I have argued on this blog that the best practical definition of infrastructure is “whatever the government wants to pay for because it benefits from the expenditure.” The adoption by the House of a $1-trillion bipartisan infrastructure bill reinforces this argument. (See Gabriel T. Rubin and Eliza Collins, “What’s in the Bipartisan Infrastructure Bill? From Amtrak to Roads to Water Systems,” Wall Street Journal, November 6, 2021.)
The standard argument for public infrastructure is that it is an investment that will yield a high rate of return in terms of future economic growth. But there are many caveats. The returns may be lower than those of the private investments displaced by government financing of infrastructure. (See Josh Mitchell, “Infrastructure Law Seen Having Small, Positive Impact on Growth,” Wall Street Journal, November 6, 2021.) More generally, if the government produces something, something else cannot be produced. And we have no credible theory showing that political and bureaucratic processes will choose the most productive investments.
The standard concept of public infrastructure may also run into a logical paradox. As the term suggests, infrastructure is conceived as underlying the whole structure of the economy, that is, of exchange and trade. This seems to imply that nearly everything is infrastructure by association. Roads need to be paved, at least in an advanced and efficient economy. So asphalt must also be infrastructure (or infra-infrastructure?). What about asphalt rollers and the steel that goes into them? The steel, aluminum, and concrete in road safety barriers? The iron used in steel manufacturing? The steelworkers themselves?
The infrastructure bill gives an idea of such infinite regress. It devotes $66 billion to rail maintenance, modernization, and expansion—mainly for Amtrak. Another $39 billion is scheduled for the modernization of public transit and its accessibility to the disabled and elderly. Energy is also part of infrastructure, at least when it is the sort favored by the government. The bill also promotes “environmental justice” and “good-paying union jobs”: is that infrastructure too? The $1.75-trillion “social” and environmental bill the Democratic Party is pushing is said to target “human infrastructure.”
Certainly, it cannot be infrastructure “all the way down.” This intriguing and humorous expression has many reported origins revolving around some old mythological belief that the earth stands on the back of a humongous turtle. But what does the turtle rest on? On another turtle, of course. And what about that turtle? “Ah, Sahib, after that it is turtles all the way down.” (In some versions, elephants stood in the chain.)
Note that, as is usually the case, much of government infrastructure propaganda is politicians’ smoke and mirrors. About half of the infrastructure bill is made of expenditures already planned, and the whole package is to be spread over the next five years. (See “Joe Biden Passes the Less Contentious Half of his Legislative Agenda,” The Economist, November 7, 2021.)
Perhaps there are some activities that only the state or the government, in the general sense of political authority, can do to support any system of social cooperation including a free society. (That’s the one-million-dollar question in political philosophy and economics.) Perhaps then we could meaningfully limit the meaning of “infrastructure” to the production of “public goods.” But it is doubtful that such things include Amtrak, electric car chargers, or broadband service which is already produced by private businesses. Or perhaps we should say that “infrastructure” is limited to fundamental conditions of social cooperation such as abstract institutions (like the rule of law) that favor exchange and individual liberty? This sort of definition is normative as opposed to the descriptive one I suggested before and would provide a guide to what the state should do.
READER COMMENTS
Floccina
Nov 9 2021 at 11:18am
Why not rather than funding rail, work to make it faster and easier to get to airports and on your jet. I’m looking at you HSA.
Pierre Lemieux
Nov 9 2021 at 2:26pm
Floccina: Who is “HSA”? (https://www.acronymfinder.com/Slang/HSA.html?)
Craig
Nov 9 2021 at 11:45am
We are also potentially at a technological inflection point with respect to driverless cars. If that is perfected; it changes everything.
Pierre Lemieux
Nov 9 2021 at 2:28pm
Craig: Do you mean that it would vindicate people who invested in electric vehicles (perhaps too early given what was known then)?
