VOTE FORWARD et al v. DEJOY et al

Filing 16

MOTION for Preliminary Injunction by AMY BOLAN, AARON CARREL, COLORADO ORGANIZATION FOR LATINA OPPORTUNITY AND REPRODUCTIVE RIGHTS, INDERBIR SINGH DATTA, DANTE FLORES-DEMARCHI, PAUL HUNTER, SEBASTIAN IMMONEN, KATHRYN MONTGOMERY, SEAN MORRISON, PADRES & JOVENES UNIDOS, LINDA ROBERSON, MARTHA THOMPSON, VOCES UNIDAS DE LAS MONTANAS, VOTE FORWARD, GARY YOUNG (Attachments: #1 Memorandum in Support, #2 Exhibit Index, #3 Exhibit USPS OIG Report (Aug. 2020), #4 Exhibit USPS, Postal Operations Manual (Excerpts), #5 Exhibit USPS OIG Report (May 2020), #6 Exhibit USPS, Mandatory Stand-Up Talk: All Employees, #7 Exhibit Leaked USPS Powerpoint, #8 Exhibit DeJoy Testimony - House Oversight Hearing (Aug. 24, 2020), #9 Exhibit DeJoy Testimony - Senate Hearing (Aug. 21, 2020), #10 Exhibit Statement of DeJoy - House Oversight Comm. (Aug. 24, 2020), #11 Declaration Professor Justin Grimmer, #12 Exhibit USPS, Congressional Briefing (Aug. 31, 2020), #13 Exhibit Chart of States Where Voters' Mail-In Ballots Are Impacted By Defendants' Delays, #14 Exhibit USPS Letter to Pa. (July 29, 2020), #15 Declaration Eitran D. Hersh, #16 Exhibit House Oversight Hearing on USPS Operations - Transcript (Aug. 24, 2020), #17 Exhibit Senate HSGAC Hearing on USPS Operations - Transcript (Aug. 21, 2020), #18 Declaration Aaron Carrel, #19 Declaration Martha Thompson, #20 Declaration Kathryn Montgomery, #21 Declaration Sebastian Immonen, #22 Declaration Amy Bolan, #23 Declaration Inderbir Singh Datta, #24 Declaration Scott J. Forman of Vote Forward, #25 Declaration Alex Sanchez of Voces Unidas De Las Montanas, #26 Text of Proposed Order)(Duraiswamy, Shankar)

EXHIBIT 13

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA

VOTE FORWARD, AARON CARREL,
VOCES UNIDAS DE LAS MONTAÑAS,
COLORADO ORGANIZATION FOR
LATINA OPPORTUNITY AND
REPRODUCTIVE RIGHTS, and PADRES
UNIDOS,

Civil Case No. 1:20-cv-02405

Plaintiffs,
v.
LOUIS DEJOY, in his official
capacity as the Postmaster General; and the
UNITED STATES POSTAL SERVICE,
Defendants.
DECLARATION OF EITAN D. HERSH
I.

Statement of Inquiry
1.

In the summer of 2020, United States Postmaster General Louis DeJoy

implemented measures that have caused delays in the delivery of mail. 1 The delays raise concern
in time-sensitive domains, such as delivery of medicines and election ballots. The concern
surrounding ballots is that mail ballots are often submitted by voters in the days just before an
election, and USPS delays would mean that some of those ballots will not reach election offices
on time and will therefore not be counted. I have been asked by Plaintiff’s counsel to investigate
a narrow empirical question: how many voters are likely to send in mail ballots in the final days
before the November 2020 election? These are the set of ballots at risk of not being counted due
to USPS delays. In brief, I estimate that over 7 million ballots, nationwide, will be submitted by

1

Alex Hufford, “What’s Going on with the Postal Service?” Lawfare, Sept 1, 2020,
https://www.lawfareblog.com/whats-going-postal-service.

