Bryan Caplan  

Question from My Last Ph.D. Micro Midterm

Edward Leamer on Housing and M... Upcoming Appearances...
Here's my favorite question from my 2009 Ph.D. Micro midterm:

People occasionally argue that Western consumers are virtually "satiated" - before long, they will have everything they want.  Assume this claim is correct, and that labor productivity continues to improve.  Describe the general equilibrium consequences for output, employment, wages, and real interest rates.  Carefully explain your reasoning.

Most of the students did poorly on this question, so let me give you an extra hint: Students would have done better if I added the following emphases in the first sentence:
People occasionally argue that Western consumers are virtually "satiated" - before long, they will have everything they want.
OK, what's your answer?

I'll post mine tomorrow.

Comments and Sharing

COMMENTS (40 to date)
Neal W. writes:

Ummm... employment wise you will see a return to single household income.

adina writes:

I'm a bit confused about the instruction to assume the claim is correct. Do we assume that people will eventually have everything that they want now (and can change their preferences, once today's wishes are fulfilled) or that they actually will feel that they've obtained everything they want, once that blessed day comes? In other words, may we invoke the concept of the hedonic treadmill, or other observations of behavioral economics?

BL writes:

Ugh, this brings flashbacks of my Micro candidacy exam. Questions that looked easy always ended up killing everybody - misinterpret/misread one point and everything you wrote is WRONG.

Matt writes:

Output growth will slow till it meets the population growth rate.

Employment will fall as productivity per person rises so that output growth stays with population growth.

Wages per person rise.

Real interest rates...I don't know. They go down?

Zac writes:

@adina- I think (could be wrong) that is what Bryan is trying to highlight with his emphasis. I assume that the students thought the claim that they should assume is true is that people will eventually have everything they want pretty soon from now as long as the relevant factors stick to trends. But the claim is only that we're virtually satiated.

I say the consequences would be a huge boom in output as people expected we were getting closer and closer to satiety. We'll have near full employment and high wages, as unused resources when we're this close to satiety would be wasteful. Demand for money should skyrocket, we would expect people to exhibit a very high time preference (after all, we will soon have everything we want, but not right now) so interest rates will be very high. Since we'll all have everything we want before long, we will rationally focus on fulfilling our remaining desires by increasing output.

I see this question as: What happens when you are almost full from dinner, maybe 1 or 2 more bites away from that bliss point of satiety? You go ahead and polish it off quickly. Runners usually sprint to the finish of a race. After all, a built-in assumption in this claim is that people are interested in having a certain bundle of things, and that it is possible to become fully satiated of your desires.

Of course this wouldn't be the first time I got something horribly wrong on a Caplan exam, so I am looking forward to his solution.

RL writes:

The thing I find interesting is that you want your students to focus on VIRTUALLY and BEFORE LONG, yet you also want them to work out a general equilibrium situation. If the difference between "virtually satiated" and "satiated" matters, and the difference between "now" and "before long" matters, it would SEEM the situation is not one of equilibrium.

(See how confusing economics can be to the non-economists)

Daan writes:

Matt is right I think,

although adding that it is hourly wages that will rise and total income per person to stay constant.

Combining this with more leisure, total utility keeps on rising. Equally with the rise in productivity

Bob Knaus writes:

Swift advanced the same arguments in Gulliver's Travels back in 1726. See Chapter VI of A Voyage to the Houyhnhnms.

From this, I think we can assume that for any future time of practical interest (say the next couple of centuries) consumers will still be virtually "satiated" - and before long, they will have everything they want.

How this affects the economic indicators Bryan lists I couldn't say for sure. But my guess would be "Not much."

Les writes:

First, the assumption of consumer satiation is in direct conflict with the economic principle of scarcity, and seems very unrealistic. However, let us make the required assumption and try to answer the question.

With increasing labor productivity and stagnant demand,there will be decreased output, employment, wages, and real interest rates.

After a brief lag, consumer disposable income will decrease, and so consumer satiation will cease to prevail.

