10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_

Summary
The e::-.:plosion in spending on television attack a dve1tisements in state supreme court elections
accelerated by the Citizens United decision has m ade courts less likely to rule in favor of defendants
in criminal appeals. State supreme court justices, already the targets of sensat ionalist ads labeling
them -soft on crime,- are under increasing p ressure to allow electoral politics to influence their
decisions, even when fundamental rights are at stake.
Citizens United (which removed regulatoty ban;ers to corporat e electioneering) has fundamentally
changed the politics of stat e judicial eledions. Outside interest groups, often with h igh-stakes
economic interests or political causes before the courts, now routinely pour millions of dollars into
st ate supreme court elections. These powerful interests understand the impo11ant role that state
supreme courts play in American go,·ernment, and seek to elect justices who will tule as they prefer
on priority issues such as environmental and consumer protections, marriage equ ality, reproductive
choice and voting rights. Although their economic and political priorities a re not necessarily criminal
justice policy, these sophisticated groups understand that ~soft on crime·· attack ads are often the best
m eans of removing from office justices they oppose.

aa

This study's two principal findings:
The more TV ads air ed dm·ing sta te s upre m e court judic ial ele ctions in a state , the
less likelyjustices are to vote in favor o f cl'iminal defenda n ts. As the number of
airings increases, the marginal effect of an increase in 1V ads grows. In a st ate with 1o,ooo ads,
a doubling of airings is associated on awrage with an 8 percent increase injustices",·oting
against a criminal defendant"s appeal.
Justices in states w hose bans on COI-pora t e and union s p e nding on el ections were
s t ruck d own b y Citizen s United wer e less likely to vot e in favor of criminal
defendants than they w ere before the decision. Citizens United changed campaign
finance most significantly in 23 of the states where there were prohibitions on corporate and
union electioneering prior to the decision. In these states, the removal of those prohibitions
after Citizens United is associated with , on a\•erage, a 7 percent decrease in justices· voting in
favor of criminal defendants.

The study is based on the work of a team of independent researchers from the Emory Uni\·ersity
School of Law. With support from the American Constitution Society. the researchers collected and
coded data from over 3,000 criminal appeals decided in state supreme courts in 32 states and
examined published opinions from 2008 to 2013. State supreme courts a re multi-judge bodies that
decide appeals collectively by majority vote; th e researchers coded individual votes from o\·er 470
justices in these cases. l11ese coded cases were merged with data from the Brennan Center for Justice
reporting the number oflV ads aired during each judicial election from 2008 to 2013. A complete
explanation of this study"s methodology is below.
The findings from this study have several important implicat ions. Not o nly do they confirm the
in fluence of campaign spending on judicial decision making, they also show that this influence
e:\tends to a wide range of cases beyond the prim ary policy interests of the contributors themseh-es.
Even more troubling, the findings reveal that the influence of money has spread from ci\i l cases to
criminal cases, in which the fundamental righ ts of all Americans can be at stake.

Background
Judicial Elections and Campaign Finance
State courts play a vital role in American democracy
State courts handle more than 90 percent of the United States' judicial business. Although ,·astly
m ore attention is paid to the U.S. Supreme Court, it decides fewer than too cases each year ,
compared with owr tOO million cases arising annually in the state courts. State courts handle the
cases that are most likely to directly touch peop)e ·s lives: child custody, divorce, consumer disputes
and criminal prosecutions.
In addition, just as the U.S. Supreme Court decides cases that have important and wide-ranging
public policy implications, so too do th e state su preme courts, deciding cases arising from state laws
and constitutional pro\-isions invohing chi! and human rights, emironm ental protections and the
criminal justice system. State supreme courts decide who can get manied to whom, who can vote,
who can drink clean water and breath clean air, who the police can detain, search and arrest and who
goes to jail and for how long.

"State supreme courts decide who can
get marTied to whom, who can vote,
who can drink clean water and breath
clean air, who the police can detain,
search and arrest, and who goes to
jail and for how long."

State supreme courts play an especially important role with respect to criminal law. Prosecutions in
state courts account for almost 94 percent of felony COil\-ictions, including an overwhelming majority
of those for serious, violent crimes1 . For exampl e, approximately 98 percent of murder cases and 99
percent of rape cases are p rosecuted in the state courts. In deciding these cases. state courts, and
especially supreme com1s, not only try to ensure that the guilty are punished and innocent go free,
but also det ermine the scope of fundamental con stitutional rights for everyone. These criminal cases
raise issues implicating tights such as privacy, freedom from unreasonable search and seizures and
confronting one"s accusers. Eve11·one, not just criminal defendants, has a stake in how these cases are
decided, because a state supreme com1·s decision to limit or narrowly interpret a defendanfs rights
under a state constitution similarly restricts those rights for evel")·one in that state.

