Gun ownership in America has remained largely stable, according to a poll released earlier this month.

Roughly three-in-ten Americans report personally owning a firearm, according to a Pew Research poll. Another 11 percent of Americans say they do not own a gun personally but share their household with someone who does–meaning roughly 106 million American adults report having a gun in their home.

This data mirrors previous results from 2017, where respondents gave identical answers to personal and household ownership questions. The most recent polling from Gallup found an increase from 2019 to 2020 but also produced numbers overall consistent with data over the last several years.

These results are surprising in light of previous reporting on unprecedented gun sales throughout 2020 and into 2021. The National Shooting Sports Foundation (NSSF), the firearms industry trade group, conducted a retailer survey in 2020 that determined 40 percent of sales during the surge were to new owners. A 2020 University of California study also found 40 percent of Californians who bought a gun because of the pandemic were first-time buyers.

There were 21 million FBI background checks on gun sales in 2020. Those checks are widely considered to be the best measure of gun sales. If, as the NSSF and University of California data suggest, 40 percent of those sales were to first-time buyers, that would equal about 8.4 million new gun owners. That represents roughly three percent of the adult population–a portion large enough that the increase should show up in public polling.

The stability in polling numbers, despite this reported spike in new gun sales and evidence many were to new owners, highlights the difficulties in capturing exact measurements of gun ownership in America. Federal law prohibits creating a national database of gun owners due to concern by many gun owners it could be used for nefarious purposes. Instead, FBI background check data and survey estimates are often used as substitutes, but they can make for imperfect proxies.

The FBI says background checks do not represent a one-to-one count of gun sales because some sales are not captured by the system, such as those made to gun-carry permit holders in certain states. Experts who spoke to The Reload said surveys, including the one conducted by Pew, present their own set of issues. The experts said the polls often undercount the level of gun ownership in America for several reasons.

“Surveys underestimate the absolute level of gun ownership, mainly because gun owners are afraid their admission of gun ownership could somehow lead to the authorities seizing their guns or even arresting them for unlawful possession,” Gary Kleck, a Florida State University professor with decades of experience studying gun ownership, said.

He said this tendency could lead to fluctuations in how many Americans report owning a gun from survey to survey but generally doesn’t make the data unusable. He cautioned that significant pushes for new gun bans could have a noticeable impact on the short-term reliability of gun-ownership surveys.

“Problems arise, however, when gun owners become more anxious about impending additional gun controls, as happened when the federal assault weapons ban was passed in 1994 – suspiciously large, sudden drops in survey-reported gun ownership occurred in national polls,” Kleck said.

But the experts also question how much is actually known about the number of new gun owners in the country.

Robert Leider, a George Mason University law professor who has written extensively on guns in America, said that while some gun owners may be unwilling to disclose that information to pollsters, the poll numbers are “likely not far off from the correct totals,” and questioned the veracity of data on new gun owners.

“I don’t think we have good data on first-time gun buyers,” Leider said. “The NSSF data comes from a retail survey which is not the most scientific source.”

Kleck agreed with Leider’s assessment of the NSSF survey. He also said the University of California study “only covered California, and its results cannot be generalized to the U.S. as a whole.” He said that because of California’s stringent gun control regime, “some respondents to [the] California survey may have been reluctant to admit to prior gun ownership given uncertainty about its legality.”

The two men said the current evidence suggesting gun ownership has expanded in recent months isn’t strong enough to assess exactly how many new gun owners were created over the past year and a half. They argued more research is needed on the topic.

The Pew poll was conducted from June 14-27. It surveyed 10,606 adults, and the margin of error was 1.5 percent.