| Summary/Abstract |
Rather than by an invisible hand, many industries are kick-started by a government policy. Despite little robust evidence, local content requirements are increasingly used to incentivize domestic manufacturing if imports are cheaper. To examine the effect of local content, we explore an unintended quasi-policy-experiment. Starting in 2013, the Indian government simultaneously held solar auctions with and without local content, providing an otherwise unobserved counterfactual. We digitize the results from the 41 auctions worth 8.65 billion $ in solar module demand and collect annual revenue and solar patents of the 113 participating firms between 2004–2020. For causal identification, we compare winners of local content with similar open auction winners in a staggered difference-in-difference estimation. While we observe an insignificant increase in the same and the following year after firms win LCR auctions, overall, we find winning local content auctions does not significantly increase firms’ solar patents or sales. We identify three reasons why the policy did not create stronger, lasting effects. First, local content did not create sufficient production to enable learning by doing. Second, local content did not generate enough revenue for re-investment into R&D. Third, local content reduced competition in auctions. The analysis underlines the predicament countries face as open auction winners, despite having won 9 times as much capacity, do not patent much (more).
|