Although Joe Biden won a majority of votes cast by labor-union members, available data suggest he did slightly worse than past Democratic presidential candidates.
That’s despite the fact that Biden ran on a labor-left agenda that includes a $15 federal minimum wage, a national initiative to increase union membership, and support for the expansive Protecting the Right to Organize Act (PRO Act) — a bill that would eliminate right-to-work laws in 28 states, essentially forcing workers into unions even if they don’t want to belong. The PRO Act would also enact a nationwide version of California’s disastrous AB5 law targeting gig economy companies like Lyft and Uber, a law that California voters just rejected.
If union voters really supported these policies, presumably they would have turned out in force for Biden. And, in a sense, they did: The 2020 exit poll for the National Election Pool, a news consortium made up of CNN, ABC News, CBS News and NBC News, found that Biden beat President Trump among union households, 56-40 percent, outperforming Hilary Clinton’s 51 percent share of those households in 2016.
But excluding Clinton, Biden won less support from union households than every Democratic presidential candidate in recent history. President Obama, for example, won 58 percent in 2012 and 59 percent in 2008. On the other hand, Trump’s numbers, while not as good as in the last election, were still on the high side of what Republicans usually get.
Though polls indicate that union members see Biden less favorably than previous Democratic candidates, definitive data on the organized-labor vote are hard to get. The National Election Pool surveyed voters from “union households,” but what does that mean? That exit poll, the main for news agencies for the last several decades, does not ask voters whether they are union members. Rather, it asks voters if they live in the same house as a union member. It’s entirely possible the people being polled are not in a union and/or voted differently from the union member they live with. To give a sense of how big this potential gray area is, the most recent national exit poll found that 20 percent of voters said they were from a union household, yet only 10.3 percent of all workers are unionized. At best, the results amount to a rough guideline for the union vote.
The Associated Press’s A.P. Votecast survey does straightforwardly ask voters if they are union members as well as members of union households. Its number for union households was also 55–42 percent for Biden. So maybe the household numbers are a good substitute. Unfortunately, this is the first presidential election for which AP did its own survey, so there aren’t data to compare this year’s results to previous elections.
There is one entity that does regularly ask individual union members and how they voted: the AFL-CIO, the nation’s largest labor federation. However, it is not big on sharing that data. AFL-CIO President Richard Trumka said last week that “initial toplines from our post-election survey show union members went 58 percent for Joe Biden.” Trumka said the incumbent got 37 percent of the union vote. That would imply that both the 2020 National Election Poll and the AP’s survey understate Biden’s union support. The AFL-CIO claimed that union voters provide a “firewall” for Biden because Trump turned his back on those voters. Note, however, that Trumka said “initial toplines.” So, were the final results any different? That’s hard to say. The AFL-CIO hasn’t released their own polling data. If the numbers were really positive for Biden, why not release them?
It’s the AFL-CIO’s polls that are the main evidence for the conventional wisdom that Trump has broader support among union voters than other Republicans. Trumka told reporters after the 2016 election that Trump’s margin of union support was 3 points higher than Mitt Romney’s in 2012, while Hillary Clinton’s came in 10 points below Barack Obama’s. The AFL-CIO apparently never released the full survey results that time either. It’s also worth noting that throughout the 2016 election, the AFL-CIO released selected poll data clearly intended to show that Clinton had stronger support among its members. Only after the election did Trumka concede that Trump did better among the rank-and-file.
As limited as they are, the National Election Pool results for union households remain the best publicly available data to compare this year’s results to previous ones. They suggest that union voters liked “Blue Collar Joe” Biden slightly less than past Democratic candidates despite the fact that he ran on the unions’ official policy wish list. That enthusiasm gap should call into question union member support for a hard-left turn in national policies.