The coronavirus pandemic collided with a presidential election in 2020, making in-person voting hazardous. In response to the situation, the federal government infused its election administration with $400 million to help states and local governments buy PPE equipment for polling places, and hurriedly set up mail-in ballot infrastructures. That was a rare intervention from the federal government, marking only the third time in the past two decades it has injected money into elections, which have traditionally been a budgetary matter for state and municipal governments.
A new report from MIT’s election lab examines the federal government’s history of election spending—and suggests ways it could consider dispersing monies to help underfunded election administrators, many of whom have to deal with unsophisticated, decades-old voting equipment that risk security threats, a thin and undertrained elections staff, and minimal capacity to instruct constituents about voting procedures. The report concludes that President Joe Biden’s proposal to budget $10 billion of federal funds over the next 10 years could be a good starting point.
“The federal government not being a full partner in the game, especially given its fiscal resources, contributes mightily to the underfunding of this area,” says the author of the report, Charles Stewart III, a professor of political science at MIT, and founder of the MIT Election Data and Science Lab (MEDSL). It’s difficult to precisely gauge how much elections cost, but average estimates stand at between $4 billion and $6 billion per election year (and the anomaly of 2020 could have reached $10 billion). Though revenue sources vary by state and municipality, they predominantly fundraise through property and income taxes.
That’s often a tall order for small jurisdictions, where local election officials consistently report being cash-strapped. With their meager funds—averaging about $8.10 to spend per voter, according to a 2017 estimate Based on 26 states—they have to cover computer voting hardware, printing and posting of ballots, officials’ salaries and training, polling place rental, voter registration databases, and election night reporting systems. Equipment is one of the “biggest ticket items,” Stewart says, and the report suggests that buying new voting machines for every district in America would likely cost between $1 and $3 billion—desperately needed in some places since the hardware has a lifetime of about 10 years before it becomes outdated, facing new security threats. Another significant cost is personnel, which is harder to price out but is a sizeable tranche. “Paying people rock-bottom wages is probably not going to attract the number and the quality of people that we would want running elections,” he says.
The report suggests there are reasons for why the federal government, specifically, should take some of the cost burden. Many of the elections for which states and municipalities are responsible are federal, as in presidential and congressional; and those elections tend to attract larger voter turnouts. For example, across five states—Kentucky, Louisiana, Mississippi, New Jersey, and Virginia—turnout was an average of 50% higher in presidential years than in gubernatorial years. What’s more, the US Congress has imposed mandates on the states that include requirements for processing overseas military ballots and registration for mail-in voting, which have increased election costs.
Then there’s the matter of cybersecurity threats, which are issues of national security. This became clear in 2016, when Russia interfered in the presidential election through server hacking, email phishing, and social media misinformation. And that was a “tipping point” for many election officials, realizing it was time for more help, Stewart says. In response, in 2018, the US government issued $805 million to states and local governments, proving a federal partnership was possible.
But it’s only been a partner-in-crisis management; an “insurer of last resort” that reactively passed out chunks of cash. States and municipalities then tend to howard those funds, Stewart says, for fear that things could get worse. It’s hard for them to plan ahead if they don’t know when, and how much, money is on the way. “If the federal government would commit to a stream, [jurisdictions could] probably use the funds more wisely,” he says.
But then, how much is appropriate? The report suggests a few options for deciphering a suitable amount. One idea is what the report calls “ballot real estate,” whereby the federal government pays for however much space—“real estate”—its election listings take up on a ballot, relative to state and local elections, which would give an estimate of what the federal government could owe for the administration of its elections in any given constituency. The problem with that, the report explains, is there are likely many more local elections than federal ones on any given ballot, which would minimize the amount the federal government pays. Administrators would likely have to weigh races by importance, increasing the price tag for a presidential election, say, versus a local election.
Another concept is one the report calls “ballot eyeballs.” For presidential and midterm elections, the federal government would pay a portion of election administration costs proportional to the number of voters who only vote in federal elections, which is probably as many as a third to a half of the electorate. I personally gravitate toward the ‘eyeballs’ [solution],” Stewart says. He deems it more manageable for estimating the federal government’s financial responsibility, which he places at about $1 or $2 billion per election “and would lend itself to an appropriations rule that can be more or less uniform and easily understood.”
As part of its 2023 fiscal year plan, the Biden administration has proposed a sum of $10 billion for election administration, spread over the following 10 years, including earmarked amounts for election security and innovation. The White House is also proposing $5 billion for the US Postal Service, specifically to augment mail-in voting costs, which proliferated in 2020 and has already become the norm over in-person voting in eight states. “I think that a billion-dollar-a-year commitment is a great place to start,” Stewart offers. Election officials will probably need to ask for more down the line, but he says it’s strategically better to start small, in order to convince Congress, which may balk at bigger wads.
Whatever the price tag, Stewart says that the urgency is on for Congress to approve a funding bill as soon as possible in preparation for 2024: “Because we’re probably going to lose a window to act in the next three or four months.”