“If Obamacare reduces labor supply, will it raise wages?”

by on February 11, 2014 at 2:46 pm in Economics, Law | Permalink

That is Greg Mankiw’s post title, Greg writes:

In a couple of recent articles written by smart economists, I have read the following claim: CBO says the incentives in the Affordable Care Act will reduce labor supply. If it does, then real wages will increase.

That sounds like reasonable, textbook economics. But I don’t think it is true. The problem is that the logic is entirely partial equilibrium. It is holding everything else constant. But that is surely not right in the long run. Lower wages mean lower income, which means lower saving, which means lower investment, which means a lower capital stock, which means lower productivity, which means lower labor demand.

Perhaps the easiest way to think about this issue is in the context of a Solow growth model. In the Solow model, the steady-state real wage is a function of technology, the saving rate, and the population growth rate. If labor supply per person suddenly falls by, say, 2 percent and stays there, the real wage will rise initially, but it will eventually return to its former level. Steady-state income per person falls by the full 2 percent.

One effect that might occur is a change in the composition of labor income. If the Act reduces labor supply primarily among the low-skilled, while not having that effect among the highly-skilled, then we might get a change in the relative wages of skilled and unskilled. But an overall increase in real wages seems unlikely.

In an increasing returns to scale model, of course, this problem becomes worse.

1 Sam February 11, 2014 at 3:06 pm

This is the same mistake immigration opponents make but in the other direction. A 2% increase in labour supply from immigration raises steady star income per person by the full 2%. The outcome improves when returns to scale are added. If the immigration is high skill (my Canadian experience) then relative wages will change favouring income equality. The labour supply restrictions from Obamacare will also tend to raise low skill wages, favouring greater equality. Combine Obamacare and highskill immigration and you’ll have higher average incomes plus greater income equality.

2 Peter Schaeffer February 11, 2014 at 3:45 pm


“A 2% increase in labour supply from immigration raises steady star income per person by the full 2%”

No. It raises total output by 2% with all of the incremental income going to the immigrants. Returns to scale could cause both immigrant and native incomes to raise. Congestion effects could cause native incomes to fall.

Given the realities of American life (gridlock, housing hyperinflation in neighborhoods with “good” schools, etc.) which do you think is larger? Congestion vs. rising returns?

3 Peter Schaeffer February 11, 2014 at 3:46 pm


The only way your model could work is if immigration caused the natives and immigrants to expand their per-person labor supply (hours presumably) by 2%.

4 Gene February 11, 2014 at 3:08 pm

It seems like a relatively silly question to me. If people drop out of the labor force and wages increase then some/many/most will be induced back in to the labor force—it is that Opportunity Cost thing-y. Employers know this and unless the exit is permanent it seems they (employers) still have the labor pricing power. Once again, people forget there are two sides to supply and demand.

5 Willitts February 12, 2014 at 12:12 am

Typical tail chasing of comparative statics. No, they are NOT induced back in because they were induced out. The people who STAYED enjoy the higher wages. The higher wages and lower quantity of labor is the result of shifting supply to a new equilibrium.

This, of course, is predicated on whether we mean a change in supply (shifting supply curve) or changing quantity supplied (shifting demand curve). People might “leave the labor force” by either method.

Workers who stay earn higher wages because their marginal productivity per dollar of revenue is greater.

6 Jay February 11, 2014 at 3:10 pm

“which means a lower capital stock, which means lower productivity”

The obvious error here is that Proggers believe all productivity comes from labor. Don’t believe me? Ask them how they derive their measure for the growth of average labor productivity from 1970 to present.

7 Michael February 12, 2014 at 9:00 am

As a “Progger”, I can tell you that you’re wrong–we don’t think productivity gains come from labor. Nor does Krugman, hence the rise of the machines problem.

8 Jan February 11, 2014 at 3:24 pm

Mood affiliation. If he had added just one more “which means” this post would count as comedy.

9 Ryan February 11, 2014 at 3:47 pm

Someone help this ignorant (talking about me) no-nothing, how we get from: “But that is surely not right in the long run.” to the assertion “Lower wages [which will be the result due to the reduced labor supply] mean lower income,…. ”

For extra points, don’t use “partial equilibrium” in your answer.

10 Ryan February 11, 2014 at 3:49 pm

Told you I was a “know-nothing” … edit!

11 derek February 11, 2014 at 4:33 pm

Mankiw is not being very clear, but it seems like he is ceding that wages will indeed temporarily go up. However, since wages, at their root, are production being distributed among all the workers, then per capita income, if not wages, will eventually go down since there are fewer workers to create production.

12 uffs February 11, 2014 at 3:24 pm

Call me crazy, but I really wish we lived in a world where wage level changes could be fully accounted for by the science that studies the economy. Perhaps I could settle for knowing simply which direction wages would move given certain policies?

Are we really this ignorant? Where’s the empirical data on any of this?

13 joan February 11, 2014 at 6:09 pm

In started 1970 women joining the labor for in larger numbers causing the labor participation rate to increase to about 68% by 1990 here is what happened to wages. http://visualizingeconomics.com/blog/2013/3/4/wages.

14 uffs February 12, 2014 at 4:30 am

Thank you.

