Do political exit polls misrepresent Latinos and other voters of color?
So argues Stanford University political scientist Gary Segura today in a piece on the Latino Decisions website; the polling firm, in which he is a principal, has been keeping tabs on the Latino voter climate in the runup to next week's election.
Segura points to language as one problem that can affect exit poll tallies on election night, and how inaccuracies tend to prevail within smaller, geographically concentrated groups of ethnic voters. In addition, he writes the exit polls tend to over-represent people of color who are middle-class and better educated, and this also affects results. He points out some previous unusual exit poll numbers, for example, Arizona Gov. Jan Brewer "attracting an above-average 28% share of Arizona’s Hispanic vote just months after signing SB1070 into law," according to the National Exit Poll from 2010.
What do do? In his post, Segura outlines three questions for reporters using exit poll data to keep in mind:
1) How were precincts selected? What criteria apart from “probability proportionate to size” were used? How were “national” precincts selected from the state samples? What did NEP do in the 19 states where there are NO state level studies?
2) What share of Latino interviews were done in Spanish? If that number is much less than 25-30%, how can your poll be accurate if between a quarter and a third of Latino voters are Spanish-dominant?
3) How reflective is the exit poll of the true population of Hispanic/Latino voters?
It's a detailed and quite interesting post; read it in its entirety here.