The term involuntary unemployment was introduced by Keynes in the General Theory as the name he attached to the phenomenon of high cyclical unemployment during the downward phase of business cycle. He didn’t necessarily restrict the term to unemployment at the trough of the business cycle, because he at least entertained the possibility of underemployment equilibrium, presumably to indicate that involuntary unemployment could be a long-lasting, even permanent, phenomenon, unless countered by deliberate policy measures.
Keynes provided an explicit definition of involuntary unemployment in the General Theory, a definition that is far from straightforward, but boils down to the following: if unemployment would not fall as a result of a cut in nominal wages, but would fall as a result of a cut real wages brought about by an increase in the price level, then there is involuntary unemployment. Thus, Keynes explicitly excluded from his definition of involuntary unemployment, unemployment caused by minimum wages or labor-union monopoly power.
Keynes’s definition has always been controversial, because it implies that wage stickiness or rigidity is not the cause of unemployment. There have been at least two approaches to Keynes’s definition of involuntary that now characterize the views of mainstream macroeconomists to involuntary unemployment.
The first is rationalization. Examples of such rationalization are search and matching theories of unemployment, implicit-contract theories, and efficiency-wage theories. The problem with such rationalizations is that they are rationalizations of why nominal wages are sticky or rigid. But Keynes’s definition of involuntary unemployment was based on the premise that reducing nominal wages does not reduce involuntary unemployment, so the rationalizations of why nominal wages aren’t cut to reduce unemployment seem sort of irrelevant to the concept of involuntary unemployment, or, at least to Keynes’s understanding of the concept.
The second is denial. Perhaps the best example of such denial is provided by Robert Lucas. Here’s his take on involuntary unemployment.
The worker who loses a good job in prosperous times does not volunteer to be in this situation: he has suffered a capital loss. Similarly, the firm which loses an experienced employee in depressed times suffers an undesired capital loss. Nevertheless the unemployed worker at any time can always find some job at once, and a firm can always fill a vacancy instantaneously. That neither typically does so by choice is not difficult to understand given the quality of the jobs and the employees which are easiest to find. Thus there is an involuntary element in all unemployment, in the sense that no one chooses bad luck over good; there is also a voluntary element in all unemployment, in the sense that however miserable one’s current work options, one can always choose to accept them.
R. E. Lucas, Studies in Business-Cycle Theory, p. 242
Because Lucas believes that it is impossible to determine the extent to which any observed unemployment reflects a voluntary choice by the unemployed worker, or is involuntarily imposed on the worker by a social process beyond the worker’s control, he rejects the distinction as artificial and lacking empirical content, the product of Keynes’s overactive imagination. As such, the concept requires no explanation by economists.
Involuntary unemployment is not a fact or a phenomenon which it is the task of theorists to explain. It is, on the contrary, a theoretical construct which Keynes introduced in the hope that it would be helpful in discovering a correct explanation for a genuine phenomenon: large-scale fluctuations in measured, total unemployment. Is it the task of modern theoretical economics to “explain” the theoretical constructs of our predecessors, whether or not they have proved fruitful? I hope not, for a surer route to sterility could scarcely be imagined?
Id., p. 243
Lucas’s point seems to be that the distinction between voluntary and involuntary unemployment is purely semantic and doesn’t correspond to any observable phenomena that are of scientific interest. He may be right, and if he chooses to explain observed fluctuations in unemployment without reference to the distinction between voluntary and involuntary unemployment, he is under no obligation to accommodate the preferences of those economists that believe that involuntary unemployment is a real phenomenon that does require an explanation.
There is a real conflict of paradigms here. Surely Lucas is entitled to reject the Keynesian involuntary unemployment paradigm, and he may be right that trying to explain involuntary unemployment is unlikely to result in a progressive scientific research program. But it is not obvious that he is right.
One might argue that Lucas’s argument against involuntary unemployment resembles the argument of physicalists who deny the reality of mind and of consciousness. According to physicalists, only the brain and brain states exist. The mind and consciousness are just metaphysical concepts lacking any empirical basis. I happen to think that denying the reality of mind and consciousness borders on the absurd, but I am even less of an expert on the mind-body problem than I am on the existence of involuntary unemployment, so I won’t push this particular analogy any further.
