David R. Henderson  

The Merits of Cochrane's Case

Testing Unflattering Claims Ab... Is an Unused J.D. a Negative S...

I posted this last weekend about an important history of thought mistake in a recent John Cochrane post. Although I said that his post was "otherwise excellent," I didn't elaborate.

Now I do.

John made the case, contra Noah Smith, that it was plausible, with enough reversals and abolitions of regulations and government spending programs, to raise the growth rate of real GDP to 4% a year for a decade or more. He didn't lay out any empirical estimates, e.g., ending this regulation or this government spending program will temporarily increase the growth rate of real GDP by x percentage points.

In an article that just scratched the surface of deregulation and government spending cuts, I showed how one could raise the growth rate of real economic wellbeing by about 3/10 of a percentage point per year for a few years. I did this in a challenge to Paul Krugman in the pages of Fortune in 1995. So John's 4 percent for, say, 5 years or so, is plausible.

My article is "The Case for Small Government," and it's in Fortune, June 26, 1995.

An excerpt:

When macroeconomists look at U.S. data on real (inflation-adjusted) GDP, they notice something interesting: The economy's growth rate seems pretty stable. Many then conclude mistakenly that economic policy doesn't matter much. This tendency is bipartisan. Robert Lucas, a libertarian/conservative economist at the University of Chicago and someone who is likely to win the Nobel Prize for economics within the next ten years, once said, "I think this economy is going to grow at 3% a year, no matter what happens. Forever." Exhibit A for liberal economists is Paul Krugman of Stanford. In these pages (Fortune, May 1), in a piece mainly devoted to defending the welfare state, Krugman wrote: "[T]he underlying growth rate of the U.S. economy has been a very stable 2.5% right through the past five Administrations."

Krugman asserted that any politician who claims he can raise the economy's growth rate "by as much as three-tenths of a percentage point is naive--or worse." Maybe, but that doesn't mean a politician can't add three-tenths of a percentage point to the growth rate of economic well-being.

I then went on to make some back-of-the-envelope calculations of the effects of cutting government spending:
Look at the numbers. Simply to have a benchmark that allows us to look at changes, assume that today's $7 trillion GDP, with its current makeup of government and private-sector production, is a measure of economic well-being. To raise the growth rate by one-tenth of a percentage point, you would have to increase economic well-being by $7 billion. Three-tenths of a percentage point, therefore, is $21 billion. Is a politician naive if he promises a $21 billion increase in economic well-being? Not at all. Privatizing the Postal Service alone could achieve that goal for one year. One more fat example along these lines: Say the Department of Defense spends $10 billion a year on a weapons system over a period of ten years. To get members of Congress to vote for that system, the Pentagon usually has to promise to have it produced in various key members' districts, even though that's hardly the most efficient use of taxpayers' money. Unconstrained by congressional pressures, let's say that the Pentagon might have been able to acquire the same quality of weapons system for $60 billion instead of $100 billion. Again, for the nation's GDP, it doesn't matter whether the government spends $100 billion inefficiently or spends just $60 billion and leaves $40 billion in taxpayers' hands. But for our economic well-being, it matters a lot. If the government procured weapons efficiently, Americans would be better off by $40 billion.
Of course, eliminating one program or streamlining procurement policies at one department would not permanently increase the U.S. growth rate. You can't privatize the Postal Service twice. But these examples just scratch the surface. The simple fact is that government has gotten so huge that just by eliminating a few programs a year you could increase the growth rate of economic well-being by three-tenths of a percentage point for at least five years.

Here's how. Assume conservatively that moving a function from the government to the private sector would lower its cost by one-third. Therefore, you would have to move only $63 billion a year in functions out of government to get to $21 billion in savings. That's only about 4% of the U.S. budget. If you did this every year for five years, you would cut the federal government's budget by about 20%.

On deregulating:
Moreover, privatizing government activities isn't the only way to cut back government and make everyone better off in the process. There's another way to achieve a higher growth rate in economic well-being: deregulate. At the height of the Interstate Commerce Commission's power, its budget was well below half a billion dollars a year. But Thomas G. Moore, a senior fellow at the Hoover Institution and a leading transportation economist, points out that the ICC's budget measures only a tiny fraction of the damage this one agency has done to the U.S. economy. By keeping rates high and restricting the items truckers could carry, the ICC caused a lot of trucks to go out half full and return empty. The ICC's so-called gateway restriction also meant that if a trucker had only two licenses, one to deliver from, say, Charlotte, North Carolina, to Indianapolis, and the second to deliver from Indianapolis to Memphis, the only way he could legally deliver from Charlotte to Memphis would be to drive to Indianapolis first, even if he had nothing to deliver there. This restriction wasted millions of gallons of fuel and thousands of man-years every year. The longer delivery times that the ICC spurred by restricting entry also induced businesses to hold much higher levels of inventory than would have been needed had transportation been cheaper.

Moore estimates that deregulation under Presidents Carter and Reagan increased shippers' economic well-being by about $60 billion, most of which was in the form of lower prices. As recently as two years ago, Moore predicted that ridding the nation of federal and state trucking regulations would save shippers as much as $20 billion a year. That is happening now. Last August, Congress eliminated almost all remaining interstate and intrastate regulation of the trucking industry, and President Clinton, the House of Representatives, and the U.S. Senate have all agreed that the ICC should be abolished.

Not surprisingly, when I wrote for Fortune, I sometimes had to make compromises. One big one I remember is in the last paragraph. The published version says "Bill Clinton is hardly Juan Peron." My submitted version said "Bill Clinton is hardly Juan Peron and Hillary is not quite Eva."

