Skip to content
 

“Threshold earners” and economic inequality

Reihan Salam discusses a theory of Tyler Cowen regarding “threshold earners,” a sort of upscale version of a slacker. Here’s Cowen:

A threshold earner is someone who seeks to earn a certain amount of money and no more. If wages go up, that person will respond by seeking less work or by working less hard or less often. That person simply wants to “get by” in terms of absolute earning power in order to experience other gains in the form of leisure.

Salam continues:

This clearly reflects the pattern of wage dispersion among my friends, particularly those who attended elite secondary schools and colleges and universities. I [Salam] know many “threshold earners,” including both high and low earners who could earn much more if they chose to make the necessary sacrifices. But they are satisficers.

OK, fine so far. But then the claim is made that “threshold earning” behavior increases income inequality. In Cowen’s words:

The funny thing is this: For years, many cultural critics in and of the United States have been telling us that Americans should behave more like threshold earners. We should be less harried, more interested in nurturing friendships, and more interested in the non-commercial sphere of life. That may well be good advice. Many studies suggest that above a certain level more money brings only marginal increments of happiness. What isn’t so widely advertised is that those same critics have basically been telling us, without realizing it, that we should be acting in such a manner as to increase measured income inequality [emphasis added]. Not only is high inequality an inevitable concomitant of human diversity, but growing income inequality may be, too, if lots of us take the kind of advice that will make us happier.

This is a cute idea but I don’t think it’s correct. I’ll explain my reasoning but first one more quote from Salam:

The idea of the “threshold earner” came to mind as I [Salam] read Michael Arrington’s post on “the rise of the gentleman hacker,” who might be understood as souped-up threshold earners, who’ve already climbed various entrepreneurial mountains.

No, I don’t think “threshold earning” increases income inequality

I don’t see how you can get from threshold earners to increased income inequality. It seems to me that threshold earning would decrease inequality, in both the examples that Salam gives. First, you have a well-educated middle class person who could be a high earner but instead gets a more moderate income (perhaps by writing for a political magazine, for example). You’re taking someone who could, perhaps, with sufficient effort make $200,000 a year but instead makes $50,000. This will tend to compress the income distribution, pulling some of the moderate high tail toward the median. Second, you have some who could possibly become a billionaire who decides to be a mere multi-millionaire instead. Again, this compresses the income distribution.

The only way I see threshold earning as increasing income inequality is someone who could earn $20,000 a year instead slacks off and earns $5,000 a year instead. I’m sure there are such people, but given the cost of living, I’m imagining there are a lot fewer people like that now than in the 1960s and 1970s. And these don’t seem to be the people that Cowen and Salam are talking about. Cowen in particular was focusing on inequality in the top 1 percent of the income distribution.

The argument that Cowen seems to be making goes like this: There is some cohort of high-powered people in the workforce who, if they all worked the same amount of effort, would have some spread in their incomes. But then some of these people work 100-hour weeks (and make more money) while others work less hard (the “threshold earners”), and the result is increased variation in incomes.

In the context of the general population, though, I see this threshold earning as a source of decreased inequality (as discussed above) in that it brings some potential incomes high in the tail down toward the median.

Normals, workaholics, and slackers

Imagine a simulation, in which people start by being categorized in percentiles of “potential income,” from the bottom (people who don’t have a serious chance of making much money, no matter how hard they work) to the top (people who could become rich, if they put in the effort and get lucky).

We’ll focus on the top qunitile: the people in the top 20% of the distribution of potential income.

Start with a scenario on which these people work normal hours and have some distribution of income, with some millionaires, some people making less $50,000 per year, and many people in between.

Now imagine that some people in this quintile are “workaholics,” people who work long hours and might become really rich. Taking a normal person in the top quintile and turning him into a workaholic will increase the inequality in the income distribution.

Now turn some normals into “slackers,” who work less hard and make less money. Taking a normal person in the top quintile and turning him into a slacker will decrease the inequality in the income distribution.

