun
buuofA]
ion a liter
Whisker
which W3

$5
warm
WMMy
bUBOfA
"AIMS
Nu

Wumm

.3

\

priorities. Thus, issuance of landmark decisions by the Court iS a crystal indication to
parties, interest groups, and lawyers that particular types of cases will be favored by the
Court, the consequence of which is that the percentage of cases addressing that type of issue
being appealed to the Courts of Appeals will increase.

Personnel changes on the Supreme Court also can have an effect beyond a changing

judicial doctrine. That is, parties realize that particular appointees to the Court can have a

distinct effect on the agenda, notably where the Court turns sharply more liberal or

conservative, and thus they commence appealing cases in the lower courts in the perceived
direction Of ideological changes taking place in the Supreme Court. Similarly, the
appointment of a new Chief Justice can have a strong effect on litigious behavior in the
Courts of Appeals, particularly when the change is from a conservative to a liberal Chief, or
fi‘om a liberal to a conservative. In these circumstances, parties again will recognize a new
emphasis on specific issues in the Supreme Court, manifested by more of the types of cases
which would harmonize with the Chief being appealed.

AS stated, these are the theoretical rationales for the increasing or decreasing of
certain issues in the appeals courts as determined by changes in the Supreme Court. The
overriding theme about each of these particular arguments is that, Since the agenda of the
Co‘-ll'ts of Appeals is determined by litigious behavior, over time the agenda of the Courts
of Appeals will fluctuate in response to various changes within the Supreme Court.

Nevertheless, the direction of causality Should not exist solely from the Supreme

Court to the Courts of Appeals. Since the Supreme Court’s agenda is a fimction of both
internal and external rules and structures, I maintain that one of the external influences on

49

 

the Court‘s 3:"
agenda is hot
endogenous V

Aceor
in'lie Courts <
oiissue may !
recognize. A:
on s“ trype 0:
of issues. but
conflict amen
SHll‘ime Cou
Mitt in M)
Eels ofraIion
for 50m justi
lUSllCCS 500“ .
%WMm
Hitasure Of CE
Warm the

Priorities of ll.

“5 fin
lift

the Court’s agenda is the agenda of the Courts of Appeals. That is, the Supreme Court’s
agenda is both an exogenous influence on appeals in the lower courts as well as an
endogenous variable which responds to litigious behavior in the courts below.
Accordingly, my next argument is that the Supreme Court reacts to litigious behavior
in the Courts of Appeals. This can occur in one of two general ways. First, a particular type
of issue may gain favor and momentum in the lower courts, which the Supreme Court will
recognize. As a consequence, the Supreme Court eventually will begin to accept more cases
on that type Of issue. On the other hand, the Supreme Court may hope to avoid certain types
Of issues, but due to "percolation" in the lower courts (Perry 1991a), usually resulting in
Conflict among the circuits or negative publicity in the legal or political communities, the
supreme Court eventually is constrained to accept some cases it would rather not. The
manner in which the Court is impelled to accept cases in this regard has to do both with the
t001$ of rational choice and the procedure by which the Court grants cert. Thus, as the utility
for Some justices decreases when potential conflicts or negative publicity increases, those
j us‘iices soon will vote to hear particular cases they might otherwise prefer to avoid. When
this Occurs over time for more than a few justices such that cert. is granted in an expanding
measure of cases pursuant to the rule of four, then the Supreme Court will be reacting to
litigious behavior in the lower federal courts. The point is, under either of these stated
seeIlalrios, the agenda of the Supreme Court will be a reflection, in part, of the agenda
priol‘ities of the appellants in the Courts of Appeals.
My final argument is that, beyond their effects on each other, external environmental
in11‘1ences additionally should have a bearing on the political actors in both courts. In

50

lit :31? 353 5
. f...
hits. Thai ‘5

353‘s Shtl‘disl i

set and relate:
whituzion
immoral t
the: court it?
:erii form
critically te

lithe followi

ignite mil:

1,, ‘ .

“lintSlZB g
In lht Agenda
h-

we loser e.

Mirror-3L m

particular, salient political events (such as the civil rights movement), public opinion, and
the state and structure of the other federal governmental institutions all Should effect the
manner in which the Supreme Court accepts cases and cases are appealed to the circuit
courts. That is, as these external forces lean in a liberal direction over time, then more liberal
cases should be heard in both of these federal courts, with the converse transpiring when
these non-judicial factors are conservative in nature.
This is the theoretical foundation for the manner in which I maintain that agendas are
set and related to each other in the Supreme Court and Courts of Appeals. It incorporates a
neoinstitutional perspective by utilizing the concepts of rational choice while embodying
institutional effects, both from within each court (that is, endogenous effects) and on the
other court (that is, exogenous effects), along with integrating other external influences on
agenda formation. By so doing, I can assert specific hypotheses, which will be stated and

empirically tested in the ensuing chapters. The manner in which I will test them is explained

in the following section.

RESEARCH DESIGN AND METHODOLOGY
The fundamental purpose of this study is to determine the nature and relationship of
agenda modifications within the federal courts. As provided in the theoretical section, I
hypothesize generally that certain changes in the Supreme Court eventually will be grounded
in file agenda of the Courts of Appeals, but that there are endogenous effects emanating from
theSe lower courts which cause certain agenda transformations in the high Court as well.
IVlol‘eover, external influences additionally are of consequence to the agendas of these courts

51

as well. In order to ascertain which of these postulated effects are empirically verifiable, I
have fashioned a research design which incorporates these theoretical constructs within the
neoinstitutional perspective. Moreover, I draw upon particular statistical methodologies
which I believe are well-suited to addressing the issues in question in this research. The

specifics of this design and the various methodologies I have assimilated into this design are
detailed in the following sections.

Data

Time-series analysis, a type of methodology which has not been utilized in pullic law

Studies all that Often, is vital and necessary to explore the long-term, systematic variations
in agendas of the Supreme Court and Courts of Appeals. Two databases were utilized for
this purpose: the United States Supreme Court Judicial Database (hereafter, the " Supreme
Court Database"); and, the United States Courts of Appeals Data Base (hereafter, the "Courts
of Appeals Database").5 Specifically, I generated series of data from each of these databases
encompassing the years in which the data overlap, from 1946 through 1988. Since the
Supreme Court Database provides data for a ntunber of years beyond the Courts of Appeals
Database, while data for the appeals courts are available for years prior to that available for
the SUpreme Court, I included appeals court data commencing in 1940 and Supreme Court

data terminating in 1995 for comparative, but not statistical, purposes.

\

94 SI‘Iarold J. Spaeth is the principal investigator for the Supreme Court Database (ICPSR
(S 22); Donald R. Songer is the principal investigator for the Courts of Appeals Database
ES ~ 8912678).

52

F0.
for 111 oft
scoring
coin; is.
only mil
(out wl
\loreox er
as ism;
éeEineano
“Elms (
mo'fiom

C

um i

For each series, I determined the percentage of cases on each court’s docket per year
for all of the issue areas to be studied. Each of these issue areas was designated strictly in
accordance with the coding scheme reported in the Supreme Court Database. Identical
coding decisions were made for Courts of Appeals cases.6 At the Supreme Court, I included
only orally argued decisions which were formally decided, even if by an equally-divided
Court, while I excluded memorandum decisions, per curiam decisions, and decrees.7
Moreover, for the Supreme Court time series, the unit of analysis in this study is the case,
as demarcated either by its citation or docket number.8 In the Courts of Appeals, no such
delineations are necessary. Nevertheless, while the Supreme Court Database includes the

universe of Supreme Court cases, the Courts of Appeals Database does not encompass an
a112:‘l10gous survey of all cases in the Courts of Appeals. In particular, the Courts of Appeals
Database includes a random sample of a certain number of appeals court cases per circuit per
Y ear from 1925-1988. In order to ensure validity in the inferences made regarding the Courts
of Appeals data, therefore, weighting techniques necessarily were employed so as to
Standardize the data among the circuits. The dependent variable for all of the analyses in this
reSearch is the agenda of either the Supreme Court or the Courts of Appeals. See Appendix

A fer details on decision rules and all issues of coding for this study.

\

63cc Appendix A for coding decisions with respect to issue areas differences between the
Dreme Court and Courts of Appeals.

Su
7Accordingly, I included DEC_TYPES 1, 5, 6, and 7 in the Supreme Court Database.
8This includes ANALU = ‘ ’ or ‘1’.

53

Methods

As with any time-series analysis, all data must be stationary in order to make valid
inferences; thus, unit root tests, including augmented Dickey-Fuller (1979) and Box-Jenkins
( l 976) tests, were used to identify potential problems. Appropriate transformations, whether
differencing or log conversions of the data, were effected to ensure stationarity. Once time
dependencies were removed fi'om the data, numerous statistical procedures could take place

K without producing spurious results.9
In chapter 3, I will determine whether causal relationships between the Supreme
Court and Courts of Appeals exist with respect to the emphasis by each court on the types
0f issues they hear. I hypothesize generally that changes in agenda at the Supreme Court
transform the agenda of the Courts of Appeals, but that there are endogenous features to
th'EBSe effects which depend in part upon the issue area. By looking at the percentage of cases
deGided over time in each court and whether such changes affect the percentage of cases
heard in the other court, I will be able to determine whether the Supreme Court motions
ratiOnal litigants simply by deciding more or less of particular types of cases over time.

MOreover, I will discern whether parties, interest groups, and their representatives in turn

\

. 9As discussed in greater detail in chapter 3 (see, e. g., p. 81, footnote 10), while
Cufferencing data is necessary to avoid spurious results in time-series analysis, this

allSformation comes at the cost of losing long-run relationships in the data. One manner
of avoiding such loss of information concerns the time-series technique called cointegration
(Engle and Granger 1987). Cointegration is a statistical method for determining whether
:onspurious, long-run relationships in equilibrium exist between variables while providing
eOr _ an error-correction mechanism to ascertain deviations from that relationship in
vq‘fllibrium (Banerjee, Dolado, Galbraith, and Hendry 1993). Nevertheless, due to the
oilfiety of other statistical methods to be employed herein, cointegration is beyond the scope

this study and will not be utilized.

54

 

rear: to th
Court just
10'»er cou

F0
“filth ml
or he do;
1969; Re

1931}. Fr:

analyllnrv

react to these signals by litigating suitably in the courts below, and whether the Supreme
Court justices themselves reciprocate by addressing any changing emphases found in the
lower courts.

For purposes of chapter 3, there are a variety of time-series methodologies available
which will uncover the causal tendencies with respect to transforming percentages of issues
on the dockets of each of these courts. First, I will employ Granger causality tests (Granger
1969; Freeman 1983); then, vector autoregression (hereafter, "VAR") will be used (Sims
1980; Freeman, Williams, and Lin 1989). Both of these techniques apply very different
means to estimate essentially the same phenomenon, to wit: to examine whether one series

0f data causes effects in another series.

Specifically, Granger causality is based on "the idea that a variable)! ‘causes’ another

Variable Y, if by incorporating the past history of X one can improve a prediction of Y over
a Prediction based solely on the history of Y alone" (Freeman 1983, 327-328). That is,
Granger causality is a simple regression of Y on itself, including its lags, and on X and its
lagS- This method employs standard f-tests, and if X provides a better explanation of Y than
Y explains on its own, then X is said to "Granger cause" Y. Consequently, since it addresses
p air~Wise exogeneity, the Granger causality method lends itself rather well to the
CleteI‘rnination whether one of these federal courts influences the agenda of the other over
time.
Nevertheless, Granger causality tests are insufficient on their own, essentially for two
reasOns. First, Granger tests address causality only in one direction, so there is no mode of
analyzing feedback in a system. Also, there is a debate as to whether Granger causality

55

«ruin
a. 81 “1.)

science.

messy}
Dreamy}
01‘ Squatic
Pit“ lags (

on the di

Marni.

"‘31 Gr

lSSUQ a

“a

l tt‘
Gem

“01131:

famed

ascertains "causation" in the statistical sense we generally are accustomed to in political
science. Alternatively, then, in this chapter I will employ VAR, which is useful in this
research in that this approach allows for feedback, since it is a reduced form (Sims 1980).
VAR does not treat any variables in the system as exogenous. Each variable is regressed on
past lags of all the variables, where fewer restrictions are imposed on the data (Freeman,
Williams, and Lin 1989). In other words, VAR relaxes the causal assumptions which are
necessary in the process of identification in structural equations modeling, such as is
necessary in two-stage least squares. VAR is able to test causal relationships within a system
of equations by imposing few restrictions on the variables, as it regresses each variable on
past lags of itself and all other variables. Accordingly, VAR is well-equipped to concentrate
on the direction of causal inferences over time, which makes this approach especially
pertinent for analyzing the dynamic, causal relationships between changes in the agendas of
the Supreme Court and Courts of Appeals.lo For these reasons, in this study I will resort to
both Granger causality tests and VAR procedures on all the series of data for each of the
issue areas in order to determine whether appropriate causal inferences can be made.
In the succeeding chapter, I seek to determine in greater detail the systematic
relationships between the several sub-issues which comprise the aggregate civil liberties
cateEOTY- In chapter 4, then, I segregate the broad civil liberties classification into its

Separate issues of civil rights, first amendment, due process, and privacy cases. Then, in the

\

I
de °Freeman, Williams, and Lin (1989) adeptly address the various criticisms ofVAR while
a monStl'ating the appropriateness and many benefits of utilizing this methodological
ppl'Oach. See also Granato and Krause (1998); Sims (1980).

 

56

same manner in which I proceeded in chapter 3 , I employ further Granger causality and VAR
tests on these sub-issues to determine whether these federal courts exist within a system for
civil liberties cases, and to ascertain whether the disaggregated civil liberties agendas behave
in the same manner the full category.

Chapters 3 and 4 incorporate the theoretical arguments that justices and parties react
to the agendas of each court along with decisional trends, but they do so by employing
bivariate statistical analyses. On the other hand, chapter 5 carries the theoretical concepts
of this research a step further by utilizing multivariate analysis in order to determine the
factors which influence agenda changed in the federal courts. Specifically, two-stage least
squares time-series regression analysis is employed, due to the endogenous nature of j udicial
agendas, as the civil liberties, economic, and criminal procedure agendas of both courts are
regressed on a variety of potential influences. These independent variables are both judicial
and non-j udicial in nature, including the past agendas and dynamic liberalism of each court,
the relative liberalism of the political branches of the federal government, public opinion,
landmark Supreme Court decisions, and the influence of the tenures of different Chief
Justices of the Supreme Court. In this manner, the time-series analysis is able to examine
Whether factors other than the judicial agenda of the other court influence changes in the

agendas of the federal courts.
Validity

By setting up my research design in this manner, I believe that threats to internal and
eXtemal Validity should be minimized. As to internal validity, since this research does not
represent an experimental design, problems such as history, maturation, testing, and

57

 

 

lli "dill:
the p055
not be a
mixers;
circuit ‘
(Cook a
Suprem
aspect

Kr his.

(.2

.13 .

instrumentation should not be of concern (Campbell and Stanley 1963). Nevertheless, while
the possibility of selection bias does exist as a potential threat to internal validity, this should
not be an issue due to the nature of the data. First, the Supreme Court data represent the
universe of cases. Second, while the Courts of Appeals data include a random sample by
circuit, the weighting index used should minimize problems with respect to selection bias
(Cook and Campbell 1979). Lastly, since there are 43 annual data points over which the
Supreme Court and Courts of Appeals data coincide, no small N problem is apparent with
respect to selection. And, I have not chosen the data on the dependent variable (King,
Keohane, and Verba 1994).

The ability to generalize obviously is an important feature of any research design,
which necessitates the minimization of threats to external validity. Again, since I am not
employing an experimental design, problems associated with reactive or interactive effect
of testing, reactive effects of experimental arrangements, or multiple-treatment inference do
not pose a problem (Campbell and Stanley 1963). Moreover, there should be sufficient
Variance of all the dependent and independent variables, especially when the different issues
are considered over the time span incorporated in this research. And, the weighting
techniques utilized should temper any potential dilemma with respect to the Courts of
Appeals data.

The most extreme threat to external validity with respect to this research, however,
Concerns the limitation of data to the Supreme Court and Courts of Appeals, an approach

Which ignores the influence of agenda effects from the numerous and assorted State Supreme

58

Connie:
meluiifis
germ"
alie‘x i333
general.
finmhm
finsaud

the em

to anal};
Dannie
30 cues
Studiei

difiif Kill)
paniculi
hem Chfi

1997. 30;

(0‘8

08
men

“Duel
"icon;
3 limes

lfo

Courts.“ To be thoroughly valid, similar analyses should be made with respect to, and
including, the state courts. Due to lack of data, however, particularly from a longitudinal
perspective, such a task is unmanageable at this time and, unfortunately, will not be
alleviated any time soon.12 Accordingly, while lack of state court data limits the
generalizability of this research to a certain extent, at this stage of data development there
is nothing that can be done. Irrespective of this potential difficulty, however, the results from
this study should be sufficiently robust to ascertain agenda effects in the federal courts, if not
the entire judicial system in the United States.
I need to point out a caveat at this stage of my research. While I would have hoped
to analyze circuit differences in this research, due to the nature of the Courts of Appeals
Database I will not be able to do so. Very simply, because this Database is limited to at most
30 cases per circuit per year, if each circuit is broken down filrther into the various issues
studied, a small N problem inevitably results, precluding valid inference. While this
difficulty would apply regardless of the type of methodology employed, this dilemma is
particularly troublesome with respect to analyses utilizing VAR, which can produce what has
been characterized as a "vanishing degrees of freedom problem" (Johnston and DiNardo

1997, 305). Accordingly, no analysis of agenda at the circuit level will be accomplished in

 

 

, "Cases from State Courts make up about 30 per cent of the Supreme Court’s docket
(O Brien 1991).

(f ”Due to the strategy behind the Hall and Brace State Supreme Courts Judicial Database
Or ’ he Oming), it appears that this problem of a lack of longitudinal data will not be remedied
any time soon, even with their substantial undertaking.

S9

 

de. Nevertheless, this should not detract from my research herein, as there is an

ince of inquiries to be realized otherwise.

60

 

AV E.)
AG

CHAPTER 3

AN EXPLORATION INTO THE CAUSAL ASSOCIATIONS BETWEEN THE
AGENDAS OF THE SUPREME COURT AND COURTS OF APPEALS

The Supreme Court sits atop the judicial hierarchy in the United States. As a result,
scholars in both the legal and political science communities insist that lower courts, whether
by norm or structure, are bound by the precedents established by the Supreme Court. While
adherence (or lack thereof) to stare decisis forms the bulk of much of the literature in

political science on the Courts of Appeals, it is not at all clear whether the lower courts heed
the Supreme Court with respect to agenda formation. First of all, the dockets of the Courts
of Appeals are not determined by the judges, so judicial norms and structures are irrelevant
in this regard. Nevertheless, the Supreme Court occupies a powerful position from its
Judicial perch. Accordingly, it seems reasonable to suspect that parties will sustain any
trends in agenda instituted by the Supreme Court, especially when it is assumed that litigants
and their representatives are rational actors. Indeed, one of the arguments I make in this
study iS that, once the Supreme Court establishes its agenda priorities, the effects reverberate
thro‘-1817l0ut the judiciary; and, one of the political institutions inspired by the Court’s agenda
preferences is the Courts of Appeals. As a consequence, the agenda of the federal circuit

Courts Should fluctuate in line with the agenda trends of the Supreme Court.

61

 

The Supreme Court theoretically is isolated from political and social pressures;
indubitably, this professional seclusion, combined with lifetime appointments, in large part
constitute the rationale that justices’ decisions are linked so closely with their own ideology
(Rohde and Spaeth 1976). Still, the Court does not exist in complete quarantine or
detachment. To the contrary, the fact that the decisions of the Court generally are not
countermajoritarian (Dahl 1956), and that some judgments potentially are influenced by
public opinion, either directly or indirectly (Mishler and Sheehan 1993), demonstrate that
the Supreme Court necessarily is not as disengaged as some might believe, hope, or fear.
From a neoinstitutional perspective, institutional structures should influence individual
political behavior, and for the Supreme Court such influences might stem both from within
and without the Court. With respect to agenda formation, the internal rule of four both limits

(Epstein and Segal 1997) and provides for enhanced prospects (Shubert 1959; 1962) for the
Court’s agenda formation. From the outside, judicial or political institutions other than the
Supreme Court might provide for constraints upon the justices’ ability to establish the
Court’ 8 agenda in line with their true preferences. Thus, the Supreme Court’s agenda is not
merely the aggregation of the justices’ proclivities over cert. in individual cases but instead
is a "function of a complex interaction of individual preferences and institutional structures
and rules" (Pacelle 1991, 13-14). To that end, I maintain expressly that one of the external
influences on the Court’s docket is the agenda trends established in the Courts of Appeals.
That is, the Supreme Court’s agenda is both an exogenous influence on lower court appeals

as Well as an endogenous variable which responds to litigation in the courts below.

62

lll
agenda of
on agend:

baedexe

males Cl‘
preferenc

the Com

Preferenc
minim
mciudim

be?“ cc

OIAPPfia

In this chapter I consider whether the agenda of each of these courts influences the
agenda of the other over time. As such, I do not take account of external or other influences
on agenda formation, but instead I inquire into the nature of agenda setting in these courts

based exclusively on the agendas established and the trends grounded in any changes therein.

AGENDA INFLUENCES IN THE FEDERAL COURTS IN GENERAL
One of the arguments posited in this study is that the choices the Supreme Court
makes concerning its agenda ripple throughout the judiciary. In particular, agenda
preferences established by the Supreme Court over time soon will be found in the agenda of
the Courts of Appeals.

As stated, I view the Supreme Court’s agenda as the accretion of individual
preferences, subject to institutional structures and rules. I assume further that, as the
institution sitting atop the judicial hierarchy, the Supreme Court affects courts below,
including the Courts of Appeals (Baum 1994; Pacelle and Baum 1992). Similarly, litigation
begets comparable litigation (Provine 1980). As a result, I posit the following hypothesis:

Hypothesis # I. The agenda of the Courts of Appeals should
respond directly to changes in the agenda of the Supreme
Court in either a positive or negative direction.

The Supreme Court has discretionary jurisdiction, while the jurisdiction of the Courts
ofAPPeals is mandatory. Thus, the Court controls its agenda, subject to institutional forces,
but the appeals courts lack analogous command over their agenda. Accordingly, their agenda
is detemfined by litigious behavior which is shaped, to a certain extent, by the Supreme
Court’s agenda priorities, as over time the Court "condition[s] the behavior of litigants"

63

 

ll’iee‘lle
litiglits
accords
increasi:
Cm t
57:07. to

l: respo

(Pacelle 1991, 80). Thus, as the Court modifies its agenda preferences, I assert that rational
litigants, interest groups, and lawyers will recognize these changes and react accordingly in
accordance with their own goals. For instance, as the Supreme Court’s agenda became
increasingly filled with civil liberties cases during the 19408 through 19603, I expect that the
Courts of Appeals also would hear and decide a greater number of civil liberties cases at a
short to moderate lag, as litigants began to increase their appellate filings in this issue area
in response to changes instituted by the Court.

Nevertheless, the Supreme Court’s place at the top of the judicial hierarchy does not
ensure total authority over the Courts of Appeals (Baum 1994). Indeed, even assuming
strong hierarchical dominance, at best it is questionable whether such authority extends
beyond the appeals court judges to litigants. Moreover, due to the disparity in the number
of cases the Supreme Court and Courts of Appeals hear each year, the circuit courts are
effectively courts of last resort; thus, the Court’s oversight capabilities with respect to the
courts below are extremely limited (Howard 1981; Songer 1991).'

Further, the extent to which rational choice theory is relevant to nonelite political
behavior is somewhat dubious, at least when compared to its pertinence to elites engaging

in high-stakes political behavior (Aldrich 1994). Thus, while some parties and lawyers

dI’iVing federal appeals might be likened to political elites (such as those involved in national

\

1Over 50,000 cases now are commenced in the Courts of Appeals each year, while
the Supreme Court, which generally decides about 150 cases per term, has issued less than
100 Written opinions in the each of the past few terms (Carp and Stidham 1996).

64

interest groups or certain professional bars), a better analogy might be one of parties to the
masses, where rational behavior is not necessarily germane.2

Under these circumstances, the agenda of the appellate courts may not be directly
responsive to the dynamics of the Court’s docket. For these reasons, I offer the following

alternative hypothesis:

Alternative Hypothesis # 1. The agenda of the Courts of
Appeals should not respond to changes in the agenda of the

Supreme Court.

Quite simply, due to the vast number of docketed appeals, limited oversight by the Court,
and the ambivalent behavior of litigants, the agenda of the Courts of Appeals may not be
responsive to shifts in the Court’s agenda.

While I argue generally that the Court’s agenda influences appellants in the Courts
of Appeals, I speculate as well that the Court as an institution is motivated by a variety of
rules and forces. Consequently, I contend that the Court’s agenda is influenced by these
various external pressures. As opposed to propounding that the Court’s agenda merely is
affected by outside influences, without specifying the nature of any of these effects which
Operate to control agenda change, I take Pacelle’s (1991) argument a step further by
maintaining that one of these external influences on the Court’s agenda is the changing
dockets of the Courts of Appeals. That is, the Court’s agenda is both an exogenous influence

on the litigious appeals (see Hypothesis # 1) as well as an endogenous variable which

reSponds to litigation in the courts below.

\

2Even if it is befitting, certainly the applicability of rational choice theory to mass
behavior is much more complex than for elite behavior.

65

I thus present my next hypothesis:

Hypothesis # 2. The agenda of the Supreme Court should

respond directly to changes in the agenda of the Courts of

Appeals in either a positive or negative direction.
Concededly, causation in this hypothesized direction should not be as strong as that proposed
in Hypothesis # 1. Moreover, there should be a greater lag in the effect by the appeals courts
on the Court’s agenda than vice versa. Nevertheless, this does not diminish the endogenous
nature of agenda change, and for this reason I offer my Hypothesis # 2.

Of course, as attitudinalists would attest, the justices are well-insulated from most
tangible constraints on their behavior, at least when it comes to their roles as decision
makers. There is reason to believe that this judicial resilience also might extend to the stage
of the agenda formation. That is, the justices may be no more influenced or constrained by
rates of appeals in the lower courts as they are by congressional threats of reversing
particular decisions (Segal 1997). For this reason, I offer the following alternative
hypothesis:

Alternative Hypothesis # 2. The agenda of the Supreme Court
should not respond to changes in the agenda of the Courts of

Appeals.

So stated, these hypotheses embody my general argument that the Supreme Court’s
agenda explains part of the agenda of the Courts of Appeals, but that an endogenous effect
additionally should be evidenced as well which stems from the preferences of parties’

appeals. In the next section, I describe some of the hypothesized relationships between the

agendas of these courts for specific issues.

66

SPECIFIC ISSUES, AGENDAS, AND THE FEDERAL COURTS

Prior to the Switch in 1937, the Supreme Court predominantly heard and decided
cases which concerned economic issues, while very few civil liberties cases reached the
Court’s docket; in part, this emphasis on economic issues was a consequence of the Court’s
mandatory jurisdiction of the day, but it also revealed the agenda preferences of the justices
once the Judges Bill of 1925 gave them more discretion to construct the Court’s agenda.
That is, the conservative justices in the majority during the late-19205 through mid-19308
were interested in deciding economic cases, particularly when the issues at bar concerned the
commerce or substantive due process clauses and the justices were able to base their
decisions on preferred Iaissez-faire doctrine (Segal and Spaeth 1993). Due to both politics
and evolving judicial preferences, this state of affairs began to vary after the Switch,
particularly once President Roosevelt’s appointees assumed the Bench. As a consequence,
over the ensuing half century civil liberties came to replace economic cases on the Court’s
agenda.

With respect to civil liberties cases, there is a sense today among many in both the
Scholarly and public communities that the Supreme Court should be the ultimate forum for
the preservation of individual rights (Adamany 1991). Indeed, Justice Stone, in his famous
footnote 4 in United States v. Carolene Products, 304 US. 144, 153, fn.4 (193 8), positively

asserted that custodian of "discrete and insular minorities" would symbolize the Court’s
appropriate judicial role in the future. Justice Stone was somewhat prescient, inasmuch as

this landmark decision soon led to an increase in the number of civil liberties cases heard by

67

 

 

the Court.3 In particular, Pacelle (1991) argued that the Supreme Court was interested in
purging its docket of economic cases to the extent possible, and the Court was able to do so
in part because a large pool of civil liberties cases was available to replace the economic
cases.4

This strategy of exchanging economic for civil liberties cases comported with the
policy goals of the justices after the Switch and enabled them to avoid the controversial
economic cases which in large part were responsible for the Court’s self-inflicted wounds
in the 19303. Hypothesis # 1 contends that, in general, a similar effect will be found in the
agenda of the Courts of Appeals as parties follow the lead of the Supreme Court.
Nevertheless, there is reason to believe that the context of agenda change in either court may
not be uniform across the panorama of issue areas. Thus, as the Supreme Court decided
greater numbers of civil liberties cases over the years, it seems intuitive that rational parties
would heed this call by the Supreme Court. But what about other issues, such as economic
cases? In other words, when the Court changed its agenda preferences for issues other than

civil liberties, how should parties in the lower courts have reacted to such changes?

‘

’Nonetheless, some empirical evidence demonstrates that the Supreme Court may not
be the great defender of civil liberties and individual rights it frequently is proclaimed to be
(See, e. g., Dahl 1956; Caldeira and Gibson 1992; Mishler and Sheehan 1993; 1996).

4In Pacelle’s (1991) terms, the bifurcated nature of the Supreme Court’s agenda
Contained an exigent agenda, which included cases which the justices needed to attend to at
Some point and usually were decided on the basis of judicial roles; and, there was the
VOIitional agenda, which contained cases which would "firlfill the policy designs or goals"
of the justices (2 8). This dual character of the Court’s agenda enabled the justices to transfer
eoonomic cases fi'om the volitional to the exigent agenda and fill the void on the volitional
agenda with civil liberties cases, while at the same time the justices were shrinking the

eXigent and augmenting the volitional agenda.

68

 

 

Essentially, the answer relates to a combination of conditions, including the saliency
of the issue, whether the Supreme Court is interested in increasing or decreasing a type of
issue on its docket, whether the Courts of Appeals truly represent the court of last resort for
particular categories of cases, and spillover effects. That is, the more salient an issue, the
greater the likelihood that parties would observe the agenda trends of the Supreme Court for
that issue. Similarly, if the Court elevated its emphasis on a particular issue, the more likely
parties would boost their appeals in that area as well. However, if the Court decreased its
emphasis on a specific issue, especially if that issue is one in which litigants need to pursue
their rights via the federal courts, then the appeals courts unquestionably would be the courts
of last resort; under these circumstances, parties would be less likely to follow the trends
established by the Supreme Court for those issues. In fact, the converse may be true, in that
this represents situations in which the Supreme Court may be compelled to settle disputes
on issues it would rather avoid. That is, it is when the Supreme Court attempts to elude
certain issues that the endogenous effect from the courts below should be most apparent.
Finally, spillover effects (Kingdon 1984) would occur when trends in one issue beget similar
inclinations in another, such that parties would be more likely to regard the Court in one
issue based upon the Court’s actions in a comparable issue.

These expectations for specific issues further enhance the general theoretical
Conditions I posited in chapter 2. The implication to be realized from the motivations listed
above and in chapter 2 is that the Supreme Court should have been persuasive to parties’
1 itigious decisions in the courts below for all issue areas. The only substantive question these

eMpectations leave open, then, is: 1) the extent to which the Supreme Court’s agenda

69

 

influenced the dockets in the Courts of Appeals; and 2) whether the Supreme Court in turn
was influenced by trends being grounded in the lower courts. Based on these theoretical
expectations, I delineate the following hypotheses for each of the following issues.

The first issue is civil liberties, in which there should be strong effects from the
Supreme Court to the Courts of Appeals but not necessarily in the opposite direction? First,
there is no questioning the saliency of, and controversy over, civil liberties cases, which
generate potentially intense beliefs among judges and others (Segal and Spaeth 1993). Thus,
as the Court heard greater numbers of civil liberties cases after the Switch in 1937, a
comparable effect seemed unavoidable among interest groups and others which perceived
an opportunity to foster their causes by bringing similar cases to bar in the lower courts.
Even when the Court began to hear fewer civil liberties cases over time, litigants would sense
that the Court might fail to perform its perceived role of guardian of individual rights,
leading to a decrease in appeals in civil liberties cases. Finally, spillover effects from similar
issue areas, such as criminal procedure or the prevailing, external political environment,
should have more of an effect on parties’ adherence to the Court’ s agenda priorities than vice
versa. These expectations form the essence of the following hypotheses:

Hypothesis # 3a. The agenda of the Courts of Appeals in civil
liberties cases should respond directly to changes in the

agenda of the Supreme Court in either a positive or negative
direction.

5For purposes of this chapter 3, the classification of civil liberties cases includes civil
I'ights, first amendment, due process, and privacy issues (see Stimson, MacKuen, and
Erikson 1995; Hurwitz and Reddick 1996). In chapter 4, I will break down this omnibus
Civil liberties issue into its corresponding branches in order to determine whether any
Systematic influences flow from the separate parts or the whole.

70

Hypothesis # 3b. The agenda of the Supreme Court in civil

liberties cases should not respond to changes in the agenda of

the Courts of Appeals.
Accordingly, parties should be influenced by agenda priorities in the Supreme Court, but the
justices unlikely will be similarly influenced by appeals rates in the courts below.

Economic cases, however, posed a somewhat different situation from civil liberties.

In fact, economic issues constituted the bulk of the Supreme Court's agenda before the
Switch. Nevertheless, once the justices were provided with discretion over the Court’s
docket and they became concerned about circumventing the controversies of the mid-19305,
the justices used this newly-found discretionary jurisdiction to resist economic cases.
Irrespective of the Court’s inclinations, however, economic issues formed, and continued to
shape, the basis for much of the case load of the Courts of Appeals. Indeed, Howard (1981)
advocated that a prime responsibility of the appeals courts was to adjudicate civil litigation,
much of which encompassed economic issues. Thus, while certain corporate parties needed
to continue pursuing their economic interests via litigation in the federal courts, the Supreme
Court clearly shied away from this area, the result being that the Courts of Appeals in fact
became the effective court of last resort for this issue. Lastly, economic cases generally are
less controversial than civil liberties. While this may signify that lower court judges will be
more likely to follow Supreme Court precedent in economic rather than in civil liberties

cases (Songer 1987; Baum 1978; Johnson and Canon 1984), such lack of salience should

have the opposite effect on parties in the lower courts, who likely would appeal less

71

controversial economic cases without nearly as much concern for the Court's priorities.‘5
Under these circumstances, while parties with economic interests continued to follow the
Supreme Court, such adherence should not be nearly as strong, either in terms of time or
effect, as for civil liberties cases. Furthermore, these are the conditions under which the
Supreme Court likely will be obliged over time to accept such cases it would rather not. For
these reasons, I assert the following hypotheses:

Hypothesis # 4a. The agenda of the Courts of Appeals in

economic cases should respond directly to changes in the

agenda of the Supreme Court in either a positive or negative

direction.

