Monday, February 23, 2015

The early retirement movement

Many times on Twitter I have revealed that I am a huge fan of Mr. Money Mustache, profiled here and a million other places. For the unfamiliar, MMM is part of a movement ("early retirement extreme," or ERE) that basically says: most people in developed countries are rich by any reasonable standard; we can live a materially abundant, happy life on a tiny portion of what we are actually spending; as a result, it's possible to retire after only a few years of full-time work (depending on your wage, of course). Keynes was right, or at least he should have been: we can do the 15-hour work week but choose not to. A lot of people on the internet hate this movement because it runs counter to some popular political and economic talking points. My assessment of the debate is that MMM is winning handily, but that's not what this post is about.

I was reminded of this recently when Josh Brown tweeted about it ("Not sure how it's a badge of honor to substitute shampoo with baking soda so you can retire at 30. What are we trying to win here?"). I think Brown is missing the point. One need not make such substitutions to live a life that approximates what MMM has pulled off. Most of us are spending far more than necessary on cars, housing, restaurants, cable TV, the latest gadgets, clothes, and so on. We're riding the hedonic treadmill; MMM gets a lot of flack for suggesting that we might be just as happy--and far wealthier--if we step off of it. Learning to distinguish between wants and needs can be extremely valuable, even if you are happy with working until 60 or 70 or 80. But more generally, the MMM people aren't trying to "win" anything. Apparently Brown values the time he spends at work more than he would value more leisure, so clearly ERE isn't relevant for him. That's fine. MMM is just pointing out that some people have a different preference, and the growth of developed economies over the last century or so has allowed those people to get the leisure time they really want.


Price and quality dispersion

Following the ERE movement has made me think a lot about the relationship between price dispersion and quality dispersion. Here's Steve Jobs:

Most things in life, the dynamic range between average and the best is at most 2:1. If you go to New York City and you get an average taxi cab driver versus the best taxi cab driver, you’ll probably get to your destination with the best taxi cab maybe 30 percent faster. And an automobile; what’s the difference between an average and the best? Maybe 20 percent? The best CD player and an average CD player? I don't know, 20 percent? So 2:1 is a big, big dynamic range in most of life.*

This is a concept I think about a lot. I think Jobs is right; but the prices of various goods might suggest otherwise. Call it the Camry Concept. You can buy a car for $200,000, but in automobile functionality terms there is no way it is 10 times as good as a $20,000 Toyota. In my utility function, there is no car worth $200,000; once a car gets me from A to B safely, comfortably, and reasonably quickly, additional improvements to the machine won't be worth a lot of money to me, even though I would find them enjoyable. Watches are even more striking. You can buy a watch for $500,000. It will basically do the same things that a $100 watch does. The extra $499,900 is the price of prestige or some other benefit that is largely orthogonal to the ostensible function of the watch.

I don't judge preferences--all of this price dispersion is fine. I get utility from lots of things that others would find pointless. But one of the things ERE people do is focus relentlessly on functionality. This is not a lifestyle for the type of person who gets a lot of utility from high price/quality ratios. The concept can be generalized some. Does a $75 cable TV package provide fives times better entertainment than getting subscriptions to Netflix and Hulu Plus? Maybe, if you watch sports. Otherwise, probably not. The ERE people ditch the cable.


The "retirement crisis"

The ERE people have basically shown that a large portion of Americans should have no problem being ready for retirement. We can all think of obvious exceptions--things happen. But MMM has shown that a 30-year working life should be plenty of time to accumulate a lot of savings, even on below-median incomes. ERE is the solution to the retirement crisis, the student loan crisis, and probably a bunch of other crises--at least at the individual level. It's the Garett Jones approach to inequality:

So let's start training ourselves and our children to delay gratification, to forego that great sound system on the new car, to eat at home a little more often.

This approach might not be of much help to policymakers, but it's a pretty good solution at the individual level. It might prevent some of this sort of thing.


What if everybody did it?

