Hayek on non-market derivations of subjective value

This morning, Don Boudreaux cites Hayek:

"Economic changes, in other words, usually affect only the fringe, the “margin,” of our needs. There are many things which are more important than anything which economic gains or losses are likely to affect, which for us stand high above the amenities and even above many of the necessities of life which are affected by the economic ups and downs. Compared to them, the ‘filthy lucre,’ the question whether we are economically somewhat worse or better off, seems of little importance."

This seems relevant to the point I was making about the fact that we do just fine allocating resources and deriving enjoyment without the price mechanism in all sorts of situations.

It reminds me also of something that John Adams wrote to Abigail in the midst of the Revolution:

"I must study politics and war, that our sons may have liberty to study mathematics and philosophy. Our sons ought to study mathematics and philosophy, geography, natural history and naval architecture, navigation, commerce and agriculture in order to give their children a right to study painting, poetry, music, architecture, statuary, tapestry and porcelain."

Perhaps the reason why it reminds me of that isn't obvious. The point for me is that the price mechanism (just like the military and the state) is a way to derive happiness in only very specific circumstances. Some of the most enjoyable pursuits of our lives are pursued with non-market, non-coercive means.

Subjective Value and Allocation

In the comment section of another post, Mattheus von Guttenberg writes:

"As far as the subjective decisions made in government, the word you're looking for is arbitrary, not subjective. Government projects do not rely on profit and loss so they have no way to really KNOW which project is profitable and which isn't (which is just another way of saying they have no way to know what the consumers want). They always operate in the dark and so any decision they make is an arbitrary stab."

and

"if you were forced into buying oranges, even after weighing the incremental gain you achieve from buying apples, then the group that is using force is expressly declaring that "Your purchase of oranges meets some higher standard than your purchase of apples, thus my use of force is justified." This is what I mean by objective value. This group assumes that the purchase of oranges meets a higher, more enlightened (or more necessary, etc.) value because it is for whatever purpose they ascribe to it. No longer are you allowed to resort to the subjective decision making process that so well fulfilled you in the earlier episode. The coercive group arrives at the arbitrary (meaning not guided by profit and loss) decision to enforce the purchase of oranges on everybody. It assumes everyone values oranges higher than apples. They are declared objectively more valuable."
.
I think where Mattheus is confused is in conflating three very different things here: (1.) the information that is used as an input to some allocative decision making process, (2.) the allocative decision making process itself, and (3.) the goal of a decision making entity. He talks as if profit maximization and the price mechanism are the same thing, and then later talks about a "subjective decision making process" as if subjective valuation and the price mechanism are the same thing. He muddies everything up, and while this post is going to be somewhat esoteric I wanted to straighten the discussion out a little.

First and foremost, the way that we as modern human beings interact with the material world is through subjective valuation of objects. Our preferences and the utility that we derive from objects determine their value to us. This is the information that we bring to the table whenever we make decisions about allocating objects. If you think about it, though, this is information that we use in a variety of settings outside the market setting. Think of all the material objects that are allocated throughout your life. Now ignore two types of allocation: (1.) market allocation, and (2.) allocation by the state. So for now, ignore buying food from the store, but think about allocation of who prepares food in your household, who eats it, how much food you give away. Ignore for a moment buying a book from a bookstore, but think about choices you make about what book to read tonight, what book to lend to a friend, what book to donate, or what book to give as a gift. Ignore the Social Security checks that your grandmother gets, but think about the money she gives in a card on Christmas or the poker game that you bet that cash in.

Mattheus speaks as if the market process and the price mechanism are somehow definitive of subjective valuation, or that if you don't use the market process you don't use subjective valuation. But if you go through this thought experiment, momentarily become a doctrinaire materialist that sees nothing but material objects and their allocation, and ignore all the market and government processes, there's still a lot of allocation going on. In Mattheus's world we'd be lost without prices and the market, but somehow we seem to manage, don't we? And everyone knows that their subjective valuation informs these decisions too. My wife and I are moving to a new apartment soon, so we're donating a few things to downsize our stuff, and we're packing other things up. What I do and don't donate depends on how I subjectively value what I own. How carefully I pack things up is a function of both how durable the object is and how much I subjectively value the integrity of that object. These are all allocative decisions I'm making quite successfully without the market.