Elections play an important role in how state court judges are selected
Given the vital role that state courts play in

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

1/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_
American democracy, the process by wh ich states
select their judges also is ex"tremely important.
Almost 90 percent of state appellate court judges
must regularly be re-elected by voters. Today,
there are four different principal systems of
judicial selection and retention:
•
•
•
•

Fii ure 1: Map of State Court Jud icia l Se lection Met hods

partisan elections
nonpartisan elections
gubernatorial appointment
merit selection plans

In the selection of judges to their highest courts,
9 states use pat1isan elections and 13 states use
nonpartisan elections. 2 In 28 st ates, the
gO\·ernor or legislatu re initially appoints judges
to the highest court, with 21 ofthose states using
some form of merit plan. For the retention of
• Parti~an Election • Nonparti~an Election
judges on the state's highest cou rt, 6 states use
partisan elections and 14 states use nonpat1isan
elections. Eighteen states hold retention elections
to determine whether th ose judges remain in office beyond their initial term, and the incumbent
judges run unopposed and must v..in majority appro,·al for retention. Kine states rely on
reappointment by the governor, legislature or a judicial nominating committee. Only three states
grant their highest court judges permanent tenure.

• Merit Selection

• Other Appointment Sy<;tems

The growing importance of money in judicia l elections
The last 20 years have marked a new era of cont entious politics and ex-ploding spending in the once
sleepy world of judicial elections. Before the 1990s, judicial elections were low-key affairs, attracting
little campaign spending and often less attention from \'Oters. The very few exceptions to this pattern,
including two aggressive campaigns in the 1980s that used the death penalty as a wedge issue to oust
justices in California and Tennessee, were viewed as outliers by most observers.

''Almost 90 percent ofstate appellate
cottrtjudges must regttlarly be reelected by voters."

But beginning in the 1990s, and accelerating in almost every election cycle since, judicial elections
have become more competiti\·e and contentious, and campaign spending on these elections has
s"k·yrocketed. Incumbent judges almost never lost their reelection b ids during the 1980s, but by :woo
their loss rates had risen highe r than those of congressional and state legislative incumbents. 3
The harder-edged, more aggressive campaigns of this new era were fueled by a flood of campaign
contributions. In the 1989- 90 campaign cycle, state supreme court candidates raised less than $6
million, but by the 2007-08 cycle, candidates raised 0\'er $45 fm· their campaigns. 4
Just as notable as the explosion in the amount of spending on state supreme court elections are the
twin transformations in how this money is raised and how it is spent. Increasingly, the money in
judicial elections flows not to the campaigns of the candidates, but rather to ind ependent e:-.-penditure
groups, whicl1 while they have an interest in who wins elections and thus becomes a judge deciding
cases, have no direct connection to the campaigns of the candidates. For example, in the 20 11- 1:: !:
campaign cycle independent expenditures accounted for 43 percent or 824.1 million of the $56-4
million spent in judicial elections during the cycle. 5
A recent spur for this ex-plosive growth in
independent expenditure spending in state
judicial races was the U.S. Supreme court's 2010
decision in Citizens United v. Fede1'{1l Election
Commission. Citizens United was the most
important and publicly controversial campaign
finance case decided by the U.S. Supreme Court
in nearly 40 years. It overruled a half a century's
wm1h of federal law by declaring
unconstitutional federal prohibitions on
corporate electioneering. The Com1"s decision
provoked unprecedented outcry for a campaign
finance case and clearly struck a public netve.
The biggest impact of Citizens United continues
to be the larger deregulation of independent
expenditures by outside groups that it has
ushered in.
Citizens United contributed to dramatic increases
in independent expenditures at the federal level
by outside groups such as Super PACs, 501(c) and
527 organizations. According to the nonpartisan
research organization Open Secrets, outside
spending on independent expenditures and
electioneering communications surged suddenly
to roughly $87 million in the 20 10 federal
elections, the same year Citizens United was
decided. This total represented a nearly sixfold
increase from the pre\ious off-year federal
election in 2006. By the 2012 presidential
election year, outside spending mushroomed
even further to au unprecedented S439 million, a
m ore than fomfold increase o,·er the previous
presidential election year.