Hypothesis # 4b. The agenda of the Supreme Court in

economic cases should respond directly to changes in the

agenda of the Courts of Appeals in either a positive or

negative direction.
That is, if endogenous effects are to be observed stemming from the lower courts to the
Supreme Court, I anticipate that it should occur with cases involving economic issues.

Criminal procedure cases are more similar to civil liberties than to economic cases,

at least for any effects derived from the Supreme Court. That is, cases involving the Fourth
Amendment present salient and controversial issues, they have received enhanced emphasis
by the justices over time, and there should be spillover effects from civil liberties cases.

Nevertheless, once the criminal defense bar realized these avenues of appeal were available,

even if the Supreme Court decreased its accent on these issues, it is likely that similar

6Note that this appraisal logically follows from what I argued concerning civil
liberties cases, where litigants are hypothesized to be more likely to follow these agenda
trends of the Supreme Court, even though the Court's impact on decisional trends lower
Courts may be less in civil liberties cases (Songer 1987).

72

appeals would persevere in the lower courts. Accordingly, as demonstrated in the ensuing
hypotheses, the effect in criminal procedure cases should be observed in both directions, as
the Supreme Court likely persuaded parties to bring these appeals, while contemporaneously
litigious behavior ought to have influenced the Supreme Court’s docket as well:

Hypothesis # 5a. The agenda of the Courts of Appeals in

criminal procedure cases should respond directly to changes

in the agenda of the Supreme Court in either a positive or

negative direction.

Hypothesis # 5b. The agenda of the Supreme Court in

criminal procedure cases should respond directly to changes

in the agenda of the Courts of Appeals in either a positive or

negative direction.

The next issue under consideration concerns union and labor cases. I anticipate that
this issue will produce effects in both directions as well, since it has been somewhat
controversial in the past, and when the relevant interest groups recognized trends in the
Supreme Court in other issues, the spillover effect should have manifested itself in these
cases in the lower courts. Still, even when the Supreme Court exhibited diminished interest
in union issues, parties would continue to pursue their labor interests, thus inducing the
Supreme Court to accept some of these cases over time. The following hypotheses, then,
arise from these conditions:

Hypothesis # 6a. The agenda of the Courts of Appeals in
union and labor cases should respond directly to changes in
the agenda of the Supreme Court in either a positive or
negative direction.

Hypothesis # 6b. The agenda of the Supreme Court in union
and labor cases should respond directly to changes in the
agenda of the Courts of Appeals in either a positive or

negative direction.

73

For both the legal and political communities, attorneys cases are relatively
unimportant, as their only tangible pertinence is to the parties at bar. Thus, if there is any
causation at all between these courts for this issue, it should emanate only from the Supreme
Court. Accordingly, I posit the following hypotheses:

Hypothesis # 7a. The agenda of the Courts of Appeals in
attorneys cases should respond directly to changes in the
agenda of the Supreme Court in either a positive or negative
direction.

Hypothesis # 7b. The agenda of the Supreme Court in
attorneys cases should not respond directly to changes in the
agenda of the Courts of Appeals.

The same can be said for cases involving both federalism and interstate relations
issues, which comprise a small fraction of the dockets in the federal courts. In fact, such
issues may be more important to the justices themselves than to any of the myriad interests
which take advantage of the federal courts. Thus, any effects at all should be noticed solely
from the Supreme Court and not in the other direction:

Hypothesis # 8a. The agenda of the Courts of Appeals in
federalism cases should respond directly to changes in the
agenda of the Supreme Court in either a positive or negative
direction.

Hypothesis # 8b. The agenda of the Supreme Court in
federalism cases should not respond directly to changes in the
agenda of the Courts of Appeals.

Hypothesis # 9a. The agenda of the Courts of Appeals in
interstate relations cases should respond directly to changes

in the agenda of the Supreme Court in either a positive or
negative direction.

74

Hypothesis # 9b. The agenda of the Supreme Court in
interstate cases should not respond directly to changes in the
agenda of the Courts of Appeals.

On the other hand, federal taxation cases seem to be most similar to those involving
economic issues. Indeed, in the early part of the time period involved in this study, taxation
cases made up about one fifth of the dockets in the Courts of Appeals, while the Supreme
Court for the most part appeared to be relatively uninterested in these cases. Moreover,
taxation issues are not all that salient, in that their interpretive foundation generally is
statutory in nature. Under these conditions, and in a similar vein as with economic issues,
where spillover effects could be observed, it appears that influence from both courts to the
other should be expected, as depicted in the succeeding hypotheses:

Hypothesis # 10a. The agenda of the Courts of Appeals in
federal taxation cases should respond directly to changes in
the agenda of the Supreme Court in either a positive or
negative direction.

Hypothesis # 10b. The agenda of the Supreme Court in
federal taxation cases should respond directly to changes in
the agenda of the Courts of Appeals in either a positive or
negative direction.

Finally, there is no a priori expectation for miscellaneous cases, a category which
includes cases fitting none of the issue areas in this study. For this reason, no testable
hypotheses are posited for these unspecified issues, and results will be reported only for
demonstrative purposes. In the subsequent section, therefore, I will test the foregoing
hypotheses in order to determine whether adjusting emphases in the agenda priorities of
either the Supreme Court justices or the parties, interest groups, and their counsel influence

inevitable changes in the dockets in the other court.

75

DATA, METHODS, AND ANALYSIS

As I explained in some detail in chapter 2, time series will be utilized to determine
the dynamic agenda influences between the Supreme Court and Courts of Appeals. In
particular, the time-series techniques I will employ in this chapter include Granger causality
tests and vector autoregression (VAR). Both of these approaches to time series are fairly
well-suited to identifying causal associations over time, yet each applies a disparate
methodological strategy to explore whether one series of data influences another. I employ
both Granger causality and VAR, as opposed merely to one of them to the exclusion of the
other, because their distinct approaches in tandem should convey results which neither could
yield on its own.

Nevertheless, before applying these various approaches, I need to ascertain whether
the data are stationary. The data, as outlined in chapter 2, include various time series of the
percentage of cases per year of several issues before the Supreme Court and Courts of
Appeals. Since these are time series, the data must achieve stationarity before appropriate
modeling can take place. Stationarity simply denotes that time plays no role in the sample
moments of mean and variance of the data. That is, stationarity implies that both the mean
and variance are unconditional, finite, and not dependent on time; moreover, shocks to the

system will dampen over time, as the system is mean-reverting.7 To the contrary, if the data

 

7Technically, this is the definition of weak stationarity. On the other hand, strict
stationarity is characterized by the joint distribution which is the same throughout the series.
Nevertheless, weak and strict stationary are identical when a normal distribution is assumed,
since the first two moments of mean and variance are the same in the normal case.
Accordingly, weak stationarity is sufficient for most time-series applications.

76

are not stationary, then applicable transformations to stabilize the data must be accomplished
before statistical techniques are employed, Since nonstationary, or integrated, data feature
infinite memory due to their time-dependent means or variances. If appropriate
transformations are not made to integrated data, the results will be spurious, since Shocks to
the system will have a permanent effect. Generally speaking, taking a natural log of the data
usually calms nonstationary variances, while differencing takes care of inappropriate means
and trends in the data. If these transformations are to be employed contemporaneously,
natural log conversions must precede differencing, since the latter results in negative values
which cannot be logged.

A unit root is associated with nonstationary data; that is, the data are integrated when
a unit root is present. For instance, in the following AR(1) representation,

Yt = pYH + ut
a unit root is present when |p| = 1. In this case, the data are integrated of the first order,
where I = (1). The goal, then, is to ensure that the data are I = (0), such that |p| < 1, and there
are a variety of processes to detect, whether or not a unit root is evidenced.8

The first procedure to determine whether the various series in this study are stationary
or integrated, even before performing statistical operations, is to eyeball the data. "In
practice, it is essential at least to look at plots of the individual series against time to establish

what are the major features of the data that the model must capture . . . Some possible

 

8At this stage of analysis and discussion, implications of the near-integrated case,
where |p| approaches but does not equal 1, are not relevant. Notwithstanding this factor,
because near-integrated processes "mimic integrated series in finite samples" (DeBoef and
Granato 1997, 622) they Should be of concern at the appropriate phase of research.

77

 

dominant features are trends in mean or variance, strong seasonal components or the long
swings associated with unit root componen " (Granger 1990, 1 1). Accordingly, viewing the
graphs from chapter] ,9 it appears that time dependencies likely are present in nearly all the
series of data; possible exceptions based simply on viewing the data include the Supreme
Court’s criminal procedure cases (Figure 1.9), the Courts of Appeals’ federalism cases
(Figure 1 . 12), interstate relations in both courts (Figure 1.13), and miscellaneous cases in the
Supreme Court (Figure 1.15), as these series appear to display a stationary character.
Since it seems very likely that various transformations will be necessary to ensure
stationarity in most of the data, one method which diagnoses more definitively whether time
series data are characterized by time dependencies is the augmented Dickey-Fuller (AD-F)
tests, which are designed to detect unit roots in time series. The AD-F test applies the
following restrictions to three separate regressions utilizing the data:
a) no intercept, no trend:
AY, = (I)Y,,l + EpAYH + 8,, where:
Ho: (I) = 0; unit root (random walk)
HA: (I) at 0; no unit root (stationarity);
b) intercept, no trend:
AYt = a + (I)Y,,, + EpAYH + 31. where:
Ho: (I) = 0; tmit root (random walk with drift)
HA: (I) 1. 0; no unit root (stationarity);
c) intercept and trend:
AYt = a + (I)YH + ZIpAYH + 6*t + e, where:

Ho: (I) = 0; unit root (random walk with drift around a deterministic trend)
HA: (I) at 0; no unit root (stationarity);

 

°See Figures 1.1 (civil liberties), 1.8 (economic), 1.9 (criminal procedure), 1.10
(union and labor), 1.1 1 (attorneys), 1.12 (federalism), 1.13 (interstate relations), 1.14 (federal
taxation), and 1.15 (miscellaneous).

78

 

where the Y,,, coefficient is observed, the null hypothesis being that a unit root, and thus
nonstationarity, is present in the data. Values for rejection of the null are not based on the
t—distribution, but instead they are predicated on the nonstandard Dickey-Fuller distribution.
Consequently, the alternative MacKinnon (1990) critical values are used for rejection of the
null, since they have been found to be more accurate than the original critical values
established by Dickey-Fuller for their distribution.

As a caveat, the tests I have delineated here are the augmented Dickey-Fuller tests,
representing modifications of the original Dickey-Fuller tests which have been determined
to provide inaccurate results when serial correlation is present. To correct for this dilemma,
the augmented D-F tests incorporate a lagged, endogenous series at higher order which is not
present in the original Dickey-Fuller tests. By absorbing any serial correlation present in the
data, the augmented Dickey-Fuller tests are preferred, as the lagged, endogenous variable in
the regressions captures the autocorrelated influences which otherwise would appear in the
autocorrelated error term. Once the errors are white noise based on the augmented version
of these regressions, the presence of a unit root will be reported accurately when the
alternative MacKinnon critical values are applied.

Accordingly, I ran AD-F tests on all the series of data in this study in levels (that is,
the raw data without differencing) to determine which series were integrated, the results of
which are disclosed in Table 3.1. In every instance save for interstate relations and
miscellaneous cases in both the Supreme Court and Courts of Appeals, the null hypothesis
of a unit root could not be rejected for all three versions of the AD-F tests at the conventional
95 percent confidence level. Thus, every series besides interstate relations and miscellaneous

79

cases exhibited a nonstationary process and, as a result, must be transformed appropriately
before statistical analyses can be estimated with the data.

The type of data transformation I implemented was first differences, a process which
removes trends by eliminating the long-term component in the data, particularly with respect
to the mean. If data are integrated of the first order (that is, when the function in the
polynomial is equal to one), then an 1(1) series will become stationary by first differencing.
Generally speaking, differencing removes the unit root, such that |p| no longer equals one.
, First differencing is accomplished as follows:

Y, - Y“ = AY,

Once the data have been differenced, AD-F tests can be run to determine whether this data
transformation was successful in removing the unit root and achieving stationarity. Of
course, if first differencing the data is effective, then appropriate time-series techniques can
be accomplished.

Ergo, I differenced all the series of data which displayed a unit root (that is, all the
data except the interstate relations and miscellaneous cases in both courts). Then, I ran AD-F
tests on the transformed data, the results of which are reported in Table 3.2. As this table
demonstrates, first differencing was successful at eradicating the integrated processes in
every instance, as the null of a unit root could be rejected universally, and at the 99 percent
confidence level. Thus, the first-differenced data are stationary and can be subjected to time
series without the threat of spurious results.

Of course, the augmented Dickey-Fuller test is not the only manner of determining
whether the data are stationarity. Nevertheless, utilizing the Box-Jenkins (1976) approach

80

confirmed the results presented above. In fact, the only issue posed by ACF and PACF
outputs from these data was whether a log transformation was necessary to calm a changing
variance in some of the series. However, taking the logs of these series (in particular, the
Supreme Court criminal procedure and the Courts of Appeals economic cases) failed to
remedy any problems. Accordingly, the only transformation of the data I incorporated to
achieve stationarity was a frrst differencing of each series where necessary.‘0

The first method to be employed here is Granger causality (Granger 1969).“ Since
this procedure examines pair-wise exogeneity, it is ostensibly appropriate for exploring
whether the agenda of one of the federal courts in this study influences agenda changes in
the other court for a particular issue. Specifically, the Granger causality analysis assumes that
"if variable X (Granger) causes variable Y, then changes in X should precede changes in Y.
Therefore, in a regression of Y on other variables (including its own past values) if we

include past or lagged values of X and it significantly improves the prediction of Y, then we

 

loWhile the data attained stationarity due to this transformation, this balance came at
the expense of losing long-term information due to differencing. For this reason, an
alternative to differencing the data has been espoused by Engle and Granger (1987),
designated as cointegration (see also Granger 1981; 1983; Granger and Weiss 1983;
Johansen 1991). While beyond the scope of this study, the cointegration method combines
two or more series of data which are integrated at 1(1), the linear combination of which will
be stationary, 1(0), if the data are cointegrated, which enables the researcher to capture the
long-run relationships while producing nonspurious results (Banerj ee, Dolado, Galbraith, and
Hendry 1993). Irrespective of cointegration’s benefits, however, the first-differenced series
for those data which exhibited a unit root will suffice. For the series in which the raw data
were stationary, I will utilize those data in levels, with no transformations.

”AS Freeman (1983) advised, while this method is referred to as Granger causality,
Wiener (1956) evidently was the first scholar to connote this concept of causality.

81

can say that X (Granger) causes Y" (Gujarati 1995, 621). The test for Granger causality
involves estimating the following regressions:

X, = 20:3,, + SIS/XM- + e, (1)

Y, = SAX“. + 251-YM- + u, (2)
where the assumption of no correlation in the error terms is maintained.12 The researcher
determines the number of lags for the right hand side variables.‘3 The null hypothesis is that
lagged terms of the Y variable do not belong in equation (1), while the null for the second
regression is that lagged terms of the X variable are not appropriate. Thus, for equation (1),
HO: 2a,Y,,,- = O; and, for equation (2), HO: BAX” = 0. If the null can be rejected in either
of these regressions using standard F -tests, then that particular independent variable (Y in
equation (1) and X in equation (2)) provides a better explanation of the respective dependent
variable than the dependent variable does on its own. Under those circumstances, that
independent variable is said to Granger cause the other. Essentially, Granger causality
attempts to determine whether the contemporaneous dependent variable is explained only
by its own past values or whether lags of the independent variable provide for a fuller
accounting; moreover, the Granger approach facilitates in ascertaining whether additional

lags of the independent variable are necessary in the equation.

 

12This assumption is necessary in order to perform Granger causality regressions
(Engle, Hendry, and Richard 1983).

'3" One relies on the lagged values of the endogenous variables to capture autonomous
(purely endogenous) sources of variation in Y and X, that is, to take into account the
predictive capability of each variable’s own history" (Freeman 1983, 333, fn. 11).

82

Granger causality tests were estimated for all of the issue areas in this study; the
results are shown in Table 3.3. Of the major issues in this study, the strongest influence
derives from the Supreme Court for criminal procedure cases, where significant values are
apparent from lags four through ten. That is, the Supreme Court’s agenda for criminal
procedure cases precedes analogous changes in the Courts of Appeals for this issue.
Nevertheless, there is no corresponding effect arising from the lower courts on this issue.
At this early stage, then, these results comport with Hypotheses # 5a but not # 5 b. Among
the remaining significant issue areas, there is no Granger causality for civil liberties cases.
And, while there appears to be some Granger causality for both economic and federal
taxation cases, the influence stems only fi'om the Supreme Court at distant lags, as there is
no apparent endogenous effect emanating from the appeals courts for these issues. AS for
other endogenous effects, two issues of consequence remain: first, union and labor cases,
where there is Granger causality at three lags originating from the Supreme Court; and
second, economic issues, where the effect of the Courts of Appeals on the Supreme Court’s
economic agenda approaches but does not exceed significance at the conventional 95 percent
confidence level for ten lags (p = .056).

Granger causality tests provide preliminary information with respect to the
hypotheses I posited in this chapter. Nevertheless, as I contended in chapter 2, this approach
is inadequate as the sole or even primary method of analysis. At the outset, while Granger
causality addresses pair-wise exogeneity, it does so in a unidirectional fashion. Thus, it is

incapable of ascertaining endogenous causation within a system, as simultaneous feedback

83

is beyond its limits.14 Moreover, Granger causality may be a misnomer, as it does not
provide for causality in the customary, statistical definition. That is, while it demonstrates
whether changes in one time series precedes comparable movements in another, this method
does not establish whether these influences are causal in nature. Finally, problems endemic
to social science data, such as missing data, lack of variation, and measurement dilemmas,
"can confound Granger causal inferences" (Freeman 1983, 336). For these reasons, it is
axiomatic that I utilize other time-series approaches in order to provide for a fuller
explanation of causality within the federal courts with respect to their agendas.

One approach which serves these purposes well and provides for information beyond
Granger causality is VAR. As Sims (1980, 152) initially contended, the "incredible"
restrictions necessary for simultaneous modeling often result in very disparate results from
competing researchers, even when the same data are employed. That is, if the exogenous
restrictions implemented a priori in structural equations modeling are incorrect, then the
results and implications of the ensuing regressions, such as two-stage least squares, will be
misleading at best (Freeman, Williams, and Lin 1989). For this reason, Sims (153)
conceived of VAR, an approach which enables the researcher to "estimate large-scale
macromodels as unrestricted reduced forms, treating all variables as endogenous." By
utilizing Sims’ approach, VAR allows for feedback in a system, since it is a reduced form,
but it does so in a manner which considers all variables in the system as endogenous. Thus,

VAR is a very different kind of approach to statistical analysis, as it does not assume any

 

1“The sole manner in which Granger causality can detect feedback between X and Y
is if both null hypotheses can be rejected at the same time (Freeman 1983).

84

variables are explanatory. Because none of the variables are deemed to be exogenous, when
each variable in the system is regressed on past lags of all variables in the system, fewer
restrictions are imposed on the data, since the causal assumptions necessary in simultaneous
equations are relaxed (Freeman, Williams, and Lin 1989). "The strength of the VAR
approach is that the investigator tests these restrictions rather than imposing them in
advance" (Williams 1987). For these reasons, VAR is rather adept at concentrating on both
the direction and nature of dynamic, causal inferences; accordingly, it is especially pertinent
for analyzing the relationships between the agendas of the federal courts.

VAR embodies vectors and matrices of time series. In its compact notation, VAR
is represented as follows:

yt = m + Alyu + Azyu + + ApyH, + e,
where: A, denotes k x k matrices of coefficients; on is a k x1 vector of constants; e, is a k x1

white noise disturbance vector with a mean of O (E(e,) = O V t) with no serial correlation but

I

possible contemporaneous correlation; and, yt = [ym y2,, yh] . Lastly, the y, term

incorporates the conventional lag operator developed for the unrestricted reduced form. As
a consequence, in VAR each variable is depicted as a linear combination of the lagged values
of itself and all other variables in the system.

VAR circumvents the identification problem endemic to traditional structural
equations modeling by assuming no exogeneity among the variables; for this reason, this
approach dispenses with interpretation of coefficient estimates, since traditional parameter

estimation (such as in OLS or MLE) is meaningless and nonsensical in VAR due to the lack

85

of restrictions placed on the data.” While neutering the interpretability of parameter
estimates is an inherent sacrifice of VAR, one of its prominent features is in its ability to
track causal inferences over time through impulse response functions, otherwise known as
innovation accounting, and variance decomposition. Sims (1980) contends that these VAR
alternatives to interpretation of the [is demonstrate which variables in the system genuinely
are most consequential.

As one of these alternative forms of analysis used in VAR, innovation accounting
traces a one standard deviation Shock in the error of each equation (that is, it traces an
impulse or innovation) through the entire system of equations; in other words, this sort of
analysis models the effects of the innovation on current and future values of all the variables
in the system. More particularly, innovation accounting pursues the effects of the various
impulses on current and future values of all the variables in the system by converting the AR
representation of the data to a moving average representation which "expresses the yt vector
in terms of accumulated past Shocks or errors in the system" (Freeman, Williams, and Lin
1989, 846). By tracking into the future the response of the variables in the system to this
shock, innovation accounting enables the VAR modeler to discover which variables "have
causal priority, create the greatest response in the other variables in the system" (Granato and

Krause 1998, 10).“

 

150f course, if the researcher believes that all of the right hand side variables truly are
exogenous and no dynamics diffuse through a system, then OLS or MLE is appropriate.

“The innovations must be uncorrelated with each other. Choleski factorization,
which creates a triangular matrix in which the orthogonalized residuals are set forth on the
(continued...)

86

Variance decomposition is the complement to innovation accounting. Pursuant to
this type of dynamic analysis, "the decomposition of forecast error variance provides an
estimate of the amount of influence variables have in the system. . . By decomposing this
forecast error, we can determine the impact of one variable’s forecast on the error in
forecasting other variables. Thus, we can measure the effects that variables have on each
other over time" (Freeman, Williams, and Lin 1989, 847).” As a consequence, variance
decomposition does not demonstrate whether effects are significant (as innovation
accounting provides), but instead this technique yields the cumulative effect of one variable
on another in a system over time, measured in terms of a proportion or percentages (Granato
and Krause 1998).

Before proceeding to the VAR analyses, two issues must be settled. First, Should
data in levels or differences be used in the VAR models? Some VAR modelers insist that
the data should remain in levels and not be differenced, although logging variables, if
appropriate, is permissible (Williams 1995). Nevertheless, other scholars who have
incorporated VAR models in their research have utilized differenced data (Granato and
Krause 1998). Therefore, which should be used in this analysis, data in levels or in
differences? It seems that the answer depends in part upon the issues in which the researcher

is interested. Certainly, if long-term forecasting via VAR is to be employed, then

 

‘°(...continued)
diagonal, ensures that the impulses are uncorrelated (see Granato and Krause 1998 for a
fuller exposition on this topic). The Econometric Views program which I utilized to run the
various VAR models in this study incorporates Choleski factorization.

l7Variance decomposition tests apply Choleski factorization as well.

87

differenced data clearly are inappropriate, since stationarity via differencing comes at a cost
of losing any long-run components in the data. Nevertheless, if innovation accounting and
variance decomposition solely are to be utilized in order to determine dynamic causality, this
suggests that stationary data might provide more interpretable results, despite the adverse
consequences of differencing. Based on this appraisal, then, since I do not expect to estimate
long-run forecasts based on VAR, yet I will observe the innovation accounting and variance
decomposition measures in this study, I will employ stationary data achieved through
differencing, where appropriate. Nevertheless, I additionally will run the VAR models with
data in levels as a variety of sensitivity analysis, in order to confirm that the results of this
approach are robust.

The second issue concerns determining the appropriate lag length, in order to ensure
that the VAR models are specified properly (Granato and Krause 1998). One manner of
choosing the suitable lag length for the models simply is to use available rules of thumb, as
follows: if monthly data, 15 lags; if quarterly data, 6 lags; nevertheless, there is no general
rule of thumb for annual data (Williams 1995). More definitive methods for settling on the
number of lags for the system of equations, particularly when annual data are adopted as in
this study, include Sims’ (1980, 156, fn. 17) modified log-likelihood test, the Akaike
Information Criterion (AIC), and the Schwartz Criterion (SBC). The Sims test statistic for

determining the appropriate lag specification for VAR models is:

LR = (T-k) [Inlzkl 41421)” ~ X2 (k as.)

88

where: T = number of observations; k = number of estimated parameters in the unrestricted

model; InIERI = natural log of determinant residual covariance in the restricted model; and,

InIZUI = natural log of determinant residual covariance in the unrestricted model. For AIC

and SBC, when a different lag length is chosen, the best-specified model has the lowest AIC
or SBC value, which is the statistic closest to negative infinity (—<>o).l8

For purposes of this study, I have applied Sims’ modified log-likelihood statistic as
the primary tool, along with the AIC and SBC statistics as secondary diagnostics, to
determine the appropriate lag length for the VAR models. That is, as a decision rule I will
use the lag length specified by Sims’ modified log-likelihood test unless both the AIC and
SBC statistics instruct otherwise. With respect to calculating the suitable number of lags for
each VAR system, I have incorporated the reduction theory of model building.‘9
Accordingly, for each system I first will compare five lags to four lags. Then, if the
unrestricted model (the one with the greatest number of lags) is not significant, I will repeat
the process and compare four lags to three lags, then three lags to two lags, until I am able
to ascertain a VAR model with the most fitting number of lags for each system. Based on

this model-building approach utilizing these several statistics and decision rules, the

 

l8While Sims’ modified log-likelihood test has better small sample properties than
AIC, both statistics are the same asymptotically.

l”Reduction theory encompasses a general to simple fiamework. With respect to
specifying the appropriate number of lags for a VAR model, this model-building process
commences with a model consisting of the greatest number of lags (in this case, 5 lags) and
employs various diagnostics as lags are dropped in order to attain a model which is both
"theoretically and statistically defensible" (Granato 1991, 135).

89

 

 

appropriate lag lengths for the respective VAR systems, in both differences and levels, are

as follows:

data in differences:
civil liberties: 2 lags
economics: 2 lags
criminal procedure: 2 lags
union and labor: 3 lags
attorneys: 2 lags
federalism: 2 lags
federal taxation: 4 lags

data in levels:

civil liberties: 3 lags
economics: 2 lags
criminal procedure: 2 lags
union and labor: 2 lags
attorneys: 2 lags
federalism: 2 lags
interstate relations: 2 lags
federal taxation: 5 lags
miscellaneous:20 2 lags

Now that the preliminaries have been confronted, I am able to pursue the VAR
analyses at issue in this chapter. For the various systems, each of which includes the time
series of identical issues in both federal courts of interest, I estimated a VAR model and
proceeded to employ the innovation accounting and variance decomposition analyses; this
was accomplished for the data in both levels and differences, as appropriate. The results for
each particular VAR system follow.

According to innovation accounting for the differenced data, the greatest effect on

the agenda of the Supreme Court’s civil liberties agenda is the Court’s own agenda for this

 

2”The interstate relations and miscellaneous series only will be estimated in levels,
since those data exhibited stationarity in levels.

90

 

issue area, as there is no appreciable effect stemming fi'om the Courts of Appeals. When a
one standard deviation shock is applied to the agenda of the Supreme Court, the only
significant effect derives from the Supreme Court’s agenda, which is illustrated by Figure
3.1a. In a VAR innovation accounting graph, the solid line represents the impulse, while the
dotted lines represent + / - two standard errors (approximately 95%) surrounding that
impulse. If either of the standard error bars crosses the zero effect margin line at any point,
there is a statistically significant effect at about the conventional .05 probability. In Figure
3.1a, Since one of the standard error bars breaks the zero line at time period three, there is
statistical significance in response to that shock to the system. That is, the Supreme Court’s
civil liberties agenda is a significant influence on its future civil liberties agenda. To the
contrary, Figure 3.1b provides no analogous significance, since the standard error bars never
cross zero. Apparently, the justices do not consider civil liberties cases in the courts below
when accepting these cases on cert. As for the Courts of Appeals, Figure 3.1c demonstrates
that parties are influenced to a significant degree by their own civil liberties appeals through
two time periods. Moreover, the Supreme Court’s agenda for these cases also has a
significant, albeit Short lived, effect on these appeals in the circuit courts (see Figure 3.1d).

These results are confirmed when the data in levels are employed, as demonstrated
by the graphs in Figure 3.2. The only difference is that the findings in levels are much more
dramatic than those when differenced data are exploited. For instance, Figure 3.2a provides
that the Supreme Court is influenced by its own agenda decisions in civil liberties for nearly
the entire 10-year period, not just for a few years. As with Figure 3. lb, Figure 3.2b exhibits
no effect whatsoever stemming from the Courts of Appeals. And, while Figure 3.2d

91

illustrates that litigants are inspired by their own litigious appeals, the greater effect stems
from the Supreme Court itself, where parties react to changes in the Supreme Court’s agenda
for civil liberties for very nearly the entire period (see Figure 3.2c).

The variance decomposition results provide further evidence along these lines, as
decomposing the forecast error variance into the future provides that the Supreme Court has
strong influence in the Courts of Appeals, but that an endogenous effect does not appear to
influence the justices in their cert. decisions for civil liberties cases. For the differenced data,
Figures 3.3a and b illustrate that roughly 15 percent of the agenda of the Courts of Appeals
is determined by the Supreme Court’s agenda, an influence which remains fairly consistent
over time, while there is a negligible effect emanating from the lower courts. Tables 3.4a
and b provide the exact percentages for the values found in these variance decomposition
figures. When the VAR models incorporate the data in levels, once again the results are
striking. Figures 3.4a and b demonstrate that, while litigation in the circuit courts do not
affect the justices, the effect deriving from the justices is so forceful that by the end of the
time period, it is forecasted that litigious appeals are influenced more so by the Supreme
Court’s agenda than by the agenda of the lower courts themselves. Again, Tables 3.5a and
b provide the particular percentages of the variance decomposition.

Economic cases additionally provide for interesting results. The innovation
accounting in differences provides that the Supreme Court only is influenced by its own
agenda for this issue (see Figures 3.5a and b). Similarly, appeals in the lower federal courts
are most influenced by the litigants’ own behavior, as the Supreme Court has a slightly
significant effect in this regard (see Figures 3.5c and d). Once again, the data in levels

92

provide for more potent inferences. AS evidenced by the various graphs in Figure 3.6, the
Supreme Court affords statistically significant influence on the economic agenda in the
Courts of Appeals. Nevertheless, while litigious appeals are most persuaded by their own
history, the influence of Supreme Court’s agenda for this issue becomes significant
throughout segments of the period. As for the potential influence of the Courts of Appeals,
innovation accounting in levels demonstrates that economic appeals provide for a Slightly
significant effect on the Supreme Court; variance decomposition analysis confirms this
modest but nonetheless important finding. It is somewhat surprising that the Supreme Court
was not influenced to a greater degree than it was by the economic agenda as set by parties
in the courts below. Still, these analyses suggest that the Supreme Court contemporaneously
inspires agenda changes in the circuit courts and is motivated by agenda trends established
in the lower courts for economic issues (see Figure 3.7 and Table 3.6 for the differenced
results, along with Figure 3.8 and Table 3.7 for the levels results in levels).

The Supreme Court also appears to be influential when it comes to criminal
procedure cases. The innovation accounting analyses provide that the justices arouse agenda
changes in the Courts of Appeals to a nontrivial, even significant, degree, at least for a small
portion of the time period when the analysis iS run in differences. On the other hand, the
justices. are not at all motivated by analogous changes in the lower federal courts. Once
again, each court is affected chiefly by its own past history (see Figures 3 .9 [differences] and
3.10 [levels]). The variance decomposition results, however, demonstrate that the Supreme
Court’s influence is more profound than appears at first blush from the impulse response
functions, as a notable portion of the criminal procedure agenda for the lower courts is

93

determined by changes in the Supreme Court’s agenda for this issue (see Figure 3.11 and
Table 3.8 [differences] and Figure 3.12 and Table 3.9 [levels]).

Similar implications are obtained for union and labor cases, as innovation accounting
demonstrates that each court is most responsive to its own past agenda to a Significant
degree; however, the Supreme Court may be slightly susceptible to agenda changes in the
circuit courts for this issue, as faintly significant effects obtain. At this stage it is
inconclusive whether the justices influence parties in the lower courts, as the differenced data
provide for barely significant outcomes, while no such inference can be made from the
impulse response in levels (see Figures 3.13 [differences] and 3.14 [levels]). Variance
decomposition analyses are a bit more sensitive to these responses and, as a consequence,
shed additional light on causal influences for union cases. Specifically, while the effects are
not immediate, both the Supreme Court and Courts of Appeals are influenced to a
considerable degree by agenda changes in the other institution, as this conclusion can be
drawn from all variance decomposition analyses, save for the decomposition of forecast error
of the Courts of Appeals when level data are employed (see Figure 3.15 and Table 3.10
[differences] and Figure 3.16 and Table 3.11 [levels]).

As for attorneys cases, this generally nonsalient issue provides for somewhat
predictable results. That is, innovation accounting demonstrates that each court is most
influenced by its own actions but not those of the other, although when data in levels are
employed the Supreme Court appears to influence changes in the Courts of Appeals (see
Figures 3.17 [differences] and 3.18 [levels]). Variance decomposition confirms these results,
as the Supreme Court does not appear to respond to the agenda of the lower courts for this

94

issue, while the data in levels provide evidence that the Supreme Court engenders changes
in the courts below. The differenced data, however, do not permit such a strong conclusion
(see Figure 3.19 and Table 3.12 [differences] and Figure 3.20 and Table 3.13 [levels])

Federalism is another of the relatively minor issue areas, so it is not surprising to
observe relatively few influences between the courts. For both differenced and level data,
the innovation accounting demonstrates no cross-institutional influences (see Figures 3.21
[differences] and 3.22 [levels]). Similarly, there is little to no effect in the variance
decomposition graphs in either direction (see Figure 3.23 and Table 3.14 [differences] and
Figure 3.24 and Table 3.15 [levels]). Nearly identical results obtain for interstate relations
cases in both impulse response and variance decomposition analyses; that is, each court
influences only itself with respect to this issue (see Figure 3.25 for innovation accounting,
and Figure 3.26 and Table 3.16 for variance decomposition [levels only]).