Comments and Sharing

COMMENTS (12 to date)
E. Harding writes:

It's obvious lots of policies (or lack of policies) can permanently lower the rate of growth (e.g., North Korea, Madagascar, Zimbabwe, Ookrayeena, Senegal, Comoros, Niger). So why can't some policies permanently raise the rate of growth? Growth is just getting to higher levels faster.

ThomasH writes:

I'm unsympathetic to the Smith-Cochran argument over whether a set of reforms exists that will raise the growth rate to 4% p.a. Let's just do the ones that raise it and it adds up to whatever it adds up to.

Two caveats:

Some desirable reforms might not raise GDP as measured. Maybe some of the benefit goes into consumer's surplus. And vice versa (a reg that allows a monopolist to discriminate more or less perfectly.)

Is it not possible that those reforms with the largest payoffs (airline regulations, the ethanol subsidy?) have already been done?

OK, three. :)

What if one of the reforms removes/introduces a measure that was partly offsetting the damage/benefit from another measure (fuel mileage standards offsetting the lack of a carbon tax?)

Let's just see what Jeb comes up with.

Kevin Erdmann writes:

The possibilities are huge with housing, which is hobbled by a constricted mortgage market that I don't see any candidates willing to loosen and constricted supply because of local regulations in the core big cities. Supply has been constricted for so long, more building could easily increase real growth by lowering inflation by nearly 1% a year for a couple of decades before we might be "overbuilt". This would also lead to growth because of higher residential investment activity. But, much of it would simply come from making valuable locations available for use. There are trillions of dollars worth of potential, high value residential space sitting in the air hundreds of feet above New York City and San Francisco, which would be worth $1 million for every $500,000 of material investment. The fact that we are sitting on all of that is the economic equivalent of sitting on vast fields of oil and choosing not to drill. Building that housing would create huge gains in real residential housing consumption with relatively small amounts of capital investment. It is not widely appreciated that real housing expenditures have been declining for decades as a proportion of PCE, even while nominal housing expenditures have remained level, because of these large city constrictions.

Sadly, in places like San Francisco, locals literally complain that prices are too high because there is too much "market rate" building. And there seems to be national unanimity about preventing mortgage markets from ever serving the median household again.

ThomasH writes:

@ E. Harding,

It is true statistically for really large differences in policy environments, but you might be surprised how poor the "fit" is between income and policy at least as measured by the World Bank "Doing Business" indicators. Scott Sumner made this point in relation to Greece: it's a lot richer than it Freedom House indicators would suggest.

That no reason to pass on doing what seems sensible, but caution about the magnitude of the benefits is in order.

E. Harding writes:


-Not Freedom House, Heritage Foundation. I'm not too surprised at the poor fit; I've noticed it before. I think a model that combined both human capital measurements and measures of institutional quality could be more powerful than one that only included one of these:

ThomasH writes:

I stand corrected. Thanks.

Memo to self: remember once check twice.

Andrew_FL writes:

@ThomasH-As long as you're checking, you misspelled Cochrane's name. It's probably a typo, and I wouldn't even bring it up normally, but there is a small chance you've confused John Cochrane the grump economist, with John Cochran, without an e, the Mises Institute Fellow.

Pajser writes:

David Henderson: "Here's how. Assume conservatively that moving a function from the government to the private sector would lower its cost by one-third. "

It doesn't seem obvious to me. What is the solid argument in favor of it? Bel et al. (2010) do not support it. Soviet GDP growth can be relevant also.

john hare writes:

I would suggest that 4% is incredibly conservative given the possibilities out there. Deregulating has a one time cost reduction of X, ignored in the piece is the year after year efficiency improvements of X/? to the rest of the country. The trucking reduced costs massively at the time of dereg, then the improved national efficiency extends for decades.

Same could be on housing mentioned by Kevin, an ongoing improvement that doesn't stop the second year after implementation. Commercial construction has similar problems as I learned to my personal cost.

How much permanent gain if labor was no longer discouraged from producing beyond certain points for fear of losing benefits? I'd say well beyond 4% on this issue alone. How about a national effort to quit lying to the non-skilled about their value to others.

How much if politicians no longer benefited from pork politics by some method? I'd bet on far more than 4% on that alone. Crony capitalism is a drag on real productivity. Honor the legitimately rich, and lower the boom on those that got there illegally and immorally.

Insurance, medical, and the litigation industries could change it by well over 4% each.

Work on educational effectiveness against cost for another major chunk.

I consider it possible to increase GDP by 10% annually for a decade or more. With effort, the list of possibilities could go on for pages and chapters.

ThomasH writes:


Thanks. It was not just a typo. I do not know anything about Mr Cochran.

[ThomasH: If it's not a typo, then readers are even more confused. Why are you referring to Mr. Cochran at all if you know nothing about him? The post is about Mr. Cochrane--with an "e" at the end--not Mr. Cochran--with no "e". The post is about a Smith-Cochrane disagreement, not a Smith-Cochran disagreement. It certainly looks like a typo when you've twice used the spelling "Cochran" without an "e" both in your comment in this thread and in your comment in the preceding thread! --Econlib Ed.]

David R. Henderson writes:

Thanks for pointing out ThomasH’s error. Sadly, John Cochran, the Austrian economist, died of cancer in May. Here’s an obit.

ThomasH writes:

I meant that my error was not just in typing, but in not knowing about the other gentleman.

Comments for this entry have been closed
Return to top