Yesterday I read this Project Syndicate comment by Kermal Dervis arguing that the returns to publicly-funded R&D have been essentially privatised – a point also made by Mariana Mazzucato including in her new volume, Mission Oriented Finance for Innovation. The comment includes some interesting suggestions such as a publicly-owned venture fund, although Nesta’s Stian Westlake has pointed out that there are some problems with the broader argument. I’d add to his list the basic argument about the diseconomies of centralised decision-taking when you’re talking about ideas, something amply demonstrated by20th century experience.

In particular, Stian argues, and I completely agree, that it is both necessary and hard to account for the full social welfare effects of publicly-financed innovation – these go beyond the financial return. The discussion yesterday on Twitter sent me back to [amazon_link id=”0231152140″ target=”_blank” ]Creating a Learning Society[/amazon_link] by Joseph Stiglitz and Bruce Greenwald. This is a book I’ve yet to work through properly, and it’s quite a demanding tome. Looking today at the key chapters only underlines the challenge of thinking through social welfare – and therefore policy implications – of innovation in the modern economy. For starters, ‘knowledge’ has the characteristics of a public good and a natural monopoly. Innovation requires accounting for risk, in imperfect capital markets – and there’s pure uncertainty too. There are complicated dynamics, path dependence, and probably instability.

[amazon_image id=”B00NIC3A1I” link=”true” target=”_blank” size=”medium” ]Creating a Learning Society: A New Approach to Growth, Development, and Social Progress (Kenneth J. Arrow Lecture Series) by Stiglitz, Joseph E., Greenwald, Bruce C. (2014) Hardcover[/amazon_image]

It leaves me thinking (a) we’re a long way from being able to make definitive policy recommendations about how to boost innovation, although I’m sure it includes state as well as market activities. And (b) there’s a lot of work economists need to do on standard welfare economics, which has a dusty 1970s (or earlier) feel to it.