The $1-trillion infrastructure deal between Congressional Democrats and Republicans may at least help us answer the question: What is infrastructure? Economic definitions, when you can find them, are not enlightening. They often provide a mere list of what are supposed to be instances infrastructure without providing a key to their common features.
Criticizing the infrastructure deal, a Wall Street Journal editorial provides a list of its own of “traditional public works” (“A Not So Grand Infrastructure Deal,” July 29, 2021):
The U.S. could use more investment in roads, bridges, cyber-security and ports, as well as for drought, wildfire and flood mitigation.
The same editorial shows how different and wider is the list concocted by Democratic and Republican horse traders. Just a few quotes:
Rail has long been an obsession of President Biden, though he’s lately become fixated with electric cars. He scored $7.5 billion for a “national network” of electric-vehicle charging stations. Even FDR didn’t get a New Deal program to build gas stations.
Energy Secretary Jennifer Granholm is also a big winner. She’ll run a green venture capital fund rivaling Kleiner Perkins with tens of billions to throw around at carbon capture, hydrogen, electric flying taxis, buses and high-speed mass-transit hyperloop. She will also be in charge of creating a “smart” grid with no less than $73 billion for transmission lines and batteries to back up heavily subsidized wind and solar power.
One of the worst parts of the deal is the $65 billion government intrusion into broadband markets. States—i.e., politicians—will get $40 billion to build out broadband in “underserved” areas. The nation’s broadband networks have been built by private companies, which invest tens of billion dollars each year, including $67 billion in a government spectrum auction that Senators plan to use to pay for their deal. …
According to a White House summary, internet providers will have to abide by rules in Mr. Biden’s competition executive order. Could the return of Barack Obama’s net neutrality rules be coming by this back door?
Whatever list one likes, the first question is whether these infrastructures need to be financed by the government. In their article for the entry “Public Infrastructure” in the New Palgrave Dictionary of Economics, Teresa Garcia-Milà and Therese J. McGuire define it as
the stock of publicly provided physical capital comprising highways, sewage and sanitation systems, water systems, school buildings, hospitals and so forth.
Other authors limit infrastructure to public capital that produces public goods or services in the technical sense, that is, goods or services that are both non-rival in consumption and excludable, which of course drastically reduces the extension of the concept. Others would insist that network effects are a necessary feature of infrastructure: for example, the deeper the internet penetration, the more useful it is. But Garcia-Milà and McGuire are right to give a more general definition because it corresponds better to what laymen, politicians, and bureaucrats think: infrastructure is simply capital equipment financed by taxpayers and operated or controlled by governments.
Reviewing the economic literature, the two authors conclude that public infrastructure is not very economically productive, except for certain region of industries that benefit from the displacement effect of government intervention. They write:
Based on the aggregate analysis, we can conclude very little. The most credible aggregate production-function estimates of the impact of public infrastructure on private output hover around zero, as do estimates of the net social benefit of public infrastructure investment.
A different approach would be to define infrastructure as any public spending that favors free and mutually beneficial exchange among individuals, like say the commercial fairs in the Middle Ages or, very generally, the peace and order and protection of property rights ideally provided by government. This could be made consistent with James Buchanan’s illuminating phrase (in James M. Buchanan and Richard A. Musgrave, Public Finance and Public Choice: Two Contrasting Visions of the State [Cambridge MA and London UK: MIT Press, 1999], p. 245):
If I observe someone with apples and somebody else with oranges, I don’t want to try to say a particular allocation of oranges and apples in a final position is better than in the other allocation. If I observe them trading without defrauding each other, whatever emerges, emerges, and that is the way I define what is efficient.
But note two caveats. First, promoting exchange by eliminating some transaction costs might be promoting inefficient exchange, trades that provide benefits lower than their costs. Second, an exchange-based definition of infrastructure is not consistent with how governments themselves use the term. A large part of their activities consists in prohibiting acts of exchange that they do not approve of. They regulate or prohibit mutually-determined hiring conditions (with minimum wage laws and trade union privileges); terms of exchange with foreigners; trade in, and consumption of, many goods and services (alcohol, tobacco, foie gras, plastic bags in grocery stores, certain drugs, guns, sex, etc.); a host of mutually agreeable financial transactions; and so on and so forth. Infrastructure is not a simple economic term but rather a concept in the economic analysis of politics or public-choice analysis.
