"Involuntary" unemployment as worsening trade-off

This is what I think an increase in "involuntary" unemployment looks like:

The vertical axis represents the quality of a job, in the eyes of someone searching (or simply waiting) for a job. "Quality" will presumably include wage, but many other characteristics as well.

The horizontal axis represents the expected duration of searching (or waiting) to get a job of that quality.

The blue curve represents a trade-off. The better the job you want to get, the longer you expect to search (or wait) before you get it. (But the trade-off probably asymptotes to some maximum, or maybe even declines if you wait too long and your skill human capital depreciates.) You pick the tangency point where your indifference curve kisses the trade-off. (Rather, you aim to pick that point, because you don't know what the trade-off will really be, and you might get lucky or unlucky.)

Now suppose the economy enters a recession. The trade-off worsens to something like the red curve. You choose a new point. You will probably choose a point as drawn, expecting to search or wait longer and get a worse quality job.

We could argue about whether the unemployment in the first equilibrium is "voluntary" or "involuntary", but there doesn't seem to be much point. We could even argue about whether the increased duration of unemployment when we move from the first to the second equilibrium is "voluntary" or "involuntary", but there doesn't seem to be much point. In both cases, a person with different preferences would have chosen differently, so you could say it's "voluntary". But the worsening of the trade-off was definitely involuntary. The unemployed person has no choice about that worsening trade-off; only whether to respond to it by accepting a worse job, or being unemployed for longer, or a bit of both.

And whether that worsening trade-off was the result of an efficient response of the economy to some real shock, or the result of an inefficient response to some daft monetary policy shock, is a quite separate question.

Update: And how that trade-off facing an individual worker would shift if all workers changed their preferences at the same time is also a quite separate question. It might shift in or it might shift out. Beware the fallacy of composition. That's why God invented macroeconomists.

In very simple supply and demand models of the labour market, the trade-off looks very different. All jobs are the same with one wage. The trade-off is L-shaped (if you twist the L clockwise through 90 degrees). If the wage is at the market-clearing level, the trade-off is vertical at 0, then goes horizontal at W. If the wage is above the market-clearing level, the L shifts right, so you have to wait for U periods, then get a job at W. If you would refuse a job at W, your unemployment is voluntary; if you would accept a job at W, your unemployment is involuntary.

But the labour market is more like the marriage market. Heterogeneity of jobs and workers matters. If people didn't care who they married, we could divide people into voluntarily single and involuntarily single. But they do care, so we can't. But we can talk about worsening trade-offs.

Just adding my twopenceworth to the debate between Roger Farmer and David Andolfatto (and Roger again). Three-way Western wars.

Update: we could draw a similar trade off for employers with vacancies, with quality of the worker on the vertical axis. But in a recession the trade-off for the employer improves. It's easier to find a better worker quicker for lower wages. Because recessions are a monetary phenomenon, where those who want to sell illiquid goods and buy money find it harder to do so, and those who want to buy illiquid goods and sell money find it easier to do so.

Comments

[Jerry Brown]

If there are 95 job openings and 100 people looking for work, there will be 5 unemployed people if I did the math right.

[Andrew Burton]

But simple models don't describe real situations. Here is a real situation.

I work in the oil & gas industry. As you know, the industry has effectively entered a recession: many people who (a) are very highly skilled (b) very specialized (c) are over 50 years old have been involuntarily terminated by the companies they worked for, in some cases for decades.

All jobs aren't the same with one wage W. An industry which could support X workers with very defined skills and experience can now support x workers, where x << X. These workers are prepared to apply the skills they have in other industries, but locating positions which overlap with their skill set takes lots of time. And potential employers are unlikely to offer entry level or early career positions to people who will likely wish to leave the workforce in less than a decade.

I don't see anything remotely voluntary about the situation these workers find themselves in.

[Ralph Musgrave]

The whole idea that the longer people job search, the better the job they find contains a big weakness, namely that if someone chooses to get a poor job quickly rather than wait for a better one, that doesn’t stop them getting a better job when it appears. I.e. job searching while working is perfectly feasible.

The only evidence on job searching while working that I know of is possibly now very dated. It was a study done by J.P.Mattila which he set out in a paper in the American Economic Review in 1974 entitled “Job Quitting and Frictional Unemployment”. He found that just over a half of those changing jobs found their new job BEFORE leaving their old job.

[David Andolfatto]

"We could argue about whether the unemployment in the first equilibrium is "voluntary" or "involuntary", but there doesn't seem to be much point. We could even argue about whether the increased duration of unemployment when we move from the first to the second equilibrium is "voluntary" or "involuntary", but there doesn't seem to be much point. In both cases, a person with different preferences would have chosen differently, so you could say it's "voluntary". But the worsening of the trade-off was definitely involuntary. The unemployed person has no choice about that worsening trade-off; only whether to respond to it by accepting a worse job, or being unemployed for longer, or a bit of both."

