Increased Productivity, OH NO!

David Beckworth asks whether the current high level of unemployment is cyclical, structural, or both.

My view is that it is both. The speculative bubble in housing resulted in substantial malinvestment. Resources, including labor, need to be shifted out of construction to the production other goods and services. Even if aggregate expenditures had remained on its 5 percent growth path, the redeployment of resources would have involved significant unemployment. This would have temporarily reduced the growth rate of potential income--perhaps even to the point it shrank, resulting in temporarily higher inflation. Eventually, as labor was reabsorbed and appropriate capital goods produced, real output would grow more rapidly and the inflation rate again fall.

However, overlaid upon this needed redeployment of resources was an unnecessary and undesirable shortage of money. While the quantity of money increased, it failed to keep up with the remarkable increase in the demand to hold money. The resulting monetary disequilibrium led to a drop in demand more or less across the board. Most firms faced falling sales and responded by reducing production. Layoffs spiked. New hires dropped off dramatically. Job openings fell dramatically.

Beckworth also agrees with the "both" answer. He focuses on the growth rate of demand, which he measures by the growth rate of final sales to domestic purchasers. This is the spending of U.S. residents on goods and services produced in the U.S. or abroad. He reproduces a chart that shows how the growth rate dropped sharply during the recession and became very negative. He notes that it is now doing better. The chart ends with a 2.5 percent growth rate.

I think the relevant criterion is gap between demand and its trend growth path and the appropriate measure is final sales of domestic product. Final sales of domestic product is spending by U.S. residents and foreigners on final goods and services produced in the U.S. It is currently 11.3 percent below its 5 percent growth path from the Great Moderation. While the gap continues to grow, fortunately, final sales of domestic product is growing, about 2.5 percent.

Rather than these long standing differences, a more interesting difference is in the source of the underlying "structural" unemployment that continues to exist, more or less hidden by the shortfall in demand. Beckworth's post focused not on a need to redeploy resources, but rather on increases in productivity. Supposedly, improved productivity has led to structural unemployment. He quotes Catherine Rampell:

Pruning relatively less-efficient employees like clerks and travel agents, whose work can be done more cheaply by computers or workers abroad, makes American businesses more efficient. Year over year, productivity growth was at its highest level in over 50 years last quarter, pushing corporate profits to record highs and helping the economy grow...

I find this sort of argument misleading.

Improvements in labor productivity allow for a given level of output to be produced with less labor. If the amount to be produced is assumed to be given, then, improved labor productivity implies a decrease in the amount of labor needed. Enhanced productivity appears to be a "problem" that is interfering with the ability of the economy to fulfill it's core purpose--create jobs. NOT!

In other words, the argument appears inconsistent with the key principle of economics--scarcity.

Output isn't given. As Beckworth, and the quoted articles note, output is increasing. Perhaps it is natural to ask why the actual increase in output has been associated with such a small increase in employment. And surely, improved labor productivity is a likely part of the answer.

However, the "answer" is never that the improved productivity has resulted in a given increase in output creating too few jobs. The increase in output isn't given. If labor productivity has increased, then output should increase more than it would have if labor productivity had not increased.

Increased labor productivity (and increased total factor productivity,) increases real output and real income. It doesn't reduce the demand for labor or other resources.

With a regime of total expenditure targeting, the effect of more rapid improvements in labor productivity (or total factor productivity) is lower inflation. The growth path of money incomes is unchanged, but the more rapid growth in productivity results in lower unit costs. As competitive firms respond by lowering prices to increase sales, the real volume of expenditures rises, allowing growing purchases to match the more rapid growth in the productive capacity of the economy. Output expands enough so that all of the available resources are utilized. Employment does not decrease.

With inflation targeting, more rapid productivity growth requires more rapid growth in demand. Nominal incomes grow more quickly, money expenditures grow more quickly, and firms sell more at an unchanged growth path for prices. Output expands so that all of the available resources are utilized. Employment does not decrease.

