Policies that Affect Growth Rates: Question for Scott

I enjoyed Scott’s original post, but I kind of agree with Bryan. Isn’t this the basic idea from McCloskey? That most growth comes from a rhetorical attitude toward innovation? Maybe if a country is tied into the global economy enough, the externalities of growth from other economies will seep into their own, so maybe today it would be especially difficult to have policies so bad that they permanently affected growth rates.

Actually, I think I have an example. My obsession lately has been housing. I am becoming more and more convinced that limits to housing in a few large cities (NY, San Fran.) is creating a persistent drag on real GDP growth. This is because the dynamics of those cities create high productivity, and artificial limits to housing supply divert that productivity to workers, and ultimately to real estate owners, there. So, as we gain national real GDP growth, a portion of that growth is used to bid up real estate in those cities.

There is a natural resource, which is the space in the air above Manhattan that is currently uninhabited, which represents a larger loss to us every year that the workers who could have lived there become more potentially productive.

We can see the effect directly by looking at how much inflation is coming from excess rent inflation. Each year, real GDP growth is lower than it should have been because nominal GDP growth came out of inflation of existing housing instead of the release of those natural resources hovering in the air over Manhattan.

Here, I discuss a paper from Hseih and Moretti that gets at this problem. They estimate a cumulative loss of nearly $2 trillion in real GDP.

Another way to look at this is in BLS data on housing expenditures. Over the past 30 years or so, nominal spending on housing has remained right about 18% of PCE, but real housing consumption has steadily declined over that time. One way to think about that is that the BLS doesn’t make hedonic adjustments for the economic rents available to households who manage to get into housing in those cities. If we could have built a thousand residential skyscrapers over the past few decades, we would still be spending 18% of PCE on housing, but we would have more housing, and the workers living in those houses would be creating consumer surplus for the rest of us instead of capturing rents and transferring much of them to landlords. In effect, the unnecessarily high cost of living in those cities is a measure of lost consumer surplus to the rest of us.

Dr. Pangloss: How I Learned to Stop Worrying About Policy and Love the Regulatory State

Does the PTO issue too many weak patents, creating thickets that may retard innovation? Is patent protection too long or short? Too wide or narrow? Is the FDA’s regulatory framework keeping drugs optimally safe or impeding pharmaceutical innovation? Is merger enforcement deterring monopolization or pouring sand into the gears of creative destruction?

Thank goodness we don’t have to worry much about any of this, or a myriad other policy questions, given that at best it implicates a mere transitory bump in levels, leaving the growth juggernaut unphased.

My only regret is that Britain didn’t institute a policy banning the burning of coal in the 18th century. We could have had all the growth of the Industrial Revolution, minus a transitory level change, without the threat of global warming.

Maybe a better position is that growth is largely unimpacted by policies within the set of commonly accepted and seen policy. Bryan’s example is probably true but we are unlikely to see policy like this, though evidently Scott disagrees.

I do think Scott is right about most policies. Yesterday, I was at a walk-in clinic and some old fellow there was in worse shape than the clinic could handle, so they must have called 911. About 6 firemen walked in the place. Only 2 or 3 could even fit in the exam room. And, even the first was redundant, because if the man really needed to be transported to the hospital, an ambulance crew would need to come. I was thinking about the waste, but it occurred to me that in some ways this was a hopeful situation. Past economic growth has been hampered by this redundancy. But, at this point, we can’t fit any more firemen in the truck, and we aren’t going to start sending 2 trucks to be redundant attendants. So, future growth should be unaffected, and if we somehow decide to address the problem, we might get a one-time boost.

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]

David Kushman has the simple answer. Growth in GDP per capita comes from growth in productivity, and we can break labor productivity down into technology, capital and how efficiently these are used. There’s a limit to how much we can improve efficiency (given technology) so changes that improve this produce level gains but only temporary growth gains. Extrapolating from the past, changes in technology through innovation do not seem to run out.

Regulations, such as occupational licensing, are about efficiency and so are primarily about level effects. Killing anyone who innovates would likely affect the rate of innovation and so would have a persistent growth effect.

The problem is that most of the suggested changes in regulation seem to be more about efficiency. But that doesn’t rule out the possibility of doing things to improve the rate of innovation; AlexR’s patents relaxation might be the best example.

This is all very interesting. The long term compounded growth rate of per capita output (or any variable) is a function of the the average per period growth rate and the volatility of per period growth rates. A very close approximation is compound growth rate = mean(rate) – stddev(rate) / 2.

It follows that if a policy can reduce the variability in single period growth rates without affecting the mean single period growth rate, that policy will increase growth just as a matter of arithmetic.

Alternately, if it is true that no policy can affect growth rates then it must also be the case that no policy can reduce the volatility in single period growth rates without also reducing the average single period growth rate.