15 rayward February 11, 2014 at 3:29 pm

The mind reels. The question is whether a reduction in labor supply will increase wages, and Mankiw responds with the truism that lower wages mean lower income. Lower wages? I thought the issue was higher wages. Mankiw then complains about the static model, but proceeds to say that if the labor supply falls by 2% so will per person income because of technology, population growth, and the saving rate. Does Mankiw teach the dismal science? Is it not possible, dynamic, that a reduction in the labor supply could increase labor efficiency (all those geezers retiring) and increase per person income? Mankiw impressed me with the speech he gave last year at his son’s high school graduation (he posted the speech on his blog). He’s a wise man, and a very good father. But sometimes he reminds me of a college freshman who has taken a few economics courses (or philosophy courses or finance courses or whatever) and is the expert in late-night dorm room discussions.

16 Al February 11, 2014 at 3:56 pm

Mankiw is talking about aggregate terms.

17 Willitts February 12, 2014 at 12:16 am

Your reeling makes a good point. There is simultaneous bandying about of micro partial equilibrium models and macro general equilibrium models. I haven’t looked at either of these arguments, but I would not be quick to accuse Mankiw of a rookie error. His hip shooting is better than our rigorous modeling.

Krugman, on the other hand, will deliberately make errors when it is essential to making his point.

18 Michael February 12, 2014 at 9:13 am

He reminds you of a college freshman? Really? May I ask what your credentials are?

19 Frederic Mari February 11, 2014 at 3:30 pm

I don’t really have a strong opinion on lower labor supply means increased wages. Maybe, maybe not, depending on a whole lot of other factors would be my prior.

But one thing that annoys me is the lower savings equal lower investment thingy. Fuck a duck. Savings have very little to do with investment. If the investment opportunity is worthwhile, the money gets created, whether there was initial savings or not. I would have thought an economist would know that.

20 CBBB February 11, 2014 at 3:53 pm

Don’t assume an economists knowns anything, especially Greg Mankiw.

21 Jay February 12, 2014 at 11:04 am

Productive post as always. What are you credentials?

22 derek February 11, 2014 at 4:36 pm

An investment is worthwhile if its return is high enough to beat the interest being charged on it (usually you would also say it has to be better than whatever the return you would get investing it somewhere else at relatively low risk, too). Since the interest rate is determined by how much money is available to be invested, reduced savings causes interest rates to rise, making fewer projects worthwhile investments.

23 Frederic Mari February 12, 2014 at 12:57 am


That might have been true during times of fixed money supply but the US has had a very low saving rates for a long time and that didn’t stop the interest rates going down.

These days, I would say that the Fed policies and ST IR fixing have a lot more to do with the IR than savings.

Even with that caveat, my personal experience of investment is that corporate leaders may use NPVs etc for projects but that’s basically ass-covering. Projects get done (or not done) not on IR or tax rates (except in the case of projects done to reduce tax liabilities) but on general ‘feeling’ of whether a given project will be ‘profitable’ i.e. usually a generally growing economy is more favourable to investments than one where IR are low.

24 Michael February 12, 2014 at 9:14 am

Savings is investment and investment is savings.

25 Michael Foody February 11, 2014 at 3:33 pm

This sort of argument, where you follow chains of causality through second and third order pull effects until you find a fourth or fifth order push effect, can obscure virtually any economic intuition.Bringing them up is probably a useful thought experiment generally and supports the ‘lowercase c conservative’ “don’t place too much trust in elegant models” view, but I think it’s a bit unlikely that he would use the same line of reasoning for policies that his political affiliation supports.

26 Al February 11, 2014 at 8:42 pm

Fair point. Mankiw’s second paragraph could be an argument for increasing the minimum wage.

27 Devin Lavelle February 11, 2014 at 3:34 pm

It seems to me those negative consequences only follow if you assume full employment, as you do in the long run, when we’re all dead.

If you assume there is a surplus of labor, no productivity will be lost shifting the folks who are choosing not to work under Obamacare out of the labor market and replacing them with those who are currently unemployed. Productivity stays exactly the same. Presumably then the decrease in labor surplus would have a minor positive impact on earnings, but for the sake of argument, let’s assume earnings remain neutral.

In the short run, while we’re alive, the impact on private investment is neutral (assuming no impact on earnings — an increase in earnings would increase private investment), since the only change is who is a shift in who is receiving the earnings. Public investment should increase, though, as less money is needed to be spent on the unemployed/impoverished (Alternately, the money could be shifted back to taxpayers, thus increasing private investment). This increase in investment would then set our economy on a more favorable trajectory into the long run, when we achieve full employment because, again, we’re all dead and, thus, no one is unemployed.

28 Mrs. Renard February 11, 2014 at 5:40 pm

“and replacing them with those who are currently unemployed. Productivity stays exactly the same.”

The above assumption is the flaw in your argument.

You assume that those that employers passed over (the unemployed) in favor of the now formerly employed (those that now choose to stop working due to benefits received under the ACA) will be able to produce at the same rate. There may be a reason why they were passed over in the first place in favor of the ones who were hired. Possibly those who were employed were hired because they were more productive? This is a plausible reason for an employer to hire person A instead of person B. If this is the case, productivity declines.

29 GiT February 11, 2014 at 6:00 pm

…and then rises again if the new employee is a younger person who has been under- or un-employed and not been given the opportunity to develop the skills he or she needs in order to some day be a highly productive borderline retiree holding on to their job for the health insurance.