Nick, this is exactly what I was trying to convey in my post. We could argue over such labels, but what's the point? The point is that economic circumstances sometimes change for the worse -- people certainly do not "voluntarily" wish such things on themselves -- but do we really need to point out such an obvious fact? In any case, it is reasonable to suppose that people choose to cope with their changing circumstances the best they can. Sometimes unemployment is the best option out of a set of not-so-good options.

It is interesting to me to see how riled up people get at this interpretation of action. Oh well.

[rsj]

Nick, this is really nice!

But I think something important is missing here, i.e. when a firm faces a reduction in sales, it chooses to lay off employees rather than lower wages. There is clear data that firms act like price takes for their labor costs, and don't give existing workers an option to continue working at a lower wage. So the number of people searching goes up in a downturn. There was a book -- "Why don't wages fall in a recession", in which many employers were interviewed.

Neither do firms lower their asking price -- it is as if they are afraid that by paying less they will get worse quality employees, even though the job market has changed.

So if in one week firms fire 100,000 workers and post vacancies for 10,000, then there will be at least 90,000 unemployed, and the matching function will determine whether that is 90,000 or 95,000. I.e. you are not really capturing what people think of when they say involuntary unemployement. Said this way, matching doesn't seem to be very important as an explanation of involuntary unemployment, at least if you define involuntary employment as the difference between the number of involuntary separations and the number of job postings. That lower bound will only be increased by adding search costs.

[Noah Smith]

I like this. Nick, I declare you the winner of the debate, which is especially cool because you weren't even in it!

[Jerry Brown]

David Andolfatto- "Sometimes unemployment is the best option out of a set of not-so-good options. It is interesting to me to see how riled up people get at this interpretation of action. Oh well." You must be assuming that employment is always an option here. Why?
Are you assuming that the new unemployed person could always find employment by just charging somewhat less than any one of the currently employed people? Even if they could get a job that way, it would just leave us with a different unemployed person.
And as an economist, are you really surprised that people get riled up about unemployment issues? You do realize as a social scientist of sorts, that unemployment can have very large negative impacts on an individual. And their family. And on society in general.

[Nick Rowe]

Jerry: "If there are 95 job openings and 100 people looking for work, there will be 5 unemployed people if I did the math right."

You did the math wrong. "Job openings" is not a number, it's a function, that depends on quality of worker, wage you would have to pay, how hard you think it is to find them, etc. And "people looking for work" is not a number either, it's a function, that depends on the quality of jobs, wage they would pay, how hard you think it is to find them, etc.

Ask a single man if he is involuntarily single. Or a single woman if she is involuntarily single. It depends. Nearly all would say "yes" if they had the right offer, or "no" if the had the wrong offer. They are not just two sets of binary numbers. But you can still talk about whether the trade-offs and opportunities are getting better or worse (though even that might depend, if it gets easier to find a bad job but harder to find a good job, or vice versa).

You think you are so clever, with your lefty snark. But you are just dragging your knuckles.

[Nick Rowe]

Jerry again: "You do realize as a social scientist of sorts, that unemployment can have very large negative impacts on an individual. And their family. And on society in general."

There's that condescending snark again: "You do realize..."

No shit Sherlock. You do realise, don't you, that David is not a hermit living in a cave somewhere who never speaks to anyone and never reads anything but math books?

By asking that non-question, all you are really doing is trying to force a performance out of David, where he genuflects at the altar of full-employment and goes through the ritual denunciations of the devil of unemployment. Which makes you feel good, because God and the Devil have been worshiped and denounced, and heresy denied, but gets us precisely nowhere in understanding what it is, what causes it, and what policies might or might not help.

[rsj]

You did the math wrong. "Job openings" is not a number, it's a function, that depends on quality of worker, wage you would have to pay, how hard you think it is to find them, etc.

No, that is a match! Firms post job openings. Workers do not show up, unasked for, looking for work. Neither is there an auction for workers. Firms know how many people they want to hire in advance, and they decide this based on their own revenues and the long term wage for the types of workers they want to hire. There is no coordinating mechanism that would cause firms to post more openings when unemployment goes up.

That's a decision made by the firm independently of the number of people looking for work. This is because each firm is a small part of the economy.

If all firms post a total of 5 openings, but there are 10 people looking, then *at least* 5 will be unemployed. Whether it is only 5, and how long it takes for the 5 to find work, is a question of search. The involuntary unemployment is not due to search but due to a coordination failure which can be thought of as due to wage rigidity or other potential rigidities.