A popular rebuttal to the ERE arguments is something like this: That's fine for you, but if everyone did it, the economy would collapse since nobody is buying anything. One of the founders of the movement (he goes by Jacob) has a note in response:

It is important to realize that a consumer economy in which people go to work in order to buy stuff is not the only form of economy. It is just the current one. 
Money can also be spent on productive assets, art, preserving nature, space exploration, eliminating hunger, maybe even eliminating war. It’s just that we've collectively chosen to spend it on cell phone upgrades, furniture replacements, fashionable vehicles, shoe collections, throwaway electronics, and so on.

I don't have a problem with the ERE people responding to the consumption question in this way. There is no economic law saying that two thirds of output must be spent on consumption (though some of the secular stagnationists seem to agree with the ERE critics...). But they're ignoring other important channels. ERE isn't just about less consumption; it's about less labor. Labor is a production input! Less labor means capital is less productive in the short run; in the long run, this means less capital as well.

When we do economic analysis of big economic changes, we have to think carefully about feedback mechanisms. We can sometimes leave everything constant when we ask what happens if a few people change their lifestyle preferences. We don't have that luxury when we want everyone to change. A useful way to think this question through is to use a model in which the relevant prices and quantities are allowed to adjust.

Suppose I take a little economic model off the shelf (for those who care, this is from page 40 of the McCandless RBC book, see description here; set depreciation=0 and set theta=0.33). This isn't the right model for the job, but I think it will illustrate the point. We produce stuff with capital (machines, buildings, whatever) and labor. We get utility (happiness) from consumption and leisure time. Think of two ways we can model the shift to ERE: people become more patient, or people increase their preference for leisure. Each concept maps directly to a parameter of the model (beta and B, respectively. Higher beta means more patient; higher B means more leisure preference).

Let's look at model steady state (long-run) outcomes for various levels of impatience and leisure preference. In the figure below, the X axis shows time spent working. The Y axis measures GDP. I plot lines for several different levels of patience, with higher beta meaning more patient (click for a larger image).


First, let's suppose the shift to ERE is all about an increase in patience. Then things look pretty good: increasing beta means more output for a given amount of labor. But ERE isn't about patience. It is explicitly about leisure time. We want to move along a given curve in the graph, not shift between curves.

Focus first on the black line. Reducing labor time from about one quarter to one tenth reduces GDP by about 60 percent (note: in this setup, all GDP is consumption in the steady state). We can even be generous: suppose we reduce labor to 10 percent while also raising beta to 0.99. GDP still falls by about 14 percent. That might be fine: in this model, all outcomes are optimal given households' preferences, so a change in output caused by a change in leisure taste doesn't bother anyone. But reason beyond the model a bit. Here's Jacob:

Money can also be spent on productive assets, art, preserving nature, space exploration, eliminating hunger, maybe even eliminating war. 

The space exploration and hunger elimination, at least, will require output. The chart shows a pretty reliable tradeoff between labor time and output--after allowing the economy to adjust thoroughly. We can somewhat mitigate the output loss through increased patience. But to keep output high, we need a really big increase in patience and/or a smaller reduction in labor supply. This is just a simple model, and the specific numbers aren't that important. What's important is that the ERE world is a world where a key input to production--labor--is supplied in lower quantities. The ERE strategy is to own productive assets--but those assets are less productive without workers, and in the long run that means fewer productive assets.

ERE works great in partial equilibrium. Mr. Money Mustache is very enthusiastic about how technology and general wealth have made his lifestyle possible. But if everyone did it, who would build the gadgets? Who would build and operate the machines that build the gadgets? Who would work for the companies in which MMM owns stock? The key point is this: the ERE world isn't just about people saving more. It's about people working less, and that's what kills it.

ERE works if we're willing to accept lower aggregate output than the counterfactual. I don't get the impression that the ERE people accept that. The robots can make it possible. Until then, it's not in the feasible set.

