Canadian Journal of Communication Vol 40 (2015) 203–221
©2015 Canadian Journal of Communication Corporation

The Capacity for Mobilizatioin in Project-Based Cultural Work: A Case of the Video Game Industry

Marie-Josée Legault
Teluq-Université du Québec

Johanna Weststar
Western University

Marie-Josée Legault is Professor, École des sciences de l’administration, Teluq-Université du Québec. Email: . Johanna Weststar is Assistant Professor, DAN Management & Organizational Studies Department, Western University. Email: .

ABSTRACT  Though dissatisfied with some management practices and working conditions, like most high-tech knowledge workers, videogame developers remain reluctant towards unionization. This article examines the factors of collective action among developers as an example, using data gathered from an international survey and interviews. We conclude that developers meet some conditions conducive to collective action but face many obstacles as well, both to collective action and to unionization proper. This does not lead us to share the belief of a decline in collective action, but rather raises the issue of conflating union action and collective action. Our study reveals how unsuited the general North American trade union system is to their situation, as it is to project-based environments and knowledge workers in general.

KEYWORDS  Film/video policy; Videogame developers; Working conditions; Project-based organizations; Unionization

RÉSUMÉ  Bien que pratiques de gestion et conditions de travail provoquent de l’insatisfaction chez les concepteurs de jeux vidéo, ils demeurent réticents devant la syndicalisation, comme les travailleurs du savoir en général. Cet article étudie les facteurs de l’action collective chez les concepteurs, en utilisant les données d’un sondage international et des entrevues. Nous concluons que les concepteurs satisfont certaines conditions menant à l’action collective mais rencontrent aussi plusieurs obstacles qui s’opposent parfois à l’action collective mais plus encore à la syndicalisation à proprement parler. Cela ne permet pas de conclure au déclin de l’action collective, mais plutôt d’interroger le bien-fondé d’assimiler syndicalisation et action collective. L’étude révèle plutôt un régime syndical Nord-américain mal adapté à leur situation autant qu’à l’organisation par projets et à l’économie du savoir en général.

MOTS CLÉS  Politique du cinéma et de la vidéo; Concepteurs de jeux vidéo; Conditions de travail; Organisation par projets; Syndicalisation


The late 20th century was marked by a decline in national unionization rates (especially in the private sector) and very low rates of unionization in emerging sectors of the economy, such as high-tech industries (Milton, 2003). Some have interpreted this phenomenon as evidence that trade unions are less relevant to highly skilled professionals, who are individualistic, mobile and career-focused (Bassett & Cave, 1993). Others call for unions to adapt and replace confrontation with greater cooperation with businesses’ economic success (Kochan & Osterman, 1994). Some have gone further and argued that new media professionals are really entrepreneurs and should not be subject to the government rules that apply to hourly employees, including unionization (Cohn, 2001). In fact, the labour laws of many U.S. states and some Canadian provinces either exclude “high technology professionals” from standards regulating payment for overtime or make particular adjustments to those standards in the name of flexibility.

Videogame developers (VGDs) are one such occupational group. Like many high-tech and cultural knowledge workers, they work in project-based environments, are highly skilled (98% have completed some form of post-secondary education), well paid (they earn 150-200% more than average workers of comparable education in Canada) and highly mobile. The videogame industry is often criticized for the violent and sexist content of mainstream games, but it has also received public attention for labour abuses. Most common are issues of work-life balance and the long and formally uncompensated hours of overtime or “crunch” (Deuze, Bowen Chase & Allen, 2007; de Peuter & Dyer-Witheford, 2005; Dyer-Witheford & de Peuter, 2006; Legault & Ouellet, 2012; Legault & Weststar, 2012, 2013; Weststar & Legault, 2012). Challenges around intellectual property and crediting standards, non-compete and non-disclosure agreements, health and safety (i.e., stress, burnout, intoxication, and musculoskeletal disorders) are other common concerns (Legault & Ouellet, 2012).

We have documented some dissatisfaction with these practices; yet, like most high-tech knowledge workers, VGDs remain reluctant towards unionization (Milton, 2003; Haiven, 2006), and this raises a number of questions. How do individuals acquire a sense of collective—as opposed to individual—grievance? How, and under what conditions, do individuals organize collectively to pursue their grievances, or interests? How, and under what conditions, will such individuals take collective action—that is, cooperative action taken by a number of individuals acting in concert and with common goals? These questions are raised by Kelly (1998) as the central problems of the field of labour relations. In this article, we apply Kelly’s mobilization theory to examine the likelihood of mobilization or collective action among VGDs. First, using data gathered from an international survey, we show that VGDs are divided on the idea of unionizing. To interpret this raw data and test Kelly’s determinants for mobilization we analyze additional survey data and interviews conducted with VGDs in Montréal.

We conclude that VGDs meet some conditions conducive to collective action. First, they have identified common problems in industry working conditions. Second, they have developed a professional community with which they identify. Yet, an examination of other conditions of Kelly’s model reveals a number of obstacles, both to collective action and to unionization proper. This does not lead us to share the belief of a decline in collective action; it rather raises the issue of the univocal nature of Kelly’s mobilization theory, which conflates union action and collective action. Our study of VGDs reveals how unsuited the general North American trade union system is to their situation and to project-based environments and knowledge workers in general.

Context and theoretical framework

In 2009 the International Game Developers Association (IGDA) asked the following question on their second survey measuring quality of life in the industry: “Some developers feel the only way to improve the quality of life in this industry is to join a union. If a vote were taken today, how would you vote?” The 2506 responses among the international sample of developers were divided in three thirds; 35 percent would vote for the union, 31 percent would vote against, and 34 percent avoided this controversial topic by choosing “no opinion or prefer not to say.” Given declining unionization rates, the degree of union support was surprising. The emerging risk society (Beck, 1992) is often seen as an economy in which individuals assume greater responsibility for protecting themselves, counting less on state support. It is assumed there are fewer manifestations of collective material interests, simply because less collective consciousness exists (Bassett & Cave, 1993; Brown, 1990).

