The Industry

Why Union Organizers Think They Got Creamed by Amazon

It wasn’t just because Amazon allegedly tampered with mailboxes and traffic lights.

An Amazon-sponsored billboard urging employees to return their unionization ballots.
Did union organizers blow it against Amazon? Elijah Nouvelage/Getty Images

It’s hard to see the results of the Amazon union vote in Bessemer, Alabama, as anything other than a creaming. The National Labor Relations Board revealed on Friday that workers at the warehouse voted 1,798 to 738 against joining the Retail, Wholesale and Department Store Union. The 1,060-vote margin of victory for Amazon, which had played hardball against the union drive, has labor experts, pro-union workers, and journalists questioning what exactly led to such a one-sided result.

The Bessemer drive was the most visible and forceful attempt in Amazon’s history to organize one of the company’s workplaces, none of which has ever unionized. The campaign in Alabama started with a group of workers contacting the RWDSU last summer about unionizing in light of grievances about Amazon’s protections against COVID-19 and grueling productivity demands. It soon blossomed into a national debate, with prominent figures like Stacey Abrams, Bernie Sanders, and Danny Glover endorsing the union effort. President Joe Biden even released a video tacitly supporting the pro-union workers. All the while, Amazon treated the drive like an existential threat, subjecting workers to mandatory town halls about the supposed dangers of unions, allegedly pressuring local officials to change the timing of traffic lights near the facility to frustrate organizers’ outreach efforts, and constantly texting workers with anti-union messaging. Mail-in voting ran from February to March, and in the end, workers rejected the union by more than a 2–1 margin.

There have been a number of postmortems written about why RWDSU was ultimately unable to win over a majority of the workers at the Bessemer plant. One of the prevailing theories is that the lack of house calls by the organizers significantly diminished the chances that workers would vote in favor of a union. A house call involves an organizer physically visiting workers’ homes to sell them on the benefits of a union. This method of canvassing has the advantage of assuring people that their voting decisions aren’t being monitored by their employers, since the interaction takes place in a private space. However, social distancing recommendations during the pandemic made visiting workers at their homes much more difficult, so the RWDSU focused more on social media outreach and calls to personal phones. This kind of virtual campaign still pales in comparison to the sort of in-person messaging that Amazon was delivering to its workers through its town halls at the warehouse. Despite the strictures of the pandemic, the union could have chosen to show up at voters’ homes equipped with PPE, just as Biden’s campaign had in the last months of the 2020 presidential election. Stuart Appelbaum, president of the RWDSU, maintains that these sorts of visits nevertheless would’ve been a bad idea. He claims that Amazon warned people in its town halls that there might be house calls and used it “as a reason that people needed to be wary of the union, and as a failure of the union to respect workers during the pandemic.” Appelbaum was worried about feeding into this argument.

Another likely major factor in the union’s defeat was the size of the bargaining unit. When the RWDSU filed paperwork to hold the union election, it originally stated that it was seeking to represent 1,500 workers. Due to a pandemic-related hiring frenzy and a move to include truck drivers and seasonal workers, Amazon pushed the NLRB instead to make 5,800 of the workers eligible to vote. This meant that the union had to reach out to thousands more people than it had been planning to, a challenge that may have diluted the support it originally accumulated among the original 1,500 people.

The timing and timeline of the drive have also been under examination. Amazon was able to reach workers early on during the election with compulsory meetings and via posters plastered throughout the facility, including in the bathrooms, pushing people to vote no. Typically, a union tries communicate with voters before their employer can in order to “inoculate” them against misleading and sensationalistic claims about the dangers of organizing. The RWDSU itself even admitted halfway through the election that people who submitted ballots early on likely voted no, because Amazon had reached them before the union could. The company also pushed workers to vote as soon as possible after the town halls in an attempt to ensure that the union’s talking points wouldn’t influence them. Even the endorsements from Biden and other leaders came fairly late in the election process.

So why couldn’t the union reach workers earlier on? Appelbaum attributes this to the difficulties of determining who exactly is working at the plant at any one time, even though it’d only been open for a year. Burnout at Amazon is high, and the union believed that the Bessemer facility had an annual turnover rate of at least 100 percent, indicating that more employees leave or are fired per year than the average number of people working at the warehouse. This made it at times unclear whether people the union had previously won over were still working there. It was also difficult to determine who exactly the recently hired employees were. The RWDSU did end up receiving a list of people who were working at the plant right before the voting period, but Appelbaum said it was too late and left the union to play catch-up with Amazon. “It took some time for a facility this large,” Appelbaum said of efforts to identify and contact eligible voters. “Amazon had complete access and a mandatory, captive audience several times a week an hour at a time where people were lectured on why they had to vote against the union.”

When asked if there was anything at all that the RWDSU could’ve done differently to maximize its chances of success, Appelbaum replied, “It took us too long to develop some momentum. We’re pushing people. It took a little longer than we had hoped for.” The high turnover rate meant that the union had to move fast; it would’ve been difficult to ploddingly build support within the ranks. The union may get a chance to continue riding that momentum, though. After it was clear that the vote was going to go in Amazon’s favor, the RWDSU announced that it would be filing objections to the NLRB about how the vote was conducted. One of its main objections has to do with a mailbox that Amazon pressed the U.S. Postal Service to install at the Bessemer facility around early February. The NLRB had already denied a previous request from Amazon to set up a drop box for votes at the facility, and the union argues that the USPS box’s presence may have intimidated workers by suggesting that the company was involved in the collection and tallying of votes.

Shortly after the vote was finalized, Amazon asserted in a blog post, “It’s easy to predict the union will say that Amazon won this election because we intimidated employees, but that’s not true. Our employees heard far more anti-Amazon messages from the union, policymakers, and media outlets than they heard from us.” If the RWDSU can convince the NLRB that Amazon violated labor laws by installing the mailbox or by engaging in any other illicit tactics, there’s a chance that the results will be thrown out and another election held. It seems difficult for the union to win in a do-over given just how lopsided this vote was, though Appelbaum thinks that more workers are now aware and motivated to organize given that the drive was able to gain some steam in the latter portion of the voting period. “We keep building now. We haven’t gone away,” he said, noting that the RWDSU held a rally in the area on Sunday. “We all understood that this was going to be a long struggle, and we were prepared for it.”