Reported below are voter turnout and roll-off figures for last night’s elections in Fairfax County, Virginia. Data was assembled from election returns available here. Turnout is calculated from the reported 696,083 eligible voters.  Races allowing a voter to pick multiple candidates from a field are denoted with an M (and total votes divided by appropriate number). An asterisk denotes uncontested races or partially uncontested races.  Roll-off is calculated as the percentage of votes cast in the top race but not cast in a lower race; since there was no gubernatorial election, the aggregate votes cast in the state Senate races is used as the top race.

A few comments:

1) Low turnout is no surprise here. You don’t have to read the endless political science studies to know that state and local elections held in an off-year are going to feature lower turnout than presidential elections. Last night, 201,044 votes were cast in Fairfax County for State Senate, the most in any race. In 2010, 303,379 voters cast votes in the U.S. House elections in Fairfax County, and 516,254 votes were cast for President in Fairfax County in 2008. There is simply less voter interest in state and local politics, and as I’ve written before, that’s a fact that I find highly lamentable. Still, it could be a lot worse: 25% of all possible voters were cast, and that’s many multiples of what some localities in other parts of the country received last night, I’m sure.

2) Um, what’s roll-off? It’s the tendency of voters to only cast a vote for the top races on a ballot, such as President, and to decline to cast a vote in lower-profile races that appear on the same ballot. The standard metric is the percentage of voters who vote in the top race but do not vote in a lower race. Roll-off tends to increase as you head down a ballot, such that in a Presidential election year you might get a 2 to 5% roll-off in a given House race, but 15% or more roll-off for a local election. For example, in Fairfax County in 2008, 516,254 voters were cast for President, but only 509,473 for U.S. Senate (1.3% rolloff), only 504,243 for U.S House (2.3%) , and only 493,642 for the lone bond issue (4.4%). The roll-off in Fairfax County last night was massive in some cases. A full one-third of votes for the soil and water conservation board were left on the table. As were 15% of the votes for School Board At-Large.

Roll-off raises a number of issue for a democracy: is it an indication of voter indifference for local government? A crowding out of local politics by state and federal issue coverage? A reason to not have concurrent federal,state, and local elections? Roll-off figures also provide us with clues as to the health of local democracy: if the number of voters who are already at the polling place but who cannot be compelled to choose a local official increases over time, that might be cause for concern.

3) What caused the roll-off last night? The political science literature regarding roll-off suggests three causes: voter fatigue from long ballots, the structure of the ballot itself, and rational voter abstention. Voter fatigue is unlikely to have had a large effect last night in Fairfax County; not only was the ballot a short three pages, but the school bond referendum, which received the third-most total votes cast, was on the last page. Therefore, a maximum of 5.1% of all roll-off can be attributed to fatigued voters not completing their ballots. But wait! The Board of Supervisor Chair was on the first page, meaning it’s highly unlikely that any of the 4.7% of that roll-off was fatigue. Which suggests that voter fatigue was no more than 0.4% of the roll-off.

The structure of the ballot may have contributed to the roll-off last night. Fairfax county uses a “pure office block” arrangement of the ballot, in which the offices are listed sequentially with the candidates listed under each office. This structure has been shown to produce more roll-off than the main alternative, the “party column” ballot, in which the offices are listed down the side of the ballot and each party has a column running across the top of the ballot, allowing voters to easily vote a straight party line, and sometimes even including an automated mechanism to do so. Another potential factor is multi-candidate races; last night in Fairfax the ballot for both the School Board At-Large race and the Soil and Water Conservation Board race asked voters to elect three candidates from a list. Some voters may have been confused and only selected one candidate, although voters who used the touch-screens would have been reminded at least once that they had not made all possible choices. And this, of course, raises the final structural issue: there is some evidence that the electronic voting machines reduce roll-off, in part because they can do things like remind you that you didn’t fill out all races.

The most likely culprit for the roll-off last night, however, is rational voter abstention, which includes several things. The most obvious is uncontested elections; voters have little incentive to vote in a race that only features one candidate. The only completely uncontested race last night in Fairfax was for Commonwealth Attorney, and it featured a 25% roll-off. Three other races featured were partially uncontested: the school board  (3 of 9 districts uncontested), the Board of Supervisors (3 of 9 districts uncontested), and the state House of Delegates (6 of 17 districts uncontested). The second reason for a voter to abstain is if they have no information about an election; if they have not been exposed to any candidate information and/or have no knowledge of the responsibilities of an office, it’s not hard to see why they might leave a voting choice blank. And information about local races is almost always less available than information about state races, which in turn is less available than information about federal races. But the most frustrating problem is that…

4) Non-partisan elections are bad democracy. I cannot emphasize this enough. The ballots in Fairfax County do not list the partisan affiliations of the candidates for local offices. The candidates for state offices have party affiliation listed right next to their name, as required by state law. But not local candidates. I can’t determine whether state law proscribes it for local elections or if it is county discretion – the state law seems to imply no party labels (“for elections for federal, statewide, and General Assembly offices only“), but it’s not clear. The crazy thing in Fairfax County is that most of the local officials are running as partisans, in that they are nominated in partisan primary elections. Only the school board elections and the soil and water board elections are deemed “non-partisan” and the candidates all run as independents. In any case, there is no excuse for this; in my mind, it’s a basic injustice against democracy.

As I’ve written before and others routinely blog about, party cues are not only the best available quick information for low-information voters, but they are damn good pieces of information as well. If voters could just see a D or R next to the names of local office candidates, not only would roll-off go down due to otherwise ignorant voters having all the information they need to make an informed choice, but many voters who did not roll-off would be given more information. The political parties do an admirable job of trying to hand out sample ballots to voters at the precincts that list their nominees and their endorsements for non-partisan races, but that can only accomplish so much. Putting the affiliations on the ballot would be a simple way to improve the quality of the voter signal in our elections.

5) Why are we even electing the soil and water conservation board? Not 1 in 10 people in the county even know what they do, I would bet, and no one has any clue how to judge the candidates, because they’ve never heard of them and don’t have any party cues to go on. It would be a lot easier to just have the Board of Supervisors appoint the whole board.

References

Bullock and Dunn. 1996. “Election Roll-Off: A Test of Three Explanations.” Urban Affairs Review, 32(1): 71-86.

Matthew J. Streb, Brian Frederick, and Casey LaFrance. “Voter Roll-off in a Low-Information Context: Evidence from Intermediate Appellate Court Elections,” American Politics Research, vol. 37,  no. 49 (2009).

Nichols, Stephen M. and Gregory A. Strizek. 1995. “Electronic Voting Machines and Ballot Roll-Off.” American Politics Quarterly 23(3): 300-318.