So what is the market? What's great about the market and prices is that they provide a mechanism for coordinating lots of people's subjective valuations. When you consider how to spend your "marginal" dollar - your next dollar - you consider the subjective value of all the potential marginal purchases you could make. A rational agent is expected to purchase whatever object has the highest marginal value (as long as your marginal valuation of the money itself doesn't exceed the marginal value of that object). If the marginal value of that object continues to be higher than the marginal value of your money, you're going to keep purchasing the object in greater quantity until it isn't anymore. Prices will adjust with supply and demand - so where the price finally reaches equilibrium we know that everybody that purchases that object places the same subjective marginal valuation on it, and that they have had the opportunity to purchase their highest subjectively valued material object that they can afford (that's available for sale at least). This ensures that objects are efficiently allocated, and that is the primary advantage of the price mechanism. We know that no one who places a higher marginal subjective value on an object is left unsatisfied, and we assume that whoever doesn't end up having the object allocated to them assigns a lower marginal subjective valuation to it. That's a good thing, because presumably maximizing total subjective valuation is something we like to have happen.

So the price mechanism is one process that uses the information provided by billions of individual subjective valuations, and it uses that information efficiently. But we know from earlier that it's not the only way to use subjective valuation information. Subjective valuation is not a decision making process, as Mattheus asserts. It is the informational input to several different decision making processes.

What about government? Does government use something different? Of course not. All valuation is subjective valuation. In the modern world at least, it doesn't make sense to go around thinking of things as having intrinsic, objective value. Nobody thinks something like "this chair is objectively worth twice as much as this desk". We may say "I find this chair to be twice as valuable as this desk", but the previous, objective valuation doesn't make sense and even governments that make value judgements don't assume any objective or intrinsic qualities about their value claims, do they? The government uses subjective valuation like every other person or institution. The difference is that it doesn't use the price mechanism to process that subjective valuation information and make these decisions. It uses a combination of electoral, political, and bureaucratic decision making mechanisms instead. Instead of the price, the government uses the median voter, or the majority or super-majority in a representative legislature, or the discretion of a bureaucrat to make allocative decisions. All of these mechanisms still use subjective value, of course. In a bad government that subjective value is the value of the leaders themselves. In a bad government electoral decision making doesn't mean much (if there are elections at all), political decision making is about rent seeking and personal gain, and bureaucratic decision making is also about rent seeking and personal gain. In a good government, the subjective valuation of the citizenry is informing each of these processes. In a good government, bureaucracies do careful studies and cost-benefit analyses of different options (or they contract the work out to organizations like the Urban Institute!). The reality of courses is going to be a mix of the two: bad and good governance practices. What actually happens in this allocative decision making process is the subject of Public Choice Theory in economics.

What are the consequences of the government not using the price mechanism? Well, it ultimately depends on what choices they're making. If they're deciding on how much steel should be produced and who should use it, eschewing the price mechanism can seriously impair decision making. We know that the price mechanism efficiently allocates resources to those who value it most. Anything other than the price mechanism is going to make mistakes. This is the old socialist calculation debate and economists are virtually unanimous on that debate. What if we're talking about a different decision - say, taking care of abused children or feeding the homeless? It's not clear how the price mechanism is going to efficiently handle these allocation decisions, right? The homeless probably place a higher marginal subjective valuation on food than I do, but their mental health condition, their ability to pay, etc. is going to be an obstacle to relying on the market to supply that food. We could rely on altruists to use the price mechanism to feed the homeless, but why would the subjective valuation of the altruist be expected to match up the subjective valuations of the homeless? A good rule of thumb is that if outside of government the market is not allocating a particular good, but some other institution like a church or a community organization is, then the government is probably not disadvantaged relative to the market in making that allocative decision. If the market were especially efficient at it, you would see more of those decisions made by the market.

The important take away point for Mattheus is that how we value things and what we do to use that valuation in decision making are two very different things.

Finally, what about profits? Profit maximization is simply a goal of firms, just like utility maximization is a goal for individuals. It's the impetus for action, in a way. Profit is just the excess of revenue over costs. You can be a state owned enterprise in a communist country and still be a profit maximizer. No one says those costs or that revenue has to be determined by market prices. You could also not care about profits at all - you could be a non-profit - and it has nothing at all to do with your participation in the price mechanism and the market economy. Non-profit organizations use the price mechanism when they purchase labor and equipment and when they sell things as well. They just don't care about profits! The Urban Institute pays me a competitive wage to keep me there, it buys paper and computers and rents the building. And we sell our services to the government and private foundations. Everything is the same as it would be for a for-profit contractor, except that all we care about is staying solvent! And we have for forty years now, and our product is in high demand in the market. Mattheus writes that since the government doesn't seek profits it doesn't know what projects are profitable. But this confuses goals. I think it's an open question whether simply being a profit seeker gives you any insight into what is profitable. But what Mattheus leaves unclear is what's so great about profits. My family doesn't seek profits. We don't try to maximize our income less our costs. What we try to maximize is our utility. There's nothing special about profit seeking. Government doesn't maximize profit either - it tries to maximize some sort of social utility, probably with a few prioritarian concerns mixed in. So?

