---

Summary

---

Introduction

Eighty-four percent of the nearly 2,900 counties that administer elections at the county level either directly elect their election official in a partisan election or ask a partisan official to appoint the election official.

While federal and state governments set the broad rules that govern elections, local governments carry them out.

Local officials are often responsible for:

• Registering voters and removing people from the registration list who are no longer eligible (e.g., they passed away or moved to a different state)
• Setting the number of polling locations and deciding where to place them
• Hiring and training poll workers
• Purchasing election equipment
• Tallying the votes and deciding whether to count a provisional ballot
• Setting the early voting schedule
• Processing mail ballots.

Each of these tasks is an opportunity for local officials to affect how easy or hard, how pleasant or unpleasant, it is to vote. Many available choices affect the experience of different citizens differently—citing a polling place in the suburbs makes the commute easier for suburban residents at the expense of city residents.

Some citizens may decide whether to vote or not on the basis of these administrative choices, voting when it’s easy and pleasant and not voting when it’s hard and unpleasant.

Many of the local officials responsible for administering elections are themselves directly elected in partisan elections or are appointed by partisan figures, raising the concern that they will use their authority to advantage their party. Eighty-four percent of the nearly 2,900 counties that administer elections at the county level either directly elect their election official in a partisan election or ask a partisan official to appoint the election official (1). One alternative—nonpartisan officials appointed in a bipartisan way—are rare.

Would appointed nonpartisan election officials make elections fairer? Existing research has focused on measuring how unfair partisan election officials are. This primer discusses the competing expectations social scientists bring to this question, before reviewing the evidence that social scientists have produced on the question so far. It closes by discussing what the evidence implies for nonpartisan election administration as a reform, and considers what researchers need to study further.

---

It is not hard to see why many social scientists and members of the public alike are worried about the partisan selection of election officials. Election officials have a lot of discretion (2).

They are also rarely covered in the media (3), so it’s reasonable to worry that may have opportunities to tilt the scales in their party’s direction, or even skirt the rules, without being noticed. Since Democratic and Republican public officials in federal and state office often have such different visions for the country, including different visions for how elections should be run, it is natural to expect that these visions are better for their party and that local election officials share the views of their party. In fact, a number of state officials have gone so far as to encourage co-partisan election officials to use their authority to advantage their party (4).

And yet, a number of countervailing forces may limit how much election officials advantage their party.

Election Officials Have a Limited Tool Kit

Many of the policies that election officials can implement target small populations or only make it modestly more or less difficult to vote. While policies like this can tilt the partisan balance of the electorate, those advantages tend to be small (5).

Candidates Are Watching What Election Officials Do

While the press and public are likely not carefully monitoring how election officials behave, candidates may. Candidates have the incentive to keep a close eye on how changes in election administration affect their chance to win. Statewide and national campaigns likely have the resources to contest egregious choices in court or make them widely known. If a choice is illegal or unpopular, the threat of campaigns bringing attention to the issue may be enough to dissuade election officials from using all of their authority to advantage their party.

Any Single Election Official isn’t Powerful Enough to Sway the Outcome of Statewide Races

There are also some costs associated with trying, so they avoid it. Some of the races that a partisan election official may be most motivated to affect are held across many jurisdictions (e.g., race for governor). Most jurisdictions make up a small share of the total votes cast in a statewide or national race. So, even if an election official were able to dramatically alter the share of people voting for one party or another in their jurisdiction, that might not amount to much when combined with the results of many other jurisdictions. Election officials might still be motivated to try, but not if there are costs.

The People Who Become Election Officials Are Not as Polarized as the General Public on Election Administration

While election officials are often elected through partisan elections or appointed by partisan officials, the types of people who seek the job are often experienced bureaucrats. The job of election official is tedious and technical, and it is often only one part of a larger suite of roles that the official is responsible for—in many jurisdictions, the person who runs elections also records deeds and marriage licenses among other tasks.

Selecting leaders based on their experience and interest in the role can lead to officials with policy preferences that differ from the public. For example, compared to the public, a smaller share of election officials want to hold elections on the weekend (6). But this also tends to limit partisan polarization on election administration issues, at least in some cases —Republican election officials are more likely to support mail voting than Republicans in the public, and Democratic election officials are less likely to support election day registration than Democrats in the public.

Even if an election official were able to dramatically alter the share of people voting for one party or another in their jurisdiction, that might not amount to much when combined with the results of many other jurisdictions.

---

Barriers to Studying the Advantage Election Officials Give Their Party

There are two main barriers to studying whether election officials advantage their party. First, Democratic and Republican election officials serve in places that have different geography, preferences about how to vote, interest in politics, etc. Democrats tend to serve in cities where space for polling places is more expensive, traffic is worse, people are closer together, and the public likes voting by mail. This means that even if election officials do not advantage their party, election administration policies and election outcomes may be different in Democratic- and Republican-controlled jurisdictions.

Second, election officials make many choices. We want to know whether election officials use these choices to advantage their party. But, we can only see some of the choices they make, and it can be hard to determine whether any particular choice will advantage the election official’s party.

Put together, these two issues mean that simply comparing turnout, the number of polling places, or some other election administration measure in Democratic- and Republican-controlled counties does not tell us whether election officials advantage their party.

The first modern empirical paper on partisan election officials demonstrates these two challenges well (7). It studies whether election officials in North Carolina purchase voting equipment that will improve their party’s performance. The authors document that certain technologies make it easy for voters to cast all of their votes for one party. The authors also argue that this equipment favors Democratic candidates in counties where most voters are Democrats and favors Republican candidates in counties where most voters are Republicans. In North Carolina, the governor selects the members of the election board, and the governor’s party holds a majority on every county election board. The paper relies on data from 1992, a year when the governor was a Republican so all election boards were controlled by Republicans.

