How to Tell If Fiscal Policy Works
Printer-friendly versionPDF version
a a
 
Type Size: Small
The Fiscal Times
January 7, 2014

There is a raging debate in the econo-blogosphere concerning the effectiveness of fiscal policy when the economy is in a deep recession. This question is important because monetary policy loses much of its effectiveness in severe downturns

Why does monetary policy lose its effectiveness in deep recessions? As the economy enters a recession, the Fed responds by lowering the interest rate to stimulate the economy. For mild recessions, that is generally enough to turn the economy around. But in severe recessions even if the Fed lowers its target interest rate to zero – the zero lower bound as it is known – it is generally not enough to bring about a robust recovery. More stimulus is needed. 

Related: The Safety Net Redistributes Income…That’s Why It Works 

The Fed can turn from traditional interest rate policy to non-traditional policies such as quantitative easing, but those polices do not have anywhere near the effectiveness of traditional interest rate policy and – as we’ve seen in the Great Recession – it is not enough to avoid a highly sluggish recovery. When monetary policy alone is not enough, can fiscal policy be used to provide the extra stimulus that is needed to propel the economy back toward full employment? 

In theory, the answer is yes. Modern macroeconomic models tell us that while fiscal policy is not a very effective policy tool during normal times, it is very effective at the zero lower bound. 

However, many observers have argued that the evidence suggests otherwise. Standard Keynesian analysis implies the sequester should have been a drag on the economy, they argue, but the economy did relatively well after the sequester was put into place. So it must be that fiscal policy does not have much impact on the macro economy. 

But this type of analysis cannot settle questions about the effectiveness of fiscal policy. To understand why, consider an example from economic history. 

Related:  Is there One Economic Model to Rule Them All? 

In economic history, there is a discipline known as cliometrics. Cliometricians use rigorous theoretical and statistical analysis to answer historical questions. One well known use of cliometrics is an attempt by Robert Fogel to determine the value of railroads to the US economy. How much worse off would we have been if railroads had never been invented? Many observers had attributed the growth of the US economy in the 1800s to the development of railroads, but would this hold up to empirical scrutiny? 

To answer this question, Fogel needed to use some sort of model to determine how the US would have developed in the absence of railroads.  How much more road development would have occurred, and how effective would it have been as a substitute for rail traffic? How much more extensive would our canal network – an alternative way of shipping heavy, bulky goods – have been? And so on. 

Surprisingly, he found that the railroads weren’t as important as many people believed. According to his estimates, in 1890 – the year he chose to analyze – the US economy would only have been 2.7 percent smaller if no railroads existed at all. 

What this illustrates is the importance of a baseline, knowledge of how the economy would have performed in the absence of a particular policy or technological development. It is not enough to simply examine the US economy after the sequester, say it did pretty well, and conclude that the sequester must not have mattered. How do we know that the economy wouldn’t have done even better, perhaps a lot better, had the sequester not occurred? 

Related: How Yellen Will Shape the Fed’s QE Exit Strategy 

Or, for another example, many people claim that since unemployment remained persistently high after the stimulus package was enacted, it must not have been effective. But how do we know that unemployment would not have been even higher without this fiscal policy action from Congress? 

People who do this type of ad hoc analysis have a baseline in mind, how else would they be able to say whether a policy worked or not? But it’s a baseline that is simply assumed rather than derived from a rigorous model (and the assumed baseline is often very convenient for the argument that is being made). 

The effectiveness of policy can only be measured by using a model to estimate what would have happened had the policy not been enacted, a baseline, and then comparing it to the actual economic outcome with the policy in place. We can argue about how good the model is, and what ought to be in it. But the analysis of the effectiveness of policy must rely upon a comparison of an actual outcome to a hypothetical baseline alternative, and it’s best if the assumptions underlying the baseline estimates are out in the open – hence the need for models and transparent assumptions. 

Many people opposed to fiscal policy on ideological grounds argue that since the economy didn’t crash after the sequester, and since unemployment remained relatively high after the stimulus package, it must be that fiscal policy doesn’t work. But don’t be fooled. That type of ad hoc analysis cannot settle the question. When the analysis is done correctly – when a baseline is estimated using modern macroeconomic models and compared to the actual outcome – the evidence on the effectiveness of fiscal policy in deep recessions is quite favorable.

Top Reads from The Fiscal Times:

 

University of Oregon macroeconomist Mark Thoma writes primarily about monetary policy its effect on the economy. He has also worked on political business cycle models. Thoma blogs daily at Economist’s View.