Widening the Lens on Voter Suppression - From Calculating Lost Votes to Fighting For Effective Voting Rights - UC Berkeley
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Research Brief
July 2018
Widening the Lens
on Voter Suppression
From Calculating Lost Votes to
Fighting For Effective Voting Rights
by Joshua Clark
haasinstitute.berkeley.edu
This report is published by the
Haas Institute for a Fair and Inclusive Society
at UC Berkeley
About the Author Report Citation
Joshua Clark is the Tides Joshua Clark. “Widening the
Senior Fellow at the Haas Lens on Voter Suppression:
Institute for a Fair and Inclusive From Calculating Lost Votes
Society. He holds a Ph.D. in to Fighting for Effective Voting
socio-cultural anthropology Rights.” Haas Institute for a
from the University of California, Fair and Inclusive Society,
Irvine, and an M.A. in Latin University of California,
American Studies from the Berkeley, CA. July 2018. http://
University of Texas at Austin. haasinstitute.berkeley.edu/
postelection2016
Editorial Direction
Romy Justilien Report URL
Ben Malley haasinstitute.berkeley.edu/
Stephen Menendian postelection2016
Copyeditor Contact
Marc Abizeid 460 Stephens Hall
Berkeley, CA 94720-2330
Charts & Graphs Tel 510-642-3326
Melissa Cetlin haasinstitute.berkeley.edu
Joshua Clark
Samir Gambhir
Ilyse MagyContents Introduction 5 Background and Overview 6 New Voting Restrictions: An Exclusionary and Partisan Agenda 7 The Challenges of Knowing Voter Suppression: An Appraisal of Recent Studies 9 The Civis Analytics-Priorities USA study.................................................................................9 The Journal of Politics Debates.............................................................................................. 10 The Milwaukee and Dane County, WI Survey of Non-Voters................................................................................................................. 11 Epilogue to 2017’s Quantitative Studies.............................................................................. 12 Concluding Implications: A More Holistic Agenda for Fighting Voter Suppression 15 Endnotes 19
This brief argues that an affirmative vision of voting rights must recognize factors currently treated as “background conditions” of voter suppression instead as causes on par with suppressive laws themselves. haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 4
Introduction
More now than at any The analysis and approach to voter suppression
contained herein is motivated by a larger commit-
time in two generations, ment to democratic principles and practices, and
the goals of making government more responsive
the right to vote in the and equally accountable to all constituents. The
United States is under suppression of votes is of course not the only
issue within this ambit. It also includes the in-
attack. fluence of money in politics, racial and partisan
gerrymandering, and much more. But due to the
Impediments to the right to vote are being erected immense interest across Haas Institute partners
across the country, and they affect voters uneven- and networks, here we place our focus on re-
ly across lines of race, class, and age. Indeed, search concerning a recent spate of restrictive
there is considerable evidence that this is precise- voting laws.
ly their purpose. Yet even where courts ultimately
Through careful engagement with empirical
confirm improprieties exist, their remedies often
studies of these laws, this brief calls on analysts,
fall short. Courts send laws back to legislators to
advocates, and donors to broaden their framing of
be amended rather than abandoned, changes to
the problem of voter suppression—and their con-
the original texts may not be implemented, and of
ception of the tools for fighting it. The brief argues
course, voters are disenfranchised as these legal
that an affirmative vision of voting rights must rec-
and administrative processes play out. For these
ognize factors currently treated as “background
and other reasons, a voting rights agenda that
conditions” of voter suppression instead as
relies principally on litigation may win some im-
causes on par with suppressive laws them-
portant victories, but will ultimately be incomplete.
selves. Taking this multi-causal view, we can see
The Haas Institute for a Fair and Inclusive Society that the best investments for fighting voter sup-
believes that pressing problems require sound pression may be projects that build capacity and
analysis, innovative strategies, and collaborative, knowledge “on the ground”: comprehensive voter
multi-pronged solutions. Grounded in our vision outreach, education, civic engagement, and com-
of an equitable society in which all can participate munity organizing and power-building. Though
and belong, our work aims not only to remove litigation continues to be necessary, if the impact
barriers that exclude, but also to build capacity, of restrictive voting laws is to be minimized in the
infrastructure, and affirmative structures for inclu- short term and eliminated in the long term, cham-
sion. This brief articulates how that vision applies pions of voting rights must also commit to foster-
to voting rights as they relate to manifold sources ing an inclusive, informed, and resolute electorate.
of voter suppression.
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 5Background and Overview
BETWEEN 2012 AND 2016, a wave of new voting The brief’s subsequent section proceeds to eval-
restrictions were enacted in states across the uate research on restrictive voting laws that has
country. The US Supreme Court paved the way shaped conversations since November 2016.
for many of these laws with its controversial 2013 The studies in question have generated vigorous
ruling on the Voting Rights Act (VRA).1 In a 5-4 debates among scholars, journalists, and other
decision, the Court rendered unenforceable key researchers over how to estimate how many votes
provisions of the VRA that required certain juris- these laws suppress. The brief examines these
dictions to secure prior federal approval—meant to debates—and the methodological dilemmas they
prevent discrimination—for any changes to voting reveal—in some detail, with two goals in mind.
or registration rules. For many states, 2016’s First is to explain why it is so hard to know voter
presidential election was the first contest regulat- suppression—why even after more than a year
ed by new restrictions on how people may regis- of extensive analysis, researchers still cannot say
ter to vote, cast ballots, and prove their identities with certainty how many votes were suppressed
at the polls. Often referred to collectively as “voter in 2016.
suppression laws,” such legislation was passed in But second, and more important, is to point to the
at least 13 states between 2012 and 2016.2 lessons about fighting voter suppression that we
Laws that eliminate same-day voter registration,3 can extract from these challenges to quantifying
or allow voters to access the ballot using a con- it. In brief, I argue that the roles of myriad other
cealed-handgun license but not a student ID,4 factors—misinformation, voter ambivalence, limit-
have always raised red flags with people serious ed civic-institutional infrastructure, and so on—in
about voting rights—and for good reason. But curtailing the exercise of the franchise remind us
since the 2016 elections, interest in voter sup- something important:
pression among the broader public has surged.
