The two questions I am asked most often (low-skilled jobs and future inflation)

The first is what kinds of jobs will be available for low-skilled Americans in the decades to come.  I’ll be writing more on that.

The second is whether inflation is due to kick up in some kind of big storm, either in the next few years or when the major bills start coming due ten or so years from now.

Probably not.  Let’s consider a few factors:

1. The future budget situation will consist, most of all, of largely of unfunded Medicare liabilities, which to whatever extent they are met must be met in real terms.  Inflation will not make that problem go away.

2. The flow of debt is large, relative to the current stock (yes, I fully agree that is a scary thought in its own right).  That means the federal government won’t gain much, and would probably lose, by trying to inflate away the value of the stock of debt.  Furthermore a lot of the current debt is quite short-term.

3. Seigniorage revenue simply isn’t a big deal these days.

4. Everything we were taught about the monetary base is wrong in a world with interest on reserves (IOR).  A large base can sit there forever.  The price level is not proportional to the base, changes in the base, etc.  It just isn’t.  The broader aggregates, such as M2, haven’t grown so rapidly.

5. If needed, the Fed could soak up lots of the monetary base by selling assets from its portfolio.  I don’t have some utopian vision of the Fed doing this remarkably well (hard to say, this is not Fed-bashing either), but of course the Fed can make mistakes in many ways and I would not focus exclusively on that way, which is in any case part of a broader program of expectations management.

6. Every market price we can possibly look at it is forecasting low to moderate inflation.  The price of gold, by the way, seems these days to be a hedge against catastrophic risk not a hedge against inflation per se.

Please do not get me wrong, it is entirely possible that inflation will go up.  Things could change.  And even if the current deck of cards is played out, I do in fact think inflation will go up somewhat, perhaps more than markets are expecting (for one thing, I am more of a pessimist on supply bottlenecks than are many observers).  That said, I do not see any ticking inflationary time bomb.  Neither market evidence nor economic theory support such a conclusion.


OK, you answered the easy question with the most predictable answer. (I agree with you, which is why I found it predictable.)

The other question is the Big Question. My guess is that we will see a return to the past in many ways. Low skilled workers are going to end up as household servants. The relative lack of household servants among middle class Americans in the recent past has been a historical anomaly. American household life will resemble a Chekhov short story more and more.

The economic reasons are obvious, but a change in social attitudes will be required to adapt to such a different way of life. Likely, it will happen gradually and unnoticed. The maid that comes once a week now comes twice a week, then comes everyday, then starts picking up the children at school, then moves into the garage apartment and works full-time. What feels socially awkward at first quickly turns into a new status symbol. What car you drive has mattered less to one's status over the past decade. The new status will be how many servants you have.

The question then is: who are these low skilled servants? This is where it gets thorny. Nobody is going to want to hire white servants, because it would be awkward. Few people are going to want to hire black servants, because it would be awkward. Hispanics will be the servants of choice, because it will be a natural, non-awkward fit for upper middle class white Americans.

But I'm obviously not solving the whole question then. What kinds of jobs will be available for low-skilled white and black American workers? Maybe they will end up personal servants too, but it's easy to see why that would take a much longer adjustment period before people become socially comfortable with that arrangement.

If America doesn't have enough servant jobs for black and white Americans because of racial sensitivities, then, since we all know that more immigration is the solution for everything, all we have to do is to import new racially insensitive masters for the servants.

You should do YouTube videos, but in black face. Or maybe wearing a big fake mustache and a sombrero. The "Esteban El Marinero Show" would be a huge hit and probably the best use if your unique talent set.

What's the competitive equilibrium of maid versus Roomba, GoogleCar etc.?

This is an excellent question. Or phrased a little more directly, how will cheap home automation not replace servants to the same degree it will in more standard businesses? Labor saving devices are labor saving devices.

I'm not convinced that household servants will soak up the excess labor.

Agreed. It is similar to saying "in the future more and more Americans will hand-assemble products in factories".


