Unions and Innovation

A useful study from Industrial Relations (65/4; Fall, 2010.).  Apparently we don’t screw things up for innovative firms.

What Do Unions Do to Innovation? An Empirical Examination of the Canadian Private Sector

Scott Walsworth
Associate Professor and Hanlon Scholar in International Business, Edwards School of Business, University of Saskatchewan, Saskatoon, Saskatchewan, Canada


This article uses Canadian national data to examine the union effect on product innovation, a firm outcome which is widely researched in the management literature but has been less prominent in Industrial Relations scholarship. Using a longitudinal sample from the employer survey of the Canadian Workplace and Employee Survey, the union effect on a firm’s ability to create or improve a product is examined. According to the commonly held view that unions impede firm performance, the results should point to a negative relationship between unions and product innovation. Interestingly, a strong negative effect is not observed. In fact a small statistically significant positive union effect is reported. This result is considered to be robust. Across various specifications the presence of a union and the intensity of the presence (firm union density) have significant and positive effects on a firm’s ability to innovate new products over a seven year period (1999-2005).

The results of this study do not imply that the presence of a union is an important determinant of product innovation. The results are noteworthy because they do not identify a negative relationship between unions and a measure of firm performance: product innovation. In this regard, the results give weight to the observation that there is very little empirical support for the popular argument that unions impede firm performance.

In Canada the demise of organized labour is often justified as a necessary adjustment to increasingly competitive markets. Indeed the signing of the 1993 NAFTA has put unions under greater scrutiny for their impact on Canada’s ability to compete internationally. That unions make firms less competitive is commonly accepted as a reasonable assessment. The results of this study and a review of the empirical literature on the union effect on other key firm outcomes such as productivity, labour costs, employment growth, sales and profitability, suggest that the popularly held negative assessment of unions is not based on a conclusive body of literature.


  • I have to get a blog post up on this but this is the general conclusion I came to in my dissertation based on a comparison between the UK, US, Swedish and German manufacturing sectors.

    The short of is that the US and UK with much much lower union density were witness to both a decline in value added and employment shares with little change in their technological composition (in terms of outputs. The German and Swedish manufactures both managed to stem their relative declines (as a share of VA), increase their profits and increase the amount of medium and high technology products as a percent of manufacturing output.

    And all this with very high union densities. In the literature two channels of causation have been suggested for why unions increase productivity and innovation. First is usually some sop to corporatist cooperation literature. The second which I find more intuitive is that higher unit labour costs require a continual search for more efficient means of production and higher value added product lines.

  • I worked for quite a number of years on the union productivity advantage. I did a couple of extensive case studies on manufacturing plants, actually designed a good portion of the Work Place and Employee survey from which this data comes from, and work quite a number of articles on industry examples.

    Once establishing that quantitatively that indeed the union advantage for innovation is real, the qualitative gets really interesting and deep into the IR cultural milieu. Basically a company within a unionized space and potentially faced with higher labor costs can either attempt union avoidance (deep south USA style), lead their own innovation programs and ignore the union or in some cases fight the union, or they can engage the union within the change process. The last can be further split into two general cases, the union and management cooperate and typically produce a management led change program or they negotiate the innovation and codify it within the collective agreement typically entitle union led.

    It is the last variety that in case study literature is more successful. And after all the research, my feeling is it is the inherent trust of workers. In a unionized environment based upon an adversarial IR system, there is not a lot of trust, and without that trust one does not get the required worker input or participation within the innovative process.

    It can be whatever sector, the worker knowledge over the actually existing labour process is a whole lot more keener than the random eye of the designers. When things go wrong, especially within a high value adding production process, the failures and offline analysis and knowledge require the trust of the operators and the designers. Trust is the starting point, but it is also the unions shop floor presence that amplifies that trust. Hidden with the democracy at work literature is that fundamental conclusion.

    The German works councils are an example of how one can go about entrenching from a public policy standpoint a legislative mechanism to aid the process.