With the wealth of data on 2016’s electoral out- Voting restrictions
comes now available, researchers also have richer
empirical bases for moving from analyzing laws’
depend on other
potential to actual effects. structural causes to
In the following section, I begin by reviewing the
case against restrictive voting reforms as it stood
suppress the vote. This
even prior to the latest round of studies. I show that opens up a broader
researchers have long agreed that these reforms
discourage and impede eligible voters from casting view on what it could,
ballots, while providing no real benefit. The impact and should, mean to
on voters is also borne disproportionately by cer-
tain subgroups, and there are clear indications that fight back against voter
this is intentional—driven by the perception that
it will deliver partisan political gains. I argue that
suppression.
these facts alone constitute a strong voting-rights
argument against the restrictive laws—irrespective
of the laws’ aggregate effect on an election.
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 6New Voting Restrictions:
An Exclusionary and Partisan Agenda
MUCH OF THE INTRIGUE around recent studies name “voter suppression laws” because the votes
of voter suppression laws has centered on how they stifle are those of eligible citizens.
many votes a new law prevented from being cast,
Researchers also regularly find that the obstacles
and whether it was enough to swing the election.
voter suppression laws put in place impinge upon
From a voting-rights perspective, these are not
certain voter subgroups more than others. For
the questions that should orient our work. Rather,
example, a law adopted in Texas in 2014 was
when we view the effective exercise of the fran-
shown to have limited the accepted forms of voter
chise as a universal right, we must at a minimum
ID in ways that disproportionately affected Black
oppose all restrictions that place an undue or
and Latino eligible voters relative to whites.7 A
discriminatory burden on any voters—no matter
2017 analysis of Alabama’s voter ID law found
who, or how many, they are.
that the same was true of it.8 In Ohio, a 2014 vot-
This section quickly reviews evidence that recent ing reform jettisoned the state’s “Golden Week”—
voting restrictions enacted across the country in which citizens can register and cast early votes
are unnecessary, discriminatory, and motivated on the same day—despite research finding that
by partisan interests. Knowing this is enough to African Americans utilized the week at far higher
know why voting-rights advocates should oppose rates than whites.9 More broadly, laws mandating
these laws. I stress this at the outset in hopes that that voters present photo IDs that are unexpired
the subsequent section—which is critical of recent or contain current addresses are greater obsta-
models for quantifying voter suppression—will not cles for the poor, the underemployed, and the
be misread as defending the laws. Rather, it is young—all disproportionately people of color.
that opposition to—and investments in counter-
Finally, there is every reason to believe that voter
ing—voter suppression laws should not be contin-
suppression laws are designed with the goal
gent upon whether they change the outcomes of
of securing partisan advantage. Numerous re-
elections. Were we to fall into that framing of the
searchers have noted that the pattern of new
problem, voting rights would already have lost.
voting regulations adopted in state legislatures
Proposals to curtail early voting, require photo demonstrates that Republicans alone champion
identification to vote, or otherwise change how the laws.10 Indeed, 12 of the 13 states with new
voters establish their identities at the polls are voting restrictions in 2016 entered the year with
usually presented as efforts to prevent voter fraud. Republican state government “trifectas”—GOP
The problem with this rationale is that there is no control of both legislative houses and the gover-
evidence that such fraud poses any threat to US norship (see Figure 1). Some of these GOP-led
elections. A White House official’s sworn decla- legislatures moved on the laws only once the
ration in January 2018 that the Trump administra- Supreme Court made the federal government
tion’s voter-fraud commission had no findings to unable to enforce pre-approval regulations on
disclose (upon its disbandment) is only the latest state voting reforms. Next, all efforts to turn up
sign that voting restrictions address no legitimate evidence of widespread voter fraud have failed.
need.5 In fact, there is broad agreement among Meanwhile, research has consistently found that
researchers and legal experts that the phenom- new voting restrictions are more likely to burden
enon these laws ostensibly combat is practically economically or otherwise structurally disadvan-
non-existent.6 New voting restrictions merit the taged groups—groups widely considered to “lean
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 7New or increased restrictive voting laws passed, 2012–2016
NH
WA
MT VT ME
ND
OR
MN MA
ID
SD WI NY RI
MI
WY CT
IA PA
NE
NV OH NJ
UT IL IN
CA WV
CO VA
KS DE
MO KY
NC MD
TN
AZ OK
DC FIGURE 1
NM AR SC
MS
GA New voting restrictions
LA
AL
(2012-2016) and state
government “trifectas.”
TX
FL
All states with new
AK
Republican trifecta state (2016) voting restrictions with
Democratic trifecta state (2016)
the exception of New
Hampshire were under
Non-trifecta state (2016)
HI
New cutbacks on when and how voters can cast ballot s full Republican control
New or stricter law and new cutbacks as of 2016.
Democrat.” Such an alignment of partisan legisla- found that its creators had consulted race-dis-
tive action with foreseeable partisan advantage is aggregated data on who in the state possessed
hard to imagine as purely serendipitous—innocent which forms of ID. According to the court’s deci-
of intent. sion, the eventual Republican law “target[ed] Af-
For the most hardened skeptics of proof of intent, rican Americans with almost surgical precision.”13
where voter suppression is concerned, we can And these are only some recent examples.14
also point to some smoking guns. In Wisconsin, a This is the voting-rights argument against the
sworn statement by a Republican staffer attests recent spate of restrictive voting laws. They serve
that GOP state senators in a closed-door meet- no legitimate purpose, while placing burdens on
ing were eager and “giddy” to pass that state’s eligible voters that are borne by some subgroups
voter ID law to give them a leg up in elections.11 more than others. Opposition to these laws
In North Carolina, e-mail correspondence from the should in no way be a partisan issue. Unfortunate-
state’s Republican Party director reveals that he ly however, they have been made one by a selec-
instructed county elections boards to curtail early tion of politicians who see electoral advantage in
voting to target certain voter subgroups—what he making it harder for some citizens to vote.