If immigration and minimum wage restrictions were eased, I would expect precisely that...

It's a bit of a leap of faith, but it's not like we don't have a couple centuries of experience with labor-saving devices already, yet there are 100 million more jobs in this country than there were in 1950. Weird- it's like it tracks population or something.

I just have trouble getting too worked up about the "what are we gonna do with all these people?" issue.

To be clear, I wasn't saying we would have "excess labor" in the future, so much as saying I don't expect to see large job growth in the personal servant industry.

All that being said, Dirk has a point and we could well see a rise in the number of cleaning services, lawn services, painting and maintenance, etc. I just don't really see Americans heading, in mass, in the direction of employing a human Maid to live with them full time. Instead, I fully expect to seem some version of Rosey the robot maid from the Jetson's in every middle class household.

And virulent cries from the Left on how unfair it is that the poor don't have their own personal robots and the rich have way to many of them, of course.

"I just have trouble getting too worked up about the “what are we gonna do with all these people?” issue."

Brian's statement reminds me of those mid-1990s productivity studies pointing out the world had personal computers for 15 years, yet the average business had nothing to show for it. It is interesting how doubts about technological and medical breakthroughs are most serious before their fruit begins to wildly proliferate.

Perhaps a leisurely examination of the Science Channels 'How It Made' program would change his eyes.

At this time, the West is more than willing to substitute *cool* technology and capital for human labor.

Engineer Dad, not only has the number of jobs continued to grow (suspiciously tracking population) despite centuries of labor-saving inventions, but a lot of the jobs that existed back then don't exist anymore. Ask Ned Ludd. Or the buggy whip makers.

It's the seen and the unseen, and a failure of imagination, IMO, that make this trope one we can return to time and again when we are running short of real things to worry about.

Yes, and because of the disparity in robot ownership rates between rich and poor, we must provide tax subsidies to low income households so they, too, can own robots.

Our labor savings devices (1) do a job, (2) give us privacy, and (3) are not underfoot during slack usage. To have servants I'd need both space and occupations for them.

(I'd rather have a 7x24 garden robot than a 5x8 gardener for these reasons.)

And who says libertarianism is not about atomic individualism?

Don't worry, my garden robot will facebook.

I find your racial theory of servants weird. I can see how some might find black servants awkward, though I doubt that would be nearly as significant as you claim (if servants do come back the concept will probably be rebranded in a way that dissociates it with distasteful realities of history). However, on what grounds would anyone object to hiring white servants?

To your larger point, the disappearance of household servants is no more anomalous than the disappearance of ditch-diggers. They have both been superseded by technology. With a modern kitchen, it no longer takes all day to cook a meal. Cleaning and other tasks have been similarly revolutionized. And improved transportation means that outside services can be obtained far more easily. People can and do outsource certain tasks (cleaning, lawn mowing, etc.) to hired help at far lower cost than a live-in servant. The cost of labor would have to drop dramatically to change this trend.

White servants will be too expensive for all but the elite

Why? Lots of white people work at McDonald's, mow lawns, etc.

The cost of labor is already sufficiently low to make live in servants affordable. Labor restrictions prevent this exchange.

I do not know, however, if servants fell out of favor because (1) low end wages increased, (2) immigration restrictions kept potential servants out of the country, or (3) they simply became unfashionable.

I will note that au pairs are still somewhat popular, and could be considered a type of servant replacement.

"I will note that au pairs are still somewhat popular, and could be considered a type of servant replacement."

No, tot really. You aren't living among middle class American's if you are seeing a lot of au pairs. ;)

I never encountered a single au pair in nearly 30 years of life in middle-class America.

I suppose labor restrictions could be part of it. But don't analogous jobs, such as live-in carers for the elderly and disabled, exist legally in the current labor market?

I've encountered a few au pairs but never bumped into any of the au pair employers. I guess the former are more my economic cohort than the latter. :)

Most au pairs seemed western Europeans wanting to do a year abroad right after high school or college. I think there's even a specific visa category for it.