Returning to a relatively salient issue, at least for the early portion of the relevant
time period, federal taxation demonstrates that this issue works within a system. For
instance, with respect to innovation accounting, the Supreme Court influences the agenda of
the Courts of Appeals to a Significant degree, particularly when level data are utilized.
Moreover, it appears there may be some feedback from the lower federal courts, as
demonstrated by the impulse response fimction in levels; however, no such effect arises
when differenced data are employed (see Figures 3.27 [differences] and 3.28 [levels]).
Variance decomposition suggests that the Supreme Court’s agenda changes for tax cases has
a strong influence on these appeals in the lower courts, hovering at about 20 percent. Solid,
long-term effects arise from the lower courts in the level data analysis, although the influence

95

seems not as convincing when differenced data removes this long-term component (see
Figure 3.29 and Table 3.17 [differences] and Figure 3.30 and Table 3.18 [levels]).

Finally, since there is no a priori expectation for any influence or direction thereof
with respect to miscellaneous cases, these results are reported merely for illustrative
purposes. Both the innovation accounting and variance decomposition depict that the past
history of each court is most influential on current and future histories, although it appears
there may be a slight effect stemming from the Courts of Appeals as well (see Figure 3.31
for innovation accounting, and Figure 3.32 and Table 3.19 for variance decomposition
[levels only]).“

DISCUSSION
The results of the Granger and VAR analyses in this chapter provide initial support

for my theory that the agendas of both the Supreme Court and Courts of Appeals affect each

 

21Earlier in this chapter, I briefly mentioned the debate among time-series scholars
concerning the use of differenced or level data in VAR analyses. Without solving this
dispute, I ran Johansen (1991) cointegration tests to determine whether the time series in this
study consisted of cointegrating vectors. If cointegrated, that would lend evidence that data
in levels should be utilized, since by definition time series must be integrated in order to be
cointegrated with each other.

Nevertheless, the results of these Johansen tests were inconclusive. Specifically,
some of the issue series definitively were cointegrated (these include: criminal procedure;
union and labor; attorneys; and, interstate relations). Others showed some but equivocal
evidence of vector cointegration (economic; federalism; federal taxation; and,
miscellaneous). Finally, civil liberties displayed no vector cointegration at all.

Accordingly, I report and rely on the VAR analyses utilizing both differenced and
level data. The results for the VAR analyses in levels and differences do not conflict with
each other; indeed, for the salient theoretical hypotheses and issues, in large part they are
corroborating.

96

other, albeit to varying degrees. The strongest conclusion which can be drawn at this stage
of the analysis is that the agenda priorities the justices of the Supreme Court establish clearly
influence their own future decisions on their agenda; additionally, parties in the appeals
courts recognize these trends and, in a rational manner, appeal cases in line with them, thus
further demonstrating the Supreme Court’s influence in setting the judicial agendas of the
federal courts, and not just their own agenda. Thus, while the Supreme Court clearly flexes
its agenda muscles in this regard, at the same time it is not entirely isolated in fashioning
agenda decisions, as it appears to be influenced by agenda priorities as determined by parties
in the federal courts below, at least for some issues areas.

As for the specific issues, the Supreme Court evidently has a potent influence on
litigious appeals in civil liberties cases. While the Granger tests were not entirely conclusive
in this regard, the VAR analyses leave little doubt that parties and their representatives
respond to priorities set by the justices concerning the civil liberties agenda. The justices,
however, do not at all consider appellate decisions on this issue when setting the Court’s
agenda. Accordingly, Hypotheses # 3a and # 3b appear to be verified, as the Supreme Court
has an influence beyond its own institutional walls but is not persuaded similarly by behavior
in the lower federal courts.

The results for economic issues also comported with the general aspects of my
theory. For instance, while the Supreme Court triggered trends in circuit court economic
appeals, it also was slightly influenced by the hypothesized endogenous effect emanating

from the lower courts. As well, while causality in this direction was not as strong as I had

97

anticipated, these results clearly were not negligible. Thus, the evidence provides support
for both Hypothesis # 4a and # 4b.

The criminal procedure results additionally provided convincing evidence of the
proposition that the Supreme Court influences parties in their litigious decisions to appeal,
as nearly every test and approach utilized in this study supported this aspect of the theory.
Nevertheless, the examination also evinces that the justices were not affected similarly by
criminal procedure appeals in the courts below. These analyses, then, ostensibly sustain
Hypothesis # 5a but not # 5b.

Union and labor cases provided support for both aspects of the theory, as not only did
the Supreme Court influence appeals below, but those same appeals constituted part of the
basis of the justices’ cert. decisions for this issue. Both Hypotheses # 6a and # 6b appear to
be bolstered by the analyses, then, as each court spurs corresponding agenda changes in the
other for union and labor issues.

There is no dispute that appeals in the circuit courts do not enter into the justices’
decisions to accept cases which concern attorneys; however, the results of this study do not
resolve the issue of whether the Supreme Court influences lower court appeals of attorneys
cases. Thus, while Hypothesis # 7b is supported, the evidence is inconclusive as to # 7a.
Additionally, the findings herein suggest that there is no cross-institutional influence in either
direction for cases concerning federalism or interstate relations issues. For these two
generally nonsalient issues, the only influence on agenda priorities appears to be each court’ 5
own past history, thus conflicting with Hypothesis # 8a and # 90 while indicating support for
Hypothesis # 8b and #9b.

98

The issue of federal taxation is another which provided support for the theoretical
proposition that agenda decisions in both of these courts cause changes in the agenda of the
other. Certainly, the Supreme Court influenced parties bringing federal tax appeals.
Moreover, as with both economic and labor issues, some feedback arose from the circuit
courts for this issue as well. As a consequence, these analyses intimate that both courts
affect the agendas of the other, thereby rendering evidence favoring Hypotheses # 100 and
# 10b.

At this stage of this study, then, the evidence appears to support my theory in general,
in that agenda trends established by the Supreme Court were complied with by litigants in
the appeals courts, as it is in their interests to do so. Furthermore, for some issues the
justices apparently were constrained to agree to hear certain types of cases they either would
rather not accept or did not realize were of consequence. While the results for economic
issues were not as profound as hypothesized, this issue in tandem with others, such as federal
taxation and union and labor cases, provide preliminary yet notable evidence of an
endogenous influence from the courts below. Thus, it appears that the Supreme Court has
an important influence beyond its institutional boundaries, as appellants recognize the
justices’ agenda priorities and respond accordingly. Nevertheless, the causality demonstrated
is not unidirectional, as the analyses in this chapter suggest that for some issues the water
seems to run uphill, at least to a certain extent, as rates of appeals in the lower courts have
considerable influence on the Supreme Court’ 5 agenda priorities over time. Thus, the agenda
of the Supreme Court in fact may be both an exogenous influence on lower court appeals as
well as an endogenous variable which responds to litigation in the courts below.

99

TABLE 3.]
AUGMENTED DICKEY-FULLER TESTS FOR UNIT ROOTS (SERIES IN LEVELS)

No Trend, —_lntercept Only Intercept and
No Interce__p__t Trend

Civil Liberties -0.238 -2. 187 - l .854

 

 

 

 

 

 

   
 
 
  
   
   
  

 

Economics -1.044 -3 .057* -4.597*
Criminal -0.495 -3 .771 ** -5 .302"
Procedure

Union and -1.613 -2.282 -2.978

Supreme Court Labor

Attorneys -1 .596 -2.709 -4. l 57*

Federalism -0.598 -3. l 74* -3 .496
Interstate -2.657** -4.81 7" -4.693 **
Relations

 

Federal Tax -1.844 -3 .3 54* -4.907**

 

   

Miscellaneous __ _

 

Civil Liberties -0.0727

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Economics -0.965 -1 .836 -2.094
Criminal -0.061 -2.088 -2.198
Procedure
Courts of Union and -.0730 -3.105* -3.207
Appeals Labor
Attorneys - l .668 -3 . l 77* -4.9 l 9
Federalism -1.884 -3.732** -3.507*
Interstate 4.456" -4.800** -4.763**
Relations
Federal Tax - l .943 - l .966 -3 .327
Miscellanesou f _ _ _ _-3.753** —3.685*
H0: unit root (data are not stationary) *pg .05
H: no unit root (data are stationary) "pg .01

Note: MacKinnon (1990) critical values are used for rejection of null hypothesis of a unit root.

100

TABLE 3.2

AUGM‘ENTED DICKEY-FULLER TESTS FOR UNIT ROOTS (SERIES IN FIRST DIFFERENCES)

Supreme Court

Ho: unit root (data are not stationary)
HA: no unit root (data are stationary)

No Trend,
_ No Intercept

-9.928**

m, ___ ___ ___ __ ._ __ i _ ________L _ ________________7
Intercept and
Trend

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Civil Liberties -10.052** -9.954**

Economics -9.102** -9.031** -8.951*"‘

Criminal -7.815** -7.730” -7.673**
Procedure

Union and —7.075** -7.074** -7.053**

Labor
Attorneys -7.692** -7.630** -7.552**
Federalism -8.44 1 ** -8.450** -8.376**
_ Fed______ -___ __ -10_—077*____ -9-95___

Civil Liberties -5.77l "”" -5.993** -5.927**

Economics -6.362"“'I -6.396** -6.492**

Criminal ~6.737** -6.741** -6.842"‘*
Procedure

Union and -5.219** -5.l62** -5.111**

Labor

Attorneys -8.606** -8.550** -8.518**

Federalism -7.744** -7.701** -7.6 l 7"

- an _ an: _ an.

Federal Tax 7.634 7.926 7.996
*ps.05
**ps.01

Note: MacKinnon (1990) critical values are used for rejection of null hypothesis of a unit root.

101

TABLE 3.3

GRANGER CAUSALITY RESULTS

 

9 Direction
' of

Supreme

Court

 

Granger

Causes

 

 

 

HA: Granger causality

*ps.05
**ps.01

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Direction
Issue Lag p-value of
__ Length I Causality
2 .415 I
3 .234 I Courts of
4 .156 i Appeals
Civil 5 .132 I Granger
Liberties 6 .138 Cause
7 .2924 Supreme
8 .578 ll Court
9 .755
10 .886
2 .427
3 .737 Courts of
4 .918 Appeals
Economic 5 .854 Granger
6 .263 Cause
7 .075 . Supreme
8 .077 Court
9 .117

 

 

 

102

 

Issue

Civil
Liberties

Economic

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lag p—value
___gfl‘ ___
2 .867 I
3 .771
4 .826
5 .535
6 .695
7 .703
8 .838
9 .879
__O___ i
2 .371 I
3 .402 I
4 .835 I
5 .920
6 .954
7 .814
8 .637
9 .340

Court

Granger

Causes
Courts of

Appeals

 

Supreme

I
I
I
I
Court
I

 

TABLE 3.3 (CONT’D)

 

Criminal

Procedure

Union and

Labor

Ho: ho Granger cauSality

HA: Granger causality

*ps.05
*‘ps.01

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.138
3 .248
4 .012*
5 .011*
6 .029*
7 .000**
8 .001**

9 002**
JE_JE1
2 .318
3 .0401
4 .120
5 .454
6 .615
7 .171
8 .314 I
9 .412 I
10 .509” I

  
 
 

 
   

Direction
of
Causality

Courts of
Appeals
Granger
Cause
Supreme

Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Issue
2
3
4
Criminal 5
Procedure 6
7
8
9
_mwwig__
2
3
4
Union and 5
Labor 6
7
8
9
l 0

 

 

103

 

I

 

TABLE 3 .3 (CONT’D)

       
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Direction : Diion
of Issue Lag p-value . of Issue Lag p-value
Causality Length Causality Lflth
2 .973 2 .468
Supreme 3 .764 Courts of 3 .665
Court 4 .85 1 Appeals 4 .693
Granger Attorneys 5 .914 Granger Attorneys 5 .508
Causes 6 .665 Cause 6 .668
Courts of 7 .768 Supreme 7 .651
Appeals 8 .761 Court 8 .595 I
9 .884 9 .287 I
2 .456 I 2 .112
Supreme 3 .469 Courts of 3 .145 '
Court 4 .385 Appeals 4 .109
Granger Federalism 5 .156 Granger Federalism 5 .248
.. Causes 6 .170 Cause 6 .322 I
Courts of 7 .239 Supreme 7 .461 I
Appeals 8 .313 Court 8 .378 I
9 .452 9 .591 I
10__ _.407U_ __W_U_ _”_______ _M__19“_ .777 I

 

 

Ho: mermr casaulity "
HA: Granger causality

‘ps.05
*‘ps.01

104

' Direction

TABLE 3.3 (CONT’D)

 

 

 

 

 

 

 

 

 

 

.835

 

.519

 

.140

 

.281

 

.191

 

.164

 

.170

 

 

 

 

 

 

 

 

 

 

 

 

I
I of Issue Lag p-value I
2 .549
Supreme 3 .693 Courts of
Court 4 .777 Appeals
Granger Interstate 5 .498 Granger Interstate
Causes Relations 6 .61 1 Cause Relations
Courts of 7 .700 Supreme
Appeals 8 .614 Court
9 .627
10 .814
I— 2 .131 ll
Supreme 3 .173 Courts of
Court 4 .136 Appeals
Granger Federal 5 .059 Granger Federal
Causes Taxation 6 . 1 83 Cause Taxation
Courts of 7 .201 Supreme
Appeals 8 .027" Court
9 .006"
1 0

 

Ho: no Granger causality

 

HA: Granger causality

‘ps.05
*‘ps.01

 

 

105

 

   

 

 

 

 

 

 

 

 

 

EcmqamnuNIgeaqaxmtsww.

 

TABLE 3.3 (CONT’D)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Direction Direction
of Issue Lag p-value of Issue
Causality Length Causality
2 .096
Supreme 3 .143 Courts of 3
Court 4 .l 10 ll Appeals 4
Granger Misc. 5 .119 II Granger Misc. 5
Causes 6 .297 Jl Cause 6
1|
Courts of 7 .505 Supreme 7
Appeals 8 . l 53 Court 8
9 .140 9 .
A. 5 I 7
10 .576 II 10 .794

 

 

 

II====E=I=
Ho: no Granger causality
HA: Granger causality

*ps.05
**ps.01

106

TABLE 3.4
VARIANCE DECOMPOSITION OF CIVIL LIBERTIES CASES (DIFFERENCED DATA)

   

 

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

Courts of Appeals

0.00
2 99.54 0.46
3 99.25 0.75
4 99.18 0.82
5 99.19 0.81
6 99.18 0.82
7 99.18 0.82 II
8 99.17 0.83
9 99.17 0.83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 11.69 88.31

3 12.35 87.65

4 14.15 85.85

5 14.96 85.04

6 15.06 84.94 ll
7 15.07 84.93 ll
8 15.10 84.90 ll
9 15.12 84.88

10

 

I___ __L _
Ordermg Supreme Court/Courts oprpeals 'i_,_______,.-,,,,,,v___- _I

107

TABLE 3.5
VARIANCE DECOMPOSITION OF CIVIL LIBERTIES CASES (LEVEL DATA)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I _ _ riance Decompositionof Courts__f Appeals

Courts of AP'Peas __

 

 

 

 

 

 

 

 

 

 

1 14.00 86.00

2 19.74 80.26

3 20.71 79.29

4 30.12 69.88
5 33.85 66.15 .
6 39.32 60.68
7 44.16 55.84 I
8 48.01 51.99 I
9 51.75 48.25 I
_ 10 _ 54.78 _ 45.22 _ EI
I

Weume Com/Courts of Arms

 

108

 

 

TABLE 3.6
VARIANCE DECOMPOSITION OF ECONOMICS CASES (DIFFERENCED DATA)

1 ‘Lmnmoofswreecm

 
 
     

   
  

 

   

 

     

 

 

 

 

 

 

 

 

 

 

I-_

1 100.00 0.00

I 2 99.69 0.31

I 3 98.14 1.86 I

L 4 94.95 5.05 I

I 5 94.20 5.80

I 6 94.27 5.73
7 94.12 5.88

I 8 93.98 6.02

I 9 93.95 6.05

I— n _ 10 395 6.05 _ _ I

_ Supreme Court _ 3 7 _ Courts of Appeals

 

 

I Variance Decomposition of Courts of Appeals J
' I

 

 

 

 

 

 

 

 

1 10.90 89.10
2 15.36 84.64
3 18.43 81.57
4 18.95 81.05
5 18.87 81.13
6 19.10 80.90
7 19.30 80.70
8 19.33 80.67
9 19.33 80.67

 

 

 

 

109

TABLE 3.7
VARIANCE DECOMPOSITION OF ECONOMICS CASES (LEVEL DATA)

 

Cums of ALPea __

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I
I--. _ __ __ ._ WSPPPCP“ . _I
I 1 16.90 83.10 I
I 2 14.44 85.56 I
I 3 21.02 78.98 I
; 4 21.94 78.06 I
5 23.53 76.47 1

6 24.26 75.74
7 24.87 75.13 I
8 25.27 74.73

9 25.58 74.42
10 5.80 ___ _ . __W

1.01de SW CW AP_ _

 

TABLE 3.8
VARIANCE DECOMPOSITION OF CRIMINAL PROCEDURE CASES (DIFFERENCED DATA)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1: 2. _ _ _ _Vnnace Decposomition of Cou_ ___ __ ___ ___w I

'_ _ ,_ a _ _ SuPeme 9"“ __ __ __ PPM ,-

I 1 100.00 0.00 I
2 99.64 0.36 I

I 3 98.94 1.06 I
4 98.91 1.09 I

I 5 98.85 1.15 I

I 6 98.83 1.17 I
7 98.83 1.17 I

l 8 98.83 1.17 I

n 9 98.83 1.17

10 98 83 1.17

I“ _ _ _ _ _ __ _, _____ __ __ _ _ __ __

I ariance Decomposition of Courts of Appeals

   

 

 

 

 

 

 

 

 

 

 

 

 

1 4.73 95.27
I 2 4.68 95.32
I 3 13.71 86.29

4 15.11 84.89

5 15.52 84.48
, 6 15.89 84.11
I 7 15.89 84.11
I 8 15.94 84.06
I 9 15.94 84.06
I 10 5-94 __ __

I

Mus MW00 of APPPP'S

111

 

L

TABLE 3.9
VARIANCE DECOMPOSITION OF CRIMINAL PROCEDURE CASES (LEVEL DATA)

fl _%ariance Decomposition of Supreme Court

 
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

1 1 100.00 000 J
1 2 98.16 184 .
f 3 96.29 371 i
1 4 94.94 506 1
i 5 94.07 593
I 6 9350 650 '
I 7 93.10 690 !
' 8 92.84 716 I
. 9 92.65 735 1
1_10 ________ __ _ __ _ __ _____ 4
.—1
—1
* 1
2 87.30 1

3 89.27 |

4 90.31 ‘

1 5 90.93 ’
I 6 91.31 |
l 7 91.54
I 8 91.69 1
l1 9 91.80 I
—1

 

112

Perm

TABLE 3.10

VARIANCE DECOMPOSITION OF UNION AND LABOR CASES (DIFFERENCED DATA)

 

Variance Decomposition of Supreme Court _

 

 

 

 

 

 

 

 

 

 

eriod

 

 

I Supreme Court Courts of Appeals i
g 1 100.00 0.00
' 2 99.51 0.49
I] 3 99.42 0.58

4 91.75 8.25
I 5 86.21 13.79 I
1 6 86.16 13.84 1
i 7 85.91 14.09 I
g 8 85.49 14.51
i 9 85.57 14.43 .
l 10 8.56 .44 !
1

—

Vanance Decomposition of Courtsof Appeals _,

1
W, ._ __#____1|

 

 

 

 

 

 

 

 

 

 

 

1-

; 1 0.35 99.65

i 2 12.48 87.52

E 3 12.42 87.58
4 13.55 86.45
5 21.13 78.87
6 21.22 78.78
7 21.22 78.78
8 21.30 78.70
9 21.22 78.78

Ordering: Supreme Court/Courts of Apals

 

113

TABLE 3.11

VARIANCE DECOMPOSITION OF UNION AND LABOR CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 95.88 4.12
3 91.45 8.55
4 89.03 10.97
5 88.05 11.95
6 87.67 12.33
7 87.54 12.46
8 87.49 12.51
9 87.48 12.52
10 87.47 12.53

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 3.41 96.59
2 2.97 97.03
3 2.93 97.07
4 2.93 97.07
5 2.93 97.07
6 2.93 97.07
7 2.93 97.07
8 2.93 97.07
9 2.93 97.07
10 2.93 97.07

 

 

 

Ordering: Supreme Court/Courts of Appeals

 

114

 

TABLE 3.12
VARIANCE DECOMPOSITION OF ATTORNEYS CASES (DIFFERENCED DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1
Period Supreme Court Courts of Appeals
1 100.00 0.00
2 99.75 0.25 ll
3 98.93 1.07 II
4 95.96 4.04
5 94.49 5.51
6 94.41 5.59
7 94.34 5.66
8 94.23 5.77
9 94.21 5.79
10 94.21 5.79
Variance Decomposition of Courts of Appeals 1|
Period Supreme Court Courts of Appeals J
1 6.52 93.48
2 6.99 93.01
3 6.94 93.06
4 7.17 92.83
5 7.35 92.65
6 7.36 92.64
7 7.37 92.63
8 7.38 92.62
9 7.38 92.62
10 7.38 92.62
Ordering: Supreme Court/Courts of Appeals

 

 

115

TABLE 3.13

VARIANCE DECOMPOSITION OF ATTORNEYS CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 99.34 0.66
3 99.36 0.64
4 99.24 0.76
5 99.23 0.77
6 99.21 0.79
7 99.21 0.79
8 99.21 0.79
9 99.20 0.80
10 99.20 0.80

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 16.83 83.17
2 20.54 79.46
3 27.20 72.80
4 28.54 71.46
5 29.73 70.27
6 30.09 69.91
7 30.31 69.69
8 30.39 69.61
9 30.43 69.57
10 30.45 69.55

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

—

 

116

 

TABLE 3.14
VARIANCE DECOMPOSITION OF FEDERALISM CASES (DIFFERENCED DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals

1 100.00 0.00

2 94.88 5.12

3 94.87 5.13

4 91.46 8.54

5 89.97 10.03

6 89.90 10.10

7 89.07 10.93

8 88.88 11.12

9 88.83 11.17

10 88.68 11.32 £-

Variance Decomposition of Courts of Appeals
Period Supreme Court Courts of Appeals

1 1.03 98.97

2 1.82 98.18

3 1.77 98.23

4 3.03 96.97

5 3.56 96.44

6 3.61 96.39

7 3.97 96.03

8 4.03 95.97

9 4.06 95.94

10 4.13 95.87

 

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

117

TABLE 3.15
VARIANCE DECOMPOSITION OF FEDERALISM CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 98.68 1.32
3 98.57 1.43
4 98.46 1.54
5 98.43 1.57
6 98.43 1.57
7 98.43 1.57
8 98.43 1.57
9 98.43 1.57
10 98.43 1.57
Variance Decomposition of Courts of Appeals
Period Supreme Court Courts of Appeals j
1 3.50 96.50
2 3.61 96.39
3 5.04 94.96
4 5.37 94.63
5 5.41 94.59
6 5.42 94.58
7 5.42 94.58
8 5.42 94.58
9 5.42 94.58
10 5.42 94.58
Ordering: Supreme Court/Courts of Appeals fl

 

118

TABLE 3.16
VARIANCE DECOMPOSITION OF INTERSTATE RELATIONS CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 99.83 0.17
3 99.20 0.80
4 99.04 0.96
5 99.04 0.96
6 99.02 0.98
7 99.02 0.98
8 99.02 0.98
9 99.02 0.98
10 99.02 0.98

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 97.67 2.33
2 97.42 2.58
3 97.42 2.58
4 97.40 2.60
5 97.40 2.60
6 97.40 2.60
7 97.40 2.60
8 97.40 2.60
9 97.40 2.60
10 97.40 2.60

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

 

119

 

TABLE 3.17
VARIANCE DECOMPOSITION OF FEDERAL TAXATION CASES (DIFFERENCED DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 98.35 1.65
3 96.74 3.26
4 96.85 3.15
5 95.64 4.36
6 95.13 4.87
7 95.20 4.80
8 95.26 4.74
9 95.24 4.76
10 95.15 4.85

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 5.46 94.54
2 16.87 83.13
3 18.45 81.55
4 18.10 81.90
5 20.56 79.44
6 21.08 78.92
7 21.08 78.92
8 21.06 78.94
9 21.26 78.74
10 21.49 78.51

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

 

120

 

TABLE 3 .18
VARIANCE DECOMPOSITION OF FEDERAL TAXATION CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 93.45 6.55
3 93.33 6.67
4 93.28 6.72
5 89.14 10.86
6 90.56 9.44
7 90.17 9.83
8 89.29 10.71
9 89.27 10.73
10 88.96 11.04

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 20.31 79.69
2 18.85 81.15
3 26.86 73.14
4 37.28 62.72
5 40.38 59.62
6 49.00 51.00
7 51.96 48.04
8 53.17 46.83
9 55.64 44.36
10 56.79 43.21

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

 

121

 

TABLE 3.19

VARIANCE DECOMPOSITION OF MISCELLANEOUS CASES (LEVEL DATA)

 

Variance Decomposition of Supreme Court

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Period Supreme Court Courts of Appeals
1 100.00 0.00
2 88.95 1 1.05
3 87.81 12.19
4 86.42 13.58
5 86.37 13.63
6 86.37 13.63
7 86.37 13.63
8 86.37 13.63
9 86.37 13.63
10 86.36 13.64

Variance Decomposition of Courts of Appeals

Period Supreme Court Courts of Appeals
1 2.58 97.42
2 10.69 89.31
3 12.33 87.67
4 13.88 86.12
5 14.12 85.88
6 14.15 85.85
7 14.15 85.85
8 14.15 85.85
9 14.15 85.85
10 14.15 85.85

 

 

 

 

Ordering: Supreme Court/Courts of Appeals

 

122

 

Civil Liberties Cases (Differences)
Response to One 8.0. Innovations :1: 2 S.E.
(a) 0))
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals

 

64

 

 

 

 

IbéPPPPee

 

 

 

 

JD

 

 

 

 

 

 

4

 

 

 

15315576510

(9) (d) '
Response of Courts of Appeals to Supreme Court Response of Courts of Appeals to Courts of Appeals

FIGURE 3.1
INNOVATION ACCOUNTING OF CIVIL LIBERTIES CASES (DIFFERENCES)

123

Clvll Liberties Cases (Levels)
Response to One SD. Innovations :1: 2 S.E.

(a) 6))
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e 5
‘1‘ .4
4.
2|
" ...........
n I'
V \
\
‘\
.2- \‘ ................................ 1‘
4 4
1 2 3 4 5 0 7 3 B 10 1 2 3 ‘ 5 C 7 B D 10
4 4
h\
3 3. x
ooooooo ‘\ ~
I ------------- ‘ ““ ‘
2 ---- ”I ‘‘‘‘‘ 2" O \‘
l ----- ’ “ § ................
\‘ .........
“\N v 1 .‘
‘
’~ ‘
e ..... A"I ‘§~.“. ................. J 9 \‘;.’-~~“
.14 .1 QQQQQQQQQQQ
.2 1 V J- 1 ‘
1 2 3 4 5 C 7 B 9 ‘0 1| 2 3 4 5 O 7 C D 10

 

(‘31 «0
Response ofCourtsoprpealstoSupremeCourt ResponseofCourtsoprpealstoCounsoprpeals

FIGURE 3 .2
INNOVATION ACCOUNTING OF CIVIL LIBERTIES CASES (LEVELS)

124

Civil Liberties Cases (Differences)

(a)

Variance Decomposition of Supreme Court
100

 

80..

 

— Supreme Court
---- Courts of Appeals

 

40.:

 

 

20..

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

 

80.

60.

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

204

 

 

 

 

12846678910

FIGURE 3.3

VARIANCE DECOMPOSITION OF CIVIL LIBERTIES CASES (DIFFERENCES)

125

Civil Liberties Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100

 

 

 

 

804

 

— Supreme Court
---- Courts oprpeals

 

 

 

 

 

 

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

80.

 

— Supreme Court
---- Courts of Appeals

 

 

 

 

 

 

0........
12345678910

 

FIGURE 3.4
VARIANCE DEOOMPosmON OF CIVIL LIBERTIES CASES (LEVELs)

126

 

Economics Cases (Differences)
Response to One 8.0. Innovations :r: 2 S.E.

(a) (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSupremeCourttoCourtsoprpeals
1o

 

 

 

 

 

 

 

 

 

 

 

 

D
/.
I
I
‘1
~§
\
I
I
I
r
r
r
r
r
1
r
r
I
I
a
o
o
|
r
|
r
o
r
r
a
r
r
r
r

 

 

 

IL

 

 

 

 

(c) (d)
ResponseochurtsoprpealstoSupremeCourt ResponseofCourtsoprpealstoCourtsoprpeals

FIGURE 3.5
INNOVATION ACCOUNTING OF ECONOMICS CASES (DIFFERENCEs)

127

Economics Cases (Levels)
Response to One 8.0. Innovations t 2 S.E.

 

0))
Response of Supreme Court to Courts of Appeals

 

 

 

 

 

Response of Supreme(ac)ourt to Supreme Court

.9 10
111
s.
4.
2.
o
.g

 

 

 

 

 

 

 

4.

 

 

-2

 

 

 

1b

 

 

 

 

 

1

 

1254511715910 1

' (c) (1!)
Response ofCourts oprpeals to Supreme Court Response ofCourts oprpealsto Courts oprpeaIs

FIGURE 3.6
INNOVATION ACCOUNTING OF ECONOMICS CASES (Law)

128

Economics Cases (Differences)

(a)

Variance Decomposition of Supreme Court

100..__\

 

 

80.

 

60.
— Supreme Court

---- Courts of Appeals

 

 

 

40.

20.

 

 

 

(b)
Variance Decomposition of Courts of Appeals

 

100

h
‘s
-
m
s
......
-----------------------------------------

80..

 

60 .
— Supreme Court

---- Courts of Appeals.

 

 

 

40..

 

20.

 

 

 

1 2 £1 4 S e 7 8 9 10
FIGURE 3.7
VARIANCE DECOMPosmON OF ECONOMICS CASES (DIFFERENCEs)

129

Economics Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100.
\

80.

 

60.

 

— Supreme Court
---- Courts of Appeals

 

 

 

40.

20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

804 ~~‘~ ____________

60..

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

 

20- \/—/

 

 

 

1 2 s 4 s s 7 a 9 10
. FIGURE 3.8
VARIANCE DECOMPOSITION OF ECONOMICS CASES (Mm)

130

Criminal Procedure Cases (Differences)
Response to One 8.0. Innovations :t 2 S.E.
(a) (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSrpremeCourttoCourtsoprpeaIs

 

4.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(e) (d)
ResponseotCourtsoprpealstoSupremeCourt RespomeofCourtsoprpealstoCourtsotAppeals

FIGURE 3.9
INNOVATION ACCOUNTING OF CRIMINAL PROCEDURE CASES (DIFFERENCES)

131

 

Criminal Procedure Cases (Levels)
Response to One 8.0. Innovations :1: 2 S.E.

(‘1 (b)
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(0) (d)
ResponseotCourtsoprpealstoSupmme Court Response ofCourts oprpealstcCourts oprpeals

FIGURE 3.10
INNOVATION ACCOUNTING OF CRIMINAL PROCEDURE CASES (LEVELS)

132

Criminal Procedure Cases (Differences)

(a)

Variance Decomposition of Supreme Court
100T

 

80.

601

 

— Supreme Court
40 ---- Courts of Appeals
I

 

 

 

204

 

 

 

(b)

Variance Decomposition of Courts of Appeals

100

 

h-----~
\
~
\
5
\‘
D--
-------------------------------------

804

60.

 

— Supreme Court
---- Courts of Appeals

 

 

40..

 

20.

 

 

 

 

1 2 3 4 s 6 7 8 s 10
FIGURE3.11
VARIANCE DECOMPOSITION OF CRIMINAL PROCEDURE CASES (DIFFERENCES)

133

 

 

 

Criminal Procedure Cases (Levels)

(a)

Variance Decomposition of Supreme Court

1001\

 

80.

60.

 

— Supreme Court
40. ---- Courts of Appeals

 

 

 

20.

 

 

------------------------------------
......
O---
_____

 

(b)

Variance Decomposition of Courts of Appeals
100

 

gggggg
------
.......

80.

60.

 

— Supreme‘Court
4O ---- Courts of Appeals

 

 

 

20.

 

 

 

12545678010
FIGURE3.12

VARIANCE DECOMPosmON OF CRIMINAL PROCEDURE CASES (LEVELS)

134

 

 

 

Union and Labor Cases (Differences)
Response to One 8.0. Innovations :t 2 S.E.

 

 

 

 

 

 

(a) . (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSupremeCourttoCourtsoprpeais
s , s

2.
1 ’l"‘\‘
o----:' \‘
l’ ‘\ "' ~‘s o" “~~---.
0 2‘ V A \‘I/Av '-..
‘\‘~‘.”’ \“ I’I ..... 0'
.1 . ‘\.l’
.2 I ‘U 1
1| 2 s 4 s s 7 a 0 1o

 

 

 

 

 

01
Id

 

 

 

 

 

 

1b
is

 

 

 

(e) (d)
ResponseofCourtsoprpeaistoSupremeCourt ResponseofCourtsotAppealstoCourtscprpeals

FIGURE 3 .13
INNOVATION ACCOUNmIG OF UNION AND LABOR CASES (DIFFERENCES)

I35

 

Union and Labor Cases (Levels)
Response to One 8.0. Innovations :1: 2 S.E.

 

 

(a) (b) '
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals
2.5 as
210. 20.
1s. 1.5.

1.0-

0.5.

 

0.0.

 

 

 

 

 

 

 

 

ID

 

 

145

 

---
.......
.‘
"

.1

 

 

 

 

 

 

(c) (d)
ResponseofCourtsoprpealstoSupremeCourt ResponseofCourts oiAppeaistoCourtsoprpeals

FIGURE 3.14
INNOVATION ACCOUNTING OF UNION AND LABOR CASES (LEVELs)

136

 

Union and Labor Cases (Differences)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a)
Variance Decomposition of Supreme Court
80.
60.
-— Supreme Court
---- Courts of Appeals
40.
20.
0 --- "4"" .
1 2 3 4 5 6 7 8 9 10
1b)
Variance Decomposition of Courts of Appeals
100R.
80. ................................
60.
— Supreme Court
---- Courts oprpeals
40.
20.
OH

 

 

123745878910

FIGURE 3.15
VARIANCE DECOMPOSIIION OF UNION AND LABOR CASES (DIFFERENCEs)

137

 

Union and Labor Cases (Levels)

(a)

Variance Decomposition of Supreme Court

 

100

 

80.

 

— Supreme Court
——-- Courts oprpeals

 

 

 

40..

20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals

 

100 ___________________________________________________

80.

 

60. .
— Supreme Court

---- Courts of Appeals

 

 

 

40.

20.