1

mail in the days right before the election, and are therefore at risk of not being counted due to
USPS delays.
II.

Qualification
2.

I am an associate professor of political science at Tufts University in Medford,

Massachusetts. I also hold the title of Associate Professor of Civic Studies at the Tisch College
of Civic Life at Tufts University. I joined the faculty at Tufts in July 2017. Prior to that, I was an
assistant professor and a resident fellow of the Institution for Social and Policy Studies at Yale
University in New Haven, Connecticut. I joined the faculty at Yale in 2011. I received my PhD
in government from Harvard University in 2011.
3.

My scholarly research focuses on U.S. elections, specifically on voter behavior,

election administration, political campaigns, and civic engagement. My research utilizes both
survey methods as well as methods for analyzing large databases such as voter registration
records. My peer-reviewed paper, “The Dynamic Election: Patterns of Early Voting Across
Time, State, Party, and Age,” (Election Law Journal, 2016) is specifically about the timing of
early voting and mail voting. Other peer-reviewed articles cover a range of topics related to voter
registration, voter ID, and election participation. Examples include “ADGN: An Algorithm for
Record Linkage Using Address, Date of Birth, Gender, and Name” (Statistics and Public Policy,
2017), “The Primacy of Race in the Geography of Income-Based Voting” (American Journal of
Political Science, 2016), and “Movers, Stayers, and Registration: Why Age is Correlated with
Registration in the US” (Quarterly Journal of Political Science, 2012). In addition, I have
written about the quality of records in state voter systems (“Voter Registration: The Process and
Quality of Lists,” in The Measure of American Elections, 2014) and the use of voter registration
databases in campaign politics (Hacking the Electorate, 2015). Articles on survey methods
include “Validation: What Big Data Reveal about Survey Misreporting and the Real Electorate”
2

(Political Analysis, 2012). One article, “Obstacles to Estimating Voter ID Laws’ Effect on
Turnout” (Journal of Politics, 2018) is about survey methods in the context of studying the
impact of voter identification laws.
4.

I have served as an expert witness in four voting cases in which I was deposed:

Judicial Watch v. King, Fish v. Kobach, Stringer v. Pablos, and U.S. v. Eastpointe. The first
three of these cases were focused on compliance with the National Voter Registration Act in
Indiana, Kansas, and Texas, respectively. In the Indiana case, I was hired by defendants; in the
other two, by plaintiffs. The fourth case, U.S. v. Eastpointe, was focused on compliance with the
Voting Rights Act. With regard to the Voting Rights Act, I also served as an expert consultant on
Texas v. Holder and Veasey v. Perry, two cases involving a voter identification law in the State
of Texas. The second of these cases was the basis for my peer-reviewed article on record linkage,
“ADGN: An Algorithm for Record Linkage Using Address, Date of Birth, Gender, and Name.”
In 2018-2019, I was hired as an expert witness by defendants in Kelly v. Verizon Pennsylvania, a
class action case for which voter registration data was used to evaluate the domicile of class
members. I have filed expert declarations in other cases as well. In 2018, I testified before the
U.S. Senate Committee on the Judiciary regarding the Facebook/Cambridge Analytica scandal.
5.
6.
III.

I am being compensated for my work at a rate of $300 per hour.
A copy of my CV is attached as Exhibit A to this report.

Analysis
7.

My goal in this declaration is to use the best available data, under a sharp

deadline, to provide the court with an estimate of the number of ballots potentially affected by
the USPS’s recent policy changes. I begin with a note of caution: it is hard to predict the future.
Patterns of voting behavior from past elections certainly offer a window into what will happen in
the 2020 election, but there is a great deal of uncertainty about how this election will differ from
3

previous ones. Most obviously, the trajectory of the Covid-19 pandemic will likely affect how
many people participate in the November election, how many people will cast ballots by mail,
and when, during the election cycle, they choose to cast those mail ballots.
8.