To the extent that variation in work habits and money goals is causing income inequality to increase, this is coming from the workaholics, not the slackers. Contrary to Cowen’s claim (see the long quoted paragraph above), if more people in the top quintile are taking the advice to chill out, enjoy their gains, and relax, this should be causing a decrease in measured inequality, not an increase.

Summary

(a) I didn’t see the evidence that threshold earning behavior is a bigger deal now than it used to be. (From an economic perspective, threshold earning made a lot more sense in the era when top tax rates were 70% or higher than now when they’re typically below 50%.); (b) I’d guess that threshold earning is compressing the income distribution (it’s people who could be making a lot who are instead closer to the median) and thus reducing the level of measured income inequality.

On the other hand, this is soooo clear to me that it’s hard for me to see how Cowen and Salam could’ve got this “wrong.” Cowen’s So maybe I’m the one who’s confused (in which case I hope someone will do me the favor of explaining the errors in my reasoning).

My friend Ubs is the ultimate threshold earner. At one point he felt he had enough money so he quit his job for a few months ‘cos he didn’t feel like working. I wonder what his reaction would be to all this.

P.S. Here’s how Salam ends his column:

And I [Salam] can’t for the life of me see how housing a somewhat smaller share of the global rich will mean that fewer Americans will spend their prime-age years in prison.

I agree. The percentage of people who are in prison is basically a political decision to lock up convicted criminals for a long time rather than giving them shorter sentences, parole, fines, community service, etc. I’d hope we could have more reasonable prison populations, whatever the tax rate structure happens to be.

P.P.S. Salam has more here. In response to his points: yes, I was using “slacker” in a playful sense to refer to those people who choose less lucrative work options or career paths.

30 Comments

  1. Reihan says:

    I should have been more explicit: by more dispersion, I mean to suggest that there are more satisficers and more maximizers. There are more opportunities to earn very, very high incomes. But not everyone who can does so. And because one can enjoy a comfortable standard of living at lower levels of income, particularly through geographical arbitrage, this suggests that a society that consist solely of highly educated people with lots of accumulated human capital would experience *more* wage dispersion in light of a (hypothesized) increase in threshold earning behavior.

    Because older age cohorts in the US are more educated than in the rest of the advanced market democracies (not younger cohorts), this helps account for some of the increase in measured inequality.

    Insofar as threshold earning is a cultural/normative shift, the salient fact is that which MTRs have declined, elasticities of taxable income vary across individuals given cultural affiliation and normative priors. As need recedes, identity becomes more important. So declining MTRs are particularly salient for maximizers, not satisficers.

  2. Andrew Gelman says:

    Reihan:

    It still seems to me that if a person in the top quintile decides to be a satisficer rather than a maximizer (in your words, "not everyone who can does so"), that this will reduce the number of people with very high incomes, which in turn will reduce income inequality.

    I'm not disputing your (and Tyler's) main point, though, which (as I see it) is that high income inequality is not necessarily a bad thing in itself, that ultimately you have to look at individuals to see what's working for them in their lives, rather than using a summary measure that's averaging over all sorts of behavior. What I got out of Tyler's column was that it makes sense, when looking at causes and consequences of inequality, to break things up and look at different groups of the population: different social classes and also different age cohorts etc.

  3. DKB @ NYU says:

    Inequality is a measurement nightmare: what's the unit (family? individual?), the population (workers only, or everyone? prisoners?), the time interval (a year? decade? lifetime?), etc. On this one, your argument makes sense to me.

    But I think the evidence is that the dispersion in (say) log earnings (wage times hours worked) is greater than the dispersion in wage rates — wages and hours are positively correlated. So in that sense, decisions on how much to work increase inequality. Not sure how that lines up with your story.

  4. TheOneEyedMan says:

    The cost of living for the poor has gone up much more slowly than the rich because marginal tax rates have fallen more and inflation is much lower on the bundle of goods consumed by the poor (http://www.slate.com/id/2123009/). That, combined with the geographic dispersal you've mentioned and globalization which increases the returns to human capital of of high effort, high ability people and you get one group of people on the poor end with the incentive to exert less effort and one group on the rich end with the incentive to exert more.