Once consumer satiation no longer applies, output, employment, wages, and real interest rates will once more recover after time absent government interference)and recovery will occur. After a lag, we will get back to normal.

AdMi writes:

I guess this depends, to some extent, on whether we should view consumption as a stock or as a flow. Because there is that possibility that being "satiated" means that our whole civilization is sublimed and wants for nothing again.

My answer probably would have assumed that the satiated consumer would be one with a very low marginal utility attached to (material) consumption. So, given that, agents would consume more leisure, they would save more to consume more security, and they'd all have blogs. I'm with Matt for the most part on the equilibrium outcomes (although I couldn't have worked out the limits) -- output growth falls, employment falls, wages rise with productivity. I figured that people would want to save more, so real interests fall. Although it is a very deflationary environment (yeah?...), so maybe that's not always true.

Gabriel writes:

Solve a Ramsey-like model with period utility of the form - (c_t - c_bar)^2, i.e. satiation at c_bar. Make assumption about TFP and population growth rates. This sounds like a macro problem to me...

liberty writes:

I had the same thought as RL.

I understand both Matt's answer (the expected/traditional answer, I should think, for a question of this sort) and Zac's answer, which is creative and sort of compelling, with its behavioral intuition.

But, I still think that you cannot really ask about an equilibrium outcome of a situation in which you only inform the student about where people WILL be in the future. Both "virtually" and "before long" describe not where people's preferences are NOW but where they WILL BE.

Hence, you invite speculation of the sort that Zac proposed...But I am not sure that you can postulate an equilibrium outcome of people's future preferences. That is, unless you know how their future preferences inform their current expectations (as Zac proposes). But I am not so confident - do these consumers even know that "before long they will have everything they want"?

It is like asking what the equilibrium outcome is of the fact that a meteor will hit tomorrow. How can we postulate an equilibrium based on today's conditions, for something that hasn't happened yet?

Alex J. writes:

Once people become satisfied, real interest rates would fall and new investment should plummet. Can we still assume continued productivity growth?

Ignoto Fiorentino writes:

I think the reason it is necessary to focus on "virtually" and "before long" is because at the bliss point itself the marginal utility of everything is zero, hence the marginal rate of substitution is 0/0 which is undefined.

So consider the situation where the marginal utility of consumption is simply very small. As consumers approach satiation with goods, they must also approach satiation with leisure. Else it would pay to substitute away from goods toward leisure.

First consider an economy with no savings or investment [i.e., assume away the interest rate problem]. In this case, agents gradually work less and less and consume more and more, asymptotically approaching the bliss point of zero work and zero marginal utility of consumption. Real wages per hour rise without limit, but real per capita income is asymptotically bounded by per capita production.

[I'm assuming here that leisure remains a good up to the physical limit of 24 hours per day; if not the analysis below would have to be adjusted. With free disposal of goods, agents would reach the bliss point of leisure, and then leisure would stop rising. Increasing productivity would simply result in production of additional goods to be thrown away. Without free disposal, however, increased productivity would result in the production and then destruction of negative utility goods.]

An economy with capital and investment is a lot more complicated. Here, it is necessary to specify whether the increased productivity is labor-saving or capital-saving [e.g., is there technical innovation, or does productivity simply increase because of the accumulation of capital. I'm going to assume there's no innovation, just capital accumulation [or alternatively assume that innovation can be stylized as a type of capital accumulation with the usual diminishing returns assumptions]. One also needs to consider the possible existence of non-renewable resources but let's ignore that here.

On these assumptions, in order to have productivity growth we need to have increasing capital per agent. Thus there needs to be a positive return to investment else agents would rather consume today at a vanishingly small but still positive marginal utility of consumption. But assuming diminishing marginal returns to capital, this return also has to fall over time. What happens to the interest rate over time then has to depend on the relative curvature of the production and utility functions and resultant elasticities of substitution between goods and labor. If increased capital intensity produces goods per hour of labor faster than workers want to substitute away from goods to leisure, then interest rates need to fall to keep investment down and balance things out. But if the marginal rate of substitution between leisure and consumption falls faster than the marginal rate of transformation between capital and labor can accommodate, then interest rates need to rise to balance things out. Or do I have that backwards?