This analysis will proceed with three estimates: a.) the number of total votes

expected in the November 2020 election, b.) the percent of those votes expected to be cast by
mail ballots, and c.) the percent of mail ballots that are cast in a timeframe in which a delay of
even one day on the part of the USPS will jeopardize the ballots being counted.
A. Estimate of Turnout
9.

The most relied-upon estimates of voter turnout in past elections come from

Professor Michael McDonald, a political scientist at the University of Florida. 2 Professor
McDonald calculates ballots cast divided by the voting eligible population (VEP). The eligible
population is calculated by subtracting non-citizens and otherwise ineligible individuals (e.g.
disenfranchised felons and ex-felons) from the voting-age population. Since 2000, the voting
eligible population has grown steadily by about 4.4% each presidential election year. In the year
2000, the VEP was 194,331,436. In the year 2004, the VEP was 203,483,455. In 2008, it was
213,313,508. In 2012, it was 222,474,111. In 2016, it was 230,931,921. The average growth in
these years is 4.4%. Assuming a 4.4% growth rate in the voting eligible population, the 2020
eligible population will be 241,114,005.
10.

How many eligible citizens will vote? In the last five presidential elections,

turnout among eligible has ranged from 55.3% (in 2000) to 62.2% (in 2008). In 2020, pollsters
are estimating very high turnout. The New York Times’ Nate Cohn recently wrote, “The 2020

2

http://www.electproject.org/home/voter-turnout/voter-turnout-data

4

election is poised to have the highest turnout in a century.” 3 Suppose turnout will not be, in fact,
higher than ever, but that it will match the recent high-water mark of 2008. In that case, 62.2% of
the estimated 241,114,005 eligible citizens will cast ballots, which equals 149,972,911. In other
words, approximately 150 million ballots are expected to be cast in the November 2020
election.
B. Estimate of Mail Ballots
11.

Of 150 million ballots, how many will be cast by mail? Here is where statistics

from prior presidential elections do not offer us much insight; Covid-19 has resulted in new state
laws that make mail balloting easier and in new preferences of citizens to avoid in-person
interactions. In recent presidential years, a quarter to a third of U.S. voters were casting ballots
before Election Day (either by mail or at early in-person locations). 4 In 2016, 21% of voters
nationwide cast a mail ballot. 5 However, all indications point to much higher use of mail
balloting in the upcoming election.
12.

What are those indications? First, in nine states, plus Washington DC, all voters

will be automatically sent mail ballots. This had been the practice in just a few states prior to
Covid-19, but new states, including the nation’s most populous state, have shifted to sending
every registrant a ballot for 2020. Another nine states are automatically sending every registrant
an application through which they can request an absentee ballot. For instance, ahead of a recent
(Sept 1, 2020) election – a highly contested statewide primary – Massachusetts sent all

3

https://www.nytimes.com/2019/07/15/upshot/2020-election-turnout-analysis.html

4

Vivekinan Ashok, et al, “The Dynamic Election: Patterns of Early Voting Across Time, State,
Party, and Age,” Election Law Journal, 15 (2), 2016.
5

Hannah Hartig, Bradley Jones, and Vanney Gomez, “As states move to expand the practice,
relatively few Americans have voted by mail,” Pew Research Center, June 24, 2020.

5

registrants a mail ballot application. Approximately three quarters of all votes cast in the election
appear to have been cast by mail ballot. 6 The states that either plan on sending all registrants
mail ballots for the November election or plan on sending all registrants mail ballot applications
contain 39% of the United States population.
13.

Another 38% of Americans live in other jurisdictions in which a registrant no

longer needs any excuse to request a mail ballot and/or in which Covid-19 is considered as an
excused reason. Again, these are not jurisdictions that affirmatively send registrants mail ballots
or applications but rather that simply permit anyone who would like to vote by mail to do so. In
primary elections held during the pandemic, we have seen cases, in this category of state, in
which the majority of voters have cast ballots by mail. For instance, an August 18, 2020 primary
in the state of Florida showed 60% of ballots cast by mail. 7 An August, 11, 2020 primary in the
state of Minnesota also showed 60% of ballots cast by mail. 8
14.

