How Growth in Service-Sector Productivity Could Eliminate Inflation

Growth in U.S. service productivity has lagged miserably since 1973. In the mid- and late 1970s and throughout the 1980s expansion, many service businesses expanded by simply adding more people. This was the equivalent of digging the Grand Canyon with millions of people armed with picks and shovels--a mighty result but a terribly inefficient one. Indeed, after achieving an average annual productivity growth rate of 2.6% between 1948 and 1973, service productivity growth fell to a dismal 0.8% in 1973-81 and 0.9% in 1981-90.

But now services are facing pressures on many fronts to cut costs and improve productivity. Foreign firms are putting pressure on service companies unused to foreign competition. Domestic competition has increased the need for cost control and productivity increases as deregulation has obscured the lines between competitors in service industries. Continued cost-cutting pressures on manufacturers are increasingly feeding through to services, since manufacturers are a major customer of many service industries. Finally, the debt hangover problems that bedevil many service providers are putting additional pressures on many service areas to get costs under control.

In the short term, employment growth may be weakened by a return to robust productivity in the service sector. In the longer term, the United States will be more competitive globally and inflation will be subdued due to a downward impact on unit costs and labor surpluses that will result.

Let’s take a closer look at how much abysmal productivity performance has affected the nation’s inflation performance and what impact improvement in productivity growth may have the next few years.

In the 1970s and 1980s, low service-sector productivity growth meant that wage increases were not matched by output growth and were therefore inflationary. If compensation in a given year rises 5%, but productivity goes up 3%, the net inflationary effect is 2%; but if there is no productivity growth, the full 5% flows through to inflation because there are no more goods and services produced to absorb the additional compensation payments.

This inflation spread to the manufacturing sector despite its rapid productivity growth in the 1980s. Manufacturing employees, of course, buy services as well as manufactured goods, so their wages reflected higher service costs. This in turn pushed up the prices of goods.

The impact of service productivity can be measured by constructing a model of inflation and wages based on raw material prices, service and manufacturing productivity, capacity utilization, merchandise imports’ share of GDP and the effect of lagged inflation. The model can simulate what inflation would have been over the 1982-90 expansion if service productivity growth had increased at its 1948-73 trend instead of the feeble rate actually achieved. Reinforced by the “virtuous circle” set in motion between lower inflation and lower wage increases, the simulation produces dramatically lower inflation of only 1.5% between 1982 and 1990, instead of the 3.9% actually recorded.

If service productivity returns to its 1948-73 trend growth rate, the model indicates that inflation will be contained within a 1.5%-2.5% range, with an average inflation rate from 1991 to 1996 of just 2.0%. Further simulations of the model indicate that even if GDP grows significantly faster than the subdued 2.0% we are forecasting, the connection between higher output and higher productivity growth keeps inflation in check.

Any external shock to the model, such as a big rise in raw material prices, is quickly damped as long as service productivity continues to grow. If service productivity growth remains at the miserable rate of the 1982-92 expansion, however, inflation would continue at between 3% and 4% the next five years within a wide range of output growth projections.

Given the strength of the competitive pressures bearing down on the service sector, service productivity can rebound at least to a level close to its postwar average. The manufacturing sector solved a similar problem in the late ‘70s and early ‘80s. In the early ‘80s, manufacturing productivity grew even faster than its long-term trend. It closed the productivity gap that had opened between actual and trend productivity from 1977 to 1981. In fact, in a pleasingly symmetrical manner, the gap that opened up over four years took four years to close. The gap between actual and trend productivity in services has opened up over 17 years.

To achieve a similar symmetry, service productivity would have to grow at a rate of 3.8% to close the gap in 17 years. It’s interesting to note that with this projection, inflation would be eliminated between 1991 and 1996, and the average inflation rate for the whole 17-year period would be zero.

With so much competitive pressure being brought to bear, and so much fat to cut after the willy-nilly hiring of the 1970s and 1980s, and with the great possibilities offered by automation and computer technology, and just plain tighter management, the service sector can revitalize its productivity performance in the 1990s. This will make a major contribution to bringing about a long-term low-inflation environment for the U.S. economy.