Probably the first time black servants as a social faux pas comes up in American literature is Tom Wolfe's "Radical Chic" in 1970. Mr. and Mrs. Leonard Bernstein are throwing a fundraising party for the Black Panthers on Park Avenue. Guests are anticipating a bit of awkwardness over the servant issue. But, the Bernsteins have already figured out the Future of American Servants: their's are from Chile!

If that does come to be, I'll insist that mine are college educated. 'Cause I can.

"If that does come to be, I’ll insist that mine are college educated. ‘Cause I can."

Psychology majors are the worst paid according to Time, so that would be your cheapest option, but anything in the Arts ought to be reasonably affordable. ;)

Would people prefer robot servants. They don't take vacation, don't require health insurance, and don't talk back. And they are very easy to scale. It should be much easier to "get a new robot" than to interview and vet a human one.

"Low skilled workers are going to end up as household servants."

Machines are cheaper, don't steal, don't incite racist political responses, and don't vote for labour protection laws.

They don't rise up and kill you in your sleep either, right?


Like Amy Chua's aunt.

Just make sure you buy the 3 Laws upgrade option. It's a much better deal than the under body rust protection.

There is still plenty of work to be done. There is still not a decent bakery or deli in the USA outside of the Washington to Boston megalopolis and the bakeries and delis in the megalopolis are dying out because the children of the owners have better less taxing ways to earn a living. I see no shortage of work that could be done.

wrong about bakeries:

your larger point, of course, is correct though.

"There is still not a decent bakery or deli in the USA outside of the Washington to Boston megalopolis"

[Rolls eyes at the parochial east coaster]

I would definitely hire white female servants. Oh, the joys of droit du seigneur!

On #4, I would say that basic teaching regarding the monetary base has pretty much been wrong for at least 40 years, if not longer. The change to paying IOR has merely made obvious what many Fed officials, bankers and Post-Keynesians have known for quite some time. Understanding how monetary policy operates, now under a permanent floor system, will help explain why QE has been so ineffective at stimulating inflation or bank lending.

But doesn't paying Interest On Reserves just mean you've countered rampant inflation by forcing the government to raise taxes or borrow more money to cover the inflationary effects of the extra money in the system?

Essentially you are adding money on the one hand and draining it out with the other. But you have to drain it out. If you can't raise taxes any higher and/or can't borrow money at a low interest rate the cycle fails. Once you have to start raising your interest rates to collect additional funds inflation will kick in, correct? Or am I completely wrong about this?

No, I think the ideas here are:

(1) There are people who will happily "park" their money if they can get a decent rate of return. They will not withdraw to spend on consumption. They will not withdraw to spend on investment goods. They will not withdraw to buy fixed assets (e.g. gold) in an attempt to preserve purchasing power.

(2) Parked money is not inflationary, because it is not being used to bid up prices.

(3) IOR is a way to satisfy the (1)-people. By increasing IOR, the Fed can incentivize banks to park reserves rather than issue loans.

You'll want to be a bit careful with Woj. He sounds like an MMTer

MMT = questionable assumptions + obfuscation

where questionable assumptions may be:
(1) The Fed can fix the yield curve forever
(2) Long-term stable Philips Curve
(3) The 1970s stagflation period was a one-time bad thing

Marris, I'm curious if point 3 then defeats the Fed's attempt to pump more money into the system if at the same time they're incentivizing banks to not spend it. Is the Scott Sumner argument? I'm not an economist, so I'm genuinely interested.

Yes, if the IOR rate was high enough, then it would keep banks from issuing loans. It's not clear why the Fed is paying this interest. Maybe to repair bank balance sheets?

Also yes, I believe that is one of Sumner's arguments. He has mentioned before that making IOR zero (or even negative) will be one way to get money into the economy. [His main argument is that the Fed should target an NGDP level forecast].

