Exploration or exploitation of knowledge from universities: does it make a difference?

Abstract

In the first step, based on the existing theoretical and empirical literature, we develop a series of hypotheses with respect to the relative importance of possible determinants of exploration and exploitation of knowledge in collaboration with universities and test them on Swiss firm data. In the second step, we investigate the impact on innovation performance of knowledge exploration versus knowledge exploitation. We obtain a clear pattern of the differences between firms that are engaged both in exploitative and explorative activities (‘exploration’-oriented firms) and purely ‘exploitation’-oriented firms. We find that exploration-oriented firms have a greater knowledge absorptive capacity, are technologically more diverse and are strongly exposed to intensive non-price competition compared with exploitation-oriented firms. We further find a positive effect on innovation performance for exploitation-oriented firms but not for those that were exploration oriented.

Keywords:

• knowledge exploration
• knowledge exploitation
• innovation

JEL Classification:

• O31

Notes

1. Versions of the questionnaire in German, French and Italian are available at www.kof.ethz.ch.

2. Teece, Pisano, and Schuen (1997) mention several reasons for the persistence of the knowledge base: firms lack the organizational capacity to develop new competences (or interest for scientific questions), some assets are not tradable (e.g. tacit knowledge), and needed inputs have to be bought at relatively high prices that reduce possible rents.

3. However, Bierly, Damanpour, and Santoro (2009) found a positive correlation for their measure of absorptive capacity (‘technological capability’) only for exploitation.

4. The robustness of this result is tested by additional estimates not shown here, in which eight dummy variables for all feasible cases in our sample (instead of EXP1, EXP2 and EXP3) are inserted in the estimation equation: (1) no exploitation/exploration; (2) exploitation/only new products; (3) exploitation/only new processes; (4) exploitation/new products+new processes; (5) only exploration; (6) exploration/exploitation/only new products; (7) exploration/exploitation/only new processes and (8) exploration/exploitation/new products+ new processes (see Table 2). We obtain significantly positive marginal effects only for the dummy variables for (1), (2) and (3).

5. Woerter (2009) found for Swiss firms a positive effect of a patent-based measure of technological specialization – which is different from the measure used in the paper at hand – on the sales shares of innovative products. Also, Bolli and Woerter (2013) found for Swiss R&D-active manufacturing firms that technological specialization is positively related with the sales share of new innovative products.