An economic approach to mitigating spillover bias in synthetic control methods: An application to the price effects of fishing quotas

Abstract Individual Fishing Quotas (IFQs) have the potential to lead to higher ex‐vessel prices by reducing market gluts and improving product quality. While IFQs have been introduced in a number of fisheries in recent decades, few causal studies investigate their price effects. Using synthetic control methods on US reef‐fish fisheries, we show that IFQs significantly increase prices for red snapper, red grouper, gag grouper, yellowedge grouper, scamp grouper, and tilefish. We also find significant positive price spillovers to adjacent, untreated fisheries. A primary challenge to unbiased estimation using synthetic control is treatment spillover, which can contaminate both never‐treated units in the donor pool and to‐be‐treated units in a staggered implementation. When control (donor) units are subject to these spillovers, this can lead to either attenuation or divergence of the causal estimate, depending on the empirical setting. To address this estimation challenge, we propose a framework relying on economic theory (market arbitrage and product substitution) and empirical analysis (e.g., market integration or demand estimation) to inform both donor selection and the timing of the treatment start date. As a result, our proposed structure improves the accuracy of synthetic control estimates, as shown by a small Monte Carlo experiment. Furthermore, by pre‐identifying contaminated units, our framework allows for the direct measurement of the spillover effects themselves.