 

 

 

 

1 2 5 4 6 6 7 is 5 10
FIGURE 3.16
VARIANCE DECOMPOSITION OF UNION AND LABOR CASES (LEVELS)

138

Attorneys Cases (Differences)
Response to One 8.0. Innovations :1: 2 S.E.
(a) (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSupremeCourttoCourtsoprpeals

1K

 

 

1.0.

0.5.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(c) (d)
Response ofCourts oprpeaIs to Supreme Court Response ofCourts oprpeals to Courts oprpeals

FIGURE 3.17
INNOVATION ACCOUNTING OF ATTORNEYS CASES (DIFFERENCES)

139

 

 

 

Attorneys Cases (Levels)
Response to One 8.0. Innovations :1: 2 S.E.

(a) (b)
ResponseoISupremeCourttoSupremeCourt ResponseofSupremeCourttoCourtsoprpeals

 

 

 

 

 

 

 

 

 

 

 

 

I; 11
1.040‘ 1.04
\
0.0. 1. 0.0.
1
0.0. ‘\ 0.0.
‘ 0
0.4. ‘ 0.4. '
01' 02' I" ‘~~‘ .....
’ .........
9-9 0.0. k? ""'_'::_-: --
\“ ' ...............
-02 -0.2. x-..“/ I
0" I '4 I U I I“ ‘l I I I
12 5070010 12345070010
M as

 

 

 

 

 

 

 

 

 

 

 

 

(6) (60
Response ofCourts oprpeals to SupremeCourt Response ofCourts oprpeals to Courts oprpeals

FIGURE 3.18
INNOVATION ACCOUNTING OF ATTORNEYS CASES (LEVELS)

140

f

 

Attorneys Cases (Differences)

(a)

Variance Decomposition of Supreme Court

100q—\

 

 

80.

60.

 

—— Supreme Court
,"" Court oprpeals

 

40.

20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

....................................................

80.

60..

 

— Supreme Court
40 _---- Court of Appeals

20.

 

 

 

 

1 2 8 4 8 8 7 8 9 10
FIGURE 3.19
VARIANCE DECOMPOSITION OF ATTORNEYS CASES (DIFFERENCEs)

141

 

Attorneys Cases (Levels)

 

 

Variance Decompositai8n of Supreme Court
100
00.
00.
40.
20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals

 

100

i...
g
- ..‘
~
‘
~
‘~
.....
-------------------------

 

60.

40.

20./

12348878910

 

 

 

 

FIGURE 3.20

 

 

— Supreme Court
---- Courts of Appeals

 

 

 

 

— Supreme Court
---- Court of Appeals

 

 

VARIANCE DECOMPOSITION OF ATTORNEYS CASES (LEVELs)

142

Federalism Cases (Differences)
Response to One 8.0. Innovations :1: 2 S.E.
(a) (b)
ResponseofSupremeCourttoSupmmeCoun ResponseofSupremeCourttoCourtsoprpeais

S

 

M

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(e) ((0
Response of Courts of Appeals to Supreme Court Response of Court of Appeals to Court of Appeals

FIGURE 3 .21
INNOVATION ACCOUNTING OF FEDERALISM CASES (DIFFERENCES)

143

Federalism Cases (Levels)
Response to One 8.0. Innovations t 2 S.E.

(a) (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSupremeCourttoCourtoprpeals

1 s

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(e) ((0
Response of Court oprpeaIs to Supreme Court Response of Court oprpeais to Court oprpeals

FIGURE 3.22 .
INNOVATION ACCOUNTING OF FEDERALISM CASES (LEVELs)

144

 

4’2!

 

Federalism Cases (Differences)

(a)

Variance Decomposition of Supreme Court
100

 

 

80.

60‘

 

‘ -— Supreme Court
40. ---- Court oprpeaIs

 

 

 

20.

 

 

.................
a
..........
O
'O
O

 

 

(b)
Variance Decomposition of Courts of Appeals

100 {F

80.

60.

 

— Supreme Court
40. 7"" Court of Appeals

 

 

 

20.

 

 

 

12348878910

. FIGURE 3.23 .
VARIANCE DECOMPOSITION OF FEDERALISM CASES (DIFFERENCES)

145

 

 

 

Federalism Cases (Levels)

(a)

Variance Decomposition of Supreme Court

 

 

1001‘;

80.

60.

 

— Supreme Court
---- Court of Appeals

 

40.

201

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

80.

60.

 

— Supreme Court
40 7---- Court of Appeals

20.

 

 

 

 

 

1 2 8 4 8 8 7 8 9 10
FIGURE 3.24
VARIANCE DECOMPOSITION OF FEDERALISM CASES (LEVELs)

146

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Interstate Relations Cases (Levels)
Response to One 8.0. Innovations 3: 2 S.E.
(a) 0))
Response of Supreme Court to Supreme Court Response of Supreme Court to Court of Appeals

1.0 , ‘ ‘ 1.0

00 00.

0.0 0.0.

0.4. 0.4.

02. 0.2.

0.0. 00

02. 01

-0.4 1 fl . , -0.4

1 2 3 4 0 0 7 0 0 10

0.1.2 0.1.2

000. 0001

0.04. 004.

000. _v.——..1..11....I..u............~ """" . 2"": :::: 0.00

-0.04. ----- “xx” .004.

-0.00 , 1 «0.00 ,

12340070010 12340070010

 

(G)

(d)
ResponseofCourtoprpeaistoSupreme Court ResponseofCourtoprpeaistoCourtoprpeals

FIGURE 3.25
INNOVATION ACCOUNTING OF INTERSTATE RELATIONS CASES (LEVELS)

147

‘T‘W J,

 

 

Interstate Relations Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100
a

80.

 

 

— Supreme Court
---- Courts of Appeals

 

 

 

40.

20.

 

 

 

 

(b)

Variance Decomposition of Courts of Appeals

 

100

 

80.

 

60 .
— Supreme Court

---- Court of Appeals

 

 

 

40.

20.

 

 

1

12348878010

 

 

FIGURE 3.26
VARIANCE DECOMPOSITION OF INTERSTATE RELATIONS CASES (LEVELs)

148

I

Federal Taxation Cases (Differences)
Response to One 8.0. Innovations :1: 2 S.E.
(a) 0))
Response of Supreme Court to Supreme Court Response of Supreme Court to Court of Appeals

3

 

 

2.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\ \ i s
9 I, ‘s I, ‘s .o"‘\‘
‘ ' m ‘V' A ‘---- A“
\ I
e ‘7 w . \
5‘ ”---.“V ’I\~~ ""' hm.
\ I ‘ ” \.’--
‘\ f ‘s ’I
\ ’ ‘
‘11 ‘\’
I
.2 -2
‘ fi 1 V I V ' I V V I V V j V I
1 2 3 4 5 0 7 ‘0 0 10 1 2 3 4 5 0 7 0 0 10

(e) (d) -
Response ofCourt oprpeaIstoSupreme Court Response ofCourt oprpeals to Court oprpeaIs

FIGURE 3.27
INNOVATION ACCOUNTDIG OF FEDERAL TAXATION CASES (DIFFERENCEs)

149

Federal Taxation Cases (Levels)
Response to One 8.0. Innovations :1: 2 S.E.

 

 

 

 

 

 

 

(a) 0))

Response of Supreme Court to Supreme Court Response of Supmme Court to Court of Appeals
3 s

2.

1. ,.\ .

I” ‘ ‘_---'v" ~~‘ ...~ ......
0 I ‘ .—-.. ..............
.‘ v v v I V v V
2 3 4 0 0 7 0 0 10

 

 

 

 

 

 

 

1.5-I ,o~‘~
I

I ------ Q.
s
‘u

1.01 \ ”

0.5.

 

-
...-"’. 3‘

 

 

 

 

 

 

 

.1“ ‘I‘.fi'1"'
12340070010 12340070010

 

(0) (6)
Response of Court of Appeals to Supreme Court Respdnse of Court of Appeals to Court of Appeals

FIGURE 3.28
INNOVATION ACCOUNTING OF FEDERAL TAXATION CASES (LEVELs)

150

 

Federal Taxation Cases (Differences)

(a)

Variance Decomposition of Supreme Court

 

1001‘

80.

60.

 

— Supreme Court
40. ---- Court of Appeals

 

 

 

20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

80.

60.

 

— Supreme Court
40 ---- Court of Appeals

 

 

 

20.

 

 

 

 

12345878910

FIGURE 3.29

VARIANCE DECOMPOSITION OF FEDERAL TAXATION CASES (DIFFERENCEs)

151

Federal Taxation Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100

 

 

\___,_L
80.

60.

 

— Supreme Court
40 --—- Court of Appeals

 

 

 

20.

 

 

 

(b)

Variance Decomposition Of Courts of Appeals
00

70-I\__\/_________/———

60.

 

 

50 . — Supreme Court
---- Court of Appeals

 

 

 

40.

.......
.................
........
...........
30. ..........

 

 

20

 

FIGURE 3.30
VARIANCE DECOMPOSITION OF FEDERAL TAXATION CASES (LEVELs)

152

Miscellaneous Cases (Levels)
Response to One 8.0. Innovations t 2 S.E.
(a) 0»
Response of Supreme Court to Supreme Court Response of Supreme Court to Court of Appeals
1.9 1.0

 

 

   

0.0. 008
0.01

0.4 .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(0) ((0
Response ofCourt oprpeals to Supreme Court Response ofCourt oprpeals to Courts oprpeaIs

FIGURE 3.31
INNOVATION ACCOUNTING OF MISCELLANEOUS CASES (LEVELs)

153

Miscellaneous Cases (Levels)

(a)

Variance Decomposition of Supreme Court

100

 

80.

60.

40.

20.

 

 

 

— Supreme Court
---- Court of Appeals

 

 

 

 

 

 

(b)

Variance Decomposition of Courts of Appeals

100

 

80.
60.
40.

20.

 

 

— Supreme Court
---- Court of Appeals

 

 

 

 

 

 

2 3 4 8 8 7 8 9 10
FIGURE 3.32
VARIANCE DECOMPOSITION OF MISCELLANEOUS CASES (LEVELs)

154

THE NATURE OF AGENDA IN THE UNITED STATES SUPREME COURT
AND COURTS OF APPEALS

VOLUME II
By

Mark S. Hurwitz

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

1998

CHAPTER 4

A CLOSER EXAMINATION OF THE RELATIONSHIP BETWEEN
THE CIVIL LIBERTIES AGENDAS OF THE SUPREME COURT
AND COURTS OF APPEALS

Civil liberties cases are among the most salient of all issues decided by the judiciary
in the United States. As Figures 1.1 and 1.16 demonstrate, these types of cases formed about
one third of the docket of the Supreme Court over the time period of this study.1 These
statistics, however, do not include criminal procedure cases. Thus, when criminal procedure
issues are included in this appraisal,2 the percentage of civil liberties cases heard by the
Supreme Court is Significantly higher. While the relative numbers were not as high in the
Courts of Appeals, they certainly were far from trivial, as depicted in Figures 1.1 and 1.17.
From a qualitative vieWpoint as well, civil liberties decisions by the judiciary have been
exceedingly important for the latter two-thirds of this century. While this trend began to be
noticeable in the late 19305 (Pritchett 1948; 1954), more particularly the drive behind the

activism of the Warren Court was fueled predominantly by civil liberties cases (Pacelle

 

1This, of course, is contrasted with the rather slight emphasis the Supreme Court
placed on civil liberties before the "Switch" in 1937 (Pacelle 1991; McCloskey 1994).

2Segal and Spaeth (1993, 106-111) included the issue of criminal procedure within
their discussion of the overall civil liberties agenda of the Supreme Court; see also Segal and
Cover (1989); Lanier (1998).

155

1991 ; Baum 1995); and, the success rate among individuals claiming civil liberties violations
rose in the 19508 and 19605 to the point where these litigants won a majority of these cases
before the Supreme Court (Adamany 1991). The Burger and Rehnquist Courts diminished
both the number and consequence of these cases on the Supreme Court’s docket, and
Adamany (1991) documented that the success rate among individuals alleging liberties
violations subsided in this recent era as well. Nevertheless, civil liberties cases continue to
comprise a brand of salience among the justices, the judicial system, and the political process
in general not found among any of the other issue heard in the federal courts.

Whether because of the salience of civil liberties issues, or possibly because they are
considered among the more interesting of issues, many scholars have researched civil
liberties as part of their judicial behavior studies. This type of scholarship, particularly those
studies which are empirically based, usually envelops a multitude of issues within the rubric
Of civil liberties. Specifically, this field encompasses civil rights, first amendment, due
process, and privacy cases.3 The prior chapter is one such example of research which
combined these various issues into a general civil liberties category, and there are an
abundance ofothers.4 On the other hand, the scholarship dissecting the various issues which
constitute the whole of civil liberties is much more deficient. In fact, the issue of civil
liberties is so broad that it is logical to disentangle the derivative issues, for two primary

reasons. First, since the issue so often is aggregated, we need to determine whether any

 

3Each of these issues is defined in both this chapter and in Appendix A, infia.
4See, e. g., Segal and Cover (1989); Reddick (1997).

156

differences indeed do exist among these issues, SO as to validate this fairly standard practice
of categorization within the subfield. Second, each of these issues is important in its own
right and deserving of its own Study, regardless of the various methodological implications.
Accordingly, in this chapter I break down the civil liberties issue into a set of separate
components in order to determine whether the results I obtained in the prior chapter for civil
liberties en masse apply to each issue. That is, I am interested in ascertaining whether, and
if so the extent to which, the agenda of the Supreme Court for specific civil liberties issues
influences appeals in the Courts of Appeals over time. Analogously, in this chapter I
examine whether endogenous forces stemming from the lower federal courts for these issues

have any effect on the agenda of the Supreme Court.

ISSUES WITHIN CIVIL LIBERTIES, AGENDAS, AND THEORETICAL SUPPOSITTONS

As described in the prior chapter, prior to the Switch in 1937, the Supreme Court
decided relatively few civil liberties cases each year. Indeed, one study which specifically
addressed the cumulative civil liberties agenda of the Supreme Court concluded that the
proportion of these decisions remained "relatively fla " from 1888 through 1936 (Lanier
1998, 10). The year 193 7, however, proved to be a monumental one for the Supreme Court,
as political science and legal scholarship have documented ad museum, for after the Switch
the Court adjusted its priorities away from conservative economic decisions in favor of more
liberal civil liberties outcomes. This shift first occurred even without a change in personnel
on the Bench, although the trend toward upholding civil liberties claims of individuals
quickly accelerated once the conservative wing of the Court was replaced with President

157

Roosevelt’s appointees. Decisions such as Palko v. Connecticut, 302 US 319 (193 7), and
United States v. Carolene Products, 304 US. 144 (1938), represent abounding examples of
the newer focus and direction from pre-Switch cases such as Lochner v. New York, 198 US.
45 (1905), Schechter Poultry Corp. v. United States, 295 US. 495 (1935), and Carter v.
Carter Coal C0., 298 US. 238 (1936).

There is no questioning the accretion of an emphasis on civil liberties in the agendas
of both the Supreme Court and the Courts of Appeals after 1937. See Figure 1.1.
Accordingly, the inquiry necessarily becomes, were the resultant trends in the agenda for all
civil liberties issues Similar, at least with respect to the Supreme Court’s agenda? As Figures
1.2 through 1.5 evidence, the answer is No, since the discrete civil liberties issues were
treated disparately on the dockets of each of these federal courts. The following sections
address in some detail each of these issues which comprise the civil liberties category.
Civil Rights

In a variety of contexts, the terms ‘civil rights’ and ‘civil liberties’ often have been
employed interchangeably. In part, this is likely because the overall issue has been
aggregated to incorporate its separate elements. Notwithstanding, I deliberately do not
utilize that practice here, as each label has a very specific and definitive meaning in this
study. For purposes of this study, then, the former classification represents a subset of the
later; that is, the category of civil liberties includes civil rights cases, along with the other

classes of liberties cases?

 

5The remaining categories within civil liberties which I will analyze, including first
(continued...)

158

With specific reference to civil rights, as detailed in the documentation for the
Supreme Court Database (1995 , 57): "Civil rights includes non-First Amendment freedom
cases which pertain to classifications based on race (including American Indians), age,
indigency, voting, residency, military or handicapped status, gender, and alienage." Figures
1.6 and 1.7 illustrate that issues incorporating civil rights comprised the majority of civil
liberties cases over the course of the relevant time period in the Supreme Court and Courts
of Appeals, respectively. AS a consequence, the trends for civil rights cases and the whole
Of civil liberties have been rather similar over time (compare Figures 1.1 and 1.2).

From a theoretical perspective, the relationship between the civil rights agendas of
the Supreme Court and Courts of Appeals should be very similar to the expectation in the
general civil liberties category, as asserted in chapter 3. To reiterate, Hypothesis #3a
provides that litigants in the Courts of Appeals should respond to changes in the civil
liberties agenda of the Supreme Court. For the same reasons cited in reference to that
hypothesis, including salience, controversy, intense beliefs, perceived role of the Supreme
Court as guardian of individual liberties, and spillover effects, parties in the federal appeals
courts Should be influenced by changes in the Supreme Court’s agenda on civil rights. This
prospect leads to the following hypothesis:

Hypothesis # 11a. The agenda of the Courts of Appeals in
civil rights cases should respond directly to changes in the

agenda of the Supreme Court in either a positive or negative
direction.

 

’(...continued)
amendment, due process, and privacy cases, are addressed in detail, infra.

159

For these same reasons, however, there is little reason to believe that the justices will be
motivated by trends in the civil rights agenda in the Courts of Appeals. Segal and Spaeth
(1993, 225) found strong evidence for attitudinal-based decision making in the area of civil
rights among the Supreme Court justices, and there is little evidence to suggest that their
agenda decisions on this issue would be any different. Accordingly, I submit:

Hypothesis # 11 b. The agenda of the Supreme Court in civil

rights cases should not respond to changes in the agenda of

the Courts of Appeals.
That is, as specified in chapter 3 for civil liberties, Hypothesis #1 1 b exhibits the expectation
that the justices unlikely will be influenced by the incidence of civil rights cases below.
Thus, while appellants should react to the civil rights agenda in the Supreme Court, an
endogenous effect is not expected.
First Amendment

This classification is somewhat self-explanatory, at least for the most part. In

particular, as described by the Supreme Court Database (1995, 57): "First Amendment
encompasses the scope of this constitutional provision, but do note that not every case in the
First Amendment group directly involves the interpretation and application of a provision
of the First Amendment. "6 Figure 1.3 demonstrates that the Supreme Court decided a greater
percentage of first amendment cases than did the Courts of Appeals; in fact, the rate for the

Courts of Appeals has been generally steady over time, while the variance and possibly the

mean have changed in the Supreme Court with respect to first amendment cases. Indeed,

 

6To reiterate, Appendix A delineates the decision rules on coding these variables.

160

after civil rights, first amendment issues formed the bulk of civil liberties cases in the
Supreme Court (see Figure 1.6), although the same cannot be said for the appeals courts (see
Figure 1.7).

The theoretical expectations are similar for first amendment cases as for civil liberties
and civil rights. That is, a strong effect stemming fiom the high Court but not in the opposite
direction Should be anticipated, as the following hypotheses depict:

Hypothesis # 12a. The agenda of the Courts of Appeals in
first amendment cases should respond directly to changes in
the agenda of the Supreme Court in either a positive or
negative direction.
Hypothesis # 12b. The agenda of the Supreme Court in first
amendment cases should not respond directly to changes in
the agenda of the Courts of Appeals in either a positive or
negative direction.
Due Process

The category of due process includes cases in which this type of claim was raised in
a non-criminal setting. The Supreme Court Database codebook provides: "Due process is
limited to non-criminal guarantees and, like First Amendment issues, need not show ‘ SADP’
or ‘14ADP’ in its LAW field" (Id). As depicted in Figure 1.4, the mean and variance for the
Supreme Court’ 5 emphasis on due process has changed over the years, while these indicators
have remained moderately stable in the federal appellate courts.

From a theoretical perspective, once again there is little reason to postulate an
endogenous effect springing from the dockets of the Courts of Appeals in due process cases,
although the influence of the Supreme Court Should be apparent. The ensuing hypotheses

set forth these expectations:

161

Hypothesis # 13a. The agenda of the Courts of Appeals in

due process cases should respond directly to changes in the

agenda of the Supreme Court in either a positive or negative

direction.

Hypothesis # 13b. The agenda of the Supreme Court in due

process cases Should not respond directly to changes in the

agenda of the Courts Of Appeals in either a positive or

negative direction.
Privacy

As the Supreme Court provided in the landmark case of Griswold v. Connecticut, 3 81

US. 479, 485 (1965): "We have had many controversies over these penumbra] rights Of
‘privacy’ . . . These cases bear witness that the right of privacy which presses for recognition
here is a legitimate one. The present case, then, concerns a relationship lying within the zone
of privacy created by several fundamental constitutional guarantees."7 While never
enumerated in the Constitution, the issue of a right of privacy has had a powerful impact on
politics in the United States, particularly as concerns the abortion cases which were decided
pursuant to a constitutional right of privacy.8 Nevertheless, despite their emphasis in the
political arena and media, cases comprising this controversial issue rarely constitute a
notable part of the judicial agendas of the federal courts. Indeed, in only four of the years

in this study did the percentage of privacy cases in the Supreme Court exceed three percent,

while this occurred only once in the Courts of Appeals. Yet, the mean and variance of the

 

7Privacy cases fall within the purview of substantive due process. See Griswold v.
Connecticut, 381 US. 479 (1965); Roe v. Wade, 410 US. 113 (1973).

8See, e. g., Roe v. Wade, 410 US. 113, 153 (1973): "This right of privacy . . . is broad
enough to encompass a woman's decision whether or not to terminate her pregnancy."

162

time series in both courts seem to have increased since the time of the Roe v. Wade decision
(see Figure 1.5). Specifically, the Supreme Court Database defines three issues within the
privacy domain, including issues concerning privacy in general, abortion and contraception,
and the Freedom of Information Act.

Theoretically speaking, for all the reasons cited with respect to the civil liberties
agenda in general, the influence of the Supreme Court on litigants in the Courts of Appeals
should be apparent in privacy cases. Accordingly, I posit:

Hypothesis # 14a. The agenda of the Courts of Appeals in
privacy cases should respond directly to changes in the
agenda of the Supreme Court in either a positive or negative
direction.

Nevertheless, privacy cases are a bit different from the other issues within the civil
liberties category. That is, the nature of the privacy issue leads to a different theoretical
expectation from an endogenous perspective. First, the right of privacy was discussed and
debated in law reviews and other legal publications prior to the Supreme Court’s ruling in
Griswold v. Connecticut, and the likelihood is that lower courts were experimenting with this
issue as well during this time. See, e. g., Northwestern L. Rev. (1960); Beaney (1962).
Moreover, once the Supreme Court espoused its decision in Roe v. Wade, it generally left the
details of this issue to be bashed out in the lower courts, only to hear selective abortion cases
after sufficient percolation below. These are precisely the circumstances under which I
postulated that an endogenous effect deriving from the lower courts would be expected, such

that the justices Slowly but surely react to a changing emphasis in the legal community and,

by implication, the lower courts. The result is a Supreme Court agenda which reflects in part

163

the agenda in the lower courts with respect to this issue (see chapter 2, p. 50). Accordingly,
I submit the following hypothesis:

Hypothesis # 14b. The agenda of the Supreme Court in

privacy cases should respond directly to changes in the

agenda of the Courts of Appeals in either a positive or

negative direction.

That is, in privacy cases the expectation is that an influence should be seen in both directions,

as the agendas of each court motivate corresponding changes in the other over time.

DATA, METHODS, AND ANALYSIS

The data and methods utilized in this chapter essentially will be the same as provided
for in chapter 3. Indeed, Since I already have delineated the particulars of the procedures
implemented in this research earlier in this study, I will not repeat them here (see ch. 2, pp.
52-58; ch. 3, passim). Accordingly, in this chapter I will detail only those issues Specifically
related to the issues within civil liberties. For instance, with respect to the data, the only
difference in this chapter from the prior ones is that the entire civil liberties category is
factored into its respective components, such that I am able to analyze the relationships
between these courts for the particular issue areas of concern in this chapter. As in chapter
3, the time series applications to be utilized herein are Granger causality and vector
autoregression (VAR).

Of course, when employing time series, the data must be examined for stationarity.
Eyeballing the various data suggests that most of the series have nonstable means and/or

variances, and the more definitive augmented Dickey-Fuller (AD-F) tests confirm this

164

impression (see ch. 3 for the applicable equations for the AD-F tests). More specifically,
when the data remain in levels, in every series within the civil liberties category, the null
hypothesis of a unit root could not be rejected in all of the variations of the AD-F tests,
indicating nonstationarity in the data (see Table 4.1). In two of the series, including Supreme
Court first amendment and Courts of Appeals due process cases, the null of a unit root could
be rejected beyond the conventional 95% confidence level in two of the AD-F variations;
but, since every variation could not be rejected, I decided to difference all the data series to
reduce the risk of Spurious results. Indeed, once the data were differenced, AD-F tests
demonstrated that the integrated problems in the level data were eliminated, as the null
hypothesis of a unit root could be rejected in every instance beyond the 99% confidence level
(see Table 4.2). Examination utilizing the Box-Jenkins (1976) approach corroborated the
success of differencing the data. Accordingly, I now am able to proceed with the various
time-series analyses?

AS in the last chapter, Granger causality (Granger 1969) is the initial method I
incorporate to determine whether the agenda Of one of these courts influences changes in the
other. To reiterate, this method addresses pair-wise exogeneity such that it enables the
researcher to determine if the dependent variable is best explained solely by its own past
behavior or by lags of the independent variable. The Granger results are presented in Table
4.3. Interestingly, these tests suggest that the Supreme Court seems to have little effect on

the agenda of the Courts of Appeals, as the only Significant result in this direction occurred

 

9Chapter 3 provides additional details along the lines of ensuring stationarity.

165

at 7 lags in privacy cases. On the other hand, the Courts of Appeals did not Granger cause
the Supreme Court’s agenda for the any of the issues of civil rights, first amendment, or due
process, which was as expected. Nevertheless, there was a rather strong Granger effect
stemming from the Courts of Appeals to the Supreme Court with respect to privacy cases,
as significant results obtained for lags six through ten, lending preliminary support for
Hypothesis #146. The VAR analyses will shed further light on these results and
implications.

VAR is a useful approach to managing feedback in a system over time without
resorting to the potentially misleading identification assumptions endemic to structural
equations modeling of simultaneous variables, such as two-stage least squares (Sims 1980;
Freeman, Williams, and Lin 1989; see ch. 3, p. 84-85). Since VAR is a reduced form, it
enables the researcher to model endogeneity in the system, since none of the variables are
considered to be exogenous. AS a consequence, while fewer restrictions are imposed on the
data, interpretation of the resulting coefficients is meaningless due to the lack of exogenous
variables. Thus, instead of resortng to the usual means of traditional parameter estimation,
the VAR approach utilizes innovation accounting and variance decomposition techniques,
which allow causal inferences to be traced over time.lo

As for the requisite preliminaries of VAR analysis, I employed Sims’ (1980, 156, fn.
17) modified log-likelihood test, along with the Akaike Information Criterion (AIC) and

Schwartz Criterion (SBC), to ascertain the appropriate lag length for model specification.

 

10For details, see ch. 3 (supra, pp. 84-87).

166

As in chapter 3, the data will be evaluated in both levels and frrst differences. Based on these
procedures, the appropriate lag lengths for the respective VAR systems are as follows:

data in differences:

civil rights 2 lags
first amendment: 2 lags
due process: 2 lags
privacy: 2 lags
data in levels:
civil rights 2 lags
first amendment: 2 lags
due process: 3 lags
privacy: 2 lags

Accordingly, having concluded these preparatory procedures, the VAR models are suitably
specified and can be estimated. More particularly, for each issue area, I ran a VAR model
and then applied both innovation accounting and variance decomposition techniques.

The first issue estimated in a VAR model is civil rights. Pursuant to the innovation
accounting for the differenced data, the only effect on the agenda of each court for this issue
is each court’s own past history, as neither court produces significant effects in the other.
Indeed, Figures 4.1b and 4.1c display that the +/- two-standard error bars which represent
roughly 95 percent confidence intervals do not cross the zero bar, indicating no causal effect.
Nevertheless, when the data remain in levels, the results are both more profound and as
hypothesized. More particularly, this analysis demonstrates that the Supreme Court’ s agenda
trends for civil rights clearly influences the dockets of the Courts of Appeals for nearly the
entire time period (see Figure 4.2c); however, an endogenous effect deriving from the lower

courts is not apparent (see Figure 4.2b). The level data analysis also establishes that the past

167

history of each court is important as well, particularly in the early part of the series (see
Figure 4.2a and 4.2d).

Variance decomposition corroborates the innovation accounting, and once again the
results are very different depending upon whether differenced or level data are analyzed.
That is, for the differenced data, there is little effect over time when the forecast error
variance is decomposed. Figures 4.3a and b display the variance decomposition in graphical
form, while Table 4.4 reveals the particular percentages ensuing from this analysis. Yet,
estimation via the level data provides that the Supreme Court has a consequential effect on
the Courts of Appeals, as approximately 35 percent of the civil rights appeals in the lower
courts are determined by the agenda in the high Court (see Figure 4.4 and Table 4.5).

The analysis for first amendment cases establishes that there is a negligible effect
stemming from the Supreme Court, and this is true whether differenced or level data are
employed; moreover, there is no influence arising from the lower courts for this issue. In
fact, the only systematic effect concerning the agendas of these courts concerns each court’s
own past agenda. That is, the justices are most reliant on their own past cert. decisions,
while litigants look to their own prior appeals in the appeals courts (see Figure 4.5
[differences] and Figure 4.6 [levels]). The variance decomposition analysis produces similar
results, as there is little influence from one court to the other on first amendment issues (see
Figure 4.7 and Table 4.6 [differences] and Figure 4.8 and Table 4.7 [levels]).

The results were more consequential for the due process cases. Indeed, whether the
data were in differences or levels, the innovation accounting illustrated that the Supreme
Court influences due process appeals in the lower courts, at least for a few years; however,

168

no endogenous effect was found in the opposite direction (see Figure 4.9 [differences] and‘
Figure 4.10 [levels]). When variance decomposition is applied, the efiects from the Supreme
Court become very noticeable, as between one-quarter and two-fifths of the agenda of the
Courts of Appeals is caused by changes in the due process agenda in the Supreme Court.
Interestingly, the level-data analysis suggests that the justices may be influenced by lower
court appeals in this area more so than indicated by the innovation analysis (see Figure 4.1 1
and Table 4.8 [differences] and Figure 4.12 and Table 4.9 [levels]).

Finally, the VAR analyses for privacy cases are consistent with the Granger results.
In particular, while there is no significant effect originating from either court when
differenced data are utilized, there is no doubt about the rather strong effect in both directions
when level data are analyzed. That is, not only are litigants responsive to changes in the
Supreme Court’s privacy agenda, but the justices themselves react to the privacy trends
found in the lower courts, and this effect appears for many lags (see Figure 4.13 [differences]
and Figure 4.14 [levels]). Variance decomposition authenticates these results, at least for
level data, as the actors in both the Supreme Court and Courts of Appeals are sensitive to
changes in the privacy agenda of the other court (see Figure 4.15 and Table 4.10
[differences] and Figure 4.16 and Table 4.1 1 [levels]). These results are consistent with the

hypothesized expectations for the privacy issue.

DISCUSSION
The Granger and VAR analyses employed in this chapter provide additional support
for my theory that the agendas of the highest federal courts in the United States influence

169

each other. As featured in chapter 3, the Supreme Court was notably influential as concerned
the civil liberties agenda of the Courts of Appeals, although an effect in the opposite
direction was not established. When the civil liberties category is broken down into its
representative issues, similar results obtain, although there are meaningful differences as
well. One of the most important findings from this chapter is that the Courts of Appeals do
not influence any part of the Supreme Court’s civil liberties agenda, except when it comes
to privacy cases. That is, the Supreme Court influenced the agenda in most of the civil
liberties issues but was swayed in return only when privacy issues were at issue, an
unobvious but crucial finding.

Before making generalizations, however, I first need to discuss the specific civil
liberties issues which constituted the bases of this chapter. To begin, the civil rights agenda
of the Supreme Court proved to be somewhat influential on litigiousness in the Courts of
Appeals in a manner very similar to the effect found in civil liberties in chapter 3, at least as
far as the VAR analyses were concerned. This is not entirely surprising, since civil rights
formed the preponderance of the civil liberties category. Furthermore, there was no
endogenous effect which derived from the Courts of Appeals. Accordingly, Hypotheses
#1 In and #1 1b appear to be borne out by the analysis.

As for first amendment cases, however, none of the analyses here demonstrated an
influence in either direction. That is, the Supreme Court seems to ignore the first amendment
agenda of the Courts of Appeals, and vice versa. While Hypothesis #12b was verified, then,
the expected effect on the lower courts from above did not materialize, thus eschewing
Hypothesis #12a.

170

The results for due process cases generally acquitted my theory, as the VAR analyses
in particular demonstrated that the Supreme Court has a powerful effect on the dockets in the
Courts of Appeals but that a reverse effect did not materialize. In fact, causality in the
direction of the Courts of Appeals was fairly convincing, particularly as demonstrated by the
variance decomposition analyses. For these reasons, the evidence provides significant
support for Hypotheses #13a and #13b.

Furthermore, the analyses clearly promoted the theoretical expectations on privacy
cases as well, as causal effects were observed in both directions. Thus, the Supreme Court
remains on the sidelines for events to occur in the legal profession, at least to a certain extent,
when it comes to privacy cases, apparently allowing for sufficient percolation to occur in the
courts below before being constrained (or strategic, perhaps) to accept particular cases on
abortion or otherwise within the privacy classification. For privacy cases, at least, it appears
that the federal courts do resemble a fimctioning system, as opposed to detached institutional
entities acting independently of one another.

This chapter accordingly builds upon the results achieved in chapter 3. First, the
agenda of the lower federal courts generally comported with the trends established by the
agenda choices made by the Supreme Court. Nevertheless, the justices seem to follow
certain agenda trends in the Courts of Appeals as well, accepting specific types of cases
when impelled to do so. What is most interesting about this result is that this endogenous
influence was obtained for one particular issue (privacy) when no such consequence could
have been gleaned fiom the analysis if this category had been aggregated in the manner it
usually is in other judicial studies and, of course, the manner it was analyzed herein in

171

chapter 3. That is, when the issue of civil liberties en masse was analyzed in chapter 3, only
an effect from the Supreme Court was discernible, but this chapter demonstrates that this is
not the entire story. Thus, the negligible influence from the Courts of Appeals for the non-
privacy civil liberties cases overwhelmed the convincing effect stemming specifically from
privacy issues.

