Firm asymmetries and sequential R&D: Theory and evidence from the mainframe computer industry

We incorporate strategic considerations into the analysis of a problem that has hitherto been treated in a decision theoretic fashion: the allocation of scarce R&D resources when R&D proceeds in stages. In doing so, we formalize a notion of ‘system complexity’ and investigate its implications for the allocation of these scarce resources. Using detailed data from fieldwork at all mainframe manufacturers in the world to investigate our theoretical predictions, we provide evidence that larger market share firms set more aggressive stage targets, as do more resource-rich firms. Our results can be seen as a verification of the mechanism underlying Arrow's ‘replacement’ effect.