Earnings

To provide some measure of changing earnings in the US, the Bureau of Labor Statistics collects data on the average weekly earnings of workers and the Employment Cost Index (ECI). Average weekly earnings are provided monthly and can be found in the two publications, Monthly Labor Review and Employment and Earnings. These figures, obtained in the same employer survey which proves estimates of employment, provide estimates of the wages and salaries of all production workers, as well as a breakdown for major industrial sectors (manufacturing, services...).  In the sub sectors such as manufacturing, workers engaged in technical, office and sales positions are also excluded. When interpreting the time-series patterns it is important to recognize that the figures are 'averages' and as a result the numbers are quite sensitive to changes in the composition of jobs. For example, the decline of jobs in the manufacturing sector has produced a shift of workers from high paying, union manufacturing jobs to lower paying, non union service sector jobs.  This shift means the average weekly earnings data for the private sector will underestimate the change in wages for specific occupations or industries.

The Employment Cost Index is provided quarterly and can also be found in Monthly Labor Review. The ECI differs from the Average Earnings data in two important respects. First, it includes information on fringe benefits - an important component of labor compensation. Second, it is designed to minimize the impact of compositional shifts - the ECI is not affected by changes in the mix of high- and low- wage jobs. The ECI, very much like the CPI measure of price level, is based on a 'basket' of occupations and industries set by the US Census every ten years. A comparison of the ECI and average weekly earnings data for the period 1980-1995 indicates very little difference in earnings growth between the two measures.

It is clear from these "stats" that earnings have increased, but we also know that prices and the cost of living has increased.  Before we leave our discussion of earnings, we need to discuss the adjustment process and then look at the adjusted data.  As you will see, the answer is quite sensitive to the measure of earnings and the means of adjustment.