Wait -- I forgot to consider the depreciation rate of capital!...

kurt writes:

Is the claim that people might have irrational expectations about the future and its effects?

Arare Litus writes:

Summary of Question: "Assume consumers are virtually "satiated" - before long, they will have everything they want. Further, labor productivity continues to improve. What are the general equilibruim consequences for output, employment, wages, and real interest rates. Carefully explain your reasoning."

Short run: Demand decreases, as wants are filled, so output decreases. Productivity increases, and output decreases, so employment decreases. Now, before we get to wages and real interest rates we note something - consumers were almost "satiated" prior to employment decrease. Those no longer working now have wants not being met. So wages drop, as people compete to work. What does this now do to real interest rates? People will notice that, as wage drops, wants are no longer being satisfied. They will save more to act as a buffer, and spend less on current wants that are more elastic in being met (further droping wages). Interest rates will decrease as money goes into the bank, and as innovation has dropped as all needs were almost satisfied. Some of unemployed, and people making less money then they planned, and people in now less wanted industries will think that perhaps they can make new and better products. They start to think. They get ideas. They go to the bank & other investors and borrow. Products, that we never knew we wanted, get innovated. Before long NOTHING CHANGES FROM NOW (when question is posed) - we have a dynamic market that changes based on current markets, change, innovation, etc.

greenish writes:

It's already happened.

KF writes:

First, I want to focus on the satiated consumer hypothesis and then focuson how the labor productivity assumption would change the argument.

I guess the first thing I would note on the satiation point is that the economy is complex with multiple different goods and a great deal of innovation. The spending habits of consumers will likely change as innovation changes the plethora of goods available. In 1900, there was significant production of horse shoes, now there is strong production of autoparts, I don't know what the next transportation will be, but something will be obsolete. Second, the key point is that people aren't satiated yet. Generally the people who say people will be satiated are the same ones who miss the new innovations that drive future changes in consumption patterns. People might be close to being satiated, but then new innovations could disturb them and they are no longer close to being satiated. But I don't think this kind of analysis (and critiquing the assumptions) really helps in a GE framework (as I understand them).

So assuming away innovation and the disequilibrating influence it will exert, people are not yet satiated but becoming close to it. Well first what does this mean. To me, it means that the marginal utility on goods that are not "essential" is moving closer to zero. "Essential" is probably a bad way to phrase it in economic terms, but I just mean that consumption to maintain their lifestyle (food, water, etc) would not necessarily have a decreasing marginal utility. Further, you could probably say that the second derivative of whatever utility functions you were looking at in class are positive. That's just math gabolygook, but I think it means that the marginal disutility of labor should increase.

Putting together an increasing marginal disutility of labor, increasing labor productivity, and a decreasing marginal utility for "non-essential" goods (and assuming away innovation), I would argue that it would be really difficult to tell what's going to happen. Increasing labor productivity could offset the marginal disutility of labor. It depends on relative preferences and how fast labor productivity is increasing. A strong enough rise in marginal disutility of labor could offset the climb in labor productivity and force employment lower and the growth rate of output per worker could decline (and eventually output per worker could decline). In this scenario, again it is difficult to tell what is happening to wages without a model and you would need to derive what exactly would happen to wages (depends on the relative strength of labor productivity growth vs. marginal disutility of labor). On the other hand, output per worker is probably still growing which would put upward pressure on nominal wages.

In terms of real interest rates, it depends on what kind of real interest rate model is being used and how you think of risk aversion and time preference. I don't really see a reason a priori why time preferences would change, but if people will be satiated in the future, it's possible that they could prefer future consumption less than current consumption. With future consumption being valued less, real interest rates would fall.

MHodak writes:

As consumers approach satiation, total demand will level off, so total output will level off, as well. Demand for durable goods would presumably drop more quickly than for non-durable goods. Households may not need newer and better refrigerators, but they will continue to need the food that goes through them, consistent with population growth.