Instead, let me try another analogy. Within the legal speed limits, drivers choose different speeds at which they drive while on a turnpike. Does it make sense to distinguish between situations in which they drive less than the speed limit voluntarily and situations in which they drive less than the speed limit involuntarily? Sometimes, there are physical bottlenecks (e.g., lane closures or other obstructions of traffic flows) that prevent cars on the turnpike from going as fast as drivers would have chosen to but for those physical constraints.
Would Lucas deny that the distinction between driving at less than the speed limit voluntarily and driving at less than the speed limit involuntarily is meaningful and empirically relevant?
There are also situations in which drivers involuntarily drive at less than the speed limit, not because of any physical bottleneck on traffic flows, but because of the voluntary choices of some drivers to slow down to rubberneck at something at the side of the turnpike but doesn’t physically obstruct the flow of traffic. Does the interaction between the voluntary choices of different drivers on the turnpike result in some drivers making involuntary choices?
I think the distinction between voluntary and involuntary choices may be relevant and meaningful in this context, but I know almost nothing about traffic-flow theory or queuing theory. I would welcome hearing what readers think about the relevance of the voluntary-involuntary distinction in the context of traffic-flow theory and whether they see any implications for such a distinction in unemployment theory.
Uneasy Money 
The rubbernecking analogy is great, because it shows that something may be involuntary at the individual level but voluntary at the aggregate level.
But perhaps with unemployment it’s the other way around. Like In Barro Grossman 1971, where AD is too low (because M/P is too low, and P is stuck), so the labour demand curve is vertical, so if W drops, some unemployed workers give up, and the “involuntary” unemployment is converted into “voluntary” unemployment.
A variant on Scott’s “Musical Chairs” model: Each day the economy creates a fixed number of chairs (jobs). The first workers in the queue at 9am get a job. So any individual worker can always get a job by getting in line earlier than the other workers (unemployment is “voluntary”). But in aggregate, they can’t all get jobs by all joining the line earlier (unemployment is “involuntary”).
The individual/aggregate thing aside, each individual unemployed worker always faces a trade-off between quality of job vs how long it takes to find one. So it’s always “voluntary” in that sense, though a Hobson’s choice sense if the quality of a job you can get now is bad enough. But in a recession that trade-off shifts, getting worse, which is involuntary (recessions are not a voluntary movement along a fixed trade-off).
LikeLike
“The first workers in the queue at 9am get a job. So any individual worker can always get a job by getting in line earlier than the other workers (unemployment is “voluntary”). But in aggregate, they can’t all get jobs by all joining the line earlier (unemployment is “involuntary”).’
I think this is a silly analogy. In no way is it representative of reality.
LikeLike
David,
The only things I know about traffic-flow theory come from having had a driving licence for over 50 years. If I want to go fast there is always someone in front of me wanting to go slow. If I want to go slow there is always someone behind me wanting to go fast.
The problem (one of many) with Lucas is that he is talking about what a person does once he no longer has a job. What should be focused on is how or why the person who no longer has a job lost the job in the first place.
People are largely unemployed because their employers sack them, generally on mass.
LikeLike
“People are largely unemployed because their employers sack them, generally on mass.”
So the question should be: Why do employers sack their employees, in particular, en mass?
LikeLike
I really liked this post and I will re-read it few times before commenting.
But on some levels, employers cannot hire more people, even if wages go down. If I run a machine shop with five machines, and the machines can only run eight hours day due to maintenance or power-supply issues, then I can only hire five workers.
How many employees can a small retail store hire? Sure, if wages go way down, maybe add an extra body or two, but soon you have people standing around. Maybe the new employees can walk the streets as human billboards. A large shop too reaches some sort of saturation point.
Meanwhile as wages are falling, so is AD. Is hiring wise?
LikeLike
Scott Sumner has a recent post which mentions (mostly in passing) that it’s the stickiness of new hires that matters for wage stickiness. I wonder if something similar applies here: rather than think “(Potential) workers won’t accept the wage that would lead to employment”, think “Employers won’t offer the wage that would lead to employment”. Put another way, employers won’t cut wages to a level where they expect hiring to be profitable. In such a case, the worker is involuntarily unemployed, in the sense that they are not even offered the job. In recessions and their aftermaths, we tend to see the ratio of job seekers to vacancies become high, and in that sense the job seekers are involuntarily unemployed. They can’t force the employer to offer them a job, even at a lower wage.
As to WHY employers might not cut wages, there could be many reasons. Unions might oppose it. Minimum wage laws could forbid it. Unemployment benefits could exceed it. Less obviously, there could be IO reasons involving game theory, efficiency wages, and the like, which I don’t know much about, but which could mean that even in a very free market and with no welfare state, you could still have involuntary unemployment.
PS Although philosophy is my field, I don’t know much about the philosophy of mind. However, the following are three distinct positions:
Eliminative materialism – we should (more or less) eliminate our concepts of consciousness, minds, wills etc. and replace them with concepts from neuroscience, behaviouristic psychology etc. Notable “eliminative materialists” like Paul and Patricia Churchland argue for a much more modest position, which is that such replacements may be rational in the future as neuroscience and psychology improve.
Reductive physicalism – this takes a variety of forms (depending on what “reduction” means) but the simplest version is that the mind is identical with the brain in its living state. Note that this doesn’t mean that we have to stop believing in minds, just as the fact that diamond is carbon in a particular state doesn’t mean that we have to stop believing in diamonds.
Non-reductive physicalism – as I recall, this is currently the single most popular view. Minds are physical, but they can’t be reduced to any “lower level” thing like brains. So you could have Ralph Hawtrey in living flesh, but also robo-Ralph Hawtrey, who is mechanical but nonetheless conscious and has the same type of memories, thoughts, feelings, intentions etc. as fleshy Ralph Hawtrey. Minds are to matter as shapes are to matter: a sphere can be made out of wax, lead, wood etc. In special conditions, you can even have spheres of liquids, gases, or (I think) plasma. What makes it a sphere is not the particular type of matter of which it is composed, but the way that the matter is arranged.
I have probably got at least part of that wrong, but I mainly wanted to clarify that while physicalism is popular among contemporary philosophers, that doesn’t mean that denying the existence of consciousness or minds is popular in philosophy today!
LikeLike
Personal example: it’s not quite the same thing I would like a permanent job in academia in a nice place to live. I have a low cost of living and I would be happy to work for a relatively low wage until later in my career. There are universities that would hire me at a wage that we could both accept. However, those job offers don’t even exist, for a variety of reasons, all of them sufficient, e.g. offering low wage jobs could erode trust between existing academics and their employers. For the higher wage jobs, there are vastly more qualified candidates than jobs.
There is a certain fussiness on my part in this situation, such that it is not truly involuntary unemployment. Yet, if we change a few details, we can imagine a similar thing happening for ALL jobs that a job seeker could do.
LikeLike
It is curiousity that in the present US macroeconomists seem to have embraced sticky wages, but on the upside.
The Beige Books of the Federal Reserve before C19 referred often to “worker shortages.” This inequilibrium persisted for years, according to the Fed, and appears to have returned.
In much of country, I wonder how worker shortages are distinguished from housing shortages. Are fast food joints along the West Coast suffering from a lack of workers…as workers can’t pay the rent and have left?
The housing issue seems like a large structural impediment, and growing worse.
LikeLike
William,
“There are universities that would hire me at a wage that we could both accept. However, those job offers don’t even exist…”
I would say this has essentially nothing to do with involuntary unemployment. You are talking about microeconomic matters, not macroeconomic matters.
LikeLike
Henry Rech,
I don’t accept a sharp distinction there. People don’t demarcate their choices into microeconomic and macroeconomic decisions.
LikeLike
William,
“People don’t demarcate their choices into microeconomic and macroeconomic decisions.”
People mayn ‘t but economists do – people just make choices – I agree..
If you are choosing between two styles of shoes, you are making a decision with microeconomic ramifications.
If you decide to buy a pair of shoes because your income has increased and if there has been a general increase in incomes, then that decision has macroeconomic ramifications.
LikeLike
Do people voluntarily rubberneck, or are there laws on every road in every state that demand you “slow down” or “change lanes”. A fun way to lose friends is to argue whether or not you are obligated to do both and how much to slow down. Do I rubberneck, yes, the law basically demands we do as we in a very uncoordinated fashion attempt to abide by it.
LikeLike
Nick, Thanks for your generous reply. The way I think about it is that in a depression a worker who offers to work for less to avoid being laid off will not save his job, because his acceptance of wage cut will not cause the employer to increase output because the effect on the employer’s costs is too small to make an output increase profitable for the employer. It’s just like one driver trying to drive faster when all the other cars on the road are slowing down.
I like Scott’s analogy, but workers being laid off are already in line, so there’s nothing they can do get to the head of the line; which may be Henry’s criticism, but I can’t tell.
Henry, Workers are laid off usually because the optimal level of output for the firm under existing demand conditions, is a lot less than the output the firm had been producing. A worker offering to work for less can’t save his job by doing so. So he doesn’t. If you’re in a traffic jam, pressing down hard on the gas pedal won’t get you to your destination any faster, so you don’t do so.
Benjamin, I think we are saying the same thing.
LikeLike
William, You are more of a philosopher than I am, and my knowledge of the mind-body problem is obviously superficial, so thanks for providing more detail on the distinctions between different types of physicalism and the eliminative materialism.
As for Scott Sumner’s idea about wage stickiness, I think the problem is that you can’t solve an aggregate demand deficiency with partial equilibrium price adjustments. The wage cut that a firm with collapsing demand for its product would need in order to make it profitable to keep its current work force employed would be so steep that it’s not plausible to believe that it could be implemented in practice. We have no way of knowing whether if it were implemented it would in fact lead to recovery.
I agree that people have personal preferences that often dictate their refusal to accept offers of employment, but the existence of involuntary employment is associated with a willingness to work at the same wage that employed people are being paid. That’s not quite the same situation as your willingness to take an academic position at the same salary that currently employed academics are receiving, because in cyclical unemployment the involuntarily unemployed have previously been employed at the same wage as those who are now employed at that wage. So, like Henry Rech, I don’t think involuntary unemployment can be dissociated from the larger macroeconomic environment.
You wrote:
“I don’t accept a sharp distinction there. People don’t demarcate their choices into microeconomic and macroeconomic decisions.”
I agree, but the point is not whether people identify microeconomic decision from macroeconomic decisions, but how their environment and alternatives change as macroeconomic conditions change. People push the gas pedal in their cars in traffic jams the same way that they do when there is no traffic. But they do not accelerate in traffic jams as they do when there is no traffic.
Benjamin, Those are good points, but I have no particular insight into those labor market anomalies.
LAL,I don’t follow your argument. Rubbernecking occurs when the road that you are driving on is unobstructed, but some event off the road causes enough people, but not necessarily all or even most people, to slow down to look at what is happening off the road. People slow down because enough other people slow down to make it difficult for other drivers not to slow down even if in the absence of any traffic constraints they would not have slowed down to look at what is happening off the road.
LikeLike
I’m denying rubbernecking happens. Or at least I’m denying that what most people describe as rubbernecking is actually rubbernecking. Instead whenever there is a road accident on the side of the road, by law you must slow down or change lanes. The behavioral explanation is redundant. This law and its ambiguity is enough to cause the traffic event we ascribe to irrational rubbernecking.
LikeLike
LAL, I will stipulate that there is an obligation for cars to slow down when there is an accident on the shoulder. But clearly, even if drivers obey that obligation, how much cars reduce their speed when there are cars on the shoulder depends on how the number of cars on the road when there are cars on the shoulder. If there are very few cars on road the speed reduction is relatively small. If there are many cars on the road, the speed reduction can be extreme and the backup in traffic quite substantial. The difference in speed reduction depending on traffic conditions can therefore be said to be involuntary. Similar effects also take place when there is a substantial separation between the roadways going in opposite directions, so that there is presumably no obligation for drivers going in one direction to slow down because of an accident obstructing traffic in the other direction.