But that doesn't matter! Because not everyone is trying to do it. You and I can still do it, or at least get as close to it as we want.

UPDATE: Here is a model description

*This is from The Lost Interview; an abridged version is in Steve Jobs, page 363.

21 comments:

  1. ERE can work as it is a new equilibrium where all of labour, output, and consumption are lower. The gadgets could be built during the short productive time.

    The comment by Jacob is incorrect, though. He is double spending his bounty: either you work less for less consumption or you do space exploration. Maybe space exploration is a more moral luxury than cable, but you cannot both save your money for retirement and use it for space exploration.

    ReplyDelete
    Replies
    1. That's basically what I'm saying. In the model, the ERE equilibrium is still efficient because it reflects the preferences of individuals... but it abstracts from preferences for space exploration. And I also think it's not obvious that all the gadgets that MMM is always testing would be around in the lower output equilibrium.

      Delete
  2. I'm generally sympathetic to the argument, but lower tax revenues might be a problem.

    ReplyDelete
    Replies
    1. Do you mean a transition problem (rigidities in state size, not in relative size to rest of economy) or a problem in that the new state is not feasible?

      Delete
  3. I retired seven years ago at 50 after my business was sold. It was a great decision, although getting a job would have been very hard, in any case. In those years, I've left innumerable comments, mainly on economics blogs, that have enriched the lives of others, and I've done that pro bono, as they say. I wouldn't have had the time to do so if I'd been working. That fact needs to part of the equation.

    ReplyDelete
    Replies
    1. Whether that needs to be part of the equation depends on what question we're asking. I don't doubt that what you say is true, but its ability to mitigate the output losses is extremely limited.

      Delete
  4. Folks who retire early do so by accumulating more capital and consuming less than they otherwise would have. That means you have MORE long run Capital, not less. More long run Capital offsets at least some of the loss in Labor (perhaps it even increases GDP).

    ReplyDelete
    Replies
    1. Hi elindbe2--the model already accounts for that! The mechanism you have in mind is illustrated by the high beta scenarios. For a given level of labor supply, you are correct--more capital accumulation, more GDP. But the key aspect of ERE is that labor supply falls. The drop in the labor supply reduces the return to capital in the short run. Eventually the capital stock falls to a lower long-run level.

      Delete
    2. By the way, as you can see, for any sizeable decline in labor supply, the discount factor must go up by a huge amount to offset the effects on the capital stock. This in part means that people have to be willing to accept a far, far lower return on capital. That's another important general equilibrium effect of ERE, ignoring the reduction in labor supply: returns on capital fall dramatically.

      Delete
  5. I have started to think of ERE as a bit of a more reasonable alternative to the idea of a universal basic income (when the argument is that "we are rich, we should dial down and not need to work" instead of a redistributionist argument)

    ReplyDelete
  6. This model is flawed, because it makes the mistake of treating labor as fungible and labor productivity as exogenous. But MMM states:

    For me, early retirement has never been about ceasing work or productive activity. Just breaking free of working to somebody else's agenda and schedule, or having the threat of running out of money influencing my decisions of what to work on...

    In other words, ERE may allow him to afford the search costs to find the most productive use of his labor. So ERE may increase his productivity. Therefore, it's not a foregone conclusion that ERE reduces productivity. By increasing marginal productivity, it may increase total productivity while reducing the total quantity of labor.

    ReplyDelete
    Replies
    1. That is not the only flaw in the model! Models simplify; that's how they work. I agree that the labor performed by people who aren't under the gun could very well be highly productive. It's a quantitative question ultimately--I think the productivity boost would have to be very, very large to overcome the results in the model. But it's a fair point.

      Delete
  7. Why are you sure early retirers wouldn't accept lower aggregate output? Surely the fact I'm willing to cut my own lifetime consumption by 50% by retiring at 37 means I can entertain the notion of cutting GDP by 50%. The whole premise of cutting my consumption is that the economy used to be 50% smaller just thirty years ago and people lived fine then, so why am I working so hard chasing that 30% improvement on a 1980s life? An economy that's 50% smaller while everyone works 50% as long seems like a perfect tradeoff to me. Sorry that I didn't understand your model and focused on the words.