Kelly (1998) notes that a collective interest can however exist in the absence of any such manifestation because the mobilization of that collective interest requires a specific supporting framework. Kelly’s mobilisation theory remains a prominent meta-model that engages with and builds on previous models of smaller scope (i.e., the models of Gamson, MacAdam, Olson, & Tilly) in an attempt to reconcile and aggregate their most common and relevant features to account for the macro-social reality of mobilization and build predictive power under a new framework (Kelly, 1998).

According to his observations, fluctuations in worker mobilization mirror the economic rhythms of capitalism, which periodically cause economic situations that provoke collective action. Throughout this article we will test the experiences of VGDs against this mobilization theory to better understand their propensity and opportunity to engage in collective action.

Mobilization theory presents four determinants that must all be satisfied to result in collective action (see Figure 1). The first determinant is derived from social movement theory. Here, the interests of individual actors must come to be: a) framed against those of a ruling group, and b) framed as collective. This is attained when people no longer believe in the legitimacy of the status quo and the attempts of ruling groups to legitimate their actions no longer succeed. More than dissatisfaction, a sense of injustice is needed to trigger collective action (i.e., the violation of established rules or a breach of equilibrium in the wage-effort exchange). As a consequence of defining the situation as illegitimate, workers in subordinate positions feel entitled to their demands and align no longer with the ruling group, but with fellow subordinates.

Figure 1: Graphical representation of Kelly’s mobilization model

Three processes are important in reaching the above determinant of collective interest: attribution, social identification and leadership (Kelly, 1998). Through attribution, the injustice is blamed on an “other” as something under his/her/its control. Through social identification, individuals aggregate as an “in-group” (us) positioned in opposition to an “out-group” (them). In most cases, leaders initiate and facilitate the social construction of attribution and social identification.

The second determinant is the organizational structure of the group vis-à-vis its capacity for collective action (Kelly, 1998). Kelly uses examples from unionized environments but is broadly referring to the connectivity and communication capacity of the group. Successful mobilization can hinge on the quality of communication channels, the degree and nature of interaction among members, and the density and strength of social networks.

In Kelly’s cumulative model, the third determinant becomes primed once the previous two are met: actors have defined their interests as a collective and reached a sufficient degree of organization. The third step is the actual mobilization. It is the process by which a group acquires collective control over the resources needed for action or the ways in which individuals are transformed into a collective actor. However mobilization itself requires additional enablers: a favourable cost-benefit assessment, leadership, and social interaction.

The fourth and last determinant is the opportunity to engage in collective action. Opportunity is based on the balance of power between the parties, the costs of repression by the ruling group (i.e., the employer), and the avenues and procedures that are available for subordinate groups to pursue their claims (i.e., alternative actions, supportive labour laws, or societal norms).

Under this model, collective action as an end result can take different forms according to the balance between perception of interests, organization of the group, mobilization, and opportunity for action. It is important to note that Kelly’s model takes union action as the height of collective action in the workplace. The actions given as exemplars are those used by trade unions: strikes, overtime bans, go-slows, working to rule, petitions, lobbying, and collective appeals (Kelly, 1998). However, as our analysis will show, a focus on traditional forms of unionization as the endpoint of successful collective mobilization may be a limiting feature in Kelly’s model.


Two sets of data inform our discussion of the propensity for video game developers to mobilize under Kelly’s model. The first is the aforementioned data from the 2009 Quality of Life (QoL) survey that was administered by the IGDA (hereafter referred to as the 2009 QoL survey). The total international sample size is 3362 and includes game developers in a variety of employment relationships; however, a number of the questions used in this article were only asked to the 2153 VGDs employed full- or part-time (Legault & Weststar, 2012; 2013). The second is a set of 53 interviews of salaried VGDs working in various studios in Montréal, Québec, conducted in the summer of 2008. The sample contains roughly equal numbers of men and women and is otherwise generally representative of the demographics of VGDs. Montréal has 80 percent of the videogame employment in Québec (Corbeil, 2012) and Québec has half of the videogame employment in Canada (ESAC, 2011).


Interests: Collective injustice and attributions of blame

Unlike traditional organizations where employees work in the same geographical location and unions are certified on an enterprise basis, VGDs show evidence of an occupationally-based collective identity. Due to the project-based nature of the industry, VGDs often have portfolio careers with high mobility (Weststar, 2013). As a result, VGDs often have weak ties to any particular studio and strong ties to the specific games they have made and the developers with whom they have worked. The IGDA is the professional association for the trade and facilitates the development of communities of practice among VGDs. In these ways, VGDs are typical of project-based knowledge workers, creative workers, and emerging technical professionals where the occupation is the nexus for collectivity (Barley & Kunda, 2006). Workers across geographical and organizational boundaries are united through the shared language and norms of their craft. This occupational community of VGDs is further reinforced through a shared culture of games and gaming. Therefore, VGDs do form an “in-group” that is defined occupationally on an international basis and is positioned against “out-groups” such as the work in other entertainment mediums or other jobs that programmers or artists might do.

We have formerly published accounts of the challenging working conditions and labour process of game development and related fields, and it is out of scope to reproduce those here (see also Chandler, 2009; Deuze, et al., 2007; IGDA, 2004; Kerr, 2011; Kline, Dyer-Witheford & de Peuter, 2003; Legault, 2013; Legault & Ouellet, 2012; McGuire & Chadwicke Jenkins, 2009). Rather, we will start from the premise that issues exist and could be perceived as injustices. The questions for mobilization theory are whether issues such as unlimited unpaid overtime (UUO) are indeed seen as an injustice against the collective and whether the source of that injustice can be attributed to another group (i.e., the employer). Regarding the first, UUO is certainly an acknowledged source of dissatisfaction. The 2004 IGDA QoL survey found that 86.2 percent of respondents could not see themselves keeping up the same pace of work due to repercussions among friends and family and a high rate of burnout (IGDA, 2004). Illegitimacy also arises because refusing overtime has consequences in the form of negative performance appraisals and exclusion from the peer network. This can damage professional reputations in an industry characterized by mobility and frequent replacement:

