Infrastructure, Stimulus, and Jobs

I’ve talked before on the subject of infrastructure as stimulus, arguing that it’s ideally used for projects with one-time costs and ongoing benefits. Tonight I want to discuss a specific aspect of this: jobs. American infrastructure projects always talk about how many jobs they will create as a benefit rather than as a cost, and even in Europe, the purpose of the Green Deal investment package is to create jobs. In contrast with this view, I believe it is more correct to view infrastructure stimulus as an unusually bad way of creating jobs to deal with unemployment. The ideal infrastructure package really has to be about the benefits of the projects to users, and not about temporary or permanent employment.

The key question when designing stimulus is, unemployment for whomst?. Unemployment is predominantly a problem of unskilled workers. The OECD has a chart of unemployment by education level, and the rates for people with tertiary education are very low: in 2019, the US and Germany were at 2%, France at 5%, and even Spain only at 8%, all standing around half the overall national unemployment rates.

In theory, this makes infrastructure a good solution, because it employs people in the building trades, who are not university graduates and who have swings in employment rates based on private residential construction. In practice, it is not the case, for two reasons.

First, infrastructure projects have a long lead time, and therefore by the time physical construction happens, the recession has ended: the Green Line Extension in Boston, funded by the Obama-era stimulus, has mostly been under construction at the peak of the current business cycle, with such a shortage of labor that the contractors had to offer workers a full day’s pay with overtime for just five hours of nighttime work to get people to come in.

And second, while the building trades have large swings in employment, they are not good targets for absorbing the mass of laid off workers in recession. It takes years to get certified. This is not the 1930s, when construction was more labor-intensive and less skilled, so that armies of unemployed workers could be put to work building bridges and hydroelectric dams. Construction today is more capital-intensive (how capital-intensive, I can’t tell, since the full capital-labor ratios for the projects we’ve delved into are buried beneath layers of subcontracting), and the workers, while not university-educated, are much higher-skill.

Swings in employment are the most common among unskilled workers who do not have a special qualification. Those are workers in retail, restaurants, sundry small non-essential businesses that depend on the state of the economy for sales. The public sector is rather bad at absorbing them, because the stuff the public sector is or should be doing – the military, police, health care, education, social work, transportation, infrastructure – employs workers who are not so interchangeable. Health care, education, and social work involve massive numbers of people in intermediate professions; the military requires long training and a long commitment and countries that use soldiers as cheap labor for civil infrastructure projects end up weak in both infrastructure and defense; infrastructure uses workers in trades that usually involve years of apprenticeship. The main employers of the workers most at risk of unemployment are private, doing things the state would not be able to provide well.

So if the point is to limit unemployment, it’s best to stimulate private-sector spending through direct cash aid, and not through large state-directed development programs. Those have their place, but in the economic conditions of the 2020s rather than the unfairly romanticized middle of the 20th century, they are not good tools for reducing the impact of business cycle on workers.

And if the point is to build infrastructure, then an infrastructure package is a great tool for this, but it must be built based on maximum value and long-term savings. The number of jobs created should under no circumstances appear in any public communications, to deter groups from extracting surplus by claiming that they provide jobs, and to deter false advertising when in reality the jobs created by public-sector construction tend to be created when the recession is over among groups that do not need stimulus by the. Instead, infrastructure should center the benefits to the public, to be provided at the lowest reasonable price; labor, like concrete and lumber, is in that case a cost, and not a benefit.


  1. Brett

    This is not the 1930s, when construction was more labor-intensive and less skilled, so that armies of unemployed workers could be put to work building bridges and hydroelectric dams.

    I don’t think that was true even back then. Most of the public works employment in the US during the Great Depression was on stuff like light construction, rather than heavy construction like building dams and bridges.

    • Vic

      It was definitely true, the PWA funded dams, bridges, tunnels, airports, etc. The Hoover Dam was a Public Works Administration project, as were the Lincoln Tunnel and Triborough Bridge, to name a few.

      • Henry Miller

        They got around the problem of the article (needing to do design before building) by way overbuilding. Where 1 foot of properly designed concrete would work they used 3 feet. This both meant for more laborers were needed, and also the design didn’t need to be as carefully checked allowing the use of less skilled labor. Today nobody will agree to not doing proper design (for good reason – sometimes what seems like overbuilding is dangerously under built) so we can’t just throw labor in as fast. We can still overbuild if we want to use more labor, but nobody wants to pay for that, and in any case it will be a few years before we can do this.

        What we could do it 5 years ago design “form based codes” which are carefully designed rules that if you follow allow you to build with a lot less design. However you are limited in what you can do: You can build a station in the tunnel subway like Spain does, but even cut and cover stations will be hard to design (not impossible). The idea is we have the parts to build the system and so long as you put them together like lego blocks anyone can build them. The only part that needs to be designed is the foundation (and that only because you need a geologist to test that the soil will support the weight) However the parts give you strict limits as to what you can do, instead of wide open design choices.

      • john

        All three were planned and started years before receiving PWA funding (hence the Hoover Dam’s name), so the lead time had already passed; in these cases PWA basically gave an infusion of cash to already ongoing (or stalled–IIRC both the Lincoln Tunnel and Triborough Bridge had either stalled or slowed for a few years before PWA chipped in) projects to ensure they were completed, which is actually a pretty good use for a bit of stimulus, and quite different from throwing anti-recession stimulus at long-term projects that haven’t started yet, which is what Alon is arguing against here. I don’t know what proportion of PWA projects were keep-current-projects-going, versus brand new start, though.