Once again, Mattheus is conflating (1.) goals (like profits) with (2.) information (like subjective valuations), and with (3.) decision making processes (like prices or elections). These three aren't the same thing. We all basically use the same information - subjective valuation. But we use that information to pursue different goals with different allocative strategies.

We've been managing to allocate material objects for millenia - sometimes with the price mechanism, sometimes without. I think the price mechanism has a distinct advantage over all other allocative processes when it comes to certain kinds of decisions, but that doesn't mean the price mechanism is the only game in town.

So where does this leave us with the state, which was in a lot of ways the subject of the discussion in the comment section. It leaves the state, as with most non-market allocative mechanisms, somewhat ill-defined. This makes sense of course - economics provides the most detailed theories of the market - it hasn't focused as much of its time on the state. I was reading some of William J. Baumol's book Welfare Economics and the Theory of the State last night, and he had a great phrase for this unfortunate condition of our theorizing about the state. His epilogue was a question about the nature of the state: "The Wreck of Welfare Economics?". A lot of times it seems that way - that the state becomes possible where the market economy wrecks and breaks down. Advocates that don't think hard enough and critics who are more interested in pot-shots than critical analysis assume that means that whenever there is an externality or market failure, we're arguing that state action is always the right solution (for an example of this, see Anonymous's comments in the last internal improvements post and the weird ideas that he inexplicably ascribes to me). Of course barely anyone is making that argument, but that won't stop lazy scholars from characterizing the situation in that way. But it's still somewhat unsatisfactory isn't it? As it stands, the theory of the state largely consists of the wreck of and the inconsistencies that turn up in "normal" economics.

Technological Singularities and Measuring Value

This week I came across this fascinating presentation by Robin Hanson, of the George Mason University Economics Department, on technological singularities and the possibility of experiencing another one by mid-century. He specifically considers the role of nanotech and AI in this singularity.

Hanson highlights four previous technological singularities: (1.) the evolution of brains, (2.) hunting, (3.) farming, and (4.) the industrial revolution. He charts out how estimated global production per capita grew tremendously at each stage, and then plateaued as the technological innovation fully defused and the prospect of exponential growth moderated. Hanson argues that it's possible we might go through another technological singularity - a prospect that seems all the more reasonable when past singularities are charted out on a log-time scale. As readers of this blog know, I'm quite optimistic about our development as a species and the break-throughs we'll have in the far (and not so far) future. Nevertheless, I do have one major reservation about Hanson's presentation, and it primarily has to do with his data.

The obvious criticism is that we only really have good data on global production after the last singularity - the industrial revolution - and for quite a bit of that period, the data is shady at best. It's not until the last half of the twentieth century that you have good data even on the developed world. Now, we have convincing proxies for earlier periods, and the task is made easier by the fact that they simply didn't produce as much back then (agricultural production accounts for the bulk of it, and knowing something about cultivation methods makes that relatively easy to estimate). But the criticism still stands.

But that's not my primary criticism. My primary criticism isn't methodological, it's conceptual. The biggest blind spot in Hanson's data work is that the way we conceive of value production now is not the way it has always been conceived. If we only count commodities that we consider commodities now, we're going to necessarily make earlier periods look more impoverished than they actually were. Value is subjective. We can conclusively demonstrate that the market economy made humanity more materially better off. And that is a very good thing. What's much harder to demonstrate is that it created more value. My guess is it did create more value, but probably less than Hanson's presentation might suggest. I'll let a selection from Keynes's General Theory explain what I mean:

"Ancient Egypt was doubly fortunate, and doubtless owed to this its fabled wealth, in that it possessed two activities, namely, pyramid-building as well as the search for the precious metals, the fruits of which, since they could not serve the needs of man by being consumed, did not stale with abundance. The Middle Ages built cathedrals and sang dirges. Two pyramids, two masses for the dead, are twice as good as one; but not so two railways from London to York. Thus we are so sensible, have schooled ourselves to so close a semblance of prudent financiers, taking careful thought before we add to the "financial" burdens of posterity by building them houses to live in, that we have no such easy escape from the sufferings of unemployment. We have to accept them as an inevitable result of applying to the conduct of the State the maxims which are best calculated to "enrich" an individual by enabling him to pile up claims to enjoyment which he does not intend to exercise at any definite time" (p. 131).

We see the period between the advent of farming and the advent of the industrial revolution as being an impoverished period. But what we really mean is that they were materially impoverished. In many ways, past ages may have been wealthier than us, in the sense that they produced more value, utility, or welfare per capita because of the way that their values were structured. We live in a world of diminishing returns. Some of these diminishing returns are technological and inescapable in any age: you will have to put in increased effort to produce more and more crops on a given plot of land, holding all else constant, no matter what age you live in. But some of these diminishing returns are psychological and self-imposed.

