The “NY Times”:http://www.nytimes.com/2010/10/06/us/politics/06immig.html?_r=1&th&emc=th is reporting on a “Pew Survey”:http://pewhispanic.org/reports/report.php?ReportID=127 on likely Hispanic voting in the 2010 midterm elections. The Times leads with the headline “Latino Vote Turnout Likely to Lag, Poll Says”, which is an absolutely correct summary of the poll data _relative to the rest of the population_; Pew reports that “51% of Latino registered voters say they are absolutely certain they will vote in this year’s midterm election, while seven-in-ten (70%) of all registered voters say the same.
But is this the appropriate comparison set for gauging the change in influence of Latinos on election outcomes? After all, we know that the number of Latinos in the country is climbing relative to the rest of the population. Thus even if Latino turnout remains the same relative to non-Latino turnout in previous election, the influence of Latinos on election results should increase. Moreover, the poll also reports that Latinos are continuing to lean strongly towards Democrats: currently Pew reports that Latinos prefer Democrats to Republicans by a 65%-22% margin, which is an even larger gap than the 67%-31% preference for Obama over McCain in 2008.
So the question is, would 50% of registered Latino voters represent a decline in turnout from previous off year elections? Before we answer that, though, we have to take into account the fact that people routinely overestimate their likelihood of participating in elections. For example, the Pew Report claims that 70% of all registered voters say they will vote in 2010. In contrast, in 2006 – another “base is fired up” off-year election, “59% of all registered voters”:http://www.infoplease.com/ipa/A0781453.html participated in the election. So let’s assume we’ll get roughly the same turnout this year. That would mean that the Pew self-reported turnout percentage inflates actual turnout by 1/6. So then the correct question is, would 42% of registered Latinos participating in the midterm election represent a decline from previous years?
Someone reading this blog must have some data on the percentage of registered Latinos participating in recent elections. If you put it in the comments section, I’ll update the post with the information. One suggestive piece of evidence can be found in 2002 _American Politics Research_ “article by Highton and Burris”:http://faculty.psdomain.ucdavis.edu/bhighton/pubs-and-papers/pubs-and-papers/2002_apr_latinos.pdf. In Figure 1, they illustrate that the gap between Anglo and Latino turnout from 1976-1996 averages about 20%, although their data come from presidential, not off-year Congressional, elections. If – and this is extrapolating a bit – Latinos generally lag behind the rest of the country in turnout by 20% points, then that would suggest that the Pew numbers for 2010 are not that out of the ordinary.
Moreover, it is important to note that Latino turnout is not likely to be uniform across the country. In particular, in a 2007 _American Political Science Review_ article (“gated”:http://journals.cambridge.org/action/displayFulltext?type=1&fid=1223876&jid=PSR&volumeId=101&issueId=03&aid=1223872&bodyId=&membershipNumber=&societyETOCSession=; “ungated”:http://journals.cambridge.org/action/displayFulltext?type=1&fid=1223876&jid=PSR&volumeId=101&issueId=03&aid=1223872&bodyId=&membershipNumber=&societyETOCSession=) Barreto demonstrates that the presence of Latino candidates on the ballot increases Latino turnout. Thus a national survey is likely to miss important local variation, especially in districts where Latinos are on the ballot.
As a final note, I would like to suggest that the reporting on this survey in the NY Times may be part of the “bubble mentality”:https://themonkeycage.org/2010/10/more_on_forecasting_and_reacting_to_2010.html developing around the 2010 election. I don’t know if 51% of Hispanics planning on voting in this election is better or worse for the Democrats than in previous elections, but the article is pitched as “yet another problem for the Democrats”. It is not until the “18th paragraph”:http://www.nytimes.com/2010/10/06/us/politics/06immig.html?_r=1&th&emc=th that the article notes that “Latinos have usually voted in lower percentages than non-Latinos, but the current gap between their enthusiasm to vote and that of the general population is wider than in the last midterm election”. To me, this is the crux of the issue, and unfortunately there is no data here. Is this difference substantial? Enough to compensate for the growth in the size of the Latino electorate? Anyone with information in this regard, please contribute in the comments section below.