himself called “mak[ing] party line changes.”12 It
is also in North Carolina that a strict voter ID law
was struck down by a federal appeals court that
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 8The Challenges of Knowing Voter
Suppression: An Appraisal of
Recent Studies
SINCE THE 2016 presidential election, public Trump won by less than 23,000. Wisconsin
interest in voting misconduct has surged, just as Senator Tammy Baldwin repeated the figures in
researchers receive their largest batch of voting a tweet. Hillary Clinton cites them in a chapter
data since most new voting restrictions took called “Why” in her election memoir.15 Yet the
effect. The result is that several new studies on study’s model was swiftly and decisively rejected
these laws’ impacts have made headlines and by most voting rights scholars and experts—and
captured considerable public attention. with good reason.16
In this section, I review the three most significant of The Civis Analytics-Priorities USA study com-
these, examining both the studies’ findings and the pares statewide turnout figures from the 2012
methodological debates they inspired. The terms and 2016 elections. It finds that, on average,
of these debates circulate far less widely than the states that adopted strict voter ID laws saw a
headlines, but I argue that they have more to teach drop in turnout, while those that made no chang-
us. In particular, the methodological challenge of es to voting laws had increased turnout in 2016.
how to distinguish restrictive voting laws’ effects on Already this is misleading, however, in that half of
turnout from the effects of other factors influenc- the ten states with new strict ID laws saw turnout
ing voting behavior demands that we reconsider go up, while a handful—Mississippi, Wisconsin,
the nature of voter suppression. The reason this and Ohio—accounted for most of the overall drop
challenge is so ubiquitous is that restrictive voting for the entire group. From there, Civis Analytics
laws are usually not singular, determinate causes treats the overall turnout increase of +1.3 percent
of lost votes. They suppress the vote precisely by in “no-change” states as the national norm—the
combining together with other voting deterrents. In amount by which each state should have in-
most cases, it is only cumulatively that these mul- creased its turnout rate in 2016. The study then
tiple causes keep eligible voters from voting. This imputes all change falling short of that level of
is what is most important, I argue, in the studies growth to voter ID laws.
discussed below: They help us to recognize the
The report says that Wisconsin’s turnout rate
co-causes of suppressed votes, and to name them
dropped by 3.3 percent, and that, if it “had in-
as voting-rights issues and necessary targets in the
stead increased by the national no-change aver-
fight against voter suppression.
age, we estimate that over 200,000 more voters
would have voted in Wisconsin in 2016.”17 At
The Civis Analytics- the national level, “If states where voter ID laws
Priorities USA study became stricter between 2012 and 2016 had
The first notable study to evaluate the impact of increased turnout by the same rate as that of
voter suppression laws in 2016 made a major states where there were no voter ID law chang-
splash. The ensuing scrutiny of its methods, how- es, we estimate that over 400,000 more voters…
ever, made plain the pitfalls of analyzing voter would have cast their ballots.”18 Fair enough;
suppression from the proverbial 30,000 feet. these statements are about math. But it is some-
thing different to attribute all divergence from the
Conducted by Civis Analytics for the super
average to voter ID laws. Doing so is particularly
PAC Priorities USA, the study offered a stun-
problematic when half of the states with new
ning headline: Wisconsin’s strict voter ID law
laws actually saw turnout increases.
suppressed 200,000 votes—in a state Donald
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 9A final contentious analytical tack in the Civis our votes don’t really matter, that the candidates Analytics-Priorities USA study is its use of a aren’t so different, or that withholding our sup- comparison between Wisconsin and Minnesota— port is the only way to get their attention.20 They which has no voter ID law—to gauge the impact could also be that new ID laws make voting more of Wisconsin’s law. Here the study looks at coun- trouble than it’s worth or that the people running ty-level turnout figures, with a particular interest in things will just turn away “people like us” anyway. counties with larger African American populations. The collective political knowledge and sense-mak- On one hand, the comparison is compelling, at ing that these conversations develop can be once affirming that drop off in counties with more just as crucial to securing the exercise of voting African Americans took place across the board, rights as any piece of litigation. The critique of but also that it was steeper in Wisconsin than in the Civis Analytics model’s indifference to this—or Minnesota. But on the other, we must wonder any—context surrounding voters reminds us that what the outcome would be were the same anal- community networks and knowledge circuits are ysis run for another neighboring state with no infrastructure that give those rights effect. ID law: Michigan. Scholars who have analyzed voter file data based on official turnout records The Journal of Politics Debates find that African American participation in that Important debates over how to quantify restrictive state declined at essentially the exact same rate voting laws’ impacts also took place in the field as in Wisconsin.19 This by no means implies that of political science in 2017, most notably around African Americans’ votes were not suppressed in a study published in The Journal of Politics. The Wisconsin. But it certainly raises questions about professional researchers involved in this work Civis Analytics-Priorities USA’s model for assess- avoided clear missteps like those in the Civis ing the impact of the voter ID laws. Analytics study, careful to control for outside Broadly speaking, the problem with this model variables as they investigated voter ID laws. Still, is that it neither takes into account, nor controls methodological disagreements persist among for, any of the other myriad factors that influence the researchers’ peers. These disagreements whether voters will or will not go to the polls. It spotlight pending challenges not only to studying can attribute variation only to the factor it is trying voter suppression, but also to combatting it. to investigate—voter ID laws. What the Wiscon- The study appearing in The Journal of Politics sin-Minnesota comparison shows in particular is was carried out by Zoltan Hajnal, Nazita Lajevardi, that the study treats voters as essentially a series and Lindsay Nielson.21 Though these research- of context-less, isolated units; it assumes they ers’ work does not look at the 2016 election, it is should behave in mostly consistent patterns irre- groundbreaking as a peer-reviewed assessment spective of place insofar as they look the same of the impact of the post-2011 generation of strict in crude demographic terms. This is what allows voter ID laws.22 Hajnal et al. use data from the for the conclusion that differences in Black voting large-sample Cooperative Congressional Election patterns in Minneapolis-Saint Paul and Milwaukee Studies (CCES) of the five election years from cannot be explained otherwise than with refer- 2006 to 2014, allowing them to focus on turnout ence to voter suppression laws. effects across different voter subgroups—particu- In bringing this criticism out, we are reminded to larly as defined by race/ethnicity. think of the ways voters are networked in local The Hajnal et al. study provides a strong and ro- communities that influence their behavior. They bust empirical case that recent voting restrictions talk, share information, and influence one anoth- are not only discriminatory in potential, but also er’s opinions, dispositions, and practices. No in actual impact. It begins by comparing turn- matter how ubiquitous social media has become, out rates of racial/ethnic subsets of the CCES we cannot ignore the significance of place in all sample for states with and without voter ID laws of this—that these networks and communication across multiple election cycles. In its straight are seated in distinct localities. This is essentially comparison, the study shows that in general elec- the difference between analyzing voters as mem- tions Hispanic turnout rates are 7.1 percentage bers of populations versus communities. points lower and Asian American rates are 5.4 Distinct civic dispositions form and spread in the points lower in states with strict voter ID laws in context of communities through the dialogue of place. African American turnout is not much differ- their members. These might include ideas that ent in voter ID states during general elections, but haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 10
is 4.6 points lower in primaries. representative at the level of state subsamples.
