Just One Minute
Balanced Fare: We Report, You Deride

Friday, April 25, 2003

Krugman Explains The "Krug Math"

His Tuesday column collected Bronx cheers from folks who wondered how he managed to keep a straight face while comparing the "cost" of ten years of tax cuts against the gain of one year's average wages. Let's review - my criticism is here, Don Luskin takes a shot, and TAPPED finds a defender of the Earnest Prof.

[Mini-Update: Max Sawicky is a late arrival, but he is not convinced by either side.]

First, let's glimpse the controversial passage:

Still, let's pretend that the Bush administration really thinks that its $726 billion tax-cut plan will create 1.4 million jobs. At what price would those jobs be created?

...The average American worker earns only about $40,000 per year; why does the administration, even on its own estimates, need to offer $500,000 in tax cuts for each job created?

Just to review, the $500,000 figure comes from dividing roughly $700 billion of tax cuts over ten years by the 1.4 million new jobs.

So, as predicted by Luskin, Prof. Krugman offers an explanation over at his personal website. Compared to the exposure of his NY Times column, this should be called the Krugman web-out-of-sight, but regardless. Here is a taste of the rationale:

I've gotten a fair bit of mail over the way I compared the annual cost of employing an average worker with the 10–year cost of the latest Bush tax cut. Some of the mail was in good faith, so here's the explanation.

I have to pause here. How can I qualify for the "good faith" mail category? I will double the obsequieosity of my next e-mail! Back to the Prof:

No, I didn't forget to divide by 10. (For God's sake: whatever you think of my politics, I am a competent economist, and know how to use numbers.) What I foolishly assumed readers would know - this isn't condescension, I really was foolish - is that no serious economist thinks that a tax cut or spending increase will have any effect on employment more than a couple of years from now. The reason is straightforward: normally the economy is operating more or less at full employment, and any demand stimulus from a tax cut will be offset by an interest rate increase by the Fed. The Fed, of course, polices the economy to prevent inflationary pressures. And eventually we will return to normal circumstances.

The only situation in which a tax cut or spending increase creates jobs is when the economy is operating below full employment, and the Fed is unable to remedy the situation.

We are in such a situation right now - or at least I think we are. The Fed, by the way, does not agree: it thinks that a good recovery is just around the corner, and that it will soon be raising interest rates; in that situation any demand push from a tax cut will simply cause it to raise interest rates faster.

Well, for a lark I would love to dredge up the talk Prof. Krugman gave in about 1995 when he explained that no reasonable economist thought that unemployment could go below about 6%. Anyway, he seems to be saying here that, assuming any reasonable mix of fiscal and monetary policy, the economy will end up back on its full employment track in a few years. A tax cut now may speed up the process of job creation by a few years, but eventually the economy will get to the target employment path regardless.

That being the case, the Bush tax cut only creates "extra" jobs for a couple of years at most; therefore dividing by ten is not appropriate.

OK, I think I get it. Don Luskin has his rebuttal, and here is mine, and I am operating at maximum pith - if all responsible economists agree that the back end of the Bush tax cut has no effect on job creation, then why include it in the calculation? Presumably the later years of the tax cut have other motivations, but by Prof. Krugman's statement, job creation could not be one of them. In which case, if the Bush people could not think the later tax cuts are there to create jobs, and Prof. Krugman does not think they are there to create jobs, then why does he include them in the calculation to show the "cost" of creating jobs?

Now, I should note that I part company with Mr. Luskin on this - his argument is that the 1.4 million new jobs, having been created, are attributable to the tax cut over the full ten years. Prof. Krugman's point seems to be that the US economy will eventually stagger to equilibrium and create those jobs on its own.

So, I suppose that the Krugman critique is that, when Bush and the White House say that this tax cut will create jobs, they are being disingenuous, sort of - yes, it will create jobs faster than otherwise, but eventually those jobs will happen. And since the Bush folks are saying that the whole $700 billion plan is needed to create those jobs a year early, Bush should be charged full freight.

Of course, Bush has also said that other good things will happen because of the tax cut, as I mentioned in my first post. If the Administration chooses to emphasize a particular feature of the plan, that does not mean the other features do not exist. For example, expanding the child credit has been pitched as "pro-family". Well, fine, but this week the talking point is jobs.

