Can Infrastructure Spending Really Stimulate the Economy?

Transportation, power, and water facilities are public goods that benefit everyone in the economy, and government provision of these goods is critical to the economy. That is not a controversial conclusion. However, infrastructure projects are often touted as excellent methods of fiscal stimulus, irrespective of the benefits of the end products they produce. This claim requires more scrutiny because the effectiveness of infrastructure projects as economic stimulus isn’t nearly as clear cut as the benefits of the infrastructure they produce.

This scrutiny is doubly important because infrastructure projects are especially appealing to politicians as a form of fiscal stimulus. The sprawling construction sites that infrastructure spending creates are a visible reminder to voters that the government is working to address a crisis. This means that concerned citizens need to be aware of the strengths and weakness of infrastructure as a form of stimulus, because politicians may default to it due to its power as a political signal.

It’s also important to keep in mind is that the question is not whether infrastructure spending boosts the economy, but whether it does so better than alternative forms of fiscal stimulus.

Overall the empirical evidence is that infrastructure spending does have a stimulatory effect on Gross Domestic Product (GDP) that is larger than some other types of spending. However, its effectiveness as stimulus isn’t without caveats. In practice, it can only achieve this level of effectiveness in very specific circumstances, limiting its use to certain instances.

Theory of Infrastructure Stimulus

The idea of infrastructure spending as an economic stimulus is rooted in Keynesian economics. In Keynesian theory, when a recession happens the economy can get stuck with sustained high unemployment and a stagnant GDP for an extended period due to a deficiency of aggregate demand. When consumers and businesses buy less stuff, businesses lose sales fire workers, those workers buy less, and the cycle continues in a self-sustaining manner.

According to the Keynesians, one option to deal with this situation is for the government to directly make up for the lack of private sector demand by replacing it with demand from the public sector financed by deficit spending. In the broadest sense, this spending can really be on anything. Keynes created a thought experiment to prove his point that, if unemployment were extreme enough, it would be useful stimulus to the economy to simply bury bottles of money in a coal mine and let people dig them up. While this is often misinterpreted as a literal suggestion, it was meant to show that any form of fiscal stimulus could have a positive effect in closing the output gap in the economy. As Keynes himself said “It would, indeed, be more sensible to build houses and the like.”

How effective stimulus is at closing the output gap depends on the multiplier effect. The multiplier effect is a name for the fact that every dollar of government spending creates some additional amount of private sector spending. For example, the government hires a person to build a road, that person goes out and spends money at a store, the owner of which hires more workers with the money, and so on. The size of this effect depends on where those dollars are spent, if dollars are given to people who are going to save them, then the multiplier effect will be small, but if the government gives those dollars to people who will spend them, allowing them to flow into the economy, then the multiplier will be larger. This can allow a fiscal stimulus to have a significantly larger effect on the economy than just the number of dollars spent by the government, allowing the economy to be brought out of recession while minimizing deficit spending.

Economic Impact of Infrastructure Stimulus

Recent estimates by the Congressional Budget Office and a meta-analysis of empirical results from economic research suggest that public investment spending does lead to a stimulating effect on private spending components of GDP and has a larger impact on GDP via the multiplier effect than other types of spending.  On paper then, the aggregate effect of infrastructure spending would seem like an appealing option for fiscal stimulus.

However, if reversing the effects of a negative economic shock by stimulating the economy is the goal, then proponents of economic stimulus generally agree on three principles of what stimulus spending should look like beyond just the sheer size of the multiplier under the best circumstances. To be most effective a stimulus should be:

• Timely – In order to stop an economy that is in a rapid downward spiral, stimulus spending must get into the economy quickly. Spending programs that take months or years to complete may take too long to have a timely impact. Delays in spending might not only reduce the impact on a current economic crisis, but might even be counterproductive if they come too late and contribute to overheating the economy.
• Targeted – In order to stimulate the economy, spending needs to get into the hands of people who will spend it quickly to multiply its impact. Usually this means lower-income households and people who are most economically distressed by the downturn. Recipients who save the money or us it to pay down existing debt can defeat the purpose of stimulating new spending, and the multiplier effect of the stimulus drops.
• Temporary – Stimulus spending needs to be limited to the period when it is needed to deal with a recession. Otherwise, permanent increases in deficit spending can lead to unsustainable government debt, crowd out private investment spending, or create undesirable microeconomic distortions in the economy.

How does infrastructure stimulus stack up here? While empirical research suggests that infrastructure spending may have a strong multiplier effect overall under the best conditions, meeting these criteria may be a challenge.

Infrastructure construction projects may take a few quarters or a few years to even get off the ground due to implementation lag. This means that the stimulus may not be timely, regardless of its total impact. Construction spending tends to peak years after a project is started, by which time the economy is often already recovering. This can create a procylical pattern, where the spending is held up during the time when the economy is suffering and then later overstimulates the economy during times when it isn’t needed. In this case, the large multiplier effect associated with this kind of spending can be counterproductive, exaggerating rather than smoothing out economic cycles. While there may be infrastructure projects ready to fully fund at the time of the crisis, there are only a limited number of those. This means there are only so many infrastructure projects that would be useful as stimulus.

Because infrastructure spending is usually for a specific budgeted amount to fund specific projects, on its face it does tend to meet the criterion of being temporary, though cost over-runs and other issues can drag this out. One caveat is that infrastructure strongly influences regional economic development patterns. If infrastructure is built solely for the purpose of providing economic stimulus, not because it provides changes to regional economic development we want, it could cause significant negative long-term effects. This is doubly important to remember as infrastructure might be rushed to provide timely stimulus in a way that doesn’t consider longer-term implications. This further limits infrastructure stimulus to projects already significantly planned out and far along in development.

Lastly, targeting infrastructure spending effectively to meet macroeconomic goals can be problematic. Such spending tends to inevitably target the heavy construction industry, which may or may not be particularly hard hit in any given recession. Furthermore investment in fixed capital, like infrastructure, is necessarily highly localized; there is no reason to expect that the regional distribution of infrastructure needs will coincide with the geographic distribution of the impact of a recession.

This can create tension between the goal of economic stimulus and actual public need for the infrastructure. Moreover, several studies have shown that in practice the distribution of stimulus related infrastructure spending is often heavily influenced by political and electoral considerations rather that either of these two goals.   While this can make infrastructure spending very appealing to policy makers and politicians, it can work counter to the economic goals of the policy.

Infrastructure: Powerful Stimulus, but Only in Some Cases

The bottom line is that, under certain circumstances, infrastructure spending can indeed stimulate broad, macroeconomic aggregates such as GDP or total employment. However, because infrastructure projects take a long time to get started, they cannot always provide stimulus in a timely manner to help during a recession. Secondly, if infrastructure is rushed and planning stages are skipped to try and provide more timely stimulus, it could have long-lasting negative consequences to regional economies that do lasting harm well after the recession ends.This means that to be effective fiscal stimulus, the government would need to provide funding for projects that are already planned and started, of which there are only so many. Because of this, infrastructure is further limited as a tool for stimulus, because those existing projects need to be located in regions most severely hit by the recession, further limiting options. Finally, the recession needs to have hit industries like construction and heavy manufacturing that are involved in infrastructure creation, or else the stimulus won’t be targeted at the people who most need it. Its strong multiplier effect means stimulus can be a powerful tool for stimulus, but these considerations mean that can only be deployed effectively in a very limited way. If these considerations are ignored then infrastructure becomes a less than ideal fiscal policy tool, or even possibly a counterproductive one.