My wife is leaving for Brussels today for a conference. While she's there she's going to be visiting a few palaces and cathedrals. Tourist attractions like these bring in millions every year. How often are such attractions built today? When we build buildings today how often do we think of them as monuments for the ages? How often do we weigh the benefits that they will produce centuries into the future against the costs of building them now? Rarely. Occasionally, but rarely. That strikes me as a necessarily impoverishing attitude. Granted, not everyone in ancient Egypt and Rome made their investment decisions based on the benefits that would accrue to future epochs. Material poverty was so desperate that such decision making was reserved for an elite few: an elite that often relied on slave labor and confiscatory taxation to build their monuments. I'm not yearning for some idyllic past that never existed. But what I am doing is pointing out that the time horizon that we use to think about what is valuable and the sheer imagination that we bring to the question of what is valuable has been, in many ways, severely curtailed, both as a result of what Marx called the "fetishization of commodities", and because of our secularization, which removed one of the primary motivations for building monuments to the ages (the prospect of eternal life). It doesn't mean that the modern market society is bad. I'm a big fan.

What it means is that it will probably have to change if we want to maintain lasting value. At some point, as a species, we need to build for the ages again. And that doesn't mean some wealthy philanthropist looking a century into the future. That means a visionary looking a thousand years into the future, and dreaming about what human society will look like then, and investing in that.

I've gone far afield from Robin Hanson. I think his talk is very good. It lays out the technical interpretation of technological change from an economist's perspective, which I enjoyed. And I appreciate his optimism about the future. I think he has a very characteristically modern understanding of value creation (at least in this presentation) which could probably be augmented. But all in all, it's an excellent piece.

Production vs. Prosperity

I want to explore a little bit of Evan's query:

"Do those who provide goods and services contribute to useful cultivation, or do they perpetuate a practice of commoditization, either of ideas, tools, or resources?"

It's a reasonable question. Many have opined that the GDP statistics we collect are a poor measure of the true "wealth" of society precisely because this summation of values doesn't necessarily correspond to the creation of actual value in society. A homeless shelter may purchase paper plates and spaghetti from the store, and those purchases will show up in GDP statistics - but the couple of bucks that they contribute to GDP will far underestimate the "value" of that purchase to society when the spaghetti is served with a friendly face, an open ear, and a safe atmosphere.

But as Christopher Hitchens has recently pointed out: "There’s also the not-inconsiderable question of capitalism’s ability to decide, if not on the value of a commodity, at least on some sort of price for the damn thing." We need not assume that GDP encompasses all value to admit that there is value to the idea of GDP (and the products that it measures).

If we set up a straw man of modern production as some monstrous, commoditizing, corrosive influence then of course we will be able to vanquish that straw man and banish it to oblivion. But that would be a hollow victory, because that straw man doesn't really represent the "modern economy" as an idea.

Something like GDP is useful if we are realistic about what exactly it is: Gross (i.e. - after subtracting off imports) Domestic (a fundamentally political concept) Product (those things which are produced for sale in a particular jurisdiction). "Production" is a much more humble thing than "culture" or "happiness" or "value". Production includes the plows and tractors that make cultivation possible, as well as the proverbial pick up truck, farm windowsill, and apple pie (perhaps the apples are home-grown, but the flour is most likely ground somewhere else and the baking tin is almost certainly made elsewhere). This is what GDP is - production that is brought forth in the context of a market. And a market is nothing if not a social collective, and therefore absolutely relevant to "culture".

But Berry and Evan have a point that is illustrated in my example of the spaghetti and paper plates used in a homeless shelter. These contribute to GDP, but they only contribute prices - i.e., market values - and not social values. A great deal of social value is produced that is totally omitted from the GDP statistic.

Charles Murray (a real conservative's conservative) used the occasion of his March 11th Irving Kristol Lecture to speak on how a civilization's happiness is grounded in it's cultural institutions, rather than in cold economic calculations alone. On the other side of the aisle the Nobel prize winner responsible for attacking the Clinton administration from it's left flank, Joseph Stiglitz, has been tasked along with Amartya Sen by French President Sarkozy with exploring new ways of measuring growth to capture social happiness.

I think Berry's point is not lost even on those who deal with the modern, competitive market every day. The task is to:

1. Understand how markets both complement and conflict with culture and value, and
2. To be more realistic about exactly what GDP is and what markets can accomplish in the first place, and not penalize them for failing to accomplish something that it was never their task to accomplish.