R&D Networks under Heterogeneous Firm Productivities

We introduce heterogeneous R&D productivities into an endogenous R&D network formation model, generalizing the framework in Goyal and Moraga-Gonzalez (2001). Heterogeneous productivities endogenously create asymmetric gains for connecting firms: the less productive firm benefits disproportionately, while the more productive firm exerts greater R&D effort and incurs higher costs. For sufficiently large productivity gaps between two firms, the more productive firm experiences reduced profits from being connected to the less productive one. This overturns the benchmark results on pairwise stable networks: for sufficiently large productivity gaps, the complete network becomes unstable, whereas the Positive Assortative (PA) network -- where firms cluster by productivity levels -- emerges as stable. Simulations show that the PA structure delivers higher welfare than the complete network; nevertheless, welfare under PA formation follows an inverted U-shape in the fraction of high-productivity firms, reflecting crowding-out effects at high fractions. Altogether, a counterintuitive finding emerges: economies with higher average R&D productivity may exhibit lower welfare through (i) the formation of alternative stable R&D network structures or (ii) a crowding-out effect of high-productivity firms. Our findings highlight that productivity-enhancing policies should account for their impact on endogenous R&D alliances and effort, as such endogenous responses may offset or even reverse the intended welfare gains.