Hajnal et al. then use regression models to control Another is nonresponse bias: “The kind of per-
for a range of possible confounding (“third”) vari- son who lacks an ID is unlikely to be accurately
ables that might be influencing the apparent rela- represented in the opt-in online CCES study.”27
tionship between turnout in communities of color Neither of these implies that voter ID laws are not
and voter ID laws. But here too they conclude that discriminatory, nor that Hajnal et al. necessarily
these laws have an effect on Hispanics more than exaggerate their impact—only that the scholars do
any other racial or ethnic group. An average His- not have the data to make their precise claims.
panic voter’s likelihood of casting a ballot drops But for those committed to advancing voting
by 10 percentage points if the individual lives in rights, these criticisms too contain a relevant
a state with a strict voter ID law, “all else equal.”23 point. Voter suppression laws pick on voter vulner-
Hajnal et al.’s regression analyses also find that abilities that manifest across spectrums of eco-
such laws decrease African American and Asian nomic, social, and civic life. If these exclusions, or
American voters’ likelihood of voting in primary forms of disconnection, make a voter both unlikely
elections. Because they have no noticeable effect to have a voter ID and unlikely to participate in an
on whites, their overall consequence is to signifi- opt-in survey, they no doubt also make her less
cantly increase the white-nonwhite turnout gap. likely to be networked with organizations and
Despite the numerous factors for which it con- institutions that defend and facilitate the exercise
trols,24 Hajnal et al.’s study quickly drew criticism of voting rights. That is, some of the same factors
for having not considered all possible confound- that foster the suppression of votes also make it
ing variables that could influence turnout. In a harder to detect. Here too, increased capacity for
response piece to be published in The Journal voter-outreach initiatives operating within commu-
of Politics, political scientists Justin Grimmer, nities targeted by voter suppression laws would
Eitan Hersh, Marc Meredith, Jonathan Mummolo, likely weaken their impact, and surely make its
and Clayton Nall make the case that there are mechanisms more knowable—and contestable.28
likely “unobserved baseline differences between
states with and without [voter ID] laws.”25 That
The Milwaukee and Dane County, WI
is, there are other factors that are both causal Survey of Non-Voters
contributors to depressed turnout and correlated Grimmer et al. end their critique of the Hajnal et
with the adoption of voter ID laws. Grimmer et al. study saying that “custom-sampling surveys”
al. ground this claim in the results of a “place- may be a way for future researchers to better
bo test,” through which they find voter turnout gauge the role of ID laws in voter turnout.29 The
effects in the states with ID laws even before final study I review here meets that call. Led by
those laws were in place.26 University of Wisconsin political scientists working
This rebuttal to Hajnal et al.’s claims reinforces with the Dane County (WI) Clerk, it too attract-
a point that should be a lesson for opponents ed considerable attention for its implication that
of voter suppression: Many or most suppressed Wisconsin’s voter ID law likely swung the state
votes cannot have their causes neatly separated, to Trump.30 But it is the study’s recognition of the
but instead lie at the intersections of multiple fac- multiple causation or multiple responsibility for
tors. Notably, Grimmer et al. cannot say what the suppressed votes, and of the prevalence of voter
additional factor (or factors) whose presence they misinformation, that I find most valuable.
identify might be. If the original study’s research- A press release dated September 25, 2017, an-
ers already account for known unknowns, the nounced, “A Survey of registered voters in Dane
critique points to unknown unknowns—all those and Milwaukee Counties who did not vote in the
other other variables. Surely the stories of votes 2016 presidential election found that 11.2% of
not getting cast are diverse; but they must also eligible nonvoting registrants were deterred by
have patterns. More research and organizational the Wisconsin’s [sic] voter ID law.” It goes on to
capacity on the ground would enable us to better say that this “corresponds to 16,801 people in
understand and contend with this multi-causality the two counties,” but that an estimate “as high
of vote suppression. as 23,252” votes would be within the confidence
Grimmer et al. have other criticisms of the Hajnal interval.31 This latter figure grabbed the attention
et al. study, as well. One is that the CCES data of journalists and activists in large part due to how
that Hajnal et al. use is not in fact designed to be close it was to the margin by which Trump carried
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 11Wisconsin (22,748)—and it only covers two of the are included in a total of non-voters “deterred”
state’s counties. from voting by the ID law.34
These findings are the result of a survey mailed The Milwaukee-Dane study finds that almost as
to 2,400 registered Milwaukee and Dane Coun- many non-voters listed the voter ID as a contribut-
ty voters who did not cast ballots in 2016. The ing (non-primary) cause as were identified as hav-
questionnaire focused on why respondents did ing been “prevented” by it from voting. Of course,
not vote, but also asked about their knowledge of we must wonder on what basis voters make these
Wisconsin’s voter ID law. An unfortunate limitation distinctions, and arrive at their rank orderings of
of the survey was the low response rate. Only causes. As political scientist Ryan D. Enos chimed
293 (around 12 percent) of the questionnaires in via Twitter, a person is often unable to name the
were returned—a very small sample. The study’s precise reason she takes (or does not take) an ac-
finding that the “burdens of voter ID fell dispro- tion; requesting an explanation after the fact tends
portionately on low-income and minority popu- to generate post-hoc rationalizations.35 But this
lations”—though consistent with all research on is only one side of the matter, and it is not safe
such laws—was also questioned due to the even to assume—as Enos’s comment might imply—that
smaller size of these subsamples.32 non-voters over-report the role of restrictive voting
The strength of a survey-based study of course is laws. Voters presumably have different ways of
that it allows the researcher to ask the voter her- thinking about how to weigh the relative import of
self why she did not vote. Surveys that pose this suppressive factors, and which is “primary.”36 And
question usually allow respondents to mark more what about those voters who knew they could
than one answer, and many respondents do.33 The not miss work or find childcare that Tuesday, and
Milwaukee-Dane study based its survey question thus never even came to know the law’s additional
on one that has long been used in Census Bureau hurdles, or that their out-of-state ID or proof of
voter studies, asking respondents to mark “yes” or residence would be rejected at the polls?
