In 2018, the number of union elections fell to nearly 1,000, the lowest level in 80 years, almost the entire era of the National Labor Relations Act
The labor movement has been in decline for decades and its traditional strategy to revive and grow has been organizing around National Labor Relations Board (NLRB) union elections. How is that going?
The NLRB election data for fiscal year 2018 is out and I took a look at the last five years to see recent union elections results. Here is the number of RC elections (where the union files for representation and official certification) and the union win rate. I’m only looking at RC elections because I’m interested in the elections that unions plan as part of their organizing programs, and the other types of elections are relatively small in number and not interesting for this analysis.
In recent decades the number of elections have fallen dramatically, from a high point of over 8,000/year in the 1970s to the low thousands in the 1980s and 1990s, and now just above 1,000 in recent years, as I discuss in this article. The average for the past five years has been 1,259 elections/year. In 2018, it was 1,055 which is an alarmingly low number. I looked back at early NLRB annual reports and you have to go back to the very early days of the National Labor Relations Act (NLRA) era in the late 1930s to see the number of elections this low, and that’s when the election machinery was first being set up. Here’s a chart from the 1945 NLRB annual report showing the first 10 years of union elections and we can see that after 1940, the number was thousands/year.
What’s interesting, and better news, is that the union win rate has held steady at around 70% for the past five years. In that article I discussed how the union election win rate was much lower in the past, from around 60% in the 1950s and 1960s, dropping to around 50% in the 1970s through 1990s, and then rising to around 70% today. Unions are much better at winning these elections today.
This higher win rate is in spite of the fact that unions have also had to deal with an increased number of Unfair Labor Practices (ULP) from employers over the years. It’s routine for employers to run a scorched earth union-busting campaign of fear and intimidation during an election. The article discusses that data.
My interpretation of the smaller number of elections, higher number of ULPs, and the higher win rate is in part that each election win requires more work, unions need to concentrate their resources on fewer elections, and they are largely winning them.
One important consequence of this smaller number of elections is the much lower number of workers voting in them. This chart shows the number of eligible voters for all elections and the average bargaining unit size for the past 5 years.
The average number of eligible voters has been about 81,000/year and the average bargaining unit size has been about 64 workers. This is a very small number of workers to have the opportunity to vote to form a union. Contrast that with the hundreds of thousands of workers who would vote every year in decades past.
According to Bureau of Labor Statistics data, union density, the percentage of all workers who are union members, is a low 10.7%, and an even lower 6.5% in the private sector. With about 136 million workers, we need to organize over 1 million new union workers every year to increase union density by close to 1%. In an election strategy, we would need to run over 10,000 elections/year to come even somewhat closer to that. That’s clearly not going to happen.
As I’ve stated in past analyses, we are unlikely to slowly grow our way back to higher union density through NLRB elections. With elections declining to numbers seen at the very beginning of the NLRA era, we have perhaps come full circle with this strategy. And let’s remember that left critics of the NLRA when it passed worried that corralling unions into government sponsored elections and certifications would domesticate and weaken the labor movement over time. It seems they weren’t wrong.
This is why many have come to the conclusion that the NLRB union election strategy is a dead end and cannot be the principal basis for union revival. Other alternative ideas have arisen. In recent decades unions have also organized with “card check” campaigns where the goal is to get recognition based on a majority of signed union cards rather than an election. But this of course still usually requires a long and resource intensive shop by shop fight with the employer. Other strategies are workers’ centers and campaigns to fight for better wages and working conditions such as the Fight for 15. The Industrial Workers of the World (IWW) has long had a strategy of “solidarity unionism” where workers simply organize and fight for better conditions without trying to win elections and bargain official contracts. The organizing.work site has case studies from these campaigns. Some folks have argued for a change in labor laws or that we should fight for European-style sectoral bargaining. Labor Notes is a good source for news and debates about various union strategies.
The fight for labor movement revival will continue to develop multiple forms but certainly must involve mass action and disruption among millions of dissatisfied workers who are trained to fight for their rights at work. There are some promising signs for this renewal, for example, this year has seen the highest number of large strikes in decades. The labor movement must continue to envision such a large scale program.