Credit Efficiency

In a post titled Debt And Transfiguration, Paul Krugman argued against the view that a high level of debt causes problems in the economy. Inadequate growth, he said, makes debt appear excessive. He thinks the people who worry about the debt have cause and consequence reversed.

Then (ha ha) he added: "I’m not denying that high debt can be a problem."

Specifically, Krugman was arguing against the view that "bad things happen when debt goes above 90 percent of GDP." As evidence he said that in both Japan and Europe, debt increased after growth slowed. And in the U.S. he said, "the only period when debt was over 90 percent of GDP was in the early postwar years, when real GDP was falling."

Now I don't know whether Krugman was referring simply to the denominator problem -- that the value of the debt/GDP ratio can rise because of lagging GDP -- or whether he has a meatier argument. But to the people who already don't agree with Krugman, that couldn't possibly make the least bit of difference.

People who already disagree with Krugman -- those, let's say, who worry what will happen when debt goes above 90% of GDP -- will be scratching their heads, wondering why he is talking about World War II debt. The problem is not some old debt from 65 years ago that everybody knows wasn't a problem. The problem is the current high level of debt, and current projections of even higher debt to come. After all, not even Krugman denies that high debt can be a problem.

Still, he has a point: Debt was over 90% of U.S. GDP once, and it turned out not to be a problem, that time.

For the 90-percenters, Krugman creates a dilemma: Why was the World War II debt not a problem? They say debt is a problem now. They know they are right. I know they are right. I think even Krugman knows they are right. But he raises a question they cannot answer. As things stand now, the only solution is for 90-percenters to reject Krugman more completely. Dilemma and all.

Krugman suffers the same dilemma. He knows that high debt can be a problem. And yes, he knows that it turned out once, at least, not to be a problem. But he doesn't know why. He doesn't know why high debt was not a problem after World War II. If he knew, surely he would tell us.

The dispute between Krugman and the 90-percenters is fought on a particular playing field. That playing field is the graph "debt relative to GDP." When the trendline reaches 90% it is a huge problem. Or it is no problem at all.

To settle that dispute, answer one simple question: What does the graph show?

Consider first what the graph does not show: It does not show that debt is high. It only shows that debt is high relative to GDP: Gee, this mountain looks big compared to that mole-hill. Yes, we have a mountain of debt. But the graph does not show that. It shows a thing we already knew: The graph shows that GDP growth has been laggard. But that was our original complaint. We don't have the growth of output, the growth of jobs, the growth of living standards, the growth of profit that we need, that we expect, that we want, that we demand. GDP is a mole hill.

What does the graph show?

The graph shows the efficiency of credit use. If an increase in debt is efficient, it produces a lot of growth. So GDP increases more than the debt, and the trendline goes down. But if an increase in debt is not efficient, it produces little growth. So then GDP increases less than the debt, and the trendline goes up. The trendline rises or falls depending on the efficiency of credit-use as measured by improved growth.

There are discussions you can find on the internet, regarding the size of the "multiplier" -- the size of the effect we get per dollar of economic stimulus spending. Formerly, the multiplier was thought to be quite large. Now it is often said to be quite small. If you come across such a discussion, it may be presented as an argument: someone is right; someone is wrong.

Here's what I think: In the days when credit-efficiency was high, the multiplier was high. In the days when credit-efficiency is low, the multiplier is low. It is not a matter of someone is right and someone is wrong. It is a matter of credit efficiency.

What does the graph show?

It shows a steep decline from the 1946 peak to about 1952. We got a lot of growth for every dollar of credit use. Credit efficiency was extremely high.

It shows a less steep decline from 1952 to 1966, the year I graduated high school. Credit efficiency was somewhat less.

It shows a mild decline from 1966 to 1974. We gained only a little growth for every dollar of credit use. Credit efficiency was quite low.

It shows flat from 1974 to 1981. At this point there was no gain from credit use. (The multiplier was about 1.) We had reached a Laffer limit.

And it shows a sharp increase after 1981. There was a loss from credit use. If growth was better in the 1980s than in the '70s, this trendline shows why: It took a lot of debt to boost economic growth. Credit use was inefficient.

After the mid-1990s the trendline loses its "bowl" shape. We'll look into that later. Meanwhile I should point out that the trendline, which disappears behind the graph border in 2009, approaches 90% in that year: last year. I know this sharp spike in the trend is part of a policy designed to boost economic growth. But low credit efficiency makes that policy ineffective.