Improving productivity measures of producing transportation infrastructure using quality-adjusted price indices

Productivity measures are based on either physical or revenue-based measures of output. Revenue based measures require the use of a price index. In this paper we propose an approach to generate a quality-adjusted producer price index for the production of roadways. We use price data for highway projects across the contiguous United States from 2005 through 2017. The data set includes approximately 5000 unique items for each project. These items are distilled to form 60 "item baskets". We redefine our output from being lane-miles to 'lane-miles of service'. The indicator of quality is the deterioration rate of a roadway, which is measured using data on pavement roughness that links to deterioration and the time between required servicing; an improvement in quality reduces deterioration and increases the time span between maintenance and rebuilding. We use a chained-Fisher price index and find that our proposed, quality-adjusted producer price index exhibits 1.7 percent lower annual growth than the unadjusted price index. Given price inflation has been overestimated in the past by failing to account for quality changes, our finding suggests the lack of productivity growth in producing roads, bridges and infrastructure, may have been significantly underestimated.