An election employee demonstrates the use of a ballot scanner in New York in 2016. (Drew Angerer/Getty Images)

Michael Windle and Matthew Caulfield were part of a team of six Wharton School students who wrote “The Business of Voting,” a report sponsored by the Penn Wharton Public Policy Initiative.

Across two counties in Virginia, last month’s primary elections brought new voting technology to the polling places, replacing some machines that had been in use for 12 years and were made with technology designed in the 1990s. This development is important not because of problems with the new machines — none were reported — but because of the urgent need for election administrators nationally to update old voting technology.

The age of America’s voting machines has been thoroughly studied, and the landscape of the voting technology industry is a major contributor to what has been deemed by an Obama-era presidential commission an “impending crisis.” Unfortunately, most election administration officials in the United States are unlikely to easily do what leaders in two counties did in Virginia last month: replace their voting machines.

A recent report by a team of researchers at the Wharton School at the University of Pennsylvania details that the market structure of the voting-technology industry is a core cause of the stagnation that keeps voters across the United States using machines more than 10 years old.

Since 2002’s Help America Vote Act, considerable vendor consolidation has brought the three largest U.S. election technology firms to reach about 92 percent of eligible voters (those who have access to some form of electronic voting machine). For comparison, the market reach held by the three largest election technology firms exceeds that of Verizon, AT&T and T-Mobile (which together hold about 84 percent of U.S. market share for wireless subscriptions) as well as Southwest, Delta, American and United (which hold a combined 68 percent of the domestic airline market share in the United States).

Additionally, election technology firms face an industry with severely limited growth (the U.S. voting population is relatively stable) and with several thousand potential customers (county and state election officials) who have limited budgets and idiosyncratic needs and desires.

Election technology firms are slow to innovate because of a lack of consistent funding for their products, and a lengthy and expensive certification process for new machines. High switching costs lead local election administrators to stay with their current vendors when making purchases or upgrades, further discouraging innovation in the market.

All this is despite voters themselves wanting newer voting machines. A November 2016 poll revealed that nearly 80 percent of Americans who voted in the presidential contest believed it was time for the United States to upgrade its election systems, and more than 80 percent thought that these updates would not only increase trust in the system, but also improve the overall election process and strengthen U.S. democracy.

Despite election administration recently being labeled a piece of our nation’s “critical infrastructure” (to the chagrin of the National Association of Secretaries of State), the United States does not have a cohesive plan or strategy to bring newer voting machines to polling places. As a result, voting technology has been updated on an ad hoc and piecemeal basis, often one county at a time, as is the case in two Virginia counties ahead of last month’s primary elections.

Potentially fruitful strategies have been identified, however. For example, significant improvements to our voting technology could come as a result of purchasing coalitions and the development of open-source systems (which is taking place in large counties such as Travis County, Tex., and Los Angeles County). Or the country’s reliance on old voting machines could be shaken via regulatory shifts toward the approval of off-the-shelf technologies and supporting modular technology over end-to-end monolithic systems.

But notwithstanding the occasional forward-thinking voting technology replacements, such as those in Augusta County (75,000 residents) and Mathews County (9,000 residents), if there is no intervention in the current market structure, swaths of the United States are destined to have old and malfunctioning equipment for many elections to come.