Quick clarification: Definitely not an MMTer. Actually an econ student at George Mason trying to combine aspects of Post-Keynesian monetary economics with Austrian and Public Choice theories of govt.

Anyways, paying IOR does not prevent the banks from making new loans. It does adjust the price of base money from ~0% to 0.25%. The main idea is that inside money (private bank money) can be lent into existence ex nihilo and thereby creates deposits. No prior saving is necessary. In order to ensure the smooth functioning of the financial payments system, the Fed promises to make unlimited reserves available at a given price (either the Fed Funds rate or discount rate). The supply curve for money is therefore horizontal and the quantity of reserves doesn't restrict lending in any practical manner.

Furthermore, the aggregate private banking system cannot alter the size of the monetary base (i.e. lend out reserves). The only option, in aggregate, is to exchange reserves for vault cash (i.e. physical currency). Paying IOR ensures banks choose to hold reserves instead of currency. The primary reason for paying IOR is not to restrain inflation, but to provide a back door bailout for the banks equaling several billion dollars per year.

As for the questionable assumptions:
1) The Fed could fix the yield curve forever, but not without imposing significant costs on society (that far exceeds any benefits). Currently though, the primary danger of negative real interest rates is not inflation but asset bubbles.
2) Post-Keynesians, including MMTers, don't believe in any type of Phillips Curve relationship or NAIRU.
3) The 1970s stagflation was not due to loose monetary policy but rather some combination of price controls, oil shocks, union wage bargaining, fiscal deficits, etc. Monetary policy is simply not effective enough to create stagflation, but fiscal policy could.

> In order to ensure the smooth functioning of the financial payments system, the Fed promises to make unlimited reserves available at a given price (either the Fed Funds rate or discount rate).

I don't think that's true. Countries like Canada have had long periods of hitting their inflation target range (1-3 percent). If reserves were provided "as needed," then this would be a very big coincidence.

I would say over the long and medium term, the central bank can target a nominal variable. Banks that don't play along get smashed. No one wants to get smashed, so everyone plays along. The result is hitting the nominal variable target.

BTW, I agree that this is one of the tenets of most Post-Keynesian schools. I think those guys are wrong.

Canada has zero reserve requirements and the banking system in aggregate usually holds zero reserves over night. If reserves are somehow necessary for and/or restrictive of bank lending, I fail to see how Canada is a good example.

As for the Fed, do you disagree that the Fed is willing to provide unlimited reserves at the discount window? If so, why has no bank ever been forced to default for being short on reserves in the post-war era (to my knowledge)? If not, than the only way to maintain the Fed Funds rate below the discount rate (w/out excess reserve and IOR) is to provide sufficient reserves at the target rate.

Canada does not use reserve requirements. It uses short term interest rates *and* a "whatever it takes" commitment to hit their monetary target. If banks believe that the CB will do whatever it takes, then they don't need the CB to *do* anything. Here is a paper covering Canada monetary policy:

The important thing is that the CB *can* pretty much do whatever it takes. They can do QE. They can (and did) open swap lines with the Fed to leverage the Fed's QE.

A bank will fail long before it becomes in danger of missing it's reserve requirements. There will be a bank run well before that. And some resolution agency will step in to sell of the bank.

Now if your theory of "reserves are provided as needed," then why was the BoC able to consistently hit it's monetary target over multiple decades. Do you think it was a coincidence?

TC--special request: can you opine on the consequences of default rather than inflation as a way of getting rid of US debt? Posting from Greece, this is a popular option for GR historically, and we know about Argentina and Russia. Since the US dollar is the de facto 'reserve currency of the world', why not default rather than hyperinflate? I honestly don't see any downside (again, Argentina, who 'got away with it'). Further, politically, the US could appease middle-class American voters by decreeing, upon default, that all US citizens holding debt would get 100% of debt instrument face value below a certain limit, say USD $200k per citizen. This would of course be a bit of a windfall for all Americans, as there would be obvious arbitrage opportunities for acquiring debt from foreigners up to this limit (at some small discount). Finally, I am reading the excellent book by John Gordon Steele "Hamilton's Blessing" and Alexander Hamilton himself stymied such a proposal for Continental US debt after the Revolutionary War, so there is a precedent for this default proposal. Any other MR readers that default is feasible as an option?

