* Chairman Greenspan believes that globalization has held down wages and unit labor costs in recent years, but that the effect is only temporary. As the integration of emerging market workers into the global labor market nears completion, he sees upward wage pressures reemerging.
* To assess the potential importance of globalization, we look at a modified version of our long-standing nominal wage growth model. It has overpredicted wage growth by an average of 1/4 percentage point since 2000 and about 1/2 percentage point in 2005 so far. This error might well be due to globalization.
* But this doesn't mean wage inflation is dead. Even if the 'globalization gap' remains in place, our model suggests that nominal wage growth should accelerate to about 4% by 2006. This would likely contribute to a pickup in core CPI inflation, to about 2-1/2% by the end of 2006 from 2% now. If the globalization gap were to disappear, a bigger inflation pickup could well loom.
Is Wage Inflation Dead?
The most noteworthy idea in Chairman Greenspan's testimony to the Joint Economic Committee of Congress was his assertion that the entry of workers in the former Soviet Union, China, and India into the global labor market had held down wages and unit labor costs, but that the associated inflation benefit was only temporary. In Greenspan's view, the 'shift in the world's workforce is producing, in effect, a level adjustment in unit labor costs.' Inflation pressure is lower while this adjustment is underway but rises back to more-normal levels thereafter.
How large might this effect be? To address this question, we have refined our long-standing model of nominal wage determination. The model explains the quarter-on-quarter annualized growth rate of average hourly earnings, AHE, by the current unemployment rate, U, the year-on-year growth rate of the headline PCE deflator lagged 6 quarters, PCE(-6), and the current 5-year median inflation expectation as measured by the University of Michigan, Infexp. (Since the inflation expectations measure is only available back to 1979, we use the current year-on-year growth rate of the core PCE deflator prior to 1978.) The main difference between this model and our previous version is that we now include expected inflation as well as lagged actual inflation.
Here is the resulting equation:
AHE = 3.9 + 0.35AHE(-1) + 0.38*PCE(-6) + 0.38* Infexp -0.70 U Sample 1964Q3 - 2005Q2, R-sqd.=0.81, DW=1.80.
To assess whether wage inflation has been unusually depressed, we have looked at the residuals from this model over the last five years. On average, the model overpredicted nominal wage inflation by 1/4 percentage point since 2000 and by 1/2 percentage point in 2005 so far. This could well be due to the globalization effect noted by Chairman Greenspan.
Even if this effect remains in place, however, our model suggests that wage inflation will accelerate to just over 4% in 2006 from 3.3% in the third quarter of 2005 (see the table below). This is because the unemployment rate continues to decline, inflation expectations are picking up, and headline inflation has risen (the impact of this factor is already 'baked in the cake' as it is lagged headline inflation that matters). In such an environment, total hourly compensation would probably grow at around a 5% rate, since non-wage labor costs are likely to outpace wages somewhat. This would be consistent with our forecast of a gradual pickup in core CPI inflation to about 2.5% by the end of 2006.
Average Hourly Earnings (quarter-on-quarter annualized)
Actual* Model Prediction 200401 1.8% 2.3% 200402 2.6 2.6 200403 3.4 2.9 200404 2.3 2.9 200501 2.3 3.0 200502 2.8 3.2 200503 3.3 3.5 200504E 3.5 3.9 200601E 3.6 4.1 200602E 4.1 4.5 200603E 4.2 4.6 200604E 4.1 4.6
* Assuming 'globalization discount' remains at 2005 level.
Source: Department of Labor. Our estimates.
If the globalization effect diminishes, the pickup could be more substantial. In this case, the table shows that our model predicts nominal wage growth of 4-1/2%. This could imply hourly compensation growth of 5-1/2% and might well result in a rise in core inflation toward 3%.
Jan Hatzius