Thus, some 77% of Americans live in jurisdictions in which anyone can request a

mail ballot (without an excuse) or are mailed applications to vote by mail or are mailed actual
ballots to cast votes by mail. In these jurisdictions, elections that have taken place since the
Covid-19 pandemic indicate that overwhelming majorities of citizens are casting mail ballots.
Even accounting for the fact that a handful of states require excuses (beyond Covid-19) to vote
by mail, it is very likely that a majority of all ballots cast will be cast by mail in November 2020.
Consistent with this view, the New York Times recently estimated that about 80 million ballots

6

https://www.masslive.com/politics/2020/08/record-high-turnout-expected-duringmassachusetts-sept-1-primary-due-to-mail-in-voting.htm
7

https://www.miamiherald.com/news/politics-government/article245268070.html

8

https://www.startribune.com/6-of-10-minnesota-primary-voters-cast-ballots-absentee-or-bymail-tripling-turnout/568966771/

6

will be cast in 2020 by mail. This would amount to 53% of 150 million ballots expected to be
cast. 9 The exact number is difficult to predict. But given that recent months have shown elections
with 60-75% of voters casting ballots by mail, an estimate of 53% of all voters casting mail
ballots, or 80 million nationwide, is an appropriate, if not conservative, estimate. 10
C. Estimate of Timing
15.

In many states, voters can request and receive absentee ballots long before

Election Day. However, past elections reveal a clear behavioral pattern in the casting of mail
ballots: many voters wait until the last minute. Consider evidence from the 2012 election. During
that election, I, as well as three graduate students at Yale University, received daily updates from
states as early in-person and mail ballots trickled into election offices. We gathered this data via
Catalist, a data vendor that is typically hired by political campaigns. Campaigns have an interest
in learning which citizens have already cast ballots so they can more efficiently target their
mobilization campaigns. My colleagues and I were able to use this data vendor to learn about the
timing of balloting for scholarly research. Our results were published in the peer-reviewed
Election Law Journal in 2016.
9

https://www.nytimes.com/interactive/2020/08/11/us/politics/vote-by-mail-us-states.html

10

Why conservative? Consider a simple, back of the envelope calculation. If 39% of the public
lives in a jurisdiction such as California (where ballots will be sent to all registrants) or
Massachusetts (where ballot applications will be sent to all registrants), mail balloting may be
used by something like 85% of voters in that category of state. Another 38% of the public lives
in a jurisdiction such as Florida or Minnesota where mail balloting doesn’t require an excuse.
Mail ballots have been used by 60% of the public in Florida and Minnesota, so suppose that 60%
of voters in this category of state uses mail ballots in November. As for the remaining 23% of the
public that lives in jurisdictions where mail balloting requires an excuse, one still expects much
higher than average use of mail ballots due to Covid. Mail ballots could be used by 15% of
voters. In that case, we calculate (.39 x .85) + (.38 x .60) + (.23 x .15) = 59.4% of ballots will be
cast by mail. 59.4% x 150 million votes = 89.1 million mail ballots. Again, these percentages are
difficult to predict and will be influenced by the trajectory of Covid-19. Nevertheless, reasonable
assumptions suggest that the 80 million estimate used above is an appropriate and conservative
estimate. The list of jurisdiction in each category of mail practice is taken from
https://www.nytimes.com/interactive/2020/08/11/us/politics/vote-by-mail-us-states.html.

7

FIGURE 1: The Timing of 2012 Ballots, by Age and Party Affiliation (Copy of Figure
A1 from Ashok, et al (2016)).

16.

