Partisan and Nonpartisan Election Administration
1
Daniel P. Tokaji
The Ohio State University, Moritz College of Law
Prepared for Election Reform Agenda Conference
University of Iowa
May 7-9, 2009
2
Please don’t cite without permission
Draft - 5/4/09
The significant discretion vested in those who run our elections makes it imperative that
they discharge their duties in an evenhanded manner. Since the contested 2000 presidential
election that inaugurated the recent wave of election administration reforms,1 the partisanship of
state and local election officials has been a prominent subject of concern. The actions of
Florida’s Secretary of State Katherine Harris, who was co-chair of the Bush-Cheney’s 2000
campaign in Florida, caused enormous consternation in Democratic quarters. There was also
concern about the actions of officials at the local level, many of whom were elected on a partisan
basis. These concerns reappeared in the 2004 election, when the spotlight focused on Ohio’s
chief election official, Secretary of State Ken Blackwell. Like Harris, Blackwell served as a cochair of the Bush-Cheney campaign. He too was accused of making decisions – like refusing to
count provisional ballots cast in the wrong precinct and infamously requiring that registration
forms be on at least 80-pound paper weight – designed to benefit his party’s candidate in a key
swing state.2 By 2008, a Democratic Secretary of State, Jennifer Brunner, had replaced
Blackwell in Ohio, but this did not eliminate concerns about partisanship. To the contrary,
1
This paper uses the term “election administration” to refer to the set of nuts-and-bolts
issues surrounding the mechanics and management of elections, such as voting technology,
registration practices, provisional voting, voter identification, polling place operations, recounts
and contests. I omit from discussion question of partisanship and nonpartisanship in other areas
of election law, such as redistricting and campaign finance regulation.
2
For more on this and other controversies that emerged in the 2004 election, see Daniel P.
Tokaji, Early Returns on Election Reform: Discretion, Disenfranchisement, and the Help
America Vote Act, 73 GEO. WASH. L. REV. 1206 (2005).
3
Please don’t cite without permission
Draft - 5/4/09
Republicans accused Secretary Brunner of being a sort of Bizarro-Blackwell, making decisions
designed to benefit Democrats at the expense of Republicans.3
Accusations of partisanship by state and local election officials have triggered proposals
to move to some form of nonpartisan or bipartisan election administration. Such
recommendations have appeared in the reports of blue-ribbon commissions4 as well as in
academic commentary by legal scholars5 and others.6 There have also been some efforts to
reform state law to implement some form of nonpartisan election administration. A group called
Reform Ohio Now put an initiative on the ballot in 2005 that would have transferred election
administration authority from the partisan Secretary of State to an appointed bipartisan board
3
See Alan Johnson, GOP Crying Foul Over Absentee-Voting Law It Passed, COLUMBUS
DISPATCH, Aug. 14, 2008 (quoting Deputy Chair of Ohio Republican Party as accusing Brunner
of having a “troubling partisan agenda”).
4
See, e.g., COMMITTEE ON FEDERAL ELECTION REFORM, BUILDING CONFIDENCE IN U.S.
ELECTIONS 50 (2005).
5
See STEVEN H. HUEFNER, DANIEL P. TOKAJI, & EDWARD B. FOLEY, FROM REGISTRATION
TO RECOUNTS: THE ELECTION ECOSYSTEMS OF FIVE MIDWESTERN STATES 188-89,
196-97
(2007); Richard L. Hasen, Beyond the Margin of Litigation: Reforming Election Administration
to Avoid Electoral Meltdown, 62 WASH. & LEE. L. REV. 937, 973-92 (2005).
6
See Robert A. Pastor, Improving the U.S. Electoral System: Lessons from Canada and
Mexico, 3 ELECTION L.J. 584 (2004).