Now what you are assuming is that there is some background coordinating process in which firms know that people will take lower wages, fire their existing labor force and/or renegotiate their wages down, thereby funding a net new 10 job openings for the 10 workers, which are then matched in a search. That would be combining the micro aspect of search with the Walrassian auctioneer price solving for the right number of job posts. It doesn't help the economy coordinate without an auctioneer, in which each firm acts as a price taker for the wage and determines total job posts with coordination with other firms.

But that problem is no different than the coordination problem without search.

You must know that the search models don't generate anywhere close to the levels of unemployment variation that we see during recessions, and that search is not considered to be the reason for the increase in employment in a downturn. Search is the reason for a permanent secular non-zero level of employment across the cycle. You know this, and yet mislead people, and then insult Jerry when he calls you on it.

[Nick Rowe]

And again: "Are you assuming that the new unemployed person could always find employment by just charging somewhat less than any one of the currently employed people? Even if they could get a job that way, it would just leave us with a different unemployed person."

Again, you are confusing a number with a schedule. You are the one who is making the assumption, and what's worse is you don't even realise you are doing it. Your (implicit) assumption is that the labour demand curve is perfectly inelastic with respect to everything. In some models it is, and in other models it isn't. And even though I can imagine a world where it is, most of the time, in the real world, it very probably isn't. And whether it is or isn't will depend on many things, in particular it will depend on monetary policy. And unless you can get your knuckles off the ground you won't even be able to think about what policies might or might not work.

[Nick Rowe]

Andrew: "All jobs aren't the same with one wage W."

Nick, in the post: "In very simple supply and demand models of the labour market, .... All jobs are the same with one wage. ....But the labour market is more like the marriage market. Heterogeneity of jobs and workers matters."

Maybe I needed to write it more clearly. Yes I know that jobs aren't all the same with one wage. That's what I am saying in this post. See the vertical axis on my diagram?

[Nick Rowe]

Thanks David, Noah, and rsj. Sometimes words don't work, and you need a picture.

rsj: personally, I think that search/matching models help us understand the POV from the individual worker and employer, and help us understand part of the determinants of unemployment in normal times, but don't help us much with explaining fluctuations in unemployment over the business cycle, which I think is a monetary phenomenon, and that sticky nominal prices and wages are part of that explanation.

[rsj]

Maybe the following example will help with the distinction:

The economy is just one firm. Workers decide to hold more money so they spend less. The firm fires some of the workers because output decreases. The problem is not the X% of workers that was fired. The firm would need re-negotiate the wages of all its employed workers downward, in order to hire back the X%.

The matching of the X% is a search problem. The overall lower wage is not search, but a coordination problem. You still need an auctioneer to solve the bigger problem. Search costs help explain why there is a persistent level of unemployment at all times. The lack of an ability to renegotiate wages for the employed workers is a we-don't-have-an-auctioneer problem. Without an auctioneer, the general schedules lie in the attic, and each firm acts as a price taker, choosing only quantities. There is no one to coordinate global market clearing prices, and no guarantee that there are enough job postings for job workers.

[Nick Rowe]

rsj: I fully agree that fluctuations in unemployment are (mostly) a coordination problem, at the macro level. And that even though a change in an individual worker's preferences may get that worker a job, it does not necessarily follow that a change in all workers' preferences will get all workers jobs. It's a fallacy of composition to say it necessarily would. It depends on the model whether it would or wouldn't.

We need a macro model to explain what determines the position of the trade-off facing an individual worker. Arguing whether we call it "voluntary" or "involuntary" gives us no help at all in building that macro model.

Barro and Grossman (1971) have a very simple model where a fall in M/P causes employment to fall. A fall in W/P, given M/P, converts the "involuntary" unemployment into "voluntary" unemployment (the participation rate falls to eliminate the excess supply of labour) but has no effect on the level of employment. Which is why the distinction is not helpful. But it is helpful to say that the fall in M/P caused the trade-off facing individual workers to worsen.

[Nick Rowe]

Ralph: you are right that my model ignores a third option, to accept a current offer while waiting for a better one. But we could say that all jobs are like that, because I can't think of anyone who might not quit if a really good offer came along. The question is why the unemployed don't all take really crap jobs temporarily at the other side of the country. Some do. Some would rather be unemployed (or fully employed searching). Plus there are fixed costs for both worker and employer, of a new hire, so you won't do it if it's only going to be very short.

[K]

Distinguishing between situations in which the unemployed could collectively lower their reserve wage and become employed, and situations in which they cannot strikes me as perfectly "useful". And calling those situations voluntary and involuntary unemployment strikes me as a good idea since that's what we've been calling them for 80 years now. Is there a good reason why we suddenly do not want words to be able to distinguish between the two?

[Nick Rowe]

K: Yes it is useful to discuss the wage-elasticity of the aggregate labour demand curve. But talking about whether unemployment is voluntary or involuntary from the perspective of the individual worker tells us nothing about that elasticity. It is easy to build a model where all unemployment is purely "voluntary" at the individual level, but the labour demand curve is perfectly inelastic. Or vice versa.

[Kevin Donoghue]

I can't see why people should object to "involuntary unemployment" as a concept. Keynes made his meaning quite clear in that instance. But if the word causes discomfort to Lucasians there's no reason why we can't follow Malinvaud's example and instead speak of "Keynesian" as opposed to "Classical" unemployment. It comes to much the same thing.

[Majromax]

@Nick Rowe:

> I think that search/matching models help us understand the POV from the individual worker and employer, and help us understand part of the determinants of unemployment in normal times, but don't help us much with explaining fluctuations in unemployment over the business cycle, which I think is a monetary phenomenon, and that sticky nominal prices and wages are part of that explanation.

I think a fuller model of unemployment that begins at the individual worker/firm level would need:

*) Searching, as you describe in this post
*) Fixed costs of hires, for both the employee (relocation) and employer (job advertisement, training)
*) Human capital, including its decay over unemployment
*) Business capital, including production functions that permit idle capacity (and consequently investment)

To capture true demand-side/monetary recessions like the 2009 recession, I think we would also need to include the possibility of firm-level default.

To capture income equality, especially to model "1% effects," we would need to introduce wealth-heterogeneous agents that receive proportional shares of capital income.

Some of the fixed cost may be replaceable with a more nuanced model, where workers have job-specific human capital (experience) and job-agnostic human capital. The former is lost upon a job switch, voluntary or otherwise. Models which permit more job-specific human capital would intuitively seem to prefer steadier employment.

I doubt this model would be fully analytically tractable, but on the other hand it also sounds simple enough to simulate with a few thousand or ten-thousand agents.

[rsj]

Nick,

Yes, BG is a good point. So let me see if I understand your thinking: involuntary unemployment (in the sense of Barro-Grossman -- hey they used the word and didn't melt) is a specific case of the excess demand for commodities when the real wage is fixed. If the real wage were allowed to fall, involuntary employment would be transformed into voluntary employment. Therefore there should be no public discussion of "involuntary unemployment", all the discussion should be of an excess demand for commodities or an excess supply of labor.

This, to me, sounds absolutely crazy. It's certainly pinning a lot on the accuracy of BG, and given that we live in a world in which the real wage is pretty rigid, it seems perfectly reasonable to discuss involuntary unemployment, regardless of what might happen in alternate universes.

Moreover, who the hell made economists the language police? Are we going to go on a rage against the use of "profits" next?

Involuntary unemployment is an accurate, meaningful description of what happens in a recession. A shift of the matching function is not. The matching function is a theoretical invention that is an artifact of a parsimonious model that doesn't describe employment fluctuations very well to begin with. The matching function is not real. Count the number of job postings and subtract the number of people looking for work. If that number is negative, then we have a real case of involuntary unemployment.

Moreover it the matching function in search is a fairly obscure mathematical construct that suggests to the public that if they lower their wage demands and search a bit longer, then they will get a job. We know that isn't true -- firms will not post more jobs if we increase our time spent searching.

We would be better off reducing our search efforts in recessions, just as we would be better of mutually disarming -- it really is a zero sum game with involuntary unemployment, which is at least one practical reason to keep this term around. In normal times, it is not a zero sum game -- it makes sense to search more. That very real difference in outcomes is what people talk about when they say "There aren't enough jobs".

[rsj]

^^the above should read involuntary unemployment would be turned into voluntary unemployment (sorry)

[Majromax]

@rsj:

> Count the number of job postings and subtract the number of people looking for work. If that number is negative, then we have a real case of involuntary unemployment.

Does this suggest that I could turn all "involuntary unemployment" into "voluntary unemployment", if I were to post a million job postings that pay wages of a penny per day?

[Kevin Donoghue]

rsj: Moreover, who the hell made economists the language police? Are we going to go on a rage against the use of "profits" next?

If we can put "equilibrium" on the bonfire, count me in.

[Kyle Johansen]

At the risk of getting caught in a tree, companies often have a 'if we don't have a job posting that fits with you, but if you think you have skills we're after, please contact us', and that's without thinking of stuff like self-employment.

I've always had a knee jerk dislike of the 'involuntary unemployment', but this blog post really made me think and gets a more nuanced view.

[rsj]

@Majormax,

I think there are many reasons why don't live in that type of world -- Even for the cheapest employee ($8/hr retail worker), it costs about $10K to train them and look for a replacement in normal times. For a nurse, it's about $50K. For a tech worker, it can be several hundred thousands. So firms have to be careful when offering lowball wages to employees because, even if the employee takes them during the recession, they will leave at the first opportunity and might create a negative atmosphere in the workforce while they are there.

This is one reason why firms don't hire over qualified employees.

For example, I was once offered a lowball wage by one firm when I was looking for work, and I knew it was a low offer but I knew the firm was going through some tough times so it wasn't that big of a deal. I was going to accept it, but then had a chance at a better offer (almost 2x). I took the better offer and then told the first firm, and they immediately countered with something even higher, which made me really hate them, because it meant that they could have afforded to pay me more to begin with, but were trying to get away with paying me less. If they would have said, "sorry we can't match that", then I would be totally fine with that firm.

The point being that when you work for a company you have to trust them to take care of you as an employee. There are so many ways that an employer can screw you if they want, and you can't monitor everything. Firms are very sensitive to this, because they too can't monitor everything that workers do, and so they need to be able to trust the worker to take care of them in return. Therefore it's in the interest of both sides to have an equitable pay. Workers will accept less pay if the firm is losing money but they will not accept low pay if the firm is making money.

[Brian]

David: "It is interesting to me to see how riled up people get at this interpretation of action. Oh well."

People get upset because calling increased levels of unemployment in a recession "voluntary" or "involuntary" has political implications. There are politicians making the argument that unemployment is purely voluntary, so doing anything to help the unemployed will just encourage laziness and discourage them from taking a job. You say that "The 'classical' idea that there is little a government can or should do to help society in a deep recession is nonsense." But when you argue against the term "involuntary unemployment", it appears to provide support for Paul Ryan and friends. I doubt that is your intent, but that is the perception.

[Brian]

Nick: "And how that trade-off facing an individual worker would shift if all workers changed their preferences at the same time is also a quite separate question. It might shift in or it might shift out. Beware the fallacy of composition. That's why God invented macroeconomists."

To me (not an economist), this seems like the question that matters. Isn't it possible to have a situation where if everyone lowers their standards, the quality of jobs available goes down even more? If the job supply curve depends on which indifference curve everyone picks, couldn't you have a situation where there is no overlap, at least for the average person? That sure sounds "involuntary" to me.

[Jason Smith]

Nice post. I really think it not only gets the most logical answer to the question "voluntary" vs "involuntary", but makes the whole argument clearer.

[Too Much Fed]

rsj said: "The economy is just one firm. Workers decide to hold more money so they spend less. The firm fires some of the workers because output decreases."

Can you say workers decide to take more "money" out of circulation from the real goods/services economy by deciding to hold more "money" or by directing more "money" into financial assets?

Can you describe that as in increase in saving?

[Nick Rowe]

Kevin: "I can't see why people should object to "involuntary unemployment" as a concept. Keynes made his meaning quite clear in that instance."

Keynes defined what *he* meant by "involuntary unemployment", but it is a very peculiar definition, and has no resemblance to anything anyone else might mean by it. Take an efficiency wage model of unemployment, for example. That looks like involuntary unemployment, because there is nothing an individual unemployed worker can do to get a job. But (absent nominal rigidities or money illusion) it is purely "voluntary" unemployment by Keynes' definition, because an increase in AD and P will not cause employment to rise.

Keynes should have called it "unemployment that can be cured by increasing AD".

[The Arthurian]

Nick: "That looks like involuntary unemployment, because there is nothing an individual unemployed worker can do to get a job. But (absent nominal rigidities or money illusion) it is purely "voluntary" unemployment by Keynes' definition..."

So you are saying that Keynes' definition of involuntary unemployment was a conservative definition, understating rather than overstating his case. And there is probably more involuntary unemployment than he might have allowed.

As I recall, Keynes was arguing against the notion that there is no such thing as involuntary unemployment. He was right, therefore, to throw out questionable cases and include only cases where the unemployment is most assuredly involuntary.

"So long as the classical postulates hold good, unemployment, which is in the above sense involuntary, cannot occur."
https://www.marxists.org/reference/subject/economics/keynes/general-theory/ch02.htm

Conversely, if anyone is involuntarily unemployed, the classical postulates do not hold good.

I really liked your post, by the way.

