Productivity Growth in Construction

Measuring productivity growth in construction is especially difficult due to the nature of production in the industry and the limitations of available data. In particular, the price indexes used to deflate output are a major problem because reliable deflators are sparse and the available data suggest productivity has declined for many decades, which is somewhat difficult to believe. This paper first reviews Bureau of Labor Statistics (BLS) estimates of productivity growth in the overall construction sector. Second, we develop measures of labor productivity growth in three industries in construction where reliable output deflators already exist: single and multiple family residential construction from 1987 to 2011, and the construction of highways, streets, and bridges from 2002 to 2011. These data, which currently cover almost one quarter of construction output, show no sign of any sustained productivity decline. Productivity growth continues to be positive when subcontractors are included as labor input. On the other hand, the new PPIs increase more rapidly than previous unreliable output deflators did; such evidence suggests that negative long-term productivity trends are more likely than the current long-term time series data suggest. Third, we analyze labor shifts among 31 industries in construction. Previous work (Allen, 1985) has shown that, from 1968 to 1978, shifts from high to low productivity industries reduced productivity in construction by 0.46 percent a year. Our results indicate that labor shifts still cause a considerable decline in productivity. Over the last two decades, shifts across industries reduced productivity by approximately 0.26 percent a year. However, most of the productivity decline, especially between 1997 and 2007, still remains to be explained. Fourth, we present evidence that land use regulation, measured in each state in each year, has a small but statistically significant negative effect on productivity growth in construction. The evidence also indicates that the regulation share of total costs, the amount by which regulation increases construction costs, is 3.7 percent. However, we estimate that increases in regulation reduced productivity growth in construction by only 0.1 percent a year.