Abstract
There is consistent evidence in the literature that an average employee's age is negatively related to firm-level innovativeness. This observation has been explained by older employees working with outdated technological knowledge and being characterized by reduced cognitive flexibility. We argue that firms can mitigate this effect through employee turnover. In particular, turnover of R&D workers is deemed a vehicle for transfer of external knowledge to the firm, which can compensate for lower cognitive flexibility and up-to-date knowledge among older workers. We use a matched employer–employee data set based on three consecutive Community Innovation Survey surveys for Sweden to test our predictions. Our results suggest (a) that overall employee age impacts negatively on product innovation activities (both in terms of propensity and success), (b) that the effect of employee staying rate (measured by the share of employees who remain in the firm from one year to the next) on innovation follows an inverted U-shape implying an ‘optimal’ level of employment turnover and (c) that this ‘optimal’ value is lower for firms with older employees. The latter suggests that firms with older employees can at least partially compensate an aged workforce by increased employment turnover.
Acknowledgements
This work has been supported by financial grants from the Swedish Research Council (Linnaeus Grant No. 349200680) and the Swedish Governmental Agency for Innovation Systems (Grant agreement 2010-07370). Martin Andersson also acknowledges financial support from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS), Dnr 2011-80.
Notes
1. Some potential sources of bias need to be addressed: the stratification Statistics Sweden employs in the CIS may work towards larger firms being included (as all Swedish firms with 250 or more employees are included as long as they are in CIS-relevant sectors). To circumvent problems of identifying firms, we have disregarded those firms that may have changed ownership structure, since that would also imply changing organizational identifier.
2. Details on the CIS survey in Sweden may be found in statistical reports such as Innovation activity in Swedish enterprises 2006–2008 by Statistics Sweden (CitationSCB 2009).
3. Note that the staying rate was set to 1 for firms without R&D-related employees, because in this case the staying rate of 100% is deemed to capture that fact there is no inflow of new ideas and thus no additional creation of variety.
4. We checked alternative specifications (see the robustness Section 4.2).
5. While this evidence is, given the significance levels, somewhat marginal, this is certainly also due to the fact that we had to split samples.