In Figure 1, I reproduce a figure contained in that scholarly article. The left-side

panels shows when mail ballots were counted by the election office and transmitted daily to
Catalist. (The right-side panels show early in-person votes). The top panels divide the population
by three age cohorts. The bottom panels divide the population by three political party cohorts.
8

17.

For the sake of interpretation, focus on the top left panel. Each point on the graph

is interpreted as the percentage of all mail voters who cast their mail ballot by the date indicated.
See, for example, the darkest line, which represents voters older than 60. By October 24, 2012,
approximately 40% of all 60-plus year-olds who cast mail ballots in the 2012 election had
already cast them. The remaining 60% cast mail ballots after October 24th. The lines representing
younger age cohorts are noticeably lower on the graph. For instance, 40% of voters under 30
years old who cast mail ballots did not end up casting them until closer to November 1. In other
words, younger mail voters cast their ballots later in the calendar than older voters.
18.

In Figure 1, a red circle highlights the key feature of this graph for the purposes of

the current declaration: there is a big jump in the graph at the end of the time series. For a very
large percentage of voters who ended up casting a mail ballot in 2012, particularly young voters,
their ballots were counted on the very last day of the pre-election period. For instance, the rightmost triangle on the upper-left graph represents voters under 30 whose mail ballots were counted
as of November 5, 2012, the day before the election. Only 75% of them had ballots counted by
that date. In other words, for about a quarter of all young people who cast mail ballots, their
ballots were counted the final day of the election.
19.

For a more recent and more specific analysis, consider the state of Iowa. In 2016,

the Iowa Secretary of State reported daily updates about when absentee ballots were requested by
registrants, sent to registrants from county election offices, and received back completed from
voters. 11 In Figure 2, I plot the number of absentee ballots received by Iowa’s election offices by
date, during the 2016 election season. Except for the last Saturday of the election season, the

11

Data available here:
https://sos.iowa.gov/elections/pdf/2016/general/AbsenteeCongressional2016.pdf

9

state did not report new ballots received on weekends. Weekend days on which no new data
were reported are indicated in gray.
FIGURE 2: Timing of Receipt of Mail Ballots, Iowa 2016.

20.

Figure 2 shows a clear pattern of more voters submitting ballots over time.

Almost no voters submitted a completed ballot in September. In October, the state received first
about ten thousand ballots a day, increasing to about thirty thousand a day.
21.

The last two dots in Figure 2 represent those ballots received on Monday

November 7 or Tuesday November 8. In Iowa, in 2016, these two dots represent 61,051 ballots
together, which amounted to 9.3% of all mail ballots cast. Assuming the 2016 Iowa ballots were
sent via First Class Mail, as is typical of Election Mail, the USPS has represented that the ballots

10

should be delivered in either 1–3 days 12 or 2–5 days. 13 Adopting the faster timeline, those 9.3%
of all ballots received on Monday, November 7 or Tuesday, November 8 were likely sent on
Saturday, November 5. If we just focus on mail ballots received by Iowa on Election Day itself
in 2016 (the last dot in the series in Figure 2), this represents 3.7% of all mail ballots cast and
received by Iowa counties. If we just focus on those ballots, a delay in mail by one day at the
very least would prevent 3.7% of mail votes from being counted.
22.

However, a one-day delay would likely impact many more ballots than 3.7% of

mail voters. In 2016, it was within the normal delivery expectations for ballots mailed on
Saturday ahead of the election to be received on November 8, Election Day, even though some
of those ballots – maybe most – were received in 2016 on the Monday before the election.
However, under the Post Office’s new procedures, ballots mailed on Saturday may not arrive
even by Tuesday, Election Day.
23.