"I honestly don’t see any downside"


Yeah that's a howler. Ray, care to elaborate?

Argentina, if anybody reads this in the future. Default is no big deal. BTW my $200k per citizen suggestion I think is too high--it should be closer to $50k I think.

I am really surprised that somebody who has been to Asia as regularly as you did not mention the most likely way the whole low-skilled saga will resolve. All you have to do is look at Korea or China.

...and the answer is?

I guess either rice paddies or line shifts at Foxconn. Not sure which is better/worse.

"what kinds of jobs will be available for low-skilled Americans in the decades to come"

Selling each other subprime mortgages didn't work out so hot last time, but I'm sure Beltway Brain Trust will come up with something better Real Soon Now. After all, Matthew Yglesias wants to import 165,000,000 more immigrants and Bryan Caplan thinks we should have Open Borders.

What could possibly go wrong?

165 million people added is the same proportion added from 1940 through 1970, a period when taxes rates were very high, taxes were hiked frequently, unions were powerful and made sure workers got most of the value of productivity gains, and the middle class American Dream was created. And the government debt burden declined sharply over the latter 25 years from over 100% of GDP to 40%.

Clearly what would go wrong is the high taxes and reduced wealth and income inequality - the conservative's worst nightmare.

That should be every American's worst nightmare. This from "a liberal."

GDP works as a proxy for happiness and progress in poorer societies. I think we all accept that there are declining returns in richer ones, where it tops out in usefulness. Sadly there is no more accepted number than GDP. There is no other consensus metric for a "better" future.

Net sum of the intensity * duration of feeling good - intensity * duration of feeling bad over all brains.

Now all we need is mobile brain scanners.


Do we get to bomb Europe and Japan back to the stone age and do China, India and Brazil volunteer to be economic backwaters again?

Do you have something for those who aren't impressed with facile comparisons?

Listening to a certain segment of American society- assuming- is like listening to someone who has been born into unfathomable wealth and simple can't conceive of what it's like to be anything but extremely wealthy? I hope that works out.

Europe and Japan were not bombed back to the Stone Age, and in any event they rebuilt and recovered from WWII within a few years (with US help of course).

"Assuming closed-borders, what kind of jobs will be available to low-skilled Americans in the decades to come?"

Is the answer a whole lot more rosy?

Maybe not much better jobs, but much better to be the poor guy living in Malibu than the poor guy in Calcutta.

How can "Furthermore a lot of the current debt is quite short-term." be true while politicians tell us that a tiny cut on March 1 will be doomsday?

The term on the debt just means how long the bonds are termed for. If debt is weighted towards being very short term, then cuts are even more important, otherwise real interest rates can rise very quickly on the debt.

Unfortunately the loss of jobs and the inflation in the US are still huge problems. Consumption has dropped and people don’t have money. Actually the Orlando Bisegna Index which measures the intensity of the economic crisis worldwide, ranks US 11th place among the countries hardest hit by the economic crisis at 13.18 points for the month of February. Hopefully things will get better soon!!

Shorter Cowen on inflation- "Government officials are smart enough to not let it happen."

At pains, I will point out that markets are always good predictors of the future until they ain't. What was the bond market and stock market predicting in 2007? In 2000? In 1994? In 1987? In 1980?

"government bonds in 1929 yielded 3.4%."

I would like to see more discussion of the impact of entrenched interests on all aspects of our economy.

I agree with #1 about medical costs being a major driver of government budgets. And here's an article about entrenched interests at work in the medical economy:

Unless someone is willing to tackle the arrangements that allow such gouging nothing is going to get better soon.

'I will point out that markets are always good predictors of the future until they ain’t. What was the bond market and stock market predicting in 2007? In 2000? In 1994? In 1987? In 1980?'