Independent expenditures a nd the
new politics of judicial e le ctions

Figure 2a: Non-Cand idate Spend ing Increa ses to mo re
than 42% Since 2001

• Pe-rcentageNon-Candidate S!M'nding

Figure 2b: Non-Candidate Spe nd ing a s a Portion of Tot a l
Spending

• Non·C.andidate

Candidate

After Citizens United, independent expenditures
and electioneering in state judicial elections have
increased just as dramatically. Independent
expenditures in these state judicial races had already accelerated e,·en before Citizens United. For
in stance, whereas only $2.7 million of independent ex-penditures was spent on state supreme court
elections in the 2001-02 election cycle, by the 2007-08 election cycle, over $ 12.8 million was spent.
Available data indicates that this politiciz..1.tion has increased e,·en further since Citizens United; the
2011- 1:::!: election cycle saw over $24 million of independent expend itures. 6

FiEure 3:2012 Total SpendinE Breakdown

The increase in independent expenditures and electioneering by outside groups has only accelerated
in judicial elections since Citizens United. Interest groups spent over $ 15-4 million on state supreme
court races in the 2011- 12 cycle, accounting for more than 27 percent of the total. 7 TI1is spending
represents an increase of over so percent compared to election cycle \\ith the next highest outside
group spending. The majority of money spent by the biggest contributors now goes toward
independent expenditures rather th an candidate contributions; 97 percent of the dollars spent by the
top 10 spenders in 2011- 12 were independent expenditures.
The flood of independent expenditures after Citizens United has not merely made judicial elections
more ex-pensive. It has t ransformed how they are conducted (largely via TV ads), altet·ed their tone
(by making harsh attacks much more common) and changed the substance of the issues addressed
(criminal justice issues, often in the form of -soft on crime~ attacks, are now commonplace). In 2012,
an estimated $33.7 million dollars was spent on n r ads in state supreme com1 elections, v..ith
unprecedented leYels of independent ex-penditures and electioneering by outside groups in
particular. 8 As is often the case, outside groups delivered messages via these ads that were more
harshly ne~ative than those put forth by the candidates and their campai~ns. Forty-four percent of

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

•

Spec ial ln tere~t<iroup~

• Polit icai Parties

• candid ates

2/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_
the ads sponsored by outside groups were attack ads. In contrast, only2 percent of candidate ads and
11 percent of party-sponsored ads were negative in tone. 9

Fieur e 4: Trends in TV Ads

e o o

• Promotto

• contra~!

• Attack

• Promote

31,122 TotalSpots

•

Contra~!

• Attack

• Promote

16,326TotalSpots

• Contra5t

• Attack

l4,762Total Spots

Effects
The Influence of Money in Judicial Elections
Other Empirical Studies
A large body of empirical evidence now demonstrates that money often manages to buy what it wants
in judicial elections. Increases in tele\ision advertising and independent e>.:penditures by outside
groups in particular, raise important concerns about their relationship to judicial decision making
and independence. The authors of this study have written e:d ensively about the worrisome
relationship betw·een campaign contributions and judicial decision making. In a previous study,
Shepherd found that contributions from vm-ious interest groups are associated with increases in the
probability that judges will vote in favor of the litigant s whom those interest groups favor. 10 In
another study, the authors specifically analyzed cont ributions from business groups and found that
campaign contributions from business groups to state sup reme court justices were correlated with
judicial decisions favorable to business interests, at least in states with partisan judicial elections. 11
Shepherd later found that the relationship betw·een business contributions and judges' voting was
st ronger in the period from 2010 to 2012, compared to 1995 to 1998. 1!:' In a separate article, the
authors also found that political party contributions and independent expenditures in suppott of
state supreme court justices were cor related with judicial decisions in favor of the position preferred
by the patty across a \\ide range of legal issues. 13

''A lm·ge body ofempirical evidence
now demonstmtes that money often
manages to buy what it wants in
judicial elections."

Public and Judicial Opinion
Moreover, 76 percent of voters believe that campaign contributions have at least some influence on
judges' decisions and almost 90 percent ofvotet·s believe that l\ith campaign cont ributions, interest
groups are tl'}ing to use the courts to shape policy. l4 Even worse, judges generally agree tha t money
matters in judicial decision making. Forty-six percent of judges believe that campaign contributions
have at least -a little influence- on their decision s, and 56 perc.ent believe "judges should be
prohibited from presiding over and ruling in cases when one of the sides has given money to their
campaign .-15 ~·Ioreowr, So percent of judges believe that with campaign contributions, interest
groups are tl')ing to use the courts to shape policy. 16
One jmist who has taken note of the role political forces have come to play in judicial selection is U.S.
Supreme Comt Justice Sonia Sotomayor. In 1-Voodward v.Alabama, a 2013 case arising from an
Alabama law giving elected trial court judges the power to set asid e sentencing determinations made
by juries, including the imposition of the death p enalty, Justice Sotomayor wrote a powerful dissent
from the Court's decision not to hear the appeal. Citing a study by the Equal Justice Initiative, 17
which found that 92. percent of the sentences overridden by Alabama judges set aside life sentences
in favor of the death penalty and that the proportion of death sentences imposed by judicial override
is elevated in election years, J ustice Sotomayor wrote that the only explanation for the functioning of
the Alabama system "that is supported by empirical elidence is one that, in my liew, casts a cloud of
illE'gitimacy owr the criminal justice system: Alabama judges, who are elected in partisan
proceedings, appear to have succumbed to electoral pressures." 18

