State of the U.S. Unions 2025
This is my annual roundup of labor data on union membership, elections and strikes. How is the U.S. labor movement doing? What should it do in this new political era?
Each year the U.S. government releases several sets of union data, covering union membership, union elections and strikes. We’ll review the data and discuss what it all means. My summary from last year can be found here.
Union Membership
In January, the U.S. Bureau of Labor Statistics (BLS) released its annual Union Membership summary. The labor movement membership rate went down again, as it does almost every year. As the BLS states,
The union membership rate — the percent of wage and salary workers who were members of unions — was 9.9 percent in 2024, little changed from the prior year, the U.S. Bureau of Labor Statistics reported today. The number of wage and salary workers belonging to unions, at 14.3 million, also showed little movement over the year.
Union membership went down 169,000 workers to 14,255,000, and the overall workforce was fairly constant, so the union membership rate dropped from 10.0% to 9.9%. The rate for the private sector is even lower at 5.9%, a slight drop from last year. These numbers are the lowest in over a century. The public sector rate also fell a bit, but remains much higher at 32.2%. The numbers for all union represented workers, which also include workers who choose to not be union members but are covered by a union contract, are slightly higher for all these.
Particularly alarming is that union membership is concentrated in so few states. Almost half of all union members are in just six states (CA, NY, IL, NJ, PA, OH), and 10 states have union membership rates less than 5% (AR, AZ, GA, LA, NC, SC, SD, TN, TX, UT). The labor movement has a very small presence in too many states, which has dire ramifications for political progress, as I’ll discuss later.
Overall, these are really alarming numbers and represent an ongoing existential crisis for the labor movement. This is especially disappointing because last year saw a big increase in organizing, which I’ll discuss more. Below is a chart of the union membership rate (also called union density) for the last century, and we see this latest decrease just continues a long term trend. Indeed, since 1965, union density has risen only five times from one year to the next.
Union Elections
The U.S. National Labor Relations Board (NLRB) released its annual summary of private sector union elections for fiscal year 2024, which covers the year from October 2023 through September 2024. I previously analyzed this data here, but will summarize the main findings.
There’s a lot of good news here. The number of RC and RM elections (see notes at the end for more on this) increased 24% from 1,320 the year before, to 1,637 in 2024. The number of union election wins rose 26%, and the union win rate rose from 76% last year to an excellent 77% this year. This continues a trend of increasing elections for the past three years after a low point during the COVID-19 pandemic. Unions have been getting better at winning elections and after many years of running fewer elections, are starting to run and win more of them. The chart below shows the trend for the last few decades.
Moreover, the last three years have seen an increase in the number of workers who won their election, with a 19% increase last year, from 78,791 to 93,517. The number of workers organized in NLRB elections last year is the highest in decades, due to a growing union win rate and an increase in the average winning bargaining unit size over time. The chart below shows this trend of workers in elections, where “Winning Workers” are the ones who won their election.
So these are some great recent trends, but this level of organizing activity is still far too low, as we’ll discuss below. It’s also important to acknowledge that there are new private sector organizing wins that come through other channels, such as achieving voluntary union recognition from employers, but even with that, overall new organizing needs to increase dramatically.
Strikes
The most prominent data set for strikes is collected by the U.S. Bureau of Labor Statistics (BLS). Unfortunately, since the 1980s BLS only tracks work stoppages of at least 1,000 workers, so it’s an incomplete picture of strike activity. A work stoppage can be a union strike or an employer lockout, but the vast majority of these are strikes.
For 2024, BLS reports that there were 31 large work stoppages that started that year, involving 271,500 workers. This is a slight drop from 33 work stoppages the year before, but well above the average of 17 for the previous decade. This is the second highest number of annual work stoppages in the last 20 years, and the fourth highest in number of workers involved.
The number of large strikes has increased in recent years, as many of us have been excited about a renewed “strike wave.” However, these recent numbers should be seen in context. In the 1950s through 1970s, there were hundreds of large strikes every year, and in more recent years, it’s usually just dozens. Here’s the trend of the average number of annual large work stoppages and workers in each decade (with the 2020s = 2020 through 2024), and we can see the huge decrease in recent decades.
The best other publicly available data set on work stoppages comes from the Cornell University ILR-School Labor Action Tracker. It tracks any work stoppages, no matter how many workers, so it’s a more complete picture of strike activity. It now has data for the last four years, finding that 2021 had 279 work stoppages, 2022 had 433, 2023 had 471, and 2024 had 359, with the vast majority of these being strikes. The number of workers who struck in these years increased from 140,000 to 224,000 to 539,000, and then dropped to 293,500 last year. The decline in work stoppages from 2023 to 2024 is largely attributed by the Labor Action Tracker to a decrease in Starbucks store strikes. A Power at Work podcast discussion of this data is here.
What Does This All Mean?
