Friday, April 12, 2013

Consumption isn't a fleeting burst of pleasure, it's a long-lived asset

I've learnt enough about the terms income, savings, consumption, investment, capital, and other major macroeconomic categorizations to pass a basic economics exam. But I've never been a big fan of the style of thinking these terms force on me. Am I just being lazy? I'll lay out my points and give an alternative.

Squarely Rooted recently commented on the somewhat arbitrary nature of the boundary economists set between consumption and saving, opining that travel should be thought of as investment, not consumption. On twinkies as savings, here is Squarely Rooted from an earlier post:
Think of a Twinkie. Twinkies are an odd product; on the one hand, they are a cheap, delicious, unhealthy snack; on the other hand, they are (at least according to legend) practically immortal. So is buying a Twinkie consumption or saving? Does it depend when you eat it? And for those who will say “but Twinkies aren’t an investment, they bear no interest, they just sit there” – so does money under the mattress, and nobody thinks that isn’t saving.
As Squarely Rooted notes, dividing the world into categories allows for discussion, but it also "pre-digests" the world for us. This is what Nick Rowe once called the Borges Problem:
We get very different results depending on how we categorise the world. And sometimes the categories we use are chosen by someone long ago who had a totally different purpose and/or a totally different theory to ours. Our way of seeing the world gets distorted by the dead hand of historical ways of seeing.
This cuts both ways. While the dominant technique of linguistically dividing up the world may distort our way of seeing things, having multiple languages can be equally problematic. Here is Steve Randy Waldman:
Our various allegiances — to schools or tribes or policy ideas — exploit the ambiguity of language to manufacture conflicts, through which we reassure ourselves that we are right and they are wrong. (And no, math doesn’t help much, because we must map it arbitrarily to the same ambiguous language for it to be of any use.) Now I will reassure myself that I am right and they are wrong.
It seems to me that looking at the world through a set of different categorical lenses can give us some great insights, as long as those lenses are coherent. But we need to be aware that there are many different taxonomies. Learning how to recognize each and being able to translate between them will probably save us eons in lost time arguing about semantics.

Back to the basic set of categories under discussion. The typical view is that income is a flow, part of which gets apportioned to consumption. Consumption spending is immediately used up, providing a sudden burst of consumptive joy before disappearing for eternity. The unspent income that remains is defined as savings, and this in turn will be invested with an eye to the future, primarily to fund consumption down the road. Most savings will go into capital, though some of it might leak into money hoardings.  Equipped with these categorizations, economists can go out into the real world and determine what falls in one basket and what falls in the other.

The line being drawn between consumption and savings is based on the distinctions between now vs. later and durable vs. nondurable. Let's try a different way of splitting up the world. Suppose that all consumption goods and experiences are long-lived durable assets. They are forms of income-yielding capital in which one invests. Canned beans bought for my pantry will provide a burst of consumptive joy several months from now. Until then, just having them in a pantry provides a stream of useful services, much like a fire alarm provides utility even though it is never used. These are what Steve Horwitz calls availability services. Both the food we keep in our larders and our fire alarm quell uncertainty and soothe us.

If I take out my employees to the bowling alley, I'm investing in organizational capital. But when I go out to bowl by myself, I'm drawing down my wealth on a one-time shot of consumptive joy, at least according to way the lines are currently drawn. Why not consider both to be long term investments?

Take travel spending. As Squarely Rooted points out, travel is a perpetuity. The travel asset that you might be considering purchasing provides an immediate burst of raw travel experience (not all of it fun), followed by a perpetual flow of memories, experiences, and knowledge that continues till death. Any one who swaps out a security, say Microsoft, from their portfolio for a travel asset has determined that on the margin, the present value of the flow of dividends provided by a voyage exceeds the present value of a flow of Microsoft dividends.

I recently splurged on an expensive meal and did so not only to enjoy the near-term burst of flavours, but also for the long-term flow of returns that my investment would produce, namely the opportunity to remember my experience and talk about it with others. Until I forget my experience a few years from now, it'll have provided me with repeating chain of returns. On the other hand, I'll probably sell out of the gold futures contract I just bought in a month or two. Which of these two swaps is the future-oriented durable one and which is about the here & now?

When someone spends their income on so-called consumption they aren't drawing down their stock of savings. Rather, they're swapping asset x in their portfolio for asset z.  The choice to buy food, travel, and get a haircut isn't a depletion or exhaustion of wealth—it's a portfolio adjustment. Consumption is a stock, not a flow. We can calculate the discounted value of all yields thrown off by a consumption good or experience over time and sum these flows up into a stock value.

We are always conducting asset swaps in order to grope towards portfolios with the highest net present value. Our personal capital is probably our greatest asset. When we earn income, all we've really done is swapped personal capital, time & effort, for a bank-issued liability. We commit to this swap because we estimate that the NPV of additional bank-issued liabilities will more than offset the lost NPV of personal capital.