I don’t really know people who won’t work the overtime. Because if you’re on a team, let’s say if the programmer refuses to work overtime, the game doesn’t get finished for that day and doesn’t get sent to the people at headquarters who have to review it every couple of days and he gets blamed. No I don’t think you really can. You can but you’d probably be fired quickly. … I know I get evaluated every six months and I know it will affect my evaluation if people perceived me as being the girl that doesn’t go the extra mile. (F-10-16-G-26-06-08-01-07)

Some interpret sufficient injustice so as to begin to speak about change through collective action:

I wanted the overtime to be justified. I wanted to be paid, and of course, you can ask the employer, but obviously he’ll say he’s entitled. And then when you call later for help, there is none, and then you don’t want to battle a giant like those huge companies on your own. Obviously a class action is needed. It takes a torchbearer. No employee will do it. That’s usually the union’s role. But we don’t have one here. (H-13-08-U-03-06-08-01-07)

It is not, however, universally considered an injustice for a number of reasons. For one, according to the 2009 QoL survey a majority of developers (64%) are poorly or not informed about labour laws in their country or region. Thus, they are far from knowing whether their situation is legitimate or not. Further, 40 percent do not know if the labour laws where they live offer sufficient protection should a grievance arise between an employer and employee. To take the example of the legislative framework in Québec, the Act Respecting Labour Standards (RSQ, c. N-1.1, ss. 52–55) states that employees may be required to work overtime, in exchange for a premium of at least 50 percent of the prevailing hourly wage, if the employer asks them to. Conversely, an employer that does not wish to pay for overtime cannot require it. Therefore, game studio practices are legally ambiguous, because supervisors do not actually ask VGDs to work overtime and maintain that it is never required, but that VGDs do it of their own volition. To avoid controversy over compensation, extra hours are called crunch time rather than overtime, thus presenting it as a project management constraint, rather than a management request. Worse, some VGDs are asked to sign timesheets showing 40 hours, no matter how many they have actually worked or no logs are kept at all.

Neither purely voluntary and willingly agreed, nor required and forced, overtime of this kind falls into the biggest category of “willingly agreed, but strongly expected” (Campbell, 2002, p. 141). In this manner the unstated expectation of management becomes rooted in the organizational and industry cultures, and indeed the occupational ethos, of making games. In the 2009 QoL survey, one-third of respondents felt that crunch was a necessary part of game development. That said, 79 percent of those who crunched often as part of regular studio practice felt that it was illegitimate (i.e., they viewed it as a failure in scheduling or flatly disagreed with the practice). The result is a conflicted response on behalf of VGDs:

You know, especially at the end of a project, they try to get people to put in just a bit more extra effort. … The company doesn’t make me do the hours. I do it because I want to. But at the same time, the constraints of working in videogames mean that it’s hard to get ahead without doing it [overtime]. (H-01-16-U-29-05-08-01-07)

As well, workers may perceive these environments as motivating and satisfying because challenging assignments are often occasions of learning and opportunities to enrich one’s portfolio (Dessler, 1999). In this way, many VGDs seem like willing conspirators in their own exploitation (McRobbie, 2002) as long as they are creatively respected:

That’s pretty [much] what seals the deal, if a project is interesting enough, people would put up with anything, they will work crazy hours if they love the project … so people will go “Oh yeah, it’s going to be a great game.” So they use that, a company uses that to make people do more work than they should do … . Sometimes they use that to exploit you so they don’t pay you as much … they know you like it, so they don’t have to pay you because they know you’ll do it anyway, they know you’ll accept it. … If I’m working for a project that I put my own personal stamp on, that I invested in, [overtime’s] sort of my choice. (H-13-11-A-17-06-08-01-07)

In a star system where reputation is the key to mobility, VGDs are all the more willing to accept poor conditions on a project if it enables them to acquire skills and eventually be associated with a hit. Thus VGDs are driven by an informal system of rewards and punishments in the form of boosts or impediments to career development, especially in studios aiming for AAA game hits. To add to this uneven interpretation of illegitimate working conditions, big stars are able to individually impose conditions and therefore often arrive at quite satisfactory, yet exclusive, arrangements (Legault & Ouellet, 2012). As well, some studios work very hard to avoid crunch or rule it with transparent policies.

Once a collective injustice is perceived, it is necessary to attribute that illegitimacy—to find someone to blame. This is again an uneven application among VGDs. Team leads, senior managers and often owners are included in the “us” of the occupational community and studio hierarchies are often quite flat. It is therefore more difficult to parse out the “them” to blame. As well, some VGDs do not blame the employer at all. For those who do, the criticisms are associated with a lack of voice in setting the schedule, lack of control over the scope of the project, and reduced budgets or staff in the face of escalating expectations:

So they have trouble coming to see us and saying: “Right, I’ve got a job to be done. How long will it take you?” When they’re planning, they put down that it’s going to take a day, when we know damn well it’ll take two or three days. So we wind up with plans that are absolutely never followed. And towards the end of production, when the deadline looms, you can’t put it off: it’s the customer’s deadline. (F-13-19-A-23-07-08-01-07)

The feeling of illegitimacy is increased in the face of successful games with large profits or when the game developers feel that they are under-resourced or are being knowingly exploited by their managers:

Team budgets are getting smaller and smaller and producers take it for granted that people will do overtime. They shorten the timeline, they do it on purpose to fit the most possible into a shorter time. (F-10-12-U-12-05-08-01-07)

Risk management is a large component of project management and involves planning to account for and mitigate threats to the project’s immutable schedule (as set by the publisher and/or senior management). Project managers have to estimate the time needed to meet the deadlines with uneven resources and avoid project failure. It is common for management to incorrectly anticipate risks, and therefore rely on crunch to save the project. For example, one programmer (F-13-19-A-23-07-08-01-07) complained that 80 percent of her time was spent on the upkeep of the computer (i.e., downed servers, slow networks, broken parts) as opposed to new work, yet the time needed for these regular events was not accounted for in the schedule. As another developer said, “there are projects that go wrong because people underestimated the difficulty or planned poorly” and he further suggested that the project-based environment is not sympathetic to such errors, “[d]eadlines don’t get pushed back because of a mistake like that” (H-06-05-U-05-06-08-01-07).

