[Image via poshmark]

A New York federal court ruled last week that the state’s ban on ballot selfies – and a New York City policy implementing that law – do not violate the First Amendment and thus may be enforced at the polls, setting up a potential conflict with a similar case from New Hampshire that could eventually find its way to the U.S. Supreme Court. Rick Hasen had the news on his Election Law Blog:

In a careful 41-page opinion in Silberberg v. Board of Elections, a federal district court in New York has upheld against First Amendment challenge a New York law that bars photographing and displaying one’s marked ballot. The court held that the state’s interest in preventing vote buying and voter coercion were compelling, and that the law was narrowly tailored to meet this compelling interest. The Court relied heavily on the Supreme Court case of Burson v. Freeman, upholding electioneering-free zones around polling places…

This case will likely be appealed to the Second Circuit, and if affirmed, it would create a conflict with the First Circuit’s opinion in Rideout v. Gardner striking down New Hampshire’s ban on ballot selfies under the First Amendment. This case could end up at the Supreme Court.

The opinion in the case is notable in two ways. First, as Rick notes, it applies traditional “strict scrutiny” principles to find that the state selfie ban is consistent with the First Amendment:

These are the Court’s findings of fact and conclusions of law after a bench trial of an action to permanently enjoin a provision of New York Election Law that prohibits the showing of a marked ballot to another person as applied to the taking of a photograph depicting an individual along with that individual’s marked ballot (known as a ballot selfie) and posting it to a social media site. Also challenged is a separate policy of the Board of Elections of the City of New York (the “City Board”) that prohibits photography at polling sites, subject to  exceptions for members of the press who obtain proper credentials.

N.Y. Elec. Law § 17-130(10), first enacted in 1890 as part of the Australian ballot movement, provides that “[a]ny person who . . . [s]hows his ballot after it is prepared for voting,
to any person so as to reveal the contents . . . is guilty of a misdemeanor.” The language of this statute and the purpose for its enactment—combating vote buying and voter intimidation by depriving the perpetrator of a means by which to verify his target’s compliance—sweep within their reach the act of taking a ballot selfie at a polling place and posting it to a social media site. Posting a photograph of one’s marked ballot to social media is indisputably a potent form of political speech, presumptively entitled to protection under the First Amendment.

For reasons to be explained, the statute as applied to ballot selfies survives strict scrutiny. The State of New York has a compelling interest in preventing vote buying and voter coercion. The State’s interest in the integrity of its elections is paramount. The law is also narrowly tailored, for a law prohibiting the display a marked ballot only for the purpose of vote buying or coercion would be ineffective.

Alternatively, the Court finds that the statute is a reasonable, viewpoint neutral restriction of speech within a non-public forum.

The City Board’s policy prohibiting photography is not a content-based restriction and is not in tension with the First Amendment. This policy is narrowly tailored to address significant state interests, such as avoiding longer wait times at the polls. Narrower restrictions are unlikely to be effective. [pp. 1-2]

Second, the court uses some familiar data – polling place wait times collected under the guidance of MIT’s Charles Stewart – and associated calculations to buttress its conclusions:

Stephen Graves, Ph. D., the Abraham J. Siegel Professor of Management at the Massachusetts Institute of Technology, testified as an expert regarding a queuing analysis he performed based on a model used to simulate the amount of time voters would spend waiting in line at polling sites under circumstances where various percentages of voters took photographs of their marked ballots…

Dr. Graves based his model upon the ten New York City polling sites that serviced the most voters during the 2016 general election. Simulations run using the model analyzed increased wait times due to photographs taken by voters while they were at the privacy booths, under the assumption that most voters who photograph their ballots will do so at this time. Several variables were needed for the analytical model: “the process and rate at which voters arrive to the polling station; the service time to vote and possibly to take a photograph; and the number of . . . privacy booths in which a voter can mark a ballot.” 

The model simulated the four hours between the polls opening at 6 a.m., through 10 a.m., the busiest period of the day, and used data provided by the City Board from the 2016 general election, including the number of votes cast by hour by borough, the number of votes cast at each sample polling site, and the number of privacy booths and scanners used at each of those sites. The model approximated voter arrival times at the polling sites by simulating random arrivals, one at a time. The simulation assumed that the time each voter took to vote, measured as the time between when the voter first occupies the privacy booth and when the next voter could occupy the privacy booth, as three minutes. This estimation was partially based upon measurements made by Professor Charles Stewart of voting times in Michigan for the 2016 general election. The model assumes that the time it takes to take a photograph while voting is 18 seconds, based on measurements made by Professor Stewart. 

Dr. Graves ran several simulations with his model with different variables for the proportion of voters who took photographs: 20%, 50%, or 100%, and adjusted the time to vote in each simulation by increasing the three minute control by the percentage in each of these three alternatives.

The simulations generated random arrival times for a predetermined estimated total number voters over the time period of the simulation… The results of the model based on Frank McCourt High School in Manhattan exemplified the findings of the simulations:

The average wait increases by three minutes if 20% of voters take photographs, by seven minutes if 50%, and by fifteen minutes if all voters take a photograph. Similarly the maximum wait time goes from 61 minutes to 67 minutes (20% take photographs), to 76 minutes (50%) and to 91 minutes (100%). Finally, the percent of voters with 30 minute or more waits increases from 49% to 54% (20% taking photographs), to 59% (50% taking photographs) and to 64% (100% taking photographs).

The Court accepts that the modelling utilized by Dr. Graves does not take account of all real world considerations present in voting in New York and utilizes some data developed in Michigan which may have its own distinguishing characteristics. But allowing for those considerations, the Court comfortably finds that permitting ballot photography at the polls would materially increase wait times at New York City polling sites. The Court further finds that a material increase in wait times would likely suppress voter turnout in certain areas of New York City. [pp. 12-15 (citations omitted) (emphasis added)]

This finding, combined with the court’s other legal analysis, led to the conclusion that the ballot selfie ban does not violate the First Amendment.

Now, in some ways, last year’s ballot selfie controversies seem almost quaint given everything that has happened in (and to) the election world since then – but the underlying issues are nonetheless still fascinating. Forgive the law school hypothetical, but I wonder whether this court’s analysis holds up in the situation where a vote-by-mail voter snaps a ballot selfie at her kitchen table, given that polling place queuing problems disappear. In other words, do selfie bans continue to be “narrowly tailored” if the risk is only to the ballot and not to an entire polling place? I also wonder if the increased focus on election security (cyber- and otherwise) post-2016 alters the reasoning as well.

It will be interesting to see what happens to this case on appeal, both to the Second Circuit and potentially the Supreme Court. Hang onto your smartphones and stay tuned …