“no” for each of a series of 12 possible reasons As the Wisconsin survey also shows, so too are
they did not vote. This question is followed by a many voters misinformed about the content of
separate one eliciting “the primary or main rea- voter ID laws. This is consistent with other recent
son” for not voting (see Figure 2). studies, including a similar survey conducted with
FIGURE 2
Reproduction of one
question from the
Milwaukee and Dane
County, WI survey of
non-voters.
As in other similar surveys, many respondents in Texans who did not vote in 2016. There the vast
the Milwaukee-Dane study reported multiple rea- majority of those who named the photo ID require-
sons for not voting. The researchers code as “pre- ment as their main reason for not voting actually
vented from voting” all of those who listed one of possessed an acceptable ID.37 Likewise, the Mil-
the voter ID responses as their primary reason for waukee-Dane study finds that most of those who
not voting, or indicated in a separate question that believed they did not have qualifying ID in fact
they lack qualifying ID. Those who listed voter ID did.38 In both cases, the studies’ authors acknowl-
as among the reasons, but not the primary one, edge that the laws, and legislatures that passed
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 12them, are largely responsible for the confusion. Still, there is room for concern over the selection
But so too do they insist that the confusion—not of cases Berman chose to present in his import-
just the law—must be urgently fought, through ant piece. The problem is that they essentially
voter education, outreach, and empowerment.39 illustrate only one type of pattern by which votes
Even better than the debates arising from the two are suppressed. It is one in which voters are abso-
previous studies, the Milwaukee-Dane study illus- lutely determined to vote, exhaust every effort, and
trates the multiple structural causes that go into still end up being thwarted.
voter suppression. These can be so intertwined This is one story, and it is a powerful one. But
and cumulative in their impact that it is hard even its depiction of voter suppression is much too
for non-voters themselves to pick one “main” narrow. Specifically, it is too narrow in how it rep-
cause. Sometimes they can. But often their choice resents the disenfranchised voter. As a result, it
is based on misinformation. In other cases, choos- misses the opportunity to make a more robust vot-
ing a primary reason is simply arbitrary. What all of ing-rights argument. Such an argument should in-
this tells those who are committed to voting rights clude voter suppression’s multiple causation, and
is that the restrictive laws are only one—and likely in doing so, must include more than only the most
not the easiest—target that could be removed to sympathetic cases of prospective voters. Most
bring about a different result. If our goal is the of those who are systematically and structurally
universal effective exercise of voting rights, we deterred from exercising the franchise are not the
are just as arbitrary when we choose to fight voter grandmother, the veteran, the survivor of a fire, all
suppression by focusing only on the laws while extraordinarily committed to voting. For many, no
ignoring other exclusionary causes. doubt, new voting restrictions are instead a final
straw in a larger structural context that already
Epilogue to 2017’s minimizes incentives and reason for faith in the
Quantitative Studies electoral process. If those voters are not includ-
ed in the analysis, we get an abridged version of
In the final months of 2017, two media events put
what voter suppression is, and a correspondingly
voter suppression in the public spotlight, if only for
abridged understanding of its solutions.43
a while. The first was the release of a cover article
in Mother Jones titled “Rigged.” In it, voting rights The case of the December 2017 special election
expert Ari Berman lays out the full case for saying in Alabama drives my point here home. There,
that voter suppression handed Donald Trump the major news outlets made room for some belat-
state of Wisconsin in the 2016 election.40 The ed discussion of the state’s exclusionary voting
second was coverage of the special election for a laws,44 but it was dwarfed by the sex-crime alle-
US Senate seat in Alabama, a state with notorious gations surrounding GOP candidate Roy Moore.
and endemic voting-rights challenges. When Moore’s opponent, Doug Jones, ultimately
emerged victorious, the storyline all but disap-
Berman’s article draws on both the Civis Analytics
peared.45 A decisive factor in the race was that
and the Milwaukee-Dane studies, but is most no-
African Americans in Alabama not only turned out
table for its long-term, on-the-ground investigation
at a higher rate than whites; they also voted at an
in Wisconsin. It profiles several disenfranchised
unprecedented level for a non-presidential-year
voters and chronicles their exasperating efforts to
election.46 So what does this say about voter
obtain voting credentials and cast ballots in 2016.
suppression in Alabama—a state that bans early
The Washington Post’s Philip Bump was quick to
voting and same-day registration, limits absentee
criticize the piece, among other things calling Ber-
voting, and instituted a voter ID requirement at the
man’s reportage “anecdotal examples of people
same time it closed dozens of offices that issue
prevented from voting.”41 But this characterization
them, disproportionately in counties with large
is patently unfair; any qualitative researcher can
African-American populations?47
recognize that the cases Berman highlights are
only a selection from a larger corpus of findings.42 We can find the answers we need in what little
This is important to note because qualitative work post-election media attention the issue received.
