Standard economic models predict that an increase in employee health care costs would be financed by an equivalent reduction in cash wages (Summers, 1989; Gruber, 1994). In this model, total compensation remains unchanged even if health care costs rise, and health care cost inflation has no effect on employment or output.
However, this view is contrary to popular wisdom. One approach to reconciling the economic model with popular wisdom is to assume that cash wages are “sticky” (that is, firms have limited ability to reduce wages in response to rising health care costs) and that rising health care costs lead to a less than one-for-one reduction in cash wages. In this case, rising health care costs will increase total compensation. The increase in total compensation creates incentives to reduce employment, since employment costs have risen but labor productivity has remained the same. The reduction in employment will also lead to a reduction in output and an increase in prices.
The extent to which an increase in health care costs affects output and employment in different industries depends on how much total compensation rises due to health care cost inflation. Industries where health care benefits are a larger share of total employee compensation will experience a larger increase in total compensation, consequently leading to a larger reduction in employment and output. For example, a 10% increase in health care costs will lead to a 5% increase in total compensation for industries with a benefit share of 50%, but only a 1% increase in total compensation for industries with a benefit share of 10%.
This section investigates how health care cost growth affects specific industries. It analyzes whether industries where benefits are a larger share of total compensation were hit harder by rising health care costs. The analysis covers the period 1987 – 2005. Since the classification of industries changed in 1997 (from SIC to NAICS), industry definitions from the SIC classification (1987-1997) were matched to those in the NAICS classification (1998-2005) using definitions that were consistent across the two time periods.7 The final dataset consists of 39 industries (matched across the two periods) over these 19 years, yielding a total of 741 observations. The share of benefits in 1987 was adopted as a baseline measure of the degree to which various industries might be affected by growth in health care costs. The rise in health care costs over time is captured by the medical care price index;8 the specific outcomes studied are employment, gross output, and value added to GDP, for each industry. Appendix 2 describes the data sources for each of the variables in the analysis. Table 8 provides summary statistics of the variables used in this analysis.