Immediately after the November elections, using the National Election Pool, national exit poll by Edison Research, Census demographic data, and the number of ballots cast nationally, CIRCLE estimated youth turnout to be 20-21 percent. On November 15, CIRCLE received adjusted (“reweighted”) exit poll data from Edison Research. The new data suggest that young voters represented 12 percent of all voters, instead of the 11 percent that was initially reported by Edison Research. That change would, in turn, raise CIRCLE’s turnout estimate to 22.8 percent of eligible citizens under 30 (instead of the 20-21 percent that we initially estimated). That turnout rate would be similar to the levels seen in 1994 and 2006, and higher than those seen in 1998 and 2002. The rise since 2002, however, is not statistically significant.
In releasing our turnout estimates immediately after the election, we emphasized that youth turnout “was fairly typical for a midterm election”: well within the normal range. That clearly remains the case. A reweighting of the exit poll data more than one week after the election is unusual, and we are not able to assess independently which results are most accurate. A more precise estimate of 2010 youth turnout will be available once the Census releases its November Current Population Survey voting data in the spring. Meanwhile, the most reliable conclusion is that youth turnout (as estimated by the CIRCLE exit poll method) has stayed between 20 percent and 23 percent in all midterm elections since 1994. Any changes are within the margin of error.
A revised fact sheet on youth turnout in the 2010 election (using the re-weighted data) can be downloaded from here: