Believe those who are seeking the truth. Doubt those who find it. Andre Gide


Sunday, March 8, 2015

Involuntary Labor Market Choices?

My pal Roger Farmer has a lot of good ideas, but he doesn't always use the best language to express them. In a recent post, for example, Roger asserts the following.
Participation is a voluntary choice.  Unemployment is not. 
The idea that unemployment is voluntary is classical nonsense.
I do not like this language. But before I explain why I feel this way, let me first describe what I think Roger is trying to say. I think he means to say that recessions are socially inefficient outcomes, manifesting themselves primarily in the form elevated levels of unemployment and not in low participation rates. The unemployed are people without good-paying jobs, but looking for good-paying jobs. Good-paying jobs are relatively scarce in a recession (especially for individuals with lower skill sets--the young, those without advanced education, etc.) If you were to interview the unemployed during a deep recession and ask them how they're feeling, most of them would are likely to reply that they are not doing well relative to when they were employed. Economists (classical or otherwise) would say that recessions are welfare-reducing events for most people. The "classical" idea that there is little a government can or should do to help society in a deep recession is nonsense.

I think this probably captures Roger's view fairly well. Notice, however, that nowhere did I employ the adjectives "voluntary" or "involuntary" to describe labor market outcomes. I did not because these labels are not useful (which I why we do not see these terms used in the labor literature). Indeed, want to go a step further and argue that the use of these labels might be worse than useless. Now let me explain why I feel this way.

Let's start with some things I think we can all agree on. First, people are endowed with some time, T. Second, there are competing uses for this time. Let me assume, for simplicity, that there are three uses of time: work (e), search (u), and leisure (n). Think of "work" as time devoted toward producing marketable goods and services, "unemployment" as searching for work, and "leisure" as producing non-marketable goods and services. Third, we can all agree that we face a time constraint: e + u + n = T.

Now, suppose for simplicity that T is indivisible: it must be allocated to one and only one of the three available time-use categories (the allocation can, however, change over calendar time). In this case, a standard labor force survey (LFS) will record e = T as employment, u = T as unemployment, and n = T as nonparticipation (or not-in-the-labor-force, NILF). [Note: the LFS never asks people whether they are unemployed or not. It asks whether they have done any paid work in the previous 4 weeks and if they have not, it then asks a series a questions relating to job search activities. If they report no job search activity, they are then classified as NILF.]

Now, Roger seems to be saying that people have a choice to make when it comes to allocating their time to either work (e = T) or leisure (n = T), but that they have no choice in determining time spent unemployed (u = T). Moreover, the idea that people may choose u = T constitutes "classical nonsense." But is this really what he means to say?

Let's start with a basic neoclassical model. In this abstraction, individuals and firms meet in a centralized market place and individuals are assumed to know where to find the best price for their labor. Put another way, there is absolutely no reason to devote precious time to searching for work. To put it yet another way, the neoclassical model was never designed to explain unemployment--it was designed to explain employment (and non-employment). And so, in the neoclassical model, where search is not necessary, individuals rationally choose u = 0.

Now, you may think this is a silly abstraction and that you want to impose (involuntarily) the state u = T on some individuals. But why? Unemployment is not idleness. Unemployment (at least the way the LFS defines it) constitutes the activity of searching for work--it is a form of investment (that hopefully pays off in a better job opportunity in a world where finding jobs is costly). Individuals not working and not searching are counted as out of the labor force (and even these people may not be "idle" because they might be doing housework or schoolwork, etc.).

So back to our neoclassical model. Since there is no unemployment, the time-allocation problem boils down to choosing between work and leisure. Depending on idiosyncratic considerations (the price of one's specific labor, wealth position, the opportunities for home production, schooling, etc.), some individuals choose work and others choose leisure. In the neoclassical model, these idiosyncratic "shocks" are largely beyond an individual's control. If the demand for your labor declines, it will cause the market price of your labor to fall. You will not like that. The shock is involuntary. BUT, you still get to choose whether to work at that (or some other) lower wage, or exit the labor force. To take another example, suppose that a source of non-labor income suddenly vanishes (involuntary). You may now be compelled to take that lousy paying job. Should we label this outcome "involuntary employment?" If so, then what next? Involuntary saving? (oops). Are all choices to be considered "involuntary?"

