The Limits of Poll Questions About Unions

With Labor Day in the rear-view mirror and the 2016 campaign approaching the final lap, it’s time to look at what information we have on unions and their effect on public opinion.

I would like to begin with what I believe, so you may judge my analysis accordingly.

1) Hillary Clinton will be the next President of the United States, winning by a substantial margin.

2) I find this preferable to a victory by Donald Trump, whom I like to compare with the drunk at the end of the bar who has a solution to all the world’s problems, and knowledge of none of them.

3) Trump will outperform his pre-election polling.

4) I will be voting for Gary Johnson, as I did in 2012.

The unions, of course, are all-in for Clinton, though Trump is said to have made some inroads among blue-collar workers. Polling suggests Trump’s support among union households is about the same as previous GOP nominees, while Clinton’s support is somewhat lower.

The Detroit Free Press reported “a poll conducted about a month ago by EPIC-MRA of Lansing revealed that about 30% of those who said there is a union member in their household plan to vote for Trump compared with 49% who plan to vote for Clinton. Another poll released last week by Suffolk University in Boston and provided to USA Today and the Free Press also showed Trump with the support of about 30% of union households compared with 57% for Clinton.

Polling requires lumping together groups of people who might not really belong in the same category. For example, citing a candidate’s polling numbers among Catholics makes the implicit assumption that an Irish Catholic male from Boston has the same voting concerns as a Catholic Filipina from Los Angeles. This is also true of the category of “union households.” It shouldn’t be shocking to learn that union members sometimes have political opinions different from their spouses and other members of their family.

“Are you a member of a union?” without reference to households will require more work by pollsters to get a proper sample size, but it will also give us a more accurate picture of the effect of unions on their members’ political views. But we shouldn’t stop there, especially in 2016.

Gallup’s annual Labor Day poll showed “a slim majority of Americans, 52%, say labor unions mostly help the U.S. economy, while 41% believe unions mostly hurt it,” which is “essentially back to what it was before the recession.”

But if you are going to ask about labor unions and their effect on the economy, why wouldn’t you differentiate between public-sector and private-sector unions? We all just had a three-day bombardment reminding us of all the things unions gave us – weekends, vacations, 40-hour work week, etc. How many of those things were the result of having government worker unions?

Perhaps Americans would say that public- and private-sector unions have an equal effect on the U.S. economy, for good or for ill. But maybe their answer would differ depending on which was discussed. Gallup also published a historical graph charting support for unions since 1936.

Looks to me like unions were riding pretty high until about 1965, then declined precipitously, hitting bottom in 1980. What happened during those years? Oh yeah, membership in public-sector unions more than quadrupled.

If polling the public about both public- and private-sector unions might yield interesting results, polling union members themselves about other unions might be even more fascinating. What does your average Teamster or United Auto Worker think about AFSCME or NEA? And vice versa. Considering the amount of money being thrown around on campaigns, it would be nice to know.