In a similar vein, while the Supreme Court influenced litigants concerning the overall
civil liberties agenda of the Courts of Appeals, the analyses in this chapter demonstrated that
this effect was most apparent for due process, privacy, and, to a lesser extent, civil rights
cases; nonetheless, the more particularized analysis illustrates that the Supreme Court’s
agenda on first amendment issues was not effectual on the agenda in the courts below.
Again, this is an instance where the aggregated issue captivated the behavior of the political
actors with respect to one of the issues’ respective elements.

At this stage, it is logical to reiterate the results for criminal procedure cases as well.
While this issue was separated from the overall civil liberties category in chapter 3, these
types of cases often have been included within the broader classification. Indeed, as reported
in chapter 3, the results for criminal procedure were very similar to those for civil liberties,
in that the Supreme Court had an influence on the lower courts’ dockets but that the reverse
effect did not materialize (see Figures 3.9-3.12 and Tables 3.8 and 3.9). If this issue were
aggregated within the overall civil liberties category, the results for the amassed issue likely
would have been even stronger. Nevertheless, severing this issue from the whole enabled

a distinct result on criminal procedure cases to be revealed.

172

Accordingly, the aggregated civil liberties analysis in chapter 3 might lead to the
implication that the Supreme Court influences the lower courts on first amendment issue,
while the high Court is not persuaded by trends in privacy cases in the appeals courts. The
results of this chapter demonstrate rather convincingly that such conclusions are
inappropriate and should not be made. As Robinson (1950) exhorted long ago, researchers
should be careful about extracting inferences regarding individual behavior (or in this case,
individual issues) from aggregate data." In particular, as the analysis here portrays, all civil
liberties issues are not treated equally in the federal courts, at least with respect to the agenda

effects between the Supreme Court and Courts of Appeals.

 

”But see Kramer (1983, 93), who contended: "[I]t is the aggregate time-series
evidence — rather than that based on individual-level . . . data - which is most likely to yield
valid inferences about the underlying individual-level behavioral effects we are trying to
measure" [emphasis in original].

173

TABLE 4.1
AUGMENTED DICKEY-FULLER TESTS FOR UNIT ROOTS (SERIES IN LEVELS)

' No Intercept Trend
[H— ___z._______#_______________________

Civil Rights . -2. 720

 

First . -3.818**
Supreme Court Amendment

 

Due Process . -2.874

 

Civil Rights

 

First
Amendment

 

Due Process

 

 

 

 

 

 

HO: unit root (data are not stationary) *ps .05
HA: no unit root (data are stationary) "pg .01

Note: MacKinnon (1990) critical values are used for rejection of null hypothesis of a unit root.

174

TABLE 4.2

AUGMENTED DICKEY-FULLER TESTS FOR UNIT ROOTS (SERIES TN FIRST DIFFERENCES)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Court Issue No Trend, Intercept Only Intercept and

___ __ No Intercept _ Trend .
Civil Rights 40.252" -lO.l44** 40.202“ .

First -6.597"‘* -6.525** -6.478"“'l

Supreme Court Amendment
Due Process -9.832** -9.852** -9.766** '
Privacy -8.783** -8.695** -8.6.3** ]
FT===== ———— — ~—— - —-

Civil Rights -6.662.*" -6.895"‘* -6.813" ’
First 5.716” -5.682** -5.6l3** l
Courts of Amendment {

Appeals Due Process -8.86l** -8.814” -8.752**

Privacy _ -6.481"”'I _ -6.424** _ -6.355*"‘

 

Ho: unit root (data are not stationary)
HA: no unit root (data are stationary)

 

’ps.05
*‘ps.01

 

Note: MacKinnon (1990) critical values are used for rejection of null hypothesis of a unit root.

175

   
   
   
   
  
  

Direction
of

  

Supreme
Court
Granger

Causes

Supreme
Court

Granger

 

Causes

 

L__

Courts of

Issue

Civil
Rights

First

Amendment

 

 

Ho no Graner causality W
HA: Granger causality

*ps.05
**ps.01

GRANGER CAUSALITY RESULTS

Lag p-
Causality Length valu

TABLE 4.3

_'T ___ ___ _

e l

 

 

 

Appeals

 

Granger

 

Cause

 

' Supreme

 

 

\OWQQM-AWN

 

.—-r
O

 

 

 

 

 

 

 

 

 

 

 

176

   
  
   
   
   
   
 
    
 
   
     
 
    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   
    
    

TABLE 4.3 (CONT’D)

 

 

 

 

 

 

 

 

 

 

 

 

l Direction ' Direction
’ of . Issue Lag p- of .
Ca_lrty Le_gth__ _L Causality
2
i Supreme 3
i Court 4
i Granger Due 5
Causes Process 6
Courts of 7
Appeals 8
9
I: r 10::
2 .410
Supreme 3 .334 . Courts of
Court 4 .409 i Appeals
Granger Privacy 5 .346 l Granger
Causes 6 .192 1 Cause
Courts of 7 .060“ Supreme
I Appeals 8 .200 I Court
I 9 .103 “
[______ ___ ___ 10 285

Ho: no Ggraner causality

 

 

H A: Granger causality

*ps.05
**ps.01

 

 

 

 

 

 

 

 

 

 

177

 

Privacy

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{gomua‘m&wr~a;
i .

 

TABLE 4.4
VARIANCE DECOMPOSITION OF CIVIL RIGHTS CASES (DIFFERENCED DATA)

r

Variance Decomposition of Supreme Court

 
    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SUPreme Court ,, f _ M __ _ fa z _ __1‘
1 100.00 mm I
2 99.72 0.27 I
3 98.50 1.50 1
fl 4 97.84 2.16 L
5 97.90 2.10 '
6 97.79 2.21 l
7 97.67 233 i
8 97.67 233 l
9 97.66 2.34 i
f_ _ -10- __ _ _ 97.65 f_ _ __ f i
L _ _ _ ___ _ Variance Decompositiono_f Appeals _ _ _ ———i
1 0.17 99.83 i
2 0-72 99.28 !
3 1.61 98.39 i
4 1.88 93.12 ‘
5 1.88 98.12
6 1.98 98.02 I
7 2.05 97.95
3 2-05 97.95 .
9 2.06 97.94
10 2.07 1 97.93

 

 

Ordering: Supreme Court/Courts of Apspeal

178

TABLE 4.5
VARIANCE DECOMPOSITION OF CIVIL RIGHTS CASES (LEVEL DATA)

Variance Decomposition of Supreme Court

 

 

 

Period Supreme Court Courts of Appeals l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 100.00 0.00 i
94.78 5.22 I
95.44 4.56 .
94.35 5.65 l
94.22 5.78 E
93.90 6.10 1
93.66 6.34 I
93.48 6.52 l
93.30 6.70 l
93.17 _ ...____.___ ._ _ _. ___. __L. l
Variance Decomposition of Courts of Appeals 1
3.60 96.40 ‘
4-35 95.65 1
14.61 85.39 |
18.42 81.58
23.89 76.”
27.52 72.48 J
30.64 69.36 ll
33.07 66.93
34.98 65.02

 

 

 

 

 

 

TABLE 4.6
VARIANCE DECOMPOSITION OF FIRST AMENDMENT CASES (DIFFERENCED DATA)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

2 95.77 4.23
3 95.71 4.29
4 94.85 5.15
5 94.79 5.21
6 94.78 5.22
7 94.77 5.23
8 94.77 5.23
9 94.77 5.23
Variance Decompositio of Courts of Appeals

 
 

 

 

 

10 49.77 5.23 _ 1
l
i

 

 

 

 

 

 

 

 

 

1 0.97 99.03 '
l 2 1.33 98.67 I
u 3 3.44 96.56

4 3.70 96.30

5 3.73 96.27 5

6 376 96.24 1

7 3.76 96.24 I

8 3.76 96.24

9 3.76 96.24

10 3.76 96.24

 

 

 

1 Order!” SW6 Wat" Appeals i

180

TABLE 4.7
VARIANCE DECOMPOSITION OF FIRST AMENDMENT CASES (LEVEL DATA)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

II Variance Decomposition of Supreme Court I
Period Supreme Court Courts of Appeals I
1 100.00 0.00 I
2 92.93 7.07 JI
3 90.77 9.23 II
4 90.21 9.79
5 90.11 9.89
6 90.09 9.91 II
7 90.09 9.91 II
8 90.09 9.91 II
9 90.09 9.91 II
10 ,5: 90.09 = 9.91 I
Variance Decomposition of Courts of Appeals I
.E Period Supreme Court J Courts of Appeals J
l 1.79 98.21 I
2 4.54 95.46
3 4.50 95.50
4 4.40 95.60
5 4.38 95.62
6 4.39 95.61
7 4.39 95.61
8 4.39 95.61
9 4.39 95.61
10 4.39 95.61 LII
Ordering: Supreme Court/Courts of Appeals 1

 

 

181

TABLE 4.8
VARIANCE DECOMPOSITION OF DUE PROCESS CASES (DIFFERENCED DATA)

_Variance Decomposrtron of Supreme Court——

c___g____uns A__p_p..:__..f

 

 

 

 

 

 

 

 

100. 00 0. 00
IL 2 99.87 0.13
3 96.44 3.56 I
4 90.28 9.72
5 86.75 13.25
6 85.97 14.03
7 85.96 14.04
8 85.96 14.04
9 85.96 14.04

 

 

 

 

10 85.96 _7 _ __ __ _ 14. 04
Variance DecompOsition of Cour-_ts of Appeals

_Cou__r_t_s_orAppeals_

 

 

 

 

 

 

 

 

 

 

 

 

26. 78 73. 22

2 22.46 77.54

3 21.80 78.20

4 23.10 76.90

5 23.55 76.45
6 23.54 76.46

I 7 23.57 76.43
8 23.58 76.42

I 9 23.58 76.42
I 1023.58 76.42

I Ordering: Supreme Conn/Courts of Appeals_ I

TABLE 4.9
VARIANCE DECOMPOSITION OF DUE PROCESS CASES (LEVEL DATA)

I

 

 

 

 

 

 

 

 

 

 

 

 

1
I _Period __Sup_reme Court I
100. 00 0.00 I

2 99.03 0.97 I
I 3 97.77 2.23 I
4 93.81 6.19 I

5 93.89 6.11 I

I 6 93.19 6.81 I
7 93.27 6.73 1
I“) 8 93.08 6.92 I
I 9 93.08 6.92
‘‘‘‘‘ 92. 97 - 7 03 I
I

 

 

 

 

 

 

 

 

 

.
‘ 1 38.33 61.67 I
I 2 39.09 60.91 I
3 38.43 61.57
4 39.71 60.29 I
I 5 39.69 60.31 I
I 6 39.81 60.19
7 39.85 60.15
8 39.89 60.11 I
I 9 39.89 60.11 I
__10 39 91 60 89 I

 

mering: Supreme_ Court/C_ourts of Appeals _

183

 

TABLE 4.10
VARIANCE DECOMPOSITION OF PRIVACY CASES (DIFFERENCED DATA)

 

 

 

 

 

 

 

 

 

 

I

I 1 100.00 0.00 I

I 2 99.96 0.04 I
3 99.94 0.06 I
4 99.95 0.05

I 5 99.94 0.06
6 99.94 0.06 I
7 99.94 0.06 I
8 99.94 0.06 I
9 99.94 0.06 I
10 99.4 0 06 I

 

 

 

__Variance “Decomposition of Courts of Appeals I
I
—I

 

 

 

 

 

 

 

 

 

 

 

 

0.37 99.63 I

I 2 0.46 99.54

I 3 1.45 98.55 I
I 4 4.38 95.62

I 5 5.48 94.52

I 6 5.48 94.52 I

7 5.63 94.37 I

I 8 5.69 94.31 I

9 5.69 94.31 1

I1”: 10 ___5.71 _____ __ 9429 _ __ _I

I

184

TABLE 4.1 l
VARIANCE DECOMPOSITION OF PRIVACY CASES (LEVEL DATA)

Variance Decomposition of Supreme Court I

 

      
     
     

 

 

 

 

 

 

 

 

 

 

 

 

1 100.00
ll 2 93.16 6.84
3 88.27 1 1.73
4 87.77 12.23
5 86.76 13.24
6 85.96 14.04
II 7 85.45 14.55
II 8 85.09 14.91
ll 9 84.81 15.19
10 .860 15.40

 

 

   
  
 
 

Variance Decomposition of Courtsof Appeals

Courts °A____f_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 8.64 91.36
2 19 03 80.97
3 33 66 66.34
II 4 35 30 64.70
II 5 37.82 62 18
6 39 48 60 52
7 40.64 59 36
8 41 41 58.59
9 42.02 57 98
___ “7 42.47 __ _. __ ___553_

Ordering: Supreme CouCourts of Appeals

 

Civil Rights Cases (Differences)
Response to One 8.0. Innovations t 2 S.E.
(a)

0))
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals

 

 

44

2. A

 

 

 

 

 

 

 

 

III

 

 

 

 

 

 

 

 

(c) (d)
ResponseofCourtsoprpeslstoSupreme Court ResponseofCorutsoprpeslstoCourtsoprpeels

FIGURE 4.1
INNOVATION ACCOUNTING or CIVIL RIGHTS CASES (DIFFERENCES)

186

Civil Right Cases (Levels)
Response to One 8.0. Innovations :1: 2 S.E.
(a) 0))
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals
6 e

 

 

 

 

 

 

 

 

 

 

-2

 

 

3.0

 

 

2.54
2.0..
1.54
1.04
0.5I

0.0

 

 

 

 

 

 

 

 

 

(c) (d)
Response of Court of Appeals to Supreme Court Response of Court of Appeals to Court of Appeals

FIGURE 4.2
INNOVATION ACCOUNTING or CIVIL RIGHTS CASES (LEVELs)

187

Civil Rights Cases (Differences)

(a)

Variance Decomposition of Supreme Court
100.

 

 

 

80I

604

 

— Supreme Court
---- Courts oprpeals

 

 

 

40.

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

---------------------------------------------------

80.

60‘

 

— Supreme Court
---- Courts of Appeals

 

 

404

 

20.

 

 

 

 

1 2 3 4 S 6 7 8 s 10
FIGURE 4.3
VARIANCE DECOMPOSITION or CIVIL RIGHTS CASES (DIFFERENCES)

188

Civil Rights Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100

 

 

804

60.

 

—- Supreme Court
---- Court of Appeals

 

40.

 

 

20.

 

 

................................................
v"
o

 

(b)
Variance Decomposition of Courts of Appeals

100

 

804

60.

 

— Supreme Court
---- Courts of Appeals

 

404

 

 

20.

 

 

 

1 2 8 4 ‘5 e 7 8 5 10
. FIGURE 4.4
VARIANCE DECOMPOSITION OF CIVIL RIGHTS CASES (LEVELs)

189

First Amendment Cases (Differences)
Response to One 8.0. Innovations t 2 S.E.
(a) (b)
ResponseofSupremeCourttcSupremeCourt ResponseofSupremeCourttoCourtsoprpeals
4

 

 

   

 

 

 

 

 

 

10

L
I
O
8
a.
N
u.
&
0
O
N
.1
0

 

 

1.5-I
1 .04
0.5.I
0.04

-0.5.

 

 

 

 

 

 

 

(9) ~
ResponseofCourtoprpeslstoSupreme Court ResponssofCourtoprpeslstoCourtoprpesls

FIGURE 4.5
INNOVATION ACCOUNTING or FIRST AMENDMENT CASES (DIFFERENCES)

190

First Amendment Cases (Levels)
Response to One 8.0. Innovations t 2 S.E.

(a) a»
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts of Appeals

4 4

 

 

 

 

 

 

 

 

 

 

 

1.5 1.5

 

 

1.0. I 1.04

 

 

 

 

 

 

 

 

 

osI " N.‘ as
‘\
‘s
\‘~
/\ “~
~~‘
Q.--
9.0 -.T=IIIOOO 9.9
"""""
P---.‘ """
s‘ "v
‘ .......
as 435
1 2 3 4 5 0 7 8 9 10

(c) (d)
Response ofCourt oprpeals to Supreme Court Response ofCourt otAppeals to Court otAppeals

FIGURE 4.6
INNOVATION ACCOUNTING or3 FIRST AMENDMENT CASES (LEVELS)

191

First Amendment Cases (Differences)

(a)

Variance Decomposition of Supreme Court

 

100._\

aoI

60.

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

20.

...............................................
O
’
C
O

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100.: ___________________________________________________

 

80.

60.

 

— Supreme Court
---- Courts of Appeals

 

 

 

40.

20.

0 I I V V V U V V

12345678910

 

 

 

 

FIGURE 4.7

VARIANCE DECOMPOSITION or FIRST AMENDMENT CASES (DIFFERENCES)

192

First Amendment Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100

 

 

80.

60.

 

—— Supreme Court
---- Courts of Appeals

 

 

 

40.

20.

-------------------------------------------
v"
I
f
a
I
I

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100

 

1'.
n
-------------------------------------------------

80I

604

 

— Supreme Court
---- Court of Appeals

 

40.

 

20.

 

 

 

 

12345678910

FIGURE 4.8
VARIANCE DECOMPOSITION or FIRST AMENDMENT CASES (LEVELs)

193

Due Process Cases (Differences)
Response to One 3.0. Innovations :I: 2 S.E.
(a) (b)
ResponseofSupremeCourttoSupremeCourt ResponseofSumeourttoCourtsoprpeals

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(c) «0
Response ofCourt: oprpeebtoSupremeCourt ResponseofcouruoprpeabtoCourtsoprpeals

FIGURE 4.9
INNOVATION ACCOUNTING OF DUE PROCESS CASES (DIFFERENCES)

194

Due Process Cases (Levels)
Response to One 8.0. Innovations :l: 2 S.E.
(a) (b)
Response of Supreme Court to Supmme Court Response of Supreme Court to Courts of Appeals
3 s

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(e) (6)
Response of Courts of Appeals to Supreme Court Response of Courts of Appeals to Courts of Appeals

FIGURE 4.10
INNOVATION ACCOUNTING OF DUE PROCESS CASES (LEVELs)

195

 

Due Process Cases (Differences)

(a)

Variance Decomposition of Supreme Court

100‘-\

80.

 

60A

 

— Supreme Court
404 ---- Courts oprpeals

 

 

 

20.

 

 

 

(b)
Variance Decomposition of Courts of Appeals

100

 

80.

-------
0“- --------------------------------------
‘
'
D

60A

 

. — Supreme Court
40. ---- Courts oprpeals

 

 

 

V.
204

 

 

 

1 i 5 A 5 6 i é 5 10
FIGURE 4.11
VARIANCE DECOMPOSITION OF DUE PROCESS CASES (DIFFERENCES)

196

 

   

Due Proces Cases (Levels)

(a)

Variance Decomposition of Supreme Court
100.
R

80(

 

60.

 

— Supreme Court
---- Courts oprpeals

 

404

 

 

20.

 

 

-----------------------------------
‘
'
’
’
-C’
........

 

(b)

Variance Decomposition of Courts of Appeals
100

 

80.

60. ....................................................

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

 

204

 

 

 

1 i 5 i 5 é % 6 s3 10
FIGURE 4.12
VARIANCE DECOMPOSITION 0F DUE PROCESS CASES (LEVELs)

197

Privacy Cases (Differences)
Response to One 8.0. Innovations :l: 2 S.E.

(a) (b) *
Response of Supreme Court to Supreme Court Response of Supreme Court to Courts otAppesIs

 

1K

 

1.0‘

0.5‘

 

0.0 ’ ___________ em I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(0) (d)
Response otCourts oprpeaIsto Supreme Court Response ofCourts otAppealstoCourts oprpeals

FIGURE 4.13
INNOVATION ACCOUNIING OF PRIVACY CASES (DIFFERENCES)

198

Privacy Cases (Levels)
Response to One 8.0. Innovations :l: 2 S.E.

 

 

Response of Supremefigourt to Supreme Court Response of Supreme (gun to Courts of Appeals
12 12
1.0 1.0.
as. as.
0.5. 0.0.

0A. 0A.

02. 01.

 

 

0D

 

 

 

 

 

 

$1 I V I I I 0.2

 

 

 

 

 

 

 

   

 

 

 

 

 

 

as 0.3
as. 0.0“
\
0.4 ,"x 0.4i
02.
0.0

$2 I V ‘V I V I I I £2 V V
12345070010 12345070010
(6) (d)

ResponseofCourts oprpealstoSupreme Court Response ofCourts oprpeals to Courts oprpeals

FIGURE 4.14
INNOVATION ACCOUNTING OF PRIVACY CASES (LEVELs)

199

Privacy Cases (Differences)

(a)

Variance Decomposition of Supreme Court
100

 

80.

60.

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

20.

 

 

 

(b)

Variance Decomposition of Courts of Appeals
100 _

 

80.

60.

 

— Supreme Court
---- Courts of Appeals

 

40.

 

 

20.

 

 

 

 

1 2 3 4 5 s i s s 10
FIGURE 4.15
VARIANCE DECOMPOSITION OF PRIVACY RIGHTS CASES (DIFFERENCES)

200

Privacy Cases (Levels)

(a)

Variance Decomposition of Supreme Court

 

100

80.

60.

40.

20.

 

 

 

Variance Decomposition of Courts of Appeals

 

(b)

 

100

80‘

60.

40.

20.

 

 

 

 

T

ééiéé‘résro

FIGURE 4.16

 

 

-— Supreme Court
---- Courts of Appeals

 

 

 

— Supreme Court
---- Courts oprpeals

 

VARIANCE DECOMPOSITION OF PRIVACY CASES (LEVELs)

201

 

 

CHAPTER 5

MULTIVARIATE ANALYSES OF THE INFLUENCES ON THE AGENDAS
IN THE SUPREME COURT AND COURTS OF APPEALS

The chapters preceding this one revealed that the federal courts exist within a system,
at least with respect to the agendas of the Supreme Court and Courts of Appeals. In
particular, each court influences changes in the agenda of the other, albeit to varying degrees
and conditional upon the issues involved. The statistical manner in which these causal
relationships were illustrated was vector autoregression (VAR). While VAR is rather well-
suited to establishing such causal connections over time, one Of the drawbacks to this
approach is that it utilizes many degrees Of freedom (Johnston and DiNardo 1996).
Accordingly, to ensure a large enough N was available, only bivariate relationships could be
examined; that is, in the prior analyses I investigated whether the agenda of the Supreme
Court influences changes in the agenda of the Courts of Appeals for one particular issue, and
Vice versa. Nevertheless, I did not test whether other potential influences enter into the
equation of agenda setting in the federal courts. In other words, are the effects of agenda the
only influence on the agendas Of these courts, or do other factors inspire changes in judicial

agendas over time? Answering this question is the primary purpose of this chapter.

202

INFLUENCES ON AGENDAS IN THE FEDERAL COURTS

As I have argued in this study, the agendas of the Supreme Court and the Courts Of
Appeals stem from an aggregation of rational decisions made by the various political actors
involved; yet, this decision is influenced by institutional features surrounding these players
in the judiciary. At the Supreme Court, each individual justice has policy preferences which
he or she would like to see implemented. Seeking to advance those preferences, the justices
endeavor to find particular cases on cert. which serve as the best possible vehicles with
respect to their preferences. Thus, each justice favors accepting certain types of cases on
cert. rather than others. Nevertheless, institutional characteristics, such as the Supreme
Court’s internal rule of four or the agenda of the courts below, act to modify these individual
preferences as they are represented in the overall agenda of the Supreme Court.

While the process is somewhat different at the Courts of Appeals, the idea is the
same. Rational litigants, their lawyers, and interest groups choose whether or not to appeal
specific cases to the circuit courts, a decision which is based almost entirely on probability
Of success. Nevertheless, this judgment is tempered by various institutional aspects both
within and without the judicial system. For instance, internal constraints include the type of
case tO be appealed vis-a-vis the ideological state of the federal courts; external forces
involve the costs of litigation, the political environment, and various social forces (Hurwitz
and Sheehan 1998).

It is the aggregation of these decisions by justices in the Supreme Court and litigants
and their representatives in the Courts of Appeals which comprise the agendas of these
respective courts. Thus, even though individuals make particular decisions which influence

203

the judicial agendas, certain patterns of these aggregate decisions may be revealed as they
are affected by broader factors stemming from the judicial and political enviromnents. Thus,
while individual decisions are fundamental with respect to judicial agenda setting, the
composite of those individual decisions is the final factor involved. Accordingly, there is
a need to look beyond the agendas of these courts as independent variables and explore the
broader determinants of the agendas within the federal courts. In this chapter, then, I analyze
such explanations for the following salient issues on the agendas Of these federal courts: civil
liberties, economic cases, and criminal procedure cases.I

In the prior chapters, I submitted various hypotheses regarding the influence that each
Of these federal courts should have with respect to the agenda of the other for the relevant
issues areas.2 Since the statistical technique I will utilize in this chapter is two-stage least
squares (ZSLS), I am able to test causality in a particular direction.3 Thus, instead Of
reiterating the hypotheses fi'om chapter 3, I will posit new hypotheses for the multivariate
analyses. These current hypotheses essentially are the same as those presented previously.

One distinction, however, is that the hypotheses presented here posit the direction of

 

1These categories of issues are defined in Appendix A. For purposes of this chapter,
the civil liberties issue embodies civil rights, first amendment, due process, and privacy
cases. The economic classification includes cases coded as economic, federal tax, and union
and labor cases. The criminal procedure issue area has no sub-issues, as it consists entirely
Of criminal procedure cases.

2F or civil liberties, see Hypotheses # 3a and # 3b. For economic, union and labor,
and federal tax cases, see Hypotheses # 4a and # 4b, # 6a and # 6b, and # 100 and # 10b,
respectively. For criminal procedure cases, see Hypotheses # 5a and # 5 b.

3The need for utilizing ZSLS will be presented subsequently in the Data and Methods
section.

204

causality as well as simply that a causal effect should be expected. That iS, the expected
outcomes will be Similar to that theorized previously (either one court will or will not be
projected to affect the agenda of the other court), but when causality is expected, a direction
will be postulated as well. Another difference is that I will incorporate the expectations for
lagged endogenous variables.

Endogenous Influences on Agenda

The first influence on the agendas of the federal courts to be tested is Simply the past
history Of agenda of each of these courts. That is, depending upon the issue, the Supreme
Court’ S agenda should be influenced to a certain extent by its own agenda and potentially the
agenda of the Courts of Appeals for that issue. Similarly, the agenda in the appeals courts
should be impelled by its own past agenda and the agenda of the Supreme Court.
Essentially, this is what I have asserted and explored in the antecedent chapters through the
bivariate VAR analyses. In this chapter, these variables on the courts’ agendas will be
included among many others in the several multivariate analyses.

For civil liberties, I contended in chapter 3 that effects Should be evidenced in only
in one direction, as the agenda of the Supreme Court should influence the agenda in the
lower courts, but that there should be no expected causality from below. That expectation
is reiterated here. With respect to the Supreme Court’s anticipated influence on the Courts
of Appeals, the direction of causality Should be positive; that is, as the proportion Of civil
liberties cases on the dockets of the Supreme Court increases over time, the consequent
proportion of civil liberties cases in the Courts of Appeals should expand as well.
Additionally, for both courts, the lagged endogenous variables should be influential in setting

205

the civil liberties agenda Of these courts. In other words, the past history of the Supreme
Court’s civil liberties agenda should positively affect the contemporaneous civil liberties
agenda of the Supreme Court; Similarly, the past civil liberties agenda in the Courts Of
Appeals Should be a positive factor in their current civil liberties agenda. These theoretical
expectations are depicted in the following hypotheses:

Hypothesis # 15a. The agenda of the Courts of Appeals in
civil liberties cases should respond directly to changes in the
civil liberties agenda of the Supreme Court in a positive
direction.

Hypothesis # 15b. The agenda Of the Supreme Court in civil
liberties cases should not respond to changes in the civil
liberties agenda of the Courts of Appeals.

Hypothesis # 15c. The agenda of the Supreme Court in civil
liberties cases should respond directly to the past civil
liberties agenda of the Supreme Court a positive direction.

Hypothesis # 15d. The agenda of the Courts of Appeals in

civil liberties cases Should respond directly to the past civil

liberties agenda of the Courts Of Appeals in a positive

direction.
Similarly, as was expected in chapter 3, for criminal procedure cases the agenda of the
Supreme Court Should positively affect the agenda of the Courts Of Appeals, while no
reverse influence Should be Observed. Moreover, the lagged endogenous variables Should

positively affect the contemporaneous agenda of each court:

Hypothesis # 16a. The agenda of the Courts of Appeals in
criminal procedure cases should respond directly to changes
in the criminal procedure agenda of the Supreme Court in a
positive direction.

206

Hypothesis # 16b. The agenda of the Supreme Court in
criminal procedure cases Should not respond to changes in the
criminal procedure agenda of the Courts of Appeals.
Hypothesis # 16c. The agenda of the Supreme Court in
criminal procedure cases Should respond directly to the past
criminal procedure agenda of the Supreme Court in a positive
direction.

Hypothesis # 16d. The agenda of the Courts of Appeals in
criminal procedure cases Should respond directly to the past
civil liberties agenda of the Courts of Appeals in a positive
direction.

On the other hand, the effect on the agenda of economic, and federal tax, and union
and labor issues (collectively referred to as the "economic" agenda) should be a bit different
than the influence of civil liberties and criminal procedure cases, as economic issues are the
most opportune issues for feedback to be expected between these courts. More Specifically,
in chapter 3, when each of the issues comprising the economic category was explored
separately, I hypothesized that causality Should be expected in both directions, as the agenda
of each court should influence the agenda of the other. In this chapter, where these issues
are examined en masse, there is no reason to believe that the expectations for the separate
issues would be any different from the expectations stemming from the aggregate measure.
Moreover, the direction of causality for economic issues should be positive, as an increasing
economic agenda in each court should lead to an expanding economic agenda in the other
court. As well, the past history of each court should be influential; but, Since I have argued

that the Supreme Court desires to avoid economic cases, the past history of the Supreme

Court should be in a negative direction. That is, after increasing its economic agenda, the

207

justices will attempt to avoid further economic cases, thus leading to a negative influence.
As a consequence, I propound:

Hypothesis # 17a. The agenda of the Courts Of Appeals in

economic cases should respond directly to changes in the

economic agenda Of the Supreme Court in a positive

direction.

Hypothesis # 17b. The agenda of the Supreme Court in

economic cases should respond directly to changes in the

economic agenda Of the Courts of Appeals in a positive

direction.

Hypothesis # 1 7c. The agenda of the Supreme Court in

economic cases Should respond directly to the past economic

agenda of the Supreme Court in a negative direction.

Hypothesis # 1 7d. The agenda of the Courts of Appeals in

economic cases should respond directly to the past economic

agenda of the Courts of Appeals in a positive direction.
Judicial Influences on Agenda

In this chapter I examine the influences on agendas in the federal courts.

Accordingly, it is axiomatic that judicial variables beyond agendas be tested to calculate their
effect on these courts’ agendas. Paramount among these potential motivations is the
dynamic liberalism of the Supreme Court. From a theoretical perspective, as the decisional
outcomes of the Supreme Court become more liberal, the justices should be more likely to
accept cases over time which would be sympathetic to promoting their policy preferences.
Those cases which are most sympathetic to a liberal judicial agenda include civil liberties
and criminal procedure issues. Indeed, as Pacelle (1991) argued, after the constitutional
crises of the 19308, the Supreme Court moved away from an agenda dominated by economic

cases and Shifted toward one based on civil liberties cases, and this trend coincided with an

208

increasingly liberal Supreme Court. Thus, augmenting decisional trends of the Supreme
Court’s liberalism should lead to a greater proportion of civil liberties and criminal procedure
cases on the docket of the Supreme Court. By the same reasoning, litigants in the federal
courts below, recognizing a liberal trend in the high Court, would appeal a greater percentage
of civil liberties and criminal procedure cases, which would be reflected in the agenda of the
Courts of Appeals. These expectations are grounded in the following hypotheses:

Hypothesis # 18a. The agenda of the Courts of Appeals in

civil liberties cases should respond directly to changes in the

liberalism of the Supreme Court in a positive direction.

Hypothesis # 18b. The agenda of the Supreme Court in civil

liberties cases Should respond directly to changes in the

liberalism of the Supreme Court in a positive direction.

Hypothesis # 19a. The agenda of the Courts of Appeals in

criminal procedure cases should respond directly to changes

in the liberalism of the Supreme Court in a positive direction.

Hypothesis # 19b. The agenda of the Supreme Court in

criminal procedure cases should respond directly to changes

in the liberalism of the Supreme Court in a positive direction.

To the contrary, increasing liberalism at the Supreme Comt would make for an
agenda consisting of fewer economic cases (Pacelle 1991), while such a liberal trend should
deter litigants from appealing economic cases as well. That is, greater liberalism in the
judiciary is not conducive to economic cases, at least not from a rational perspective. Thus,
if any effect is observed which derives from the influence of the liberalism of the Supreme
Court, it Should be negative. These predictions are incorporated in the following hypotheses:

Hypothesis # 20a. The agenda of the Courts of Appeals in
economic cases Should respond directly to changes in the

liberalism of the Supreme Court in a negative direction.

209

Hypothesis # 20b. The agenda of the Supreme Court in
economic cases should respond directly to changes in the
liberalism of the Supreme Court in a negative direction.

Another potential judicial influence stems from the liberalism of the Courts of
Appeals. While I have argued throughout this study that the dynamic agenda of the Courts
of Appeals influences the Supreme Court to accept some types of cases, there is no rationale
for conceiving that the decisional trends of the lower courts have any influence on the high
Court in any manner, including the justices’ decisions with regard to setting the Supreme
Court’s agenda.4 Nevertheless, it seems rather logical to presume that decisional outcomes
in the Courts of Appeals could lead to rational litigants appealing certain types of cases over
others in these courts. More particularly, as the liberalism of the Courts of Appeals
increases, a greater proportion of civil liberties and criminal procedure cases Should be
appealed to the circuit courts, as those issues are amenable to a liberal judiciary. To the
contrary, increasing liberalism among the Courts of Appeals Should inspire fewer economic-
based cases from being appealed. Accordingly, I propose the following hypotheses:

Hypothesis # 21. The agenda of the Courts of Appeals in civil
liberties cases should respond directly to changes in the
liberalism of the Courts of Appeals in a positive direction.

Hypothesis # 22. The agenda of the Courts of Appeals in
criminal procedure cases should respond directly to changes

in the liberalism of the Courts of Appeals in a positive
direction.

 

4For this theoretical reason, and for the statistical reasons stated subsequently in the
Data and Methods section, no hypotheses are submitted for any potential effect on the
Supreme Court’s agenda which might originate from the liberalism of the Courts of Appeals,
nor will any such tests be accomplished.

210

Hypothesis # 23. The agenda of the Courts of Appeals in
economic cases Should respond directly to changes in the
liberalism of the Courts of Appeals in a negative direction.