like Berman’s is—for reasons discussed above— LaTosha Brown of Black Voters Matter Fund was
uniquely able to access many of the patterns and a leader in the effort to mobilize Black voters in
processes by which votes get suppressed. It is the Alabama—around, and in spite of, restrictive laws.
knowledge of these—and not the number of sup- Surely the restrictions in many cases worked
pressed votes—that must guide voting-rights work. in tandem with voters already being ambivalent
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 13about whether the election mattered; but home- grown power-building groups like Brown’s tackle both. Of the restrictions, Brown told the New York Times, “I do think that very committed, focused people will find a way [to vote]. But is that fair?”48 Assuming it is not, then neither would it be fair if only those who work unimaginably hard to vote were treated as deserving of voting-rights energy and resources. The equal and effective exercise of voting rights means creating whatever structures, capacity, and infrastructure are necessary to facili- tate the participation of everyone. haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 14
Concluding Implications:
A More Holistic Agenda for
Fighting Voter Suppression
THIS BRIEF HAS proceeded by way of three
t i v e V ot i n g L a
steps. It began by arguing that the case against s t r ic ws
new restrictive voting laws lies not in the number Re
of votes they suppress, but the facts that they
address no legitimate need, impose burdens dis-
Disconnection from
criminatorily, and arise from partisan interests. I Organizations
showed that these facts are well established by
research and reportage including, but also pre- Voter
Ambivalence
dating, the recent round of post-2016 studies.
Next, the brief turned to those studies, review-
ing their findings on voter suppression laws’
impacts, and the debates those findings have
sparked. It highlighted the various methodologi-
cal challenges to isolating restrictive voting laws’ Misinformation
independent suppressive effect—challenges that
arise from the laws’ entanglement with, and reli-
ance upon, other subtler exclusionary structures,
both formal and informal.
Finally, I argued that these challenges—headaches
for quantitative researchers—can and should be
cultivated for their lessons on fighting voter sup-
pression. Those lessons start with a broadening
of the framing of what count as “voting-rights
issues,” by way of the recognition that voter sup-
pression has multiple causes or responsibilities.
This recognition opens up our understanding of
the factors suppressing the vote, bringing into
view as co-causes that which has until now been
considered mere context. From these multiple
causes spring entry points for multiple modes of
intervention—and solutions.
Figure 3 illustrates the proposal. On the top is
a conventional framing of the problem of voter
suppression. In the illustration, the laws are an
encompassing and unbroken barrier—determinate
for an unknown number of voters. It is notable that FIGURE 3
other factors do not go wholly unacknowledged Conventional framing of the problem
here. But they are imagined like clouds or a fog,
of voter suppression (top), and an
traversing the boundary between those who are
and are not affected by the laws, and essentially
alternative proposal (bottom).
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 15incidental to them. If anything, these factors are a lessons about “wins” in the courts: First, they are
nuisance in that they obstruct the view of restric- usually only partial; and second, voting rights can
tive laws’ causal impact, and the people being still “lose on the ground.” As Michael Wines has
constrained by it. reported for the New York Times, the reasons
The bottom graphic on Figure 3 is a depiction include such factors as “foot-dragging by states,
of my proposal for replacing this framing. It rep- confusion among voters, [and] the inability of
resents voter suppression as a net woven of inter- judges to completely roll back bias.”50 When state
locking strands. Not a perfect mesh, the distances administrators fail to comply with or implement the
between parallel strands vary, making for larger decisions of the courts51—or even the letter of the
and smaller openings through which voters might restrictive laws themselves—knowledgeable and
pass or get ensnared. The strands’ intersections strong citizen advocacy is the needed next step
are clearly the most perilous terrain, and yet still to turn back laws’ suppressive effects. Communi-
no part of the net is determinate—one can con- ty-based organizations are the best vehicle for this
ceivably fight through it. But what is more import- type of comprehensive voter education.
ant for our purposes is that a net can be snipped Civic-engagement and outreach infrastruc-
in many places, and from many angles. And, as ture. As my analysis of recent research above
with any net, here each strand is dependent on noted, social, economic, and civic exclusion and
the others for its strength, and thus snipping any disconnection are partners to restrictive laws in
one weakens the entire structure. the work of suppressing votes. One way to sever
At present, most of the energy and resources ex- the ties of these co-causes is by bolstering orga-
pended in the name of tackling voter suppression nizations that conduct outreach with the young,
continues to go toward fighting restrictive laws the poor, and other constituencies with limited
in the courts. This is important—indeed crucial— contact with civic networks and resources. This
work. But it should not overshadow the essential will include many who are mistrustful of institu-
voting-rights role of a variety of other kinds of tions, for whom outreach will no doubt require
initiatives as well. I conclude here with a few that innovative tactics and perhaps new technologies.
stand out based on the previous sections’ analy- The goal however should be to create and sustain
sis. Just as the “strands” in Figure 3 above do not connections and belonging in a community of
represent all of voter suppression’s causes, nor voters. Many organizations already do this kind of
is the following meant to be an exhaustive list of work; yet it has not been widely appreciated as
possible modes of intervention. fundamental to voting rights.
Comprehensive voter education. When sur- Given adequate resources and capacity, these
veys reveal that most of those who did not vote organizations can also serve as community-level
because of an ID law actually possessed accept- sites for receiving and collating reports of mis-
able ID, clearly misinformation is a serious prob- conduct and anomalies at the polls. The need for
lem. Educating voters on new voting laws should such a presence is evident when we consider
be the responsibility of the state governments that an article like Berman’s. Why does it take nearly
pass the laws. Where they show little inclination a year—and a uniquely committed voting-rights
to do so, civic organizations should make de- reporter—to bring forth the stories of voters dis-
mands at both the state and local level for robust enfranchised in a US presidential election? If the
and equitable public education campaigns. problem is that most media are disinclined to
cover voter suppression, what would make such
But the kind of voter education needed to over- negligence impossible? One answer is to build
come voter suppression also goes well beyond local infrastructure for eliciting, aggregating, and
informing citizens about the content of new re- publicizing people’s stories. Crucially, this is but
strictions. First, such initiatives must be targeted another capacity for the same civic-engagement
and tailored to the voter groups that these laws institutions and networks that double as forces
affect. Among other things, this will involve their for preventing voter suppression in the first place.