30 Mrs. Renard February 11, 2014 at 7:15 pm

Based on the above, you believe that eliminating more of the “highly productive borderline retiree(s)” will make the economy more productive in the long term? Would we see this same gainful effect from burning down a few factories or office buildings? After all, the construction industry is in dire need of jobs.

“if the new employee is a younger person who has been under- or un-employed and not been given the opportunity to develop the skills he or she needs in order to some day be a highly productive borderline retiree holding on to their job for the health insurance….”

In the meantime, as we await the blossoming of this hypothetical “under-employed” employee, you agree productivity is lost?
With this loss comes lost savings, lost investment, etc. and the loss compounds in the economy over the long term.
With lower investment, less new employees with untapped potential will be hired in the future.
These new hires represent a kind of investment, after all.

In the no-ACA future the highly productive, now early-retiree’s productivity fuels investment in new employees. The result is increased productivity in the long term.
In the ACA future, in your best case scenario, the highly productive retiree is replaced by an eventually, equally productive new employee. Productivity stays the same in the long term.
The no-ACA future is still more productive. Short-term or long-term, productivity is lost.

Another view is that the “highly productive borderline retiree” represents an investment that is destroyed by the ACA.
Justifying destruction in the name of having something to replace discards the value of what was destroyed in the first place.

31 GiT February 12, 2014 at 1:35 am

But long term un- and under- employment also destroys and discards value. Labor market hysteresis – it’s been somewhat topical lately. Maybe you’ve heard of it.

But then I wouldn’t expect someone whose argument boils down to the “broken windows fallacy” fallacy to get the whole underemployment equilibrium thing.

32 Nathan W February 12, 2014 at 4:00 am

Human capital and physical capital are very different. They accumulate differently, they “depreciate” differently, and are different in many other respect.

I think the reasoning of long term gains arising from more young adults actually getting a chance to start a career is pretty solid.

33 john personna February 11, 2014 at 3:41 pm

Mankiw blogs without comment so that he can say things that have obvious responses. As in, “were you for a higher minimum wage yesterday?”

34 collin February 11, 2014 at 3:43 pm

If the labor supply decreases, then wages don’t go up? Oh Brother that low wage workers get higher wages on the benefit of the capital owners! Now, will the ACA labor supply drop increases wages? The effect is too minor to mention although it could mean more teenagers back in the workforce and the a long run decline in part time labor.

Why is there such an assumption that only low wage workers leaving? (This trend of part time has been occurring a long time here.) I know a lot married couples where one earns the income and the other gets the benefits. I suspect that we going to see a continued slow movement back to one income families and ACA supports that. My prediction is economist will call this the “Home School Revolution”:)

35 Willitts February 12, 2014 at 12:21 am

You’re getting confused in all the details. The wage of the marginal worker goes up because marginal productivity per dollar of revenue is rising. However, the total wage bill, i.e. QL x W, will decrease. In other words, aggregate wages will decline.

I think. I’m sure elasticity of labor supply and demand influences the outcome. But Mankiw is thinking in a GE Rube Goldberg.

36 Matt February 11, 2014 at 3:47 pm

All Mankiw is saying is that the ACA-induced reduction in labor supply implies a deadweight loss. That shouldn’t be too controversial.

37 john personna February 11, 2014 at 4:51 pm

We hear that these newly independent workers were actually more interested in insurance coverage than take home pay. You suggest that we suffer a loss as those folk do no longer work with that fear motivation. OK fine.

38 TMC February 11, 2014 at 6:00 pm

“newly independent workers” Is that ACA speak for unemployed?

39 john personna February 11, 2014 at 10:36 pm

Unemployed are looking. An established definition.

40 Turkey Vulture February 11, 2014 at 10:50 pm

Is that the source of most of the fewer hours worked? I believe the CBO report said it was the subsidies that were most important. I am not even sure they took the “they just needed access to an insurance plan” people into account. But people spinning the report as a good thing are arguing as if this is the entire effect of the ACA as predicted by the CBO.

41 Willitts February 12, 2014 at 12:23 am

Yes, a deadweight loss will result in worker surplus declining. Depending on the relative elasticities of labor supply and demand, workers might bear the majority of the tax incidence.

42 John February 12, 2014 at 2:00 pm

But if the works are leaving voluntarily then this is a purely ficticious deadweight loss. Haven’t the terms of the exchanges changes in a way that allows a better trade on both sides?

Or is this purely a case that the economic analysis is rejecting Becker’s old houshold production approach and you must be part of the labor ofrce to be part of the exonomy?

43 mavery February 13, 2014 at 12:53 am

To assume this is a deadweight loss, you need to start with the assumption that ACA was itself a reduction in net welfare. Which is fine if that’s the assumption you want to make. But it’s an assumption, not a logical conclusion of basic economics.

The thing here is that the shift is a result of all things NOT being equal. New regulations, new taxes, new marketplaces, etc. give people different options. The shift in labor force participation and the new equilibrium are a result of those changes. A very real scenario is “GDP is net lower” but “People are happier/have better lives because of healthcare”. There are both plausible, simultaneous consequences, and assessing the relative magnitude of the effects is the only honest way to evaluate the law within the context of economics.

