Two economists say in new paper that TARP worked

By Ezra Klein
Sunday, August 1, 2010

TARP is horribly unpopular. The stimulus is pretty unpopular. But does that mean they were bad policies? Not according to a new paper by Mark Zandi, Moody's chief economist and a former adviser to both the McCain and Obama campaigns, and Alan Blinder, a Princeton economist who has served as vice chairman of the Federal Reserve's Board of Governors.

Zandi and Blinder offer the first comprehensive estimate of our full response to the crisis: Absent the financial rescue and the stimulus, "GDP in 2010 would be about 6 ½ percent lower, payroll employment would be less by some 8 ½ million jobs, and the nation would now be experiencing deflation." I spoke to both men last week about how they got to their estimates, why the economy is so bad if the response was so good, and where we go from here. What follows are edited excerpts from the two interviews.

Q. This paper is heavily based on the model Moody's uses for economic forecasting. But these models have come under some criticism: Some say that they're just abstract equations and that you can get whatever answer you want by tweaking the numbers. So what is this thing? Who uses it?

Mark Zandi: I developed the model almost 20 years ago. I'm an economic consultant. I've got clients in many large, private-sector institutions. And we provide macroeconomic forecasts for them. The model has been used by banks to stress test themselves. Bank of America, J.P. Morgan, SunTrust and others used it to run scenarios on their solvency under different conditions. The business community uses it to figure out how to account for macroeconomic changes in their own budgeting and planning.

And how is it built? What are the numbers based on?

Zandi: I've gone back to every recession and depression and looked at the policy efforts to address the downturn, and try to at least capture the different ways in which policymakers have tried to generate a recovery. And what we've done in the Great Recession, some of it is unique, but most of it has been done many times before. Tax cuts, emergency unemployment benefits, aid to state government: These are things we've done every single time.

Alan Blinder: Different models do give different answers. That's why we say we welcome others to try and estimate this. But you can't make anything come out that you want. These models are fitted to real data. They're not just made up. They describe how the U.S. economy worked in the past.

CONTINUED     1           >

© 2010 The Washington Post Company