How Money in Elections Influences Judicial Decisions: A Note on
Causation
These same concerns regarding rapidly growing levels of money in judicial elections and political
pressure on judges also arise for the recent surge in independent expenditures by outside groups and
the television adve1tising that they fund. The increasing cost of campaigning for state supreme court
might affect the politics of judicial elections and judicial decisionmaking in at least two ob\ious,
impo1tant - and troubling- ways.
First, outside groups can get what they want by pa};ng for television advertising th at helps
sympathetic judges win office and thereby shape the ideological composition of the state judicial)'.
Outside groups can first detennine which candidates are most likely to decide cases as the groups
prefer and which candidates they want to oppose. n1ese outside groups can then fund television
advertising campaigns that help their fa\·ored candidates, and attack their opponents, thus helping
favored candidates win and retain judicial offic,e o\·er candidates less sympathetic to the group's
in terests. In this way, independent e:-..'Penditures and telelision adwrtising help decide jud icial
elections and shape the judicial)' in the direction that outside groups prefer.

"The only answer that is
Stlpported by empirical evidence

is one that, in my view, casts a
cloud of illegitimacy over the
crimina/justice system: Alabama
judges, who m·e elected in
pm·tisan proceedings, appea1· to
have succumbed to electoml
pressures
-Justice SOtomayor

Second, less obviously, judicial candidates may foresee the imp01tance of su ch independent
expenditures and TV attacks ads a nd thus consciously or unconsciously bias their decisions in order
to insulate themselves from such attacks. Judici al candidates may be tempted to lean toward the
preferred positions of wealthy outside groups, either to draw their support or at least avoid their
opposition in subsequent elect ions. What is mo re, judicial candidates may want to do what they can
to decide cases in ways that do not leaw them vulnerable to campaign attacks through negative TV
ads.

Why Independent Expenditure Groups Often Feature Criminal Justice
Issues in Their Ads
This study explores the effect of increasing independent expenditures follO\dng Citizens United on
judicial decision making by examining criminal appeals. Pre\·i ous empirical studies on judicial
selection and criminology establish that criminal cases are particularly effective in n10tivating voters
and more likely to be considered by judges with electoral considerations in mind. 19 Independent
expenditures are more negative overall than advertisements by candidates, who prefer not to -go
negative- if they can avoid doing so. Candidates for judicial office are frequently concerned about
appearing aggressively negative, wishing instead to convey an image of -judicial temperament - in
their campaigns. Thus, the best means of pa};ng for a sensationalist attack advertisement invohing a
violent, bloody fact pattern may be an independ ent expenditure by an outside group not directly
connected to the benefitting candidate.
Because this combination of funding and electioneering techniques is so effective, it should be
particularly worrisome to those concemed abou t the influence that money in judicial elections can

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

3/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_
ha\·e on judicial decision making. Th ese concerns are particularly relevant in appeals arising from
criminal cases, in which the liberty (and in capital cases, the life) of the defendant, as well as th e
constitutional rights of all residents of the state in question, are at stake .
Both judicial ca ndidates and outside groups are well aware of the power of attack advertisement s that
portray judges as "'soft on crime.~ For example, during a 2004 \Vest Virginia Supreme Court election,
an outside group called An d for the Sake of the Kids, which was funded by Massey Coal Company
CEO Don Blanke nsh ip ran an 1V ad alleging that an incumbent justice voted to release a "'child
rapist Mand then -agreed to let th is convicted child rapist work as a janitor in a West Virginia school.··
Similarly, an ad in a 20 12 Louisiana Supreme Court race claimed that a candidate had Msuspended
the sente nce of a cocaine dealer, of a man who killed a stat e trooper, two more drug dealers, and over
half the sentence of a child rapist. MFear-provoking adve11isements such as these, funded by outside
groups ,,.;thout public accountability, can S\\"ing an election and put judges on notice that their
ju dicial careers m ay be at stake each time they consider voting in favor of a defendant in a criminal

Attack: Candidate Bride;et McCormack "foue;ht
to p rotect sexu al p red ato rs"

