Tag: consumption smoothing

An Obscure National Holiday

I was informed earlier today that today is, in fact, national underwear day.   That said, here is a piece I wrote on life-cycle consumption not too long ago:

Back in the day, Modigliani and Brumberg (from their perches in Urbana-Champaign!) posited that individuals smooth out their consumption over the course of their lifetimes. In other words, total individual consumption expenditures are pretty stable, or smooth, from year-to-year, rather than having individuals curb consumption in one year to pay for big expenditures in the next. The big-picture implication is that individuals base their consumption spending on their expectations of lifetime earnings.  So, if I expect to make a lot of money years from now, I will spend at higher levels now, even if I don’t have it yet. As a result, the young and the old spend more than they make, whereas the middle aged make more than they spend.

The Modigliani and Brumberg work is now known as the Life Cycle Hypothesis, and it is a seminal contribution for a number of reasons.  First, it is a micro model that has significant macro implications –aggregate consumption depends on (expected) lifetime income, not current income.  It also implies that government deficits are a source of fiscal “drag” on economic growth.  You can check out more on Modigliani and his contributions at The New Palgrave Dictionary of Economics (available at campus IP addresses; otherwise, Google it).

Even if people spend the same total amount of money every year, however,  they will probably be some variation in the items they actually spend it on.  And empirically, of course, this turns out to be the case. Exhibit A: The Atlantic Monthlyhas a fascinating set of figures showing how U.S. consumer spending on various goods and services ranging from booze and smokes to lawn and garden services to men’s furs vary by the age of the consumer.

Presented without comment
Send Grandpa some new drawers

The figures are instructive.

First off, it appears that men pour increasing amounts of money into their undergarments as they age, reaching “peak underwear” at around age 50.  The average male aged 45-54 will drop about $120 on his drawers during that ten-year stretch. After that, underwear spending falls like a stone, and by age 75 or 80 it appears that most men are only spending a couple bucks a year on those closest to them.

At the same time, however, there is a decided uptick in spending on sleepwear/loungewear. I wonder what’s going on?  (Seems like a job for the Economic Naturalist).

In addition to these brief insights, the graphs seem to corroborate some intuition about how spending changes. For example, it seems that people in their late 20s and early 30s start dropping money on childcare services, which temporarily cuts into the amount spent going out boozing. I guess kids and the nightlife are substitutes, not complements.

It is also noteworthy and possibly surprising that 70-year olds spend as much on the sauce as 20-year olds do.

Or, perhaps that isn’t surprising.

As a bonus, some clever interns at The Atlantic have peppered each graph’s url with sometimes amusing, sometimes trenchant, and sometimes bordering on subversive commentary.

Well played all around.

Consumption Smoothing with a Non-Zero Probability of a Robot Uprising

True, but how will it affect your current consumption patterns?

One of the assumptions underpinning the life-cycle consumption hypothesis discussed in the previous post is that consumers have some beliefs about how long they (the consumers) will live.  If I expect to expire at age 50, for example, I might choose to spend my money stockpiling quality underwear at an earlier age.  But if I expect to die that young I would also expect to have lower lifetime earnings, which would potentially affect both my consumption levels and the levels of “excess savings” that I carry with the intention of bequeathing it to my little ones, if any (savings, not little ones). The important point, of course, is that the change of the expected terminal date would also cause a change in lifetime consumption patterns.

Another possibility that we should probably consider is that a robot uprising would pit machine against man, and wipe out civilization as we know it.  This would be different than simply expecting to die young, of course, because now not only will I not be around, my offspring won’t be around, either.  This little wrinkle completely wipes out any bequest motive I might have.  The comparative static results for both of those, I imagine, point toward greater consumption today. I would further conjecture that the greater the probability of a future uprising, the more it affects your current expenditures.

On this point, you probably have a lot of questions, as do I.  In particular, is this robot uprising likely to be anticipated by some but not others, is it common knowledge and we all see it coming, or is it a completely unanticipated shock to everyone?  These are important  because if I anticipate a possible robot uprising before the credit markets do, I can borrow heavily before interest rates spike. This is clearly the appropriate strategy and it would allow me to consume at levels well beyond what my lifetime earnings would support. In a macro model, of course, this would come out in the wash because my increased consumption would be offset by the lenders’ decreased consumption.

At any rate, if you think I’m just some lone professor thinking about these issues, you would be wrong. As just one example, Cambridge philosopher Hew Price is also very publicly concerned.  But rather than sitting idly by and awaiting the robot apocalypse, Professor Price is helping to get a jump by founding the Center for the Study of Existential Risk. And here’s why:

“It seems a reasonable prediction that some time in this or the next century intelligence will escape from the constraints of biology.”

Escape from the constraints of biology? I’m not sure what that means, but it sure doesn’t sound so good.  What’s more:

[A]s robots and computers become smarter than humans, we could find ourselves at the mercy of “machines that are not malicious, but machines whose interests don’t include us”.

Those snippets are taken from this BBC piece.

Regular readers of the Lawrence Economics Blog know that I worry what robots might be up to, and I have warned you that if push ever came to shove, you can bet that the robots won’t fight fair.

Robots taking your job may well turn out to be the least of your worries.

Consumption Smoothing and Peak Underwear

Back in the day, Modigliani and Brumberg (from their perches in Urbana-Champaign!) posited that individuals smooth out their consumption over the course of their lifetimes. In other words, total individual consumption expenditures are pretty stable, or smooth, from year-to-year, rather than having individuals curb consumption in one year to pay for big expenditures in the next. The big-picture implication is that individuals base their consumption spending on their expectations of lifetime earnings.  So, if I expect to make a lot of money years from now, I will spend at higher levels now, even if I don’t have it yet. As a result, the young and the old spend more than they make, whereas the middle aged make more than they spend.

The Modigliani and Brumberg work is now known as the Life Cycle Hypothesis, and it is a seminal contribution for a number of reasons.  First, it is a micro model that has significant macro implications –aggregate consumption depends on (expected) lifetime income, not current income.  It also implies that government deficits are a source of fiscal “drag” on economic growth.  You can check out more on Modigliani and his contributions at The New Palgrave Dictionary of Economics (available at campus IP addresses; otherwise, Google it).

Even if people spend the same total amount of money every year, however,  they will probably be some variation in the items they actually spend it on.  And empirically, of course, this turns out to be the case. Exhibit A: The Atlantic Monthly has a fascinating set of figures showing how U.S. consumer spending on various goods and services ranging from booze and smokes to lawn and garden services to men’s furs vary by the age of the consumer.

Presented without comment
Send Grandpa some new drawers

The figures are instructive.

First off, it appears that men pour increasing amounts of money into their undergarments as they age, reaching “peak underwear” at around age 50.  The average male aged 45-54 will drop about $120 on his drawers during that ten-year stretch. After that, underwear spending falls like a stone, and by age 75 or 80 it appears that most men are only spending a couple bucks a year on those closest to them.

At the same time, however, there is a decided uptick in spending on sleepwear/loungewear. I wonder what’s going on?  (Seems like a job for the Economic Naturalist).

In addition to these brief insights, the graphs seem to corroborate some intuition about how spending changes. For example, it seems that people in their late 20s and early 30s start dropping money on childcare services, which temporarily cuts into the amount spent going out boozing. I guess kids and the nightlife are substitutes, not complements.

It is also noteworthy and possibly surprising that 70-year olds spend as much on the sauce as 20-year olds do.

Or, perhaps that isn’t surprising.

As a bonus, some clever interns at The Atlantic have peppered each graph’s url with sometimes amusing, sometimes trenchant, and sometimes bordering on subversive commentary.

Well played all around.