Otherwise, you’re just starting from the premise that the original steady state was the “optimum” and therefore any deviation is deadweight loss.

44 Peter Schaeffer February 11, 2014 at 3:51 pm

This debate shows how divorced from reality, economic debates are. The same folks (on the left) who assume that the economy will automatically scale up in response to massive increases in immigration, are shamelessly willing to claim that a reduction in labor supply won’t have the equivalent opposite effect.

45 mpowell February 11, 2014 at 4:47 pm

I’ve seen a lot of actual analysis that the negative economic impact of increased immigration are borne almost exclusively by recent previous immigrants. Who disproportionately tend to favor more immigration for non-economic reasons. This is different from, ‘free borders will be great!’, but its real economic data making the case, at least, for current immigration levels plus some uknown extra. ‘Both sides do it’ is not actually an argument for anything.

46 Devin Lavelle February 11, 2014 at 5:02 pm

They’re divorced from reality when they assume things that aren’t true — like full employment.

47 Willitts February 12, 2014 at 12:25 am

You’re saying that the concept of full employment doesn’t exist?

I don’t think I need to remind you that full employment does not mean everyone is working.

48 Nathan W February 12, 2014 at 4:03 am

There are always unemployed, no? People between jobs, creative destruction, and so on? Perhaps this is a game of how we define our terms.

49 Peter Schaeffer February 11, 2014 at 7:48 pm


I have seen similar claims. They amount to “Low-skill immigrants and low-skill natives are complementary, not competitive. The proof is that there aren’t any natives in construction and meat-packing.”

This ignores the fact that in the quite recent past, construction and meat-packing were dominated by natives until they were driven out.

For a serious comment on the subject, see “Borjas, Grogger, and Hanson: Immigrant and Native Complementarity” (http://economistsview.typepad.com/economistsview/2008/03/borjas-grogger.html). Quote from Mark Thoma.

“George Borjas, Jeffrey Grogger, and Gordon Hanson have a new paper, and it’s not good news for the Ottaviano and Peri result that immigration can cause native wages to increase due to strong complementarities between native and immigrant labor:

Immigrant-Native Complementarity Revisited, by George Borjas: I’ve often been asked what I think about the Ottaviano-Peri finding that there are strong complementarities between comparably skilled immigrants and natives—complementarities that lead them to conclude that immigration raises wages for many natives.

I’ve always been a little skeptical of the Ottaviano-Peri evidence. … Here’s the abstract to our new paper:

In a recent paper, Ottaviano and Peri (2007a) report evidence that immigrant and native workers are not perfect substitutes within narrowly defined skill groups. The resulting complementarities have important policy implications because immigration may then raise the wage of many native-born workers. We examine the Ottaviano-Peri empirical exercise and show that their finding of imperfect substitution is fragile and depends on the way the sample of working persons is constructed. … As an example, the finding of immigrant-native complementarity evaporates simply by removing high school students from the data… More generally, we cannot reject the hypothesis that comparably skilled immigrant and native workers are perfect substitutes once the empirical exercise uses standard methods to carefully construct the variables representing factor prices and factor supplies.

English translation: The Ottaviano and Peri data includes currently enrolled high school juniors and seniors. They classify these high school juniors and seniors as part of the “high school dropout” workforce. Their finding of immigrant-native complementarity disappears if the analysis excludes these high school juniors and seniors.

Things that seem too good to be true usually aren’t.”

50 Peter Schaeffer February 11, 2014 at 7:59 pm


A related point is that each new cohort of native workers is far (far) less skilled than is commonly guessed. The usual analysis of immigration amounts to

“Low-skill immigrants are clearly complementary to everyone I know… All of whom went to Andover, Choate, Dalton and on to the Ivy League.”

In real life, the U.S. high school graduation rate is 75% (at best, see Heckman) and worse, the “graduates” have shockingly minimual skills. At least 1/3rd of the “graduates” are below-basic (NAEP) in math or reading. Around half are below-basic in science.

Roughly half of each cohort of 18 year olds have skills (don’t have skills) that make them directly competitive with low-skill immigrants. You shouldn’t be surprised that LFP (Labor Force Participation) has plunged for young people and blacks as immigration has soared.

Of course, if an immigrant drives a native out of the labor force (statistically correct in many cases), immigration is essentailly a pure drain on the economy.

51 Nathan W February 12, 2014 at 4:07 am

For your last statement, and without addressing the rest of the issue, I would argue that you can say that may negatively impact conditions in certain segments of the labour market, but strictly in terms of output, cheaper access to inputs should increase output. According to your logic though (I think), this could negatively impact average conditions in the economy if the immigrants don’t work with sufficiently excessive vigour relative to the native worker in those same segments of the labour market.

52 mavery February 13, 2014 at 1:00 am

That analogy shows how divorced from reality you are…..

On one hand, you have people who WANT to work (and people who want to hire them!) but can’t legally. One the other hand, you have people who’d RATHER NOT work but do so out of necessity to obtain employer-provided healthcare (there are the folks opting out of the labor market according to the CBO).

My view in both cases is that we’re giving people more choice over how much labor they want to provide.

53 Yu Feng February 11, 2014 at 4:15 pm

The issue is with this statement “If labor supply per person suddenly falls by, say, 2 percent and stays there” Implicitly this is assuming people are likely to quit at a similar rate, regardless of the productivity. I would argue ZMP workers are more likely to drop out of labor force, while not necessarily worse off economically, since they did it by choice. So the average productivity of the labor force will increase. Hence higher wages.

