Mises Daily

Home | Mises Library | Filling the Holes in Krugman's Analysis

Filling the Holes in Krugman's Analysis

January 15, 2009

Tags Big GovernmentBiographiesThe FedOther Schools of Thought

Although many free-market economists were aghast that Paul Krugman won the Nobel (Memorial) Prize in Economics, I have come to realize that he is every bit as brilliant as that august award indicates. For some time now, Krugman has said we are in "depression economics" mode, where the normal rules of scarcity and tradeoffs don't apply. In this universe, it makes sense to have one group of workers dig holes, and another group fill them back up. Sure, when all is said and done, there is nothing tangible to show for this effort, but at least it "creates jobs."

So what I've come to realize is that in these last few months Krugman has implemented his own private-sector stimulus plan. He has been working furiously, cranking out fallacious articles and blog posts, which then provide work for people like Bill Anderson and me, as well as thousands of other bloggers who still can't understand why it's bad for families to save more. A clever chap, this Dr. Krugman, no?

Today my make-work will fill in two holes in a recent Krugman blog post. The first flaw is his belief that output generates employment (rather than vice versa), and the second is his belief that government spending is a measure of real output.

Krugman Thinking Backwards

In his post, Krugman goes through some "stimulus arithmetic" to see how much spending the incoming Obama administration needs to avert a serious recession:

The starting point for this discussion is Okun's Law, the relationship between changes in real GDP and changes in the unemployment rate. Estimates of the Okun's Law coefficient range from 2 to 3. I'll use 2, which is an optimistic estimate for current purposes: it says that you have to raise real GDP by 2 percent from what it would otherwise have been to reduce the unemployment rate 1 percentage point from what it would otherwise have been. Since GDP is roughly $15 trillion, this means that you have to raise GDP by $300 billion per year to reduce unemployment by 1 percentage point.

We already see the problem. Regardless of whatever correlations Okun may have found, it is quite obvious that to increase real output — to crank out more units of goods and services — you must first get more people working to create the products. In other words, higher real GDP is associated with lower unemployment, because more people are working and thus producing more output.

But because Krugman ignores supply and insists on viewing everything through the prism of aggregate demand, he thinks that government spending gives businesses the incentive to hire workers. That is how he manages to reverse cause and effect and think that changes in output ("real GDP") drive changes in unemployment. By making sheer "spending" the primary consideration — rather than focusing on the important economic problem of channeling scarce resources into those lines most desired by consumers — Krugman worries about "multipliers" and how tax cuts aren't nearly as potent in boosting GDP as government spending:

Now, what we're hearing about the Obama plan is that it calls for $775 billion over two years, with $300 billion in tax cuts and the rest in spending. Call that $150 billion per year in tax cuts, $240 billion each year in spending.

How much do tax cuts and spending raise GDP? The widely cited estimates of Mark Zandi of Economy.com indicate a multiplier of around 1.5 for spending, with widely varying estimates for tax cuts. Payroll tax cuts, which make up about half the Obama proposal, are pretty good, with a multiplier of 1.29; business tax cuts, which make up the rest, are much less effective.…

Let's be generous and assume that the overall multiplier on tax cuts is 1. Then the per-year effect of the plan on GDP is 150 x 1 + 240 x 1.5 = $510 billion. Since it takes $300 billion to reduce the unemployment rate by 1 percentage point, this is shaving 1.7 points off what unemployment would otherwise have been.

I confess that I have not delved into this "multiplier" literature, but I am very skeptical that the models being estimated have adequately dealt with the causality problem mentioned above.

In any event, we don't need to rely on sophisticated econometric critiques. Suppose the government borrows an extra $750 billion and spends it on "infrastructure." Assume for the sake of argument that the loss of this money from private capital markets, as well as households' increased savings (because they fear higher future tax burdens), does not totally cancel the increased government spending. In other words, suppose that total "spending" really does go up because of the government's move. Unemployment really does go down.

Even so, does this mean the plan was a success? Hardly. This is because aggregate measures of gross output are bogus when they include government expenditures. Even though purists can (rightfully) point out all sorts of methodological problems, it sort of makes sense to add up how much a family spends on rent, dining out, clothing, and other items, in order to measure the family's total "consumption" over the year.

The justification for adding up expenditures to get a gauge of total production is that a household will only spend its limited dollars on things that are more valuable than the dollar bills needed to purchase the items. So if a man spends $10,000 on a car, it sort of makes sense (though not really) to say he is consuming ten times as much as someone who buys a $1,000 computer. Correspondingly, we can justifiably claim that the car manufacturer and computer company together must have produced $11,000 worth of goods.

Government Expenditures Don't Measure Real Output

But what happens when the government spends $10 million on refurbishing an airport? Does that mean the community in question is now that much richer, as if a thousand new cars had appeared from heaven? Of course not. Politicians and bureaucrats don't have the incentive to economize on their spending, to make sure every dollar counts. On the contrary, President-elect Obama has specifically said that state and local governments need to "use it or lose it" when it comes to possible stimulus aid. They are going to throw those billions at anything that moves.

In the final analysis, even if a government "stimulus" plan managed to bring down unemployment, it wouldn't represent a true economic recovery. To the extent that the newly hired laborers — such as the construction workers fixing up the airports — seem genuinely richer, it would only be at the expense of taxpayers, whose share of the federal debt had increased.

Although a bit crude, it may clarify things to imagine three groups in the economy:

  1. the taxpayers
  2. the unemployed
  3. the capitalists

What Krugman wants to do is have the capitalists give (say) $750 billion of their wealth right now to the unemployed. The unemployed are definitely fond of this plan. But why would the capitalists do it? It seems as if they're out $750 billion, doesn't it?

Ah, the capitalists go along because the government promises that over the coming years, it will force the taxpayers to fork over the equivalent of (say) $800 billion of their wealth to the capitalists.1 Say what you will about this game, but it has not made all three groups richer. Notice that the capitalists and the unemployed can refuse to go along with the stimulus plan. But the taxpayers have no choice; armed men ultimately throw them in jail if they refuse to play along.

In conclusion, Paul Krugman reverses cause and effect in his analysis, and he also fails to note the difference between private and government expenditures. But hey, at least he provided me with an hour's worth of employment.

  • 1. Note that this $800 billion figure is in present discounted terms. It is higher than the $750 billion that the capitalists lend out for two reasons: first, I assume that the capitalists reap a net subsidy from the existence of the US Treasury; second, the government itself sucks up resources in its own operations to effect the transfers.

Follow Mises Institute