Aaron Smith of the Pew Internet & Public Life Project has released a new report based on Pew’s polling that provides the best current information on how we use the Internet and social media to engage in politics. These are some points that struck me as especially interesting:

• “39% of American adults took part in some sort of political activity in the context of a social networking site such as Facebook or Twitter.” That would include, for example “liking” political material (which 23% of adults have done) and encouraging other people to vote (21%).
• For most political activities–online and offline–levels of engagement are strongly correlated with income and education. The gap is also seen in the use of social media, such as Facebook and Twitter, to communicate about politics, but the disparity is smaller in that domain. “In other words, even as the impact of income on political participation is more modest in the context of social networking sites, socio-economic distinctions related to education still play a prominent role in these spaces.”
• Age matters for social networking but not for the other forms of online political engagement: “There are only modest age-related differences when it comes to getting involved in political groups or in-person activities, speaking out publicly in online spaces, or speaking out publicly in offline spaces.” But two-thirds of young adults had used social media for political communication, as compared to just 13% of seniors.
• Not only has the use of social networking sites and tools expanded, but the proportion of users who do something political with social media tripled between 2008 and 2012. That surprises me because I might have expected that as social media reached mass scale, it would draw more apolitical users.
• A key question about social networking is whether it leads to anything, because some social networking behaviors (such as clicking to “like” a post) are so quick and easy that they hardly seem to count. According to Smith, “18% of social networking site users say they have decided to take action involving a political or social issue because of something they read on those sites.” He also finds that those who use social media for political purposes are more engaged in other ways. Overall, it looks as if social media do not detract from other forms of engagement but probably contribute modestly.

I am always worried about over-reporting in surveys of civic behavior–our own as well as other organizations’. One way to test the level of over-report is to ask respondents whether they voted, because we know how many ballots were actually cast. This survey was conducted last summer, before the election, so turnout could not be measured. But respondents were asked whether they were registered to vote, and 75% said they were. The official voter registration rate in November was around 70%, but that may be an overestimate because Alaska, Colorado, Maine, Michigan and Mississippi actually count more people as registered than they have eligible resident citizens. (Citizens who have moved or died are still on their rolls.) Also, many people register after the summer of an election year. So the 75% registration rate reported in this Pew survey is high. It suggests either a bias in their sample toward active citizens or a social desirability bias that causes people to claim they engaged when they did not. These problems are absolutely typical in our business, and I am always more interested in differences within a poll than what we would call the “point estimates” (e.g., “39% of American used Facebook or Twitter for political purposes”). But if you want to estimate the rates of online civic participation, I would cut the reported numbers down by a quarter or so. They remain pretty impressive.