4
Please don’t cite without permission
Draft - 5/4/09
5
Please don’t cite without permission
Draft - 5/4/09
failed in 2005. That proposal was handily defeated.7 In Wisconsin, however, good government
advocates succeeded in instituting a Government Accountability Board in 2007, whose members
are selected through a process designed to ensure bipartisan support.8
There is some empirical research regarding the effects of partisanship on the decisions
make by election officials which, by implication, sheds some light on what might be considered
“best practices” when it comes to the institutions that run our elections. But there is not a lot9 –
in my judgment, not enough to be confident in precisely what institutional reforms will work best
at the state and local level. We may know (or think we know) that partisan election
administration is a problem, but it is harder to know with any certainty what should be done
about it. In the remainder of this paper, I summarize the research that exists and make some
suggestions on what additional research, both empirical and legal, would be useful in developing
reforms to election administration institutions at the state and local level.10
7
See State Issue 5: November 8, 2005, available at
http://www.sos.state.oh.us/SOS/elections/electResultsMain/2005ElectionsResults/051108Issue5.aspx (accessed May 4, 2009).
8
See HUEFNER, ET. AL., supra note 5, at 115-17.
9
See David C. Kimball & Martha Kropf, The Street-Level Bureaucrats of Elections:
Selection Methods for Local Election Officials, 23 REV. OF POL’Y RES. 1257 (2006)(“Despite the
long-term existence of nonpartisan governing and its wide use among city councils, there is a
dearth of empirical evidence to support policy changes, especially for local election
officials.”)(citation omitted).
10
I omit, for purposes of this paper, reforms that might be made at the federal level to the
6
Please don’t cite without permission
Draft - 5/4/09
Here’s what we know: At the state level, partisan election administration – at least in the
sense that party-affiliated elected or appointed officials run elections – is the rule. According to
Rick Hasen, 33 states have a chief election official who is elected through a partisan election
process.11 Other states have an appointment process but, in many of those states, the chief
election official is appointed by the state’s governor (who is of course elected through a partisan
process).12 At the local level, about two-thirds of jurisdictions elect their election officials, and
party-affiliated officials run elections in almost half of local jurisdictions.13 David Kimball and
Martha Kropf have identified over 4500 local jurisdictions in the U.S.,14 and reported the
following selection methods:
Selection Method for Local Election Authority
Selection Method
Share of Jurisdictions
Voter Representation
Election Assistance Commission (“EAC”), the federal administrative agency with responsibility
over election administration. For more on this subject, see Daniel P. Tokaji, Voter Registration
and Institutional Reform: Lessons from a Historic Election, 3 HARV. L. & POL’Y REV. ONLINE
(Jan. 22, 2009).
11
Hasen, supra note 5, at 974.
12
Id. at 974-75.
13
David C. Kimball, Martha Kropf & Lindsay Battles, Helping America Vote? Election
Administration, Partisanship, and Provisional Voting in the 2004 Election, 5 ELECTION L.J. 447,
453 (2006).
14
Kimball & Kropf, supra note 9, Table 1.
7
Please don’t cite without permission
Draft - 5/4/09
Individual Elected by Voters
61%
45%
Elected Board of Elections
2%
1%
Appointed Board of Elections
22%
31%
Appointed Individual
15%
22%
About 46% of local election jurisdictions had party-affiliated election authorities (20%
Republican, 26% Democratic, and 0.1% other), while 14% had bipartisan and 29% nonpartisan
local election authorities.
Survey evidence suggests that this reality doesn’t square with the public’s views on who
should be running elections. Michael Alvarez and Thad Hall report on a national survey that
found strong support for nonpartisan boards. Of the general population, 66% thought that local
or state election officials who run elections should be nonpartisan, while only 19.6% thought
they should be partisan.15 Among registered voters, the percentage favoring nonpartisan election
administration was even higher, at 72.6%. Only 1.5% of the general population and 0.9% of
registered voters favored a partisan single elected official, the model that predominates at the
state level, as opposed to a nonpartisan elected or appointed board.