I thus propose the following definition: Infrastructure is whatever the government wants to pay for because it benefits from the expenditure. “Government” is defined as politicians, bureaucrats, and occasionally a majority of voters who approve a bundled package of complex measures that have unknown future consequences and that the ordinary voter is not motivated to study and understand anyway.
This definition has the benefit of explaining what governments do in practice, instead of speculating on what their infrastructure spending would be in the Garden of Eden where everything is good and free.
READER COMMENTS
Thomas Lee Hutcheson
Aug 3 2021 at 7:22am
I’m less interested in defining “infrastructure” than in the rules that guide public investment and ideally the institutional framework for applying those rules. It would also be good to separate the question of collection of user/congestion fees from the issue of public ownership. In general I think we need to shift to more state and local finance of public investment, backed up by an “Infrastructure” Bank that would condition its financing on positive NPV’s.
Jon Murphy
Aug 3 2021 at 8:46am
Could you expand on this point? I’m not sure I understand it given I’m inclined to disagree. If an action taken reduces transaction costs, then on the margin it would create more efficient trades. I don’t understand why inefficient trades would occur if the transaction costs are truly eliminated (as opposed to merely shifted onto the government).
Pierre Lemieux
Aug 3 2021 at 10:37am
Jon: It is Carl Dahlman’s argument that transaction costs are a fact of nature just like transportation costs and that and that pretending otherwise “is much like stating that a world in which apples are costly to produce is inoptimal compared to one in which apples are a free good.” You are right to emphasize the qualification that a “real,” costless elimination of transaction or transportation costs would allow new trades and thus be efficient (disregarding the fact that the value of transportation assets, such as ships and airplanes, would fall to zero, as it should be disregarded). Even a costly elimination of transaction costs is efficient if the resources to achieve this goal are committed voluntarily or if a previous involuntary use of resources (in slavery, for example) is eliminated.
Pierre Lemieux
Aug 3 2021 at 11:01am
P.S.: “Voluntary” remains to be defined…
Joel N Pollen
Aug 3 2021 at 11:30am
This is a great point, Pierre. I think many folks see it as self-evident that more roads and bridges are a good thing, because they make economic activity more efficient. This is true, but another way of putting it is that roads and bridges are subsidies of transactions that are otherwise not worth doing.
Subsidizing minimally productive transactions is not inherently a good idea. To believe that it is requires some further stipulation. Perhaps we think that the transaction is made artificially harder by some other policy, or maybe we think the transaction is inherently valuable regardless of its narrow productivity contribution (as many feel about, say, solar power, or liberal arts education). But in the case of “infrastructure,” these additional conditions are rarely specified. In fact, a great deal of infrastructure has a negative value, because the cost greatly outweighs the small productivity of the additional transactions that it allows. This is true of a great many freeway extensions, road widenings, interchanges, roundabouts, new suburban developments, and light rail lines, which probably constitute the bulk of government infrastructure spending (in the layman’s sense of “infrastructure”).
Jon Murphy
Aug 3 2021 at 12:06pm
Thanks for your comment. Pierre’s reply to me and your follow up made his point much clearer to me. I had misread the post!
robc
Aug 3 2021 at 1:47pm
One big problem with “infrastructure” is that they are often built with federal dollars and then turned over to localities, who are responsible for ongoing maintenance. It is often a bad deal for the local governments – what seems like a good idea, a “free” road or bridge turns out to be a anchor on the budget.
It works great for 5 or 10 years but 20 or 25 years down the road the maintenance costs outweigh the benefits. By that point, the politicians who leveraged the gift into a reelection or promotion are long gone.
Phil H
Aug 3 2021 at 11:18am
“Infrastructure is whatever the government wants to pay for”
Ha. Yes, probably. I’m just in the middle of translating a book about “digital infrastructure” in which the author is keen to label certain kinds of software as infrastructure (including stuff as simple as online survey software). He’s not a government employee, but he needs to please the government right now.