In equilibrium, the returns on all assets are equilibrated through arbitrage. Investing $10 in cigarettes should provide the same prospective flow of services as investing $10 in a trip to somewhere, $10 on a massage, or $10 in Microsoft. If the yield on some consumptive asset, say a massage, exceeds the economy wide rate of return, individuals will sell their Microsoft shares and go off to the masseuse. This process continues until the price of massages has increased to the point that it no longer makes sense to conduct this arbitrage.

What about the classical bias against so-called consumption? Say that rather than swapping out bank-issued liabilities for Microsoft shares we swap them for a long-lived travel asset. Economists would say that we have high time preference and are sacrificing future consumption for present. Society says we've splurged on silly consumption. It seems to me that as long as we've made this decision by appraising each asset's future earnings stream, who cares if we've chosen Microsoft or travel? All we've done is clocked the two NPVs against each other and purchased the one with the best yield relative to its cost.

Microsoft shares have one advantage over a travel asset. They are liquid. Travel experience can't be resold. In committing your entire portfolio to travel, the problem isn't that you've sacrificed future consumption for present consumption, nor that you've sent money down a black hole. Rather, you've rendered your portfolio less liquid. People who reallocate their portfolios towards travel are asset rich, but liquidity-poor.

While we can't directly remonetize our travel asset, we can indirectly remonetize it. If travel and good food improve our spirits and productivity, then we can recombine these benefits with our labour and resell the total product at a higher price than before. So investing in consumption assets can be a great idea. But in general, it's probably not a good idea to invest everything one owns in illiquid consumption assets. Liquid assets will always be good in a bind.

Getting back to the Borges problem, how does reconceptualizing things this way lead to different results? The old lines between capital, land, and labour aren't so important, nor is the axis between the household and firm. Nor does the distinction between durable and non-durable goods concern us. All we have are millions of yield-generating assets that are constantly moving in or out of individual's portfolios. The main difference between these assets is their swappability, or their liquidity. It's a good platform on which to start thinking about moneyness.

Even if you don't agree with me on any of this, I hope you see how the Borges problem operates. How we choose to linguistically parcel up the world influences the way we take in and sort data, and the data we generate is the base for our actions, policies, and institutions. We've built up an incredibly large edifice based on our initial categorizations. Hopefully we've gotten them right.


  1. Macroeconomists need to discover the Irving Fisher/Jack Hirshleifer view of borrowing, lending, saving, dis-saving, investing, and dis-investing. It's all explained with indifference curves, budget lines, and production-possibilities curves, and it has almost nothing in common with the over-simplified definitions offered in macroeconomic texts.

    1. Care to elaborate more? I don't know anything about Hirshleifer's views. Do they involve treating consumption as a long-lived yield producing asset?

  2. Since I can't draw a picture here, it's hard to get the idea across. There are probably lots of old, cheap copies of Jack's textbook "Price Theory and Applications" on the market, and the chapter on Intertemporal Choice explains it pretty well. There are also a few good links that come up by googling 'intertemporal choice'. Every intermediate micro book spends a few pages on it.

    Anyway, consumption can be a 1-time thing (strawberries) or a long-lived thing (light bulbs). Draw a regular-looking indifference curve diagram, but instead of putting goods x and y on the axes, put "this year's corn" on one axis and "next year's corn" on the other. If you want more years, just keep adding axes. Then draw a budget line along with indifference curves, and you're on your way (assuming one knows what budget lines and indifference curves are). If you want to incorporate productive activity between time periods (aka investment and disinvestment) then you need to draw in a production possibilities curve PPC. The individual's productive activity consists of finding the point on his PPC that reaches the highest budget constraint (a tangency of the PPC and the budget constraint), and then finding the point on that budget line that reaches the highest indifference curve (a tangency of the budget line with an indifference curve). Once the diagram is drawn, various distances in the graph give precise meanings to the otherwise vague concepts of borrowing, lending, saving, dis-saving, investment, and disinvestment.

    1. Here's a picture of the Irving Fisher diagram:

      From halfway down this post:

      Do macroeconomists need to discover it? Or rediscover it? I agree that one simple picture says more about consumption, saving, and investment than....a lot of things. The only thing wrong with it is that it only has 2 dimensions!

    2. Just add a third axis for "period 3 goods", a fourth axis for "period 4 goods", etc. Problem solved!

    3. How do I think about complex goods like houses that provide repeating bursts of consumption over long periods of time? I noticed that Hirshleifer uses corn, and Nick you use apples. Once these things are eaten they are gone. But most of my consumption goods in my house stick around forever... books, sofa, etc. Buying these things now gives me C today and C tommorrow... how do I unpack this using a Fisher diagram?

    4. The key phrase is "dated consumption claims". A house provides shelter this year, shelter next year, etc, and those dated consumption claims can be broken down to days, hours, etc. (Then we get to more complicated things like uncertain consumption claims, and state-dependent consumption claims.)

      A house or a lightbulb or an acre of land all give streams of dated consumption claims. Since the land lasts forever, those claims are valued like a perpetuity. So a house that yields $50 of pleasure per year will be valued by the consumer at 50/.05=$1000 as a purchase price.

