The dark side of innovation policy uncertainty

Using a state-of-the-art large language model, we construct an innovation policy index from public policy statements and measure the uncertainty surrounding innovation-related policy. Our theoretical model, framed in the real option framework, reveals that policy uncertainty has a dual effect on innovation in the presence of moral hazard: it encourages project initiation but also increases the likelihood of project exit through bankruptcy or premature abandonment. Furthermore, we identify a novel adverse selection mechanism, whereby more capable innovators may be more likely to abandon projects if their outside options are sufficiently valuable. Empirically, we find that regions with higher policy uncertainty and lower emphasis on innovation exhibit significantly reduced patenting activity.