The True Value of the Infrastructure Bill

Investment in US Infrastructure has long been a topic heavily discussed by politicians and the media. The most recent large investment project has been the Infrastructure and Jobs Act passed by the Biden administration in November 2021, commonly referred to as the Infrastructure Bill.  

The sticker price of this bill has been touted as $1.2trln over 8 years, but only $550bln is newly announced spend on infrastructure (the remaining $650bln is spending that is normally allocated each year). This number still seems like a hefty investment into the decaying US infrastructure. However, how much improvement will this $550bln (or approximately $70bln a year) generate, especially when considering the historic underinvestment in infrastructure spending?

US Infrastructure Deficit

US infrastructure has been significantly underfunded for the past several decades. The current infrastructure deficit is estimated to be up to $5.6trln. This deficit has been ever present, as annual spending on infrastructure is insufficient to cover the cost of yearly wear and tear. The deficit does not include any costs related to new projects (new airports, railroads or highways), but just the maintenance of the current infrastructure. Why is historic underfunding important?

Let's take an example of everyday infrastructure – a home. If you have a minor pipe leak, fixing it immediately will generally entail much lower costs than waiting for the problem to grow. However, sometimes it might make sense to do a general gut renovation, rather than patchwork. Similarly, these types of choices can also be applied to larger infrastructure projects such as roads, bridges, railroads, etc. - is it better to do a smaller fix or wait for a bigger renovation. Thus, we can ask what should be the optimal and most cost-effective approach to tackling infrastructure issues.

Many studies have been conducted by civil engineers and economists that attempt to model the optimal maintenance schedule for different types of infrastructure. For example, Zhang and Wang (2014) look at the optimal maintenance strategy for maintaining metal girders on rail bridges. These girders have 5 levels of deterioration, with the fifth level being critical to safety. There are multiple possible maintenance strategies available - we can choose how long we let the girders deteriorate before stepping in to fix them. In this study, for example, the authors find that allowing the girders to deteriorate to the worst state costs approximately 9 times more to fix than if the optimal and most cost-effective maintenance strategy was followed. However, the optimal strategy is very often not feasible due to budget limitations.

The example above shows that focusing on the amount being invested is not the correct metric when judging the infrastructure spend. Instead, we should focus on what we are actually getting in return for this investment. If we were to follow the optimal repair strategy, it might appear that we are not making a significant investment into rail bridge repairs. On the other hand, by waiting until the last moment to undertake the repairs, the repair costs would be far greater, making it seem as if we are investing a lot into infrastructure, while really we would end up with the same rail bridge as before. 

In the context of the Infrastructure Bill, the proposed spending today of $550bln, although sounds significant, will most likely not get us a lot of working infrastructure (“functioning rail bridges”), especially if we had instead spent the same amount over the past decade to maintain it properly. The debates around increasing investment in infrastructure have been a constant topic in US public discourse, with the last significant bill being the American Recovery and Reinvestment Act of 2009, which had a $105bln earmarked for infrastructure.

Estimation of Delay

So if the United States had decided to spend on infrastructure several years ago, how much cheaper would it have been? The answer to this question is not immediately available. Furthermore, the extra cost caused by deterioration is not the only source of loss – for example, delayed road repairs increase traffic congestion, which further generates a real economic and environmental cost (idling in traffic burns extra fuel, polluting the environment, as well as costing time). 

Focusing purely on the extra cost caused by the additional deterioration due to delays in maintenance spending, one estimate by Common Good in 2015 suggests that delaying just rail fixes by 6 years costs an extra $22.5 billion, or 30% more than if the repairs were done immediately. The 30% increase in costs only takes into account the increase in material and overhead costs, and not any additional impacts of wear and tear. Thus, the $66bln in spending on rail in the Biden Infrastructure bill, would approximately be worth $51bln in 2015 terms, if we just control for the annual cost increase for the same repairs. Similar analysis can be performed to other categories of infrastructure spend in the bill - $110bln for roads and bridges, $65bln for the power grid, $65bln for broadband, $55bln for water infrastructure, $39bln for public transit, and the rest on ports, airports, environmental clean up and electric vehicles. To estimate conservatively just based on the increase in material costs and overhead, in order to achieve the impact of the $550bln being spent, we would need only $423bln in 2015. 

If we were to attempt to extrapolate further, for example to 2010, the estimate would not be a simple linear extension, because infrastructure deterioration costs are exponential, that is every additional year of delay increases the costs of fixing infrastructure by more than the previous year. This question can be modeled out, but I have not been able to find any research that addressed this.

Conclusion

It appears that the Infrastructure Bill spending is going to make a very small dent into the real infrastructure needs of the US. Due to the significant spending delays and underfunding over the last several decades, the impact of this bill will be far lower than the sticker price. 

Interestingly, spending on infrastructure is estimated to have one of the higher rates of return of fiscal stimulus in the US economy – Moody's Analytics found that a $1 increase in infrastructure spending leads to a return of $1.44 to the US economy. Using this value, the approximate impact of the new spend in the Infrastructure Bill on US GDP will be an increase of around $100bln each year, or around 0.4% higher GDP. The current spending bill is far too little (or far too late) for the current US economic needs. Due to its limited size, the effects of this Bill will most likely not be seen or felt by most people, which will lead it to be judged (incorrectly) as a failure.