  5. Reihan says:

    We're on the same page. I'm surprised that you saw the "decreasing measured inequality" piece as the main takeaway. Just wrote a new post on this, delighted to have caught your reply before publishing.

  6. Mark Palko says:

    I may be completely off here, but this sounds a bit like Simpson's paradox. Within professional subgroups (MBAs, engineers, doctors, etc.) it does increase inequality, but in the aggregate it pushes people toward the median.

  7. How useful is this nuance? I get it as posture (for an "ironic" conservative like Reihan, it has the appearance of a costless posture to built conservative bonafides) but how big a slice should it be of our attention pie?

    I've seen Reihan as an embedded rationalist secret agent within an ideological space on the side of angels in the past. Here I'm not so sure.

  8. Mark Palko says:

    p.s. I'm wondering if there's a better comparison than Simpson's when an increase in subgroup variability produces a decrease in overall variability.

  9. Daniel Coelho says:

    Cowen "slaker" argument refers to why so few had so great income growth. The inequality source, as you can see in his article, is demografic moviments, living costs and mainly the financial sector. Cowen's point is centered in finance sector behaviour and the bad incentives it faces, really worth reading.

  10. Andrew Gelman says:

    Daniel:

    You write that Cowen explains why so few had so great income growth. But if more people had great income growth–moving some people up from the 80th percentile, say, to the 99th percentile, or from the 99th to the 99.9th–then inequality would've increased even more. I'm not dissing Cowen's whole article, I just think he's wrong to claim that threshold earning is increasing income inequality.

  11. Kevin says:

    You're confusing the two issues Cowen warns about:
    "The data show a big difference between two quite separate issues, namely income growth at the very top of the distribution and greater inequality throughout the distribution."

    Clearly your point is correct when we think about the entire distribution, but Cowen spends a couple paragraphs explaining how the distribution between the 0-99th percentiles has not changed (“there was no increase of inequality after 1993 in the bottom 99 percent of the population”). Cowen is trying to explain the rise of the super-super-rich in the top percentile.

    Thus the number of threshhold earners can make a difference on the distribution at the very top. If the distribution of 0-99 is unchanged from 1990 to 2010, but more people are threshold earners in 2010, then we would expect the top 1% of wages to be much higher in 2010 relative to the bottom 99% because the super super rich guys are likely not the threshold earners.

    Once you understand the context, Cowen's point is clear.

  12. Andrew Gelman says:

    Kevin:

    This might be what Cowen was thinking but I think the math still doesn't work out under any reasonable model. See my comment to Daniel above.

    To me, the logic of Cowen's argument is that the "workaholics" are driving the inequality, in the entire distribution and also at the high end. If there were no "workaholics" (or if the highest income tax rate were, say, 90% (for real, not just in Greg Mankiw's 30-years-into-the-future calculations), so that there was much less financial incentive to make more money), then there'd be a lot less measured inequality (again, in the whole distribution and also at the high end). So it's fair to label workaholism (or whatever you want to call it) as a contributing factor to measured inequality.

    But "slacking," or "threshold earning," goes the other way, even if you just consider the high end. Again, suppose that nobody at the high end of potential income slacked. Then you'd have some more ultra-rich people and more variance at the high end. Imagine doing a simulation. Take a bunch of "slackers" making 50-75K and spread some of them up into the stratosphere. Then you're shifting the whole high end up. And when you move some people up, that's changing the intermediate percentiles. I don't see how this will be decreasing measured income inequality, even at the upper epsilon percent. Sure, I suppose it's possible to construct a model that would do this, but I think it would be a pretty weird model. The most natural way to do this would, I think, simply increase the variance at high percentiles as well.

  13. Steve Sailer says:

    Are there more people satisficing than a half century ago? I don't know. One thing that's obvious about the appeal of the show "Mad Men" (an appeal which is limited to, say, the upper quintile or so of the population) is that Manhattanite professionals are shown having more leisure than Manhattanite professionals have today.