If we extrapolate from Iowa, what is the implication for 2020? Out of 150 million

likely votes cast in 2020, at least 80 million will probably be cast by mail. If 3.7% of mail ballots
are delayed by one day, this amounts to 3 million ballots. If 9.3% of mail ballots delayed by one
day, this amounts to 7.4 million ballots. These are ballots for which delays at the U.S. Post
Office might make the difference between a vote counted and a vote disqualified. 14

12

USPS, Mail and Shipping Services, https://www.usps.com/ship/mail-shipping-services.htm
(stating First Class Mail is delivered in 1–3 days).

13

USPS OIG, Processing Readiness of Election and Political Mail During the 2020 General
Elections (Aug. 31, 2020), https://www.uspsoig.gov/sites/default/files/document-libraryfiles/2020/20-225-R20.pdf (stating First Class Mail is delivered in 2–5 days).
14

Note that, in reaching this conclusion, I have adopted the 1–3 day delivery timeline for FirstClass Mail. Under the longer 2–5 day delivery timeline, all ballots sent between Thursday,
October 29 and Election Day could be disqualified based on even a one-day delay—resulting in
an even higher number of ballots at risk of disqualification.

11

IV.

Conclusion
24.

Turnout in the 2020 election is expected to be high. Use of mail balloting in the

2020 general election is expected to be unprecedentedly high. On account of Covid-19, more
states are encouraging mail ballots and voters are using mail ballots like never before. Upwards
of 80 million votes – the majority of all votes in the presidential election – are expected to be
submitted by mail. Past behavior indicates that voters, especially younger voters, tend to wait
until the last possible moment to cast their ballot. Data from both 2012 and 2016 indicate high
rates of ballots received by election offices in the final day or days of the election season.
Extrapolating from past elections, in states where ballots must be received by Election Day,
perhaps 9% of Americans who will cast mail ballots in 2020 may submit those ballots on the
Saturday right before the election. Prior to recent changes in practice at the U.S. Post Office,
these ballots would be received by election offices on time. Following the recent changes, delays
– even of just one day – in mail delivery puts a significant number of ballots at risk of not being
counted.

I declare under penalty of perjury that the foregoing is true and correct.
Executed this 7th of September, 2020:

_________________________________
Eitan D. Hersh
Associate Professor
Department of Political Science
Tufts University

12

EXHIBIT A

Curriculum Vitae
Eitan D. Hersh

Contact
Department of Political Science, Tufts University
108 Packard Hall
Medford, Massachusetts 02155
Tel: [phone]
Email: [e-mail]
Web: www.eitanhersh.com

Employment
Associate Professor, Political Science
Tufts University, 2017-Present
Assistant Professor, Political Science
Yale University, 2011-2017

Education
Ph.D. in Political Science, Harvard University (2011)
M.A. in Political Science, Harvard University (2010)
B.A. in Philosophy, Tufts University, summa cum laude (2005)

Books
Eitan Hersh. 2020. Politics is for Power: How to Move Beyond Political Hobbyism, Take Action, and Make Real Change. New York: Scribner.
Eitan Hersh. 2015. Hacking the Electorate: How Campaigns Perceive Voters.
Cambridge: Cambridge University Press.
— Winner, Best Book in Information, Technology, and Politics, APSA 2016

