When It Comes To an Amazon Union, Voters Side With Workers

By Prerna Jagadeesh and Devi Ruia

Last month, all eyes were trained on an Amazon warehouse in Bessemer, Alabama, where workers participated in a historic unionization effort to secure better working conditions at the infamously brutal Amazon facility. The union drive failed, in part because of union-busting tactics used by Amazon against its workers — including threatening them with losing their jobs, interfering with traffic patterns to stop workers from talking with union organizers, spreading misinformation about the union, and hiring a detective agency to spy on employees. Yet even though the unionization effort did not succeed, new polling by Data for Progress shows that the American people are very much on the workers’ side — and that Amazon’s anti-worker tactics have made a negative impression on people’s opinions of Amazon and their willingness to buy from it.

In a national survey of likely voters, Data for Progress gauged voter opinions on Amazon, the Bessemer union drive, and union legislation. Data for Progress also oversampled voters in Arizona in order to ascertain the attitudes of Senators Kyrsten Sinema and Mark Kelly’s constituents towards the Amazon union drive and unionization, as Senators Sinema and Kelly don’t yet support legislation that could have radically changed the outcome in Bessemer: the Protecting The Right To Organize (PRO) Act. Data for Progress polling found that voters nationwide support the PRO Act by a strong 30-point margin, and Senators Sinema and Kelly’s constituents shared this strong support for the PRO Act by a margin of 22 points

 
image1
 

Data for Progress also tested how voters in Arizona and nationwide feel about Amazon in light of its union-busting tactics. Since February 2020 alone, at least 37 charges have been filed with the National Labor Relations Board (NLRB) against Amazon across 20 different cities. We found that voters who learned about these union-busting tactics became markedly less favorable towards the company: when thinking about Amazon running afoul of NLRB standards, voters hold an unfavorable view of Amazon by a 41-point margin. With voters in Arizona, that margin of unfavorability increases to a 47-point-margin.

 
image2
 

Voters don’t just disapprove of Amazon in the abstract: their opposition to its union-busting tactics compels them to vote against Amazon with their wallets. After learning about tactics used by Amazon in response to the unionization effort in Bessemer, Alabama, national voters say they were less likely to buy from the Amazon site by a 33-point margin. Arizona voters say they were less likely to buy from Amazon by a 28-point margin. 

 
image3.png
 

And although the Bessemer union drive failed, our polling shows that voters are very much on the size of Amazon workers in their efforts to secure better working conditions through unionization. Voters support unionization efforts by Amazon employees by a 37-point margin, even after knowing unionization efforts in Bessemer, Alabama failed. Arizona voters support Amazon workers across the country unionizing by a 22-point margin.

 
image4.png
 

Amazon is an immensely powerful company, with vast resources at its disposal and an effective monopoly over American consumers; nonetheless, when ordinary Americans are made aware of how Amazon used threats, surveillance, and intimidation to stop its workers from exercising their right to form a union, voters heartily disapprove of Amazon and are considerably less likely to do business with it. Voters in Arizona especially support the PRO Act, and the onus is on Senators Sinema and Kelly to deliver on behalf of their constituents and vote in support of the PRO Act.


Prerna Jagadeesh (@PrernaJagadeesh) is a writer at Data for Progress.

Devi Ruia (@DeviRuia) is a press assistant at Data for Progress.

Methodology:

From April 21 to 25, 2021, Data for Progress conducted a survey of 1,730 likely voters nationally using web panel respondents. The sample was weighted to be representative of likely voters by age, gender, education, race, and voting history. The survey was conducted in English. The margin of error is ±2 percentage points. This survey included oversamples of likely voters in Arizona and West Virginia. The results of this oversample are reported here. 

NB: subgroups with a n-size less than 50 (<50) are not shown on these cross-tabs. We choose not to display N<50 subgroups because the sample is too small to have statistical significance. We did, however, take samples of these subgroups for representational and weighting purposes to accurately reflect the electorate makeup. Some values may not add up to 100 due to rounding.