Permanently Stupid
-- Posted by Neil H. Buchanan
In the early 1980's, The Harvard Lampoon ran a parody issue of Newsweek magazine. Among the more memorable pieces therein was an opinion piece by a columnist named George Fwill entitled, "Why I Like the Feudal System." Almost thirty years later, the real-life columnist George F. Will continues to challenge the limits of parody. Most recently, he has argued that cutting off unemployment benefits for the long-term unemployed (the affected population of former workers currently numbering over 1.2 million people, growing to 2 million within two weeks) is just plain good economics.
In a recent appearance on ABC, Will made one familiar argument (People will just be lazy if you pay them to do nothing) and one relatively novel one (People will not spend their unemployment benefits, because those benefits are not "permanent income"). I take issue with both arguments in my most recent FindLaw column, published yesterday: "The Debate Over Unemployment Benefits: Economics is Complicated, Heartlessness is Simple." In that column, I focused most of my attention on the dole-as-work-disincentive argument, offering only a brief critique of the particularly weird use of the permanent income argument to oppose unemployment benefits. I will expand on the latter here.
The Permanent Income Hypothesis (PIH) has long been a staple of undergraduate economics courses. I always taught it either in basic macro or in intermediate macro, in large part because it is such a good example of how a theory that relies on people acting like "rational economic men" holds up in the real world. The short answer: not well.
The PIH starts by noting the difference between income that is transitory and income that is expected to continue into the indefinite future. A one-time $10,000 bonus (say, for 20 years of dedicated service) is different from a $10,000 salary increase, assuming that the job is likely to continue for the rest of the worker's career (an increasingly fantastic assumption in the U.S. economy). The psychological question is: How will a person divide an increase in income between saving and consumption, depending upon whether the income is permanent or transitory? It is easy to imagine that there will be a difference, but it is not so easy to know what that difference will be.
Consider the psychology of a person who has just been given a $10,000 one-time bonus. Should she spend it all, save it all, or a little of both? The answer will, of course, vary from person to person. One person might say, "I've been living just fine on my salary, but I've always wanted to go to Europe, so I'll spend the bonus on the trip of a lifetime." Another might say, "This is a good way to make up for not having started a college fund for my kids. Into the bank it goes!" Those same people might well have different answers if the bonus is only $100.
The adherents of the PIH, however, make a strong prediction. People who know that the bonus is a one-shot deal will understand that their lifetime incomes have risen only a tiny bit. (If you have 25 years of earning in front of you, this is -- interest aside -- like a mere $400 increase in your annual salary.) They will thus rationally put the money in a savings account and adjust their consumption only by a tiny amount each year.
Of course, this is really an empirical question. Given the differences in how people think about finances, they will surely act in ways that cannot be predicted in advance -- and that do not at all reflect the "rational" decisions that the theory requires. The classic evidence offered to support the PIH is the apparent response to two tax cuts in the 1960's. In one, the tax cut was announced as a temporary one-shot deal. In the other, the tax cut was announced as permanent. In the aggregate, most people apparently saved the first tax cut, but they spent the second one. There has been a lot of econometric work dedicated to building up a stronger edifice of empirical evidence for the PIH, with some success. Still, given that every change in income is of a different size, in different circumstances (the 60's being a time of sustained prosperity and low unemployment, for example), this is one of those empirical questions that is almost surely impossible to nail down as a general matter. Context matters. A lot.
This basic question arises all the time in policy analysis. Is a one-shot $3oo tax rebate going to be spent or saved? Because tax cuts are analytically no different from government support payments (merely "negative taxes"), it might be tempting to apply the idea to unemployment benefits. And since unemployment benefits -- even if extended for the duration of a multi-year economic near-depression -- are known by all to be temporary, one might try to predict whether unemployment benefits will be spent or saved based on the PIH. For example, Will said last Sunday: "[W]hat stimulates is the consumer and savers' sense of permanent income. And everyone knows that unemployment benefits are not permanent income."
Again, however, the PIH is not a reliable predictor of people's behavior in all (any?) circumstances. Even if we have evidence that an across-the-board tax cut was saved by most people, that tells us next to nothing about how people who have been jobless for over a year will act. A law professor might view a $300 check as something to drop in his bank account (although he might also view it as an opportunity to spend a night at an inn in the Finger Lakes region); but someone who has been living on meager unemployment benefits for months and months will spend it immediately.
Not spending it, after all, is only an option if there is some other source of money on which to rely to buy food, medicine, etc. People who have run through their savings, lost the equity in their homes (and have probably lost the homes as well), and tapped out every source of support are in no position to say, "But this unemployment check isn't permanent income; so I should not adjust my spending because of it."
Whatever one might think of the predictive value of the PIH in general, therefore, it is simply irrelevant to the situation of the long-term unemployed. Whereas we need empirical evidence to predict the effect on consumption of certain types of income changes, there is no doubt how the long-term unemployed will respond to receiving an extension of benefits. They will spend it all. That is known as "economic stimulus."
Even to describe the misapplication of the PIH to this situation is to ridicule it. That is not to say that Will understood the actual theory or the contexts in which it should and should not be applied. It is a talking point that somehow bubbled up into his discussion one day. It has the advantage of being obscure and sounding important. Even so, it is a ridiculous argument to offer in support of a callous effort to score political points on the backs of the people who have been most harmed by the recession.
