Deriving a Trans-Log Cost Function for Pharmaceutical Distribution

Average wholesale drug prices (AWP) continue to rise at twice the rate of general inflation. Consolidation among distributors, which should increase efficiency, has done little to decrease the upward trend in prices. To better understand the cost structure of pharmaceutical distribution, we derive a transcendental logarithmic (translog) cost function using publicly available data. To our knowledge, this is the first application of production cost economic analysis performed on distribution centers using public data. We collected data for 92 pharmaceutical distribution centers (DC), representing ninety-five percent of industry sales, from four databases. We calculated five cost related variables for each DC based on US Census data, corporate financials, and site specific data. We specified Cobb-Douglas and complete translog models. Both models converged using various non-linear techniques to minimize residual errors. We discuss implications for logistics researchers, policy makers, and mangers.