The more experience employees have, the more they tend to blame overtime on poor project management as opposed to “the way it is in games.” Many commented on the inability to refuse customer requests after the contract is signed. Resources are assigned to the project according to the parameters of the contract terms, which are grounded in the terms of the agreement. Customer change requests should theoretically have an impact on contract terms (i.e., extended budget or time), but in practice that is uncommon:

In the other cases of overtime, when the publisher says: “Oh, can we have this?”—“Can we have that?”—“We don’t like that.”—“This doesn’t work anymore.”—“We’re gonna change this” So that has a huge impact on the production because it’s not something that’s planned and it’s usually something that comes very late and the reason why it happens is usually that the … licence holders or any sort of third party owner of this intellectual property might only get involved towards the very end of the project, so then that’s when things start getting really messy. (F-12-16-A-16-06-08-01-07)

These quotations demonstrate a challenge with attributing blame in a project-based environment. Local management of the project is the responsibility of the producer or project manager; however, they are subject to the decisions of senior management within the studio and also parties external to the studio—the most important being the client or editor when the studio is a second or third party. As the point of blame becomes more removed from the developer, it becomes easy to see the problems as too big, systemic, unchallengeable, or “just the way it is”:

[When you consider the question of hours …] It’s not just the company, it’s the whole industry. The industry is aggressive, highly competitive. You always have to try and stand out. Of course, the company I’m with is one of the top five in the world. Just to stay in the top five, you have to be demanding, have a great catalogue that will attract players, that will sell, that will be fun, so there’s a lot. (F-01-01-U-31-07-08-01-07)

Unlike mass production, the full details of the process and, to some extent, the exact outcome that can be achieved, are unknown. Every game created must be different from those preceding it and make full use of available technological possibilities. The uncertainty inherent in estimating the time needed to achieve a creative result makes many salaried VGDs sound like entrepreneurs deciding how many hours to work based on the importance of product quality:

So if I didn’t do it [overtime] and no one else did it, it wouldn’t show in the final product and we have a certain amount of pride and a certain attachment to the final product, the common goal. So it’s not just repetitive work delivering a certain number of products, it’s the quality of the final product. (F-18-02-U-22-07-08-01-07)

In this same entrepreneurial mindset, some VGDs include themselves or their team members in the blame for failed projects or long hours:

It’s a young industry, so we still don’t think about how to properly plan a game yet. We wouldn’t need to do as much overtime if we’d plan things better. Generally we don’t really know what we are doing a lot … Like we know how to make the game, but … things change all the time and right at the very last minute and I think it’s maybe lack of experience, we still sort of rush in, rush in, right to the last, last bit. So that causes these extra crunch times. (F-05-20-U-25-06-08-01-07)


On one hand, VGDs are well organized under Kelly’s definition in that they have established structures and the capacity to communicate quickly and broadly. The IGDA is an international professional association that commits to “advocate on behalf of our membership to ensure quality of life, perpetuation of our craft and preparing the next generation of developers” (n.p.). Under the IGDA banner are about 90 local chapters, which exist in most cities with game development clusters, and special interest groups (SIGs) on key topics. Face-to-face meetings and Facebook discussions are facilitated through the semi-autonomous local chapters. SIGs tend to operate through email distribution lists and forums, although they will also host panels and meetings at game conferences. The industry also has a number of trade associations such as the Electronic Software Association (US & Canada), The Independent Games Development Association (UK) and the European Games Developer Federation. However, these groups tend to focus less on workers’ issues than on building a competitive global industry. The industry has also spawned a plethora of online electronic magazines and blogs that report on all aspects of game development and game play (i.e., Gamasutra). These sites publish articles written by game developers and act as open forums for discussions and opinions. Through these channels and social media, word travels fast (Shirky, 2008).

On the other hand, VGDs face an organizational challenge to mobilizing because existing groups and associations rely heavily on volunteers to conceive and execute activities. This severely limits organizational capacity. Like their medieval forebears, modern guilds focus on sharing knowledge—networking, providing services, and helping their membership anticipate and capitalize on changing industry trends. This requires building close ties with employers and does not facilitate the “us-them” dichotomies required by Kelly’s mobilization theory and found in “traditional” labour relations (Benner, 2003). For instance, IGDA membership fees are often paid by studios. Many VGDs do not see the IGDA as capable of making changes to the working conditions in the industry, nor do they seem to demand that intervention:

I see professional associations as more for providing tools, training, advice, things like that. I see them more as a community of people working in the same occupation who can talk and discuss the subject. I don’t really see them as backing me in case of problems. A professional association isn’t like a trade union, either. It’s really a group of people who do the same job, who may be able to give me cues here and there for getting ahead, tools to do the job better. (F-01-20-U-06-06-08-01-07)

Without the steward systems typical in trade unions, it is difficult to accurately monitor on-the-ground issues at individual studios. Professional associations also do not have the legal backing to engage in more than public peer pressure. That said, depending on the personalities and inclinations of its executive director and volunteers at any given time, the IGDA has shown leadership on important debates such as working conditions. They have issued public statements to rogue studios, collected and published data on the conditions of the industry (such as the QoL surveys).


Much of the lack of mobilization seen among VGDs can be attributed to the cost-benefit analysis of Kelly’s model. Many game developers see few benefits and perceive many costs to becoming unionized. One strong hindrance is the high individual bargaining power of VGDs that is rooted in a favourable job market. For the time being at least, many VGDs, especially those with highly demanded skills and reputations, don’t see any added value to a union:

I think that right now, people don’t feel they need a union. Why? Because there’s a lot of work. You don’t need to defend yourself. Even though there are disparities between some … people who do the same job, there’s still great satisfaction with pay, because it’s driven by market pressures. We’ve got the long end of the stick. (H-12-16-16-A-04-06-08-13-19)

Like other new media professionals, VGDs also struggle to see benefit in a union because they have a weak commitment to any particular employer or employment arrangement (Batt, Christopherson, Rightor & van Jaarsveld, 2001). This is a manifestation of the project-based industry structure. VGDs move frequently from project to project, team to team, and studio to studio. As such, many do not perceive themselves as having labour issues that warrant attention because they will not be in that environment long enough for it to matter. It is a classic case of Hirschman’s (1970) “exit” over “voice” response to unfavourable conditions. High mobility across employers also does not fit with the traditional North American model of enterprise unionism:

… we do change companies a lot, so if you work hard and try to get one company to implement something and then you just move to the next one, then you have to work hard to get it done again, so I think that probably in the long run, it’s better through politics and setting standards on having something that is more universal. (F-05-20-U-25-06-08-01-07)

Besides leaving for another studio, another kind of “exit” response may be found in the common yearning for creating one’s own independent small-scaled studio. This trend seems to be building in the industry, driven in part by new technologies to allow for digital distribution and easier access to market for small developers. However, it materialized as a growing trend after the 2009 IGDA survey.1 Whether, how and to what degree this trend is rooted in major structural changes in the industry or in developers’ longing for autonomy and/or dissatisfaction towards studios that hire them is a very interesting issue, though far beyond the scope of this article.

Returning to the quote above, to address the need for “something that is more universal,” for some time, unions in the film and television industry have promoted an alternative organizing model in the form of industry- or occupation-wide certification. This model fits the ideological frame and working realities of project-based occupational communities (i.e., the actor’s union, the screenwriter’s guild). It allows for portable rights and benefits and is able to account for freelancer professionals (Batt et al., 2001; Amman, 2002; Legault & D’Amours, 2011). Such options for mobilizing or unionizing may not be well known or well understood among VGDs who seem to rely on a generic notion of industrial unions that legitimately does not fit their realities. That said, there remain real challenges to work citizenship and representation in the context of national and international job market mobility (Legault & D’Amours, 2011), not the least of which is the issue of portable rights when workers are increasingly being asked to move (Carré, 2010). Internationally mobile VGDs are well aware of these challenges and the barriers they pose for local or national systems of collective representation or rights:

Well, the thing I want to emphasize was that the industry is very international and it’s a little bit tricky to look at it only in national level … like people that work for [studio] and then go to [Asia] lose their civil rights, or people that are from Sweden and move … don’t have the same child care …, but they still have the same family. You know, there’s just so many things related to people crossing borders constantly. … For me, investing so much in retirement that I’ll never be able to collect on because it’s part of a national system … It doesn’t belong to me; really, it’s paying into a system that will pay back out to me. (F-08-11-I-01-08-08-01-07)

An equally large barrier to unionization among VGDs is the perceived costs of unions. Like computer programmers (Milton, 2003), many VGDs harbour misgivings about unions, especially the fear of compromising creativity and innovation:

People talk about VGD unions, but it’s a pipe dream. … The union is kind of anti-passion … It brings everyone down to the same level, gives everyone the same working conditions. And in terms of innovation, it would be even worse than today, I think. It could really put the brakes on ideas and people’s commitment … It’s employee commitment that gets a game out. So if your employees only work from eight to five, nothing’s going to get done. Montréal’s reputation is going to suffer. (H-06-16-G-23-07-08-01-07)

Interviewees voiced a plethora of negative anecdotal and second-hand experiences that signify a deep-rooted disinclination toward unions and powerfully reinforce negative perceptions. Traditional or typical union models that protect seniority, stability and equality, establish job descriptions, and set up pay raises unconnected to individual merit are perceived as antithetical to learning-oriented meritocracies. Many contended that it benefits the least ambitious and stifles creativity. These perceptions are similar to those of computer programmers who feel that union members have nothing in common with high-tech workers (Milton, 2003).

Though this stereotyping can be lamented, its pervasiveness cannot be denied. In this way the dominant image of industrial unions can blind workers to more accurate and promising comparisons. As with industry-wide certification systems, film and television unions are again a more useful comparison for new media professionals. The performing arts provide numerous examples of compensation systems within unionized environments that account for merit through mixed allocations of fixed and variable pay and allow for “above-scale deals” (Amman, 2002, p. 126-127; Legault & D’Amours, 2011).

Social interaction

Where Kelly sees social interaction as necessary to build the message and momentum of a mobilizing drive, as discussed above, social interaction can also reinforce the costs or negative tendencies toward collective action and unions in particular. Returning to the 2009 QoL survey, employed respondents were asked how they thought the people at their company would vote in a hypothetical union certification. Just over one-quarter felt that their coworkers would vote against a union, while about 20 percent felt their coworkers would be in favour of a union (16% said the vote would be split 50-50 for and against and 38% had no opinion or preferred not to say). When considered in contrast to the figures presented earlier where one-third said that they themselves would vote in favour, this data shows that workers in the videogame industry perceive more negativity toward unionization on behalf of their co-workers than actually exists.

Mobility also plays a role in that it can reduce the opportunities to develop the social fabric required for local collective action. Though VGDs are connected through an online community with occasional face-to-face events, constant turnover and team reorganization:

reduces the opportunity for repeated cycles of exchange, risk-taking, and achievement, experiences that would strengthen the willingness of trusting parties to rely on each other and expand resources brought to the exchange. (Milton, 2003, p. 39)

Without time to build shared experiences of continued unjust or illegitimate treatment and with the continued thought that somewhere else might be different or better, it becomes challenging to develop the needed ideology for an “us versus them” framing of struggle.

And to date, no leaders have emerged from the industry to sufficiently unite and mobilize the myriad of experienced illegitimacies under a common theme for change. Individual developers have written articles and provided commentary, groups have pursued class action lawsuits for specific violations, and roundtables have been held at conferences, but there has not been lasting leadership. The year 2004 marked a great deal of interest due to the conflagration of a highly popular blog decrying working conditions at Electronic Arts (discussed below), a number of class action lawsuits, and the launch of the Quality of Life committee within the IGDA; however, the vibrancy of that movement has ups and downs (Hyman, 2008).