Unions had some big wins last year. This includes the UAW victory with 5,000 workers at the Volkswagen plant in Chattanooga, 27,000 Fairfax County, VA school employees who organized a union, and 10,000 nurses at the Corewell hospital chain in Michigan organized with the Teamsters. The number of organized graduate student workers continues to increase and has more than doubled since 2012. There were also gains in major contracts after strikes by the Machinists at Boeing and the east coast Longshore workers, and a strong strike vote of American Airlines flight attendants. And the Starbucks organizing continues, with over 500 store election wins in the last few years. The overall increase in union elections, wins and strikes compared to a few years ago is really encouraging and exciting.
Some more good news is that unions are extremely popular, with a 70% approval rating measured by Gallup in their annual poll, and union approval rates at 65% or higher for the last five years. Moreover, a major survey several years ago found that about half of all non-union workers want to be in a union, which is tens of millions of workers.
But all of this increased organizing and strike activity, and high union popularity, has not yet turned around the decline of the union membership rate. A major reason is that the obstacles for successful organizing remain high. It’s difficult to organize due to intense and routine employer union busting. Employers were charged with violating the law in over 40% of union elections, and they spend hundreds of millions of dollars annually on “union avoidance” consultants.
The recent NLRB has taken some significant steps to protect the right to organize, for example with its “Cemex framework” for union recognition, and its faster election timeline rules. But unions still have to find ways to scale up successful organizing within tough legal terrain. And of course the NLRB will certainly get worse under the Trump administration.
An Economic Policy Institute analysis of the recent union membership numbers says,
The disconnect between the growing interest in unionization and declining unionization rates can be explained by the fact that there are powerful forces blocking the will of workers: aggressive opposition from employers combined with labor law that is so weak that it doesn’t truly protect workers’ right to organize.
But meaningful labor law reform, like the PRO Act, seems impossible in the near term. Our political conundrum is it’s a challenge for unions to grow within the current system, but meaningful labor law reform will likely come only when the labor movement is large and disruptive enough to get it. The small union presence in many states means that elected officials there face no political consequences when voting against worker and union interests. Many major progressive political goals will be difficult if not impossible to achieve without a larger labor movement.
Let’s Go All In, But It Will Be Tougher Now
This problem is intertwined with the alarming growth of an authoritarian MAGA political culture led by Trump. Union members voted for Harris by 17% over Trump in the 2024 election. If the labor movement was much larger, with a more substantial presence in more states, that could have changed the election result.
Given the size of the overall workforce, unions need to organize over one million workers per year to meaningfully increase union density, and current organizing levels are about 1/10 that amount. A recent analysis of labor movement finances shows that unions could hire tens of thousands more organizers, and they should.
But we will never have enough union staff to reach the millions of workers who want a union. We must help train more workers to organize themselves and build their own workplace unions. The Emergency Workplace Organizing Committee is a promising model, with hundreds of volunteers successfully helping workers launch their own organizing campaigns, and connecting many of them with unions.
There should also be more interest in supporting pre-majority unions. In this framework, workers are organizing at workplaces to win improvements, but may be unlikely to get recognition and a contract in the near term, especially with a less functional NLRB. We need a massive labor movement salting program to place thousands of workers in strategic workplaces to organize new unions. And there needs to be much more striking for union recognition and to win gains. The Labor Action Tracker counted 13 work stoppages last year with a demand for union recognition, the same as the year before. That will likely increase if the NLRB ceases to provide an effective process to get recognition.
Last year I said we should go all in, taking advantage of a more favorable environment for organizing during the Biden administration, and continuing the organizing surge of the past few years. We still need to accelerate our organizing, but now with Trump and Musk, we have new assaults on unions and a potentially crippled National Labor Relations Board and Department of Labor. This will be a wild year, but we know that unions are always a critical defense against authoritarianism. MAGA can’t stop workers who are determined to organize, so it’s time to throw down, with increased action and solidarity.
Notes on Union Elections
In the past I would typically track NLRB RC elections, but now it’s necessary to also look at RM elections as well, because of the NLRB’s 2023 Cemex framework. Why? Both of these are “representation” elections that determine if the workers have majority support to join and be represented by a union. RC petitions are filed by workers and their union at a workplace when they want an election to get union certification so they can bargain a contract. RM petitions are filed by management to test whether there is majority worker support for a union.
But under Cemex, the NLRB changed the rules in an important and good way. If a majority of workers demand voluntary union recognition from their employer, the employer can no longer just ignore this and force the workers to file an RC petition for an election. Now under Cemex the employer has two weeks to either grant union recognition, or file an RM petition to force the workers to have an election. In most cases they will file an RM petition. In the past, RM petitions have been fairly rare, but now they will become much more common. In 2024, of the total number of representation elections, 1,494 were RC and 143 were RM, or about 9% of the total. In contrast, over the last decade only 1.3% of all representation elections were from RM petitions. Unfortunately, voluntary recognitions are not included in the NLRB’s annual summary, so it’s unclear how many there are. And of course this framework will likely be overturned in the new Trump administration.
