Learning-by-failing. An empirical exercise on CIS data

Failure to innovate has been only recently recognized as one of the key elements in determining successful firms' innovative performance. However, as this literature focuses only on the determinants of firms' failure, it neglects the role of failure in spurring innovative activity. In this paper, the relationship between innovative performance and failure to innovate is empirically tested, through a two step econometric model, on the 2008 CIS Innovation survey dataset. The main results of the paper are, first, that failure is negatively correlated to the firms' experience (proxies by R&D), and to the acquisition of direct external knowledge (through productive links in product and process innovation). Indirect learning from the failures of similar firms is moderated by firms engagement in R&D and in searching for external knowledge. The second step reveals that failure in turn has a positive impact on performance in term of percentage of turnover from new to the market innovative products. Finally, an additional test is performed on still ongoing innovation (rather than abandoned), and the results show a minor impact on innovation activity.