Research

Are ‘Complementary Policies’ Substitutes? Evidence from Public Spending on R&D (JMP)

This paper tests whether grants (direct funding) and tax incentives (indirect funding) for R&D are complements or substitutes. Governments have a long history of subsidising private R&D, often through a mix of support mechanisms. However, there is little evidence regarding how such policies interact. I develop two quasi-experimental research designs — one combining regression-discontinuity and instrumental variables methods and another combining difference-in-differences and regression-discontinuity methods — to disentangle the effects of R&D tax incentives and grants on innovation. I build a database of R&D, performance, and innovation outcomes by linking several micro-business datasets to study this question for UK firms. Preliminary results show that they are complementary for producing some types of innovations, but surprisingly, not all. Continued work explores why this is the case, what this means for spillovers, and how policy substitution steers the direction of innovation.

Disentangling Uncertainty and Salience: The Case of Solar Self-Consumption (with Eoghan McKenna)

Uncertainty and salience can have dramatically different policy implications, but the distinction between them is not always entirely clear. We use high-frequency data on solar energy generation and electricity consumption to disentangle rational habits from endowment salience effects. We test how providing real-time information regarding onsite solar generation affects household electricity consumption substitution patterns. Preliminary results suggest that households behave rationally towards uncertainty but do not fully optimize with respect to endowments that are imperfectly salient. With increased salience, the endowment elasticity increases by 4 percent on average and up to 18 percent depending on time of year and day. Continued work is applying econometric and machine learning methods to derive salience-adjusted self-consumption counterfactuals and welfare effects. The methods can be applied in other settings with misoptimizing consumers.

Pass-Through as a Test for Market Power: An Application to Solar Subsidies (with Arthur van Benthem) (Submitted(previously “The Surprising Pass-through of Solar Subsidies“), NBER Working Paper No. 23260.

We formalize pass-through over-shifting as a simple yet under-utilized test for market power. We apply this test in the market for solar energy. Specifically, we estimate the pass-through of solar subsidies to solar system prices using rich micro-level transaction and subsidy data from California. Buyers of solar systems capture nearly the full subsidy, while there is more-than-complete pass-through to lessees. We conclude that solar markets are imperfectly competitive by ruling out alternative explanations for over-shifting, and reinforce this conclusion with a test of solar demand curvature. This procedure can serve to detect market power beyond the solar market.

“Identifying collaboration and product market rivalry effects on innovation” (very preliminary)