29 June 2011

What infrastructure in particular?

Time | Fareed Zakaria | Are America's Best Days Behind Us?

The problem with the U.S. government is that its allocation of resources is highly inefficient. We spend vast amounts of money on subsidies for housing, agriculture and health, many of which distort the economy and do little for long-term growth. We spend too little on science, technology, innovation and infrastructure, which will produce growth and jobs in the future.
This article is all over the place, but this passage in particular stood out to me as an example of something I hear a lot which bugs me. Lots of people tell me the key to growth (or one of them, to be generous) is infrastructure. What infrastructure are we talking about here?

I understand in general how infrastructure can create growth. If I live in some one-cow-town in India I'm not going to start making ... I dunno, awesome footstools shaped like integrated circuits .... unless I have some way of receiving parts, and orders, and electric power, and sending products out. So yeah, I would appreciate some more infrastructure to make that possible.  Pretty much any infrastructure improvements in a place like that would probably improve growth.

But we're not that place.  What do you think of when I say "infrastructure"? Bridges and highways and phone lines and ports and things, right? What specifically are we lacking in America that is holding people back from starting and growing businesses?

Something like WiMax? Or driverless cars?  Those would result in plenty of growth opportunities. But if that's what you want, say we need more spending on wide area wireless internet or automated vehicles, not just infrastructure in general.  If you mean we need money for electric vehicle charging, or updated air traffic control, or better electric grids, then say so.  When people try and sell me on "infrastructure" in the abstract, I'm pretty sure we'll end up spending my money repaving sidewalks and giving better pensions to bus drivers and building light rail lines which will never cover their operating expenses to say nothing of capital costs and planning 55mph trains that proponents will call "high speed" between two podunk towns no one goes to in the middle of California.

Are there some some places we could use a wider bridge or an extra highway?  Sure.  But there are also a lot of places where we'd build those things wouldn't cause any growth at all.  I'm not so confident Congress could tell the difference between one and the other.  Infrastructure isn't some silver bullet to spur growth.

"Infrastructure" isn't some monolithic pile of stuff.  We can't throw money at it to make the pile grow.  Infrastructure is a collection of discrete things, some of which may be worthwhile and some of which won't be.


  1. Let's zoom in on the example of a train line between "two podunk towns no one goes to".

    Suppose the citizens of those towns spontaneously got together and decided to build such a rail line by themselves: they dig a mine to get metal, they build refineries to refine it, etc. etc. and finally lay down the tracks and build the train cars. That would be pretty good, right?

    That's pretty unlikely though, in the absence of leaders spurring them to do it.

    We live in a capitalism, and the way for leaders to spur people to do things, is by invoking fiat money. If Uncle Sam pays the two towns to build that rail line, there are only two differences from the imagined, spontaneous scenario: 1. greater efficiency is achieved since the townspeople don't actually have to reinvent the entire wheel (they can import refined metal, etc.), and 2. numbers change places in databases (this is the "money" part).

    It seems you're saying the whole project is no good because of 2. numbers changing places in databases.

  2. Actually I think such a project (like this one between Corcoran and Borden, CA, or this one in Iowa) is no good because there are not enough people willing to pay enough money to use them. They are value-destroying propositions with or without fiat money.

    If the federal government takes my money and used them to pay for the construction of a rail line, the debt service on the capital costs, and subsidize the ongoing operations, and that rail line still never makes money it does not matter what form the money takes. They could be dollars, or gold doubloons, or bitcoins, or shiny plastic beads.

    My problem is that these projects are consuming resources without fulfilling a sufficient amount of consumer desires, not that "numbers change places in databases."

  3. In 2007, the total amount of physical U.S. dollars in circulation was roughly about 700 billion. Then we gave the too big to fails a 700 billion dollar bailout. By your reasoning, the government must have seized just about all the U.S. dollars in the whole world. Funny because I don't remember my own physical dollars being seized. Must just have been lucky.

    Numbers in databases.

  4. I am not claiming fiat money, or fractional reserve banking, or anything else effecting monetary policy is a good or bad idea. I make no claims about what sort of money we should use. I am saying that wasteful infrastructure is wasteful *no matter what type of money we are using.*

    Resources will go into constructing a train between Corcoran and Borden. I am not interested now in how those resources are paid for. Could be by gold, could be by some illusory numbers in databases that people are only fooled into thinking are valuable. It doesn't matter to me for the purposes of this discussion, because no matter what sort of money they are purchased with the resources used for this train can not be used for other purposes, and so can not fulfill other peoples desires for other things. If the amount of desire fulfilled by the train is less than the amount of desire left unfilled by the reallocation of these resources, it is a bad project.