Lectures on Macroeconomics, No. 6
By Arnold Kling
The main theme of this lecture is economic policy and labor market adjustment. My conjecture is that in our post-industrial economy, conventional Keynesian policies do not operate as they do in an industrial economy.
Another issue is that the natural impetus of the political process is to impede adjustment. Therefore, in a post-industrial economy, the presumption that government ought to do something about unemployment can be dangerous. The government may be more likely to do something wrong than to do something right.Commenting on an earlier lecture, Matt Yglesias wrote a blog post entitled Fight Recession by Causing Inflation?
If I’m reading him right, that’s what Arnold Kling is calling for here. Not just a shift in emphasis away from inflation-fighting and toward expansion, but specifically an effort to create inflation to allow real wages to fall in lieu of mass layoffs. I’ll file that under “provocative.”
In the Great Depression in the United States, I am pretty confident that monetary inflation would have been a good idea. So, without any apologies, I officially call for inflationary monetary policy in 1930.
Speaking of the Great Depression, Amity Shlaes passes along a note from Lee Ohanian with some relevant facts.
In 1939, total hours worked per working age person – including the hours of those on government payrolls – were 21 percent below 1929 level….
Why was there so little recovery in employment? Because wages in many sectors of the economy were far too high. For example, manufacturing wages relative to trend were about 16 percent above trend in 1939.
First, note that Ohanian uses a measure of employment–hours worked per working-age person–that is quite similar to the one that I referred to in my last lecture. My thinking with that kind of measure is that it is better than the unemployment rate because it gets rid of labor force participation rate changes. Hours are a better indicator than total employment because employment is affected by “labor hoarding”–at the end of a boom businesses are slow to fire unnecessary workers.
Second, if Ohanian is correct that real wages were too high, and if inflation would have lowered real wages, then his data support the view that more inflation would have helped to alleviate the Great Depression.
Of course, the point that is important to Shlaes is that the employment measure calculated by Ohanian shows that the New Deal did not bring about anything close to a full recovery to the 1929 peak. I would argue that the U.S. economy has yet to fully recover from the Dotcom recession that began in 2000, based on a similar indicator of labor capacity utilization.
Notwithstanding my views on the Depression, I do not think that inflation is the universal solution to every adjustment problem in labor markets. Let us look at the most severe recessions since the 1930’s–the oil recession of 1974-75, the Volcker recession of 1980-1982, and the Dotcom recession.
With the oil recession, the economy needed to adjust to higher oil prices. Government impeded the adjustment by using rationing, including the infamous queues at gasoline stations, rather than prices. Otherwise, there was plenty of inflation in the late 1970’s, and I see little evidence that this helped with labor market adjustment.
What the late-1970’s inflation gave us was the appointment of Paul Volcker as Federal Reserve Chairman, and what ensued was his eponymous recession. The Fed raised interest rates, leading to cutbacks in the interest-sensitive sectors of the economy such as housing.
The subsequent recovery was not accompanied by inflation. Instead, the Volcker recession produced a twenty-year trend of disinflation. Interest rates came down only with a lag, so that ex post, real interest rates were high. However, part of the reason that rates came down slowly was that expectations about inflation only declined slowly, so that ex ante real rates were generally lower than they were ex post. (If the interest rate is 8 percent, and everyone expects 5 percent inflation, then the expected real interest rate is 3 percent. If inflation actually turns out to be 4 percent, then the realized real interest rate is 4 percent.)
I should point out that housing is sensitive to nominal rates as well as to real rates. That is because high rates on fixed-rate, amortizing mortgages produce payment shock to borrowers. Regardless of what is happening to inflation, it is harder to make the payments on a 10 percent mortgage than a 6 percent mortgage.
(Suppose that we equate the real interest rate for both borrowers at 4 percent. That is, the 10 percent mortgage is accompanied by 6 percent home price appreciation and income growth, and the 6 percent mortgage is accompanied by 2 percent appreciation and income growth. In that case, in the early years of the mortgage, the high-rate borrower will have a higher payment/income burden and accumulate more equity than a low-rate borrower.)
Because the Volcker recession was caused by high interest rates, the cure was simply to let interest rates drift back down, once the Fed Chairman was satisfied that we were on a disinflationary path. Otherwise, there was no real macroeconomic policy involved. As rates eased down, some people returned to work in the interest-sensitive sectors related to housing, business investment, and consumer durables. A drop in oil prices helped the rest of the economy.
