Innovation at firms

Innovation at firms

/ Wpis z serii: ASSA 2014 brought to you by GRAPE :) /

By now, everyone accepts that innovation is THE key to economic development. This affirmation seems to apply to every country in every period, but to be even stronger in the present, with the outstanding technological growth of the last 20 years, the internet revolution and the important increases in productivity that innovation introduced. Then, it is not surprising that several of the papers presented in the American Economic Association annual conference dealt with this topic. Two of the contributions stand-out by their merits: “The causal effect of labor unions on innovation” by Bradley et al. and “Why do Regions Vary in their Response to Crowdfunding? The Young, Restless, and Creative" by Agrawal et al. The first offers a new view into a traditional conception of innovation, namely innovation at the workplace. The second provides an analysis of crowfunding, a platform that allows a diffuse innovation process by granting credit to entrepreneurs, we review this article here

As any firm engaging in R&D could tell, innovation is very much a random process. Even though the discovery of a new technology is influenced by the effort put on by the researcher, the final success or economic profitability of the research is to a large extent stochastic. Moreover, the simple act of research is completely different from other economic activities, as the results are only visible in the long run. Then, it is natural to ask what factors could improve the chances of coming up with a bright and profitable idea. This task was undertaken by Bradley et al., who examine the effects of the unionization level in firms’ innovativeness.

In theory, the effect of unionization could go either way. By providing job security, researchers (and plant level workers) can feel more comfortable expressing their ideas and proposing innovation projects. In addition, this environment seems more prone to tolerate short term failures, such as those present in the innovation process, in the search for a long term benefit. However, the same job stability provides incentives to shrink, not only on the scientists, but also on the assistant personnel, which decreases the productivity of workers. Moreover, as unions capture a part of the rents generated by the workers, the most talented researchers would obtain higher wages in firms without union, and will have incentives to migrate there. Finally, when unionized, the employee cannot commit him/herself to not ask for a raise once the firm incurred in some sink costs (buying the equipment for the lab); foreseeing this problem, firms might underinvest in R&D as well. 

In their research, the authors use a regression discontinuity design to determine the influence of unionization in quantity and quality of the innovations. In their research, negative effects prevailed, with an impact that increases in each year during the first three years. Moreover, the innovations seem to be of worst quality (measured by the number of citations per patent) in the same period of time. The authors found also evidence that all the causal mechanisms described before are jointly responsible for this negative effect. 

Do these results indicate that we should engage in a deunization process? Well,… not that fast. Some of the results are pretty much dependent on the particular characteristics of the American labor unions. In particular, each firm can decide whether to unionize or not, if this option is not presented, then talented researchers have less incentives to migrate to other firms. On the other hand, it is not clear that deunionization will reduce the reverse hold-up problem and increase expenditure in R&D. The results of Bradley et al’s research are a first step and provide us with an interesting starting point to think about the influence of the work environment on innovations.