Pur Autre Vie

I'm not wrong, I'm just an asshole

Friday, December 30, 2011

The Inevitability of Coordination of Economic Activity

Incidentally, I think my previous post on time and money points the way to a fairly deep point about society - the extent to which "nominal" coordination is necessary in order to achieve "real" outcomes, and the trade-offs that arise.

Some libertarians take the view that people should simply be left to their own devices. Coordination will spontaneously emerge thanks to the invisible hand, etc. etc.

In my view, this is unrealistic both as a description of how coordination is achieved and as a concept of how it could be achieved in an ideal world. I think coordination is a central problem of modern society and that it is inherently political. Of course, the government need not be the institution doing the coordinating. But coordination is pervasive, and no matter who does the coordinating, it will favor some and not others.

By way of example, think about a futures market. The idea is that you have defined contracts specifying the commodity, the quantity, the quality, and the time and place of delivery. The seller undertakes to provide the commodity at that time and place, and the buyer undertakes to pay for it. (The contract doesn't specify the price, which is established on an exchange.)

Very well. This can all be arranged without government help (although of course it probably helps to have a general background of legally-protected property rights and contract enforceability).

But who defines the contracts? You might think that you could have infinitely many contracts, varying continuously along all dimensions, but that's not the case. For contracts to be traded anonymously on an exchange, you want liquidity, which means you want to coordinate on a few specified contracts to the exclusion of others. So you have a few grades of each commodity, and a few places of delivery (famously, West Texas Intermediate crude oil is deliverable at Cushing, Oklahoma). As Canadian grain farmers are gaining the ability to sell their grain to anyone, and not just the Canadian Wheat Board, they will have to consider which futures contract makes the most sense for them, since they can't each write a contract that meets their precise needs. I read a Wall Street Journal story indicating that exchanges are competing on the basis of what contracts they support - some support Canadian-dollar-denominated contracts, others don't. (Sorry, no link - it was in the December 23 issue, I think.) Farmers are thinking about how much they care about liquidity vs. how much they care about the quality of the hedge. (A contract deliverable in Chicago is probably a better hedge than a contract deliverable in Los Angeles, even though the grain itself might be delivered to Toronto. I am making this example up, but you get the idea.)

Now, note, no one is forced to use the futures market, as far as I know. It's purely voluntary. And yet I'm guessing that lots of people who don't trade futures pay attention to the prices on the exchanges, and I'm guessing that farmers know how to optimize the grade of grain that they ship to market, based on the relative prices of the different contracts (or maybe just based on the contract specifications generally). That is, a lot of "real" activity goes into conforming economic activity to the "nominal" coordination achieved by the futures market. Another way of putting it is that conforming to the market standard may be voluntary in the legal sense, but it is compelled by economic logic. And therefore a standard that is in no way compulsory ends up affecting behavior, just as daylight saving time affects when people wake up (even people whose jobs do not require them to show up at work at any particular hour).

Anyway, my bigger point is, contracts could be defined well or poorly, but they have to be defined, and people will have different preferences about those definitions. My broader point is that a lot of human activity benefits tremendously from coordination of one kind or another, and I think libertarians are naive to think that all of this can be achieved in a decentralized way. The question facing modern society is not whether or not to have a leviathan, but how to choose (or design), monitor, and control that leviathan for our collective benefit. (Again, the leviathan need not be the government - it could be the clearinghouse, or the CGPM, or whatever.)

But so, these decisions implicate people's interests, and so they are political, and so we probably need political institutions to deal with them. Often, the government is well-situated to play this role. If someone is going to crush your liberty, it might as well be someone subject to public scrutiny and competitive elections, and bound by a written Constitution and norms of fairness and democracy. But, you know, not always. We have other good institutions with roles to play - the market, academia, etc. And so you end up with the "ideological division of labor" among the institutions at our disposal. I think some libertarians imagine they can avoid playing this game, but they can't. Society will be coordinated. But how and by whom, and in whose interest?

[Update: Another way of making this point is to think about whether socially-imposed conventions such as time zones, currencies, language, standardized weights, etc. make the economic calculation problem easier or harder. If the answer is "easier," then perhaps the economic calculation problem is not the all-purpose libertarian argument that it might seem to be on first glance.]

Some Ramblings on Time and Money

Tyler Cowen at Marginal Revolution links to a piece in Wired about a new proposed calendar and timekeeping convention (he also links to a Cato post promoting the scheme). I will ignore the proposed calendar and focus on the timekeeping convention.