Scholarly Articles
Aaron Kaufman and Eitan Hersh. 2020. “The Political Consequences of
Opioid Overdoses.” PLOS ONE 15 (8): 1-10.
Laura Zatz, Eitan Hersh, Kimberly Gudzune, Anne Thorndike, Matthew
Goldenberg, and Sara Bleich. 2020. “Physicians’ Political Party Affiliation and
Clinical Management of Obesity,” Clinical Obesity: 1-9.
Justin Grimmer, Eitan Hersh, Marc Meredith, and Clayton Nall. 2018.
“Obstacles to Estimating Voter ID Laws’ Effect on Turnout.” Journal of Politics
80 (3): 1045-1051.
Bernard Fraga and Eitan Hersh. 2018. “Are Americans Stuck in Uncompetitive Enclaves? An Appraisal of U.S. Electoral Competition” Quarterly Journal
of Political Science 13 (3): 291-311.
Eitan Hersh and Yair Ghitza. 2018. ”Mixed-Partisan Households and Electoral Participation in the United States” PLOS ONE 13(10): 1-18.
Stephen Ansolabehere and Eitan Hersh. 2017. “ADGN: An Algorithm for
Record Linkage Using Address, Date of Birth, Gender, and Name” Statistics
and Public Policy. 4(1): 1-10.
Eitan Hersh and Brian Schaffner. 2017. “Post-Materialist Particularism:
What Petitions Can Tell Us about Biases in the American Policy Agenda,”
American Politics Research doi.org/10.1177/1532673X17722989.
Ryan Enos and Eitan Hersh. 2017 “Campaign Perceptions of Electoral Closeness: Uncertainty, Fear, and Over-Confidence.” British Journal of Political Science. 47(3): 501-519.
Eitan Hersh and Matthew Goldenberg. 2016. “Democratic and Republican
Physicians Provide Different Care on Politicized Health Issues,” Proceedings of
the National Academy of Sciences. 113 (42): 11811-11816.
Vivekinan Ashok, Daniel Feder, Mary McGrath, and Eitan Hersh. 2016
“The Dynamic Election: Patterns of Early Voting Across Time, Space, Age,
and Party,” Election Law Journal. 15(2): 115-128.
Eitan Hersh and Clayton Nall. 2016. “The Primacy of Race in the Geography of Income-Based Voting: New Evidence from Public Voting Records.”
American Journal of Political Science. 60 (2): 289-303.
— Winner, Pi Sigma Alpha award for best paper presented at MPSA (2013)
— Winner, MPSA Emerging Scholar award (2013)
Ryan Enos and Eitan Hersh. 2015. “Party Activists as Campaign Advertisers: The Ground Campaign as a Principal-Agent Problem.” American Political
Science Review. 109 (2): 252-278.
Stephen Ansolabehere and Eitan Hersh. 2013. “Gender, Race, Age and
Voting: A Research Note.” Politics and Governance. 1 (2): 132-137.

Eitan Hersh. 2013. “The Long-Term Effect of September 11 on the Political Behavior of Victims’ Families and Neighbors.” Proceedings of the National
Academy of Sciences. 110 (52): 20959-20963.
Eitan Hersh and Brian F. Schaffner. 2013. “Targeted Campaign Appeals
and the Value of Ambiguity.” Journal of Politics. 75 (2): 520-534.
Stephen Ansolabehere and Eitan Hersh. 2012. “Validation: What Big Data
Reveal About Survey Misreporting and the Real Electorate.” Political Analysis
20(4): 437-459.
Stephen Ansolabehere, Eitan Hersh, and Kenneth Shepsle. 2012. “Movers,
Stayers, and Registration: Why Age is Correlated with Registration in the U.S.”
Quarterly Journal of Political Science 7(4): 333-363.
Eitan Hersh. 2012. “Primary Voters Versus Caucus Goers and the Peripheral
Motivations of Political Participation. Political Behavior. 34(4): 689-718.
Bernard Fraga and Eitan Hersh. 2011. “Voting Costs and Voter Turnout in
Competitive Elections.” Quarterly Journal of Political Science 5(4): 339-356.

