The price-concentration relationship in grocery retailing: Retesting Newmark

A study by Newmark is one of two price‐concentration studies of U.S. grocery retailing that has failed to find a positive relationship. We examined four possible sources of experimental error in this study. The most important source of error was the failure to exclude noncompeting grocery retailers in the construction of the concentration measure. The relationship was also sensitive to controlling for a subgroup of observations from one state, to alternative data sources, and functional form. When these changes were made, concentration was found to be positively associated with the cost of groceries across cities. The relationship was significant despite a small nonrandom sample of cities and unscientifically sampled prices. [Econ Lit citations: L130, L810, L100] © 2002 Wiley Periodicals, Inc.