The Corner

When Keynesians Attack

If you missed David Einhorn’s op-ed in The New York Times yesterday, take a moment to get caught up. I won’t summarize it here, because you really should read the whole thing.

Economist Dean Baker, who thought the biggest problem with the $800 billion stimulus bill was that it should have been much larger, has written an intemperate response to Einhorn that resorts to name-calling (Einhorn’s concerns about America’s debt problem amount to “looniness”) and accuses Einhorn of being either clueless about inflation statistics or deliberately misleading.

Knowing that Einhorn has a day job and no ready platform from which to reply, I e-mailed him a link to Baker’s post last night and asked for his thoughts. Here is his response:

The NY Times piece was space constrained. I elaborated on my speech at the Ira Sohn Conference. [The entire speech can be found here.] The section on inflation says:

Now, government statistics are about the last place one should look to find inflation, as they are designed to not show much. Over the last 35 years the government has changed the way it calculates inflation several times. For example, under the current method, when the price of chocolate bars goes up, the government assumes people substitute peanut bars. So chocolate gets a lower weighting in the index when its price rises. Even though some of the changes may be justifiable, the overall effect has been a dramatic reduction in calculated inflation. According to, using the pre-1980 method CPI would be over 9%, today compared to about 2% in the official statistics. While the truth probably lies somewhere in the middle, this doesn’t even take into account inflation we ignore by using a basket of goods that does not match the real world cost of living.

For example, we all now know that healthcare, which is certainly a consumer good, is about one-sixth of our economy and its cost has been growing at a rapid pace. So what is the weighting of healthcare in the CPI? About 6%. The government doesn’t count the part which the consumer doesn’t pay out of pocket. So, if your employer has to pay more for your health insurance, it doesn’t count, even if it means you have to accept lower wages. Similarly, Medicare cost increases don’t count, even though everyone has to pay higher taxes to fund them. Income and payroll taxes, which are part of the cost of living, are not counted in the CPI either.

On the other hand, one-fourth of the index is comprised of something called owners’-equivalent-rent. This isn’t something that anyone actually pays for. If you own your house, the government assumes you are foregoing rental income. The amount that you could receive from a hypothetical renter – the government implicitly assumes you rent it to yourself – is counted in the basket. So, rising taxes, which you do pay, don’t count; the fast rising cost of healthcare, which someone else pays on your behalf, doesn’t count; but hypothetical rents which you don’t pay, and conveniently don’t rise very quickly, have a huge weighting.

The simple fact is that if your goal is to never see inflation, you won’t see it until it is rampant.

I believe that every time the government has adjusted the inflation calculation, the result has been lower reported inflation. It doesn’t seem like a coincidence to me. While I am not going to argue each adjustment (and as I said, some of them may be justifiable), if we used housing prices rather than owners’-equivalent-rent, inflation would have been much higher during the housing bubble. Greenspan probably would not have been able to hold rates at 1% for so long after the emergency passed in the face of higher headline inflation. We probably would have cut the bubble off sooner so there would have been less damage.

There are those who don’t believe deficits matter at all, since we can print money. Printing a little money will only have a little effect. If you print too much, at some point, you will have a problem. If deficits don’t matter at all, why should we bother to tax anyone? Of course, it is because deficits do matter. As has happened many times in history and most recently with Greece, at some point deficits matter a lot.

Finally, as I understand it (correct me if I am wrong), when the government calculates various statistics like GDP, first it calculates nominal GDP. From that it subtracts an estimate of inflation called “the deflator” or something like that, which it then subtracts to get to real GDP. So, if inflation is estimated to be higher, real GDP comes out lower. I believe it is the same with the other statistics I cite.

My two cents: Baker’s argument — that Einhorn’s criticisms of the way the government measures inflation are irrelevant, because other, more comprehensive measures of inflation do not indicate we are currently seeing a lot of it — is an exercise in aggressively missing the point. Einhorn never said that we are currently seeing rampant inflation. He said that our warning systems are flawed, and that other flaws in our system — our structural deficits; our reliance on pro-cyclical securities ratings; our ability to avoid default by printing money — have created the conditions for rampant inflation to manifest itself very quickly in the event that foreign creditors no longer wish to finance our government. Meanwhile, Congress is getting ready to pass a new stimulus package without coming up with any way to pay for over $80 billion of it. That’s the kind of irresponsibility Baker is providing intellectual cover for.


The Latest