As demand levels off, so will investment, which means that job growth will level off. The effects of leveling job growth would be accentuated by continued increases in productivity (especially in the durable goods sectors). Investors trying capture the benefits of increased productivity in an environment of stagnant demand would lead to a net decline in demand for workers, which could slow wage growth, or even lead to wage decline.

Lower wage growth could coincide with lower demand on the part of the workers. In principle, their lower demand for stuff (satiation) should translate into a lower demand for jobs and aggregate wage growth. So, the economy could equilibrate with the same number of jobs with lower wage growth. (In a realistic case, some workers would be experiencing above average wage growth while others would see wage decreases or layoffs. Lower "wage growth" results would be for the "average" worker.)

Investors will be hurt by declining growth in demand, but helped by increasing productivity. To the extent that those two effects cancel out, demand for capital would remain constant, leaving real interest rates relatively unchanged.

So, where do I pick up my PhD?

Chris Milroy writes:

My guesses:

1) Output: as it approaches the per-person satiation point, output growth converges down to the population growth level, while total output converges to the replacement value of durable goods plus consumption.

2) Employment: we expect total employment to drop off as satiation approaches, since output becomes effectively fixed in any satiated short-run and productivity increases. However, I would expect this stock of labor to be divided among people in smaller chunks, since the relative value of consumption and leisure becomes skewed toward leisure at or near satiation.

3) Wages: since workers would have to compete for ever smaller numbers of jobs, wages would drop (and, incidentally, the maximum willingness to pay of employers would fall with the demand for goods and thus the marginal productivity). At some point, wages would hit a bottom, and then rise with productivity growth.

4) Real interest rates: as the satiation point approaches, the difference between returns to capital now and returns in the future gets large--thus, interest rates rise as people try to produce before the satiation point and then fall after the optimal investment has been made.

Tired writes:

Scarcity is postulated away. Economic theory has no response. Next question, plese.

jb writes:

If nothing else, Bryan, I think the comments here suggest that your question isn't nearly as clear as you think it is.

We have answers ranging from "it's a trick question" to "demand drops and then recovers" to "demand skyrockets" to "people stop working" to "people invent new stuff" to "it's already happened"

For me, the problem with this concept is food and health care - those will have to continue to be consumed. Electricity will still need to be generated, and natural gas/propane/etc. Water will need to flow. Cars will still need to use gas to get to the grocery store. Kids will still have to go to school. People will still get bored, and seek entertainment. Your stuff will break, and it will need to be repaired. If you have loans or rents to pay, they will still have monthly payments. I'd still have to pay for insurance (health, life, disability, car)

But that would be it, roughly - I'd see my personal expenses drop by about 10% without any need for a nightstand, or a motorcycle, or an XBox, or a cool stereo, or a guitar. Assuming everyone else has about the same amount of discretionary consumption spending, we would probably all stop buying stuff, and either invest the money, pay off our loans faster, or switch the money to service consumption instead of product consumption. (I'm assuming that we can buy all the food we want, and it's exactly the food we want, no need to 'upgrade' to fancier food)

Investing just transfers the money to some other person (most of the time), so that wouldn't really matter. assume that half of the "free money" goes to loans, half to services. Early loan payoffs would cause interest rates to drop, as banks would have too many assets. Service businesses (restaurants, dry cleaners, etc) would boom, at first.

Phase 2: Entrepreneurs would borrow money cheaply to start new service businesses. There would be a significant inflation in service worker wages. The costs of service businesses would rise to the consumer, eating up more of their surplus.

Meanwhile, loans would continue to be paid off, and there would be no factories, no product companies to replace them. The need for new stuff would drop to the population growth rate + breakage rate. Manufacturing labor would continue to decline, probably fairly rapidly. Those people would move into the service sector, equalizing prices a bit. But this would probably look like a fairly nasty recession to the people caught in the middle of it.

Interest rates would continue to drop, and more entrepreneurs would build service businesses. There would be service innovation - new ways to automate and computerize services, to reduce labor costs. To the extent that new products were being made, it would be in this arena - replacing humans with robots. This would also put downward pressure on service wages.