This is not the way we (as economists) want to go, in my opinion. In my view, it makes more sense to view choices as voluntary and responsive to the incentives imposed on individuals by the economic environment. If we want to view anything as "involuntary," it would be exogenous changes to the environment that reduce material living standards.  If circumstances change for the better, welfare increases. If they change for the worse, welfare declines. In either case, people can be expected to allocate their scarce time toward the activities that promise the highest expected payoff. What room is there left for the "voluntary/involuntary" distinction? None, in my view.

Let's stick with the neoclassical model for a bit longer, but tweak it the way I did here to permit multiple equilibria. Now, this is right up Roger's alley. All individual choices here are rational and "voluntary."  But this doesn't mean that the economy operates perfectly all the time. Indeed, the economy might get stuck in a bad equilibrium, where employment is low, non-employment is high (and unemployment is still zero). What would Roger suggest here in the way of labels? Is this a model of involuntary leisure?  How does this label help us understand anything? I argue that it does not.

Alright, so I don't find the "involuntary leisure" label useful. So what? Well, I don't want to make too much of this, but I think such labels can lead to muddled thinking. The label "involuntary" suggests that individuals may not respond to incentives (after all, they evidently have no choice in the matter). I think it's better, from the perspective of designing a proper intervention, to view the individual's circumstances as beyond their control, but to respect the fact that they are likely to respond to altered incentives. We are economists, after all -- why would we not interpret the world this way? People demonstrably do respond to incentives! 

I could go on and talk at length about abandoning the neoclassical assumption of centralized labor markets and replacing this construct with a decentralized search market. There is a big literature on labor market search and I'm not about to review it here. If you're interested, read my Palgrave Dictionary entry on the subject here. Suffice it to say that I find no value in interpreting an individual's state of unemployment as "involuntary" either. There are all sorts of jobs out there and I think people rationally turn "ill-suited" job opportunities down to search for better matches (the way I did, when I lost my construction job in the 1981 recession). Sometimes, people get "discouraged" and exit the labor force. These are all choices that people make relative to the circumstances they find themselves in. If we want to design programs to help the unfortunate (some of whom are employed or out of the labor force), then we want to design a system that respects incentives. 

What's that you say? You don't believe that incentives matter? Not for the unemployed? This is what I call nonsense. Consider, for example, the well-known "spike" in unemployment exit rates at the point of unemployment benefit exhaustion (see David Card here: "In Austria, the exit rate from registered unemployment rises by over 200% at the expiration of benefits..."). We see clear evidence that the unemployed do respond to incentives--they do have choices, especially in an economy with so many competing uses for time. Interpreting unemployment as "voluntary" does not mean that we are to have no compassion for the the unemployed. We feel bad for anyone (employed or out of the labor force too) who face terrible circumstances beyond their control. What it means is that we should measure economic welfare based on consumption (material living standards), not time allocation choices. It means is that we understand and respect the fact that people make choices based on the incentives they face. It means that a well-designed policy should respect these incentives.

Let me sum up here. Commentators attach the label "involuntary" to unemployment to emphasize the fact that the unemployed are not typically happy with their circumstances. Fine. But then can the same not be said of many people who find themselves "involuntarily" employed (the working poor, for example) or "involuntarily" out of the labor force (looking after a sick relative, for example)? If so, then how can one unequivocally proclaim that "participation is a voluntary choice, unemployment is not?" It makes no sense to me. I want to ask Roger to stop using bad language. 

36 comments:

  1. Part of me agrees, but another part says "hang on!".

    In a recession, it gets harder to find a willing buyer (of illiquid goods as well as labour), and easier to find a willing seller (ditto). The search/matching trade-off worsens for sellers, and improves for sellers. And it's the opposite in a boom. A very crude way of talking about this is to say there is excess supply in a recession, and excess demand in a boom. And a crude way of talking about that is to say there's "involuntary unemployment".