[Patrick]

FWIW, it occurs to me that for whatever jobs are available in a recession, the employer's search problem gets harder/more costly during a recession - they tend to get pickier and they have to sort through more potential candidates. Bit of nasty negative feedback.

[Avon Barksdale]

Nick, I love your post!

As far as people complaining that, because unemployment is voluntary, this observation somehow supports a hard right wing agenda to eliminate unemployment insurance, they need to calm down a bit.

Unemployment insurance does increase unemployment – yes, sorry, it's true, you do get more of what you pay for. So, labour markets do have a different equilibrium with unemployment insurance than without. But that doesn't mean that unemployment insurance is bad even if unemployment itself is voluntary. Taking a terrible job during a recession (the involuntary economic shock) might not be optimal either. Perhaps we should view unemployment insurance as Pareto optimal risk sharing with moral hazard. The trick is to keep the moral hazard down and find out just how much unemployment responds to insurance – that would be a solid scientific approach to the trade-off embedded in the problem, which might keep the emotions down a bit.

[Nick Rowe]

rsj: in the Barro-Grossman model, if we put W/P in the vertical axis in my diagram, the trade-off at full equilibrium is initially L-shaped, (rotated 180) with the vertical bit on the vertical axis. If M/P drops and W/P stays the same, the trade-off shifts horizontally right, which is a worsening. If W then drops (involuntary unemployment gets converted to voluntary unemployment in B&G's terminology) the trade-off shifts left (good) but also shifts down (bad).

Comparing the 3 positions of the trade off, we can say that both 2 and 3 are worse than 1, but cannot say which of 2 or 3 is worse, because the two trade-off cross. Which makes sense to me. In either 2 or 3 (whether W is sticky or flexible) we get a worsening trade-off when M/P falls. It's "involuntary" by my definition, in both 2 and 3.

[Nick Rowe]

Kyle: I tend to agree. "Vacancies" is a very fuzzy number. Whether you look for a worker depends on what you think you might find. "Unemployment" is also a fuzzy number. Whether you look for a job depends on what you think you might find.

Brian: "Isn't it possible to have a situation where if everyone lowers their standards, the quality of jobs available goes down even more? If the job supply curve depends on which indifference curve everyone picks,..."

We call it "labour demand curve", not "job supply curve". But yes, it is possible. In some models it will be true. It depends on monetary policy in particular. But with a sensible monetary policy it won't happen. Economies with higher populations tend to be able to create labour demand in proportion, so they don't have higher unemployment rates or worse wages (except in Malthusian economies where land is fixed so the lower land/labour ratio reduces labour productivity and real wages).

[Nick Rowe]

Thanks Jason. Again, pictures beat words.

Arthurian: Keynes' definition of "involuntary unemployment" was both too conservative and too liberal. He should have called it "deficient AD unemployment". But in Keynes defence, he was trying to create the concept of AD, so we can look back on his definition as an awkward stepping stone that can be discarded after we have used it to cross the river.

Patrick: Yep, if you are trying to pick the best candidate, it gets a lot harder when unemployment is high. I am right now on a hiring committee for 2 positions in Global and International Studies. We have hundreds of applications, many of them from stellar candidates. Reading them all, and picking the best, is a very hard job. There are 6 of us on the committee. The fixed costs of hiring someone are large. But then, if an employer wanted, he doesn't have to pick the best candidate; just look through the applications till he finds one that's good enough.

Avon: thanks! Yes, there's always a spike in the unemployment duration data at exactly the point that EI runs out, which strongly suggests that people respond to incentives. If they can't find a good job by the time it runs out, they take a crappy one. But this does not mean that EI is necessarily a bad thing. Matching the right worker with the right job is a productive investment (comparative advantage and all that) and we won't get the right amount of investment if the unemployed are borrowing-constrained, which they probably are. (Though one suspects that in some cases they know there aren't any better jobs, and EI is just holiday pay, and the whole EI system is just treated like an export industry, that workers, firms, and provincial governments successfully work together to manage. That's the moral hazard angle.)

[Nick Rowe]

One thing I should have said in my post: saying that individual unemployed workers face a choice, and their choice depends on their preferences, does NOT mean that increases in the unemployment rate during recessions are caused by mass outbreaks of contagious increased pickiness. It is much more likely that they faced a worsening trade-off.