“estimates of the impact of public infrastructure on private output hover around zero”
As ever, my problem with this is: most developed countries seem to have public infrastructure. There are very few that got rich without public investment in big physical stuff. And countries that invested in lots of infrastructure remain middle income, even after collapse and lots of bad governance (I’m thinking of Russia). There seems to be value to it, some kind of positive externality. I don’t know what it is, but empirically, that observation is hard to shake.
Another data point: I live in China, experiencing one of the most dramatic and sustained economic booms in all of history, and they do lots of public infrastructure.
Joel Pollen
Aug 3 2021 at 11:45am
It’s hard to disentangle how much infrastructure spending causes wealth and how much it’s the other way around.
For example, I’m almost certain that the 40 foot wide roads with sidewalks and curbs and storm drains and subterranean sewer lines in the neighborhoods in my area have a negative productive value. They are incredibly expensive to build and have to be maintained frequently and ultimately replaced every 30-50 years. Admittedly, the cost of construction is usually covered by a private developer, but they often receive large subsidies and tax breaks from various levels of government. Moreover, the maintenance and replacement costs nearly always fall on the municipality, which in turns relies on state and federal funds to make road-building ends meet. So in the main these should be thought of as public works.
At virtually all times the traffic on these enormous roads is near zero, and almost no one parks on the street except for the occasional backyard party on the weekend. The storm sewers obviously achieve something, but roadside ditches often achieve the same result at a much lower cost.
People definitely want these things in their neighborhood (or at least, some do) but I think it makes more sense to think of these infrastructural amenities as a luxury good consumed by those who can afford it (with a generous subsidy from government) rather than an economically productive stock of capital.
Chap
Aug 3 2021 at 1:07pm
Joel –
Your assessment is really spot on, but I would push back a little on the street, sidewalk, and sewer subsidies, at least where I am. The costs are real, but they are baked into the final price and the end user loses some affordability. Additionally, these and other requirements are in the zoning and design guidelines as a way to control the final product under the guise of maintaining a minimum standard of quality but rather negatively impacting affordability.
I have seen granite countertops required in new multi-family construction, as if the local government should dictate what surface I chop carrots on.
Storm water controls are another example where commercial development is held to a higher standard than single lot, single-family residential with the perception that it’s too expensive for a homeowner to meet that standard. Commercial, multi-family and large scale residential developments all have to meet that with no subsidy so the costs go to the end user.
Tax breaks for developments are not good, and usually it’s a shell game by the local development authority to tout the value of money coming into their area. This often occurs as other areas where the development may be better suited do not want it for reasons of perception similar to what I have outlined above. If you need tens of millions of dollars to make ends meet in a development, it’s not a good proposition to begin with. Ironically, the original land owner often profits the most because they can sell at a inflated value without the risk.
robc
Aug 3 2021 at 1:57pm
Agreed, I made a similar point up above, but after you.
I am in favor of turning over infrastructure to local neighborhoods, POAs where they exist or small taxing districts otherwise. I wonder how much of that would still be built if it was being paid for directly by the local neighborhood. Roads would narrow for sure, except in really high end neighborhoods, or high status neighborhoods anyway (I know some super high end neighborhoods that are practically [and literally in some cases] gravel roads). Storm sewers might survive, but then again, I think there would be lots more septic tanks and less sewers in the distant suburbs.
Phil H
Aug 4 2021 at 1:23am
@Joel: Yeah… I think I have two things to say to that:
(1) It feels like infrastructure could be productive in a lot of ways we haven’t thought of yet. One of the principle benefits of wide streets, for example, is safety. Wide streets prevent road traffic deaths and space us out more, reducing the spread of contagious diseases. Getting the early death rate down seems to me to be one of the key indicators of development, and boosts GDP because the investment in small children can be recouped.
(2) If some infrastructure is ultimately just a consumption good, it’s not obvious that that matters. If the wide streets are a consumption good, and they are consumed (by house prices going up and people buying and living in those areas), then the investment has still served its purpose.
@robc: right, but that sounds like a worse and poorer world.
Joel Pollen
Aug 4 2021 at 2:19am
Phil, thanks for replying.