being delivered in linguistically and culturally ap- Where they are not able to help voters through
propriate ways. Second, for voter education to the “net,” their work to amplify disenfranchised
counter the formidable forces suppressing the voters’ stories seems preferable to coverage of
vote, it must not only convey information but also quantitative studies. This is, again, because the
equip individuals and communities to be effective latter tend to become consumed by technical
civic advocates.49 We have learned two important debates over methodology, and by partisan ones
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 16When we strengthen efforts to activate and empower voters, we fight voter suppression by building knowledge, resilience, and advocacy capacity. Grounded in our commitment to effective voting rights, and in a body of empirical research, the Haas Institute sees the problem of voter suppression as requiring a multi-pronged and holistic approach. haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 17
about whether or not a voting law swung the elec- rest on decisions made at the discretion of city, tion, rather than staying focused on voters’ rights. county, and state officials.54 These may include Voter empowerment and activation. A related matters such as registration deadlines; the timing but distinct factor that works in concert with restric- and scheduling of early voting; the hours, loca- tive voting laws to suppress the vote is voter ambiv- tions, and staffing of polling places; and much alence. By ambivalence, I mean to refer to various more, depending on the state. These matters are levels of doubt about whether one’s vote matters, anything but trivial to voters, and—especially when the legitimacy of the electoral process, and wheth- set in tandem with new voting restrictions—can er any candidate represents one’s interests or have significant effects on who ultimately casts community, among others. Such doubts of course ballots. transcend voter subgroup. But they are a partic- Those who wish to fight voter suppression ularly logical response for those who are rarely should thus support citizen advocacy around the courted by electoral campaigns due to being habit- administration of elections. Here I refer to con- ually characterized as unreliable or “low-propensity” stituent-driven demands aimed below the level of voters. That such characterizations are perceived legislative action—demands whose content will by—and disempowering of—their subjects is borne vary depending on the particular needs of the out by the limited research exploring non-voters’ electorate in a given jurisdiction. Not incidentally, reasons for sitting out the 2016 election.52 The each of the strategic modes of intervention listed effects are strongest on precisely those groups above also builds constituencies’ capacity to for- targeted by restrictive voting laws.53 Nonetheless, mulate, target, and organize campaigns around those shaping mainstream conversations on voter such demands. suppression rarely permit that voter ambivalence Organizations committed to voting rights should be considered within the issue’s ambit. be supported in work to assess local constit- This omission is wrongheaded. The voter who is uents’ voting-access needs—as they relate, for ambivalent is much more likely to be disenfran- example, to differences of age, ability, and identity. chised by new voting laws than is the voter who Subsequent advocacy will then be equipped to is determined and certain her vote matters. Princi- apply pressure and to promote inclusive reforms pled voting-rights advocates should not be in the enabling everyone to reach the universal goal of business of writing off the former—of saying that effective enjoyment of the right to vote.55 she didn’t try hard enough or probably wouldn’t The Haas Institute sees all political power-build- have voted anyway. Of course, to a large extent it ing at the community level, especially with young is the responsibility of campaigns and candidates people and communities of color, as voting-rights to inspire a sense of commitment and determina- work. When we strengthen efforts to activate tion to vote. Still, strong community and political and empower voters, we fight voter suppression organizations on the ground are best positioned by building knowledge, resilience, and advocacy to address voter ambivalence and restrictive laws capacity. Investing in these capacities among in a sustained manner that builds constituency groups disproportionately targeted by restric- power and transcends individual candidates or tive laws also fuels a virtuous pro-voting rights election cycles. cycle: Once empowered, the groups will become Citizen advocacy around election adminis- constituencies capable of holding lawmakers tration. The three strategic entry points above accountable to high standards of inclusiveness are derived directly from my analysis of recent in future laws and policies related to voting and studies of restrictive voting laws and challenges to registration. calculating their impact. As I have explained, the Grounded then in our commitment to the universal debates surrounding those studies draw attention enjoyment of effective voting rights, and in a body to these entry points as promising immediate, of empirical research, the Haas Institute sees the affirmative modes of intervention that are not con- problem of voter suppression as requiring a multi- ventionally understood in voting-rights terms. pronged and holistic approach. By pursuing it, we But there are also several additional institutional will not only be removing barriers—working from a structures that are not captured here, but that defensive or rearguard position. We will be build- work in concert with restrictive voting laws to ing capacity, infrastructure, and structures that dissuade certain constituencies from voting. Many affirm and advance the voting rights of everyone— policies and protocols for administering elections structures for an inclusive electorate. haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 18
Endnotes
1 Shelby County v. Holder, 570 U.S. general election. David A. Graham,
___ (2013). “Ohio’s ‘Golden Week’ of Early Voting
is Dead, Again,” The Atlantic, August
2 ACLU, “Voter Suppression Laws:
23, 2016.
What’s New Since the 2012
Presidential Election.” Georgia and 10 See for example Daniel R. Biggers and
Rhode Island also had restrictive Michael J. Hanmer, “Understanding the
voting laws on the books dating to Adoption of Voter Identification Laws
2011. in the American States,” American
Politics Research 45(4): 560-588
3 Aaron Blake, “North Carolina governor
(2017); and Keith G. Bentele and
signs extensive Voter ID law,” The
Erin E. O’Brien, “Jim Crow 2.0?: Why
Washington Post, August 12, 2013.
States Consider and Adopt Restrictive
4 Rebecca Leber, “In Texas, You Can Voter Access Policies,” Perspectives
Vote With a Concealed Handgun on Politics 11(4): 1088-1116 (2013).
License—but not a Student ID,” New Political scientists William Hicks, Seth
Republic, October 20, 2014. McKee, Mitchell Sellers, and Daniel A.