Liberalism of the respective federal courts may not be the sole judicial influence on
agendas. Indeed, the potential effect of particular Chief Justices of the Supreme Court can
be explored in time series analysis. Since the Chief Justice is first among equals (Baum
1 998), most judicial scholarship on the effect of the Chief s influence has encompassed some
aspect of his formal duties, such as opinion assignment (Slotnick 1979; Brenner and Spaeth
1988; Maltzman and Wahlbeck 1996a), or has addressed the voting patterns or
considerations on a particular court (Spaeth 1963; Brenner, Hagle, and Spaeth 1989; Hagle
and Spaeth 1992; 1993; Maltzrnan and Wahlbeck 1996b; Gerber and Park 1997). Yet, what
I seek to accomplish is a bit different from this research, as I examine whether the Chief
Justice or his particular Court has an effect on the overall process of building agendas in the
federal courts. Moreover, Since the data analyzed in this study persist beyond the tenures of
the particular Chief Justices at issue, multivariate estimation allows for a kind of
multiplicative analysis of the Chiefs influence on agendas.

The potential influences of Chief Justices Warren and Burger will be analyzed in this

chapter. Since Chief Justice Vinson’s term ended before the available data for this chapter,

any possible effect from his tenure cannot be examined.’ And, while Chief Justice

 

5While the Supreme Court Database includes data from the Vinson Court and the
Courts of Appeals Database commences well before this time, other variables in this study,
including political liberalism and public mood, do not provide data points early enough in
time to include Chief Justice Vinson’s influence on agendas. These variables will be
discussed subsequently in this section and in the Data and Methods section.

211

Rehnquist’s term commenced during the time for which some data is available, the Short
time period thereof, from 1986 through 1988, is insufficient for time series analysis, once
appropriate data transformations, including differencing and lags, are incorporated.
Notwithstanding, his term will be coded and used to set up part of the analysis (see the Data
and Methods section).

Nevertheless, the potential influence of the Warren and Burger Courts can, and will,
be explored in this chapter. In particular, the liberal Warren Court is known as changing the
direction of the Supreme Court by protecting individual rights and liberties and moving
toward a decidedly liberal position. The Burger Court reversed this trend set in the Warren
Court by supporting civil liberties claims at a much lower rate; however, the Burger Court’s
retreat from the stances of the Warren Court were even more pronounced in the area Of
criminal procedure (Adamany 1991).

This brief historical perspective of the Warren and Burger Courts provides sufficient
information such that testable hypotheses on their likely effect on agendas can be asserted.
More particularly, the Warren Court Should positively influence the civil liberties and
criminal procedure agendas in both the Supreme Court and the Courts of Appeals, while a
negative effect should emerge with respect to economic cases, as stated:

Hypothesis # 24a. The agenda of the Courts of Appeals in
civil liberties cases should respond directly to the influence of
the Warren Court in a positive direction.

Hypothesis # 24b. The agenda of the Supreme Court in civil

liberties cases Should respond directly to the influence of the
Warren Court in a positive direction.

212

Hypothesis # 25a. The agenda of the Courts of Appeals in
criminal procedure cases Should respond directly to the
influence of the Warren Court in a positive direction.

Hypothesis # 25b. The agenda of the Supreme Court in
criminal procedure cases Should respond directly to the
influence of the Warren Court in a positive direction.

Hypothesis # 26a. The agenda of the Courts of Appeals in

economic cases Should respond directly to the influence of the
Warren Court in a negative direction.

Hypothesis # 26b. The agenda of the Supreme Court in
economic cases should respond directly to the influence of the
Warren Court in a negative direction.

Nevertheless, the effect of the Burger Court Should be in the opposite direction for
criminal procedure and economic cases, as this Court generally sought to neuter or even
reverse the direction and influence of the Warren Court in these issue areas. Moreover, while
the Burger Court was not necessarily as conservative as some scholars and politicians had
thought or hoped in civil liberties issues, it certainly was more conservative than the Warren
Court, particularly once President Reagan appointed Justice O’Connor in 1981 (Mishler and
Sheehan 1993). Thus, I expect that the Burger Court’s influence on civil liberties cases be
negative as well:

Hypothesis # 27a. The agenda of the Courts of Appeals in
civil liberties cases Should respond directly to the influence of
the Burger Court in a negative direction.

Hypothesis # 27b. The agenda of the Supreme Court in civil
liberties cases should respond directly to the influence of the
Burger Court in a negative direction.

Hypothesis # 28a. The agenda of the Courts of Appeals in
criminal procedure cases Should respond directly to the

influence of the Burger Court in a negative direction.

213

Hypothesis # 28b. The agenda of the Supreme Court in

criminal procedure cases Should respond directly to the

influence of the Burger Court in a negative direction.

Hypothesis # 29a. The agenda of the Courts of Appeals in

economic cases should respond directly to the influence of the

Burger Court in a positive direction.

Hypothesis # 29b. The agenda of the Supreme Court in

economic cases should respond directly to the influence of the

Burger Court in a positive direction.
Accordingly, the assorted hypotheses asserted in this section represent the theoretical
expectations of various judicial influences, including the dynamic liberalism of the Supreme
Court and the Courts of Appeals, and the effect of the Warren and Burger Courts, on the
agendas of the Supreme Court and Courts of Appeals.
Exogenous Judicial Interventions

Time-series analysis pemTits the researcher to model independent shocks to series of

data, in order to ascertain whether such changes in policy or other external event affect the
system (Box and Tiao 1975). In this chapter I specifically will test whether a number of
interventions of a judicial nature influence agenda changes in the federal courts. Pacelle
(1991) maintained that the Supreme Court systematically alters its agenda preferences by
issuing landmark decisions in particular issue areas. AS he contended (34-35): "Landmark
decisions are important tools for justices to utilize in prospectively building future agendas
and structuring the nature of agenda change. . . The doctrinal changes created by the
landmark decisions virtually assure the justices the opportunity to address Similar types of
cases in order to continue the development of doctrine in these recently changes areas. . . A

landmark decision is a means by which justices create conditions inviting litigants to bring

214

cases that will fill in the nascent doctrine or Signaling litigants that such cases are no longer
welcome." In this regard, it is expected that as a greater percentage of specific kinds of cases
are brought before the justices, the opportunity for the Supreme Court to pursue its
predilections on agenda and policy is elevated.

One of the keys to intervention analysis is that the exogenous Shock be readily
identifiable (Mills 1990). AS Pacelle (1991, 35) asserted, landmark decisions provide such
a possibility: "The results of landmark decisions on the agenda can be measured empirically
through changes in the levels of agenda allocation that can be ascribed to the landmark."
While the specifics of coding these landmark interventions are described subsequently in the
Data and Methods section, it suffices here to profess that I will examine landmark'decisions
espoused by the Supreme Court in issue areas of civil liberties, criminal procedure, and
economics to establish their effect on the agendas of the Supreme Court and the Courts of
Appeals over time.

I expect to corroborate Pacelle’s findings for the Supreme Court’s agenda, as
landmark decisions in each issue area should positively influence its docket for that
particular issue. That is, the proportion of civil liberties cases on the Supreme Court’s
agenda Should increase as a result of landmark civil liberties decisions; and, an analogous
expectation should ensue for criminal procedure and economics cases when landmarks are
handed down in those issue areas. The expected effect from landmark cases need not be
confined to the high Court’s agenda, as rational litigants, their lawyers, and interest groups

should recognize the intended, future course of the Supreme Court based on its landmarks;

215

consequently, the agenda of the Courts of Appeals should follow suit. Ergo, I assert the
following:

Hypothesis # 30a. The agenda of the Courts of Appeals in
civil liberties cases Should respond directly to landmark civil
liberties decisions issued by the Supreme Court in a positive
direction.

Hypothesis # 30b. The agenda of the Supreme Court in civil
liberties cases should respond directly to landmark civil
liberties decisions issued by the Supreme Court in a positive
direction.

Hypothesis # 31a. The agenda of the Courts of Appeals in
criminal procedure cases should respond directly to landmark
criminal procedure decisions issued by the Supreme Court in
a positive direction.

Hypothesis # 31b. The agenda of the Supreme Court in
criminal procedure cases Should respond directly to landmark
criminal procedure decisions issued by the Supreme Court in
a positive direction.

Hypothesis # 32a. The agenda of the Courts of Appeals in
economic cases Should respond directly to landmark
economic decisions issued by the Supreme Court in a positive
direction.

Hypothesis # 32b. The agenda of the Supreme Court in
economic cases Should respond directly to landmark
economic decisions issued by the Supreme Court in a positive
direction.
External Influences on Agenda
All of the potential influences discussed thus far have one theme in common, to wit:

the past history of agendas, the liberalism of the Supreme Court and Courts of Appeals, the

tenure Of various Chief Justices, and landmark decisions are judicial-oriented. Nevertheless,

216

behavior in the federal courts may be influenced by forces outside the judicial branch and,
as a consequence, some of those potential forces should be analyzed as well.

The first of these potential influences concerns the non-judicial branches of the
federal government, sometimes referred to as the political branches. Indeed, the political
actions taken by both Congress and the President have variously been argued to influence
judicial behavior. Nevertheless, Segal (1997) persuasively theorized and demonstrated that
the Supreme Court ostensibly disregards the actions of Congress, whether statutory or
constitutional interpretation is at issue. Primarily for this reason, I expect that the agenda of
the Supreme Court will not be affected by the actions taking place in the political
environment.

On the other hand, litigants in the Courts of Appeals may be amenable to influence
from the political branches of government in a manner the justices of the Supreme Court are
not. Nonetheless, as contended by Hurwitz and Sheehan (1998), such influence may not be
uniform across distinct types of parties in reaction to a transformed political environment.
That is, rational litigants strive for the type of arena which is most responsive to their
interests, yet disparate litigants have discrete goals. Thus, when the political branches
become more liberal over time, corporate litigants, which generally bring appeals based on
economic issues, are more likely to pursue their claims in the courts, Since the political
environment has become less sympathetic to their interests. To the contrary, individual
litigants, whose appeals ordinarily are in the nature of civil liberties and criminal procedure

claims, likely would decrease their lower court petitions as the political environment grows

217

more sensitive to their interests. The following hypotheses incorporate these theoretical
expectations:

Hypothesis # 33a. The agenda of the Courts of Appeals in
civil liberties cases should respond directly to the liberalism
of the political branches in a negative direction.

Hypothesis # 33b. The agenda of the Supreme Court in civil
liberties cases should not respond to the liberalism of the
political branches.

Hypothesis # 34a. The agenda of the Courts of Appeals in
criminal procedure cases should respond directly to the
liberalism of the political branches in a negative direction.

Hypothesis # 34b. The agenda of the Supreme Court in
criminal procedure cases Should not respond to the liberalism
of the political branches.

Hypothesis # 35a. The agenda of the Courts Of Appeals in

economic cases Should respond directly to the liberalism of

the political branches in a positive direction.

Hypothesis # 35b. The agenda of the Supreme Court in

economic cases should not respond to the liberalism of the

political branches.‘S

One final non-judicial concept to be analyzed in this chapter is public opinion. AS

described in Hurwitz and Sheehan (1998, 6): "The . . . variable we use to measure social

change is the public mood index (Stimson 1991). While this variable is not a precise

measure of social change, it does provide an indication of the perception of the general

 

‘These expectations are consistent with those asserted in Hypotheses # 20a and # 23,
where an increasingly liberal federal judiciary Should lead to fewer economic appeals in the
lower courts; and, Hypotheses # 18a, # 19a, # 21, and # 22 are consistent with these
predictions as well, as a more liberal judiciary Should lead to enhanced civil liberties and
criminal procedure dockets in the Courts of Appeals.

218

public towards public policy issues." While some judicial scholars have asserted that public
Opinion has no influence on Supreme Court, whether individually or in the aggregate (Segal
and Spaeth 1993), others have found a direct link between public opinion and the decisions
of the Supreme Court, whether in the aggregate (Mishler and Sheehan 1993; Fleming and
Wood 1997) or individually (Mishler and Sheehan 1996).

Certainly no consensus exists within the judicial subfield on the effect of public
Opinion on the Supreme Court (see, e. g., Norpoth and Segal 1994). Nevertheless, these
studies incorporating public mood have concerned judicial decision making, yet they have
not considered the agenda-setting function of the Court. Accordingly, in this chapter I will
examine whether public opinion, as expressed by Stimson’s (1991) public mood index,
influences the agenda of the Supreme Court. Furthermore, I will base my theoretical
assumptions on the various studies which uncovered an influence of public opinion on
decision making in the Supreme Court. Accordingly, as public Opinion (expressed by the
public mood index) becomes more liberal over time, the justices would likely choose cases
embracing issues more in line with a liberal agenda; of course, these are the civil liberties and
criminal procedure cases. Conversely, a public more comfortable with liberal political
outcomes Should result in a Supreme Court agenda with fewer economic cases. These
expectations are expressed in the following testable hypotheses:

Hypothesis # 36. The agenda of the Supreme Court in civil
liberties cases Should respond directly to changes in public
opinion in a positive direction.

Hypothesis # 3 7. The agenda Of the Supreme Court in
criminal procedure cases should respond directly to changes
in public opinion in a positive direction.

219

Hypothesis # 38. The agenda of the Supreme Court in
economic cases Should respond directly to changes in public
Opinion in a negative direction.

No hypotheses are submitted for any potential influence of public Opinion in the
agenda of the Courts of Appeals. While the statistical reasons for ignoring this influence are
explained subsequently in the Data and Methods section, there are theoretical reasons for not
expecting an effect of public Opinion on lower court appeals. First, it is litigants, not judges,
who set the agenda of the lower courts, and litigants’ decisions to appeal likely have no basis
in public Opinion but instead are grounded on a rational evaluation of their probability Of
success on appeal (Songer, Segal, and Cameron 1995; Hurwitz and Sheehan 1998). Even
assuming, arguendo, that this calculus includes a consideration of factors which affect
appeals court decision making, it is not likely that appeals court judges are persuaded by
pubic Opinion. Scholars have contended that these judges are bound, at least to the certain
extent, by legal precedent (Howard 1981 ; Songer, Segal, and Cameron 1994; Johnson 1987),
or they pursue their own ideologies behind the cloak of legal doctrine (Wasby 1970; Songer
1982; 1987; Songer and Sheehan 1990). Since it is likely that decision making on the Courts
of Appeals is most fully explained by both the attitudinal and legal models in tandem
(Songer and Lindquist 1996; Hurwitz and Reddick 1996), then public opinion generally
should not enter into these judges’ decisions. As a consequence, there is no reason to believe
that rational litigants, primarily concerned with prevailing on appeal, would consider public

Opinion as part of their calculus to appeal. That is, public opinion apparently provides no

service to parties in enhancing their probability of success on the merits.

220

The theoretical expectations of this chapter having been expressed in the form of
these varied, testable hypotheses, the ensuing section describes the manner in which these

concepts are operationalized and tested.

DATA AND METHODS

I am interested in examining the various influences on agendas in the Supreme Court
and Courts of Appeals. In this section I explain the manner in which the data I adopt were
operationalized, measured and coded, and then I illustrate the statistical methods by which
I intend to analyze the influence of the myriad independent variables on the dependent
variables.
Coding of Data

Since the particulars of the various agenda data have been described in the prior
chapters, I simply will assert here that the categories of issue areas include: the civil liberties
issue, consisting of civil rights, first amendment, due process, and privacy cases; the criminal
procedure issue, consisting simply of criminal procedure cases; and, the economics issue,
consisting of economic, federal tax, and union and labor cases. Figure 1.1 depicts the civil
liberties agendas of these courts, while Figure 1.9 illustrates criminal procedure cases over
time. See Figure 5.1 for the aggregate economic agendas of these courts.

The liberalism of the Supreme Court, which stems from the Supreme Court Database,
measures the annual percentage of liberal decisions espoused by the Supreme Court. A

similar annual value for the Courts of Appeals, which is derived from the Courts of Appeals

221

Database, represents the liberalism of the federal appellate courts. Figure 5.2 presents the
dynamic liberalism of these courts.

Chief Justice Warren’s tenure ranged from 1953 through 1969, whereas Chief Justice
Burger commenced his term in 1969 and retired in 1986. Finally, Chief Justice Rehnquist
assumed this role in 1986. These variables take the form of a dummy, where a 1 was entered
for any year in which these justices sat on the Court in the role of Chief, zero otherwise.
Thus, there is a separate variable for each of these Chief Justices.

The landmark decisions stem from Pacelle’s (1991) coding in Appendix 4 of his
book. In particular, Pacelle lists a number of key Supreme Court decisions in various issue
areas. For any year in which a landmark decision was espoused, that landmark category was
coded as a 1, while any year in which no landmark was issued was coded as a zero. For
instance, Brown v. Board of Education, 347 US. 483 (1954), and Griswold v. Connecticut,
381 US. 479, 485 (1965), both appear as landmark cases in Pacelle’s Appendix 4. The
Supreme Court Database codes Brown as a civil rights case, while Griswold is codified as
a privacy case. Since both of these categories fall within the civil liberties classification of
this study, the civil liberties landmark variable was coded as a 1 in each of 1954 and 1965,
the year (not the term) in which these decisions were released, respectively. The other civil
liberties landmark cases from Pacelle’ S Appendix 4 form the remainder of this variable. See
Appendix B for details of the cases utilized as landmark civil liberties, criminal procedure,
or economics cases.

The rationale for coding landmarks in this fashion is to examine their effect
collectively. That is, from 1953 through 1988, according to Pacelle there were nine years

222

in which landmark civil liberties decisions were issued. Instead of testing the influence of
each of these landmarks one at a time,7 by coding this landmark variable as a series of Is for
each year a landmark was issued, I am able to test the idea that these decisions systematically
were issued by the Supreme Court as they sought to transform the nature of its agenda.
Accordingly, this coding scheme analyzed with time-series analysis provides a mechanism
to test whether these landmarks affected the agendas of both the Supreme Court and Courts
of Appeals in a cumulative sense over time.

AS for the political liberalism variable, I rely on the latent policy liberalism score for
the President, the House of Representatives, and the Senate as provided by Stimson,
MacKuen and Erikson (1995). I averaged the values identified by Stimson, et. al., for both
houses of Congress and the President for each year to devise a relative score of political
liberalism for the non-judicial branches of the federal government. AS expressed in Hurwitz
and Sheehan (1998, 7): "The rationale for averaging the scores emanates from the logic of
Hammond and Miller (1987) who asserted that all three institutions must be in agreement
before any policy outcome can be obtained." This political liberalism variable enables me
to test whether the political branches have any influence on the agendas of the federal courts.
See Figure 5.3 for the manner in which this measure has changed over time.

To reiterate, public opinion is operationalized by utilizing Stimson’s (1991) public
mood index. "This index is a composite, aggregate measure of American views about the

role of government that can be arrayed along liberal-conservative lines" (F lemming and

 

7Testing each landmark decision individually likely would induce unacceptable levels
of collinearity.

223

Wood 1997, 470). The longitudinal nature of Stimson’s index, and the manner in which it
was devised by his compiling the results of several public opinion polls into a relative score
of liberalism of the populace, make it appropriate for use as a proxy for public opinion in
general. See Figure 5.4.

Data Transformations for Time-Series Analysis

AS I have indicated in this chapter, time series will be utilized to appraise the
influence of these variables on the agendas of the federal courts. Indeed, time-series
techniques are necessary to explore systematically the dynamic causes of agenda fluctuation
over time. In this chapter, however, I will not employ the available but more complex time-
series procedures. Instead, the analysis in this chapter will proceed by applying regression
techniques. However, as I will discuss subsequently, estimation by ordinary least squares
(OLS) is not appropriate in this study due to the endogenous nature of agenda in the federal
courts. Instead, two-stage least squares (ZSLS) will be utilized.

Nevertheless, before applying any least squares method to time-series analysis,
whether OLS, ZSLS or otherwise, a few preliminary inquiries must be settled. More
specifically, in order to obtain consistent estimates, the dependent variable must be modeled
as an AR process with white noise (that is, no serial correlation) in the error term. There are
Six dependent variables to be analyzed in this chapter: the civil liberties agendas of the
Supreme Court and of the Courts of Appeals, the criminal procedure agendas of the
Supreme Court and the Courts of Appeals, and the economic agendas of the Supreme Court
and Couris of Appeals. I determined via Box-Jenkins procedures (1976) that each of these
dependent variables in first differences can be represented adequately by an AR process;

224

indeed, each series is an autoregressive of the second order, save for the civil liberties agenda
in the lower courts, which is best modeled as an AR(1) process. Moreover, the error terms
exhibit only white noise. Accordingly, I am able to estimate these models with least squares,
Since ZSLS estimators will be consistent under these circumstances, so long as the
endogenous nature of the system is addressed.8

While these prerequisites have been satisfied, I am not quite ready to confront the
endogeneity issue. First, I must ensure that stationarity and balance are achieved for each
equation. Valid inferences in time series regression models can only be made when the data
are stationary. AS I explained in chapter 3 (76): "Stationarity simply denotes that time plays
no role in the sample moments of the mean and variance of the data. That is, stationarity
implies that both mean and variance are unconditional, finite, and not dependent on time;
moreover, shocks to the system will dampen over time, as the system is mean-reverting."
If the data are not stationary, then appropriate transformations must be made to ensure that
spurious results do not transpire. Balance, on the other hand, signifies that all of the
variables in the equation must be integrated of the same order; since stationary data are
integrated as 1(0), if each variable is stationary in the system, then there is balance.

The agenda series were analyzed previously in chapter 3, and all of them required to

be first-differenced to achieve stationarity.9 As for the independent variables, Box-Jenkins

 

8If a moving average process best represents a dependent time-series variable, then
maximum likelihood procedures Should be employed, not least squares.

9The several series which comprise the aggregated economic series in this chapter
displayed evidence of a unit root in chapter 3, and taking the first difference there resulted
(continued...)

225

analysis demonstrated that all of these series were integrated of the first order, thus
exhibiting nonstationary processes. Once again, however, first-differencing resulted in data
integrated as 1(0); that is, the differenced data were stationary, and any consequent regression
equation would be balanced. Thus, Spurious estimates should not be expected.

One final data transformation must be pondered before moving to the estimation
stage, and that consideration is whether lags of the variables need to be incorporated into the
system of equations. The dependent variables will be represented by the contemporaneous
agenda (in first differences) of the Supreme Court or Courts of Appeals for the various issues
areas. It makes no substantive sense to include contemporaneous independent variables,
however, Since this is time-series analysis. Thus, with the sole exception of the lagged
endogenous, dependent variables, each unrestricted model will implement one, two, and
three lags of the independent variables, whether exogenous or endogenous. To ensure
validity, the respective lags for these independent variables must be formulated from the
differenced data.

Lagged endogenous variables are also appropriate in the unrestricted regressions,
since the past history of each dependent variable is a probable influence on the current
measure of the dependent variable at issue. The number of endogenous lags of the dependent
variable which is befitting to each model is conditional upon the type of AR process endemic

to the particular dependent variable. With the sole exception of the civil liberties series in

 

9(...continued)
in stationary data for each of these series. Similarly, the aggregated economic series for both
courts in this chapter needed to be differenced to achieve stationarity.

226

the appeals courts, each dependent variable was appropriately modeled as an AR(2);
consequently, each of these equations will incorporate two endogenous lags of the dependent
variable. Since the civil liberties agenda in the Courts of Appeals was represented by an
AR(1) process, however, only one endogenous lag will be incorporated into that regression.
Of course, the lags of these endogenous variables will be generated from the differenced
series.

With these time-series transformations and other considerations regarding data now
complete, it is appropriate to move on to the issue of estimating a system of equations which
feature endogeneity.

Estimation of an Endogenous System

OLS assrunes that causation exists solely in one direction, yet this assumption is
violated when there is endogeneity, or feedback, in a system of equations. If classic OLS is
performed on such a system, the OLS assumption that the independent variables are
distributed independently of the error term is violated. Once this occurs, the result is biased
and inconsistent estimators. Thus, classic OLS is inappropriate to estimate a Simultaneous
system of equations which exhibits endogeneity (Gujarati 1995).

Accordingly, when there is endogeneity, it cannot be assumed that the right-hand side
(rhs) variables are uncorrelated with the error term. Maneuvering around this simultaneity
problem necessitates that the researcher engage in structural equations modeling. In that
regard, a reduced form needs to be calculated. A reduced form expresses endogenous
variables solely in terms of exogenous variables which are determined outside the model.
Once a reduced form is established, such that the endogenous variables are expressed only

227

in terms of truly exogenous variables, then the equation can be estimated, and unbiased and
consistent estimators will result, so long as the system is identified.

Arriving at the reduced form, therefore, is the first step in the process of estimating
the system of equations. But, there remains the SO-called identification problem, which
concerns whether parameter estimates can be obtained from the reduced form coefficients,
which can be accomplished only if the system is identified. "For an equation to be identified
. . . it must be shown that the given set of data will not produce a structural equation that
looks similar in appearance to the one in which we are interested" (Gujarati 1995, 659).
Thus, for parameter estimates to be attained, additional information is needed in each
equation to set it apart from the other.

The practical manner of determining whether a system is identified is by the rank and
order conditions of identification. The order condition is a necessary but not sufficient
condition.10 In a model of G Simultaneous equations, an equation is identified if it excludes
at least G-l variables, endogenous or otherwise. If there are G-l exclusions, then it is just
identified (the minimum needed to estimate the system); more than G-l exclusions provides
for over-identification, which iS more restrictions than are necessary; but, if there are less
than G-l exclusions, the equation is unidentified, which precludes meaningful estimation.
The rank condition of identification is both a necessary and sufficient condition.l ' In a model

of G equations, an equation is identified if and only if at least one, non-zero (G-1)*(G-1)

 

10Order refers to the order of a matrix (that is, the number of rows and columns).

”Rank refers to the rank of a matrix, which is the order of the largest, square sub-
matrix whose determinant is not zero.

228

determinant is contained in the array of coefficients with which those variables excluded
from the equation in question appear in the other equations.

The manner in which the reduced form of the system is calculated is ZSLS, a
technique which provides for unbiased and consistent estimators for the system. The basic
idea of ZSLS is to ensure that the dependent variable in each equation in the system is the
only endogenous variable in the equation, such that it is a function solely of the error term
and the exogenous variables. ZSLS accomplishes this feat by finding a proxy (via an
instrumental variables procedure) for the endogenous variables which is not correlated with
the error term, which are purified of any influence of the stochastic error.

More particularly, in ZSLS each endogenous variable is regressed on all exogenous
variables in the system; that is, the reduced form is estimated in this first stage. The
predicted values are saved from these regressions, which become instrumental variables for
the endogenous variables in those equations in which the endogenous variables are on the
rhS. When the second stage of the regressions is estimated utilizing the instrumental values
for the endogenous variables, these instruments are not correlated with the error term, and
thus the OLS assumption is no longer violated.

I have argued throughout this study that the agendas of the federal courts exist within
an endogenous system, and the VAR analyses in chapters 3 and 4 provided considerable
evidence in this regard. Accordingly, Since there is feedback within this system, I cannot
estimate these models with OLS, otherwise biased and inconsistent coefficients will result.

Thus, I necessarily must calculate a reduced form and identify the system of equations so that

229

_1“‘—"=:-r-_

the multivariate analyses are able to manifest meaningful parameter estimates which are
unbiased and consistent.
The civil liberties system is represented by the following equations:

(1) Y1: at + plYt-l + [5sz + Baxz + P4X3 + [35X4 + péxs + P7x6 +
psx7 + psxs + u;

(2) x11: Y + 6IXrt-I + 6zY + 63X2 + 54X3 + 55Xa + 56x5 + 57x6 +
53X7 + 69X, + v,

where: Y = civil liberties agenda, Supreme Court; X1 = civil liberties agenda, Courts

of Appeals; X2 = Supreme Court liberalism; X3 = political liberalism; X4 = Chief

Justice Warren dmnmy; X5 = Chief Justice Burger dummy; X6 = Chief Justice

Rehnquist dummy; X7 = civil liberties landmark decisions; X8 = public mood; and,

X9 = Courts of Appeals liberalism.”
The criminal procedure and economic systems are nearly identical, to the system referenced
above, the only distinction being that the agenda and landmark decision variables are
specified for that particular issue; that is, in the criminal procedure system, the agenda and
landmark variables include those coded for the criminal procedure issue, and the same
decision rule applies for the economic system.

In order to calculate the reduced form, the endogenous variables, which include the
variables for the civil liberties agendas of both the Supreme Court and the Courts of Appeals,

need to be regressed on all the exogenous variables in the system in separate regressions.

The predicted (y-hat) values are saved from each of these regressions, and those predicted

 

”For the sake of Simplicity, I have not included all the lags for each of the variables,
but the number of lags Specified earlier in this chapter apply to this system of equations.

230

values become the proxies, or instrumental variables, for the endogenous variables, which
then are regressed on the exogenous variables for that equation.'3

Once the reduced form is calculated and the predicted values are saved, stage two of
ZSLS requires that these instrumental variables be used as proxies for the endogenous
variables in the original equations in the system. Unless the system is identified, however,
specific parameter estimates will not be obtainable. Nevertheless, I am able to utilize ZSLS
because the equations in these systems are ‘just’ identified. That is, there is one variable in
each equation which is excluded from the other equation in the system, the minimum
required to proceed with ZSLS. In particular, public mood is excluded from equation (2),
while Courts of Appeals liberalism is excluded from equation (1 ), as these are the identifying

variables for each equation. Consequently, the order condition of identification is satisfied. ‘4

 

1:"It is intuitive that when X causes Y and Y in turn causes X, there is feedback in the
system. Under these conditions, instrumental variables need to be generated to rid each
equation of any correlation between the variables and the error term. The same concern
arises when the rhs variables include lagged endogenous, dependent variables, whether or
not a system of equations is specified (Fair 1970; Greene 1990). Instruments are needed here
as well, in order to ensure that these variables are not correlated with the error term;
otherwise, inconsistent parameter estimates will result (Ostrom 1990).

1"AS an aside, the order condition is based on prior theoretical expectations that
certain variables are truly exogenous and appear only in one equation. This is a tenuous
point, at least according to Sims (1980), who argued that such theoretical suppositions are
not credible. For this reason, he introduced the VAR approach to time series, a reduced-form
system which assumes that all variables are endogenous in the system.

Notwithstanding Sims’ counsel, 1 will proceed with the SEQ modeling in this
chapter. First, the theoretical presumptions I employed for excluding particular variables in
the various systems are defensible. Second, to utilize multivariate analysis I need to engage
in appropriate SEQ modeling, Since VAR is inadequate in this regard for the data involved.
Finally, I incorporated VAR in the prior chapters, and the results fiom the SEQ modeling in

(continued...)

231

The rank condition is satisfied as well, Since at least one non-zero determinant exists (in this
case, they are scalar terms)"

Since the equations within the system are identified, the system itself is identified,
signifying that the second stage of ZSLS can proceed, and unbiased and consistent estimates
will be Obtained. In fact, all aspects of OLS interpretation apply to the second stage of ZSLS,
except that the R2 statistic is meaningless; thus, it will not be reported in the analysis.
Accordingly, I will run each model with the instrumental variables as proxies for the
endogenous variables, lagged appropriately.

A few caveats need to be made regarding the dummy variables for the Chief Justices.
First, while Chief Justice Rehnquist’s tenure was included in calculating the reduced form,
it will not be included in the second stage regressions of ZSLS. AS stated previously, there
is not sufficient data to incorporate the influence of his tenure on agenda due to the time-
series transformations of differences and lags. However, this variable was used to create the
instruments in order to provide the proxies with as much independent information as
possible, such that the second stage regressions would converge closer to the true parameter

values. Furthermore, while the Warren and Burger variables will be tested, due to the nature

 

"’(...continued)
this chapter will provide a kind of sensitivity analysis, such that the VAR and ZSLS analyses
can be compared.

lSln the prior section, I provided the theoretical reasons for excluding public mood
from the equations where agenda of the Courts of Appeals is the dependent variable, and for
excluding Courts of Appeals liberalism where agenda of the Supreme Court is the dependent
variable. The current section above, then, provides the statistical rationale for excluding
these variables.

232

of coding these dummies, the variables are nearly perfectly collinear (I r I > .9). Thus, they
cannot be run in the same regressions. Accordingly, for each dependent variable, two
regressions will be reported, one with the influence Of Chief Justice Warren’s tenure, the
other with that of Chief Justice Burger.

Finally, I will utilize the reduction process of model building, as described in chapter
3. In particular, I will commence each unrestricted model with all the available independent
variables (utilizing instrumental variables where appropriate), and eliminate ineffectual
variables while applying various diagnostics, in order to achieve a final model which is valid

both theoretically and statistically (Granato 1991).

ANALYSIS

For each system of equations 1 will present the results of four regressions, each of
which is the second stage of a ZSLS estimation. Two of these models per system include the
Supreme Court’s agenda for that issue as the dependent variable, while the other two
embrace the agenda for Courts of Appeals as the dependent variable. The only difference
between the models with the same dependent variable is that one contains the variable for
Chief Justice Warren, the other for Chief Justice Burger. As will be seen, the results
generally comport with my theory that the agenda functions in the federal courts act within
an endogenous system, as the dockets of the Supreme Court and the Courts of Appeals help
to comprise the agenda of the other court for some issues but not others. Other factors enter
into this equation as well, but the influence of the agendas of each of these courts on the
agenda Of the other is considerable. On a technical note, every model to be reported

233

performed rather well, and several diagnostic tests revealed no serial correlation problems
in any of them; that is, the error terms exhibited nothing but white noise.‘6
Civil Liberties System

I begin by discussing the model where the dependent variable is the Supreme Court’ S
agenda, with the Burger variable included in the model. Every independent variable in this
regression was Significant beyond the conventional 95% confidence level. In particular, it
was expected that the Supreme Court’s civil liberties agenda would cause changes over time
in the agenda of the Courts of Appeals for this issue, but that an endogenous effect Should
not be anticipated. For this model, this forecast was partially true. True to form, the
Supreme Court’s past civil liberties agenda at a lag of one year proved to influential with
respect to its contemporaneous agenda. Nevertheless, I expected that the Courts of Appeals
would have no influence on the Supreme Court’s agenda, but at a lag of two and three years
there is a significant effect, and it is in the negative direction. Thus, as the civil liberties
agenda in the lower courts increases, the Supreme Court decreases the civil liberties cases
on its agenda, at least where the Burger Court’s influence is concerned.

While I found that, as expected, an increasingly liberal Supreme Court accepted a
greater proportion of civil liberties cases, the influence of the Burger Court was incorrectly
signed, as this variable also was associated with an increasing civil liberties agenda.
Landmark civil liberties cases indeed led to a stronger civil liberties agenda, thus confirming

the expectations asserted by Pacelle (1991) and tested herein. On the other hand, in a

 

”Since lagged endogenous variables are included in these models, the Durbin-Watson
statistic is inappropriate and will not be reported.