    ReplyDelete
    Replies
    1. Well, we can both admit it's crude :)

      But why not just say that there's an elevated level of unemployment? What does the adjective do to clarify ideas?

      Delete
    2. The adjective tells us that unemployment has increased because it has become harder to find a job with the same characteristics (wage, distance, etc.) than it was before. It tells us something about the constraint facing those who want a job, and how it has shifted.

      Delete
  2. David
    I am happy with the way you characterize my beliefs in the first paragraph. But not with anything after that.

    The idea of involuntary unemployment was introduced by Keynes in the General Theory. But you already knew that. It is defined as a situation where (in modern language) the ratio of the marginal disutility of work to the marginal utility of consumption is not equal to the real wage. That seems a pretty accurate description of the equilibrium outcome of labor search models.

    Bob Lucas cast a spell over the profession in a series of papers in the 1970s. You are accurately summarizing Bob's view. That view was tied to a three decade long campaign by economists predominately located in Chicago, Minnesota and Rochester (at the time) to discredit Keynesian economics. Tom Sargent reputedly advised his students not to read the General Theory. That was a tragic mistake and we are still suffering from the consequences.

    You are right to assert that the important distinction is between equilibria that are Pareto optimal and those that are not. You are wrong to assert that term 'involuntary unemployment' has no useful meaning.

    I accept your categorization of the allocation of time between three competing ends. Every family, and every member of that family, chooses every day whether they will choose to participate in the labor force. As long as they are in the labor force, they may be employed or unemployed. Those who are unemployed do not choose that state. They must wait for a job offer to appear. In some states, that job offer may take a couple of days to arrive. In others, it make take a couple of years. The activity of waiting for a job, even when it involves active search, can meaningfully be called involuntary unemployment.

    The dismissal of 'involuntary unemployment' from the lexicon of the modern economist was introduced as part of a deliberate attack on Keynesian economics. It is time to roll back that attack. 'Involuntary' unemployment is a useful way of distinguishing unemployment that is part of a social optimum, from unemployment that is not.

    ReplyDelete
    Replies
    1. Roger, I did not write this post to justify past challenges to Keynesian thinking. I think you're getting caught up in past wars.

      Look at this way. I understand your point of view (indeed, the mathematical models you work with make that view very plain). I have a lot of empathy for your view. My understanding and empathy are in no way enhanced by your use of the adjective "involuntary." If it is an involuntary state, how do explain the spike in exit rates at UI exhaustion? How?

      Delete
  3. David,

    it seems to me that you have in mind a world in which someone who is searching meets potential matches, bargains or receives a wage offer, and then chooses to accept an offer or to continue searching.

    What if we imagine a world in which search need not result in any potential matches at all? Can I than call an unemployed person "involuntarily" unemployed? In that case the very act of search makes the lack of employment involuntary, I search because I'd prefer to be employed yet I'm not, at no point did I choose to keep searching instead of taking a possibly poor job offer.

    I wonder if this isn't closer to what Farmer is thinking. Firms pay a cost to recruit, if too few pay the cost then the number of job openings is too low to absorb those searching even if every job offer, no matter how poor, is accpeted. People search because they can't know a priori that they'll be one of the ones with no offers. The ones without offers are surely involountarily unemployed, no?

    ReplyDelete
    Replies
    1. Yes Adam - I agree with your interpretation

      Delete
    2. Yes, Adam, I think that IF a person does not meet any potential matches at all then the term "involuntary" is appropriate.

      But ask yourself this: Do you honestly believe that people do not alternative employment opportunities? I admit that most, if not all of them might be lousy along some dimension, but they exist nevertheless. And people can rationally and justifiably turn them down (voluntarily). No?

      I am curious as to why the label seems only to apply to the unemployed. Why not those who are compelled to be out of the labor force? Those that are compelled to pump gas to feed their family, rather than returning to school? Why are these labels not widely used in these cases? The asymmetric treatment here makes no sense to me.