To your (1):
I concede that extremely wide roads with expensive sewers and curbs and sidewalks that are all apparently underused could have some hard to measure or unforeseen benefits. But I don’t think that is a strong case for these things being necessary for becoming a rich country. The costs are large and much more certain than the benefits.
More specifically, I am skeptical of your claim that wide streets are safer. My personal experience and experts I have read on the topic suggest that the main traffic effect of ultrawide suburban streets is that people drive faster, because the environment feels more like a freeway and less like a small suburban street. I think most, or maybe all, of the theoretical safety benefit is equilibrated away by people’s willingness to drive faster (I think this is the Peltzman effect). Regardless, the safety or efficiency benefits of 40 foot wide suburban streets that see paltry numbers of cars are dwarfed by the massive expense in constructing them.
This may feel like a cherry-picked example, but I think these principles are broadly applicable to a lot of infrastructure in the US, and probably other places also.
To your (2):
It seems obvious to me. Consumption of luxury goods is correlated with national wealth, but it’s certainly not the cause. Driving a Lamborghini and eating $500 steak dinners doesn’t make you rich. If building all these roads and sewers didn’t cause our wealth, then, like fancy cars and expensive food, we can safely do without them.
Now, if individual homeowners paid the full cost of the infrastructure out their front door, I would have no problem with it, just like I don’t mind when rich people buy Lamborghinis. But they don’t pay the full cost. A great deal of infrastructure spending comes from the state and federal government and not from property taxes. In most American municipalities, the share of property tax devoted to roads and sewers doesn’t even come close to paying for them. If people want to keep their huge roads and sewers and see their property taxes triple, fine. But we don’t want government funds paying for people’s Lamborghinis.
robc
Aug 4 2021 at 8:58am
You are right, it “sounds” like it, but it isn’t. And it is easy to prove.
In that hypothetical world, I have two choices, I can pave my street OR I can buy something else. If I value the something else more than pavement, then the world is better and richer if I buy the something else. If I value the pavement more, then its exactly the same after I spend on pavement.
In the current world, we don’t get the choice. Whether my street is pavement or gravel is decided at some distant (well, across-town) location. In my idea, it still isn’t at the household level, but it is much more like the Dumbar’s number level of choice. A true community decision.
Also, for some reason, private purchasing of road paving services is much cheaper than government road paving services. You would think that buying in bulk would be cheaper, but apparently it isn’t. So in my case, you could get the pavement AND some smaller something else, if you so chose. A richer and better world.
Jens
Aug 4 2021 at 10:44am
Even assuming that “the world” depends directly on “your value”, it is not “the world” of Phil H (or at most it is very, very unlikely). So you don’t talk about the same thing even though you use the same word.
robc
Aug 4 2021 at 4:41pm
If I spend my money, and Phil spends his money, and Jens spends his money, then it is a win-win-win.
The world is the sum of the 3 of us (and possibly 7 billion other people) spending as we want. And while there could be some theoretical negative externalities, I am unaffected by whether you or Phil have a paved or gravel road in your neighborhood, and I doubt you care whether I do either.
Pierre Lemieux
Aug 3 2021 at 3:56pm
Phil: A few rejoinders:
Garcia-Milà and McGuire’s article show that, empirically, that observation is not hard to shake at all!
You have to recall why economic growth has any normative value. The reigning potentates in Saudi Arabia and other Arab producing countries (as well as in Venezuela) have made a lot of infrastructure investments in oil. Who has benefited?
Vast sums in foreign aid were invested in underdeveloped countries’ infrastructure before the 1980s. There were no results except for the reigning rulers. Now compare that to Hong Kong, where there was zero public infrastructure investment.
The history of Chinese emperors for two millennia strongly suggests that government investment in infrastructure has little impact on the incomes of ordinary people. There was a difference after the death of Mao, but this difference is now being undermined.
Phil H
Aug 4 2021 at 1:38am
Er… this once again demonstrates to me what a poor rhetorical strategy rhetorical questions are. To the extent that I know the answer to any of your questions, they all seem to back up what I think. If you have information to the contrary, you’re going to have to share it explicitly and directly. I disagree with you. Asking me pointed questions isn’t going to change that. Only data and solid arguments will.
“Saudi Arabia…other Arab countries (as well as in Venezuela)…infrastructure…oil. Who has benefited?”