234

somewhat countermajoritarian fashion, when the political branches became increasingly
liberal, the Supreme Court heard relatively fewer civil liberties cases over time, at least when
the Burger Court’s influence is encompassed. Finally, public mood proved to be significant
as well, but surprisingly in a negative direction. That is, as the populace became more
liberal, the Burger Court became less interested in civil liberties cases. Table 5.1 displays
the results for this model.

The findings were remarkably different when the Warren Court variable was included
in the regression of the Supreme Court’s civil liberties agenda. In fact, as demonstrated in
Table 5.2, the only significant variable was public mood and, contrary to the model with the
Burger variable, this coefficient was positive; notwithstanding, the influence Of the Warren
Court also was very nearly statistically Significant, with a p-value of .053.'7 All the other
variables in this model either were unwarranted or insignificant. The implications of these
findings will be addressed subsequently in the Discussion section.

As expected, the influences on the civil liberties agenda of the Courts of Appeals
were quite different than those on the Supreme Court’s agenda. In the model including the
Burger variable, the past civil liberties agendas Of the Supreme Court and Courts of Appeals
were influential and in the expected positive direction. Additionally, the liberalism of the
Courts of Appeals was a significant influence on their civil liberties dockets but,

interestingly, in a negative direction. As illustrated in Table 5.3 , no other variables achieved

 

17Since all but two of the civil liberties landmark decisions were espoused by the
Warren Court, there is some multicollinearity between these variables, which likely neutered
the overall effect of each. Nevertheless, the regression results as reported for this model
displayed no serial correlation.

235

 

statistical significance. The Courts of Appeals model with the Warren variable achieved
nearly identical results, the only difference being that the past agenda history of the Courts
of Appeals was insignificant, although the Supreme Court’s agenda was influential in the
expected positive direction. See Table 5.4

Criminal Procedure System

While many researchers include criminal procedure cases within the rubric of an
overall civil liberties category (see the references in chapter 4), the results from the models
employing a separate criminal procedure variable were notably distinct from the models with
civil liberties as the dependent variable, thus substantiating the argument for disaggregating
these issues in judicial studies. In any event, Table 5.5 presents the results for the model of
this issue in the Supreme Court, with the Burger variable included as exogenous. In
particular, the significant coefficients were for the past Supreme Court criminal procedure
agenda and landmark criminal procedure decisions, and each was in the expected positive
direction. Also, as hypothesized, there was no effect in either direction stemming from the
agenda of the appeals courts. Nearly identical results were Obtained for the model employing
the Warren variable, with the influences stemming from the past agenda history of the
Supreme Court and from landmark criminal procedure cases. See Table 5.6.

Once again, the factors which influence agenda in the Courts of Appeals were rather
unlike those which affect the Supreme Court. For the model with the Burger variable, one
of the consequential variables proved to be the Supreme Court’s criminal procedure agenda,
in the expected positive direction. The other Significant variables included the liberalism Of
the Supreme Court the Courts of Appeals, and the liberalism of the political branches; the

236

 

intriguing aspect of these outcomes is that all of these coefficients were negative, yet only
the political liberalism variable was predicted to be in this direction. Indistinguishable
results obtained when the Warren variable was incorporated into this model. See Tables 5.7
and 5.8, respectively.

Economics System

If the dockets of the Courts of Appeals influence the Supreme Court’s agenda at all,
as I have argued it should be in the area of economic cases. Indeed, the findings of this
analysis generally support these expectations. That is, in the Supreme Court economics
model with the Burger variable, at three lags the Courts of Appeals’ agenda significantly
influenced agenda changes in the Supreme Court in the expected positive direction.
However, the agenda in the Courts of Appeals at one lag was a negative influence on the
Supreme Court’s agenda. Additionally, as the Supreme Court became more liberal over
time, the economic agenda increased, which was not as expected, although landmark
decisions had the anticipated positive influential effect in setting the high Court’s economic
agenda. While the Burger Court’s influence led to a weaker economic agenda in the
Supreme Court, a more liberal public mood led to a more prominence for this issue. See
Table 5.9.

The results were a bit different, however, when the Warren Court variable was
included. In this instance, the economic agenda of the Supreme Court was influenced only
in a positive direction by the agenda in the Courts of Appeals, while it was affected
negatively by the Supreme Court’s past economic agenda. Moreover, Supreme Court
liberalism was related to both an increasing and decreasing economic agenda, depending on

237

the number of lags. And, contrary to the findings on the Burger Court’s influence, a more
liberal public mood led to fewer economic cases on the agenda of the Supreme Court. See
Table 5.10.

When the economic agenda of the Courts Of Appeals with the Burger variable was
analyzed, the past economic history was both positively and negatively influential, once
again the difference being conditional upon the number of lags. Furthermore, as the
liberalism of the Supreme Court and Courts of Appeals increases, the dockets in the lower
courts followed suit with economic issues, yet this result was not as expected. Finally, when
this model incorporated the Warren variable, the only significant variables were the Courts
of Appeals’ past history and the Supreme Court’s liberalism, and both of these variables

were incorrectly signed. See Tables 5.11 and 5.12.18

DISCUSSION
This analysis demonstrates that a variety of factors influence agenda setting in the
Supreme Court and Courts of Appeals. Some of these factors were expected, while others
were not. Nevertheless, it should be clear by this stage of this study that these factors are
conditional upon the type of issue area involved. Furthermore, as I have argued throughout,
the federal courts do seem to act within a system when agenda is concerned, as the Supreme

Court causes changes in the cases being appealed in the lower courts, while the justices

 

18For each Of these regressions, landmark decisions had to be removed in order to rid
the models of serial correlation. While I would have preferred to retain this variable due to
theoretical considerations, it was far from Significant before being extracted from the models.

238

 

 

 

themselves are influenced by some, but not all, actions in the courts below with respect to
agenda.

With respect to the multivariate influences on agendas, I tested a number of models
in order to ascertain the extent to which some of the potential factors influence agenda
setting. In the civil liberties system, for instance, the past history of the Supreme Court’s
agenda in this area has an influence on current cases, while a liberal Supreme Court is
amenable to a more prominent civil liberties agenda. Moreover, as Pacelle (1991) asserted,
landmark cases do have an influence on the Supreme Comt’s dynamic agenda. One
interesting result was that as public opinion became more liberal, the Warren Court’s
influence on civil liberties increased, while civil liberties cases decreased under the influence
of the Burger Court. Instead of being an unexplained anachronism, this result generally
comports with the findings ofMishler and Sheehan (1993). They reported that the Supreme
Court was directly influenced by public Opinion until 1981, when its decisional trends
became much more conservative than the public mood. This latter time period, of course,
coincides with the Burger Court; thus, it is logical to think that the Burger Court, particularly
after Justice O’Connor was appointed to the Bench, had a civil liberties agenda which was
not as liberal as the Warren Court, which rode the wave of increasing public liberalism for
most of his tenure.

At the Courts of Appeals, clearly the Supreme Court’s agenda was strongly
persuasive on civil liberties appeals, even more SO than such appeals in the lower courts.
That increasing liberalism in both courts led to relatively fewer civil liberties cases might be

explained by Howard’s (1981) explication, where corporate litigants need to pursue their

239

 

interests in the Courts of Appeals, regardless of other factors. Thus, higher judicial
liberalism leads to lower civil liberties appeals simply because in a relative sense, economic
issues necessarily continue to be appealed.

As for criminal procedure cases at the Supreme Court, the only factors which affect
this agenda include the Supreme Court’s past agenda and landmark decisions. That is, the
justices do not seem concerned with anything but their own actions in setting the criminal
procedure agenda. This appears to be consistent with attitudinal theory (Segal and Spaeth
1993), where the Supreme Court is sufficiently insulated so as not to be concerned with
exoteric influences. Yet, it also comports with Pacelle’s (1991) neoinstitutional theory on
agenda transformation, such that the Supreme Court takes tangible steps to drive its agenda
in the direction of its preferred policy outcomes, but that such transition is subject to inertia
stemming from the Court’s institutional features. Notwithstanding, the fact that no other
variables were significant might be explained by the fact that many of the Supreme Court’s
criminal procedure cases originate in State Supreme Courts.

Once again, the Supreme Court has an appreciable influence on the agenda in the
Courts of Appeals for criminal procedure cases, particularly when compared to the effect Of
the lower court’s own agenda history for this issue. Nevertheless, the Supreme Court’s
agenda is not the SOle factor in criminal procedure appeals. In particular, elevating liberalism
within both the judiciary and the political branches leads to fewer of these appeals. While
this result was expected only for the political branches, one possible explanation for this
result could be that more criminal appeals are made by prosecutors than is commonly
believed. That is, as the judiciary becomes more liberal, it is more likely to rule against the

240

 

govermnent in criminal cases; thus, a prosecutor who lost a motion in the District Court, such
as a motion by the defense to exclude evidence pursuant to the Fourth Amendment, might
be less inclined to appeal the adverse decision and instead place prosecutorial resources
elsewhere (the efforts Of Independent Prosecutor Kenneth Starr notwithstanding!) Since this
study examined only the relative proportion ofcases on the dockets Of the appeals courts, and
not whether such appeals were being made by the government or the accused, a future line
of research would be to ascertain the types of litigants appealing such criminal procedure
cases, and whether an increasing judicial liberalism affects disparate parties in criminal
matters distinctly.

Rather interesting results ensued fi'om the economic system models. At the Supreme
Court, it seems as if the justices do seek to avoid economic cases, particularly where the
Warren Court was involved. First, there was an endogenous effect arising from the appeals
courts for this issue, but it was negative when influenced by the Warren Court and both
negative and positive with the Burger Court. Thus, the feedback effect was present, but the
direction of this effect was mixed. Additionally, when the Supreme Court issued a greater
proportion of economic decisions, the Warren Court justices reacted by moving away from
these types of cases in the future, thus potentially confirming part of the findings of
Tanenhaus, et. al. (1963). Similarly, the Warren Court was less likely to hear economic
cases when public opinion became more liberal, and landmark economic cases had no effect
on Chief Justice Warren’s liberal tenure. The story is very different for the Burger Court,
however, as a more liberal populace led to a more prominent economic agenda, again
comporting with the theory espoused by Mishler and Sheehan (1993) as the Burger Court

241

became increasingly counterrnaj oritarian in the 1980s. Moreover, increasing Supreme Court
liberalism led to higher rates Of economic cases on the docket when influenced by the Burger
Court, possibly because the conservative nature of this Court was evidenced by a desire to
move away from the liberal tendencies of the past. Finally, contrary to the results of the
Warren Court, landmark economic decisions led to relatively greater numbers of economic
cases on the Supreme Court’s agenda, again demonstrating the conservative temperament
Of the Burger Court.

The final models concerned the influences on the economic agenda of the Courts of
Appeals. The most conspicuous result here is that increasing judicial liberalism surprisingly
led to higher economic appeals. Again, this could be explained by Howard (1981), in that
corporate interests will continue to pursue their economic interests in federal court,
regardless of the ideological nature of the circuits.

Accordingly, a number of conclusions can be extracted from this chapter. First, the
process of setting the agendas of the federal courts is more complicated than previously
surmised, as a multitude of factors enter into the decisional calculus of both the Supreme
Court justices and litigants in the Courts of Appeals. Moreover, except for criminal
procedure cases, the Supreme Court is not as insulated from outside influence as might
otherwise appear; although, the actions of the political branches rarely if ever influence their
agenda decisions, thus apparently confirming Segal’s (1997) thesis. Furthermore, the federal
courts do operate as a system. A hierarchy with respect to agendas seems to exist (Baum
1994), as the Supreme Court has a good deal of influence on parties in the lower courts when
it comes to their decision to appeal. On the other hand, the direction of causality is not

242

merely in one direction, as the this analysis demonstrates once again that the water does
indeed flow uphill. This river does not always flow in this arduous direction, and its strength
iS not entirely consistent. Nevertheless, flow uphill it does, as the agenda in the Courts of

Appeals does have a certain influence on the agenda of the Supreme Court.

243

Percent Cases on Docket

Percent Judicial Liberalism

.;
<3
<3

 

 

 

 

 

Court:

Supreme Court

Courts of Appeals

1940 1946 1952 1958 1964 1970 1976 1982 1988 1994

1943 1949 1955 1961 1967 1973 1979 1985 1991

FIGURE 5.1
AGGREGATE ECONOMIC CASES, 1940-1995

1M

 

lvv

 

 

 

1952 1956 1960 1964 1968 1972 1976 1980 1984 1988

 

Court:

Supreme Court

Courts of Appeals

1954 1958 1962 1966 1970 1974 1978 1982 1986 1990

FIGURE 5.2
JUDICIAL LIBERALISM, 1952-1990

244

 

.L

.

 

-‘

... y
"\_l‘ . ‘-

 

i‘vvi

Percent Political Liberalism

Liberalism

 

80

701

601 W

 

 

401

 

301

 

 

 

 

I1 I I I I I I I I I I I I I I I I I I I I I I I III I I I I IIIII

1954 1958 1962 1966 1970 1974 1978 1982 1986 1990
1956 1960 1964 1968 1972 1976 1980 1984 1988 1992

FIGURE 5.3
LIBERALISM OF THE PRESIDENT AND CONGRESS, 1954-1992

150

 

1401

1301

 

1201

 

 

110.

1

 

 

100
1952fi '19'56' ' '19'60 ' '19-64 ' '19'68' ' '19'72' ' '19'76' ' 19100 ' '1904' ' 1938' “1992

1954 1958 1962 1966 1970 1974 1978 1982 1986 1990 1994

 

FIGURE 5.4
PUBLIC MOOD, 1953-1992

245

TABLE 5.1
SUPREME COURT CIVIL LIBERTIES (BURGER) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source | SS df MS Number of Obs = 22
--------- +------------------------------ F( 9, 12) = 5.07
Model I 947.911169 9 105.323463 Prob > F = 0.0056
Residual | 249.479752 12 20.7899794 R-squared =
--------- +------------------------------ Adj R-squared =
Total | 1197.39092 21 57.0186153 Root MSE = 4.5596
Coef. Std. Err. t P>lt| [95% Conf. Interval]
SupCt
CivLib
Agenda
(1 lag) 2.401253 1.000953 2.399 0.034 .2203634 4.582143
CtsApp
CivLib
Agenda
(2 lags) -6.53211 1.584282 —4.123 0.001 -9.983963 -3.080256
CtsApp
CivLib
Agenda
(31ags) -5.223624 1.540452 -3.391 0.005 —8.579981 -1.867267
Public
Mood
(1 lag) -l.394307 .5801842 -2.403 0.033 ~2.65842 -.l301941
SupCt
Liberalism
(2 lags) .3242274 .1340073 2.419 0.032 .0322507 .6162041
SupCt
Liberalism
(3 lags) .43314 .1341943 3.228 0.007 .1407558 .7255242
Political
Liberalism
(2 lags) -1.025765 .2939995 -3.489 0.004 -1.666335 -.3851949
CivLib
Landmark
Decisions 11.01698 5.01122 2.198 0.048 .0984722 21.93549
CJ
Burger 20.91058 5.527964 3.783 0.003 8.866184 32.95498
Constant —6.688335 3.0082 -2.223 0.046 -13.24264 -.1340306
246

TABLE 5.2

 

 

 

 

 

 

 

 

 

SUPREME COURT CIVIL LIBERTIES (WARREN) - ZSLS REGRESSION RESULTS

Source | SS df MS Number of Obs = 23
--------- +------------------------------ F( 7, 15) = 2.26

Model I 561.907359 7 80.2724798 Prob > F = 0.0874
Residual | 532.341346 15 35.4894231 R-squared =
-- ------- + ------------------------------ Adj R-squared =

Total I 1094.24871 22 49.7385775 Root MSE = 5.9573

Coef. Std. Err. t P>|t| [95% Conf. Interval]

SupCt
CivLib
Agenda
(2 lags) .5452691 .504253 1.081 0.297 -.5295208 1.620059
Public
Mood
(3 lags) 1.497352 .5361513 2.793 0.014 .3545721 2.640131
SupCt
Liberalism
(1 lag) -.2949136 .1573387 -1.874 0.080 —.6302731 .0404458
SupCt
Liberalism
(3 lags) .2676368 .1599822 1.673 0.115 -.0733572 .6086307
Political
Liberalism
(3 lags) -.1082416 .244785 -0.442 0.665 -.6299885 .4135054
CivLib
Landmark
Decisions -4.638528 5.133775 -0.904 0.381 -15.58091 6.303853
CJ
Warren 7.02768 3.347657 2.099 0.053 -.1076809 14.16304
Constant -1.918059 1.970051 -0.974 0.346 -6.117124 2.281006

 

247

 

 

'FABLEfL3
COURTS OF APPEALS CIVIL LIBERTIES (BURGER) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

Source I 88 df MS Number of obs = 30
--------- +------------------------------ F( 8, 21) = 2.33
Model | 154.485167 8 19.3106459 Prob > F = 0.0578
Residual I 174.2095 21 8.2956904? R-squared =
--------- +------------------------------ Adj R-squared =
Total | 328.694667 29 11.3342989 Root MSE = 2.8802
Coef. Std. Err. t P>|t| [95% Conf. Interval]
CtsApp
CivLib
Agenda
(1 lag) 1.690862 .7984015 2.118 0.046 .0304952 3.351229
SupCt
CivLib
Agenda
(2 lags) .707872 .3097456 2.285 0.033 .0637208 1.352023
SupCt
CivLib
Agenda
(3 lags) .6721199 .2677231 2.511 0.020 .1153591 1.228881
CtsApp
Liberalism
(2 lags) -.2744219 .1131385 -2.426 0.024 -.5097063 -.0391376
SupCt
Liberalism
(1 lag) -.1270054 .0660947 -1.922 0.068 -.2644568 .0104461
Political
Liberalism
(2 lags) -.113749 .0674891 -1.685 0.107 -.2541003 .0266024
CivLib
Landmark
Decisions -2.759482 1.920598 -1.437 0.166 -6.753584 1.23462
CJ
Burger -2.499635 1.595211 —1.567 0.132 -5.817059 .8177885
Constant 1.540711 1.166826 1.320 0.201 -.8858353 3.967258

 

248

 

 

 

TABLE 5.4
COURTS OF APPEALS CIVIL LIBERTIES (WARREN) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of Obs = 30
--------- +------------------------------ F( 8, 21) = 1.94
Model | 139.525181 8 17.4406476 Prob > F = 0.1076
Residual l 189.169486 21 9.00807075 R—squared =
--------- +------------------------------ Adj R-squared =
Total I 328.694667 29 11.3342989 Root MSE = 3.0013
Coef. Std. Err. t P>It| [95% Conf. Interval]
CtsApp
CivLib
Agenda
(1 lag) 1.115185 .7026003 1.587 0.127 -.3459525 2.576322
SupCt
CivLib
Agenda
(2 lags) .7360705 .339288 2.169 0.042 .0304825 1.441658
SupCt
CivLib
Agenda
(3 lags) .6661405 .2908103 2.291 0.032 .0613673 1.270914
CtsApp
Liberalism
(2 lags) -.289255 .117363 -2.465 0.022 -.5333248 -.0451852
SupCt
Liberalism
(l lag) -.1293552 .0688713 -1.878 0.074 -.2725808 .013870
Political
Liberalism
(2 lags) -.0936502 .0686588 -1.364 0.187 -.2364341 .0491336
CivLib
Landmark
Decisions -2.55458 2.141196 -1.193 0.246 -7.007441 1.898281
CJ
Warren 1.130857 1.459369 0.775 0.447 -1.904067 4.16578
Constant -.3393006 .8218128 -0.413 0.684 —2.048354 1.369753

 

249

 

TABLESfi
SUPREME COURT CRIMINAL PROCEDURE (BURGER) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Source | SS df MS Number of Obs = 24
--------- +-~---------------------------- F( 9, 14) = 3.78
Model | 270.672413 9 30.0747126 Prob > F = 0.0131
Residual | 111.440916 14 7.96006541 R-squared =
- -------- + ------------------------------ Adj R—squared =
Total I 382.113329 23 16.613623 Root MSE = 2.8214
Coef. Std. Err. t P>|t| [95% Conf. Interval]
SupCt
CrimPrO
Agenda
(1 lag) .7051993 .4311168 1.636 0.124 -.2194542 1.629853
SupCt
CrimPro
Agenda
(2 lags) 1.589786 .4944845 3.215 0.006 .5292218 2.65035
CtsApp
CrimPrO
Agenda
(2 lags) .3498132 .2823002 1.239 0.236 -.2556606 .955287
Public
Mood
(3 lags) -.2153294 .2049949 -1.050 0.311 -.6549996 .2243409
SupCt
Liberalism
(1 lag) -.1276587 .0816872 -1.563 0.140 —.3028603 .0475429
SupCt
Liberalism
(2 lags) -.1262913 .0771529 -1.637 0.124 -.2917678 .0391851
Political
Liberalism
(2 lags) .1840851 .123375 1.492 0.158 -.0805279 .4486981
CrimPrO
Landmark
Decisions 5.965853 1.680484 3.550 0.003 2.361573 9.570133
CJ
Burger 1.332141 1.586333 0.840 0.415 -2.070206 4.734487
Constant —3.142295 1.30227 -2.413 0.030 -5.935386 -.3492039
250

TABLE 5.6

SUPREME COURT CRIMINAL PROCEDURE (WARREN) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of Obs = 24
--------- +---------------------—-------- F( 9, 14) = 3.75
Model I 270.062259 9 30.006917? Prob > F = 0.0135
Residual I 112.05107 14 8.00364784 R-squared =
--------- +~----------------------------- Adj R-squared =
Total I 382.113329 23 16.613623 Root MSE = 2.8291
Coef. Std. Err. t P>It| [95% Conf. Interval]
SupCt
CrimPrO
Agenda
(1 lag) .6032359 .4514168 1.336 0.203 -.3649569 1.571429
SupCt
CrimPrO
Agenda
(2 lags) 1.628223 .4864374 3.347 0.005 .5849186 2.671528
CtsApp
CrimPro
Agenda
(2 lags) .3533992 .2828096 1.250 0.232 -.2531669 .9599654
Public
Mood
(3 lags) -.2295875 .2139066 -1.073 0.301 -.6883716 .2291967
SupCt
Liberalism
(1 lag) -.124815 .0837463 -1.490 0.158 -.3044329 .054803
SupCt
Liberalism
(2 lags) -.1218237 .0794726 -1.533 0.148 —.2922755 .048628
Political
Liberalism
(2 lags) .1555798 .1284016 1.212 0.246 -.1198143 .4309738
CrimPro
Landmark
Decisions 6.103779 1.782531 3.424 0.004 2.28063 9.926929
CJ
Warren -l.466116 1.854323 -0.791 0.442 —5.443244 2.511011
Constant -1.679075 1.109761 -1.513 0.153 -4.059277 .701126

 

251

 

TABLE 5.7
COURTS OF APPEALS CRIMINAL PROCEDURE (BURGER) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

Source | SS df MS Number of obs = 29
--------- +------------------------------ F( 8, 20) = 5.23
Model I 632.139262 8 79.0174078 Prob > F = 0.0013
Residual I 302.286263 20 15.1143131 R—squared =
--------- +------------------------------ Adj R-squared =
Total I 934.425525 28 33.3723402 Root MSE = 3.8877
Coef. Std. Err. t P>|tl [95% Conf. Interval]
CtsApp
CrimPrO
Agenda
(2 lags) .4182399 .2784668 1.502 0.149 —.1626316 .9991114
SupCt
CrimPrO
Agenda
(2 lags) 1.757767 .5797495 3.032 0.007 .5484304 2.967103
CtsApp
Liberalism
(1 lag) -.4018319 .1745004 -2.303 0.032 -.7658334 -.0378304
SupCt
Liberalism
(1 lag) -.2831633 .1079078 -2.624 0.016 -.508255 -.0580717
SupCt
Liberalism
(2 lags) .1522106 .0973469 1.564 0.134 -.0508516 .3552727
Political
Liberalism
(3 lags) -.2410784 .0926032 -2.603 0.017 -.4342453 -.0479114
CrimPro
Landmark
Decisions .3685859 2.201326 0.167 0.869 -4.2233 4.960472
CJ
Burger -3.631799 2.018419 —1.799 0.087 -7.842148 .5785502
Constant 2.370184 1.693702 1.399 0.177 —1.162816 5.903184
252

 

r

TABLE 5.8
COURTS OF APPEALS CRIMINAL PROCEDURE (WARREN) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of Obs = 29
--------- +------------------------------ F( 8, 20) = 4.32
Model I 591.977419 8 73.9971773 Prob > F = 0.0037
Residual I 342.448106 20 17.1224053 R-squared =
--------- +------------------------------ Adj R-squared =
Total I 934.425525 28 33.3723402 Root MSE = 4.1379
Coef. Std. Err. t P>It| [95% Conf. Interval]
CtsApp
CrimPrO
Agenda
(2 lags) .2893345 .300449 0.963 0.347 -.337391 .9160601
SupCt
CrimPrO
Agenda
(2 lags) 1.372793 .5793089 2.370 0.028 .164376 2.58121
CtsApp
Liberalism
(1 lag) -.4354393 .1860464 -2.340 0.030 -.8235252 -.0473533
SupCt
Liberalism
(1 lag) -.2284206 .1092191 -2.091 0.049 -.4562478 —.0005935
SupCt
Liberalism
(2 lags) .1927757 .1001697 1.924 0.069 -.0161747 .4017261
Political
Liberalism
(3 lags) —.2217234 .097801 -2.267 0.035 -.4257327 -.0177142
CrimPro
Landmark
Decisions 1.129918 2.826274 0.400 0.694 -4.765586 7.025422
CJ
Warren 1.777616 2.483542 0.716 0.482 -3.402962 6.958194
Constant -.6106669 .9835611 -0.621 0.542 -2.662339 1.441006

 

253

 

 

 

 

TABLE 5.9

SUPREME COURT ECONOMIC CASES (BURGER) - 28LS REGRESSION RESULTS

Source

|
+
Model I
Residual I
+

|

1366.65694
143.669745

1510.32669

10 136.665694
10 14.3669745

20 75.5163344

Number of obs
F( 10, 10)
Prob > F
R-squared

Adj R-squared
Root M88

21
9.51
0.0007

3.7904

 

SupCt
Econ
Agenda
(2 lags)
CtsApp
Econ
Agenda
(1 lag)

Coef.

-.3376565

-3.0263l3

Std. Err.

.3940922

.4957621

-0.857

-6.104

P>ltl

0.412

0.000

[95% Conf.

-1.215749

-4.13094

Interval]

.5404358

-1.921686

 

CtsApp
Econ
Agenda
(3 lags)

1.113441

.4722629

2.358

0.040

.0611737

2.165708

 

Public
Mood
(1 lag)

1.00415

.4350803

2.308

0.044

.0347305

1.973569

 

Public
Mood

(2 lags)
SupCt
Liberalism
(1 lag)

1.364231

.4931607

.4654154

.1442493

2.931

3.419

0.015

0.007

.3272205

.1717534

2.401241

.8145681

 

Political
Liberalism
(2 lags)

.4235871

.2059591

2.057

0.067

-.0353184

.8824926

 

Political
Liberalism
(3 lags)

-.6286893

.2954129

-2.128

0.059

-1.28691

.0295317

 

Econ
Landmark
Decisions

26.08482

7.426789

3.512

0.006

9.5369

42.63273

 

CJ
Burger

-45.30227

8.191714

-5.530

0.000

-63.55455

-27.04999

 

Constant

139.116

28.11665

4.948

254

0.001

76.46818

201.7638

 

TABLE 5.10
SUPREME COURT ECONOMIC CASES (WARREN) - ZSLS REGRESSION RESULTS

Number of Obs
Fl 10, 13)
Prob > F
R—squared
Adj R-squared
Root MSE

24
10.23
0.0001

Source

|
4.
Model I 1713.11514 10 171.311514
Residual l 217.738221 13 16.749094
+
l

 

83.9501461

1930.85336 23 4.0926

 

SupCt
Econ
Agenda
(1 lag)
SupCt
Econ
Agenda
(2 lags)

Coef.

-l.997094

-.495326

Std. Err.

.5568904

.4203347

-3.586

-1.178

P>|t|

0.003

0.260

[95% Conf.

-3.200183

-l.403404

Interval]

-.7940053

.4127519

 

CtsApp
Econ

Agenda
(1 lag)

-l.614127

1.029736

-1.568

-3.838737

.6104834

 

CtsApp
Econ
Agenda
(2 lags)
Public
Mood

(3 lags)

2.059957

-.9253076

.4616899

.3588276

4.462

-2.579

0.001

0.023

1.062536

-1.700507

3.057377

-.1501078

 

SupCt
Liberalism
(1 lag)

.4826201

.1090519

4.426

0.001

.2470278

.7182123

 

SupCt
Liberalism
(3 lags)
Political
Liberalism
(2 lags)
Econ
Landmark
Decisions
CJ

Warren

Constant

-.2349683

.1186633

2.571399

20.25499

48.67144

.1060733

.323551

9.89631

24.61597

57.36921

-2.215

0.367

0.260
0.823

0.848

255

0.045

0.720

0.799

0.425

0.412

-.4641258

-.5803262

-l8.80828

-32.92457

-75.26721

-.0058108

.8176528

23.95108

73.43456

172.6101

 

h...

 

TABLE 5.11
COURTS OF APPEALS ECONOMIC CASES (BURGER) - ZSLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of obs = 29
--------- +------------------------------ F( 8, 20) = 4.68
Model I 656.083021 8 82.0103776 Prob > F = 0.0024
Residual I 350.784595 20 17.539229? R—squared =
--------- +---—--—----------------------- Adj R-squared =
Total | 1006.86762 28 35.9595577 Root MSE = 4.188
Coef. Std. Err. t P>|t| [95% Conf. Interval]
CtsApp
Econ
Agenda
(1 lag) —1.269145 .3653715 -3.474 0.002 —2.031296 -.506993
CtsApp
Econ
Agenda
(2 lags) 1.853498 .5866641 3.159 0.005 .6297379 3.077258
SupCt
Econ
Agenda
(1 lag) -.5096901 .2685029 -1.898 0.072 -1.069777 .0503972
CtsApp
Liberalism
(1 lag) .5280064 .2455275 2.150 0.044 .0158449 1.040168
CtsApp
Liberalism
(2 lag) .4332623 .2356771 1.838 0.081 -.0583516 .9248762
SupCt
Liberalism
(1 lag) .4457493 .1080841 4.124 0.001 .2202898 .6712088
Political
Liberalism
(3 lags) .1532316 .1012338 1.514 0.146 -.0579384 .3644016
CJ
Burger 7.874826 6.157512 1.279 0.216 —4.969519 20.71917
Constant -21.10224 25.26749 -0.835 0.413 -73.8093 31.60482
256

TABLE 5.12
COURTS OF APPEALS ECONOMIC CASES (WARREN) - 2SLS REGRESSION RESULTS

 

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of Obs = 29
--------- +---——------------------------- F( 9, 19) = 4.49
Model I 684.897746 9 76.0997496 Prob > F = 0.0028
Residual I 321.969869 19 16.9457826 R—squared =
--------- +------------------------------ Adj R-squared =
Total | 1006.86762 28 35.9595577 Root MSE = 4.1165
Coef. Std. Err. t P>|tI [95% Conf. Interval]
CtsApp
Econ
Agenda
(1 lag) -.8994285 .4069391 -2.210 0.040 —1.751162 —.0476952
CtsApp
Econ
Agenda
(2 lags) .9911143 .5179674 1.913 0.071 -.0930041 2.075233
SupCt
Econ
Agenda
(2 lags) .2725736 .3133541 0.870 0.395 -.3832841 .9284313
CtsApp
Liberalism
(1 lag) .3728426 .2605733 1.431 0.169 -.1725436 .9182287
CtsApp
Liberalism
(2 lags) .3083945 .2277748 1.354 0.192 -.1683437 .7851327
SupCt
Liberalism
(1 lag) .3389151 .1091471 3.105 0.006 .1104676 .5673626
SupCt
Liberalism
(2 lags) -.1407538 .1029997 -1.367 0.188 -.3563348 .0748271
Political
Liberalism
(3 lags) .1522349 .0995358 1.529 0.143 -.0560959 .3605658
CJ
Warren -9.189894 5.18704 -1.772 0.092 -20.04649 1.666705
Constant -11.21169 14.14189 -0.793 0.438 -40.811 18.38762

 

257

 

CONCLUSION

The analysis in this study examined a general theory encompassing
neoinstitutionalism. That is, it was argued that the behavior of political actors in the federal
courts is driven by a rational desire to pursue their goals, but that such aspirations and efforts
are influenced by the institutions in which their decisions are made. In particular, it was
theorized that the Supreme Court justices have policy goals which they would like to pursue,
and they strive to implement those policies into society, in part by setting the agenda of the
Supreme Court. Indeed, as Schattschneider (1960, 68) plainly asserted: "the definition of the
alternatives is the supreme instrument of power." However, the individual justices cannot
always attain their preferred policies with respect to the Supreme Court’s agenda, Since their
behavior is shaped by institutional rules and fimctions, including the fact that no case can
come before the high Court unless a petitioner files a writ of certiorari or appeal and at least
four justices agree to hear the case. This internal constraint is fashioned further by the fact
that the agenda function of the federal courts exists within a system, such that appeals in the
Courts Of Appeals at times impel the justices to choose cases it might otherwise prefer to
disregard. That is, not only is the behavior of the Supreme Court forged by an endogenous
process fiom inside its own walls, but additionally endogeneity exists from the outside as
well, as lower court appeals by litigants influence some of the aggregate agenda decisions
made by the Supreme Court.

258

 

 

 

 

If the justices of the Supreme Court potentially are influenced by factors both within
and without the institution, surely the appellants in the appeals courts Similarly will be
constrained by institutional rules and actions under which they must act. Specifically, in an
effort to increase its probability of success on the merits, rational litigants, their legal
representatives, and interest groups, which are the political actors setting the agenda in the
appeals courts, will be influenced by factors inside and out of the judicial system as they seek
to pursue their own goals within the judiciary. One of those influences is the agenda of the
Supreme Court, which signals litigants that the federal courts are becoming more supportive
of certain types of cases while shunning others. Recognizing this trend set by the Supreme
Court’s changing agenda, parties act accordingly with respect to their appeals.

This, then, is the general framework under which the this study proceeded. In chapter
3, the initial empirical chapter, bivariate VAR analysis was utilized to test whether the
agendas of these courts influenced agenda decisions over time. The findings there illustrated
that perhaps the most important factor in setting the agenda of one of these courts was the
past agenda history of that court, irrespective of the issue area involved. Moving beyond this
feature, however, the analysis also exhibited that the Supreme Court has a strong influence
on lower court appeals. Thus, the power of the justices’ decisions on the Court’s agenda
reach beyond their own courtroom. Nevertheless, for ceratin issues, particularly those which
the Supreme Court either would rather avoid or was not aware of, appeals in the circuit
courts have a decided influence on building the agenda of the high Court. While this

persuasive ability of the lower court appeals was not as compelling as that garnered by the

259

Supreme Court in its ability to affect litigants’ decisions to appeal, nonetheless the influence
deriving from the lower courts was far from trivial on some issues before the Court.