5 “Second Declaration of Charles C. Smith also present evidence that voter
Herndon,” Dunlap v. Presidential identification laws are significantly
Advisory Commission on Election more likely to be championed in
Integrity, US District Court for the electorally competitive states, further
District of Columbia, Document 39-2 suggesting that their adoption is
(2018). a partisan electoral strategy. See
“A Principle or a Strategy?: Voter
6 See the studies and judicial opinions Identification Laws and Partisan
compiled in the Brennan Center for Competition in the American States,”
Justice’s comprehensive review memo, Political Research Quarterly 68(1):
“Debunking the Voter Fraud Myth,” 18-33 (2015).
January 31, 2017.
11 Ari Berman, “Rigged: How Voter
7 Matt A. Barreto and Gabriel Suppression Threw Wisconsin to
R. Sanchez, “Accepted Photo Trump,” Mother Jones, November/
Identification and Different Subgroups December 2017.
in the Eligible Voter Population,
State of Texas, 2014,” Expert report 12 Colin Campbell, “NC Republican Party
submitted on behalf of plaintiffs in seeks ‘party line changes’ to limit early
Veasey v. Perry, June 2014. This voting hours,” The News & Observer
version of the Texas law was struck (Raleigh, NC), August 17, 2016.
down in part by the Fifth Circuit 13 The court’s decision further argued
Court of Appeals, but is among those that the state’s proffered rationales for
for which an interim amendment the law “cannot and do not conceal the
was ordered rather than wholesale State’s true motivation.” North Carolina
elimination. State Conference of the NAACP
8 NAACP Legal Defense and Education v. McCrory, 831 F. 3d 204 (4th Cir.
Fund, “NAACP Legal Defense Fund 2016).
Files Response to Alabama’s Motion 14 For a discussion of several earlier
for Summary Judgment in Alabama examples, see Aaron Blake,
Photo Voter ID Law Case,” Press “Republicans keep admitting the voter
Release, November 6, 2017. ID helps them win, for some reason,”
9 Darrel Rowland, “Judge rules Ohio The Washington Post, April 7, 2016;
voter rights violated,” The Columbus see also Dara Kam, “Former Florida
Dispatch, May 24, 2016. Here an GOP leaders say voter suppression
initially successful legal challenge was was reason they pushed new election
later overturned in time to ensure there law,” The Palm Beach Post, November
was no “Golden Week” for the 2016 25, 2012.
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 1915 Hillary Rodham Clinton, What other state-level electoral laws that
Happened, Simon & Schuster, 2017. may encourage or discourage voter
See p. 420. participation. Ibid., p. 367.
16 See for example, the reactions 25 Justin Grimmer, Eitan Hersh, Marc
compiled in Rick Hasen, “Some Words Meredith, Jonathan Mummolo,
of Caution About New Study Finding and Clayton Nall, “Comment on
Voter ID Cost Clinton WI and Election,” ‘Voter Identification Laws and the
Election Law Blog, May 9, 2017. Suppression of Minority Votes’,”
forthcoming in The Journal of Politics,
17 Civis Analytics and Priorities USA,
p. 2. The version of this article
“Voter Suppression Analysis,” Press
referenced herein is dated August 16,
Release, May 3, 2017.
2017.
18 Ibid.
26 Ibid., pp. 7-8.
19 Bernard L. Fraga, Sean McElwee,
27 Ibid., p. 3.
Jesse Rhodes, and Brian Schaffner,
“Why did Trump win? More whites 28 Hajnal et al. themselves note this
– and fewer blacks – actually voted,” knowledge gap, saying, “It could
Monkey Cage/The Washington be that more minorities do not have
Post, May 8, 2017. Fraga et al. find the requisite ID, that the costs of
that the voter participation rate for obtaining an ID are too high for
African Americans dropped by 12.4 minorities to bear, that passing these
percentage points in Michigan and laws sends a signal to minorities that
12.3 points in Wisconsin. These they are not wanted at the ballot box,
figures differ from those of the or some combination of the above.
US Census Bureau’s Voting and We simply do not know.” Hajnal et
Registration supplements, which show al., “Voter Identification Laws and the
a smaller drop in African American Suppression of Minority Votes,” p. 377.
turnout in Michigan from 2012 to 29 Grimmer et al., “Comment,”
2016. However, the Census Bureau’s p. 12.
estimates have a number of known
methodological problems, and in any 30 Kenneth R. Mayer and Michael G.
case, official ballot counts for Wayne DeCrescenzo, “Supporting Information:
County (Detroit) and Flint, MI support Estimating the Effect of Voter ID on
Fraga et al.’s findings. Nonvoters in Wisconsin in the 2016
Presidential Election,” September 25,
20 See the findings from focus groups 2017.
with drop-off voters in Milwaukee:
Civic Engagement Fund and Brilliant 31 Kenneth R. Mayer and Scott McDonell,
Corners Research & Strategies, “Voter ID Study Shows Turnout Effects
Breaking Away: Exploring the Third in 2016 Wisconsin Presidential
Party Millennial “Protest” Vote of Election,” Press Release, September
2016, June 2017. 25, 2017.
21 Zoltan Hajnal, Nazita Lajevardi, and 32 The researchers themselves state
Lindsay Nielson, “Voter Identification that their “estimates cannot be
Laws and the Suppression of Minority extrapolated to the state of Wisconsin
Votes,” The Journal of Politics 79(2): as a whole,” making the small sample
363-379 (2017). size somewhat less problematic.
Mayer and DeCrescenzo, “Supporting
22 The authors rightly note that any Information,” p. 2.
inconclusiveness in peer-reviewed
research on voter ID laws comes from 33 For example, the post-2016 election
studies that pre-date this wave of strict wave of Pew Research Center’s
voting restrictions. See ibid. American Trends panel survey
(n=4,183) identified 472 non-voters.
23 Ibid., p. 368. Of these, 1/3 selected more than one
24 Hajnal et al. control for age, education reason for not voting; 1/6 chose at
level, family income, nativity, gender, least three reasons. It is a tiny sample,
marital status, having children, being a but 14 out of the 20 who mentioned
union member, owning a home, being problems with voter ID, mail-in ballots,
unemployed, and religion, as well as or their registration also mentioned
haasinstitute.berkeley.edu Widening the Lens on Voter Suppression 20You can also read