As people continued to pay off their debts, their debt payments would represent less of their takehome pay. They would reach service satiation at some point, and start working less, and consuming more leisure. Artistic and media-related services (TV shows, movies, music, games, etc) would experience a boom.

Over time, most non-creative service jobs would be automated away. This includes all menial labor - road repair, extracting oil, growing food, transport, etc. This would be exceptionally deflationary, I think. People would start spending their money trying to find new ways to stave off boredom - adventure dinners, travel, tutors, hobbies, etc, etc, etc.

Finally, I think it would all dissolve into everyone spending their time trying to be entertained, and entertaining others, for status, rather than money.

Student writes:

As a student who took this exam, the wording of the problem did seem a bit ambiguous. We were told to assume current "virtual satiation" and that "before long", taken to mean in the short run, we would reach total satiation. The answer it seems you wanted however assumes that we never do reach complete satiation, but the question tells us to assume that complete satiation will in fact be reached.

Jason Briggeman writes:

My guesses... There will be rapid disinvestment, seeing as future ROI will be -99%. Output and employment plunge. Demand for loans tanks, so the real interest rate falls. The effect on wages is ambiguous because demand for labor drops but at the same time it will be harder to get anyone to work.

Devin Finbarr writes:

Assuming people desired leisure, individuals would start working less hours, retiring early, taking sabbaticals, etc. A lower supply of workers would drive up wages for remaining workers. This would drive up costs for employers, thus making products more expensive. As products became more expensive, people at the margin, who were almost at the point of satiation, would need to continue working in order to afford their desired products. Thus the system would reach a new equilibrium, in which total labor was less.

Snark writes:

Discarding the more realistic assumption of non-satiation of preferences (axiom of dominance), virtual satiation would result in a Walrasian equilibrium, where all markets clear. Output would plateau, interest rates would fall toward zero, and unemployment would probably increase, as those who reach their bliss point pursue more leisure or seek out better job opportunities. As unemployment rises, wages must increase.

And because economics is so frustratingly counter-intuitive, I’m probably dead wrong on every count.

Ignoto Fiorentino writes:

A follow-up: what if the marginal utility of leisure time is strictly bounded away from zero? [That is, if even at 24 hours per day of leisure, people wish they had more time to do the things they want to do, because they are not immortal.] In that case, rational consumers could never approach satiation in goods.

See also Scott Gordon, The Economics of the Afterlife, Journal of Political Economy, 1980, vol. 88, issue 1, pages 213-14 [], who extends a similar argument to show that life in Heaven must be infinitely intense in experience but fleetingly short in duration.

Against the grain writes:

No expert but here is a simple minded answer:

Western consumers are virtually "satiated." They will have everything they want. Before long there will be no need for any labor and there will be no production because people will have everything they need. Labor productivity continued improvement is a false statement since people have everything they need so labor provides no marginal benefit and therefore has to go to Zero! Before long: output Zero, employment Zero, wages Zero, and real interest rates are irrevelent since there is no need to lend or be lent to.

I suspect that I have avoided the question by framing the answer from the perspective veiw after before long, when the equilibrium is in place.

qgambit writes:

Tired is right. Economics is the study of choices under scarcity. Postulating scarcity away turns this into a philosophy question.

I guess one can image what a post scarcity economy could look like if indeed satiation were possible.

1)Output would depend on population size and how much output per person is needed to satiate all.

2)Employment would be zero. All output would be produced by capital. All human time/effort would be spent on hedonistic pursuits (how could one have everything they want otherwise?).

3)Capital stock would need to be autonomous and self maintaining (think the robots in WALL-E).

4)Wages and interest rates would not exist. People could consume all they wanted when they wanted to.

If this sound utopian, its because any post scarcity society would necessarily have to be ("they will have everything they want").

Zac writes:

I read the question as "if a group of people believe they have nearly everything they want and that they will have everything they want soon, what are the behavioral implications? What would be the consequences of this behavior on output, wages, etc"

It seems a lot of people are trying to answer the question that starts "suppose a group had everything they want..." but this a) doesn't make any sense and b) doesn't seem to be the question that is asked.