      Delete
    3. David,

      Let's imagine a world in which there are 3 people searching for jobs and only 1 job opening (I am the employer who's recruiting).

      I may first encounter searcher 1, make a low wage offer which is refused and then move on to encounter searcher 2 with the same result.

      Now, searcher's 1 and 2 have both had the chance to be employed but chosen to continue searching (they do not know there is only me recuiting).

      I then encounter searcher 3, make the same low wage offer and this time it's accepted because since searcher 3 has spent two "periods" searching without encountering any potential match he/she has figured out that there is a dearth of openings.

      The question is, are searchers 1 and 2 involuntarily unemployed. In both cases they were offered a job and chose to continue to search, but on the other hand we know from the aggregate number of searchers and rectuiters that 2 of the 3 searchers would be unemployed. Had searcher 1 taken the job then the others never even meet a potential match, however they did do that.

      I think what I'm tryiing to say is that Farmer is using the term "involuntary unemployment" to describe an aggregate outcome, not an individual outcome. In the example I think we can say that the unemployment of at least 2 of the 3 searchers was "involuntary" even though in the actual event all 3 got to encounter a potential match and make a choice to accept or reject a wage offer.

      The key to all this is that firms pay a cost to recruit and thus aggregate recruting activity can be too low to possibly employ all those searching. Thus, regardless of individual choices we know for sure that someone will search and not find a job and that person's unemployment is unambigously "involuntary". That is the situation Farmer is using the term to describe and I think it is entirely appropriate language.

      Delete
    4. Adam P, I do not agree at all with either your premise or your conclusions.

      First, suppose there is only one job and three workers. Are you kidding me? What world are you describing? The situation is the exact opposite. It is as Alchian is reported to have said (according to Jim Rose below):

      "Alchian said there are always plenty of jobs because to suppose the contrary suggests that scarcity has been abolished."

      So, first off, let's get serious. There are always more things to be done than there are people to do them. Throughout history it is *leisure* that has always been wanting, not work.

      Second, understanding what I just said to be true in no way suggests any moral statement concerning the choices people make in terms of how they choose to allocated time. Yes, of course, equilibria are generically suboptimal. Unemployment may be too high OR too low (it is easy to derive such results). We already have labels for this: suboptimal unemployment rate (why do we need the adjective "involuntary?").

      Third, how do you propose to decompose "voluntary" vs "involuntary" unemployment in the data? Oh, and while you're at it, please do the same for employment. Yes, if you find the label involuntary unemployment useful, I think you must also find the label involuntary employment useful too, no? If not, why not?

      These are just silly, outdated labels, in my mind. They do more to confuse thinking that to clarify it. My 2 cents worth, at least.

      Delete
    5. David, surely you accept that not everyone who wants a job is actively searching, right?

      So why do you deny that there can be firms with jobs they'd in principle like to fill but aren't actively recruiting?

      Point me to where in my example i said there was only 1 job that needed doing? i said only one firm was actively recruiting.

      Again, this is Farmer's point. If there is a cost to recruiting firms can have a job they'd in principle like to fill (if recruiting were costless) but still not actively recruit because the expected cost exceeds the expected benefit. After all, when people who want a job stop searching surely it can only be because the expected cost exceeds the expected benefit?

      And again, if the number of firms recruiting is less than the number of people searching then someone must fail to find a match.

      Delete
    6. David, surely you accept that not everyone who wants a job is actively searching, right?

      If you define "want a job" as "wanting a good, fulfilling, and high-paying job with great career prospects but realize that such jobs are presently or forever out of reach" then yes, I accept that.

      So why do you deny that there can be firms with jobs they'd in principle like to fill but aren't actively recruiting?

      I don't deny that. I have published papers that have this property.

      Point me to where in my example i said there was only 1 job that needed doing? i said only one firm was actively recruiting.

      OK, sorry. Then you accept that there are many jobs that need doing. Far many than there are people on the planet. Good, I agree.
      Again, this is Farmer's point. If there is a cost to recruiting firms can have a job they'd in principle like to fill (if recruiting were costless) but still not actively recruit because the expected cost exceeds the expected benefit. After all, when people who want a job stop searching surely it can only be because the expected cost exceeds the expected benefit?