Craig
Nov 9 2021 at 2:54pm
Well EVs are associated with driverless features because Tesla is currently a leader in both. It doesn’t necessarily mean an ICE car can’t be driverless though. My point though is that to the extent driverless features exist will change the infrastructure that is needed. For instance, let’s say there are 100% driverless cars. Would I ever fly from South FL to Atlanta again? I could just go to sleep and have the car take me up I-75, right? Likewise, the Acela? Who would take it?
Or perhaps its not quite so perfect where one might be comfortable with a ‘driverless’ lane where trucks can be sent down the lane and drivers need to handle first/last mile.
One study I saw said that if all of the vehicles were 100% optimized, clearly a theory, that about 80% of the cars in Manhattan wouldn’t need to exist at all. If memory serves correctly this was an MIT study, but combine driverless cars with Uber/Lyft sharing technology and the future could be a driverless future where we have 80% less cars than we have today. We theoretically might wind up needing far less infrastructure as the case may be and these technologies are theoretically right over the horizon.
Phil H
Nov 10 2021 at 7:15pm
I don’t see why driverless technology would make people give up their own cars. People are still going to want to have swankier cars than their neighbours, and keep their own stuff in their car, as they do now. Driverless will be great, but I don’t see it massively reducing car ownership.
Craig
Nov 11 2021 at 7:36pm
Well to this day there are people into the equestrian culture and thoroughbreds, etc and of course horses have fallen out of favor. So for sure conspicuous consumption in automobiles will continue. The question of course is one of degree.
Look at it this way. Pre-pandemic, many families had two cars, they’d drive them to and from work, to the soccer game for junior, to Publix, etc. And yes, they would spend more on these cars than the necessity of basic transportation would necessitate.
BUT, at the end of the day it was a massive expenditure in machines that ultimately spend the vast majority of every single day completely out of use. Just sitting around in your driveway or in the parking lot at work.
Today the average NYC taxi does 70k miles per year. Right now of course there’s a person in them and they need to get paid. Take the person out and suddenly it has occurred to some that a driverless fleet could achieve much, much higher utilization rates making the cost per mile plummet to where you might very well just rely on that versus having 2 cars sitting in your driveway waiting around for ‘just you’
This might take time to filter through of course, perhaps families will opt for one really, really nice car and rely on shareservices as ‘car 2’
We’ll see what happens, I mean, obviously threshold one is that driverless technology becomes available and ubiquitous. Right now its still in its infancy.
David Seltzer
Nov 9 2021 at 12:36pm
“And we have no credible theory showing that political and bureaucratic processes will choose the most productive investments.” I believe we have credible counterfactual examples in the form of Amtrak, electric chargers, USPS, etc…etc… . Using the definition of theory as being formulated to explain or predict an outcome, it seems reasonable to predict inefficient outcomes when governments, paraphrasing Hayek, direct resources from wood to steel, thereby favoring the buyer of automobiles at the expense of book buyers.
Pierre Lemieux
Nov 9 2021 at 2:23pm
David: I wanted to make a milder statement that could not fall prey to some rare exceptions. (For example one could claim that government subsidization of solar energy brought some technological breakthroughs, albeit perhaps non-economical at this time.) I said something analogous to ” we have no credible theory showing that cows can fly.” Logically, this does not contradict your stronger claim that “we have credible theories that cows cannot fly.” (But perhaps, somebody has heard from his cousin that a friend of his made a cow fly by attaching a big helium balloon to it.)
nobody.really
Nov 12 2021 at 4:19pm
No–but I suspect people have heard of Gov. Gavin Newsom and Elon Musk. To compete with Wisconsin dairy, California paid SpaceX to have three cows sent into low-Earth orbit in a Falcon Heavy; the cows would then be used in a publicity campaign. Alas, once they returned, the cows could no longer stand, so the whole campaign was scrapped and the cows went to the slaughterhouse.
Newsom has tried to avoid adding any publicity to this boondoggle–but nothing he could do would completely kill the story of the man who fired the herd shot ’round the world.
Pierre Lemieux
Nov 14 2021 at 3:51pm
Ha! ha!