Attack: Candidate Bill O' Neill " sympathetic to
ra pists"

Attack: "Whe n he was a dist rict attorney,
Incumbent J ustice David Prosser covered up "

State:Michigan
Sponsor: State Repub!lcan Party

State:Ohio
Sponsor. State Republican Party

Sponsor. Greater Wisconsin Committee (progressive)

Results
Skewed Justice: Empirical Analysis
This study find s that increases in television cam paign advertising, often funded by independent
expenditu res, are associated \'>"ithjustices voting against criminal defendants in ways that call into
greater question the fundament al fairness of the criminal justice system.

fl1e more 1V ads aired during state supr eme com1.judicial elections in a state, the
less like ly jus tices are o n a,·erage to vote in fa,·or of crimin al defe ndan t s .
Jus tic es in states w h ose bans on corporate and unio n s p e nding on e lections until
the-y were. struck down by Citizens United were le-ss likely on ave-rage- to vote- in
favor of c1iminal defe ndants than ti1e~· were b e fore ti1e decision.

Data
To e:-.:plore wh ether the increase in TI7 ads has in fluenced state supreme court justices· rulings against
criminal defendants, this study compiled data from several different sources. First, a team of
independent researchers from Emory UniYersity School of Law collected and coded data from almost
3,100 criminal appeals decided in st ate supreme courts. Because justices' votes are likely to become
fodder for future 1V attack ads in only the most heinous cases, the researchers examined only cases
im·ohing certain violent crimes tracked by the FBI's Uniform Crime Reporting Program: mu rder,
robbery, "iolent aggraYated assault, and rape and other sex crimes. The cases were randomly selected
from state supreme court published opinions from zooS to 2013 from 32 states. w The researchers
coded the individual votes from over 470 justices in each of these cases. :n All coding went through a
two-step quality control process; ultimately, over 25 percent of the cases were coded by at least 2
researchers to confirm reliability.
The researchers coded whether the justice, sitting as a member of a multi-judge appellate panel,
voted in favor of the criminal defendant on appeal. Follo\\ing other data on judicial decisions, a \'ote
in fa\·or of a defendant is defined as any vote that improves the defendanfs position-whether it is
O\'erturning any part of a criminal comiction or reducing a defendant's sentence. In addition, the
researchers coded data on the offense for which the defendant was convicted, the number of victims
im·olved in the crime and whether any \"ictims were juveniles. The researchers also collected data on
each stat e's ban on corporate and union indepen dent e!>.."J)enditures p1ior to Citizens United, data on
indh"idual judge characteristics, and data about state judicial selection and ret ention methods.
These d ata were merged \\ith data from the Brennan Center for Justice, -su)ing Time- project. Since
2000, the Brennan Center has collected all available tele\ised state supreme com1 campaign ads that
were aired in states holding supreme court elections. This data on 1V ad aitings are calculated and
prepared by Kantar lldedia/ CMAG, which captures satellite data in the nation's largest media
markets. The authors compiled the Brennan Center's data measuring the number oflV ads aired
during each judicial election from 2008 to 2013.

0

Learn More

To learn more about this report- particularly the
data that it draW!; from- visit the spet:ial reporn
and collaboratJons page of tht' NatJonallnstJtute for
Money in State Politics. The American constitution
Societystronglyencourageinterestedpartiesto
reviewthedata,study t heissuefurther, and
contributetoandexpandthelmportant
conversation about fair courts.
NIMSP istheon1ynonpartisan,nonprofit
organizatlonreveallngthelnlluenceof campalgn
money on state-level elections and public policy in
ati SOstates.The organization encourages
transparency andpromotes "indepefld{'nt
iflvestigationofstate-levelcampaigncontributions

Methodology

byjournalists, academ icres.earcht'~,public­

inlerest groups,governmentagencies,

policymake~.studentsandthepublicatlarge."
The analysis tests wheth er the threat of future TV att ack ads influences justices to cast more '"toughon-crimeM votes. It measures th e th reat of future attack ads in two different ways. First, it uses the
number of televised campaign ads that aired in
the most recent supreme court election in each
state; this measure assumes that recent airings
Figur~ s : States with T~ levised Campaign Ads Concern~d
detern1inejustices" estimates of the likelihood of
With Stat~ Supreme Court Elections 2008- 2013
futu re attack ads. Figure 5 repol1s states in which
1V campaign ads concerned ,...;th state supreme
court elections aired in the years :woS-13. The
awrage number oflV campaign ads aired in
eacl1 state per year was 3,650; the minimum was
30 and the maximum was 17,830.