Mike Griswold writes:

As satiation is approached, the growth rate of output needed to reach satiation will fall below the rate of productivity increase. Output will eventually equal that which is needed to maintain satiation plus population growth (a level higher than it is now). Workers will substitute leisure for labor at a rate equal to their productivity growth, so "employment" will fall. Real wages will hit a permanent high plateau. Real interest rates will equal productivity growth.

KF writes:

Oh I forgot to add something, not sure if it was addressed yet.

When we say that consumers are satiated, that means who? Is everyone satiated (or on their way)? No. The young who have little capital, likely don't have all the things they want and the old might have an overabundance. An aging population is more likely to be satiated than a young population, but as the old die off, there is a process for more people to want things and build up capital necessary to be satiated. So the writers who say that there is ROI=-99%, that is ridiculous. People will still need "non-essential" goods in this environment, particularly younger people who don't already have them.

Richard Pointer writes:

Inglehart (1990) postulates that satisfaction should allow people to pursue non-material goods. He finds that those who hold post-materialist values seek more public goods.

Ignoring some of his more ambitious claims, the marginalist would say that virtually satisfied people should seek the good with the next highest value.


Marginal utility of current output

The before "long" means that the economy will eventually shift into new sectors that satisfy new wants. But in the mean time:

Current sectors will look much like farming did over the past 100 years; the percentage of the population involved in those industries will shrink but the return to those industries will match productivity increases.

Or as Mike Griswold said! Ha! I shouldn't have wasted my time.

Back to grading.

Redbud writes:

Q: People occasionally argue that Western consumers are VIRTUALLY "satiated" - BEFORE LONG, they will have everything they want. Assume this claim is correct, and that labor productivity continues to improve. Describe the general equilibrium consequences for output, employment, wages, and real interest rates. Carefully explain your reasoning.

Positive A: Sounds wonderful. Abundance and satisfaction, with growing productivity. So output grows, and seeks new markets for surplus. Employment holds, as Westerners stay satisfied. Surplus is sold, used in promotion, or donated. Wages are flat, as all is well ;-) plus some. Real interest rates are flat, as is domestic demand.

Negative A: Sounds rough. Collapsing demand, with growing productivity. So markets shrink, causing employment to shrink. Wages are flat, as times are tough. Real interest rates are flat, as is domestic demand.

Minimal solutions from a non-economist in Kansas. Enjoy!

liberty writes:

ooh.. although I don't disagree with what I said before, I must say that Arare Litus nailed it. That is the answer you were looking for (or should have been looking for).

Floccina writes:

"virtually "satiated" - before long" Reminds me of:
How much is enough?
A little bit more.

So I would say not so much.

I would also say that if employment falls the unemployed would become further from satiated. Also demand from the poor countries should continue.

So output should continue to grow although people should work less by choice as they get closer to satiated for things. Employment should stablize. Wages should grow and real interest rates should fall.

Troy Camplin writes:

Assuming things wear out and that new technologies are always coming online and that new consumers are born, people will remain desirous of goods, so we should see increases in employment and wages from the increased demand in goods. The demand for loans for new products will keep interest rates mostly stable because demand will keep pressure on them being up, while if they get too high, fewer loans will be made, and they'll come down again.

Or must we assume an unchanging technology, wear on products, and population growth?

Richard Pointer writes:

that should have said:

Marginal utility of current output is less than the Marginal utility of current leisure.

Could this be a general description of the current economy?

Jason Briggeman writes:

@KF: Your answer admirably shows how "satiated" is not a technical term and has no remotely clear meaning. I took 'satiated' to mean that we've built the machines we need to satisfy our every material desire and it takes a trivial amount of work to proceed. (You may note that several other commenters interpreted the question similarly, as when they say Bryan is postulating that scarcity no longer holds.)

Assuming the other N-1 questions on Bryan's exam were better specified and that he allows a student to answer X

Jason Briggeman writes:

...where X is less than N-1, maybe the right answer is to skip this question.

Comments for this entry have been closed
Return to top