What worked and what didn’t: A history of organizing at Starbucks

In this two-part series, Nick Driedger takes a look at previous attempts to organize Starbucks. This installment covers the Canadian Auto Workers’ campaign on the lower mainland of British Columbia, 1996-2007.​​

With organizing at Starbucks once again making the news, we should brace ourselves for another round of excitement about “new and novel” organizing approaches. These often include media-heavy strategies and attempts at certifying one shop at a time in small units of only a few dozen workers. For those of us who cut our teeth organizing at places like Starbucks, it pays to revisit these previous attempts, to assess their effectiveness, and look at the debates that happened back then. What follows is an in-depth, nuts-and-bolts study of exactly what happened in two previous campaigns, what lessons were learned, and how the organizers involved changed their approach based on what worked and what didn’t.

One organizing drive was run by the Canadian Auto Workers (CAW) on the lower mainland of British Columbia, from roughly September 1996 to April 2007, when they voluntarily let their certification lapse at the stores due to there being no actual union presence there. The other was started by the Industrial Workers of the World (IWW) in 2004 and spanned various cities across the United States and Canada before petering out in 2017. 

The CAW was a Canadian private sector union, originally a split from the United Auto Workers in the US, that years after the Starbucks campaign merged with the Communications Energy and Paperworkers Union (CEP) to form Unifor. The CAW had a long tradition of progressive unionism and would consider their unionism a kind of social movement unionism. The IWW is a union of a few thousand voluntary members with few certifications and a younger-skewing membership which tends to work in a lot of lower-paid service jobs like the kind you find at Starbucks.

The two campaigns covered a similar amount of time — 11 years for the CAW and 13 years for the IWW — and had a similar membership: the CAW had about 150 workers paying dues in shops at the height of their presence while the IWW had between 100-200 members at its peak. 

The CAW ran their campaign like they would any other drive: they identified issues and ran certification elections and then bargained contracts at the workplace. By contrast, two early failed attempts at labor board certification had the IWW’s Starbucks Workers Union (SWU) turn their back on elections. The SWU signed Starbucks workers up anywhere the company had a presence, not limiting themselves to what would build a viable legal bargaining unit. They then mostly used direct action at work to pressure the employer to settle on issues of concern.

This first installment will look at the CAW drive and what it shows about the limitations of Wagner Act unionism – the dominant form of legal unionism since the passage of the National Labor Relations Act in the US in 1935. The second installment will look at the IWW’s campaign and assess the strengths and weaknesses of its “solidarity unionism” approach. 

A common union-avoidance strategy from Starbucks

The challenges the two unions faced were similar. Both complained that high turnover and a relentless employer presented difficulty in building the campaign.

The union avoidance strategy on the part of Starbucks was consistent between the two campaigns despite the CAW and IWW having very different organizing strategies. In both cases, Starbucks used the small workplace and socially tight-knit workforce to their advantage.

For example, a barista from the IWW campaign in Connecticut reported: 

May 2nd 2011, [name redacted] led a light version of a march on the boss at our store meeting. Workers were extremely polite, well-spoken but direct about the issues of scheduling and the manager’s attendance record, which was more than she could handle. She flipped out, cried and ended the meeting, which was extremely off-putting to everyone in attendance.

[An IWW member] did follow-up with coworkers immediately after the meeting to encourage and reassure them, as well as on the days that followed. [The] Manager was obviously aware of how much of an idiot she made herself look and caved on our demands, but the damage to peoples’ boldness was done.

IWW Starbucks Workers Union national conference call notes, June 11, 2011

John Bowman, the main union representative on the CAW campaign, reported that when they did their organizing at Starbucks on the other side of the continent: 

[The] anti-union campaign played on the personal relationships between the manager and employees and the image of staff as “partners.” It was very personal. Managers would meet with employees and express that their hearts were broken.  “[You] Should have come to us about problems instead of inviting outsiders” kind of approach. Some people did rescind support as a result. They put out some written materials — not a ton, but some — giving their official response. They generally argued they could resolve the problems internally, more on a personal [level].

Author interview with John Bowman, February 28, 2020

There may well have been some genuinely hurt feelings. Starbucks at the corporate level may train its managers to think of themselves as having real relationships with employees. Either way, Starbucks used those feelings to push an anti-union sentiment. In a workplace made up of small stores where workers deal with each other in close quarters and see management several times a day in small numbers, this kind of emotional pressure was often a useful tool for isolating union activists and cutting the union off. As one fired worker in Atlanta reported: 

[SWU member] has lost support from coworkers because the company has retaliated against people who associate with him

IWW Starbucks Workers Union national conference call notes, October 26, 2010

Small shops present a special challenge for organizers that large shops don’t, in that open conflict can be very stressful and wear people down over time due to it being much more personal and direct. There is no escape from the conflict that union activism brings when the workplace is so small you have to interact with the management constantly every day. Often management doesn’t need to change people’s minds so much as make them feel negative emotions and get tired out.