The study's second measure of the threat of
futu re attack ads is the nonexistence or removal
of a ban on corporate an d union independent
expenditures. Prior to the Citizens United
decision, 23 states had bans on such spending. As
most 1V ads sponsored by interest groups are
funded by independent expenditures, the
availability of such spending can dramatically

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

4/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_
increase the possibility of future 1V attack ads.
Figure 5 identifies which states had a ban on
independent e:-.:penditures from either
corporations or both unions and corporations at
the time of the Citizens United decision. Citizens
United also may have increased independent
expenditures e,·en further by opening the door to
independent expenditure-only committees,
known at the federal level as Super PACs, that
can r~eive uncapped contributions from their
donors. 22

Figure 6: States with Bans on Corporate Independent
Expendit ures Pri or to Citizens United

Control variables in the estimations include
various case, judge and state characteristics that
might also influence judicial voting. Case
characteristics include the crime for which the
defendant was convicted (murder, aggravated
assault, rape, robbery), the number of v;ctims,
and whether any of the ,;ctims were children. It
also includes a measure of the underl};ng
st rength of the case. This control ' 'ariable is
impot1ant because some cases are so strong (or
• C.Orpor.lte/Union S.ln
weak) that justices will ,·ote in fa,·or of or against
the criminal defendant t-egardless of their
ideological predisposition or the influence of 1V
ads. To create a measure of case strength , we estimate how many of the justices hearing a case would
be predicted to vote in favor of the defendant based on certain quantifiable case and state
characteristics. The difference between this predicted number and the actual number of justices
,·oting in favor of the defendant provides a measure of case strength-the more justices voting in
favor of the defendant compared to the predicted number, the stronger the defendant's case.

• Corpor.lte B.an

The study also takes into account th e justices· political party affiliation to control for the role of
ideology on justices· voting in criminal appeals. It includes indicators for retention method for state
supreme court justices to measure whether the method of reel~tion or reappointment affects judicial
vot ing. All estimations also include state and year fixed effects to capture systematic differences in
the criminal appeals process across st ates and general trends in TV ads. The analysis estimates a
series of ordinary probit models with t-statistics computed from standard errors clustered by case.

Results
The More TV Ads, The Fewer Votes in Favor of Defendants
The analyses reveal that TV ads are associated
with justices casting fewer votes in fa,·or of
defendants in criminal appeals. The first analysis
finds that the more 1V ads aired during state
supreme court judicial elections in the state, the
less likely justices are to \'Ot e in favor of criminal
defendants. The results are statistically
significant across n umerous estimations that
alter the specification and include various
combinations of control variables, e nsuring
robustness.
To illustrate the results in an intuitive way,
Figure 7 shows the relationship between the
number of TV ads aired and justices' likelihood of
voting in favor of criminal defendants. The figure
reports, for different numbers of tele,;sed
campaign ads, the expected change in the
justices' probability of voting in fa\·or of the
criminal defendant if ads increased by 100
percent, holding all other variables at their mea n
,·alue. That is, the figure shows that in a state that
aired only 2,000 ads, a doubling of airings would
be expeded to d~reasejustices· mting in favor of
defendants by 2 percent, or change a justice's
vote in 2 percent of cases. However, as the
number of airings increases, the marginal eff~t
of an increase in TV ads grows. In a state \\;th
10,000 ads, a doubling of airings would change a
justice's V'Ote in 8 percent of cases.
The analysis also explores whether the
relationship between televised campaign ads and
ju dges' likelihood of ,·oting in fa,·or of defendants
,.al)' across political party. As a baseline,
Republican justices are, on average, slightly less
likely to vote in fa,'Or of defendants than other
justices; in this sample, Republiean justices voted
in fa,·or of the defendant in 27 percent of cases
but Democratic justices voted in favor of
defendants in 3 1 percent of cases. Howe,·er, the
analysis indicates that TV ads exacerbate this
difference. Figure 4 shows that, although
campaign ads are related to decreases in voting in
favor of defendants across all parties, the
relationship is more pronounced for Republican
justices. Even starting from different baselines,
the relationship between campaign ads and
judicial voting is stronger for Republicans than
either Democrats or Independents.

Figure 7: The Relationship betw een Televised Campaign
Ad s and Voting in Favor of Criminal Defendants

NumM r ofTV Campai1n Ads

-

• PI'fcer~tage

Notes: Tile figure reports the m.lrgln.ll effect on the probability of votir~g in f.Jvor of defend.lnts for .l proportional ch.Jnge in
the number of televi~ c.lmp.lign .ld5, ev.lluated .lt different numbers of TV ad5 and hold ing .lll other v.lri.lbles at t heir me.Jn

Figure a: The Relationship betw een Televi sed Campaign
Ads and Voting i n Favor of Criminal Defendants Across
Political Parties

NumM r ofTV Campai1n Ads

• Independent

• Democratic

• Republlc.ln

Notes: The figure reports the expected probability of voting in f<lVOJ o f defend.lnt s in sute/ years with .lnd without b.lns on
corpor.lte independent expenditu res, holding <lll other ll.lri able5 .lt their me.1n.