OK, a mini-update paints me into a tighter corner, and thanks. From the WaPo, describing Bush's trip to Ohio to promote the plan:

[Bush] linked it to short-term job creation and economic stimulus, even as his aides discussed the possibility of postponing some of the plan's stimulative measures. Bush said job creation is the "whole purpose" of his plan.

What's next, you are going to read his lips? Normal political oversell, ignore him!

Now, an excerpt from the CEA report itself seems to confirm part of Prof. Krugman's point:

On average over end-2002 to end-2007, job creation as a result of the package would be 140,000 higher than otherwise. This indicates that the proposal would bring forward a good deal of the job creation that would otherwise have occurred in 2005 and beyond (and add some as well).

As noted above, the statistical model used for the projections does not include any supply-side effects under which lower tax rates would be expected to boost labor supply and further improve job creation. Corporate income tax relief would likewise be expected to lead to positive supply-side effects through improved allocation of capital across the economy and thus higher growth and job creation—again, however, this is not reflected in the numerical projections.

Emphasis added. Well, as I wriggle about, let me point out that the CEA seems to think that the tax code can be changed to improve the long term equilibrium levels of employment in the US. My guess as to Prof. Krugman's model is that the government influences the economy through fiscal, monetary, and regulatory policy. Over the long run, the first two are neutral, so it is regulatory policy that will affect the equlibrium level of full employment (and I am thinking of the European labor markets as I say this).

I suspect that most of us would agree that taxes can have both a fiscal and a regulatory impact. For example, instead of a Social Security payroll tax of 15% capped at $80,000, the Feds could attempt to raise the same revenue with a flat tax of $12,000 per employee. I don't think anyone would argue that, even in the long run, the effect on employment of the two tax schemes would be the same. So, not all taxes are fully employment-neutral.

That said, the CEA has not attempted to estimate these "regulatory relief" effects of the tax plan on employment, and the 1.4 million figure "new jobs" seems to be as Prof. Krugman described - something that the CEA thinks would happen eventually anyway. If we are meant to take seriously Bush's statement that the tax cut is all about jobs, and the CEA is formally estimating 1.4 million new jobs roughly one year earlier than otherwise, I suppose that, in this narrow sense, Prof. Krugman is rrrriiii.... Oh, I can't say it.

So dividing by ten never made sense. I try to learn something new every day, and this blog certainly helps in that process. And why is Prof. Krugman dividing by one? Modesty, I suppose - the CEA report suggests that eighteen months might be right, Prof. Krugman's own comments suggest that he is more pessimistic about the duration of the slump than others, but apparently he thinks someone out there believes that the economy will be back on track in a year.

I will close with two questions.

When an enthusiast for the Clinton Administration says things like "Rubinomics led to the creation of 20 million new jobs", does P. Krugman leap to his feet and explain that most, if not all, of those jobs were pre-ordained?

And in his website piece, the Earnest Professor tells us that "Nobody, and I mean nobody, who knows any economics thinks that the tax cut will have an effect worth mentioning on employment 5 years from now, let alone 10." OK, I think he means by that that the tax cut will not have a significant effect on employment after five years, or ten.

So in his "Yes, We Have No Bananas" column where he refinanced his home mortgage and predicted hyperinflation in the US, he also quoted the warning that "a fiscal crisis threatens our future standard of living". Isn't it fair to infer that employment might suffer in such a scenario? For example, I thought Argentina was an example of a "banana republic" experiencing an economic collapse. What is the threat to our standard of living posed by the Bush plan, and where is the consistency with his current "no effect" argument? Or is he subtly arguing that we will operate at a new equilibrium with the same full employment, but (much) higher interest rates and inflation, and for other reasons that is a bad thing? We await (with trepidation) the painless collapse.

UPDATE: I'm back, like an alcoholic. NO, Bush is the alcoholic, says Krugman.

It's O.K. to run a deficit during a recession, as long as the deficit is clearly temporary. But both the numbers and the administration's search for excuses tell us that there's nothing temporary about the red ink. On the contrary, we'll probably be on a deficit bender until the baby boomers retire — and then it will get much worse.

Trust me: we're going to miss Rubinomics. Maybe not today, and maybe not tomorrow, but soon, and for the rest of our lives.

Originally published in The New York Times, 1.17.03

Same question - if the only consequence of deficits is higher interest rates, and employment is not affected, what is the worry?

Comments: Post a Comment