Opportunity – Employer retaliation and alternative action

The power of the employer and fear of reprisal is critical in the decisions of all workers considering unionization (Godard, 2008). The 2009 QoL survey asked employed respondents about how they thought their studio’s management would react to a unionising initiative. A small proportion of the respondents said their managers would welcome the union (6%) or would not care (11%). One-third preferred not to voice their opinion on this question. The majority (52%) felt management would oppose the initiative and 15.5 percent of those felt that management’s opposition would be aggressive and take the form of threats and harassment. In a young population that is not well informed about their labour rights, this fear could be heightened. Many VGDs maintain anonymity in online posts that are critical of their employer or the industry in general. This perception is not unfounded:

It’s the problem of being seen as a [trouble-maker]: Don’t cause too many problems because … arbitrarily … fires people sometimes. It just seems that if it’s at the end of a project and if it’s gone really well, everyone’s safe, but if it hasn’t, they’ll fire the producer and the designer and someone else. [Without any explanations?] It happened before and they would just say “we didn’t work well together,” “work didn’t go fast enough,” “the project didn’t go that well, it’s your fault.” … Without a warning. (F-10-16-G-26-06-08-01-07)

This example concerns individual reprisals, but they are not the only ones to consider. Regardless of the advantages gained from collective actions, they may have a perverse effect. Following the class action wave in California, Electronic Arts transferred hundreds of developers to Florida and Canada, wishing to avoid its new liability to pay them overtime (Feldman & Thorsen, 2004). Such a retort can chill a movement and stall would-be union organizers in a context where the threat of outsourcing always lies in the background:

A lot of people say, “Oh, if the game industry is unionized, it will move to China, period, and that’s the end of that. They’ll pay people who live to work, rather than people who work to live, and … ” You know, they often tell us it’s impossible to unionize and employers would go elsewhere. Everyone would love to find a solution, but no one is very well informed. (F-03-18-U-13-06-08-01-07)

The discussion so far paints a rather dismal picture of the capacity for VGDs to mobilize under Kelly’s model insofar as mobilizing is conceived as forming a union. However, there is evidence of discontent and the desire for action, as well as explicit evidence of VGDs coming together in new forms of collective action. Over the past 10 years, there have been a number of online campaigns against abusive employers. These have the advantage of targeting a large audience and protecting the anonymity of activists through the use of avatars or taglines. As Shirky (2008) argued, social networks remove two important obstacles to collective action: the limits to circulation of information and the constraint of physical gathering to deploy collective expression. The capacity to instantly and internationally share strategic information and to coordinate collective action allows for quickly constituting a redoubtable stock of evidence in cases of media or legal action. The most famous case is the protest against unlimited unpaid overtime written by a woman “widowed” by her then-fiancé’s long hours at an Electronic Arts studio. Her blog as EA Spouse went viral and resulted in thousands of online comments about VGD working conditions. This built momentum for the filing of three successful class action lawsuits for unpaid overtime and prompted EA to change some of its internal practices (Legault & Weststar, 2013; Peticca-Harris, Weststar & McKenna, in press).

Closer to a democracy of the multitude model (Hardt & Negri, 2004) and emblematic of the alter-globalization movement, many VGDs reject any transcendental hierarchy of command in collective action, which collides with well-established union approaches. They prefer to collectively produce social organization in temporary coalitions where the various social actors collaborate instead of being imposed an order by any external authority. This ethos is embedded in the prior socialisation of a majority of VGDs in gamer communities where players collaborate in massively multi-player online games and “mod” the source code of games to create new variations of gameplay that are shared (as derived from the collaborative open source movement). To join issue-based coalitions that disband when no longer needed is a type of job action that is more consistent with their beliefs; moreover, their skills, resources, and communication channels enable them to form effective issue-based networks (Milton, 2003). The EA Spouse mobilization provided VGDs with the feeling that “another kind of job action is possible”—one that is emerging, spontaneous, non-permanent, non-hierarchical, and controlled by actors themselves.

Part of this belief is poorly based on a context that may change and as such, is misleading. Similar to the threat of exit or more traditional employee voice mechanisms, the power of social media strategies first relies on the existence of a so-called “supplier market.” Employers are likely to respond to the publicized concerns to preserve their recruitment and retention in a tight labour market. It also relies on the fact that many game studios are heavily state-funded and can do without bad publicity. Should this context change, such a strategy would cruelly reveal its weakness.

As noted above, the IGDA has also relied on peer pressure to promote good behaviour. They will release public statements about the importance of good working conditions when problems are brought to their attention. They also work collaboratively with employers to create and enforce industry standards. This approach seems to align more closely to the identity of many VGDs:

There are initiatives like the IGDA that attempt to formalize things like getting your name in the credits. For example, there are people who’ve done fifty percent of a game and they don’t get credited … When you apply for a job, the idea is really there … People will say, “How many titles have you delivered, how many projects have you worked on?” It’s good to have your name [in the credits], because your reputation is based on credits. Those standards are developed by the IGDA, for example. It’s a kind of association, but it’s not a union. (H-12-16-16-A-04-06-08-13-19)

According to our respondents, big studios also have open-door policies, both as union-prevention strategies and under the influence of the high-performance work systems (HPWS) managerial approach (Butler, 2009). Within a HPWS, management actively seeks employee collaboration and commitment by encouraging them to express their ideas and reducing conflict. Many managers allocate about 30 percent of their time to answering employee questions and solving problems with working conditions. A number of women we interviewed mentioned this was how they obtained practical arrangements to help balance their personal life and work life.

Under Kelly’s mobilization theory, actions such as collective lawsuits, negotiation with managers, social media, and the IGDA all represent alternative mechanisms through which VGDs can pursue their claims. In this way, they can act as a hindrance to more full-fledged mobilization, such as unionization, because they act to diffuse issues over the short-term and reinforce individualized solutions. Though some positive change has come from the above-mentioned initiatives, and the consciousness about poor working conditions has been raised in the industry, real change has been slow. The Ubifree movement—where VGDs in France attempted to form a virtual union—was quickly silenced with only cursory appeasements from management, and there continue to be EA Spouse-like outcries online about abusive working conditions at various studios (Legault & Weststar, 2013). Developers themselves comment on the fleeting nature of these web-based movements and there is growing popular critique of the ability of social media to promote real engagement and lasting change. Developers easily post a supportive comment on a blog but seem reluctant to engage more fully to push for real changes.