Justices Less Likely to Vote in Favor of Defendants After Citizens United
The analysis also explores whether the U.S. Supreme Court's decision in Citizens United had any
impact on justices· votes in criminal appeals. The Court's decision immediately was followed by a
dramatic increase in both the actual number of TV ads aired during judicial elections and the threat
of future TV ads. However, Citizens United changed campaign finance most significantly in the 23
st ates that had bans on corporate or union indep endent e:q>enditures prior to the ruling; 27 states
had no such ban and thus were not as affected by the Citizens United decision. The analysis
empirically e:-.:ploits this variation across states and the resulting differential effect of the decision to
isolate the impact of corporate a nd union independent expenditures in judicial elections on justices·
,·otes in criminal appeals .
The results from this analysis indicate that unlimited corporate and union independent expenditures
are associated \',.;th a d~rease injustices \·oting in favor of defendants. n1e results are statistically
significant across different sp~ifications and with different control variables. Figure 9 shows that
unlimite-d independent spending is associated \\ith, on ave-rage, a seve-n percent decre-ase injustices·
,·oting in favor of criminal defendants. That is, the results predict that, after Citizens United, justices
would ,·ote differently and against criminal defendants in 7 out of 10 0 cases.

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

7%
of cases
''The results predict that, after Citizens
United, justices would vote differently
and against criminal defendants in 7
Ollt of100 cases."

5/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_
Conclusion
In states \\ith more advertising and perhaps more competith·e electoral en\-ironments, elected judges
at·e more likely to be electorally sensitive to bein g seen as wsoft on crimew and therefore less
sympathetic to criminal defendants when they decide criminal appeals. At the margin, whether
consciously or unconsciously, they prefer to avoid a judicial vote in a criminal case that can be the
basis for attack advertisements funded by independent e:\:pe nditures.
Indeed, the analysis set forth abo,·e demonstrates that as television ad,·ertising in a state goes up,
st ate's judges are more likely to decide criminal appeals against criminal defendants. The analysis
also demonst rates that Citizens United exacerbated the influence of money in judicial elections
influence on judicial decision making. In the 23 states that had bans on corporate or union
independent expenditures, Citizens Unitecfs lifting of these bans is associated with a decrease in
justices voting in favor of defendants.
These findings are likely to be only a preview of escalating trends in judicial campaign finance and
elections. There has been only one presidential election cy cle since Citizens United. Outside groups,
whether funded by corporations , unions and wealthy individuals have only begun to professionalize
their operations and \\;11 only grow more sophisticated in the years to come.

Attribution
About the Authors
Joanna Shepherd
J oanna Shepherd teach es Torts, Law and Economics, Analytical Methods for
1...'1\'<·yers, Statistics for Lawyers, and Legal and Economic Issues in Health Policy .
Before joining Emory, Professor Shepherd was an assistant professor of economics at
Clemson Unh·ersity. Much of Professor Shepherd's research focuses on topics in law
and economics, especially on empirical analyses of legal changes and legal
institutions . She has published broadly in lawre\;ews, legal journals, and economics
journals. Professor Shepherd received a BBA from Baylor University and a PhD from
Emory University.

MichaelS. Kang
MichaelS. Kang's research focuses on issues of election law, voting and race,
shareholder voting, and political science. His work has been published by the Yale
Law Journal, NYU Law Re\;ew, and Michigan L."lw Review, among others. Professor
Kang received his BA and JD from the Unh·ersity of Chicago, where he served as
technical editor of the University of Chicago L.1 w Review and graduated Order of the
Coif. He received an MA from the Unh·ersity of Illinois and his PhD in government
from Harvard University. After law school, Professor Kang clerked for Judge Micl1ael
S. Kanne of the US Court of Appeals for the Seventh Circuit and worked in private
practice at Ropes & Gray in Boston before joining the Emory Law faculty in 2004.

