Robert Samuelson’s Troubled TARP Arithmetic

We know that arithmetic is not the strong suit of the Washington Post and Robert Samuelson drives this point home again today with his discussion of the TARP . Samuelson tells us that TARP is now projected to cost just $19 billion and that the final cost may actually be lower. He also tells us that the alternative to TARP , bank nationalization would have been far more costly. And, he said that without TARP the unemployment rate “would be 11 percent or 14 percent; it certainly wouldn't be 8.9 percent.”

Okay, let's take these in turn. First, the idea that the TARP cost almost nothing is based on some very shoddy accounting. Samuelson apparently does not understand the idea of money carrying an opportunity cost.

Suppose the government lent me $1 trillion for 10 years at 1 percent annual interest. In the Robert Samuelson world, the government is earning a $100 billion profit on this investment ($10 billion a year for 10 years). Economists familiar with opportunity costs would instead see this as a huge loss to the government, since it is giving me an enormous loan at an interest rate that is several percentage points below the market rate.

We saw how this worked with the TARP when Warren Buffett reported earning twice the money on his investment in Goldman Sachs which was half of the size of the investment from Treasury. Buffett got the market rate of return on his investment, the difference was a subsidy from taxpayers to the shareholders and executives of Goldman. The same story was true with the other TARP loans, as well as the even larger amount of money lent through the Fed as well as the guarantees provided by the FDIC.

This gets back to the comparison with the option of nationalizing the bankrupt banks, which Samuelson asserts would have been far more costly. Each year, the large banks are pulling over $100 billion a year out of the economy in profits. They also pay their executives tens of billions of dollars each year. Let's say that this sum comes to around $150 billion a year in total or 1 percent of GDP.

This money would not be pulled out of the economy if the banks had been nationalized. This is money that would have been available for other purposes (e.g. it could have paid for higher wages for ordinary workers) rather than supporting the consumption of bank shareholders and executives. The way this would work practically is that the Fed could stimulate the economy more with lower interest rates (think of some future point when the economy is closer to full employment) allowing for workers' wages to raise, because we do not have $150 billion or so in consumption by these shareholders and executives.

If we take the discounted value of this sum over the next thirty years it would come to more than $3.5 trillion. This can be viewed as the cost of the TARP and related rescue programs compared with nationalization. (Samuelson tells us that nationalization would have been complicated, so was TARP . Life's tough.)

Finally, Samuelson tells us that without the TARP unemployment would be “11 percent or 14 percent: it certainly wouldn't be 8.9 percent.” This is incredibly bad logic. These numbers are based on a counter-factual in which the government and the Fed let the financial system collapse and then did nothing by way of response. These are undoubtedly reasonable projections of the unemployment rate under such circumstances, however that is not a plausible counter-factual.

If Samuelson paid attention to what he was writing he would note another possible response, bank nationalization. If the Fed had taken over the bankrupt banks and then flooded the system with money (as it did with the TARP and related Fed liquidity programs) then we would not have seen the rise in unemployment from these projections.

Samuelson's analysis would be comparable to noting that a particular fire hose was used to put out a school fire, saving dozens of children. Samuelson would then tell us that this fire hose saved dozens of children. While this would literally be true, if that particular fire hose did not exist, the firefighters would have extinguished the fire with the other one they had on the truck. In other words, the alternative was not that the children would die, the alternative was that they would use a different hose.

In the same vein, the alternative to TARP was not that we sit around with a collapsed banking system waiting for the economy to sort itself out on its own. The alternative was a different set of monetary actions to boost the economy. It is silly to tout this no-hose story as the counter-factual to TARP.