Monday, December 17, 2012

Lack of progress in Macroeconomics

In today's Financial Times, an article by Jeffrey Sachs presents all kinds of arguments to conclude that Keynesian economists have been wrong during the crisis, that fiscal and monetary policy tools have not worked and that the US and other advanced economies need a different kind of policy (education, infrastructure,...).

The issues raised are complex and there will always be a debate on the mix of economic policies that are right to help growth in a given circumstance. But the article does two things which are difficult to justify: first it reaches conclusions about what we have learned without citing any proper empirical evidence (the "it is so obvious" syndrome). And, second, it presents arguments which are inconsistent. Let me take on both of these issues separately.

The article argues that Keynesian policies have not helped and cannot help in the current economic environment. This has been and still is a debate where one needs to go to the data and come up with some clever way to produce a proper test. Jeffrey Sachs knows this and opens up his argument by saying:

"We can’t know how successful (or otherwise) these policies have been because of the lack of convincing counterfactuals." 

Yes, empirical tests in macroeconomics are challenging. But then he says


"We should have serious doubts. The promised jobs recovery has not arrived. Growth has remained sluggish. The US debt-GDP ratio has almost doubled from about 36 per cent in 2007 to 72 per cent this year."

Is that a test? Is this the counterfactual that was so difficult to build? That was easy!

And if you want to know what is the cause of low economic growth in Europe, it is not austerity:

"The crisis in southern Europe is often claimed by Keynesians to be the consequence of fiscal austerity, yet its primary cause is the countries’ and eurozone’s unresolved banking crises."

I wonder what type of counterfactual he built to reach this conclusion.

And as it is also obvious that austerity is not responsible for the UK slowdown, it is all due to

"the eurozone crisis, declining North Sea oil and the inevitable contraction of the banking sector"

That was another quick conclusion to reach.

Let me move now to the second point I wanted to make, one of inconsistencies in the logic used. The article criticizes monetary policy for ignoring the risk that current low interest rates might be creating another bubble. Here is what I find interesting: the only mainstream models that I know where monetary policy can affect real interest rates are Keynesian models (the ones he is criticizing). It is only in models where prices are sticky that you have the central bank, who sets the nominal interest rate, having an effect on the real interest rate and economic activity. But even in these models, this effect is transitory. As soon as prices adjust, real interest rate go back to equilibrium, out of control of the central bank.

The narrative that suggests that the Federal Reserve is responsible for low interest rates since 2003 is not supported by any economic model I know. The drift in real interest rates that we have witnessed for more than 10 years has to be more fundamental than a switch to expansionary monetary policy.
















And nothing more fundamental than supply and demand which in this market is represented by world saving and investment. It is well documented that there has been a significant shift in global saving during the last decade. One that justifies the global imbalances that we have witnessed and also the fact that global interest rates have decreased. And after 2008 a global crisis that resulted in yet another increase in the supply (saving) and a decrease in demand (investment) has brought these rates to even lower levels. Central banks are not responsible for this trend. And if you do not believe in Keynesian models they cannot be even responsible for the small movements around the trend.

Economists will never agree on everything and the excuse we (economists) have is that it is impossible to build empirical tests that definitely prove or disprove alternative theories. But we can do better than this article. We can at least build on a set of knowledge that is commonly accepted rather than reopening the debate starting from zero whenever we want to make a clever or controversial point.

Antonio Fatás


on December 17, 2012 |   Edit