Dear taxpayers: Thank you for your support.

Photographer: Daniel Acker/Bloomberg

Angel Investing, Government-Style

Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.
Read More.
a | A

Should the government be an angel investor? Mariana Mazzucato thinks so. The University of Sussex economist and author of the bestselling book "The Entrepreneurial State" argues that government funding is essential to the research and development that creates the companies of tomorrow. That includes funding for basic and applied research -- an example would be the Defense Advanced Research Projects Agency (DARPA) funding that created much of the Internet -- and also funding for startups that develop their technology in-house.

Talk to an American conservative, however, and you're likely to get a very different answer. You'll probably hear about Solyndra, the government-backed solar company that went belly up in 2011 after receiving $535 million in government loan guarantees.

You probably won't hear about Tesla -- a company that obtained a $465 million low-interest loan from the U.S. and is a wild success, pushing the boundaries of energy-storage technology while achieving an enormous market capitalization. Despite success stories like Tesla, it's failures like Solyndra that stick in the minds of people who assume that the government always gets everything wrong.

Well, now people like Mazzucato have a little more evidence in their arsenal. A recent paper by Sabrina T. Howell, an assistant professor of finance at New York University, examines what happens to companies lucky enough to get government loans. Her findings support the notion that the "entrepreneurial state" can play a big role in helping innovative companies succeed.

There's a problem with simply showing that companies that get government support do well. The government may simply "water the green spots" by funneling resources to companies that didn't need them and would have done just as well -- or better -- without them. But Howell's research gets around this problem using a popular form of statistical analysis called "regression discontinuity design."

When the U.S. Department of Energy's Small Business Innovation Research program tries to decide which companies to support, it ranks them and gives support to all the applicants that fall above a certain cutoff. But that cutoff is arbitrary; a company that barely makes it isn't much different from one that falls just below it. It's just luck. Howell compares the performance of the lucky companies with the unlucky ones. This is the kind of simple but powerful statistical analysis that has been gaining popularity in the economics profession in recent years.

It's hard to measure success, so Howell looks at a number of different characteristics. First, she looks at startups' success in getting venture capital down the road. She finds that even very small grants can hugely boost a firm's chances of obtaining VC funding. This finding should calm the fears of people who think that government angel investing is crowding out VC. It looks like the two are acting as complements, with government taking on the super-early-stage risk and handing things off to venture capitalists in later stages. Howell finds that grant recipients are more likely to go public or to be acquired, allowing private individuals to capture much of the return from the government's activity. 

Howell also tracks grant recipients' real performance. Government grants substantially increase companies' chances of making something that actually sells and increase the amount of patents they generate. Those companies are not, however, less likely to fail. Government grants, it seems, can help with growth, but they can't create good ideas out of thin air. 

Why would government funding help make young companies more likely to grow? One possibility is that grants may simply tip the scales in a winner-take-all market, giving lucky companies the jump they need to out-compete the unlucky ones. An alternative explanation is that very young companies are severely financially constrained. These companies are the youngest of the young, with unproven technologies and business plans. Even venture capitalists, who seek out risk, may balk at the uncertainty. 

Howell finds support for the second theory. The effect of government grants is much stronger for younger companies and those with less-proven technologies. In addition, the effect is stronger when financing conditions in the economy are tight or when venture capital itself isn't doing well. In other words, government grants really do look like they are providing financing that the private sector can't. 

So here we have a bit of evidence in favor of the theory of the "entrepreneurial state." It's just one paper, of course, and there will be others. But the U.S. should take the idea seriously, especially now that public-sector research funding has been slashed. If government really is as big a driver of innovation as these economists think, we could be in big trouble if this trend continues. 

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Noah Smith at nsmith150@bloomberg.net

To contact the editor on this story:
Paula Dwyer at pdwyer11@bloomberg.net