The U.S. Senate approved its $1.2 trillion infrastructure package today and moved the bill to the House. But will spending so much money on roads, rail lines and other infrastructure significantly boost the economy, as proponents and some economists argue? In fact, this type of spending produces little short-term stimulus and only modest increases in long-term productivity and economic growth, extensive research shows, so the gains might not be worth the cost.
Improving a transportation network does promote economic activity by increasing the mobility of people and products. With greater mobility, people can search for jobs over a larger geographic area, resulting in better job matches. Workers may then be more satisfied with their jobs and become more productive. Making airports and seaports more efficient helps expand international commerce, which also leads to long-term economic growth. And spending on water and sewerage systems can boost productivity by improving public health in both the short and long term.
When it’s easier and faster to ship products to customers, the productivity of a business’s equipment, such as delivery trucks, rises. Less fuel consumption and shorter delivery times lower costs, making the operation more profitable. This encourages businesses to invest more and hire more people.
Indeed, most of the jobs created by infrastructure spending don’t come from building a project but from the benefits generated by a project once it’s finished. Construction projects are not very labor-intensive these days, but many politicians get this wrong.
But there are many problems with infrastructure projects. A big one is that they can take a long time. Former President Obama faced this reality in 2009, when he touted infrastructure spending to help end the recession but there were very few “shovel-ready projects” on tap. A large amount of planning, design and engineering work must be done before a major project moves forward, so it may be years before ground is broken. Legal, environmental and other challenges can delay and sometimes stop a project. So, the benefits are often far down the road.
Moreover, funding infrastructure projects can be difficult. Most are not directly paid for by a sudden federal spending spree, as with the Senate’s Infrastructure Investment and Jobs Act. Often state and local governments borrow the money and then taxpayers service the debt, which tends to be unpopular. User taxes, such as fuel taxes and tolls, are the most efficient way to pay for infrastructure. But sales, income and other types of taxes often finance projects, and these levies reduce the returns on working, saving and investing. This can reduce economic activity and offset some of a project’s benefits.
In many cases, politics results in projects that should never have been built—the famous bridges to nowhere and rarely used light rail systems. This spending is a drag on economic growth. And even if there’s a good case for a project, the return on infrastructure spending falls as the existing infrastructure expands. While building the interstate highway system has paid many dividends, this doesn’t mean we should build a second system.
Many studies have aimed to measure the direct impact of infrastructure spending on gross domestic product, and as the research has gotten better over the decades, the impact it’s finding is shrinking. Some of this work estimates the elasticity, or responsiveness, of GDP to changes in the public capital stock, or infrastructure. A more elastic GDP means that infrastructure spending is making a bigger mark. Early research estimated the elasticity at around 0.40. This means that a 1% increase in the public capital stock would raise GDP by slightly less than one-half of a percentage point.
Later studies found serious problems with that research. The methodology had overestimated the elasticity and newer work corrected those problems. A comprehensive meta-analysis of all the research put the short-term elasticity at between just 0.08 and 0.13; long-term elasticity was between 0.12 and 0.16. Taking the largest estimate, a 1% increase in the public capital stock would raise GDP by only 0.16% over time.
Other research calculated what is called the government infrastructure spending multiplier. A multiplier greater than 1 means that $1 of government spending raises GDP by more than $1 and thus earns a positive rate of return. A multiplier less than 1 means that the increase in GDP is less than a dollar. This negative return indicates that the extra government spending “crowded out” some private spending, perhaps because of higher interest rates or a stronger dollar in currency markets.
A wealth of research uses data from the U.S. and other countries covering the late 1950s to 2016 to estimate multipliers. Controlling for other factors that might have an impact on the economy, such as monetary policy, these studies find that government-spending multipliers typically fall below 1. Some estimates are close to zero. Once again the research shows that the impact of government infrastructure spending is modest at best and usually negative.
Before the pandemic, the last big infrastructure and stimulus spending bill from Washington was the American Recovery and Reinvestment Act, aimed at stabilizing the economy during the deep recession of 2008-2009. The package of spending hikes and tax cuts eventually cost $831 billion, including more than $120 billion for infrastructure. More than 12 years later, we can assess whether the spending had any impact.
Most of the analysis has focused on the state and county levels. It was hard to find much of a short-term impact because of the long time it takes to plan and complete infrastructure projects. One estimate using county-level data found that an additional $1 million in spending added six construction jobs at a cost of $150,000 a job. Overall, researchers found that the stimulus package did little to boost a region’s economic activity.
Both the history of U.S. infrastructure spending and the research into its effect on economic performance are clear: Such spending has, at best, only a small positive impact. It would be wise for President Biden and Congress to tone down their predictions for any infrastructure package that gets signed into law. And though “jobs” is in the title of the bill, they should certainly stop promising that building infrastructure will generate lots of jobs. Long delays, poor project choices and higher taxes to pay for everything will mute the economic impact.