  14. Noni Mausa says:

    The ability to adjust income like this is itself a luxury. Teens living at home can do this – 1950s TV wives could do it by taking a hobby job – and people accustomed to wealth can do it, either because they have savings or because their family and social milieu provides many opportunities for above-average income positions.

    That this article could be written at all is an indication of the presumed wealth of the people Cowen is discussing.

    In the 70s people at most salary ranges could move fairly easily from job to job. I remember it clearly, as a college kid I could support myself, try new things or move from town to town, confident that something would turn up. Not so now. At that time, the idea that educated, experienced people would search two years or more for any work at all was completely foreign.

    Cowen's article demonstrates that the mobility and luxury of choice of the 70s live on for the well-to-do. Not so for the rest of us.

    Noni

  15. Andrew Gelman says:

    Noni:

    Not quite. It's not about being well-to-do. Rather, as Salam points out, the key is having minimal responsibilities. When Ubs quit his job because he wanted more free time, he did not have a lot of money. What he had was low living expenses, so he could afford to live on his savings for awhile. He preferred a less expensive lifestyle that required him to work less.

  16. George N. Wells, CPI says:

    Is this an economic decision or a psychological one? Looking back on my career I guess I would have to call myself, in the context of this discussion only, as a "Slacker." However, my personal reasons were, and still are, more based in my psychology than in my economics.

    My Myers-Briggs classification is an ISTJ with a very strong "I" score and I am an introvert. I do well in work that allows me to work independently with minimal person-to-person responsibilities. In my working career (I'm now retired) I avoided promotions to varous positions because of the requirement for more group-based work over individual productive work. Eventually,I became a very successful consultant in the Supply-Chain-Management field because I could limit the amount of interpersonal relationships required and still be highly productive — actually I was more productive in consultancy than when I worked for corporations as an employee/manager.

    The use of the terms Slacker/Workaholic/Normal seem prejorative to me. I also feel the basis for the behaviors is deeply rooted in psychology and not in rational economic decisions.

  17. Andrew Gelman says:

    George:

    All decision are ultimately psychological. But I do think that economic factors affect a person's psychological decision making. If you know (or believe) the government is taxing your marginal income at 90%, you might decide not to work so hard. This doesn't have to be the product of a formal utility calculation; it can be something more indirect, such as: knowing that you're only keeping 10% of your money, you don't enjoy spending time at work so much. But the economic processes are still having an effect on your decisions.

  18. Jason McCullough says:

    Why on earth are you bothering to respond to something in the National Review?

  19. Hyena says:

    "Threshold earning" really has to come in amongst the upper quintiles, no one else can do it without state support because it's impossible to find high-end part-time work so that you can make, say, $20k per year and work only half the week.

  20. Tyler Cowen is just rehashing the conservative inequality is good argument. The core of the argument, any way they go at it, is that there is a direct hydraulic relationship between effort and income. But that's so clearly nonsense it's laughable. To paraphrase a famous Larry Summers quip: just look around you. Working hard is a necessary but by know means sufficient condition for becoming rich (ignoring inheritance wherein it wouldn't even be necessary but the evidence very clearly shows that inequality is driven by earned income). As you go up the income distribution the positive relationship between hours and income breaks down.

    I honestly don't see how anyone could not arrive at the conclusion that inequality is a result of Superstar Economics and political capture (i.e. the rich getting a pass on gaming the tax code). All the other theories have been shot down and Cowen's is so clearly bogus it's a joke.

    I am by no means claiming that I know everything. There are many total mysteries left out there in economics but this one seems to be fairly well resolved in my opinion. Also, I am by no means saying that Cowen et al are stupid. They're just willfully ignoring the evidence and turning out clever arguments to support there ideology.

