To the Editor.
—Dr Goodman and colleagues1 should be commended for their sound methodological approach in delineating wide regional variations in physician supply. As Dr Schroeder2 points out in his accompanying Editorial, population benchmarking exposes a stark 3.5-fold difference between highest and lowest levels of per capita generalists and 2.75-fold difference among specialists. These significant discordances in resource allocation have implications regarding cost and efficacy of medical services delivery. However, we take issue with the authors' position that benchmarking is an alternative to needs-based or demand-based workforce planning. Rather, benchmarking should be used as one of several tools within an overall framework for analyzing physician specialty supply and requirements.Although Goodman et al identify current overages and shortages in regional markets, their analysis does not offer the policymaker levers with which to make desired changes. Other workforce planning tools allow for modifications in projected utilization and clinician productivity,3 incorporation of