Both campaigns also faced the challenge that Starbucks is not a mostly franchise operation like McDonalds; locations are owned by one big company which mean they can bring a lot of weight to bear on one store or even city to drive the union out.

In the face of these challenges, the CAW and IWW took very different organizing approaches. 

Autoworkers on bar

In September 1996, the University of British Columbia campus newspaper The Ubyssey reported that an organizing drive had started at Starbucks with six stores submitting authorization cards: 

Employees at six of the American coffee giant’s Vancouver outlets signed union cards for the 200,000 strong Canadian Auto Workers union, and will go before the BC Labour Relations Board for certification this Friday. UBC political science student Steve Emery, who works at the Hornby street location across from the Courthouse, initiated the movement. “Any multinational corporation that makes as much profit as Starbucks does and pays its employees minimum wage has to expect that their employees should not be satisfied,” he said.

In the same article, Bowman explains what was in it for the CAW:

“Any union that wishes to continue to be a major union over the next decades has no alternative but to organize workers in the sector” he said. “it is pretty clear the trend is job creation in the service sector and away from manufacturing and resource sectors”. 

This is a fairly candid statement from the union that they need to maintain their membership in order to continue being a social force. Doing so is a matter of diversifying in order to survive the decline of their core industries in the automotive manufacturing sector. 

The workers didn’t initially pick the CAW either. Falling in with that union was largely an accident. They were initially looking at just forming their own union — something that was not unusual in the pre- Wagner Act days — but the staff at the Labour Board discouraged them, explaining how difficult and technical the process is. Laurie Banong, a Starbucks worker and organizer with the CAW campaign, in an interview with the Canadian labor studies journal Labour/Le Travail a year after certification,  explains:

We went to the old labour board office on Howe Street [in Vancouver]. We wanted to see if we could organize our own union because I didn’t think there was such a thing as a service sector union. Because I thought, you know, I hadn’t heard of any company of this nature that was unionized. So I thought I had to go and start my own…. We went into the [labour board office] and talked to a secretary, realized that [starting our own union] was nuts, and on the way out we ran into an organizer from the CAW [Canadian Auto Workers] who was there on business. Because we didn’t know what to do, he overheard us talking, he’s like, “Here’s my card.” … And so Steven and I talked to him for like an hour just that same day. And I’m like, “This sounds pretty ok.” And then he [the CAW representative] talked to everyone, then we got everyone else together from the store…

“Barristas [sic] of the World, Unite!” Labour / Le Travail, Volume 42 (Fall 1998), 335-343.

The organizing drive relied heavily on the workers taking on the work of signing people up — it was not staff-driven. Bowman would mentor and lead the workers through the organizing process, but explains it was on the workers on the job to bring each other together and to get the cards signed: 

We had meetings in groups with small groups store by store. We put the onus on the people inside to obtain membership cards. We had the initial meetings from a few stores. They would decide then if they would take cards. 12-15 employees per store, they were fairly fast campaigns — 2-4 days/store. We initially applied for 2-3 stores at once… Workers on the job were expected to keep signing people up as the new hires came on; that didn’t happen much. Starbucks did not hire a lot of new people upon certification. After the first few stores they became more aware and ran an anti-union campaign.

Author interview with John Bowman, February 28, 2020

Key to the CAW’s initial success was pointing to other contracts that they had negotiated in the service sector. They had collective agreements at White Spot restaurants and Kentucky Fried Chicken. Bowman explained to workers that the goal was to get similar contracts at Starbucks.  Banong echoes that wages were a key issue for signing authorization cards, although as an organizer putting in the volunteer work, it was a personal sense of empowerment that drove her forward: 

LB: We had our little song and dance, “We want dignity. We need lives, blah blah blah.” But then we said we are going to ask for ten dollars an hour. That’s what got most people signed up.

 L/LT: Is that what did it?

 LB: I think that’s the biggest factor. I’m like, “You can make more money, it’ll be easy, if we all work together.” It was basically just the dollar signs that got their interest. Other people had personal vendettas against shitty managers. And wanted to see them crushed — that was one of my motivations. “I’m gonna make you pay.” But we’ll say it was a moral victory.

“Barristas [sic] of the World, Unite!” Labour / Le Travail, Volume 42 (Fall 1998), 335-343.

