A new working paper from various university researchers suggests that age-verification laws aren’t effective.

Age-verification laws differ by state, but typically require visitors of websites with over a third of explicit content to submit ID as proof of age. Forms of ID can range from a digital ID to facial recognition. Since 2022, 19 states have passed age-verification laws, all of which are in effect except Georgia’s, which will be in effect as of July 1.

In January, the Supreme Court heard a case about age verification, Free Speech Coalition v. Paxton. Its decision, which will likely come this summer, will impact current and future age-verification laws.

Free speech and digital privacy experts agree that children shouldn’t have access to porn, but have told Mashable since 2023 that age-verification laws won’t work for that aim. Among other reasons, some porn sites that are not U.S.-based may not feel the need to comply with the law, and people can use VPNs to pretend like they’re in another location.

Now, researchers from New York University’s Center for Social Media & Politics and several other universities looking into whether these laws impact search behavior have found the same thing.

Research suggests age-verification laws don’t work

Through analyzing Google Trends data, researchers found a 46.6 percent traffic reduction of searches to Pornhub, the biggest platform compliant with the laws. Pornhub has blocked most states with age-verification laws because of the burden of complying to the laws. In Louisiana, where Pornhub still operates while complying, traffic has dropped 80 percent, its parent company Aylo confirmed to Mashable. (Pornhub continues to operate in Louisana because the state has a digital ID that is fairly commonly used, called LA Wallet.)

While searches for Pornhub dropped in these states, researchers saw a 48.1 percent increase in searches for a large non-compliant platform, XVideos (which didn’t respond to Mashable’s request for comment), and a 23.6 percent increase in searches for VPNs. This occurred in the states with age-verification laws on a rolling timeline based on when the laws were enacted.

This shows that when it comes to internet regulation that primarily tries to impact access, there are unintended consequences, said Zeve Sanderson, co-author and executive director of the NYU Center for Social Media & Politics, in an interview with Mashable. One is the substitution effect, which undermines the potential efficacy of the policies. This is seen here with explicit websites, but was also evidenced by TikTokkers flocking to RedNote when TikTok users thought the ban was imminent.

“In Louisiana last year, Pornhub was one of the few sites to comply with the new law. Since then, our traffic in Louisiana dropped approximately 80 percent. These people did not stop looking for porn,” Aylo told Mashable in an emailed statement. “They just migrated to darker corners of the internet that don’t ask users to verify age, that don’t follow the law, that don’t take user safety seriously, and that often don’t even moderate content. In practice, the laws have just made the internet more dangerous for adults and children.”

Aylo stated that it has publicly supported age verification, but the way many jurisdictions have chosen to implement it is “ineffective, haphazard, and dangerous.” It pointed to the safety and privacy risk of requiring websites to collect a large amount of highly sensitive personal information (from people’s names to the content they consume). Like free speech advocates Mashable has previously spoken to, Aylo stated the best solution to prevent children from seeing explicit content is device-level filters.

Studying search trends for porn sites

Researchers used a method called synthetic control to analyze the Trends results. Basically, they created a “digital twin,” or synthetic version of the state to show what could’ve happened if the state had not passed these laws. To make sure their results were robust, they also used a multiverse analysis and adjusted the dates they used. Their results held.

As the authors explain in the paper, though, there are some limitations to using Google Trends to measure policy efficacy. For one, Google Trends data doesn’t measure website visits or actual web traffic. It also can’t account for actions like typing in a URL on a browser and going to a website directly.

Researchers were also unable to differentiate search results by age, which is the main component of these laws. The question of whether these laws are impacting minors’ behavior is probably unanswerable, however, because of ethical and legal concerns of exposing minors to explicit content.

Google Trends does measure search behavior, and previous research cited in the paper confirms that it is reliable for tracking population-level behavioral patterns. These researchers also conducted a correlation analysis and found a strong correlation between Google Trends volume and actual Pornhub and XVideos traffic data from SimilarWeb.

The study has yet to be peer-reviewed or published in a journal, but the authors wanted to publish it as a working paper now because the Supreme Court decision still hasn’t come down and because states continue to debate this topic, Sanderson said.

“Given how quickly this regulatory space is moving, we wanted to be able to contribute to it when it was most meaningful, not on the timeline of journal publication,” he continued. The results will be updated once SCOTUS makes its ruling.

For now, however, these initial findings propose that, as experts warned, age-verification laws don’t achieve their intended purpose.

This “first large-scale empirical analysis of these of these laws suggests that they’re likely not reaching their stated goals,” Sanderson said, “and if anything, could be incentivizing riskier behavior.”