The authors find that boards appear to advantage their party—Republican-controlled county boards are more likely to use the technology that makes it easy to vote straight ticket in counties where most of the citizens are Republicans than in counties where most of the citizens are Democrats. Yet, when the authors ask boards how long they have used their current technology, a large majority say they have used it since before 1984 when the governor was a Democrat and Democrats controlled all county boards. In other words, counties that vote more for Republicans were more likely to have technology that enabled straight-ticket voting, but partisan county boards did not strategically select technology to advantage their party. In fact the Democratic county boards mostly used the same technology in prior elections to their disadvantage.

Elections may be run differently in Democratic– and Republican-controlled jurisdictions for reasons unrelated to the party of the election official.

This captures the basic problem: elections may be run differently in Democratic- and Republican-controlled jurisdictions for reasons unrelated to the party of the election official.

Election Officials Do Not Noticeably and Consistently Advantage Their Party

Two recent papers address these concerns. Both find that election officials do not substantially and consistently advantage their party.

The first paper studies polling places in North Carolina (8). As discussed above, the governor appoints county election board members in North Carolina and always chooses a majority from their party, so every county has an election board with a majority that favors the governor’s party. The paper studies two elections, one when Democrats controlled the election boards (2012), and one when Republicans controlled the election boards (2016). It asks whether election boards moved the polling places of members of the opposing party more often than they did for members of their own party. It also studies whether election boards move polling places further from their opponents and closer to their allies. By comparing polling place moves that benefit allies and hurt rivals in 2012 and 2016, the paper inoculates against the risk that polling place moves may just be more common for Democratic or Republican citizens—Democrats run elections one year and Republicans run them in the next, so we can see whether different partisan leadership affects the behavior of the board.

The paper finds that county boards do not use their authority citing polling places to advantage their party. County boards move the polling places of their allies and rivals at nearly identical rates. In the simplest analysis, election officials change the polling places of roughly 16% of their partisan allies and 15% of their partisan rivals. County boards also do not strategically move the polls further from their rivals and closer to their allies. Instead, polling place moves appear unrelated to partisan politics.

A complementary recent paper extends this research to study 21 states over 20 years where elections are run at the county level by a single person who is directly elected in a partisan election and oversees all or nearly all election administration (9). The paper asks whether Democratic presidential candidates perform better in counties where Democrats run elections rather than Republicans. To answer this question, it compares the performance of Democratic presidential candidates in places where Democratic election officials narrowly won their election to places where Republican election officials narrowly won. These places should be similar in all ways other than who runs elections, so this analysis isolates the effect of partisan control of local election administration from other differences in the counties that have Democratic or Republican election officials.

The paper finds that election officials do not noticeably advantage their party. Instead, Democratic and Republican officials elected in close elections oversee presidential elections with similar outcomes and turnout. The authors find that Democratic presidential candidates perform at most 0.3 percentage points better in counties with a Democratic election official than in counties with a Republican official, and they may perform worse in elections run by their own party. In other words, the evidence is not consistent with election officials substantially advantaging their party.

Both of these studies have weaknesses: the first focuses on only one election administration choice. The policy choice they focus on—where to cite polling places—seems like a potentially important one and one that election officials have a lot of discretion over. Still, the fact that election officials don’t move polling places away from their rivals does not imply that election officials are not using other tools to advantage their party.

The second study cannot rule out that election officials offer small or rare but determinative advantages to their party. The study uses a research design that isolates the advantage election officials offer their party, but this research design would only reliably find that election officials advantaged their party if they were able to increase their party’s performance by roughly one and a half percentage points on average. Some elections are determined by less than 1.5 points in one county. If election officials are offering less than a percentage point advantage to their party in their jurisdiction, this paper may miss this.

Democratic and Republican officials elected in close elections oversee presidential elections with similar outcomes and turnout.

These two studies together suggest that election officials are not generally advantaging their party. Their weaknesses make clear that we need additional complementary research that investigates when and where this pattern holds and can tell us about the effects in close races.

---

Conclusion

Put together, the evidence suggests that most partisan election officials are not using their authority to advantage their party.

This may be because election officials don’t have the tools necessary to change the outcome, they are worried about potential consequences, they only administer the election for a small share of the electorate, or they are more motivated by public service.

While the evidence suggests that partisan election officials do not systematically advantage their party, it’s not obvious that this is permanent. If we rely on the personal motivations of election officials to produce neutral policies, new people can be elected or appointed with different motivations. On the other hand, if election officials are highly constrained because their toolkit is limited, we do not need to worry as much about election administration that advantages one party or another in the future.

While the evidence suggests that partisan election officials do not systematically advantage their party, it’s not obvious that this is permanent.

Assuming that partisan election officials will continue to be neutral, would nonpartisan election officials still be better? The main argument in favor of a partisan process for selecting election officials is about representation—the two parties have different views on how elections ought to be run, and people broadly know what these visions are, so using party membership as a shorthand for the types of policies a person will implement could lead to elections run the way that the community wants. The problem is that this doesn’t seem to be happening. Regardless of the party in charge, elections are run in a similar way. It is also not clear that elections are an area where we want individual communities tailoring the rules to local tastes.

The common argument against partisan election officials is that the public is suspicious of parties being involved in running elections. This may be true, but it is also not obvious. On the one hand, explicit partisan involvement in election administration leads to an ever-present concern that election officials will at some point illegitimately use their authority to advantage their party. On the other hand, partisan officials may also be especially good messengers to their own co-partisans that the process is fair. These considerations point to appointed nonpartisan election officials as a promising institutional design, but one that could have unintended effects. More research needs to be done to assess which of the existing institutions engender more trust and operate most effectively.

---

References

---