Just One Minute has some comments about Paul Krugman's latest offering in the New York Times.

Is it unfair to ask whether the Earnest Prof now foresees a full recovery? Does he think that investors generally foresee such a recovery? Does he have any guidance for us on whether, or if, normal conditions of supply and demand might return to the bond market?

As usual with Professor Krugman's economic columns, they make me yearn for his comparatively sensible ravings on politics, and vice-versa. Here's what Krugman said:

So you can't claim that interest rates will be far below historical levels because inflation is gone. And on the other side, we need to think about the impact of budget deficits.

That last sentence will send the deficit apologists to battle stations (sorry, I can't avoid politics completely). For many years, advocates of tax cuts have insisted that the normal laws of supply and demand don't apply to the bond market, and that government borrowing--unlike borrowing by families or businesses--doesn't affect interest rates. But there's no argument among serious, nonideological economists. For example, a textbook by Gregory Mankiw, now the president's chief economist, declares--in italics--that "when the government reduces national saving by running a budget deficit, the interest rate rises."

Of course, classical supply and demand theory would indicate that Krugman is right. But we've had a lot of real world experience with interest rates and budget deficits and the theory needs a little revising.

To demonstrate, let's look at the budget deficit when Clinton took office. For the year 1993, the on-budget deficit was about $300 billion. In 2000, seven years later, the on-budget surplus was $86 billion. The amounts for both years come from here.

Now let's take a look at the interest rates for those years. Krugman focuses on the 10-year treasury, which is a good interest rate to use, since it is the basis for many other rates (especially home mortgages). Since the fiscal year runs from 10/1 of one year to 9/30 of the next, I will look at the interest rate halfway through that year--i.e., March 1st (or the nearest date after if March 1st falls on a weekend.

The 10-year treasury as of 3/1/93 was at 5.98%. The 10-year treasury as of 3/1/00 was at 6.26% Both interest rates were obtained from here. So we can see that despite a dramatic change in the budget status--from a $300 billion deficit to an $86 billion surplus, interest rates did not decline, in fact they went up, the exact opposite of what Krugman and Mankiw would predict.

Okay, so going back to the same sites, I checked the deficits and interest rates for every year from 1993-2001. Then I correlated the two using Excel.

Correlation is a measure of how much two different variables appear to be related. For example, suppose we were to take everybody's in a office's height and weight, and correlate the two. We would expect there to be some correlation; that the tall people would weigh more (on average) than the short people. The correlation would not be perfect (there are, after all, short fat people and tall skinny people) but in general tall people do weigh more than short people. This would result in a positive correlation--that on average the taller you were, the more you weighed and the shorter you were, the less you weighed.

We could also look for negative correlations. Let's say we asked 100 people chosen at random how many years of school they attended and how many hours a week they spend working at manual labor. We would expect that the more years of schooling one had, the fewer hours a week they would spend on manual labor, and the fewer years of schooling they had the more hours a week they spend on manual labor. Again, the correlation is not perfect because there are folks with masters' degrees working as auto mechanics and high school dropouts running software companies, but in general the more years of schooling, the fewer hours a week spent doing heavy lifting.

Correlations can run from -1.00 (perfect negative correlation) to +1.00 (perfect positive correlation). A correlation of 0.0 simply means that the two variables have nothing to do with each other. For example, we might expect that correlating years of schooling with weight would result in a correlation close to 0.0 nowadays. Back when most college graduates were men that might have been different, since men weigh more than women.

Okay, so now that we have correlation defined, what is the relationship between budget deficits and interest rates? It's (-.48). That is, there is some correlation between budget deficits and interest rates, but it is not strong and in general it's negative--that higher budget deficits are frequently associated with lower interest rates, and vice-versa.

Now, it is important not to confuse correlation with causation. The classic example of this is if you correlate the number of churches in a town with the number of crimes in a town, you will come up with a fairly strong, and positive correlation. This does not mean that building more churches will result in more crime. Rather, both variables are being influenced by a third variable, in this case, almost certainly, the population of the town. More people mean more churches and more crime.

What is the third variable in our budget deficit and interest rate scenario? As Bill Clinton once said, it's the economy, stupid! A strong economy leads to lower budget deficits, but also to higher interest rates. A weak economy leads to higher budget deficits and lower interest rates.