Harold Meyerson writes in The American Prospect that the union household vote in the 2016 Presidential election skewed toward Trump according to race.
“Whites from union households preferred Trump over Clinton, 52 percent to 40 percent,” wrote Meyerson, citing a “network exit poll” whose “cross-tabulations have not been previously published” but no further source.
Much of this margin can be ascribed to white, blue-collar union members, but we cannot ignore the possibility that teacher union members and their families accounted for more Trump votes than previously estimated.
The K-12 public school teaching force is 82 percent white, and teacher union members constitute more than 20 percent of all union members in America. If white teacher union households really went overwhelmingly for Clinton, it would require an overwhelming majority for Trump from all other white union households. Is that plausible? Or did a substantial number of teacher union members, particularly in those battleground states, secretly vote for Trump?
It was clear in the primaries that union households were not lining up strongly behind Clinton, and that her margins of victory were being provided by African-American voters. But it’s impossible to tell if these turned into Trump votes in any significant numbers.
I’m not even sure a survey of teacher union members would accurately answer this question. How many members would openly admit to a union pollster that they voted for Trump?
There is a lot we don’t know about the attitudes of the union rank-and-file, and surveys usually fail to distinguish between members and households, or by race, or by public- or private-sector, or even by size of their local.
In 2005 I did discover something extraordinary about local teacher union presidents: the larger the local, the more politically liberal the union president was.
Fewer than 50 members = local presidents 44% conservative, 49% liberal
50-149 members = local presidents 40% conservative, 54% liberal
150-499 members = local presidents 34% conservative, 63% liberal
500-999 members = local presidents 26% conservative, 70% liberal
1000+ members = local presidents 14% conservative, 82% liberal
Of course, we aggregate categories to avoid being overwhelmed by permutations, but in doing so we sacrifice some understanding of the motives, strengths and preferences of smaller groups. Was union membership trumped, so to speak, by race? We still don’t know for certain.