Discussion and conclusion

This article sought to examine the propensity for videogame developers (VGDs) to engage in collective action under the determinants of Kelly’s (1998) mobilization theory. Survey and interview data indicate that VGDs have identified common problems in their working conditions and some define these problems as illegitimate. VGDs do identify as an “in-group” through the shared norms, experiences, and values of their occupational community. However, the model reveals considerable obstacles to collective action: the problem is not unanimously blamed on employers; VGDs are somewhat indifferent to legislative protection; alternative explanations of crunch time are accepted by many; labour shortages reinforce strong individual employee bargaining power; and managerial open-door strategies enable them to solve some problems to their satisfaction. When individual or collective action are equally likely to resolve issues, the former may be the most efficient because it does not have coordination costs. VGDs already work long hours and juggling priorities are only likely to coalesce when they believe that collective efforts will be successful and they cannot achieve the same results by acting alone. After all, many feel that they already have voice within their organizations and that their employers will be responsive to their individual requests.

There are also organizational obstacles to traditional union action. The group considered in Kelly’s model is stable and locally defined. In the case of VGDs, the in-group exists, but it is not employer-based; it is mobile and defined internationally by membership in the industry and the occupation. What is worse, the mobility of VGDs reduces the benefits of strictly local action, for which they pay the price without enjoying the results. Therefore, unionization through the usual enterprise-based certification system has very few advantages. As well, the hiring and compensation system is based on reputation, with a clear emphasis on the recent portfolio. The inherent meritocracy among VGDs conflicts with an egalitarian union ideology. These last two obstacles limit the ability to mobilize workers by limiting the potential gains.

Industry-based certification is an alternative avenue in both cases: it enables workers to enjoy benefits throughout the industry anywhere in the country, and North American industry-based certification for artists demonstrates the possibility of systems that use a mix of fixed compensation and variable merit pay. Though these “old media” union models in their current form are quite suitable for new media industries (Amman, 2002), it is important to note that these entertainment unions formed and gained their foothold in a different environmental context, one of few employers, high vertical integration of firms, extreme geographical clustering in Hollywood, and in a time of growing labour power (Gray & Seeber, 1996). These unions have maintained relevance because they have done much to adapt to meet the restructuring and technological changes of their industries, but they also owe some success to their historical presence. This legacy and favourable labour context does not exist for the videogame industry. VGDs are fearful of reprisals by local employers and significant stakeholders because venture capital has already demonstrated that it is extremely mobile. VGDs take the prospect of operations shifting to another country seriously and that undermines their belief in possible change through unionization. As well, even industry certification systems do not protect internationally mobile VGDs.

In short we see a group that holds a collective consciousness and is prepared to engage in forms of mobilization, but not in the univocal manner of traditional enterprise-based unionism. Therefore it is important to question Kelly’s conflation of collective action with union action, and of union action with Wagner-era industrial unions (Kelly, 1998). In the project-based organizations of the knowledge economy, of which videogame development is just one example, the conditions can be hospitable to collective action, but the usual enterprise-based union certification system is poorly suited to the structure of the industry and to workers’ most pressing problems. The primary effect of these structural changes is not to make collective action obsolete, but to make the traditional model of unionization less attractive.

VGDs perceive common interests among all workers in the industry, including self-employed workers and consultants, because their precarious status often draws them close to employees, especially with respect to intellectual property, and because they often alternate between contractor and employee status (Haiven, 2006). While unions may be reluctant to engage in individualized bargaining or similar services, and to lead organizing campaigns that are far more expensive, those that wish to organize and keep members will find they have to do it. In fact, a “new craft unionism” may be needed to meet the new exigencies of the employment market (Haiven, 2006, p. 111; see also Stone, 2004), taking the union outside the boundaries of traditional collective bargaining and the National Labor Relations Act template because whole areas of our economy will not play by those rules.

Unions in the movie industry have adopted trade-based rather than employer-based practices and such seems to be the need—if not the wish—in the new media arena (Amman, 2002), as these could better fit the videogame industry. But more deeply, opting for alternative modes of collective action may suggest that unions face a demand for a change in their purpose and use. In a new project-based context, there is space to reconsider some well-established norms embedded in the general unionization model including enterprise-based certification and the centralized decision-making processes that drive bargaining processes and job actions, among others. Both unions and mobilization theories that explain their outreach need to account for structural economic changes that do not make collective action obsolete, but rather call for a change.


The authors would like to acknowledge the support of the Social Sciences and Humanities Research Council of Canada (SSHRC) in the completion of this research.


  1. We analyse this more recent trend in the forthcoming outcomes of our research.


Amman, John. (2002). Union and the new economy: Motion picture and television unions offer a model for new media professionals. WorkingUSA, 6(2), 111–131.

Barley, Stephen, & Kunda, Gideon. (2006). Contracting: A new form of professional practice. Academy of Management Perspectives, 20(1), 45–66.

Bassett, Philip, & Cave, Alan. (1993). All for one: The future of the unions. Fabian Pamphlet, No 559. London, UK.

Batt, Rosemary, Christopherson, Susan, Rightor Ned, & van Jaarsveld, Danielle. (2001). Net working: Work patterns and workforce policies for the new media industry. Washington, DC: Economic Policy Institute.

Beck, Ulrick. (1992). Risk society: Towards a new modernity. Newbury Park, CA: Sage Publications. (Original work published in 1986)

Benner, Chris. (2003). “Computers in the wild”: Guilds and next-generation unionism in the information revolution. International Review of Social History, 48(S11), 181–204.

Brown, Phelps. (1990). The counter revolution of our time. Industrial Relations: A Journal of Economy & Society, 29(1), 1–14.

Butler, Peter. (2009). Non-union employee representation: Exploring the riddle of managerial strategy. Industrial Relations Journal, 40(3), 198–214.

Campbell, Iain. (2002). Snatching at the wind? Unpaid overtime and trade unions in Australia. International Journal of Employment Studies, 10(2), 109–156.

Carré, Françoise. (2010). Quelles politiques publiques pour la citoyenneté au travail? In M. Coutu & G. Murray (Eds.), Travail et citoyenneté. Quel avenir? (pp. 397–420). Québec City, QC: Presses de l’Université Laval.