    ReplyDelete
    Replies
    1. These are all fair points. I'm thinking specifically of things like Jacob's stated preference for space exploration, etc. Generally big public expenditures require a lot of output. Moreover, the broader point is this: You are able to do what you do because aggregate output is high. Yes, you are willing to consume less. But that does not mean that your current level and composition of consumption would be possible if everyone were doing what you do. In some sense, ERE relies on a lot of other people being productive. Otherwise the assets that you rely on to provide non-labor income are less productive.

      In other words, cutting your lifetime consumption by 50% is easy when everyone else keeps working, operating the machines that bring you your retirement income. If everyone else tries to do what you do, you might have to cut lifetime consumption by much more than 50%, and its composition might be different. For example, are we sure we'll keep producing cheap, efficient bicycles? etc.

      Delete
    2. Bicycles seem like the sort of mature technology that we could keep running almost on autopilot, and would want to as more people switched to lower cost transport. I imagine us losing things like semiconductors and airplanes that take a whole planet's demand to make the factories profitable. And public services like parks.

      I pretty much have to face that feeling of "what if everybody else did it?" every day, living in America: I can only have free checking because others pay $30 overdrafts, I can only have Android because others paid for Apple, I can only visit parks because others are wealthy enough not to miss the taxes. The guilt over what you describe is part of why I plan to return to work in a few years, even though I love the leisure and my health is improving fast.

      Delete
  8. Ryan - thanks for the article. I'm sorry too that unfortunately much of it went right over my head. That said I did still find the ideas I did grasp to be interesting.

    I agree that MMM ideas do seem to be "winning" in that they seem to be on trend with the movement gaining momentum. I don't however see ERE becoming a mass movement so for me concerns about the potential impact on capital seems overstated. In contrast I can't help but think that the kinds of impacts you write about may be more relevant to the shifts that relate to the much more significant baby boomer transition into retirement. There are far more baby boomers than ERE folks. I can't help but think those boomers will have a more significant impact on capital than MMM followers will ever have.

    ReplyDelete
    Replies
    1. Agreed! It was just a fun hypothetical to explore.

      Delete
  9. Thanks for the analysis, Ryan. I've always thought the "what if everyone did it" argument is an interesting thought experiment that's unfortunately used as a straw man by ERE detractors to try to invalidate or distract from the very simple and powerful ERE idea of spending less, focusing on things in life that matter and investing wisely. I'm reading "The Simple Life" by David Shi, which traces the history of frugal, more purposeful living in America, and it's clear ERE is just the latest in a long history of such movements, which never seem to gain a huge following. If it did catch on it would be very gradual and we'd have plenty of time to watch its effects.

    I'm curious if you think stay-at-home moms might be a useful analogy here. There are over ten million of them in America, and it's clear they do a lot of good for the country that doesn't specifically add to the GDP. If 10M people suddenly became MMM disciples we'd certainly hear a chorus of "Stop! Not everyone can do it!", so why don't we hear a similar cry for these moms to get to the office? Imagine the boon to the GDP with all of the new day cares, fast food business, car sales, office buildings, etc.

    ReplyDelete
    Replies
    1. And to be clear, I'm not necessarily in disagreement with the idea that if large numbers of people did ERE that it could be self-defeating due to a paucity of workers. On the surface that sounds like a pretty reasonable argument, although I think it's way too complicated to know with any certainty what would happen. I can't help but think if all of humanity were suddenly interested in the kinds of things MMM is - health, family, travel, intellectual pursuits, etc - that the world would not in fact fall into a state of squalor because we aren't trading in our cars at as fast a clip. My main beef with the argument is that it's both mostly irrelevant and it gets used a lot.

      Delete