      I agree with this, but so what? I do not see where "involuntary" unemployment enters the picture if, as you implicitly admitted, there are far many more jobs to do than there are people.

      Read Lucas' statement below, posted by Jim Rose. He says, and I agree, there is a voluntary and involuntary aspect to unemployment. Involuntary in the sense that nobody chooses bad luck over good. But still, given bad luck, people have choices. Not great choices, but choices nevertheless and this is what makes labor market choices "voluntary."


      And again, if the number of firms recruiting is less than the number of people searching then someone must fail to find a match.

      Delete
  4. What about "classical" rationing, in which no additional work is demanded, although supplied, at the "going wage" due to e.g. market power or other (real) wage frictions?

    ReplyDelete
    Replies
    1. In the classical setting, rationing can lead to underemployment, but no unemployment (which the LFS defines as full time active job search).

      Delete
  5. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. Hi David
      Here is a LINK to my thoughts on your post.

      Delete
    2. Thank you, Roger. Good fun here. Just like the old days. :)

      Delete
  6. This makes a lot of sense to me. For the most part, we have no control over the situation with which we are faced. And "involuntary" doesn't mean "faced only with really terrible decisions."

    However, I think it is possible to be unable to find a job at any legal price. Lots of young people would like to work, for money, but can't. Nothing they can do would add value greater than minimum wage, and there's only so many commission-based jobs and waiting jobs out there.

    Now, one can say "well, they could work for nothing," and that's sort of true, but you have to call it volunteer work or an internship.

    I wonder what the inability to find paid work, because of minimum wage laws or child labor laws or some such, would be called?

    ReplyDelete
    Replies
    1. Suppose my labor is worth $5 an hour and the minimum wage is $10. Then there is no point in searching for work--I will not be unemployed, I will exit the labor force. I suppose it would make sense in this case to label the person non-employed for reasons beyond his/her control. But the same would hold true for a person who is non-employed because they have to look after a sick relative. Who cares about these labels? I still don't get it.

      Delete
  7. Isn't the issue here that, if unemployment is caused by wages which are stuck above the market-clearing level, then both employers and the unemployed would be better if they could negotiate at a lower wage? And thus *employers* are responding to incentives by not hiring, but it doesn't really make sense to ask if the unemployed are responding to incentives, since the lack of market-clearing prevents them from even entering the market.

    (I'm not sure how to translate this into the language of search models)

    ReplyDelete
    Replies
    1. Basil, the Keynesian view of unemployment is that it is the outcome of a coordination failure. Sticky wages play no role in the analysis (indeed, Keynes wrote how he believed flexible wages would make matters worse). These are the types of models Roger works with. Individuals respect incentives but collectively, these incentives are all screwed up (even if markets clear in the conventional sense).

      Delete
  8. Adding my twopenceworth, with a picture: http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/03/involuntary-unemployment-as-worsening-trade-off.html

    ReplyDelete
    Replies
    1. I don't have much to disagree with there, Nick.

      Delete
  9. I like Nick Rowe's definition of "involuntary unemployment".

    ReplyDelete
    Replies
    1. What is the definition of "voluntary" employment then? And I thought he made it clear that it was pointless to label such things. The issue is how the environment (trade-offs, in his language) changes and (my point is) how people respond to this new environment.

      Delete
  10. I am a bit boggled listening to this discussion, since it leaves out the fact that your worker, a living creature, does not have the unrestricted choice of how he will divide his time. He can work, not work, or search for work? Really? This conflates "working at a job" with "acquiring the necessities of life," but also, and very oddly, assumes that while he is searching or not participating, he still has access to the necessities of life in enough quantity that his leisure, or his searching, (or even his working!) is not interrupted by death.

    Absent a guaranteed income or a decent system of social supports, the above discussion is nonsense, having left out an essential element of the system.