During the long recovery (from the mid-1980’s to 2000, we had only one, relatively mild, recession, the “economy, stupid” of 1992), the required labor market adjustments were gradual. Over this period we continued to phase out (through retirement) cohorts of less-educated workers while bringing into the labor force cohorts of more-educated workers. Personal computers spread rapidly, at first causing Robert Solow to famously complain (in 1987) that they were “everywhere but in the productivity statistics.” However, by the latter part of the 1990’s, it was clear that productivity growth was rising, and many economists concluded that computers were at least partly responsible.
The Dotcom recession is the one that I think is most relevant to circumstances today. It was a post-industrial recession, by which I mean that the main driver was not an inventory correction among manufacturers of consumer durable goods. There were major job losses in the service sector, and these were not temporary layoffs.
I would argue that standard Keynesian remedies were tried. Government spending rose and taxes were cut. Interest rates were kept low, with short-term real rates below zero (meaning that inflation was higher than the interest rate).
With all of this stimulus, what do we have to show for it, other than a housing bubble? As the table in my previous lecture showed, the Dotcom recession seemed much worse than the other two major recessions, particularly in terms of its duration. Of course, we do not know what would have happened had fiscal and monetary policy been to provide a weaker stimulus, or none at all.
My tentative judgment is that the Keynesian remedies are less appropriate to the post-industrial economy. Fiscal and monetary policy may be good for kick-starting the durable goods sectors, but the Dotcom crash required more complex adjustments, involving a more educated labor force.
From 2002 through 2007, the economy fell short of providing the employment opportunities for highly-educated workers that the Internet bubble was able to generate. Meanwhile, we overheated the traditional housing and consumer durables sectors. Some of the demand for consumer durables was met by foreign producers, as our imports shot up. One could argue that the demand for housing also was met partly by foreign labor, in the form of illegal immigrants.
Another sector that overheated in this period was the financial sector. The mortgage origination, securities-trading, and risk-disguising industries ballooned. These excess financial services employees now need to find other lines of work. I cannot see how a Keynesian policy of creating inflation to reduce real wages can help with that.
When the market tells an industry to shrink, the natural response of those in that industry is to sound the alarm. “This would be a disaster,” they say. “If X is allowed to fail, then that will affect Y and Z as well.”
It is true that most important industries are connected to other industries. Can you name a major industry that is completely isolated, so that it could suffer a catastrophe without affecting any other industry? Perhaps there is such an industry, but offhand it is easier for me to think of examples of industries that have important interconnections than to come up with an example of an industry that is fire-walled off from the rest of the economy.
Listening to the plea not to let industry X fail is probably a bad idea. A dynamic economy is one in which human and physical capital are chasing new opportunities, not holding onto lost causes. Politically, on the other hand, I can see where a bailout is a winning policy. The threatened industry is organized and visible. The alternative uses of capital are diffuse and unseen.
The economy needs to adjust as market conditions change. Instead, the political process encourages denial. It is natural to seek to prop up declining firms, but the result is something like Japan’s lost decade(s), where “zombie banks” misallocated capital in a massive way.
We have created the expectation that government can ensure high rates of labor utilization. In an industrial economy, where there is a lot of low-skilled labor and economic activity is heavily concentrated in a few key sectors, inflationary policies may in fact have some effect in terms of bringing real wages into better balance, either across sectors or in the economy as a whole.
My guess, however, is that in a post-industrial economy, the necessary adjustments are too subtle and complex to be helped much by inflation-producing macroeconomic stimulus. The problem is not to bring about a general reduction in real wages, or even a reduction in real wages in a key sector. In today’s economy, an enormous variety of career changes, business start-ups, and business failures will be needed in order to bring labor markets into balance.
In theory, wise technocrats could help guide workers in declining industries to appropriate re-training and career development. In practice, technocrats are not that wise. But it is much worse than that. Instead of giving the technocrats the mission of making the adjustment process more efficient, politicians will give them the mission of delaying the adjustment process and resisting the signals coming from the market. Thus, the expectation that government should help could have an ironic effect: the more that the public asks government to relieve the distress in labor markets, the longer it may take for labor markets to adjust.
Previous lectures in this series:
2. Misconceptions about Labor Markets
3. Unemployment as an Adjustment Problem
4. Why Wage Cuts are Rarely
5. The Dotcom Recession was, by one indicator, rather severe