54 Peter Schaeffer February 12, 2014 at 12:33 am


Presumably, if we shrunk the labor force down to most productive 30 million, average productivity would go up even more. However, per-capita GDP would plunge. Given the high taxes on the 30 million, their living standard would fall as well. This is not an hypothetical a point as you might imagine.

Several countries match the U.S. in per-hour output. However, hour worked are much lower. As a consequence they lag badly in per-capita GDP and taxes are indeed, much higher.

55 mpowell February 11, 2014 at 4:42 pm

The argument for free trade was partially based on the observation that the labor share of national production was relatively contant over time. But over the last two decades, it has been dropping. Mankiw is right that long term trends are as important as short term ones. But we should be careful with our assumptions. And generally speaking, assuming something equivalent to the academic concept of Full Employment is probably not at all helpful in these kinds of debates. It just assumes what you are needing to prove. Which is how things like ACA enduced lower employment will shift partial equilibria like the labor share of income. I don’t have an answer to this question and I haven’t seen an argument one way or the other. So I think right now nobody knows how the ACA will impact wages.

56 Brian Donohue February 11, 2014 at 5:05 pm

Not buying Mankiw’s logic here.

57 Julian February 11, 2014 at 5:15 pm

“Lower wages mean lower income, which means lower saving…”. Not buying this part of the argument. If the only reason you are working is because you need health insurance, chances are you are not saving much in the first place.

58 Willitts February 12, 2014 at 12:31 am

Of course you buy it – it’s simple math. When wages decline, the aggregate wage bill declines and thus aggregate savings (which is some fraction of aggregate wages) declines. This should not be controversial.

Resist the temptation to endow your representative worker with particular characteristics. Macro economics deals with very generalized people. They are not trying to predict how John Smith will behave but how the entire labor market will behave.

What percentage of workers are working ONLY for the insurance? Damned small.

59 Nathan W February 12, 2014 at 4:11 am

But the ones who will leave the labour market due to improved health insurance presumably value that insurance quite highly. That group of workers is almost definitely not saving much at all. I think it is correct here to focus on the attributes of some particular segment of the labour market (or group of workers) endowed with particular characteristics, because they are the ones we are talking about.

Otherwise stated, we are not talking about the entire labour market, we are concerned about the ones who will drop out of the labour market. I would argue that we should be very interested in their particular characteristics as opposed to average characteristics across the entire market.

60 Willitts February 12, 2014 at 11:18 am

That’s a fair point. I can’t see, though, how to resolve the analytical problem. You can endow the marginal group with any sort of characteristics you like and achieve the theoretical result you prefer. The market wage is still determined in the aggregate, and therefore the proportion of these marginal workers to the whole is crucial.

It is not compelling to isolate the effects of the law on a targeted group when we care about total effects. After all, we could have a net zero or net negative effect on both employment and health care coverage with a badly crafted law.

61 Nathan W February 13, 2014 at 3:13 am

I’m arguing that the best way to find the total effect is to focus on the marginal group. However, and this is probably not surprising, I struggle to think of a practical a priori way to do so for analytical purposes.

62 Spencer February 12, 2014 at 12:15 pm

If the people leaving the labor force are largely the elderly I do not think your assumption that they are low savers will be correct.

Of course we do not know if they will be mainly the elderly, but I think the odds favor that it will be people waiting to switch over to medicare in a short time.

63 Nathan W February 13, 2014 at 3:14 am

I don’t intend to suggest that they have no savings (if they are about to retire I certainly hope they have some savings), but rather that in a significant share of cases, savings rate on current income is low, zero or negative due to health-related needs.

64 dirk February 11, 2014 at 5:15 pm

“But that is surely not right in the long run.”

So he leaves open that it might be true in the short run. But if it is true in the short run, then you don’t have “lower wages mean lower income, etc.” His long run logic requires assuming his implicit short run scenario.

65 Paul February 11, 2014 at 5:43 pm

I wish he wouldn’t use a term like “written by smart economists.”

66 Charlie February 11, 2014 at 6:02 pm

He says lower capital stock, but he should be talking about capital per worker. Capital per worker determines the wage. It’s not clear that will be lower.

67 derek February 11, 2014 at 6:27 pm

This is weird. Wages can’t rise because if they do the folks working less will lose their subsidy. The result would be similar to the $500 allowable income for those on welfare. Short term low skilled jobs where you are by necessity replaceable.

In some jurisdictions the employment costs imposed by government make it very common for people to hold down three part time jobs. In this situation the employee would not have the option of doing a third one because it would cost thousands in subsidies.

I think it is evil, and those spouting the spin are only to be scorned.

68 Ed February 11, 2014 at 6:36 pm

Surreal.Yet massive amounts of low skilled immigration over the last 40 years is sold as a benefit to low income.

69 Willitts February 12, 2014 at 11:26 am

I agree with you that it is objectionable but it can be quite real. Immigrants enter and improve overall welfare by providing labor services to employers who profit. Consumers are better off. Part of the surplus is taxed and transferred to the low skilled unemployed natives. This latter group is less preferred as labor to employers and they have a strong preference for leisure and a high tolerance for poverty. Think of them as “retired.”

