Thread by @JWaldfogel

Colleagues and I have a new paper on GDPR whose abstract has prompted a little Twitter heat. I’ll add my perspective, consisting mostly of cool water. 1/n
@JanssenReb @reinholdkesler Michael Kummer
nber.org/papers/w30028

GDPR aims to improve privacy in various ways, some of which are costly to app developers. As a result, GDPR induced a large number of apps to exit rather than face the costs of coming into compliance.

This exit spike is probably good news all around. The exiting apps had very little usage and, on top of that, were not compliant with privacy regulation. An extinction event, but for intrusive creatures.

Far more important for consumers, in the long run, is GDPR’s effect on app entry. To understand the possible effect of depressing entry, it’s important to step back to think about how innovation works.

Apps are “nobody knows anything” products. As with creative products and many others, it is very hard to predict which newly launched apps will eventually find favor with consumers.

In such contexts, society better finds more “winners” (valuable products) – along with a bunch of additional “losers” – when producers are able to take more draws from the innovative urn (i.e. when more new products enter).

Using this perspective in a 2018 paper, Luis Aguiar (@luisaguiarw) and I quantified the benefit of a digitization-fueled tripling in new music entry, finding that the welfare benefit is an order of magnitude larger than the standard “long tail” measures.

Back to GDPR: by reducing the number of new products entering, GDPR is like “digitization in reverse.” The number of new entering apps falls by roughly half, and – this is the important part – the number of new apps eventually garnering large numbers of users falls similarly.

It’s almost as if app success were entirely unpredictable. As a result, we calculate the long run effect of GDPR to be a substantial negative hit to the well being of app users in the conventional economic sense (consumer surplus).

Does this mean that GDPR is bad? No: Answering that question requires a quantification of GDPR’s benefits. Our point is just that, whatever the benefits of GDPR’s pro-privacy provisions, they come at a cost in foregone innovation. n/n