There’s also some evidence – though not a lot – that partisan election administration may
affect the decisions election officials make. Here are the empirical studies that I’ve found on this
15
R. Michael Alvarez & Thad Hall, Public Attitudes About Election Governance (June
2005 manuscript), available at
http://www.cppa.utah.edu/publications/elections/Election_Governance_Report.pdf.
8
Please don’t cite without permission
Draft - 5/4/09
subject (and I’d be grateful for anyone letting me know of others that I may have missed):
$
Guy Stuart examined the use of centralized voter lists to purge felons from Florida’s
voting rolls. He found Florida counties with Republican election officials tended to be
more aggressive in purging voters from the rolls than those with Democratic election
officials. This is consistent with partisan motivation, since Democrats are generally
believed to be more seriously harmed by overly aggressive purges.16
$
Kimball, Kropf & Battles found some evidence of an interaction between local election
officials’ partisan affiliation and provisional voting practices. Democratic officials were
slightly more likely to implement a more generous rule with respect to counting “wrong
precinct” provisional ballots. In jurisdictions with a Democratic election authority, the
number of provisional ballots counted as Democratic vote share increased. By contrast,
in those with a Republican election authority, the number of provisional ballots decreased
as the Democratic vote share increased.17
$
Anna Bassi, Rebecca Morton, and Jessica Trounstine examined the relationship between
the party of local registrars and turnout in state elections. They found a strong positive
correlation between the party of the registrar and the turnout of members of that party in
gubernatorial elections. Bipartisan boards, on the other hand, tended to mitigate this
effect.18
16
Guy Stewart, Databases, Felons, and Voting: Bias and partisanhip of the Florida Felon
List in the 2000 Elections, 119 POL. SCI. Q. 453 (2004).
17
Kimball, Kropf & Battles, supra note 13, at 457-58, 459.
18
Anna Bassi, Rebecca Morton & Jessica Trounstine, Reaping Political Benefits: Local
9
Please don’t cite without permission
Draft - 5/4/09
$
Joshua Dyck and Nicholas Seabrook examined the moving of voters to inactive status to
Oregon which, in its “vote by mail” system, results in the voter effectively being
disenfranchised. They found that Republican local election officials were more likely to
move Democratic registrants to inactive status than Democratic officials.19
Collectively, these studies provide some evidence to support the claim that partisan
election officials sometimes act to benefit their parties, rather than discharging their
responsibilities evenhandedly. In conjunction with the anecdotal evidence described at the start
of this paper, the empirical research helps make the case for moving toward some form of
nonpartisan or bipartisan election administration.
Reformers might also point to evidence of public skepticism regarding the honesty and
fairness of elections. From 1996 to 2000, National Elections Studies data showed a sharp
increase in the percentage of the public believing the election was “somewhat unfair” or “very
unfair.”20 And a 37-country survey of conducted in 2004 found that the United States came in
next to last (just ahead of Vladimir Putin’s Russia) in the percentage of citizens who rated the
last national election “very honest” or “somewhat honest,” while a higher percentage of
American citizens rated their last national election “very dishonest” than those in any of the other
Implementation of State and Federal Election Law (July 22, 2008 manuscript).
19
Joshua J. Dyck & Nicholas R. Seabrook, The Problem with Vote-by-Mail, Paper
prepared for presentation at the Annual Meeting of the Midwest Political Science Association
(Mar. 18, 2009 manuscript).
20
Hasen, supra note 5, at 943.
10
Please don’t cite without permission
Draft - 5/4/09
surveyed countries.21 Of course, at the time of this survey, the last presidential election had
taken place in 2000. Still, this evidence bespeaks a lack of confidence on the part of many
citizens in the integrity of election administration.
That said, I think there’s a need for more research on the impact of partisan election
administration. While there’s public concern about the honesty and fairness of American
election administration, it’s not entirely clear that moving to nonpartisan election administration
will actually improve public confidence. And while it may seem obvious to some that partisan
election officials are more likely to make decisions that benefit their parties, there’s a need for
more empirical research to augment that cited above and to verify that this is in fact the case.