Other Writing
Eitan Hersh, “How Democrats Can Repair Their Generational Divide,” Politico,
March 12, 2020.
Eitan Hersh, “‘We All Really Need To Do Hard Things”’ The Progressive,
January 31, 2020.
Eitan Hersh, “Listen Up, Liberals: You Aren’t Doing Politics Right,” New
York Times, January 27, 2020.
Eitan Hersh, “College-Educated Voters are Ruining American Politics,” The
Atlantic, January 20, 2020.
Eitan Hersh, “What Rich Centrists Could Learn from Howard Schultz about
Political Power,” Niskanen Center, January 16, 2020.
Eitan Hersh, “Politics is for Power, Not Consumption,” Boston Review,
November 4, 2019.
Eitan Hersh, “The Political Beliefs and Civic Engagement of Physicians in
an Era of Polarization,” Introduction to special issue, guest editor of special
issue, Journal of Health Politics, Policy, and Law 44, 2019.
Eitan Hersh, “In a Deep Blue or Red State? You Can Still Influence Politics,”
New York Times, October 29, 2018.
Eitan Hersh, “Political Hobbyists are Ruining the Country,”New York Times,
July 2, 2017.
Eitan Hersh, “The Most Dangerous Hobby,” Boston Globe, May 3, 2016.
Occasional essays in media outlets such as the Washington Post’s Monkey

Cage and FiveThirtyEight, see https://fivethirtyeight.com/contributors/eitanhersh/.
Justin Grimmer, Brian Feinstein, Eitan Hersh, and Daniel Carpenter, “Are
Close Elections Random?” Working Paper.
Stephen Ansolabehere and Eitan Hersh. 2014. “Voter Registration: The
Process and Quality of Lists.” In The Measure of American Elections, eds.
Barry C. Burden and Charles Stewart III. Cambridge: Cambridge University
Press.
Stephen Ansolabehere and Eitan Hersh. 2011. “Who Really Votes?” In
Facing the Challenge of Democracy, eds. Paul M. Sniderman and Benjamin
Highton. Princeton: Princeton University Press.
Stephen Ansolabehere, et al., “Voter Registration List Quality Pilot Studies,”
Caltech/MIT Voting Technology Project Report, June 8, 2010.

Teaching
“Information, Technology, and Political Power”
“U.S. Elections: Rules, Strategies, Outcomes”
“Advanced Seminar on U.S. National Elections”

Referee Service
American Journal of Political Science American Political Science Review, American Politics Research, American Sociological Review, British Journal of Political Science, Economics and Politics Election Law Journal, Electoral Studies,
Journal of Elections, Public Opinion and Parties International Journal of Public Opinion Research, Journal of Politics, Journal of Urban Affairs Legislative
Studies Quarterly, Political Analysis, Political Behavior, Political Psychology,
Political Research Quarterly, Political Science Research and Methods, Politics
and Governance, Proceedings of the National Academy of Sciences, Public Opinion Quarterly, Research and Politics, Review of Economics and Statistics, State
Politics and Policy Quarterly.

Invited Talks
I have given presentations at annual meetings of the American Political Science Association, Midwest Political Science Association, and Society for Political
Methodology.
I have given invited talks at Boston University, Columbia, Chicago, Dartmouth, Harvard, Harvard Kennedy School, Harvard Medical School, McGill,
MIT, NYU, UCLA, University of Massachusetts-Amherst, Michigan, UNC, UPenn
Annenberg School, Princeton, Stanford, Stanford Law School, Washington UniversitySt. Louis, and Yale Law School.

Consulting & Service
Editorial Board, Election Law Journal, 2020Member, Brookline, MA, Select Board’s Taskforce to Reimagine Policing, 20202021
Academic Advisor, 100 Million: Nonvoter Study, Knight Foundation, 2019-2020.
Witness, United States Senate Committee on the Judiciary, May 16 2018 (Facebook/Cambridge Analytica scandal)
Statistical Consultant, Texas v. Holder (Voter ID case)
Statistical Consultant, Veasey v. Perry (Voter ID case)
Expert, Judicial Watch, et al v. King, et al, Indiana Voter Registration Case
Expert, Fish v. Kobach, Kansas Voter Registration Case
Expert, Stringer, et al v. Pablos, Texas Voter Registration Case
Expert, United States v. Eastpointe, MI, Voting Rights Act case
Expert, Collins v. Adams, Kentucky Voter ID case
Commissioner, Commission on Youth Voting and Civic Knowledge, CIRCLE,
2012-14