[Peter T]

About 10% of the C18/early C19 populations of England and France were malnourished to the point that they literally could not work - a little light begging was about the most that their bodies could support. This is still the case for a similar proportion of the population in India or similar places. Are they "involuntarily unemployed"? Or maybe just facing a worse trade-off? There is a floor amount of nutrition, clothing, housing etc that renders anyone suitable for employment. This floor is particular to times and places (as Adam Smith observed, lack of shoes is not the issue in Scotland that it was in England). In the US or Canada it is quite high. Also, almost all employment takes place within organisations structured by location, capital, expertise needed and so on (and often also by class, accent, race, appearance, gender and other unwritten requirements). So it makes no sense to "model" unemployment as an intersection of two abstractions, as this explains nothing at all. It is simply hand-waving when it's not feeding some agenda. When we say "there are only 100 jobs and there are 120 people" we take as read the background that there is, in the current organisation of production, places for only 100. The background structure is the explanation. Why ignore it?

[Nick Rowe]

Peter: "When we say "there are only 100 jobs and there are 120 people" we take as read the background that there is, in the current organisation of production, places for only 100. The background structure is the explanation. Why ignore it?"

No. When we say "there are only 100 jobs and there are 120 people" we are explaining nothing. We are saying nothing about what determines the number of jobs and the number of people who want jobs, and whether or not one or the other would change if the other changed, and what might cause them to change. If instead you tell me "there are 100 shovels and 120 people who want work, and the technology is one shovel one worker and a worker without a shovel is useless, and the stock of shovels is fixed", then you are actually telling me something, and you actually have an explanation for the level of employment. (But you still have no explanation of unemployment, because that will depend on the elasticity of the labour supply curve, and whether wages adjust, and the method by which those 100 jobs get allocated to workers.)

"About 10% of the C18/early C19 populations of England and France were malnourished to the point that they literally could not work - a little light begging was about the most that their bodies could support. This is still the case for a similar proportion of the population in India or similar places. Are they "involuntarily unemployed"? Or maybe just facing a worse trade-off?"

What you are describing there sounds more like a Malthusian equilibrium. And it's a very nasty place, and most of human history (and animal history) is like that. If a deer can't produce enough food to feed itself, it starves. And there's a nasty positive feedback loop, because the weaker it gets the less food it can find and the weaker it gets. It did not starve because it was "involuntarily unemployed". We could have 120 jobs for every 100 workers, but if real wages are too low, especially for the sick and disabled, some people will starve. You need welfare (or poor houses) to fix that problem, not "full employment" monetary policy.

[Nick Rowe]

And even in the 100 shovels case, that would not explain why employment falls and unemployment rises in recessions, unless recessions are caused by loss of shovels. Bad monetary policy will cause a drop in employment of shovels as well as a drop in employment of workers. And we do in fact observe more idle machines as well as more idle workers in a recession.

[Min]

Ralph Musgrave: "The whole idea that the longer people job search, the better the job they find contains a big weakness, namely that if someone chooses to get a poor job quickly rather than wait for a better one, that doesn’t stop them getting a better job when it appears."

Also, longer waits may result in ever poorer jobs and job prospects down the line.

[Peter T]

"But you still have no explanation of unemployment..."

If I observe that 1. the fish stocks of Newfoundland will no longer support a large fishing industry and associated employment; 2. that the political economy of Canada will not transfer money or other resources to Newfoundland in sufficient quantity to employ the present population; 3. that the resources of Newfoundland, relative to other parts of the global system, are inadequate to support full employment among the present population; 4. that the present population of Newfoundland is, for various reasons, slow to emigrate; and 5. that the government of Canada is, for various reasons, willing to provide a basic subsistence to unemployed Newfoundlanders then

I have explained unemployment in Newfoundland. I could do the same analysis (with different factors operating, of course) for Ireland, Latvia, Spain or any other place you care to name.

The 10% I mentioned is not a "malthusian equilibrium". It's a given political economy operating as it did. The food was there, and plenty of 'unemployed" persons accessed it (eg all those on the pension rolls of the court or aristocracy). When for political reasons, it was no longer acceptable to keep so large a fraction of the population below subsistence, the resources were made available.

Looking at it as an individual trade-off obscures the factors actually operating.

[Nick Rowe]

Peter: "Looking at it as an individual trade-off obscures the factors actually operating."

It does not obscure them. It is silent on the question. It says nothing about what determines the position of the trade-off facing the individual, and why it shifts. You actually need a macro theory for that. My diagram is not a theory of unemployment, and does not pretend to be.

Reading between the lines of what you say about Newfoundland, I see a labour supply and demand curve where the market-clearing real wage is below EI benefits, so in equilibrium wages equal EI benefits and some are unemployed. If EI benefits were cut, wages would fall, we would see a movement along the labour demand and supply curves, and unemployment would fall. A theory of unemployment needs to do more than simply restate the fact that there is unemployment; it needs to tell you what things would cause the unemployment rate to change, what other things would change too, and why.