The succeeding chapter 4 then disaggregated the civil liberties category to see if the
representative issues were affected dissirnilarly on agenda decisions, and to ascertain whether
the general theory espoused here would apply within this classification of issues as well. In
particular, the bivariate VAR analysis demonstrated that, while the justices were not swayed
by lower court appeals in cases defined by this salient issue in the aggregate, not all of its
fragmented parts were so affected. Indeed, while similar results obtained for the aggregate
civil liberties class as for most of the issues of which it is comprised, the issue of privacy is
one in which the high Court is influenced by appeals in the lower courts. It appears that the
justices View the controversial privacy cases, which include the abortion cases, in a similar
manner in which they view economics cases, as they seek to avoid them and accept only
these cases on cert. when compelled to do so; in particular, such compulsion seems to stem
from the agenda of the Courts of Appeals.

The final empirical chapter employed multivariate analysis in an exploration of a
variety of influences on judicial agendas. This examination illustrated that the respective
agendas of the federal courts are influenced by factors in addition to those examined in the
prior chapters. Nevertheless, agenda trends over time were found to be influential at both
levels of the federal judiciary, depending upon the issue. Moreover, the factors which
influence the Supreme Court, for the most part at least, were very different from those from

which litigants take their Signals. Yet again, the results from this chapter were consistent

260

 

 

 

with my general theory that the federal courts act within a system on agenda decisions, as
each court influences the other.

Several factors outside of agendas were found to be important influences on agendas
as well. For instance, the actions of the political branches were much more portentous to,
and influential on, litigants in the Courts of Appeals, while actions taking place on Capitol
Hill and the other end of Pennsylvania Avenue were not at all salient with respect to the
Supreme Court’s agenda. Furthermore, the Warren Court’s actions and influence were very
different than those of the Burger Court. Finally, one important result is that the decisional
trends of the Supreme Court were not nearly so influential as hypothesized. Indeed,
appellants’ decisions to appeal were influenced more so by the Supreme Court’s trends in
agenda than with the dynamic liberalism stemming from above.

Accordingly, the agenda-building process for each of these federal courts is not as
Simple as determining whether four justices agreed to hear a case, or whether a litigant paid
the proper filing fee to the clerk of the circuit court. To the contrary, various individual
decisions made by the relevant political actors are influenced by a multitude of factors, and
the aggregation of those decisions are affected as well by some of those same factors and by
others. The repercussion is that there are multiple considerations for setting the agendas in
the federal courts. Perhaps, then, the most important message to be derived from this study
is that, for agenda purposes at least, the federal judiciary in fact "articulates as a system"
(Frankfurter and Landis 1927, 3). AS a consequence, future judicial studies which seek to
ascertain the influences on agenda of either of these federal courts necessarily Should
incorporate the agenda influences of the other court within the federal system.

261

APPENDICES

 

APPENDIX A

CODING DECISION RULES

Issue Areas

To generate the data for this study delineating the various issue areas, a number of
coding decision rules needed to be incorporated. In particular, while the Supreme Court
Database and Courts of Appeals Database are similar in many regards, there are some
particular issues in those Databases which are not coded identically (Songer 1997).

Accordingly, to ensure the issues are coded analogously, I operationalized the data as
described below.

First, for purposes of this study all issue areas as provided in the Supreme Court
Database remained as a given. Thus, the following values are delineated as issues on the

Supreme Court’ 5 agenda, in accordance with the codebook for the Supreme Court Database:

Criminal Procedure: if Value = 1;

Civil Rights: if Value = 2;
First Amendment: if Value = 3;
Due Process: if Value = 4;
Privacy: if Value = 5;
Attorneys: if Value = 6;
Union/Labor: if Value = 7;
Economic: if Value = 8;
Federalism: if Value = 10;

Interstate Relations: if Value = 11;
Federal Taxation: if Value = 12;
Miscellaneous: if Value = 13.

Note: Judicial power cases (Value = 9) are dropped from the

analysis, since there is no analogous issue coding in the
Courts of Appeals Database.

263

Next, data from the Courts of Appeals Database needed to be recoded in order to
conform to the decision rules in the Supreme Court Database. For example, in the Courts
of Appeals Database, General Category 2, which includes specific case type codes 201-299,
defines civil rights issues. However, not all of these specific case type codes within General
Category 2 would be coded as civil rights in the Supreme Court Database. For instance, case
type 203 is "due process rights in prison." This is coded as a civil rights case in the Courts
of Appeals Database, whereas the Supreme Court Database would code this as a due process
case.

After consultation with Harold Spaeth, I made the following transformations to the
data for the agenda on the Courts of Appeals, in accordance with the codebook for the Courts
of Appeals Database:

Criminal Procedure: if case type = 101-158, 201-202,206-208, 773;

Civil Rights: if case type = 204, 209-299, 904, 912-914, 921;
First Amendment: if case type = 301-399;

Due Process: if case type = 203, 205, 410-412, 499, 505-507, 774;
Privacy: if case type = 413, 501-504, 599;

Attorneys: if case type = 903;

Union/Labor: if case type = 601-610;

Economic: if case type = 701, 710-772, 799;

Federalism: if case type = 902;

Interstate Relations: if case type = 901;
Federal Taxation: if case type = 702-706;
Miscellaneous: if case type = 910-911, 915-916, 999.

For purposes of Chapters 3 and 5, civil liberties includes the aggregate of the
following categories: civil rights; first amendment; due process; and, privacy. In Chapter 4,
these issues are disaggregated into their separate parts.

For purposes of Chapter 5, economic cases include the aggregate of economic,
federal taxation, and union/labor cases. In Chapter 3, these issues are disaggregated into
their separate parts.

Specific Database Considerations

Supreme Court Database. For the Supreme Court data, I included only orally argued
decisions which were formally decided, even if by an equally-divided court, while I excluded
memorandum decisions, per curiam decisions, and decrees. Accordingly, this included
DEC_TYPES 1, 5, 6, and 7. Moreover, the unit of analysis in this study is the case, as
demarcated either by its citation or docket number; this included AN ALU = ‘ ’ or AN ALU

=61’.

264

Courts of Appeals Database. The Courts of Appeals Database utilizes a Specific sample
across circuits and years. Thus, this data were weighted in order to account for variations
in the dockets of the several circuits. The specifics for weighting the data are provided in the
codebook for the Courts of Appeals Database.

According to the codebook for the Courts of Appeals Database (1997, 42): "Up to
two case types . . . are coded for each case, though the majority Of cases have only one case
type. NO decision was made in coding about which issue was the most important when two
or more case types were present. Therefore [sic], CASETYPl should not be considered
more important then CASETYP2." For this reason, in this study I added CASETYPI and
CASETYP2 together for each issue area, in order to account for all issues in the Courts of
Appeals. AS a consequence, some double counting of issues was inevitable.

265

APPENDIX B

LANDMARK DECISIONS

Landmark Civil Liberties Decisions:

Shelley v. Kraemer, 334 US. 1 (1948).

Brown v. Board of Education, 347 US. 483 (1954).
Roth v. United States, 354 US. 476 (1957).
Monroe v. Pape, 365 US. 176 (1960).

Baker v. Carr, 369 US. 186 (1962).

New York Times v. Sullivan, 376 US. 254 (1964).
Griswold v. Connecticut, 381 US. 479 (1965).
Levy v. Louisiana, 391 US. 68 (1968).

Lemon v. Kurtzman, 403 US. 602 (1971).

Roe v. Wade, 410 US. 113 (1973).

Miller v. California, 413 US. 15 (1973).

Landmark Criminal Procedure Decisions:
Adamson v. California, 332 US. 146 (1947).
Mapp v. Ohio, 367 US. 643 (1961).

Gideon v. Wainwright, 372 US. 335 (1963).
Miranda v. Arizona, 384 US. 436 (1966).
Camera v. Municipal Court, 387 US. 523 (1967).
See v. Seattle, 387 US. 541 (1967).

Terry v. Ohio, 392 US. l (1968).

Benton v. Maryland, 395 US. 784 (1969).

Landmark Economics Decisions:

Feres v. United States, 340 US. 135 (1950).

National League of Cities v. Usery, 426 US. 833 (1976).
Garcia v. San Antonio Metropolitan Transit Authority, 469 US. 528 (1985).

Source: Pacelle (1991, Appendix 4).

266

LIST OF REFERENCES

LIST OF REFERENCES

Adamany, David. 1991. "The Supreme Court." In Gates, John B. and Charles A. Johnson,
eds. The American Courts: A Critical Assessment. 5-33. DC: Congressional
Quarterly Press.

Aldrich, John A. 1994. "Rational Choice Theory and the Study of American Politics." In
Dodd, Lawrence C. and Calvin Jillson, eds. The Dynamics of American Politics:
Approaches and Interpretations. Boulder: Westview Press: 208-233.

Aldrich, John A. 1995. Why Parties? Chicago: University of Chicago Press.

Atkins, Burton and Henry Glick. 1976. "Environmental and Structural Variables as
Determinants of Issues in State Courts of Last Resort." American Journal of Pol itical
Science. 20(February): 391-425.

Banerjee, Anindya, Juan Dolado, John W. Galbraith, and David F. Hendry. 1993. C0-
Integration, Error-Correction, and the Econometric Analysis of Non-Stationary
Data. Oxford: Oxford University Press.

Barclay, Scott W. and Thomas A. Birkland. 1997. "Law, Policy Making and the Policy
Process: Closing the Gaps." Presented at the Annual Meeting of the Midwest
Political Science Association, Chicago, Illinois.

Barker, Lucious. 1967. "Third Parties in Litigation." The Journal of Politics. 29:41-

Baum, Lawrence. 1977. "Judicial Impact as a Form of Policy Implementation." In Gardiner,
John A. ed. Public Law and Public Policy. New York: Praeger.

Baum, Lawrence. 1978. "Lower Court Response to Supreme Court Decisions: Reconsidering
a Negative Picture." Justice System Journal 3:208-219.

Baum, Lawrence. 1979. "Judicial Demand—Screening and Decisions on the Merits: A
Second Look." American Politics Quarterly. 7: 109-119.

Baum, Lawrence. 1994. "Specialization and Authority Acceptance: The Supreme Court and
Lower Federal Courts." Political Research Quarterly 47:693-704.

268

Baum, Lawrence. 1995. The Supreme Court (F ifih Edition). DC: Congressional Quarterly
Press.

Baum, Lawrence. 1998. The Supreme Court (Sixth Edition). DC: Congressional Quarterly
Press.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American
Politics. Chicago: University of Chicago Press.

Beaney. 1962. "The Constitutional Right to Privacy." Supreme Court Review. 212.

Boucher, Robert L. and Jeffrey A. Segal. 1995. "Supreme Court Justices as Strategic
Decision Makers: Aggressive Grants and Defensive Denials on the Vinson Court."
The Journal of Politics. 57:824-837

Box, G.E.P. and GM. Jenkins. 1976. Time Series Analysis: Forecasting and Control
(Revised Edition). San Francisco: Holden-Day.

Box, G.E.P. and G.C. Tiao. 1975. "Intervention Analysis with Application to Economic and

Environmental Problems." Journal of the American Statistical Association. 70: 70-
79.

Brace, Paul and Melinda Gann Hall. 1993. "Integrated Models of Judicial Dissent." The
Journal of Politics. 55: 914-31.

Brown, A. Lee. 1981. Rules and Conflict: An Introduction To Political Life and Its Study.
Englewood Cliffs, NJ: Prentice-Hall.

Brenner, Saul, Timothy M. Hagle, and Harold J. Spaeth. 1989. "Reasons for the Defection
of the Marginal Justice in Minimum Winning Coalitions on the Warren Court."
Western Political Quarterly. 42: 409-425.

Brenner, Saul and John F. Krol. 1989. "Strategies in Certiorari Voting in the United States
Supreme Court." The Journal of Politics. 51(November): 828-840.

Brenner, Saul and Harold J. Spaeth. 1988. "Majority Opinion Assignments and the
Maintenance of the Original Coalition on the Warren Court." American Journal of

Political Science. 32 (February): 72-81.

Caldeira, Gregory A. and James L. Gibson. 1992. "The Etiology of Public Support for the
Supreme Court." American Journal of Political Science. 36:635-664.

269

Caldeira, Gregory A. and John R. Wright. 1988. "Organized Interests and Agenda Setting
in the US. Supreme Court." American Political Science Review. 82:1109-1127.

Caldeira, Gregory A. and John R. Wright. 1990. "The Discuss List: Agenda Building in the
Supreme Court." Law & Society Review. 24: 807-836.

Caldeira, Gregory A., John R Wright, and Christopher J .W. Zorn. 1996. "Strategic Voting
and Gatekeeping in the Supreme Court." Paper presented at the annual meeting of the
American Political Science Association, San Francisco, California.

Campbell, Donald T. and Julian C. Stanley. 1963. Experimental and Quasi-Experimental
Designs for Research Boston: Houghton Mifflin Company.

Carp, Robert A. and Ronald Stidham. 1991. The Federal Courts (Second Edition). DC:
Congressional Quarterly Press.

Carp, Robert A. and Ronald Stidham. 1996. Judicial Process in America (Third Edition).
DC: Congressional Quarterly Press.

Cook, Thomas D. and Donald T. Campbell. 1979. Quasi-Experimentation: Design &
Analysis Issues for Field Settings. Boston: Houghton Mifflin Company.

Dahl, Robert A. 1956. A Preface To Democratic Theory. Chicago: University of Chicago
Press.

DeBoef, Suzanna and Jim Granato. 1977. "Near-integrated Data and the Analysis of Political
Relationships." American Journal of Pol itical Science. 41 : 619-640.

Dickey, DA. and WA. Fuller. 1979. "Distributions of the Estimators for Autoregressive
Time Series with a Unit Root." Journal of the American Statistical Association. 74:
427-431.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper Collins
Publishers, Inc.

Engle, Robert F. and C.W.J. Granger. 1987. "Co-Integration and Error Correction:
Representation, Estimation, and Testing." Econometrica. 55: 251-276.

Engle, Robert R, David F. Hendry, and Jean-Francois Richard. 1983. "Exogeneity."
Econometrica. 5 1 (March): 277-3 04.

270

Epstein, Lee and Jeffrey A. Segal. 1997. "Assessing Cross-Institutional Constraints on
Supreme Court Agenda Setting." Presented at the Annual Conference on the
Scientific Study of Judicial Politics, Emory University, Atlanta, Georgia.

Fair, Ray C. 1970. "The Estimation of Simultaneous Equation Models with Lagged

Endogenous Variables and First Order Serially Correlated Errors." Econometrica.
38(May): 507-516.

Fine, Toni M. 1997. American Legal Systems: A Resource and Reference Guide. Cincinnati:
Anderson Publishing CO.

Flemming, Roy B. and B. Dan Wood. 1997. "The Public and the Supreme Court: Individual
Justice Responsiveness to American Policy Moods." American Journal of Political
Science. 41 (April): 468-498.

Frankfurter, Felix and James M. Landis. 1927. The Business of the Supreme Court. New
York: Macmillian.

Freeman, John R. 1983. "Granger Causality and the Time Series Analysis of Political
Relationships." American Journal of Political Science. 27(May): 327-358.

Freeman, John R., John T. Williams, and Tse-Min Lin. 1989. "Vector Autoregression and
the Study of Politics." American Journal of Political Science. 33: 842-877.

Gerber, Scott D. and Keeok Park. 1997. "The Quixotic Search for Consensus on the US.
Supreme Court: A Cross-Judicial Empirical Analysis of the Rehnquist Court
Justices." American Political Science Review. 91(June): 390-408.

Granato, Jim. 1991. "An Agenda for Econometric Model Building." Political Analysis. 3:
123-154.

Granato, Jim and George Krause. 1998. "Information Diffusion within the Electorate: The
Asymmetric Transmission ofPolitical-Economic Information." Political Institutions
and Public Choice Program Working Paper 98-01, Michigan State University, East
Lansing, MI.

Granger, C.W.J. 1969. "Investigating Causal Relations by Econometric Models and Cross-
Spectral Methods." Econometrica. 37: 424-438.

Granger, C.W.J. 1981. "Some Properties of Time Series Data and their Use in Econometric
Model Specification." Journal of Econometrics. 1 6: 121 -1 30.

271

Granger, C.W.J. 1983. "Cointegrated Variables and Error Correction Models." Discussion
Paper, 83-13a, University Of California, San Diego.

Granger, C.W.J. 1990. Modeling Economic Series. In Granger, ed. Oxford: Clarendon Press.

Granger, C.W.J. and AA. Weiss. 1983. "Time Series Analysis Of Error Correcting Models."
In Karlin, S., T. Amemiya, and LA. Goodman, eds. Studies in Econometrics, Time
Series and Multivariate Statistics. 255-278. New York: Academic Press.

Green, Donald P. and Ian Shapiro. 1994. Pathologies of Rational Choice Theory: A Critique
of Applications in Political Science. New Haven, CT: Yale University Press.

Greene, William H. 1990. Econometric Analysis. New York: MacMillan Publishing
Company.

Grossman, Joel and Austin Sarat. 1975. "Litigation in the Federal Courts: A Comparative
Perspective." Law & Society Review. 9: 321-346.

Gujarati, Damodar N. 1995. Basic Econometrics (Third Edition). New York: McGraw-Hill.

Hackman, Nathan. 1966. "Lobbying the Supreme Court—An Appraisal of ‘Political Science
Folklore’." Fordham Law Review. 35: 15—50.

Hagle, Timothy M. and Harold J. Spaeth. 1992. "The Emergence of a New Ideology: The
Business Decisions of the Burger Court." The Journal of Politics. 54 (February):
120-134.

Hagle, Timothy M. and Harold J. Spaeth. 1992. "Ideological Patterns in the Justices' Voting
in the Burger Court's Business Cases." The Journal of Politics. 55(May): 492-505.

Hall, Melinda Gann and Paul Brace. 1993. "Integrated Models of Judicial Dissent." The
Journal of Politics. 55(November): 914-935.

Hall, Melinda Gann and Paul Brace. 1995. "Studying Courts Comparatively: The View from
the American States." Political Research Quarterly. 48: 5-29.

Hall, Melinda Gann and Paul Brace. State Supreme Courts Judicial Database. Forthcoming.

Hammond, Thomas H. 1986. "Agenda Control, Organizational Structure, and Bureaucratic
Politics." American Journal of Political Science. 30(May): 379-420.

Hammond, Thomas H. and Gary J. Miller. 1987. "The Core of the Constitution." American
Political Science Review. 81 (December): 1155-1174.

272

 

 

Howard, J. Woodford. 1981. Courts of Appeals in the Federal Judicial System. Princeton:
Princeton University Press.

Hurwitz, Mark S. and Malia Reddick. 1996. "Ambivalence at the Courts of Appeals: Does
the Legal or Attitudinal Model Apply?" Paper presented at the annual meeting of the
Midwest Political Science Association, Chicago, Illinois.

Hurwitz, Mark S. and Reginald S. Sheehan. 1998. "Environmental Influences on
Participation in the United States Courts of Appeals." Paper presented at the annual
meeting of the Midwest Political Science Association, Chicago, Illinois.

Johansen, Saren. 1991. "Estimation and Hypothesis Testing of Cointegration Vectors in
Gaussian Vector Autoregression Models." Econometrica. 59(November): 1 55 1 -l 5 80.

Johnson, Charles A., and Bradley C. Canon. 1984. Judicial Policies: Implementation and
Impact. DC: Congressional Quarterly Press.

Johnson, Charles A. 1987. "Law, Politics, and Judicial Decisions Making: Lower Federal
Court Uses of Supreme Court Decisions." Law & Society Review 21 :325-340.

Johnston, Jack and John DiNardo. 1997. Econometric Methods (Fourth Edition). New York:
McGraw-Hill.

Jucewicz, Joseph and Lawrence Baum. 1990. "Workload Influences on Supreme Court
Acceptance Rates, 1975-1984." Western Political Quarterly. : 123-135.

King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing Social Inquiry.
Princeton University Press.

Kingdon, John W. 1984. Agendas, Alternatives, and Public Policies. Harper Collins
Publishers.

Kingdon, John W. 1985. Agendas, Alternatives, and Public Policies (Second Edition).
Harper Collins Publishers.

Kramer, Gerald H. 1983. "The Ecological Fallacy Revisited: Aggregate- versus Individual-
level Findings on Economics and Elections, and Sociotropic Voting." 77(March): 92-
1 1 l.

Krehbiel, Keith. 1991. Information and Legislative Organization. Ann Arbor: University of
Michigan Press.

273

Krol, John F. and Saul Brenner. 1990. "Strategies in Certiorari Voting in the United States
Supreme Court: A Reevaluation." Western Political Quarterly. 43 (June): 335-342.

Lanier, Drew Noble. 1998. "The Times They Are A Changin’: Longitudinal Analyses of
United States Supreme Court Agenda-Setting and Liberalism, 1888-1989." Paper

presented at the annual meeting of the Midwest Political Science Association,
Chicago.

Light, Paul C. 1991. The President ’s Agenda: Domestic Policy Choice from Kennedy to
Reagan (Revised Edition). Baltimore: Johns Hopkins University Press.

MacKinnon, JG. 1990. "Critical Values for Co-Integration Tests." Working Paper:
University of California, San Diego.

Maltzrnan, Forest and Paul J. Wahlbeck. 1996a. "May It Please the Chief? Opinion
Assignments in the Rehnquist Court. American Journal of Political Science.
40(May): 421 -443.

Maltzrnan, Forest and Paul J. Wahlbeck. 1996b. "Strategic Policy Considerations and Voting
Fluidity on the Burger Court." American Political Science Review. 90(September):
581-592.

McGuire, Kevin T. and Gregory A. Caldeira. 1993. "Lawyers, Organized Interests, and the

Law of Obscenity: Agenda Setting in the Supreme Court." American Political
Science Review. 87: 717-726.

McKelvey, Richard D. 1976. "Intransitivities in Multidimensional Voting Models and Some
Implications for Agenda Control." Journal of Economic Theory. 12 (June): 472-482.

McKelvey, Richard D. 1981 . "A Theory of Optimal Agenda Desi gn." Management Science.
27 (March): 303-321.

McCloskey, Robert G. 1994. The American Supreme Court (2d Edition). Chicago: University
Of Chicago Press.

Mills, Terrance C. 1990. Time Series Techniques for Economists. Cambridge University
Press.

Mishler, William and Reginald S. Sheehan. 1993. "The Supreme Court as a
Countennajoritarian Institution? The Impact of Public Opinion on Supreme Court
Decisions." American Political Science Review. 87(March):87-101.

274

 

Mishler, William and Reginald S. Sheehan. 1996. Public Opinion, the Attitudinal Model,
and Supreme Court Decision Making: A Micro-Analytic Perspective. The Journal
of Politics. 5 8: 169-200.

Moe, Terry M. 1985. Control and Feedback in Economic Regulation: The Case of the
NLRB." American Political Science Review. 79(December): 1094-1116.

Murphy, Walter F. 1964. Elements of Judicial Strategy. Chicago: University of Chicago
Press.

Norpoth, Hehnut and Jeffrey A. Segal. 1994. "Popular Influence on Supreme Court
Decisions." American Political Science Review. 88(September): 711-716.

Northwestern University Law Review. 1960. "Griswold, The Right to be Let Alone." 55 Nw.
U. L. Rev. 216.

O’Brien, David M. 1991. Constitutional Law and Politics. New York: Norton Press.

Olson, Susan M. 1984. Clients and Lawyers: Securing the Rights of Disabled Persons.
Westport, CT: Greenwood Press.

Ostrom, Jr, Charles W. 1990. Time Series Analysis: Regression Techniques (Second
Edition). Newbury Park, CA: Sage Publications, Inc.

Pacelle, Jr, Richard L. 1991. The Transformation of the Supreme Court ’s Agenda. Boulder:
Westview Press.

Pacelle, Richard L, Jr., and Lawrence Baum. 1992. "Supreme Court Authority in the
Judiciary." American Politics Quarterly 20: 169-191.

Palmer, Jan. 1982. "An Econometric Analysis of the US. Supreme Court’s Certiorari
Decisions." Public Choice. 39-387-398.

Perry, Jr., H.W. 1991a. Deciding to Decide: Agenda Setting in the United States Supreme
Court. Cambridge: Harvard University Press.

Perry, Jr., H.W. 1991b. "Agenda Setting and Case Selection." In Gates, John B. and Charles
A. Johnson, eds. The American Courts: A Critical Assessment. 35-59. DC:
Congressional Quarterly Press.

Plott, Charles R. 1976. "Axiomatic Social Choice Theory." American Journal of Political
Science. 20(August): 51 1-596.

275

 

Plott, Charles R. and Michael Levine. 1978. "A Model of Agenda Influence on Committee
Decisions." American Economic Review. 68(March): 146-160.

Pritchett, C. Herman. 1948. The Roosevelt Court: A Study in Judicial Politics and Values,
193 7-194 7. New York: The MacMillan Company.

Pritchett, C. Herman. 1954. Civil Liberties and the Vinson Court. Chicago: University of
Chicago Press.

Provine, Doris Marie. 1980. Case Selection in the United States Supreme Court. Chicago:
University of Chicago Press.

Reddick, Anna Malia. 1997. The Applicability of Legal and Attitudinal Models to the
Treatment of Precedent in the Courts of Appeals. Unpublished Ph.D. Dissertation,
Michigan State University.

Riker, William H. 1990. "Political Science and Rational Choice." In Alt, James E. and
Kenneth A. Shepsle, eds. Perspectives on Positive Political Economy. Cambridge:
Cambridge University Press.

Riker, William H. 1993. Agenda Formation. Ann Arbor: The University of Michigan Press.

Robinson, W.S. 1950. "Ecological Correlation and the Behavior of Individuals." American
Sociological Review. 15: 3 5 1-3 57.

Rohde, David W. 1972. "Policy Goals and Opinion Coalitions in the Supreme Court."
Midwest Journal of Political Science. 16: 208-224.

Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University
of Chicago Press.

Rohde, David W. and Kenneth A. Shepsle. 1978. "Thinking About Legislative Reform." In
Rieselbach. Legislative Reform: The Policy Impact. Lexington, MA: Lexington
Books.

Rohde, David W. and Harold J. Spaeth. 1976. Supreme Court Decision Making. San
Francisco: Freeman.

Rowland, OK. and Bridget Jeffrey Todd. 1991. "Where You Stand Depends Upon Who
Sits: Platform Promises and Judicial Gatekeeping in the Federal District Courts." The
Journal of Politics. 53(February): 175-185.

276

 

Schattschneider, BE. 1960. The Semisovereign People: A Realist ’s View of Democracy in
America. New York: Holt, Rinehart, and Winston.

Schlesinger, Joseph A. 1966. Ambition and Politics: Political Careers in the United States.
Chicago: Rand McNally.

Schlesinger, Joseph A. 1991. Political Parties and the Winning of Oflice. Ann Arbor:
University of Michigan Press.

Schubert, Glendon. 1959. "The Certiorari Game. " In Quantitative Analysis of Judicial
Behavior. New York: Free Press.

Schubert, Glendon. 1962. "Policy Without Law: An Extension of the Certiorari Game."
Stanford Law Review. 14: 224-327.

Segal, Jeffrey A. 1988. "Amicus Curiae Briefs by the Solicitor General During the Warren
and Burger Courts." Western Political Quarterly. 41: 135-144.

Segal, Jeffrey A. 1997. "Separation of Powers Games in the Positive Theory of Congress and
Courts." American Political Science Review. 91(March): 28-44.

Segal, Jeffrey A. and Albert D. Cover. 1989. "Ideological Values and the Votes of US.
Supreme Court Justices." American Political Science Review. 83(June):557-565.

Segal, Jeffrey A. and Harold J. Spaeth. 1993. The Supreme Court and the Attitudinal Model.
New York: Cambridge University Press.

Sheehan, Reginald 8., Mark S. Hurwitz, and Malia Reddick. 1996. "A Model of Decision
Making at the United States Courts of Appeals, 1953-1988." Presented at the Annual
Conference on the Scientific Study of Judicial Politics, Washington University, St.
Louis.

Shepsle, Kenneth A. 1978. The Giant Jigsaw Puzzle: Democratic Committee Assignments
in the Modern House. Chicago: University of Chicago Press.

Shepsle, Kenneth A. 1995. "Statistical Political Philosophy and Positive Political Theory."
Critical Review. 9(Winter-Spring): 213-222.

Sims, Christopher. 1980. "Macroeconomics and Reality." Econometrica. 48: 1-48.

Slotnick, Elliot E. 1979. "Who Speaks for the Court? Majority Opinion Assignment from
Taft to Burger." American Journal of Political Science. 23 (February): 60-77.

277

Songer, Donald R. 1979. "Concern for Policy Outputs as a Cue for Supreme Court Decisions
Of Certiorari." The Journal of Politics. 41 : 1 185-1 194.

Songer, Donald R. 1982. "Consensual and Nonconsensual Decisions in Unanimous Opinions
of the United States Courts of Appeals." American Journal of Political Science
26:225-239.

Songer, Donald R. 1987. "The Impact of the Supreme Court on Trends in Economic Policy
Making in the United States Courts of Appeals." The Journal of Politics. 49: 830-
841.

Songer, Donald R. 1991. "The Circuit Courts of Appeals." In Gates, John B. and Charles A.
Johnson, eds. The American Courts: A Critical Assessment. 35-59. DC:
Congressional Quarterly Press.

Songer, Donald R. 1997. "United States Courts of Appeals Data Base, 1925-1988." SES -
8912678.

Songer, Donald R., Charles M. Cameron, and Jeffrey A. Segal. 1995. "An Empirical Test of
the Rational-Actor Theory of Litigation." The Journal of Politics. 57: 119-129.

Songer, Donald R. and Stefanie A. Lindquist. 1996. "Not the Whole Story: The Impact of
Justices’ Values on Supreme Court Decision Making. " American Journal of Political
Science 40: 1049-1063.

Songer, Donald R., Jeffrey A. Segal, and Charles M. Cameron. 1994. "The Hierarchy of
Justice: Testing a Principal-Agent Model of Supreme Court--Circuit Court
Interactions." American Journal of Political Science 3 8:673-696.

Songer, Donald R., and Reginald S. Sheehan. 1990. " Supreme Court Impact on Compliance
and Outcomes: Miranda and New York Times in the United States Courts of
Appeals." The Western Political Quarterly 43:297-316.

Spaeth, Harold J. 1963. "An Analysis of Judicial Attitudes in the Labor Relations Decisions
of the Warren Court." The Journal of Politics. 25(May): 290-311.

Spaeth, Harold J. 1 995. "United States Supreme Court Judicial Database, 1 946- 1 994 Terms."
ICPSR 9422.

Springsteen, Bruce. 1992. "If I Should Fall Behind." Lucky Town. New York: Columbia.

278

 

Stidham, Ronald, Robert A. Carp, Donald R. Songer and Donean Surratt. 1992. "The Impact
of Maj or Structural Reform on Judicial Decisionmaking: A Case Study of the United
States Fifih Circuit." The Western Political Quarterly. 45: 143-52.

Stimson, James A. 1991. Public Opinion in America: Mode, Cycles, and Swings. Boulder,
CO: Westview Press.

Stimson, James A., Michael B. MacKuen, and Robert S. Erikson. 1995. "Dynamic
Representation." American Political Science Review. 89(September): 543-565.

Tanenhaus, Joseph, Marvin Schick, Matthew Muraskin, and Daniel Rosen. 1963. "The
Supreme Court’s Certiorari Jurisdiction: Cue Theory." In Schubert, Glendon, ed.
Judicial Decision Making. New York: Free Press.

The Third Branch. 1988. "Congress Signs Supreme Court Docket Bill." 20(July): 5.

Ulmer, S. Sidney. 1972. "The Decision to Grant Certiorari as an Indicator to Decision ‘On
The Merits’." Polity. 4: 429—447.

Ulmer, S. Sidney. 1983. "Conflict with Supreme Court Precedent and the Granting of
Plenary Review." The Journal of Politics. 45: 474-477.

Ulmer, S. Sidney. 1984. "The Supreme Court’s Certiorari Decisions: Conflict as a Predictive
Variable." American Political Science Review. 78: 901-91 1.

Wasby, Stephen L. 1970. The Impact of the United States Supreme Court: Some
Perspectives. Homewood, IL: Dorsey Press.

Wiener, N. 1956. "The Theory of Prediction." In Beckenback, E.F., ed. Modern Mathematics
for Engineers. 165-190. New York: McGraw-Hill.

Williams. John T. 1995. "Time Series Analysis." Indiana University Notes.

279

 

LIST OF CASES

 

LIST OF CASES

Adamson v. California, 332 US. 146 (1947).
Baker v. Carr, 369 US. 186 (1962).

Benton v. Maryland, 395 US. 784 (1969).

Brown v. Board of Education, 347 US. 483 (1954).
Camera v. Municipal Court, 387 US. 523 (1967).
Carter v. Carter Coal Co., 298 US. 238 (1936).
Feres v. United States, 340 US. 135 (1950).
Garcia v. San Antonio Metropolitan TransitAuthority, 469 US. 528 (1985).
Gideon v. Wainwright, 372 US. 335 (1963).
Griswold v. Connecticut, 381 US. 479 (1965).
Lemon v. Kurtzman, 403 US. 602 (1971).
Levy v. Louisiana, 391 US. 68 (1968).
Lochner v. New York, 198 US. 45 (1905).
Mapp v. Ohio, 367 US. 643 (1961).

Marbury v. Madison, 1 Cranch 37 (1803).

Ex parte McCardle, 7 Wall. 506 (1869).
Miller v. California, 413 US. 15 (1973).
Miranda v. Arizona, 384 US. 436 (1966).

281

Monroe v. Pape, 365 US. 176 (1960).

National League of Cities v. Usery, 426 US. 833 (1976).
New York Times v. Sullivan, 376 US. 254 (1964).

Palko v. Connecticut, 302 US. 319 (1937).

Roe v. Wade, 410 US. 113 (1973).

Roth v. United States, 354 US. 476 (1957).

Schechter Poultry Corp. v. United States, 295 US. 495 (1935).
See v. Seattle, 387 US. 541 (1967).

Shelley v. Kraemer, 334 US. 1 (1948).

Terry v. Ohio, 392 US. l (1968).

United States v. Carolene Products, 304 US. 144 (1938).

Wade v. Mayo, 334 US. 680 (1948).

282

 

I“...1
I I

I III. I

- I

991

Iii/Iii“

r

Ill/Ill

016

I

I

I
9 I’ll
T ”

ill

I!
N (ill

II”

I

31293

Ill/Ill