Thomas Lee Hutcheson
Nov 10 2021 at 6:03am
Ideally we should not be debating what “infrastructure” is. What we should debate is how to chose and finance public investments with net present values greater than (or equal to zero), where the are “public” because private investors would have difficulty (even with proper pigou taxes and subsidies) capturing enough of the returns carry out the investment.
Jon Murphy
Nov 10 2021 at 2:46pm
That seems like question begging, no? One must decide on what is investment before one can decide on how to fund investment.
Pierre Lemieux
Nov 10 2021 at 4:27pm
Jon: You are right. If I want to force X and Y to invest in infrastructure or even only to persuade them to vote to force Z to pay for it, I should know the meaning of the two terms.
Thomas Lee Hutcheson
Nov 11 2021 at 8:12am
But “what is investment,” is a potentially interesting question. What is “infrastructure,” is not. But you may substitute “expenditures with present costs and future benefits” for “investment” if you prefer.
Pierre Lemieux
Nov 14 2021 at 3:48pm
Assuming that those who will reap the future benefits pay the present costs (or get the capital as a voluntary gift). If somebody forces me to pay the present cost in order for him to reap the future benefits, it looks more like slavery.
Jose Pablo
Nov 10 2021 at 10:23pm
Cost-benefit analysis done by an agent with no skin in the game (like, for example, a government bureaucrat) are a scam. Just a way to rationalize what he/she (or their boss) has already decided that “make sense” to do. It is beautifully explained in one of Pierre’s favorites quotes of De Jasay:
The two statements “the state found that increasing group P’s utility and decreasing that of group R would result in a net increase of utility,” and “the state chose to favor group P over group R” are descriptions of the same reality.”
“Net Present Values” are a great example of “teaching birds to fly”: from observing market prices you reverse engineer a discount rate and armed with it you show investors why they should be doing an investment they are not willing to do.
“Private investors would have difficulty (even with proper Pigou taxes and subsidies) capturing enough of the returns carry out the investment.”
This is almost always the case. How to “monetize” part of the huge utility that using your browser, or your WhatsApp brings you, is far from trivial. The Amazon idea of letting you buy stuff online and sending it to your home was a money losing idea for a very long time (still is in many parts of the world), you can argue that this amazing service has survived only because it was “packaged together” with AWS (a money-making machine).
No reason for the government to intervene. Don’t worry too much, if left alone, investor will figure it out.
Robert Schadler
Nov 10 2021 at 4:40pm
Good piece.
Two general concerns about the understanding of two key words: public & infrastructure.
For the Brits, public means open or available to the general public, as in public schools (with very expensive tuition) and public house (aka pub), where anyone can enter but has to pay up for whatever is ordered.
In the US, public has come to mean government funded and usually owned, as in public schools in the US where taxpayers pay for them.
Infrastructure, at least in the last year, has come to mean almost anything and everything that Biden & Co wants the US government to fund.
The concept of infrastructure is a good one: those material and cultural elements that bring people together and facilitate their common efforts. A road, the internet, common language and work habits and countless other things. And almost all of which can exist or be created without government funding. There is a great deal about infrastructure that one can favor w/o an implication that the government should pay for it.
Pierre Lemieux
Nov 11 2021 at 10:29am
Robert: You are totally right about the ambiguities of “public.” In this post I used it mainly and a bit carelessly to mean “government,” except in the expression “public goods.” Further confirming your point, the concept of “public good” (at least in its standard, Samuelsonian sense) does not imply government production or even financing. It is important to distinguish “public” as “government” and “public” as accessible to, or benefiting, the general public.
Jose Pablo
Nov 10 2021 at 5:01pm
“public infrastructure is that it is an investment that will yield a high rate of return in terms of future economic growth.”
If it yields a high rate of return, why should it be “public”? Are not private investors much better at assessing “rates of return”?
Or perhaps we should say that “infrastructure” is limited to fundamental conditions of social cooperation such as abstract institutions (like the rule of law) that favor exchange and individual liberty?
That is a very interesting and highly beneficial definition of “infrastructure”!! … just don’t expect governments or politicians adopting it any time soon.
Comments are closed.