Bargaining and leverage

In this kind of drive, the union develops by moving through various stages that are outlined in legislation. The first step is “organizing,” which in this model means signing authorization cards (cards that are used as proof of majority support in the shop). British Columbia at this time had card check certification, a staple policy agenda of the left wing of the Democrats in the US and of the New Democratic Party in Canada. Under this system, workers sign union authorization cards, and the card becomes the vote for the union. In the US, where there is no card check certification, the card is treated as a signature on a petition requesting an election on whether or not to unionize. In jurisdictions in Canada with card check certification, it’s a fast process and often happens before the boss knows there is a union campaign. This limits the boss’ ability to run a counter campaign to try to intimidate workers into not signing cards (or voting no). The CAW had a very developed organizing program that could sign cards very quickly in a member-led process. 

After the authorization cards are signed comes “recognition.” This step is where the union wins the legal right to represent the workers and the employer is compelled by legislation to address their concerns and sit down to bargain with the union. 

Unlike the process of organizing, which is often led by members, and certainly was in the CAW’s Starbucks organizing, the process of bargaining is very technical. It’s the process of creating a document legally binding on both parties. Membership is always at the table but technical staff tend to play a key role. So while there was a lot of delegation to the workers to get cards signed, which involved getting people together to discuss complaints they had about their pay and conditions, bargaining went very differently. As Banong explains: 

LT: Once the CAW got involved, how much did the grassroots stay involved in the process and how much was it taken over by people who are professionals?

 LB: We would sit at this big, huge table as long as this room. And there would be the ten of us on the union side and four guys from Starbucks. And just … our representative, and their lawyer would talk to each other the entire time. And sometimes we would pass notes, or lean over and whisper. The entire session was them just kind of arguing back and forth. A lot of it was machismo…. Like we were the fire hydrant and they were like circling us. They kind of strutted their stuff. Back and forth for a while. But the thing that really dragged us down was that they refused to talk money issues, any kind of money issue, until every other issue had been decided.

“Barristas [sic] of the World, Unite!” Labour / Le Travail, Volume 42 (Fall 1998), 335-343.

As Banong indicates here, while bargaining is a very technical process, it is fundamentally about power. This contradiction is key to understanding the problems with the Wagner Act system. In order to win at the table, you need leverage; in order to get leverage, you need the workers to be confident in taking on the boss; in order to achieve that, they need to lead their own struggles. The technical process at the table undermines that: the workers are not deploying their own power and learning from their mistakes, instead they are advising union staff who lead the bargaining process. 

Banong describes how bargaining played out:

We were all demoralized, we were all depressed, we were all bored senseless…. It was like a siege. We were slowly being starved out. We started organizing in October [1996], we started bargaining in late November, December [1996] and we did not get done until July [1997]. Six months of bargaining. At that point we’d lost members of our bargaining committee, we’ve got staff turnover during all this time, they thought they could just stall us. It could’ve worked. But we wouldn’t let them. But by the same token, I also knew that we wouldn’t win if we went on strike either. Because, we could walk out, we’d be locked out forever.

“Barristas [sic] of the World, Unite!” Labour / Le Travail, Volume 42 (Fall 1998), 335-343.

Generally the point of bargaining under the Wagner Act system is to use the threat of a strike or a lockout to break an impasse. Bowman explains that the all-out strike was not used on the understanding that it would not be that effective:

The base reality of the whole thing, and it’s a problem with all multi-location employers, [is that] we did not have enough stores to succeed at negotiating the kind of collective agreement to get what we wanted. We could have struck the stores we had and kept them out forever but we would not have gotten much more than we did. That was the problem we faced.

Author interview with John Bowman, February 28, 2020

The inability to mount a strike made it difficult to meaningfully bargain because there was no serious threat to the company’s bottom line. What kept them from being able to strike was having only 12 of 50 stores but also being very thin on the ground in the stores they had organized. This made a strike less than practical.

Instead, the union engaged in “unstrikes.” Says Bowman:

[With the] unstrike we had them not wear uniforms, didn’t do any overtime — it was basically a work-to-rule. The employer moved a little bit because of it at the bargaining table. It did not get the movement we would have liked. 

Author interview with John Bowman, February 28, 2020

According to a labor board report, the “unstrike” did manage to put some pressure on the employer: 

During collective bargaining for the first collective agreement Starbucks advised the Union that it had done an extensive market analysis and learned that its employees’ wages and benefits were the highest in the industry.  It took the position that no increases were warranted.  After a three month ‘un-strike’, Starbucks agreed to increases. 

