Why Doesn’t Keynesian Stimulus Work?

In an op-ed in the Wall Street Journal award-winning economist Edmund Phelps argues against Keynesian stimulus:

Among economists and policy makers it is widely thought that fiscal stimulus—increased public spending as well as tax cuts—helped pull employment from its depths in 2010 or so back to normal in 2017. The new tax cuts on personal income are thought to be increasing demand further.

But is there evidence that stimulus was behind America’s recovery—or, for that matter, the recoveries in Germany, Switzerland, Sweden, Britain and Ireland? And is there evidence that the absence of stimulus—a tight rein on public spending known as “fiscal austerity”—is to blame for the lack of a full recovery in Portugal, Italy, France and Spain?

A simple test occurred to me: The stimulus story suggests that, in the years after they hit bottom, the countries that adopted relatively large fiscal deficits—measured by the average increase in public debt from 2011-17 as a percentage of gross domestic product—would have a relatively speedy recovery to show for it. Did they?

As the accompanying chart shows, the evidence does not support the stimulus story. Big deficits did not speed up recoveries. In fact, the relationship is negative, suggesting fiscal profligacy led to contraction and fiscal responsibility would have been better.

Isn’t this a fluke? Don’t history and theory overwhelmingly support stimulus? Well, no. First, the history: Soldiers returning from World War II expanded the civilian labor force from 53.9 million in 1945 to 60.2 million in 1947, leading many economists to fear an unemployment crisis. Keynesians—Leon Keyserling for one—said running a peacetime fiscal deficit was needed to keep unemployment from rising. Yet as the government under President Harry S. Truman ran fiscal surpluses, the unemployment rate went down (from 3.9% in 1946 to 3.1% in 1952) and the labor-force participation rate went up (from 57.2% to 58.9%).

Standard economic theory points to other ill-effects of fiscal stimulus. Economist Franco Modigliani showed in 1961 that fiscal deficits have a negative impact on both the supply of labor and the supply of saving. I explored this territory further in a 1965 book on “fiscal neutrality.” The deficit may cause people to feel richer, and thus to work and save less. The Keynesian James Tobin showed in 1955 how consumer demand can simply crowd out investment demand. Another possibility, suggested by Robert Mundell’s early work, is that the effect of a country’s fiscal stimulus is diffused over the global economy, so that the deficit-running country itself feels little of the effect.

If fiscal stimulus is not effective in combating a recession, what about monetary stimulus—increasing the supply of money or reducing the cost of money in relation to the return on capital? We can perform a similar test: Did countries where monetary stimulus in the years after they hit bottom was relatively strong—measured by the average quantity of monetary assets purchased by the central bank from 2011-17—have relatively speedy recoveries? This is a complicated question, but preliminary explorations do not give strong support to that thesis either.

I don’t know if “most economists” believe it but my account of Keynesian stimulus (counter-cyclical deficit spending to offset a shortfall in aggregate demand) would be somewhat different. I think that Keynes was right on the theory but it’s very difficult to make work in practice. In other words I believe that a properly timed stimulus carefully applied can reduce the depth of a recession. The problem is not that it can’t but in practice stimulus is almost always applied too late and rarely applied prudently.

In the specific case of the ARRA, I think that the stimulus provided by the Bush Administration, although probably inadequate, was responsible for reducing the depth of the recession of 2008-2009 and that the stimulus of 2009 was mostly misapplied—plain old pork barrel politics rebranded as fiscal stimulus—and might even have contributed to the shallowness of the recovery.