    5. That's a cool term.

      What I think I was trying to say in this post is that spending one's income on so-called consumption doesn't result in an irreversible drawdown of wealth. It just means that you've swapped cash (and related "dated cash claims") for a whole portfolio of what you call "dated consumption claims". It's just a portfolio realignment. As you point out, a lightbulb provides a stream of claims that can extend far into the future, so when someone purchases the particular set of dated consumption claims associated with a lightbulb, they aren't necessarily making a decision to consume now at the expense of future consumption.

      That being said, there is an element of irreversability since its tough to resell lightbulb claims. It's fun to think about a sci-fi world in which markets for all sorts of dated consumption claims existed so we could easily liquidate already-acquired consumption goods. I suppose that something like this already exists for housing?

      Nick, when you talk about Walrasian/Arrow Debreu General Equilibrium theory here, "lot's of time periods, lot's of different goods", dated consumption claims are what you're referring to?

    6. If you buy the house for $1000, expecting to get $50/year of shelter in perpetuity, then your wealth is unaffected. But as you live in the house you are giving up the $50/year that you could have rented it for, so it's the consumption of the house that draws down wealth, but the purchase of the house leaves wealth unaffected.

  3. JP: "Nick, when you talk about Walrasian/Arrow Debreu General Equilibrium theory here, "lot's of time periods, lot's of different goods", dated consumption claims are what you're referring to?"

    Yes. Or, at least, that's one way of looking at it. And we could make it "state-contingent dated consumption claims" too.

    Ownership of a house is ownership of a bundle of such claims. A house would be like a square PPF, for n time dimensions (where the house lasts for n periods), with dated 'somewhere to stay warm and dry in period t' on the axis.

    And you are quite right that almost every 'consumption' good is an investment good too, because it lasts for more than one instant.

    Mike: "Just add a third axis for "period 3 goods", a fourth axis for "period 4 goods", etc. Problem solved!"

    Yep. And double the 4 dimensions to 8, for both apples and bananas. Etc.

    1. Right, so the choice to buy a consumption good is different from the decision to consume. Consumption is just the using up of dated consumption claims that inhered in a certain consumption good.

      What do you think about this? Highly liquid financial assets also provide a bundle of consumption claims. We know that we can easily sell these assets in an emergency at time + ?. Knowing this soothes us, just like a fire alarm soothes us. When we buy the asset at t we have to price in the discounted flows of dated consumption claims at t+1, t+2... etc.

    2. Correct. It's just like pricing an annuity, plus you have to throw in extra terms if the annuity payment is growing or shrinking, or if there is a convenience yield.

  4. Not only that, but investments usually require consumption to build them, and sometimes fail, but are failed investments anything other than consumption?

  5. This way of looking at "consumption" goods goes all the way back to William Stanley Jevons. He had a very simple diagram that was meant to show both the period of "gestation" and period of "use" that was very similar to the "hayekian triangles" from "Prices and Production" here...

    I always interpreted this as meaning that there is no such thing as consumption spending. There is only investment spending and then "consumption consuming", though the time period may be so short as to be irrelevent, like with fast food. What's important here is that the way Austrians complain about the lack of importance placed on capital structure, there is an equally important "consumption structure" which has been neglected. I keep cash in my portfolio to stay liquid for uncertain circumstances, but others trade cash for flashlights and canned goods because of uncertainty as well.

    I will say, when I first read your articles on consumption as a portfolio good, I was reading it much more psychologically, as though the experience itself were a part of your portfolio, and the "having had" of the experience would yield utility in the future in just as important a way as the experience itself at the moment of the "consumption".

    Would this explain why low earning young people seem to have higher time preference and high earning old people have low time preference, a dearth of utility yielding consumption experiences in the portfolio that has reached a highly diminished yield by the time of old age?

    1. John,

      Thanks for the Jevons tip.

      I like the phrase "consumption structure". It jives with what Mike, Nick and I were talking about above, in which people hold portfolios of "dated consumption claims". And you're entirely right that in the face of uncertainty investors may choose to hold a portfolio filled with dated consumption claims (flashlights, canned goods) and/or "capital" goods (machinery) and/or financial assets (cash/stocks).

      "I was reading it much more psychologically, as though the experience itself were a part of your portfolio, and the "having had" of the experience would yield utility in the future in just as important a way as the experience itself at the moment of the "consumption"."

      My train of thought on this goes something like this... when we talk about consumption we mean goods & services. It's easy to think of goods (flashlights & canned goods) as items in someone's consumption structure that they use up via "consumption consuming". Services is tougher. The only way I can transfer the idea of services into someone's consumption structure is by capitalizing them as a flow of relived memories. Either that or the concept of services is too irregular to be thought of as more than a one time shot of pure consumption. Still not sure, but I do like the idea of capitalizing services. How can we not capitalize experience earned while consuming travel services?

  6. Being a big fan of Borges, I'm smitten by your incipient term "Borges problem" and have added it to my list of favorite terminology ever! (Also in the list, "weakly godlike AI", "childfree", "POTUS to PRIME", "whales humping", "camelCase"...)

    I tried to narrow down the term into 140 characters. It was hard! What do you think? :)

    THE BORGES PROBLEM: categories are CONTINGENT, inescapable (because language), dangerous (blinding), unconscious (historical)--yet DECISIVE.