The whole world! Certainly most Saudi citizens are rich in global terms. And the whole planet benefited from low-cost oil. (Venezuela has messed it up spectacularly, of course; I don’t defend Chavez or the new government at all.)
“Hong Kong, where there was zero public infrastructure investment…”
That’s just not true. The Hong Kong government builds sewers and lots of other infrastructure. It owns most of the housing!
“The history of Chinese emperors…government investment in infrastructure has little impact on the incomes of ordinary people.”
In the Tang golden age, the emperor built the largest city in the world, and it pulled in riches from all of Asia. (1) Infrastructure > (2) golden age.
“There was a difference after the death of Mao, but this difference is now being undermined”
2020 GDP growth: 6.11%. What’s the argument here?
Pierre Lemieux
Aug 4 2021 at 12:28pm
Phil: On China and what causes economic growth, Scheidel’s book is enlightening. See my review but the book is better: “Let’s Travel That Road Again,” review of Walter Scheidel, Escape from Rome, Regulation 43:1 (Spring 2020), pp. 59-61.
Pierre Lemieux
Aug 4 2021 at 12:42pm
On Hong Kong and development, an interesting Economist article: https://www.economist.com/books-and-arts/2017/10/05/meet-the-invisible-hand-behind-hong-kongs-rise. On what economic growth means, see my post “What Is Economic Growth?”
Phil H
Aug 4 2021 at 1:55am
Oh, and just on this one:
“Vast sums in foreign aid were invested in underdeveloped countries’ infrastructure before the 1980s.”
I don’t know much about this, but I was very interested in the idea in a recent Scott Alexander piece that foreign aid absolutely can work if it’s large-scale and sustained. Related link: https://developingeconomics.org/2018/11/12/historicising-the-aid-debate-south-korea-as-a-successful-aid-recipient/
So, I don’t know. All of the examples you give look very wobbly to me.
Craig
Aug 3 2021 at 12:47pm
“Rail has long been an obsession of President Biden”
Liberals love trains!
Pierre Lemieux
Aug 3 2021 at 3:58pm
Craig: It’s because these false “liberals” (the real ones are classical liberals) have misinterpreted the “long train of abuses and usurpations” in the Declaration of Independence.
Craig
Aug 3 2021 at 8:59pm
Another aspect of infrastructure that currently gets overlooked is that projects now seem to get litigated instead of built. Yes, absolutely, blame the lawyers.
Frank
Aug 3 2021 at 1:34pm
“When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.” “The question is,” said Alice, “whether you can make words mean so many different things.” “The question is,” said Humpty Dumpty, “which is to be master—that’s all.”
Niko Davor
Aug 4 2021 at 11:02pm
Lemieux takes a “both sides” tone and stresses that both parties are to blame. That’s true, both parties are to blame; there’s tons of fault to be found on both sides. However, on spending, the Democrats really are much worse.
Trump did plenty wrong fiscally. He promised not to address social security spending, he failed to make big health care changes, he caved in negotiations with Democrats, he let spending go crazy in response to the pandemic, but outside of the pandemic, he basically kept the spending status quo, and didn’t push any big new spending initiatives. The Democrats aren’t keeping the status quo. In 2021, they have pushed for $6 trillion in completely new spending and are debating even larger spending on single payer health care and student loan forgiveness. This is unprecedented levels of new spending.
Compare this op-ed to Pierre Lemieux’s September, 2020 op-ed criticizing Trump over wasteful pre-pandemic spending:
https://www.econlib.org/one-thing-rationally-ignorant-voters-do-not-know/
Lemieux criticizes the Trump Administration for $600 billion extra spending in 2019 vs Obama’s last year of 2016. Most of this increase is automatic increases in interest payments on a larger debt and social security payments with a growing retired population. That is spending that the government is already committed to. Lemieux’s criticism doesn’t even mention new spending that the Trump Administration specifically increased; I presume because outside the pandemic, which admittedly is a big exemption, the Trump Administration didn’t push for big new spending.
It’s fair to find fault with the Republicans, but the Democrats really are worse, and suggesting both parties are similar on budget issues is unreasonable.
Comments are closed.