  2. Basil Marte

    Many people habitually compress all evaluations onto a single yay/boo scale (this is called the halo/horns effect). This is how people end up saying “these people deserve jobs for being nice/moral/etc. (insert romanticized image)”, “these people deserve being admitted to university for being Aryan”, and so on, which in turn can get rolled into evaluations of infrastructure projects (also including “this design is good because it busts unions”).

  3. Matthew Hutton

    Could you use stimulus spending to build cycle lanes or something? Or is that also hard?

    Otherwise this is a very interesting post.

      • fjod

        Pop-up lanes are usually not good cycle infrastructure though. In some cases they are worse than nothing, forcing riders into the uneven gutter of a road and giving them a false sense of security when the lane disappears at a junction (which is where cyclists are really at danger). Building actual cycling infrastructure requires the same technical skills as realigning/rebuilding roads.

        • WL

          The amount of money for bicycle infrastructure (the on street ones) is pretty cheap even including junctions.
          It gets really expensive in America when they try rearranging the road to save some parking/car lane and end up needing to move some utilities. Or if delayed for 10+ years by environmental review.

          • fjod

            You do usually need to rearrange the road for good cycle infrastructure though – and sometimes this includes utility relocation. Painted lanes on the same camber as the main carriageway, which end at junctions and are full of drain/utility covers, are not what you should be aiming for – they are neither safe nor comfortable.

  4. Tom the first and best

    The part of an economic downturn in private sector housing construction could be with off-the-shelf shovel-ready or near shovel-ready (land already owned or subject to pre-planned and notified compulsory acquisition at short notice, designs done, only a small amount of planning approvals left to do, etc) public/social housing and counter-cyclical stimulus. That does require a lot of pre-planning by committed and planning funded public housing authorities.

    The environmental parts of the Green New deal are more structural stimulus, to restructure the economy to be more environmentally realistic, rather than counter-cyclical stimulus.

  5. Frederick

    Politicians who tout their public transit projects as “job creating” couldn’t care less about job creation; they use such phrase because they think it’s what their voters want to hear.

    This is especially true in America, where the voters either don’t care about public transit, or don’t believe that they can deliver a satisfactory piece of public transit infrastructure. Job creation might be the only real merit of such project.

    • Henry Miller

      Having listened to the hearing on transit a few months ago I’ve concluded you are wrong. Politicians care about jobs only – tranist is for most just a way to get union people paid (union is important here), if anything comes out of it that would be bad as it means that they can’t keep sending money to that project.

      There are exceptions, a few politicians do care about transit. About the same number on both sides of the aisle were asking questions related to actually getting people around. They are a minority though. The vast majority care only about paying off their cronies.

      • Eric2

        The problem is that the US doesn’t have a real social safety net (Republicans oppose one and Democrats don’t have the votes). So you basically have to work or become homeless. But with technological change, growing numbers of people have become borderline incapable of supporting themselves by working. So any politician who promises new jobs will automatically gain a large following of people who are justifiably paranoid about the possibility of becoming unemployed.

  6. ericson2314

    Whether they care about actual “job creation” or just the optics hardly matters, because both of those trump good public transit which, as you say, neither the voting public or politicians here seem to genuinely care about.

  7. Ericson2314

    Very happy to read this. The fact the US politicians continue to insist on glamorizing work is a peculiar thing. Certainly, they worry that infrastructure isn’t actually popular, and jobs are. They are probably also defeatist on some level that the material culture can actually change.

    > So if the point is to limit unemployment, it’s best to stimulate private-sector spending through direct cash aid, and not through large state-directed development programs.

    Yes absolutely. The resistance to this is very telling. The small-business-idolizing Right hates loosing it’s coercive bludgeon. The Left is more interesting. On one hand I think anything so monetary and abstract peeves the new left romantics. At the same time, the backlash against Keynesian has only partially abated, and people still can’t let go of a fear that “just” boosting demand (creating jobs being two-sided) actually works, and won’t conjure up inflation demons from the 1970s or whatever.

    > The main employers of the workers most at risk of unemployment are private, doing things the state would not be able to provide well.

    Well, if the private sector won’t retain then, they are not necessarily productive there either. We should hand out the cash firstly because people deserve a basic standard of living, and secondly, because only through boosting aggregated demand (be it planned or market-directed) will there be enough work to do.

    Finally, I think it is worth pointing out that infrastructure and basic income go hand-in-hand because it’s precisely in raising demand that we find the bottlenecks that are most productive to address via infrastructure. Likewise for carbon dividend to shift demand and break suburbia’s fecundity.

  8. Gregory Homatas

    Hello Alon

    Thought you might be interested in commenting and looking into why the MTA consultants report on the estimated cost of reactivating the Rockaway Beach Branch of the LIRR appears so high. Seems out of line people say with similar projects. Here is a link you can follow;

    Something gives here. Maybe you can shed some light on it and comment.

    Advocates for this project believe this will spur economic growth to southern Queens and the Rockaways and provide a one seat ride from midtown Manhattan to JFK airport.

    Greg Homatas

  9. sacolema

    Not sure if this is really relevant, but I thought I’d leave a quiet nod to the Jobs Guarantee Program (p. 123):
    As a note, most of the ideas come from Mitchell, Wray, and Watts (2019). Also, the references to previous letters refers to the following:
    Apologies for the poor writing and organization, as the letters were written when I was younger.

  10. Robert Fizek

    I must note that there is NO indication that anyone is really accounting for the climate impacts in this infrastructure work?!?
    It’s truly massive! And involves much wasting of -maintenance deficient- yet functional existing structures.
    We really don’t have time to keep DOING things without planning and choosing the the least carbon costly design, and construction means, and methods, and ideally even NOT-DOING -reusing, rehabbing, updating existing built infrastructure and architecture…
    -to not keep adding to the problem, and give us a bit more time to get our acts together on how to build better.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.