    I was so captivated by the possibility of such projects, I took a job at Labour Canada and worked wih management and unions of many companies large and small trying to at these notions within our Canadian IR system, trying to grow a home grown version of policy that could improve and diffuse such policies. I think the best we could achieve were promoting Interest based bargaining and a few other tools within the various IR jurisdictions in Canada.

    That said, not every company is willing to work with its union to achieve such advantages. But I seen it in action and I do believe very much in the findings, and it could lead to a competitive advantage if one could ever implement such infrastructure within the workplace culture of every shop floor willing to try.

    It is not co-opting the worker, but the key is empowering the worker that the theory and practice resolves to.


  • That is all fine but in the aggregate data what we see at least for Sweden and Germany is successful rationalization, increasing diversification into higher tech products along side decreasing employment and decreasing compensation shares.

    So the question is to what end has this trust been put? I get the decreasing employment but why the decreasing compensation shares of (specifically) value added?

    That to the side if you ever feel like co-writing an article on this let me know. You can be the good cop I will be the bad cop.

  • Yes Travis, I am sure that was also a good part of the causation, and how rationalization and some of the other notions you mention are instrumental in bringing about success, and are negatively correlated with worker trust.

    And I would argue that when the outcomes result in pounding away at worker interests, the whole process of innovation and change lose that critical aspect of worker trust and potentially a lot of co-opting of the worker.

    I still strongly believe these aspects that go against the worker produce environments where worker involvement in the innovation and change process, is drastically curtailed.

    I guess it also must be tempered with the concept of what do we mean by innovation. Is it product innovation, production process or service delivery innovation, or organizational innovation (or some combo of all three which is typically the case).

    It is a huge subject and back in the late 90’s and early 2000’s I did strongly believe this is what is at the heart of a successful innovation within the workplace- an active and involved union where management is not left to its own devices and a countervailing structure is in place to ensure that the innovation process is in the best interests of workers. Whether it be, somekind of product innovation, that will secure jobs, or a production process innovation i.e. a process change that increases health and safety which results in compensation costs.

    The whole area as far as I can determine is quite immature within the practice and theory but does hold so much potential for change within a workplace culture can translate into productivity enhancements and concomitantly translate into worker improvements within the workplace. Especially within a transforming workplace of high value adding workplace settings.

    It hold some the potential that can contribute to increasing the competitive advantage of a nation, as most IR systems are nation specific.

    Now how well that translates into traditionally non-union occupations and industries is a whole different kettle of fish.


  • Hey Travis, can you post a citation for your dissertation? I am collecting an annotated bibliography on this stuff and would like to add a reference. Thanks! Thanks also Andrew for this interesting and useful post.

  • Hi Jim,

    When it is offically published I will send the link along in the meantime I can send you a PDF of the relevant chapters if you like. Send me an email.

  • When is it getting published? No pressure- lol

    send me one as well.

    I’ll would be glad to write something on this. Are you still in academia? Maybe we can get a hold of the WES data through the RDC, I do believe the DLI has stocked all years of the WES data in the RDCs. I worked 4 years on the design of that survey but did not get much of a chance to research it. I would like to explore the union linkages further. Recall that this survey was a two stage sample, employers were surveyed and then a small systematic sample of workers were selected.

    The linkage was a key aspect of the survey, as worker outcomes could be linked to the employer outcomes and the reciprocal.It had limited public sector coverage which was too bad.

    Being a longitudinal survey offered some unique aspects for research, although it ramped up the technical and methods skills needed to ascertain statistically robust outputs.

    It was such a cool experience. Sadly I never got much of a chance to research the file.

  • Yes still stuck in the academy.

    Yes we should write something. The topic is still quite “hot” in Quebec and a perennially important subject to industrial relations in general. I will send you an email and we can work-up something formal.

Leave a Reply

Your email address will not be published. Required fields are marked *