    My favorite is Greg Mankiw's paper about taxing people based on their height. Yes, that would be a decent way to tax. In fact taxing based on anything that is positively correlated with income would do the trick: beauty, gender, race, etc. His argument is just a clever way to say "See! Look at how arbitrary and unfair progressive taxation is." And I know his point is that the logic of utilitarianism leads to that conclusion. But the height thing is just a straw man. It's simply beside the point. Progressive taxation is a normative issue. There's also no way to prove an optimal age of sexual consent.

    You all need to, at some point, stop playing their game and just ignore these arguments. It's always the same old sausage in a new wrapper. These people are mocking you and in so doing wasting your time.

  21. Jonathan says:

    It seems to me the issue is the metric. Suppose there are three groups, each 1/3 of the population, with capabilities of 1,2 and 3 in income. Now suppose half the threes behave as twos, so the twos are now half the population. The standard deviation of income declines from 0.8 to .69. That's Andrew's point. But the Gini coefficient rises from 2/3 to .678, right? What has happened is that the high earners have moved farther from the mean than before because the mean has fallen. And this tail behavior is one metric of inequality — how far are the highest earners from the mean? Under this metric, inequality rises with "slackers."

  22. David Shor says:

    I see it more as a general equilibrium point, that the wide-spread presence of "satisficers" makes "Maximizers" make a lot more money. I see this a lot in my cohort:

    Financial math is, for a math major, relatively easy. The only issue is that it isn't terribly interesting to most math majors and that the life-style is very hectic.

    If math major just wanted to maximize income, then they would all become quants, and then quants really wouldn't make much money. Instead, because the overwhelming majority of math majors are "satisficers", quants make unholy salaries.

    This isn't an isolated example. It's relevant with Lawyers, Doctors, Professors, etc. As you pointed out, this has nothing to do with the poor and middle class.

  23. Jonathan says:

    Ignore my previous comment, the math of which is subject to New Year's Day hangover. The point is that if we measure inequality by the ratio of the highest group of earner's income to the mean (as some do) slackers cause inequality to rise by lowering the mean without affecting the earnings of the highest group of earners, particularly when the slackers come from the group just below the highest group, so that they don't affect the composition of the highest group at all.

  24. Noni Mausa says:

    Above a certain base income you would be correct. When ubs quit his job he had a low cost of living, granted, and savings. Enough savings so he could make his move without risking destitution.

    Most people probably can't do this. "…the typical American family…has about $3,800 in the bank. No one has a retirement account, and the neighbors who do only have about $35,000 in theirs. Mutual funds? Stocks? Bonds? Nope. The house is worth $160,000, but the family owes $95,000 on it to the bank. The breadwinners make more than $43,000 a year but can't manage to pay off a $2,200 credit card balance…" [link below]

    And that was in 2006, before any of the poo hit the propeller. Even if this median "family" is only one guy — that's still only $4k in savings, $70k in equity and $35k in pension savings. Sounds like a lot till you try to live off it.

    A US buddy of mine, in a well paid technical position, had $40k in a 401-k when the economy crashed and he was laid off. Half that money evaporated, the other half went to living expenses from September to March. Eleven years savings, gone pffft.

    Threshold people can sit on the threshold only if they are Buddhist monks, or else are assured that the door behind them isn't locked and they can get back in the house when they need to.

    http://www.washingtonpost.com/wp-dyn/content/arti

  25. Andrew Gelman says:

    Jason:

    It's always a tough call, how much time to spend engaging with the general public vs. scholarly research. Perhaps I do too much of the former.

    To answer your specific question, I read Marginal Revolution regularly. From there I followed a link to Salam's discussion.

    But, yeah, I agree that there's no point in responding to this sort of thing.

    Leroy:

    I wrote about Mankiw's height tax here.

    David:

    Yes, I actually thought about this when I was writing my blog on this. You might be right, but it still seems like a bank-shot, second-order kind of effect. I still think that people dropping out of potential riches will decrease, not increase, income inequality. Under the right assumptions, though, I suppose it could go the other way.

    Jonathan:

    But "satisficing" would lower the incomes at the highest quantiles. That's the whole point: there are some potential Bill Gateses out there wren't bothering to put in the hours and thus aren't hitting those income thresholds.