LikeLike
There is also an obligation to move over (change lanes). There are several websites that show all 50 US states enforce such laws. That is an important part of this phenomenon that I think needs an analog in a model trying to incorporate ideas from traffic. I don’t see why for example, in a law that requires only that people drive 15mph fewer, traffic would flow any differently than just 15mph fewer. Involuntary only in that some people would rather not obey the law. Voluntary in that most people would choose to obey it and drop their speed exactly 15mph. There is no minimum speed limit but it is merging traffic to fewer lanes that causes the difference in slowing speed. Very similar traffic patterns emerge during rush hour when there is a popular exit and many cars try to queue for it. Interestingly on highways in Texas, I dont find myself having to slow down for traffic stopped or slowed going in the other direction.
LikeLike
LAL, There may be some sense in which people are influenced by the traffic regulations about slowing down and changing lanes in certain circumstances when cars are stopped or damaged but are not obstructing traffic, but I doubt that most people are aware of their existence and it seems far-fetched to think that those regulations are almost ever enforced. I have been driving for over a half century and wasn’t aware of such regulations. But that is really beside my main point, which is that the way people drive is affected by interactions between other drivers on the road over which no individual driver has any control. So your ability to decide at what speed you will drive on a highway is radically constrained by how other drivers are responding in a rubbernecking situation. If traffic flow is sufficiently heavy relative to road capacity, drivers will involuntarily drive at a slower speed than they would have chosen to drive if traffic were lighter. That proposition holds irrespective of the regulations imposed by the traffic authorities.
LikeLike
I’d be willing to bet, that now I’ve stated it, you will start to (re)notice the roadside signs that instruct it. Just as when you look for them you’ll realize there are speed limit signs you have long since ignored on your daily routes having learned in other ways to abide by them (the invisible hand). I’m harping on this point about the law because it is in my view at the heart of the analogy with recessions: it is not the usual case while driving, there are many choices in front of the actors, there is a public effort to coordinate behavior. It is always true that we interact with each other on the road just as we must engage with others’ price and labor choices in an economy. To me this distinction between rubbernecking vs the law is a perfect analogy between the sides of the debate you have presented. Not surprisingly, I agree with Lucas and you believe traffic behavior is driven by animal spirits.
LikeLike
I won’t take you up on your kind offer of a wager about requirements to slow down in rubbernecking episodes; it is indeed quite possible that you are right that I have never noticed the signs instructing drivers to do what you say they are required to do. And I will admit that I regularly and knowingly drive at speeds above the speed limit that is clearly marked. However, my experience is that I am generally passed by more cars than I pass on I-95, so I doubt that I am one of the fastest drivers on I-95, and I think it is at least 40 years since I received a ticket for exceeding the speed limit. But, again, this discussion about the legalities of rubbernecking is all beside the point I was making, and which you have not disputed: that drivers are involuntarily driving at speeds less than the speeds at which they want to drive in rubbernecking episodes, even if one takes into account the legal requirement for drivers to slow down in those episodes.
LikeLike
Hmm, you are right. I don’t think I can dispute it in a meaningful way, but I also can’t affirm it. For me either seems like question begging. I would begin modeling traffic as rational agents on infinite or closed loops with costs imposed to jump to a parallel loop through speed reductions increasing in magnitude as a function of traffic densities on the loops. Randomly, penalties are imposed on agents if they don’t change lanes around certain points on the line. In this kind of model (assuming: behavior can be well defined, equillibria reached [ie uniform density flows across lanes and speeds normally distributed], etc) no agent is ever voluntarily going the speed they want in the usual case because they must abide by the nash equilibrium from the game to play optimally. Or should I say they always play voluntarily because rational agents enjoy playing optimally? My worldview doesn’t seem to allow a distinction. Anyway, I have thoroughly enjoyed arguing this, thanks for the article and engaging my comments.
LikeLike
LAL, thank you very much for engaging me in the discussion, especially after the long hiatus in my response. Your questions were very helpful to me in trying to clarify my own thinking. I think a key underlying difference in how we’re looking at these issues is that you’re looking for an equilibrium and assuming that the players find the equilibrium, and optimize accordingly. I on the other hand, don’t think that players in a non-repeated game find the equilibrium, so that in the actual observed outcomes they’re not typically optimizing.
LikeLike
David,
“Workers are laid off usually because the optimal level of output for the firm under existing demand conditions, is a lot less than the output the firm had been producing.”
Yes, totally agree.
LikeLike