What is surreal is not that this works for everyone but that it developed primarily from allegedly well meaning policy. We have permanently consigned a tranche of humanity to depravity to the extent that preferences are acquirable.

70 JRPtwo February 11, 2014 at 6:37 pm

Even if wages rise in the short run for those still working, won’t their taxes go up to pay for the subsidized health care?

71 Willitts February 12, 2014 at 12:34 am

If a wedge is created by the imposition of a tax, then the wage paid by the employer will be higher than the wage received by the worker; that difference is the tax and yes it grows with the tax rate.

It’s hard to tell at this point whether they are talking about a tax wedge or shifting labor supply and demand curves for some other reason. It is irrelevant whether the tax is imposed on employers or workers – the incidence is the same.

72 Turkey Vulture February 11, 2014 at 6:59 pm

I think most arguments for it raising real wages are coming from a lump of labor perspective. That is the error.

73 Rusty February 11, 2014 at 7:10 pm

Not really buying the lower wages part (assuming he is referring to aggregate wages). These people leaving the workforce are going to be replaced, and the replacements are going to come from a smaller labor pool, which definitely would indicate higher per-employee wages. In order to assume that aggregate wages would be substantially lower (to the degree that the investment effect offsets the smaller labor pool) one would have to believe that a substantial number of those jobs go unfilled. Not sure I understand how Mankiw can assume that.

74 Willitts February 12, 2014 at 12:38 am

Yes, he is talking about aggregate wages and no, they will not be replaced because the quantity of labor in the new market equilibrium is LOWER.

This is simple: shift labor supply up/left, equilibrium wage rises and equilibrium quantity falls. Depending on the relative supply and demand elasticities, the wage bill could go up or down. If demand for labor is relatively elastic, the total wage bill will fall.

75 David Cushman February 11, 2014 at 7:27 pm

From a steady-state equilibrium in the Solow growth model, if the amount of labor suddenly falls (workers exiting the labor force), then the immediate effect is a rise in the real wage because the capital-labor ratio is now higher. (The real wage = the marginal product of labor which is a positive function of the capital-labor ratio.) But the eventual new steady state, determined by the saving rate, s, the labor supply growth rate, n, and the rate of technological progress t, is the same as the old one, assuming s, n, and t remain unchanged. In that case, the real wage is back where it started from. I believe this is what Mankiw meant. But suppose the exiting workers now save less, as they now have less income, suppose the government deficit tends to rise to fund the insurance subsidy. Then the economy’s overall saving rate falls assuming everyone else in the economy does not save at an offsetting higher rate. In the new steady state, the real wage is now lower because the capital-labor ratio is lower.

76 Matt Young February 11, 2014 at 7:30 pm

An economy that uncovers shortages set the price level higher.

77 tt February 11, 2014 at 7:42 pm

therefore if she weighs more than a duck….she’s a witch

78 Ray Lopez February 11, 2014 at 7:44 pm

Agree with Rayward and Derek upstream. Mankiw plays games equating long and short terms, and asserting the Solow model, which is fun to study, easy to derive from the simplest principles (Cobb function, and how labor and capital rates of change are proportional to a constant times the existing size of labor and capital, which is reasonable, namely: L'(t) = k L(t) and C'(t) = k C(t), as well as other simple but reasonable homogeneous first order differential equation assumptions).

However, the Solow model is just a model, and while in the long run it may be true (or not, see below), it does not tell you when the long run is (hence Keynes “it the long run we are dead”). It’s analogous to a physicist saying in the long run the second law of thermodynamics predicts a ‘heat death’ for the earth and life will extinguish: true yet meaningless from a practical point of view.

Finally, it amazes me that people who say technology is exogenous as per the Cobb-Douglas function, which has no real microeconomics basis in reality (see: http://en.wikipedia.org/wiki/Cobb%E2%80%93Douglas#Difficulties_and_criticisms), are just throwing up their hands and denying that you can engineer technology. But you can engineer innovation, through a sound patent system (not today’s but along the lines I have blogged about here) and make the rate of technology increase. It’s not just cutting prices and free trade and hoping Good Samaritan investors will invent (i.e. like today’s Nobel Prize winners, who invent for the fame or not for the money but for the love) that drives growth, but you can engineer growth by incentivizing people to invent. You, reader, can program. Yes you can, even if you are a liberal arts major. But nobody ever gave you a good reason to code, or society made it seem hard, so you ceded that task to geeks like me.

79 Nathan W February 12, 2014 at 4:23 am

I appreciate your inclusive tone regarding programming. Most good programmers I meet are convinced that natural talent (maybe plays a small role), whereas I argue that a solid education, long years of hard work, trial and error and progressively complex problem solving (especially an interest in it) are more likely the basis of their skills.

Then again, some people really are just naturals in programming.

I think “exogenous” usually means that we don’t really understand it well enough to model it, or that trying to endogenize the variable will make the math too complicated. I think it’s the first case for technical progress. We can define it, and measure it according to that definition, and then argue about various reasons for why it is higher or lower, or grows faster or slower, in different places. At the end of the day, I have opinions about the various arguments regarding drivers of technological progress (solid public education, access to capital and an environment which financially and socially rewards innovators), but I tend to find that evidence props up pre-existing opinions of researchers which are made clear in the way they write their results.