  26. Jonathan says:

    Andrew: it depends on who satisfices. Assume a tournament model, so that a small set of winners get outsize shares. Then when Bill Gates becomes a Starbucks clerk, some guy we never heard of sells an operating system to IBM for the PC and becomes Bill Gates. Alternatively, all you have to assume is that it's the guys several steps below Gates who satisfice — Guys who would have earned $1 million per year who barely shpw up at all in the top percent. But the guys at the top are so driven that they appear there no matter what.

  27. MWS says:

    I think you are correct to question Cowen's argument. But I think the pattern he identified can help us understand increased income inequality, just not for the reason Cowen states.

    Take the following toy model. A population has two properties, "ability" and "insatiability". Each property has two values, high and low ability, and high and low insatiability. Each property is distributed so 10% of the population has high ability and 10% high insatiability. The two are independent.

    Consider the case where income only depends on ability. Say High ability earns $1M and low ability earns $50K. The millionaires are concentrated in 10% of the population

    Compare that with a case where income also depends on insatiability. Say insatiability affects income as follows. Highly insatiable people will earn all they can. Low insatiability people are "satisfied" with earning $100K.

    When combined with ability, low ability workers are unaffected by their insatiability. They earn $50K no matter what. But for high ability workers, only the highly insatiable worker earn $1M, the rest earn $100K.

    The resulting distribution of income has 90% earning $50K, 9% earning $100K, and 1% earning $1M. That's more inequality, fewer people at the high end are millionaires. I think that's what Cowen is right about.

    But where he's wrong is in how he interprets the math. He makes a lot of errors there.

    First, by labeling the threshold property by a willingness to work, he's prejudging the worth of the people. If instead we labeled the property "immorality", and theorized that a high ability person who is highly immoral has the opportunity to earn $1M (say by a Ponzi scheme like Bernie Madoff) while a moral person of whatever ability is limited to making $100K, we end up with the exact same income distribution and the same level of inequality. The distribution of income tells us little about the properties that lead to that distribution.

    Second, the key insight is that inequality is caused by people having to pass through a series of gates to win the big bucks. You have to pick the right parents, have enough intelligence, go to the right schools, pick the right career, have the right connections, be immoral enough to cheat (you need sharp elbows), and so on. These aren't independent, and few of them have much to do with "how willing to work" you are. Inequality rises the more gates must be passed.

    What's hard is teasing out the conditional probability distributions that model how each of these properties affects income. Cowen is making a hypothesis that some people have some property (whether it is insatiability or morality) that limit their maximum income. That's is one of many possibilities. But there's only anecdotal evidence of it in what he wrote, and thus its only a hypothesis. His idea is like the silly idea that the reason we have high unemployment is due to people not wanting to work.

    IMHO, the anecdotal evidence I've seen tells me that other factors have a much bigger impact than this idea of a "threshold earner" being due to a willingness to work. Indeed, Cowen's post provides plenty of evidence of this. He acknowledges that most of the inequality is due to a few traders on Wall Street. They seem to have managed to get most of the increased GDP through huge bonuses driven by financial engineering. If you didn't choose to work on Wall Street, you didn't become a big winner. Since only a few can pass through that Gate, inequality is increased.

    Mark

  28. Steve Roth says:

    Just to say that threshold earners aren't just at the bottom end of the spectrum. I'm an example of one who's well into the upper end.

    An example of someone like me thinking through the problem:

    If I could have $250,00 a year in income for life without working (inflation adjusted, so it goes up each year), and leave a little chunk for my kids, I'd have everything I need — *in spades.*

    That's a boatload of money even by my quite high standards, especially with no work. No problem paying for a nice house, reliable and comfortable cars, private schools and post-secondary for the kids, nice (though not extravagant) vacations on a regular basis… I just don't need any more than that.

    Now to get a reliable quarter million a year, I need $4 to $6 million. (Assuming 4% after tax returns, and spending down most of the capital over my life, leaving some but not much for my kids, depending when I die.)