    Noni

    ReplyDelete
    Replies
    1. Try not to be so boggled, Noni. Absent a decent system of social supports, many people are "forced" to work. So why then do we not have a category called "involuntary employment?" My whole point is about language and how it potentially obfuscates the real issues at hand.

      Delete
  11. I would stay away from "voluntary" and "involuntary" simply because using these words can open the door to moral judgments, even if they were helpful in clarifying certain points. Economics really must always remain a positive science, otherwise we lose legitimacy of our results. As such, using more technical or clinical language would, I think, be preferred.

    ReplyDelete
    Replies
    1. Yes, that's a valid concern as well, I think.

      Delete
  12. I think you are spot on at the end about "unhappiness" in employment outcomes, and the overall theme of voluntary/involuntary being vague.

    So, in light of this, would it be better to use "involuntary" as "actively searching for a job, but currently not employed" and "voluntary" as "not actively searching for a job and not currently employed"?

    This is how I have always thought about involuntary/voluntary unemployment, and it really didn't occur to me the implications of the words involuntary/voluntary.

    Though my definitions (and better words associated with the definitions would be useful) would basically just sort people into 2 categories, which may not be useful for thinking about employment. But, I think it's what many (though not all) people think about when they say "involuntary" unemployment. It's a "I would rather be employed than not" situation (though with actively searching as a condition).

    ReplyDelete
    Replies
    1. HG, once you start along the voluntary/involuntary distinction, you run into all sorts of problems. How does one measure "voluntary" unemployment, for example? Or "involuntary employment?" Who cares how their labor market choices are labeled? We should care more about their material living standards (consumption).

      Delete
  13. Robert Lucas in a famous 1978 paper argued that all unemployment was voluntary because involuntary unemployment was a meaningless concept:

    "The worker who loses a good job in prosperous time 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 undesirable 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."

    I agree that we all make choices subject to constraints. To say that a choice is involuntary because it is constrained by a scarcity of job-opportunities information is to say that choices are involuntary because there is scarcity. Alchian said there are always plenty of jobs because to suppose the contrary suggests that scarcity has been abolished.

    Lucas elaborated further in 1987 in Models of Business Cycles:
    "A theory that does deal successfully with unemployment needs to address two quite distinct problems.

    One is the fact that job separations tend to take the form of unilateral decisions – a worker quits, or is laid off or fired – in which negotiations over wage rates play no explicit role.

    The second is that workers who lose jobs, for whatever reason, typically pass through a period of unemployment instead of taking temporary work on the ‘spot’ labour market jobs that are readily available in any economy.

    Of these, the second seems to me much the more important: it does not ‘explain’ why someone is unemployed to explain why he does not have a job with company X. After all, most employed people do not have jobs with company X either.

    To explain why people allocate time to a particular activity – like unemployment – we need to know why they prefer it to all other available activities: to say that I am allergic to strawberries does not ‘explain’ why I drink coffee.

    Neither of these puzzles is easy to understand within a Walrasian framework, and it would be good to understand both of them better, but I suggest we begin by focusing on the second of the two."

    ReplyDelete
    Replies
    1. Thank you for this, Jim. My view on the matter has evidently been greatly influenced by Lucas.

      Delete
  14. I like your parallels of unemployment with the marriage market.

    Would it make sense for people to go around saying that they are involuntarily unmarried or involuntarily single?

    Of course you could go out and marry the first person you meet in the street, if they will have you, an important proviso, but I think both sides would know that that would turn out to be a pretty poor match.

    The notion of involuntary marriage and involuntary single illustrates the notion that talking about search and matching as voluntary unemployment and involuntary unemployment just doesn't help clarify what you attempting to understand.

    ReplyDelete
  15. Starvation, a voluntary choice!

    ReplyDelete
    Replies
    1. Well, first off, one can choose to starve while looking for work or while out of the labor force. But seriously, in wealthy economies, people do not starve. They are, however, depending on circumstances, compelled to work at menial jobs just to support themselves and their families. I would call this "involuntary employment." But you and Roger appear to have no compassion for the working poor.

      Delete