Furthermore, even if it were conclusively established that partisan election administration
diminished public confidence and led to partisan decisionmaking, there would still be the
question of what to do about it. Hasen has suggested that state chief election officials be
nominated by the governor and approved by a supermajority of the state legislature (something
like 75%), to ensure a consensus candidate who is above the political fray.22 Wisconsin has
actually implemented an institutional reform designed to ensure evenhanded election
administration, in the form of a Government Accountability Board whose members must be
approved by a two-thirds supermajority of the state senate.23 Chris Elmendorf proposes a
21
Caroline Tolbert, Todd Donovan & Bruce E. Cain, The Promise of Election Reform, in
DEMOCRACY IN THE STATES: EXPERIMENTS IN ELECTION REFORM 1, 7 (2008).
22
Hasen, supra note 5, at 984.
23
HUEFNER, FOLEY & TOKAJI, supra note 5, at 132.
11
Please don’t cite without permission
Draft - 5/4/09
different institutional fix, in the form of advisory commissions like those used in the U.K.24 All
these are promising ideas. But how well will they work? We really don’t know yet.
This leads me to suggest three general areas in which more research would be useful.
The first is more research on how partisan and nonpartisan election administration affects the
decisions that are made. The studies identified in the bullet points above provide a helpful
starting point, but more research along these lines would be valuable in confirming the intuition
of many observers that partisan election officials make partisan decisions. The second area in
need of further research is how partisan and nonpartisan election administration affects public
confidence. There is survey evidence that Americans lack confidence in their system of election
administration and think nonpartisan election administration is a good idea, but it would be
helpful to “connect the dots” by demonstrating that people in jurisdictions with nonpartisan (or at
least less partisan) election administration have greater confidence in their systems. Third, we
need more research on what sorts of institutional arrangements work best. This might take the
form of qualitative, comparative assessments of how different ways of running elections at the
state and local level actually work in practice, of the type that my Moritz colleagues tried to do in
From Registration to Recounts.25
In the meantime, one might fairly ask, what if anything should policymakers do? Should
we try to muddle through with existing institutional arrangements, until we get a better sense of
what their problems are and what should be done about them? On this point, I confess to being a
24
Christopher S. Elmendorf, Representation Reinforcement Through Advisory
Commissions: The Case of Election Law, 80 NYU L. REV. 1366, 1395-1404 (2005).
25
HUEFNER, TOKAJI & FOLEY, supra note 5.
12
Please don’t cite without permission
Draft - 5/4/09
bit conflicted. My ordinary instinct is to be cautious, in keeping with what I’ve termed the
“Moneyball” approach to election reform which resists seat-of-the-pants judgments about what
reforms should be implemented – something we’ve seen far too much of in recent years – and
instead favors studious assessments of the evidence before changing election laws.26 At the
same time, the lack of evidence on what institutional arrangements work best is partly due to the
lack of experimentation, particularly at the state level. Although there are a variety of
institutional structures at the local level, the vast majority of states still have either an elected or
appointed partisan official running elections. It’s difficult to know what sort of institutional
structure will work best when there has been so little variation and, therefore, comparative
analysis of different alternatives.
In sum, there’s a bit of a chicken-egg problem here. We don’t know what institutional
arrangements will produce the best election administration to the shortage of research on the
subject. But part of the reason for this paucity of research is that most states have the same type
of state-level structure. For this reason, I’m inclined to believe that states should experiment
with different institutional arrangements, as Wisconsin has done and as Reform Ohio Now tried
to do, even without clear evidence on what the “best practices” are. Only through such
experimentation, and through empirical and legal researchers carefully studying their results, will
we ultimately develop election administration institutions that will better serve the needs of all
citizens.
26
Daniel Tokaji, The Moneyball Approach to Election Reform (Oct. 18, 2005), available
at http://moritzlaw.osu.edu/electionlaw/comments/2005/051018.php.
13