About the American Constitution Society

~-

Th e Ame rican Cons titution Socie ty for Law & Poli~· (ACS), founded in 2001
and one of the nation·s leading progressive legal organizations, is a rapidly growing
network of lawyers, law students, scholars, judges, poliC)makers and other concerned
individuals. ACS embraces the progress our nation h as made toward full
embodiment of the Constitution·s core values and believes that law can and should be
a force for improving the lives of all people. ACS is revitalizing and transforming legal
and policy debates in classrooms, courtrooms, legislatures and the media, and we are
building a diverse and dynamic n eh,·ork of progressi\·es committed to justice. By
bringing together powerful, relevant ideas and passionate, t alented people, ACS
makes a difference in the constitutional, legal and public poliC)· debates that shape
our democracy. For mor e infonuation about the organization or to locate one of the
more than 200 lawyer and law student chapters in 48 states, please \;sit acslaw.org.
All expressions of opinion in this report are those of the author. ACS takes no
position on specific legal or poliC)· initiatives.

Endnotes
1 Matthew R. Durose, Donald J. Fa role, Jr. Ph.D. and 5ean P. RosenmeOle, Felony 5entences in State Courts, Table 1.6 (2009), http:r lwww. bjs.gov/inde~.cfm'ty=pbdetail&iid=2152
2 Roy A. SChotland, New Cllollenges to Stores· Judicio/ selection, 9S Geo.L.J.1077, 1085, (2007).
3 Melinda Gann Hall and Chris w. Bonneau, Does Quoliry Matter.1Chollengers in Store supreme coun Elections, 50 Am. J . Pol. S.Ci. 20, 21 {2006).
4 James sample, et al., The New Politics of Judicial ElectKlns, 2000.2009: Decade of Changeat 5 (Char1es Hall ed., 2010).
http://www.brennarn::enter.org/content/resource/the_new_politiu_ofjudicial_elections
5 Alicia B.:mnon, et al., The New Politics of J ud ic ial Elections, 2011-2012 at 5 (laurie Kinney and Peter Hardin ed., 2013).
http:/rwww.brennancenter.org/Sites/default/files/publlcatlons/New%20Pohtlcs%20of%20Jud1Ciai%20Electlons%202012.pdf
6 Ibid.
7 1bid.,3- 4.
8 lbid.,19-20.
9 lbid.,22.
10 Joa nna M. Shepherd, Money, Politics, ond Impartial Justice, 58 Duke L.J. 623, 671)-72& Tables:7-8 (2009).
11 Michael Kang and Joanna M. Shephe rd, The Partisan Price of Justice: An Empirical Analysis of Campaign Contributions and Judicial Decisions, 86 N.Y.U. L. REV. 69 (2011).
12 Joa nna Shepherd, Justice at Risk: An Empirical Analysis of Campaign Cont ributions and Judicial Decisions, Ame rican Constitution So<:iety for Law and Policy (2013).
13 Michael Kang and Joanna M. Shephe rd, The Partisan Foundations of J udicial Campaign Finance, 86 S. CALL. REV. 1239 (2013).
14 Greenberg Quinla n Rosner Research Inc., Justice at Stake-State Judges Frequency Questionnaire 5 (2002), http:r lwww.justiceatstake.org/files/JASJudgesSurvt'yResults.pdf
15 Ibid.
16 Ibid.
17 Equal J ust ice InitiatiVe, The Death Penalty in Alabama: J udge Override, at 14, {2013), http: /eji.org/files/CNerride_Report.pdf
18 Woodward v Alabama, 511 U. S. 405, 408 (2013).
19 5ee Billy Corriher, Center for American Progress, Criminals And Campaign Cash: The Impact Of Judicial Campaign Spending On Criminal Defenda nts, (20 13).
20 Alabama, Alaska, Arizona, AOlansas, California, Colorado, Connecticut, Florida, Georgia, Id aho, Illinois, Iowa, Kansas, Kentucky, Louisiana, Maine, Michigan, Minnesota, Mississippi, Missouri,
Monu na, New Jersey, North Carolina, Ohio, Oklahoma, Oregon, Pe nnsylvania, Tennessee, Texas, Virginia , West Virginia, Wisconsin
21 The resear(hers coded cases from the highe st court of criminal a ppeals in Texas and Oklahoma because the supreme courts in those states do not hear crimina l appeals.
22 However, in other analyses not presented here, the authors found that their gene ral findings remain substantive ly urn::hanged when they cont rolled separately for the changes in state law
regardmg contribution limitS to independent E>xpenditur@-only <Ommittee. They plan to present t hese analyses in forthcoming academic work.

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

6/7

10/22/2014

FireShot Capture - Skewed Justice - http___skewedjustice.org_

Join the Conversation
'I

f

in

Cl 2014 Am~rl{.lln Conslrtuhon SooE"Iy for L oow ;and Polt<y All nghts resef'tled

chrome-extension://mcbpblocgmgfnpjjppndjkmgjaogfceg/fsCaptured.html

7/7