This paper examines the relationship between infrastructure and productivity, focusing on roads as the largest component of public capital. The author, John Fernald, an economist at the Federal Reserve Board, uses data from 1953 to 1989 to analyze how changes in road growth affect the productivity of U.S. industries. The study finds that vehicle-intensive industries benefit disproportionately from road-building, with productivity growth changing more in these industries when road growth changes. This suggests that causation runs from infrastructure to productivity. However, there is no evidence that roads offer an above-average return at the margin, indicating that the massive road-building of the 1950s and 1960s provided a one-time boost to productivity rather than a sustained source of rapid growth. Additionally, the paper explores the impact of congestion, finding that it becomes empirically important after 1973, affecting road services and productivity. The findings suggest that while roads were exceptionally productive before 1973, they are not exceptionally productive at the margin.This paper examines the relationship between infrastructure and productivity, focusing on roads as the largest component of public capital. The author, John Fernald, an economist at the Federal Reserve Board, uses data from 1953 to 1989 to analyze how changes in road growth affect the productivity of U.S. industries. The study finds that vehicle-intensive industries benefit disproportionately from road-building, with productivity growth changing more in these industries when road growth changes. This suggests that causation runs from infrastructure to productivity. However, there is no evidence that roads offer an above-average return at the margin, indicating that the massive road-building of the 1950s and 1960s provided a one-time boost to productivity rather than a sustained source of rapid growth. Additionally, the paper explores the impact of congestion, finding that it becomes empirically important after 1973, affecting road services and productivity. The findings suggest that while roads were exceptionally productive before 1973, they are not exceptionally productive at the margin.