I found myself frustrated by her conversation with Russ Roberts. She repeatedly points out that government has funded successful innovations. The implication, in her view, is that we should pay higher taxes in order to fund more innovation.
But the question is whether the government or the private sector is structurally more suited to spending money on innovation. It seems to me that Russ could not get Ms. Mazzucato to discuss the issue at that level. Trading anecdotes back and forth about particular private-sector and public-sector successes and failures is not constructive. I gave up listening well before the podcast ended.
I like to break down innovation into experimentation, evaluation, and evolution. I think that government has a disadvantage at all three.
Any one organization is limited in the number of experiments on which it can focus. When the government was focused on landing on the moon, it succeeded. But it is hard for any one organization, not just government, to focus on a lot of experiments at once.
When it comes to evaluating experiments, everybody is overly optimistic about their own projects. But in the private sector, the subjective evaluation of the project champions is subordinate to the third-party evaluation of consumers. In the public sector, the legislators and bureaucrats who champion a project also are doing the evaluation.
Finally, when it comes to evolution, government has no natural mechanism to shut down unworthy projects. When a private-sector project is not producing value, it loses money and gets terminated.
Twenty years ago, I wrote an essay in which I explained why, contrary to what some pundits were arguing, middle managers should not be encouraged to take more risks inside corporations. In essence, I made a “skin in the game” argument. That argument applies very strongly to the case of government-funded risk-taking, in which the cost of failure is not borne by the decision-makers.
In biotechnology (which is the field I know best), most of the big successful innovations involve government research grants and private venture capital. This system works because of support from many other institutions (it’s not just the NIH and the VCs, there’s also the universities, the journals, the patent system, the FDA, …). Classifying institutions as “public-sector” or “private-sector” just buries the key questions about whether these institutions are working well or not.
I believe her point would be that we could make government more suited to spending money on innovation by looking at where government has successfully sparked innovation (DARPA, early NASA) and by training bureaucrats more like MBAs.
I had a similar reaction, although I did listen through the end. She kept arguing a straw man relative to the points Russ was trying to make. Just talking past him with such a simplistic argument. The issue comes down to whether the government has the proper incentives to successfully expand its funding of innovation.
If we are looking for a structural reason why government might be suited to spending money on innovation, it would be for the reciprocal of why it isn’t suited.
Is a pattern of innovation that only follows profit motive justifications sufficient, or is there also value in forcing some fundamental forms of research without a clear path to profit?
The other justification for research is that the government is a huge consumer in the market.
The only value of the moonshot program was that it tricked the federal government as a consumer to demand significant improvements in performance from the market. It demanded that the computer industry solve problems. It demanded material research improvements. It even demanded improvements in food packaging.
Does the federal government demand improvements in how highways are paved? Is it a force in improvements to software? Not really.
It is difficult politically to force the government to demand things of the private sector, but when it does, it can be powerful, because the government is a huge consumer, and public consumption is usually much less demanding than private sector consumption. If we can just do better there, we’ve gotten somewhere.
That interview was replete with bold but highly questionable assertions: the internet and GPS would not exist absent government “investment”, unions ended child labor and got us the weekend, bailouts saved “capitalism”, and so on.
Russ avoided getting distracted with those, and raised the fundamental issue of opportunity costs, incentives and knowledge a few times. But Mariana kept emphasizing benefits without considering costs or opportunity costs.
She could use some more Hazlitt: “The art of economics consists in looking not merely at the immediate but at the longer effects of any act or policy; it consists in tracing the consequences of that policy not merely for one group but for all groups.” 😉
The mentioned government investments were aimed toward the satisfaction of government requirements, to show up the Russians, to increase the ability to spy on and kill people. Those investments weren’t made to benefit consumers, those are after the fact benefits, turned into consumer values by the commercial sector.
The result is that the consumer benefits are somewhat negated by the absence of any returns for landing on the moon or the ability to guide expensive missiles to designated targets.
On the other hand, investment in R&D in the commercial sector is aimed at creating value for consumers without the cost of expanding the ability of government to spy on, control, and kill people.
