Determining the cost of innovative products and services requires information on the actual resources used to produce them, including staff time, supplies, and overhead. This approach is called "micro-costing" or "direct measurement." This method requires information on the cost of labor, including salaries and the employer cost of benefits. There is additional information available on the VA intranet at http://vaww.herc.research.va.gov/include/page.asp?id=labor.
Labor costs differ across the US. Labor in more expensive markets, such as San Francisco or Boston, can cost 70% more than the national average. Researchers need to account for regional differences, otherwise they run the risk of claiming that providers in high cost labor areas (e.g., San Francisco) are less efficient than low cost labor areas (e.g., Kansas City). One way to adjust for local variation is to use facility-level fixed effects, as long as the data don't span too many years. Another way is to use the Medicare Wage Index.
Health care costs are more expensive in geographic areas that have higher wages (e.g., Boston, San Francisco). Because labor is a major input into the health production function, this geographic variation can confound cost analyses. There are two common ways to control for geographic variation. The first involves including a dummy variable (fixed effect) for each medical center. The second method involves including the Medicare Wage Index in the analysis. Medicare creates the Wage Index to keep track of labor costs in geographic markets. HERC linked the Medicare Wage Index to VA hospitals using the station (sta3n) and substation (sta6a) identifier.