Certain Employees of Starbucks Corporation v. National Automobile, Aerospace, Transportation and General Workers Union of Canada, Local 3000, 2001, CanLII 33075 (British Columbia Labor Relations Board)

Banong fills in the details:

L/LT: How did you fare on the wage question?

 LB: We got the starting wage to $7.75.

 L/LT: Up from?

 LB: $7.50.

 L/LT: Were you disappointed? 

LB: Yeah. But we also know this is our first contract. Now that we are guaranteed the hours, we’ll get more money….

“Barristas [sic] of the World, Unite!” Labour / Le Travail, Volume 42 (Fall 1998), 335-343.

Another thing that created pressure was that the CAW managed to organize Starbucks’ pastry bakery, which supplied all 50 stores with baked goods. Successfully signing this business up led to movement at the table according to Banong, but shortly after organizing the plant, Starbucks shut it down. Retaliatory closings are illegal under the Labor Relations Acts, but even unionized businesses retain the prerogative to run their operations as they see fit. The company successfully managed to make the case that they were planning on this closure so all along, thus avoiding an Unfair Labour Practice charge.  

The union suffered another legal setback at the Board. Shortly after the second round of bargaining in 2001, the Cambie Street location applied for a partial decertification. The union argued that management was behind the move, citing anonymous tips that Starbucks lawyers and management were coordinating the action.  The CAW put a lot of energy into defending legal recognition in the unit — there are at least four cases on file with the British Columbia Labour Relations Board. Cases like these are staff-intensive and almost always involve a lawyer for the union. All of these are costs a union will usually measure against the dues the shop brings in. Especially since, as in this case, the union’s argument was unsuccessful.

Even more effective than its legal response, Starbucks’ strategic response to the union was to generalize any changes made in bargaining: the wage increase that the union stores negotiated was immediately passed on to workers in non-union stores.  Bowman believes that as soon as that happened, the writing was on the wall for the campaign.  If the union could not deliver something more than what the non-union stores received, there was no meaningful incentive to sign with the CAW – after all, a big part of the CAW’s organizing approach had been to argue they could get a higher wages in a contract. 

Peak and decline

At its height around 2001, the CAW drive at Starbucks had 12 stores out of about 50 in the Lower Mainland of BC.  The contract covered around 150 workers.

By the end, the campaign had none of the original people, and the CAW had failed to sign enough new members over the years to keep a presence on the job. Being involved in the union and even being a member of the union are not the same as being in the bargaining unit. Unions can become the certified representatives in the shop and receive dues for all workers, but formally inviting workers to take out a union card and actively participate in union membership, or take on a position such as shop steward, is a separate matter. Also if a union can’t give a sense of progress to their activists then eventually morale drops off.

As we saw Banong describe, one of the challenges in the shops was getting the leaders on the job to take ownership over bargaining. Member-driven bargaining can help but it’s also just a fact that conventional bargaining is a technical field where expertise matters and it’s very easy to drift into the workers being spectators to their own involvement. 

Starbucks also capitalized on the CAW’s own success by handing any wins off to non-union shops, extending the free rider problem (where non-dues-payers enjoy the products of union members’ struggles) to all of the shops in the lower mainland. The incentive to join the campaign was removed by handing those victories off to others. 

At the end, while the CAW campaign at Starbucks had legal recognition and a contract — all the formal stuff was in place — they didn’t have a union on the job. During hearings about the Cambie Street location, the union conceded on the stand that the store in question, while permitted a shop steward as per the collective agreement, did not have one and may never have had one.  Bowman says there were very few grievances filed according to his recollection. Only one stood out and it was on scheduling and eventually settled before going to arbitration. According to him, “they were generally not the kind of workforce that filed a lot of grievances.”  In 2007, the union withdrew from Starbucks and allowed their representation rights to lapse though this was largely a formality as the presence on the job had been gone for quite some time. 

Overall the successes of the campaign were not negligible. They won a fifty cent increase right out the gate in response to the drive and another twenty five cents at the table (unions often fail to claim pay raises that happen as an incentive not to organize). For workers making $7.50 per hour that represents a ten percent pay hike. They also did win better language around scheduling and the standard discipline and grievance language that comes with a union contract. The Wagner Act model however presents real challenges to a union looking to organize companies like Starbucks. The company managed to marshal tremendous resources in order to stop the campaign from spreading and to limit the concessions they had to give out. When the union certifications began to fall apart what was exposed was little shop floor presence and insufficient union activists to keep the formal legal machine running. 

In the next installment we will look at how the IWW tried to move past these limits. 

Nick Driedger is the Director of Labour Relations and Organizing for the Alberta Union of Provincial Employees and a frequent contributor to Organizing Work.