80 Ray Lopez February 12, 2014 at 6:56 am

Nathan W: “Most good programmers I meet are convinced that natural talent (maybe plays a small role), whereas I argue that a solid education, long years of hard work, trial and error and progressively complex problem solving (especially an interest in it) are more likely the basis of their skills. ” – that, in a nutshell is what the debate over technology creation being exogenous or not is all about. The Solow model assumes that “natural talent”, which is randomly occurring, is the driver for technological innovation, whereas I argue that innovation can be taught, like programming. If you adopt the first model, then there’s no need for patents (“inventors invent”, regardless of reward), and that’s the traditional view. If you adopt my model, you need more better patents (more incentive) to get to the next level.

BTW Mankiw himself is from the “first model” camp, and he believes a large population is the key to innovation (he has said so on his blog), since if geniuses are randomly distributed, say one in 10 million, then a larger population means more geniuses will arise. But world population has grown to 7B and soon to be 10B but as TC and others point out we have a Great Stagnation in technology. So either Mankiw and the “first model” is wrong, or, as perhaps Mankiw would argue, the 7B people on the planet have not been properly educated and given the tools to innovate. We have an education problem, not an incentive problem, they might argue, whereas I would argue we have an incentive problem, not so much an education problem.

81 Ricardo February 12, 2014 at 11:29 am

20+ years in the software industry has convinced me that no, not everyone has the ability to write software. I’ve seen lots of intelligent, well-intentioned, hard-working people who just couldn’t get it.

82 JonFraz February 11, 2014 at 8:26 pm

I would think this would most likely affect the higher skilled workers: people who have enough savings (or a spouse with a good job) that they can live without wage/salary income. The poor and modestly skilled won’t have that sort of cushion: they’ll still need to work to keep a roof over their heads.

83 weareastrangemonkey February 11, 2014 at 8:54 pm

Mankiw, doesn’t seem to know much about economics.

The ACA improves life for those people who are not working. Hence, employers need to offer higher wages than they would otherwise to induce workers to accept a job. Capital does not decrease one for one (as it would in a Solow model with a reduced population) because the unemployed workers are still around to accept jobs should a good enough job come along. This means that as the number of vacancies (stock of capital) falls it becomes easier for employers to find and keep new employees. This increases the returns to capital and so partially offsets the fall in the labour supply leaving workers with higher wages.*

That is, it is utterly wrong to claim that capital stocks will decline in tandem with the unemployment rate. Mankiw treats one extra unemployed person as being the same as one less person. But it is obvious from the above analysis that unemployed people are not the same as no people.

These results follow from the standard models of labour markets (see Rogerson Shimer and Wright 2005 for a survey). Mankiw must be aware of these results as must TC, this is textbook stuff. Why have they ignored it?

*Note that the higher wages do not mean that GDP per capita won’t fall. However, it is possible that GDP per capita can increase. Workers with better unemployment benefits can afford to be more picky about what jobs they accept. This means that people rather than rushing to take a job which they do not fit very well wait for a job that is a good fit. This can, under some conditions, increase productivity (see Acemoglu and Shimer 1999 for example) but the jury is very much out on whether it does in fact increase productivity.

84 weareastrangemonkey February 11, 2014 at 8:59 pm

“Mankiw, doesn’t seem to know much about economics.” That is not true, don’t know why I said it. I still think he has got it wrong on this one.

85 Matt Young February 11, 2014 at 9:19 pm

Mankiw uses Solow, so the answer is automatic. Before the change, the distribution of prices must be bell shaped, Solow requires it. Then Mankiw claims a set of unsustainable low priced producers make the distribution skewed, not bell shaped and mean price less than median. Solow then requires bell be restored, the mean gets higher, the median lower. That is a higher average price. Built in by the assumptions.

86 Matt Young February 11, 2014 at 9:33 pm

So where did Mankiw start? From the original equilibrium or the middle disequilibrium?

87 Dan K February 11, 2014 at 10:00 pm

Probably too late on this one to get a response, but here’s my thinking / question. I’ve been thinking of the wage impact of this from a “change in relative negotiating power between capital and labor” perspective. I know (think?) that the share of wages going to capital has increased dramatically relative to historical norms. Could both sides of this be “right” when talking about the relevant populations that they implicitly are talking about.

Mankiw: Overall output (or income) goes down by the 2% in the long run
Me: Labor gets a bigger share of a slightly smaller pie, so wages – if not total income – go up (also the whole massive un/underemployment thing, and the offsetting labor mobility thing)

So yes there may be a “loss” to the economy, but there would be permanent distributional shifts – in addition to the utility improvements I see from decoupling health coverage from employment

88 Turkey Vulture February 11, 2014 at 10:56 pm

What is the level of labor-market-distorting welfare benefits at which real wages will begin to fall with an incremental increase in benefits? They can’t increase forever even if you posit that they would increase here. Eventually nothing is being produced.

89 Alec Wright February 11, 2014 at 11:01 pm

My A in macroeconomics didn’t prepare me for this post and comments. Personally I think the economy, even if it does take a hit, owe’s it to labor that if they get cancer they should be cared for. They are after all, the ones doing the crap work. Sesame street prepared me for that.

90 Turkey Vulture February 11, 2014 at 11:18 pm

“Cared for” can mean a lot of different things, from hospice to every experimental cure in existence.

