Does red tape not matter for firm productivity?

This page in:

A new World Bank working paper finds that the answer, counterintuitively, is 'yes'. De Rosa et al. look at a large sample of firm-level surveys completed in 2009 and find that:

...while the payment of bribes is negatively associated with the productivity of the bribing firm, time spent dealing with bureaucratic requirements per se appears to be irrelevant.

This result is quite surprising, considering that managers of firms in many countries spend large chunks of time on pointless bureaucratic procedures. Should we give up on trying to slash regulatory burdens, contra this blog post by World Bank economist Mary Hallward-Driemeier? Not quite so fast!

There is one aspect of the De Rosa paper that leaves me unconvinced. It looks only at firm-level data from a single year. (Footnote 7 confirms this point.) But firm-level data of this type suffers from 'survival bias'. We only get responses from the firms that have managed to survive the particular regulatory and economic environment they find themselves in. In poor regulatory environments, only those firms that have managed to adapt themselves to burdensome regulation will show up in the sample. But a country that slashes red tape might see a productivity boost not only in existing firms, but also in new firms that are more suited to take advantage of a better regulatory environment.  

Of course, this doesn't mean that the conclusion the authors reach is necessarily wrong. But to prove the point that red tape doesn't matter for firm productivity, they would need some kind of panel data to assess the impact of changes in business regulation on firm productivity over time. In the meantime, policymakers in countries with poor regulatory environments ought to keep their focus on slashing red tape.

(Thanks to Christian von Drachenfels for the pointer.) 


Authors

Ryan Hahn

Operations Officer

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000