The space program and a few other big projects are always cited. But then they had good PR and marketing. But while the space race was chugging along, both the CONEX box, specifically the locks that made containerization possible and the PLC (Programmable Logic Controller) were innovated by the private sector. Both had far greater impact on prosperity and the GDP.
The PLC, just over 50 yrs old, is a good example of the private market killing projects when a better alternative comes available. BTW, PLCs run pretty much every manufacturing line in the world today. It cut months off the retooling for a new model year in Detroit. GM had issued a request for proposals, many were working on the problem, but Dick Morley was unaware of the GM request when he laid out the design on January 1, 1968. He was just tired of solving the same problem. But he went for robustness, which let to his newly formed company to cause Digital to kill their program.
As Morley related:
“The real test went something like this. Our competition was the PDP-14. It was designed to GM specifications for a controller. I guess GM thought they could design computers as well as automobiles. Stan Schoonover of Landis had early experience with both our Modicon unit and the Digital Equipment PDP-14 and he selected the 084 Modicon (now Schneider) unit as his preferred component. Well! The Digital salesman rankled. Stan held his ground. So the Digital man brought down Ken Olsen to see Stan. At the time, Mr. Olsen was president and cofounder of Digital Equipment Corp. (the Bill Gates of the late ’60s). Stan showed Mr. Olsen the 084 setup running a small control problem. The unit had welding cables wrapped around it with the welder being used and it ran fine. Stan then took his Coke and poured it over the 084 – it still ran fine. The experiment also ran next to a huge motor starter. Stan said, “When I can do this to your PDP-14, I will buy some.” Mr. Olsen soon cancelled the PDP-14 program.The PLC has two things going for it: ladder logic and good hardware. It matches the mind of the user and the environment of the application. Blue collar software and hardware did the job. It has no blue screen of death.”
http://www.automationmag.com/opinion/distorted-realities/877-christening-the-controller-what-you-dont-know-about-the-birth-of-the-plc
Now had government been financing Digital to produce a PLC?
I’m gonna play devil’s advocate and try to counter some arguments, see what we can get out of this conversation.
> But it is hard for any one organization, not just government, to focus on a lot of experiments at once.
Is there any quantitative evidence about this? It seems to me it ain’t that hard to find some organizations that are pretty good at several things at the same time: See Amazon’s ecommerce/AWS/electronics divisions, Google’s search/email/autonomous vehicles businesses, Virgin several branches, some Japanese firms like Mitsubishi, Sony… Pretty sure they didn’t start all their divisions at the same time but we wouldn’t require that from the government either.
> But in the private sector, the subjective evaluation of the project champions is subordinate to the third-party evaluation of consumers.
That is only true of markets with symmetry of information and perfect competition, it doesn’t hold in a lot of cases. Some examples: Natural monopolies, where you are stuck with the de facto option (try switching to Facebook’s competition); Oligopolies, which tend to prevail in necessity good markets such as electricity providers and where providers often fix prices (I know it’s illegal but gains are so high they outperform disincentives from sanctions, just search “fix pricing scandals” in google).
> When a private-sector project is not producing value, it loses money and gets terminated.
Again, only in cases of perfect competition and symmetry of information, it doesn’t hold true in, for example, oligopolies whose coordinated failure could pose a systemic risk: see the financial sector during the ’08 crisis. See also any public bailout or treasury loan in which the risk is transferred to the taxpayer; a.k.a no “skin in the game”.
This is not to say, of course, that the government should be a player in every market but that there might be some situations in which it could outperform private institutions at innovation.
In the real world there is never perfect competition or perfect symmetry of information. There are only degrees.
I think Arnold was trying to say that *in general* whether consumers are willing to buy matters more to a profit-seeking firm than to a government agency, and that failure to please consumers is more likely to result in termination of a profit-seeking firm project than a government agency project. I think he is correct in that.
Your financial crisis example is an interesting one. In a purely private financial system, those firms would have indeed gone out of business. Instead, some were saved (e.g. Bear Stearns) and others (e.g., Lehman Brothers) were allowed to go bankrupt, at the decision of the government.
Milton Friedman taught us that when we spend money on ourselves we want the best quality for the lowest price. When one (government) spends someone else’s money for the benefit of a third party you get neither quality or low prices.