91 Willitts February 12, 2014 at 12:40 am

To euthanasia
To neglect in a public hospital
To neglect from long waiting lines
To placebos

92 Nathan W February 12, 2014 at 4:26 am

Why work for a system that will leave you high and dry when you most need it? Many Canadians continue to place very high value in the institution of public health care, not because it is good or bad for the economy, but because it is good for people.

Then again, Alec raises a very valid point. Tax dollars can’t fund every latest innovative and expensive method, and in any case not enough doctors are familiar with the latest methods to implement them for everyone.

93 Nathan W February 12, 2014 at 4:27 am

Sorry, Turkey Vulture’s point about innovative/experimental cures is what I mean to say.

94 Lord February 11, 2014 at 11:29 pm

The CBO has it that total wages are reduced but total hours are reduced more, delivering higher wage rates but less total wages. The Fed will adjust that to keep total wages approximately constant for constant prices. Fewer will work but they will be higher paid, but not by much, and probably not noticeable.

95 Willitts February 12, 2014 at 12:41 am

I feel dirty after reading these comments.

96 Brandon Berg February 12, 2014 at 3:38 am

Isn’t it pretty obvious that those induced to leave the labor market will tend to be skilled workers? The preconditions of being induced to leave the labor market by Obamacare are having a job which provides health insurance and also having enough money to quit working and still pay all other living expenses plus the Obamacare premiums. There aren’t a lot of low-wage workers who meet those criteria.

97 Brandon Berg February 12, 2014 at 3:39 am

If I had to guess, I’d say that this will mostly subsidize early retirement for upper-middle-class workers in their late 50s and early 60s.

98 Nathan W February 12, 2014 at 4:29 am

Perhaps many low wage workers will quit their third part time job, while the upper middle class folks you refer to will take early retirement.

Very interesting though, I hadn’t been thinking of those people.

99 Layperson February 12, 2014 at 8:20 am

This is fantastic! According to Mankiw raising wages will decrease demand for labour but decreasing supply of labour will not raise wages.

100 CBBB February 12, 2014 at 10:09 am

Microeconomics is absolutely fine though, the micro people have it ALL figured out

101 Boonton February 12, 2014 at 1:47 pm

Lower wages mean lower income, which means lower saving, which means lower investment, which means a lower capital stock, which means lower productivity, which means lower labor demand.

Perhaps the easiest way to think about this issue is in the context of a Solow growth model. In the Solow model, the steady-state real wage is a function of technology, the saving rate, and the population growth rate. If labor supply per person suddenly falls by, say, 2 percent and stays there, the real wage will rise initially, but it will eventually return to its former level. Steady-state income per person falls by the full 2 percent.

I think the fallacy here is that the theory is that some portion of people are working not for cash pay but for health benefits. Hence savings is irrelevant. Person A works and earns $6K per year and ‘spends’ it in the form of having healthcare. New law says person A gets healthcare for $6K a year from some mix of taxes and reduced spending on other health care. Person A decides to not both with work anymore but all that’s happened is consumption has been rearranged. Person Z who was working for pay now has a bit less or person X who was on Medicare and not working because he was happy with its benefits now is willing to work to make up his lost benefits.

Labor supply hasn’t really been decreased since the law doesn’t make it illegal for anyone who wants to work to work. All that happens is when anything changes, taxes, benefits, whatnot some people who didn’t care to work (or work more) change their mind while other people who were working a lot decide to cut back.

Suppose a UFO landed and offered free health care to thousands of people? In that case you very well may get a decrease in labor supply since Person A is now happy to quite the working world and enjoy the UFO’s free clinic while Persons X and Z are unaffected. But that’s an example of an exogeneous change…some amount of health care ceases to be made by the economy and is being brought into the system by the mysterious UFO. It shouldn’t be surprising then that the economy will produce a bit less than it otherwise would and if the economy produces a bit less then people are going to work a bit less. But the ACA doesn’t import health care from some mysterious ‘outside’ and doesn’t direct the system (meaning the entire economy) to produce less or more than it otherwise would. Hence the amount of people working isn’t going to change, but of course the distribution probably will in some way that’s going to be almost impossible to really detect in the stats.

102 David Cushman February 12, 2014 at 1:47 pm

In the blog entry, Mankiw has corrected the sentence starting as “Lower wages mean lower income . . .” to “Lower labor supply means lower income . . .” So now that part makes sense.

Layperson: Mankiw does state that the reduction in labor supply will initially raise the real wage. It’s in the long run that the real wage falls back to its original level because of the eventual decline in the capital-labor ratio (from a temporary fall in the rate of investment).

103 John Trevor February 16, 2014 at 10:43 pm



104 neil March 1, 2014 at 1:21 pm

“Those who make peaceful evolution impossible make violent revolution inevitable.”
— John F. Kennedy

105 nike air max 95 March 13, 2014 at 5:07 am

”I really like to improve in each match,” Berdych said. “Our goalkeeper kept us in the game and we scored late, when it is hard for anybody to come back into it. It’s a great end to an important run of fixtures.” Jose Fonte scored the only goal as Southampton improved their prospects of securing European football by winning a tight encounter against Hull at the KC Stadium.

Comments on this entry are closed.

Previous post:

Next post: