Editor's Note: The authors prepared this piece as part of Infrastructure Week 2017. This national collection of events, meetings, and publications serves to convene leaders inside and outside government about infrastructure’s contributions to the broader economy. 1

  • 1. To learn more, visit http://infrastructureweek.org/

We’re going to rebuild our infrastructure, which will become, by the way, second to none. And we will put millions of our people to work as we rebuild it.” From the very first night of his election win, President Trump was clear about his intention to usher in a new era in American infrastructure. Since assuming office, the president and his cabinet continue to use the figure of $1 trillion over ten years to demonstrate the scale of their vision.

By any measure, one trillion dollars is a lot of money. Given the well-documented maintenance and modernization backlogs in a range of infrastructure sectors, federal attention is welcome. Infrastructure spending has the added benefit of helping to support millions of good-paying jobs.

But just how historic would a $1 trillion federal infrastructure program be?

Despite the allure, judging the historic nature of this figure is impossible without more context. This analysis provides additional background by using historical data from the Office of Management and Budget (OMB) on federal spending on physical infrastructure over the past eight-plus decades. The exercise confirms that while a $1 trillion infrastructure program would rank among the larger sustained periods of federal spending in relative terms, it would likely fall short of record spending under the New Deal.

Methodology and Approach

While there are several recent analyses tracking levels of federal infrastructure spending, changing budget classifications and differences in infrastructure categorizations make it difficult to conduct long-term comparisons. The Congressional Budget Office, for instance, is frequently cited for federal transportation and water infrastructure spending, but these data do not capture a broad range of federal activities, including energy and housing, that align with the 2017 conversation. Similarly, a variety of other organizations like the World Economic Forum and OECD, and media outlets like the Economist and Wall Street Journal typically concentrate on a more limited time horizon.

To get a more complete understanding of physical infrastructure spending at the federal level, this analysis uses the “physical resources” superfunction data from historical OMB budgets, which have greater consistency over a longer period of time. This category exists from 1940 through 2017, and includes the categories President Trump and his cabinet have mentioned as part of their ideal package: energy; natural resources and environment; transportation; and community and regional development. Since this analysis is also primarily interested in direct capital spending, it excludes the sizable Commerce and Housing Credit within this budget category.

This analysis views these levels of federal spending relative to earlier New Deal era programs in the 1930s, such as the Public Works Administration. The enormous size and number of federal infrastructure projects during this period figure prominently in the nation’s collective mindset. There are the epic visual reminders like the Hoover Dam, but there are also the less visible, but still significant efforts such as rural electrification and municipal water systems. Since OMB did not report data in the 1930s, this analysis required manually going through historic budgets and economic literature to construct comparable measures. While it is difficult to precisely define infrastructure investment during this decade, particularly given fluctuations in federal spending and economic output, this analysis attempts to highlight the relative height of spending during the New Deal. To do so, it includes peak spending in 1933 and average spending across 1933 to 1937. The figure below highlights an example of what spending just before the peak in 1933 looked like (and how they visualized the data).