Chandler, Heather. (2009). The Game Production Handbook (2nd Edition). Hingham: Massachussets, Infinity Science Press.

Cohn, Jonathan (2001, February 9). The Jungle. New Republic, 224(8), 23–27.

Corbeil, Vincent. (2012). L’emploi dans l’industrie du jeu électronique au Québec en 2011. Un portrait sommaire de la situation. Montréal, QC : Techno compétences - Comité sectoriel de main-d’œuvre en technologies de l’information et des communications

Dessler, Gary. (1999). How to earn your employees’ commitment. Academy of Management Executive, 13(2), 58–67.

Deuze, Mark, Bowen Martin, Chase, & Allen, Christian. (2007). The professional identity of gameworkers. Convergence: The International Journal of Research into New Media Technologies, 13(4), 335–353.

de Peuter, Greig, & Dyer-Witheford, Nick. (2005). A playful multitude? Mobilising and counter mobilising immaterial game labour. FibreCulture Journal. URL: [August 24, 2013].

Dyer-Witheford, Nick, & de Peuter, Greig. (2006). “EA Spouse” and the crisis of video game labour: Enjoyment, exclusion, exploitation, exodus. Canadian Journal of communication, 31(3), 599–617.

Entertainment Software Association of Canada (ESAC). (2011). Canada’s entertainment software industry in 2011. Report. Secor Consulting, Montreal, QC.

Feldman, Curt, & Thorsen, Tor. (2004). Employees readying class-action lawsuit against EA. Gamespot News. URL: [August 24, 2013].

Gray, Lois, & Seeber, Ronald. (1996). Under the stars: Essays on labor relations in arts and entertainment. Ithaca, NY: Cornell University Press.

Godard, John. (2008). Union formation. In P. Blyton, E. Heery, N. Bacon & J. Fiorito (Eds.), SAGE Handbook of Industrial Relations (pp. 377–406). Thousand Oaks, CA: Sage.

Haiven, Larry. (2006). Expanding the union zone: Union renewal through alternative forms of worker organization. Labor Studies Journal, 31(3), 85–116.

Hardt, Michael, & Negri, Antonio. (2004). Multitude. War and democracy in the age of empire. New York, NY: Penguin Books.

Hirschman, Albert. (1970). Exit, voice and loyalty: Responses to decline in firms, organizations and states. Cambridge, MA: Harvard University Press.

Hyman, Jeff. (2008, May 13). Quality of life: Does anyone still give a damn? Gamasutra URL: [August 24, 2013].

International Game Developers Association (IGDA). (2004). Quality of life in the game industry. Challenges and best practices. URL: [August 24, 2013].

Kelly, John. (1998). Rethinking Industrial Relations. Mobilization, Collectivism and Long Waves. London, UK: Routledge.

Kerr, Aphra. (2011). The culture of gamework. In M. Deuze (Ed.), Managing media work (pp. 225–236). Thousand Oaks, CA: Sage.

Kline, Stephen, Dyer-Witheford, Nick, & de Peuter, Greig. (2003). Digital play. The interaction of technology, culture and marketing. Montréal, QC: McGill–Queen’s University Press.

Kochan, Thomas, & Osterman, Paul. (1994). The Mutual Gains Enterprise. Boston, MA: Harvard Business School Press.

Legault, Marie-Josée. (2013). IT firms’ working time (de)regulation model. Work, Organization, Labour and Globalisation, 7(1). URL: [August 24, 2013].

Legault, Marie-Josée, & D’Amours, Martine. (2011). Représentation collective et citoyenneté au travail en contexte de projet. Les cas des artistes interprètes et des concepteurs de jeux vidéo. Relations industrielles–Industrial Relations, 66(4), 655–677.

Legault, Marie-Josée, & Ouellet, Kathleen. (2012). So into it they forget what time it is? Video game designers and unpaid overtime. In D. Jemielniak & A. Marks (Eds.), Managing dynamic technology-oriented business: High-tech organizations and workplaces (pp. 82–102). Hershey, Pennsylvania: IGI Global, coll. Information Science Reference.

Legault, Marie-Josée, & Weststar, Johanna. (2012). More than the numbers: Independent analysis of the IGDA 2009 Quality of Life survey. Commissioned by the International Game Developers’ Association (IGDA). URL: [February 23, 2015].

Legault, Marie-Josée, & Weststar, Johanna. (2013, January 9). Are game developers standing up for their rights? Gamasutra. URL: [August 24,2013].

McGuire, Morgan, & Chadwicke Jenkins, Odest. (2009). Creating games: Mechanics, content, and technology. Boca Raton, FL: A K Peters–CBC Press.

McRobbie, Angela. (2002). Clubs to companies: Notes on the decline of political culture in speeded up creative worlds. Cultural Studies, 16(4), 516–531.

Milton, Laurie. (2003). An identity perspective on the propensity of high-tech talent to unionize. Journal of Labor Research, 24(1), 31–53.

Peticca-Harris, Amanda, Weststar, Johanna & McKenna, Steve (In Press). The perils of project-based work: Attempting resistance to extreme work conditions in video game development. Organization.

Shirky, Clay. (2008). Here comes everybody. The power of organizing without organizations. New York, NY: Penguin books.

Stone, Katherine. (2004). From widgets to digits: Employment regulation for the changing workplace. Cambridge, UK: Cambridge University Press.

Weststar, Johanna. (2013). Understanding videogame developers as an occupational community. Unpublished manuscript. URL: [February 23, 2015]

Weststar, Johanna, & Legault, Marie-Josée. (2012). Facts and discussion about hours of work in the video game industry. In A. Bracken & N. Guyot (Eds.), Cultural perspectives of video games: From designer to player (online). Oxford, UK: Interdisciplinary Press.

  •  Announcements
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Current Issue
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Thesis Abstracts
    Atom logo
    RSS2 logo
    RSS1 logo

We wish to acknowledge the financial support of the Social Sciences and Humanities Research Council for their financial support through theAid to Scholarly Journals Program.