    So when my business is worth $5 million, I'm going to sell it and start working on cool stuff that might or might not be remunerative.

    Slacker?

  29. Ubs says:

    I realize I'm more a conceptual example than an actual data point, but for the record, the most savings I've ever had is about $20,000, which is roughly what I had each time I quit my job. The first time I quit for about a year; the second time I quit for about six years while pursuing a "career" as an artist. But even if I had $20,000 when I quit, there were times when I was close to zero and still chose not to look for a real job. There were many times during my starving artist phase when I didn't know where rent was coming from more than about one month in advance. I think it's a personality thing. Some people can't handle that kind of uncertainty. Also, it helps a lot to be an unmarried childless non-homeowner.

    I think you guys are making the income inequality question more complicated than it needs to be. Being a threshold earner makes your income less than it would otherwise be. If it would otherwise be above average, it decreases inequality; if it would otherwise be below average, it increases inequality. So if most threshold earners start out above average (which I suspect they do) then it decreases inequality. The same thing applies to workaholics in the other direction. If a workaholic is above average, his workaholism increases inequality; if he's below average, it decreases it

    I would add that we're talking about income inequality here, as opposed to wealth inequality. Which is really more relevant? I know lots of people who earn more than I do but spend beyond their means. This past year I grossed about $30,000. My spending was such that my savings crept up a little bit, and when it was around $10,000 I cut back my hours. If some other guy grosses $100,000 a year but spends to match so that his net worth is near zero because his mortgage matches his home value and he's carrying credit card debt, then who is wealthier?

    I suppose you could say that he is wealthier because he's driving a nicer car, taking a vacation in Hawaii, and going out to dinner every night, while I'm sitting in my little house reading books from the library and shopping at Safeway. But then you're saying that the essence of wealth is consumption of material goods. That's not what I value. But then, I guess that's why I'm a threshold earner in the first place.

    I used to criticize people whose spending increased to match increases in their income. If a person got by on $50,000 for a few years, and then her income went up to $80,000, I figured she ought to now be saving $30,000 a year, and if she instead ramped up her lifestyle then she was being an undisciplined spender. But later I realized that I essentially do the same thing. The only difference is that my "spending" takes the form of reduced hours. Each of us is reacting to an increase in earning power by spending more on what he values most, but for me what I value most is having more free time by way of working fewer hours.

    This makes me wonder if what is really most relevant is earning potential, not actual income. If I'm capable of earning $100,000 a year, but I'm slacking and earning $30,000 in a lazy job with lots of free time instead, then I'm essentially getting $100,000 worth of life satisfaction. If B is also capable of earning $100,000 and she chooses to actually do so while happily spending it as fast as she makes it, she is also getting $100,000 worth of life satisfaction. But suppose C and D are only capable of earning $30,000. Maybe C wishes he could have more time off but he can't because he has to work long hours just to get by. And maybe D wishes she could afford to spend a lot more of buying fun toys and vacations but she can't because she can't earn enough no matter how hard she works. The focus on measuring income groups me with C and D, but I think really I'm more like B but with different spending habits.

  30. Andrew Gelman says:

    Steve:

    Yes, I was just using "slacker" as an amusing shorthand. I think "threshold earner" is a better term.

    Ubs:

    What you say makes sense. There are a few more complexities, though. I know people who make more money than they could possibly need, but they still want to make more. (Hey–this describes me too!) One problem, I think, is money's fungibility and imperishability. The imperishability is not complete–after all, there's inflation (sometimes) and the possibility of economic catastrophe (I have an uncle who's been predicting this since back when he started hanging out with off-the-grid dudes back in the 70s!)–but still . . . the point is that money is appealing, partly because there's no logical stopping point. There's never a compelling reason not to accumulate more. Or maybe this is pretty much irrelevant given that most Americans at all income levels are probably living at the edge of or beyond their means. You and I are just lucky enough not to have expensive tastes relative to our wealths.