Of course, improvements in labor productivity or total factor productivity are unlikely distributed proportionately across all types of labor, much less all productive resources. This implies that if there is an improvement in labor productivity in one field of employment, the reason why output may not increase enough to leave everyone employed is bottlenecks of complementary resources whose productivity has not been enhanced.

Structural unemployment is unemployment due to a change in the jobs that need to be done. (Not unemployment due to the end of scarcity so there less need for labor.) I believe that the usual term to describe unemployment due to changes in the jobs that need to be done because of an improvement in productivity is technological unemployment.

I don't want to dismiss the hardship resulting from technological unemployment. Nor do I wish to dispute the wisdom of the ages that firms respond to crisis with productivity enhancing changes. Sure, it would have been profitable to make these changes all along, but the need to cut losses enough to survive may focus the minds of managers more than generating greater profits for the owners.

Yet it is essential to always remember the basics. Improved productivity means that output will be higher than it otherwise would. The reason why some are left unemployed by the process is shortages of products or resources that have not had enhancements in productivity. If we have a situation where rapid growth of labor productivity in some sectors has led to higher technological unemployment, we should observe a higher growth rate of output and shortages and bottlenecks somewhere in the economy.

While it is true that job openings have increased over the low point of 2.3 million in July 2009, the current level (March, 2o10) of 2.7 million is well below the 4.8 million before the recession. Hires have also increased during that period, from 3.9 million per month in June 2009, to 4.2 million per month in March of 2010.

When vacancies begin to get close to their their pre-recession level, and hires remain lower than that level, then it will be time to point out that any remaining unemployment is largely structural or technological.

During the heyday of Keynesian fine tuning, the unemployment rate and the level of real output were targeted. If unemployment was high and real GDP low, then expansionary policies would be utilized. If structural or technological unemployment have increased, then responding by efforts to expand money expenditures would be a mistake. If the process of redeploying resources has temporarily reduced the productive capacity of the economy, then responding to lower output by expanding money expenditures would be a mistake. Oddly enough, if technological unemployment due to more than normal increases in factor productivity were the "problem", noting that output continues to grow as usual would not a sign that all is well. Output should grow more than usual. From a Keynesian perspective, an expansionary policy would be in order. (George Selgin's arguments for the "productivity norm" has convinced me that an expansionary policy is unnecessary, as explained above.)

Currently, the Fed follows a neo-Keynesian approach, adjusting the federal funds rate (as long as it isn't "too' low) in order to reduce any gap between real output and the productive capacity of the economy and to keep the expected increase in the core CPI rising 2 percent from its current value. This approach requires an estimate of the productive capacity of the economy. If there is structural unemployment due to a redeployment of resources, the productive capacity of the economy will be at least temporarily depressed. If there is technological unemployment associated with enhanced total factor productivity, then the productive capacity of the economy will be permanently enhanced. Trying to figure out the underlying level of unemployment and its exact source appears to be necessary given the Fed's "output gap" approach to policy.

In my view, targeting unemployment, real output, or output gaps is a mistake. Like Beckworth, I think the least bad approach is to target demand (i.e. total spending, i.e. cash expenditures, i.e. nominal expenditures, i.e. final sales of domestic product.) Real output, the unemployment rate, and the price level should all be left free to adjust according to market forces.

P.S. The purportedly technologically unemployed worker in the story was actually employed at Walmart. While she was only working part time, surely an expansion in demand would allow her to put in more hours. And that, of course, shows the real pain of "creative destruction." Some businesses fail. The owners lose. And it isn't just the entrepreneurs that bear the burden. Some people end up in new jobs that have lower pay and benefits than their old jobs. Yes, there is unemployment associated with technological change, but that is not the real problem. We all benefit over the generations for putting up with creative destruction. But the costs are not distributed fairly. Some people, through no fault of their own, must start over. Maybe as a cashier at Walmart.