In the early 1980's, The Harvard Lampoon ran a parody issue of Newsweek magazine. Among the more memorable pieces therein was an opinion piece by a columnist named George Fwill entitled, "Why I Like the Feudal System." Almost thirty years later, the real-life columnist George F. Will continues to challenge the limits of parody. Most recently, he has argued that cutting off unemployment benefits for the long-term unemployed (the affected population of former workers currently numbering over 1.2 million people, growing to 2 million within two weeks) is just plain good economics.
In a recent appearance on ABC, Will made one familiar argument (People will just be lazy if you pay them to do nothing) and one relatively novel one (People will not spend their unemployment benefits, because those benefits are not "permanent income"). I take issue with both arguments in my most recent FindLaw column, published yesterday: "The Debate Over Unemployment Benefits: Economics is Complicated, Heartlessness is Simple." In that column, I focused most of my attention on the dole-as-work-disincentive argument, offering only a brief critique of the particularly weird use of the permanent income argument to oppose unemployment benefits. I will expand on the latter here.
The Permanent Income Hypothesis (PIH) has long been a staple of undergraduate economics courses. I always taught it either in basic macro or in intermediate macro, in large part because it is such a good example of how a theory that relies on people acting like "rational economic men" holds up in the real world. The short answer: not well.
The PIH starts by noting the difference between income that is transitory and income that is expected to continue into the indefinite future. A one-time $10,000 bonus (say, for 20 years of dedicated service) is different from a $10,000 salary increase, assuming that the job is likely to continue for the rest of the worker's career (an increasingly fantastic assumption in the U.S. economy). The psychological question is: How will a person divide an increase in income between saving and consumption, depending upon whether the income is permanent or transitory? It is easy to imagine that there will be a difference, but it is not so easy to know what that difference will be.
Consider the psychology of a person who has just been given a $10,000 one-time bonus. Should she spend it all, save it all, or a little of both? The answer will, of course, vary from person to person. One person might say, "I've been living just fine on my salary, but I've always wanted to go to Europe, so I'll spend the bonus on the trip of a lifetime." Another might say, "This is a good way to make up for not having started a college fund for my kids. Into the bank it goes!" Those same people might well have different answers if the bonus is only $100.
The adherents of the PIH, however, make a strong prediction. People who know that the bonus is a one-shot deal will understand that their lifetime incomes have risen only a tiny bit. (If you have 25 years of earning in front of you, this is -- interest aside -- like a mere $400 increase in your annual salary.) They will thus rationally put the money in a savings account and adjust their consumption only by a tiny amount each year.
Of course, this is really an empirical question. Given the differences in how people think about finances, they will surely act in ways that cannot be predicted in advance -- and that do not at all reflect the "rational" decisions that the theory requires. The classic evidence offered to support the PIH is the apparent response to two tax cuts in the 1960's. In one, the tax cut was announced as a temporary one-shot deal. In the other, the tax cut was announced as permanent. In the aggregate, most people apparently saved the first tax cut, but they spent the second one. There has been a lot of econometric work dedicated to building up a stronger edifice of empirical evidence for the PIH, with some success. Still, given that every change in income is of a different size, in different circumstances (the 60's being a time of sustained prosperity and low unemployment, for example), this is one of those empirical questions that is almost surely impossible to nail down as a general matter. Context matters. A lot.
This basic question arises all the time in policy analysis. Is a one-shot $3oo tax rebate going to be spent or saved? Because tax cuts are analytically no different from government support payments (merely "negative taxes"), it might be tempting to apply the idea to unemployment benefits. And since unemployment benefits -- even if extended for the duration of a multi-year economic near-depression -- are known by all to be temporary, one might try to predict whether unemployment benefits will be spent or saved based on the PIH. For example, Will said last Sunday: "[W]hat stimulates is the consumer and savers' sense of permanent income. And everyone knows that unemployment benefits are not permanent income."
Again, however, the PIH is not a reliable predictor of people's behavior in all (any?) circumstances. Even if we have evidence that an across-the-board tax cut was saved by most people, that tells us next to nothing about how people who have been jobless for over a year will act. A law professor might view a $300 check as something to drop in his bank account (although he might also view it as an opportunity to spend a night at an inn in the Finger Lakes region); but someone who has been living on meager unemployment benefits for months and months will spend it immediately.
Not spending it, after all, is only an option if there is some other source of money on which to rely to buy food, medicine, etc. People who have run through their savings, lost the equity in their homes (and have probably lost the homes as well), and tapped out every source of support are in no position to say, "But this unemployment check isn't permanent income; so I should not adjust my spending because of it."
Whatever one might think of the predictive value of the PIH in general, therefore, it is simply irrelevant to the situation of the long-term unemployed. Whereas we need empirical evidence to predict the effect on consumption of certain types of income changes, there is no doubt how the long-term unemployed will respond to receiving an extension of benefits. They will spend it all. That is known as "economic stimulus."
Even to describe the misapplication of the PIH to this situation is to ridicule it. That is not to say that Will understood the actual theory or the contexts in which it should and should not be applied. It is a talking point that somehow bubbled up into his discussion one day. It has the advantage of being obscure and sounding important. Even so, it is a ridiculous argument to offer in support of a callous effort to score political points on the backs of the people who have been most harmed by the recession.