Eyeballing the S&P500 pe chart, I'd say high expected stock returns circa 1980, average to high 1987, average 1994, low 2000, average 2007.

Which is pretty much what happened.

rmark, what "the market is predicting" is in the price. *You* are using the p/e ration to guess when the market price is off (when p/e gets too high).

Roughly 1980 pe = 8, 1987 pe = 12, 1994 pe = 18, 2000 pe = 25, 2007 pe = 18. The long term S&P500 average pe has been around 15, so the market was valued for a future return higher or lower than on average. I'm not guessing if the market price was off, I'm only noting it was above or below average in the years mentioned above and that returns in the years following were pretty much as the pe suggested.

"During 1928, the price-earnings ratio for 45 industrial stocks increased from approximately 12 to approximately 14. "


#4. If the fed is unloading huge quantities of low interest instruments into an inflationary environment, it may selling at, say, 30% off par. That loss would crush you or I, but does it not matter to the Fed?

Also, when treasury pays on bonds held by the Fed, the Fed sends the money back to Treasury. What happens when the Fed unloads trillions to the public? How will Treasury make up the loss of revenue/absence of obligation?

Last, if the Fed is shedding assets it will also stop buying US Treasuries. what interest rates will these real auctions produce?

I really don't know, I haven't been initiated into the Macro Mysteries. Do you wear fez's?

The Fed's Massive Accumulation of Treasuries Won't be Unwound Instead the Fed will raise IOR to set interest rates and allow its balance sheet to decline over time as bonds mature. Paying IOR above the yield on current assets is the more likely outcome whereby the Fed takes considerable losses since it doesn't use mark-to-market accounting. You are correct though that the Fed can operate indefinitely with negative capital merely deferring payments to the Treasury until some future date.

If the Fed were to unload its bonds to the public, Congress would have to decide whether to allow larger fiscal deficits or how to adjust policy (tax hikes vs spending cuts) to maintain the desired budget level. Even if the Fed sells, interest rates on Treasuries will still be tied to the expected short-term (Fed Funds) rate plus a risk premium. I suspect the risk premium aspect would rise, but the Fed could likely manage expectations of rate changes to a degree that doesn't result in significantly higher yields. It's important o remember that before the crisis the yield curve had actually inverted as world demand for long-dated AAA ("safe") assets was relatively high.

Thank you for your reply.

Is there a political split on who trusts self-reported happiness? And why? Is it as simple as confirmation bias?

(I think the question underlies both jobs and inflation fears, those being about current and future unhappiness.)

I see a proliferation of and return to artisanal crafts but this time on the high-end. Examples could include coffee, chocolate, and other foods. There is also a trend towards more beauty in everyday production - the aesthetic revolution you can call it. Basically, I see a return towards making delicious and beautiful things by hand and this should soak up a lot of manual labor.

Is there really a shortage of low skill work? It's been my understanding from things I've read that we are still adding low skill (and low wage) jobs. Where the problem is, is with semi-skilled middle-tier jobs.

"Everything we were taught about the monetary base is wrong in a world with interest on reserves (IOR). A large base can sit there forever. The price level is not proportional to the base, changes in the base, etc. It just isn’t. The broader aggregates, such as M2, haven’t grown so rapidly."

What happens when interest rates rise? If the market begins demanding more interest, banks will not let excess reserves sit at the Fed for a measly 0.25% return.

The answer I get to this question, and the one that the author here is implying, is that the Fed will simply raise the IOER to a level sufficient to deter new lending. Ok, I get that. But how long will the public tolerate a policy that just hands the banks hundreds of billions of dollars in order to prevent them from lending? Dancing With The Stars and The Real Housewives and professional sports can't keep Americans distracted forever, right?

Second, what happens when the Federal Reserve starts losing money every year to finance these enormous IOER payouts to the banks? Is that not an unconstitutional appropriation of money?

Comments for this post are closed