[Nick Rowe]

I see the latest e-axes missive has filed the 3 posts by Roger, David, and me under "semantics". Which I think is correct. We are arguing about the meaning of "involuntary" unemployment (or what would be a useful meaning of the word, if any). I think we should drop the word, and instead talk about worsening trade-offs. Then we can argue about what causes that trade-off to worsen at some times (and improve at other times), which is a substantive question. I think it is caused (for business cycles) by an excess demand for the medium of exchange. And the policy response should be to change monetary policy.

[Peter T]

"If EI benefits were cut, wages would fall, we would see a movement along the labour demand and supply curves, and unemployment would fall."

Worked so well for Greece, Latvia, Ireland...I understand that in the Central Sahara, where there are no unemployment benefits at all, idleness is unknown...

Snark aside, an analysis of the particular situation in Newfoundland or anywhere else along the lines above directs policy attention to numerous areas of consideration (industrial policy, mobility of the population, external transfers and so on). Cutting EI benefits is simply a stock response generated from simple assumptions, most of them empirically weak (and one that has led to great suffering and sometimes, in the past, to actual famine). I would prefer a government that pays attention to the mechanics of actual situations myself, but YMMV.

[Nick Rowe]

Peter : that wasn't *my* theory. That was my interpretation of *your* theory of unemployment in *Newfoundland*.

My theory of recent unemployment in Greece Latvia and Ireland is: bad monetary policy causing a fall in AD.

[Kevin Donoghue]

Nick: "My theory of recent unemployment in Greece Latvia and Ireland is: bad monetary policy causing a fall in AD."

Control of Irish monetary policy was handed over to Frankfurt when Ireland committed to joining the Eurozone. It doesn't make sense to say monetary policy was bad, since it could only be good if, by some fluke, the Irish economy was in synch with the Eurozone core. Like it or not, if you want to discuss Irish policy you've got to look at fiscal policy.

[Nick Rowe]

Kevin: OK, but joining the Euro was bad monetary policy. And even given that mistake, the ECB has followed bad monetary policy. It could not have been good for Ireland, but it could have been not as bad.

[Oliver]

Late to the show here, but this seems very wrong.

Involuntary employment is a macro term. It is a macro constraint faced by individuals, just as Jerry Brown said in the first comment. It is involuntary in the sense that not every individual who wants a job can find one. This does not mean there doesn't exist a policy stance that could change that. If that were the case, we wouldn't need economics / economists.

But relegating the responsibility for the macro outcome to the lack of flexibility among employees is the real bogeyman here. This is the point that Keynes rejected, if I'm not mistaken.

Now suppose the economy enters a recession. The trade-off worsens to something like the red curve. You choose a new point. You will probably choose a point as drawn, expecting to search or wait longer and get a worse quality job.

The point of involuntary unemployment is not that trade off for the individual worsens or that individuals may take time to adjust to a new economic reality - a time during which they might be temproarily unvoluntarily unemployed. The point is that, macroeconomically speaking, there is a (growing) number of people who have no trade-off to speak of. And that a: lowering overall wages actually worsens that problem in the short run and b: there is no reason to believe that is will take care of itself, even in the long run.

According to your model, most of the unemployed are likely to be found among people who have a good job before the recession and refuse to accept a worse one during or after it. Reality tells me that most of the unemployed, especially the long-term unemployed, belong to the class of people whose jobs could not actually become any more menial. And / or whose wages could not drop any further without the trade off becoming one of joining the informal / illegal sector or emigrating instead. Neither of which can possibly be a macroeconomic goal.

[Jacques René Giguère]

Oliver: "And / or whose wages could not drop any further without the trade off becoming one of joining the informal / illegal sector or emigrating instead. Neither of which can possibly be a macroeconomic goal."
Isn't that the model Germany has in mind for Greece?

[Oliver]

Jacques
Yes, my point exactly. But then again, according to Wolfgang Schäuble and many others, things were all hunky dory before these Syriza punks showed up and spoiled the show.

[Nick Rowe]

Oliver: if I put a gun to your head, and give you a choice between your money or your life, does that mean you voluntarily give me your money?

If there are 100 people, and I can only mug the 10 who run away from me the slowest, does that make it voluntary?

I would say "no" in both cases. I have worsened the trade-off facing you, both as individuals and collectively.

[Oliver]

I'm glad you feel 'no' is the answer, but tto me the term trade-off implies it is voluntary (trade being a voluntary exchange of things). So there's a bit of a contradiction there.

But more importantly, any model of unemployment should at least mention the gun / the mugger. If not, the only logical conclusion is that it is the person's own fault if she becomes unemployed / is mugged / robbed. And that is exactly how the unemployment problem in southern Europe is being framed at the moment.