The basic idea is to force everyone to use Greenwich Mean Time, with no adjustment for daylight saving time (which would be abolished). So if it is 11 a.m. on Sunday in London, it is 11 a.m. on Sunday everywhere. (There would be no international date line, as the entire world would simultaneously pass from one day/date to the next.)

Obviously, people would not be compelled to keep their current daily schedules (e.g., getting up at 7 a.m. and going to sleep at 11 p.m.). Rather, it is anticipated that many people would wake up in the "astronomical" morning and go to sleep at "astronomical" night, regardless of the nominal time. (So for instance, a New Yorker might wake up at noon and go to sleep at 4 a.m., as this would correspond to 7 a.m./11 p.m. as currently denominated.)

What I find fascinating about this suggestion is something that its proponents seem to find uninteresting: the question of how "nominal" time may or may not have "real" consequences. Here's one of the proponents, Richard Henry, describing how he responded when a child told him she didn't like his calendar scheme:

And I said, ‘Why?’ She said, ‘My birthday is always going to be on a Thursday.’ I said, ‘You’re free to celebrate when you want! What the devil difference does it make what it says on the calendar?’
All right, so, one might ask Henry why he cares about timekeeping conventions at all. What the devil difference does it make what it says on the clock? I have a digital scale in the kitchen, and by pressing a button I can switch back and forth between metric and imperial units. With digital timekeeping, one could in principle do the same thing. And then it is very difficult to see why the "nominal time" would matter at all. (Already it is easy to find websites to convert between time zones, and in fact in some places there are multiple clocks on the wall for this purpose - some watches even have this feature.)

The answer, presumably, is coordination. The shift to time zones apparently took place because trains kept slamming into each other (or at least that's what I remember reading in Nature's Metropolis by William Cronon). (Previously, a local official computed noon the old-fashioned way and set a clock accordingly - therefore, time varied continuously, not discretely, with longitude. Or more precisely, there were a lot more discrete time zones, so that it verged on continuity. In jest, I once imagined exactly such a system, not realizing it was the status quo before the 19th century. Anyway you can see how difficult it would be to formulate train schedules in such an environment.)

So then to me, the interesting discussion is how much coordination matters - how much do nominal changes in timekeeping conventions affect real behavior? The answer seems to be "a lot," but it would be fun to look into this more closely.

But so anyway, it seems to me that these guys want it both ways. On the one hand, they claim that with their timekeeping convention Russia could coordinate banking hours across the country. On the other hand, if you don't like any of the perceived consequences of the system, then those consequences are treated as merely nominal changes that will not compel any change in behavior. Celebrate your birthday anytime you want, regardless of what the calendar says! What the devil difference does it make?

One could just as easily say, "What the devil difference does it make to banking hours in Russia? Just dictate their hours in Moscow time!" That is, if you want all the banks to be open at once, you hardly need to reformulate your entire conception of time in order to make it happen. In Soviet Russia, clock adjusts you for daylight saving time!

All right, so, all of this is by way of setting up an interesting analogy to monetary policy, a metaphor famously drawn by Milton Friedman (as related by Paul Krugman in the New York Times Magazine):

Back in 1953, Milton Friedman offered an analogy: daylight saving time. It makes a lot of sense for businesses to open later during the winter months, yet it’s hard for any individual business to change its hours: if you operate from 10 to 6 when everyone else is operating 9 to 5, you’ll be out of sync. By requiring that everyone shift clocks back in the fall and forward in the spring, daylight saving time obviates this coordination problem. Similarly, Friedman argued, adjusting your currency’s value solves the coordination problem when wages and prices are out of line, sidestepping the unwillingness of workers to be the first to take pay cuts.
This is pretty fascinating, as it again touches on the question of how nominal values can affect real outcomes. I want to think about it some more.

Anyway, I was a little surprised to see the abolish-time-zone scheme promoted on the Cato website - are time zones and daylight saving time now seen as socialist endeavors? Certainly I think there is a strong conservative distrust of any claim that nominal changes can have real-world consequences. I suspect that there is a psychological thread that connects the gold-bug obsessions of the modern GOP, sentiment against daylight saving time, and ultimately the desire to abolish time zones entirely. If the government can't meddle with time or money, we will be on our way to true freedom! Or at least off the road to serfdom.