U.S. midterm results and recent federal teen-health data offer snapshots of voting and well-being, but neither establishes whether social media caused changes in political behavior or youth mental health.
Tyler Cowen, in a Nov. 5 post on Marginal Revolution, called the election results moderate and argued that improving teen indicators challenge broad criticism of social media. He wrote, “Neither the political nor the mental health trend is guaranteed to continue.”
The Associated Press reported that Democrats won control of Congress in the Nov. 3 midterms. The AP Voter Poll, based on more than 100,000 voters, found that about seven in 10 voters under 30 backed the Democratic candidate in their local House race; those voters made up about one in 10 of the electorate. Roughly two-thirds of moderates supported Democratic House candidates, compared with about six in 10 moderates who backed Kamala Harris in 2024.
AP also reported that Abdul El-Sayed won Michigan’s Senate race and described him as a progressive favorite. The reported results do not establish whether social-media exposure shaped voters’ choices.
The Centers for Disease Control and Prevention’s (CDC) 2025 Youth Risk Behavior Survey found that five of six mental-health and suicide-risk indicators improved from 2023. The share of high school students reporting persistent sadness or hopelessness fell to 33% in 2025 from 40% in 2023. It was 30% in 2015, leaving the 2025 measure above its earlier level.
About one in three students used social media every hour or more, and the CDC found no statistically significant change in that measure from 2023 to 2025. The survey tracks reported health indicators but does not establish what caused them.
A separate CDC analysis found that the proportion of emergency-department visits involving suspected suicide attempts among people ages 12–17 was 20.8% lower in 2025 than in 2021. The measure rose from 73.3 visits per 10,000 emergency-department visits in 2024 to 78.9 in 2025. These are visit-based figures, not population rates.
A randomized study assigned some Facebook and Instagram users to deactivate their accounts for six weeks before the 2020 election. Researchers found no significant effects on measured affective or issue polarization. The trial covered two platforms and one election period, so it cannot settle questions about other services, different election periods or longer-term effects. Its authors cautioned against extending the findings to other years or longer periods.
