SEBASTIAN MALLABY: I think we can probably get started. Thanks for coming. First thing to say is please turn those things that go "bleep" off. This meeting, unlike most Council on Foreign Relations meetings, is on the record. Normally we have a please plagiarize rule which is the rule that says, you know, you can use all the information but just don't attribute it to us. We've reversed this, perhaps reflecting my journalistic background.
I'm Sebastian Mallaby. I direct the Center for Geoeconomic Studies here at the Council. The point of the Center for Geoeconomic Studies is to look at stuff that has an interaction between foreign policy and economic capital flows, climate change and so forth. And the topic today is a good example of this. Clearly, if the consensus in favor of trade cannot be repaired, if the polling data that we see on hostility toward globalization cannot be turned around, the consequences will go beyond economic growth, suffering -- there will also be foreign policy consequences as well. And it seems to me if one result of less multilateral energy in trade is a growth of bilateral and regional trade deals, the prospect of the world breaking up into sort of regional blocs might well, in the end, have significant foreign policy knock-on effects.
And this is also a good topic that we have today, because I'm sure all of you have already read every 531 pages of Alan Greenspan's book by now. You've had, after all, five days, and it's reasonable to expect you to have done it. If you haven't read it, you'll have read the 841-word version that Robert Samuelson published this week in The Washington Post, which explains that, you know, we've been in a golden moment in terms of economic growth with this productivity gain between around 1995 and 2005 being unusually high because of the absorption of the simultaneous effects of China's entry into the global labor market and also the sort of delayed effects of the IT revolution finally percolating into productivity gains. And Greenspan's point is that this moment may be ending. And if so, we face slowing growth at a time precisely when the baby boom or baby bomb is going to hit us.
And so thinking about policies that could maintain a political consensus domestically for pro-growth policies is extremely relevant. And one of the people who's thought hardest about how to think about that consensus domestically is Professor Bob LaLonde who's right here from the University of Chicago. We have a man bite dog angle here, of course -- a Chicago university economist recommends intervention in the markets. And as he's going to explain to you, that's what he's doing but for reasons that complicate the picture.
Bob is a professor at the Harris School of Public Policy at the University of Chicago, and he's the author of the report that we're rolling out now which is "The Case for Wage Insurance."
So, Bob, your starting point here is that when people dislike globalization and they dislike pro-growth economic policies, they actually have a pretty good reason in some cases. Can you explain that?
ROBERT J. LALONDE: Yeah. I think the way that those of us who have supported pro-growth policies, trade and so on have often missed the point that people who fear job loss are right. That there are, although the evidence is overwhelming that the economy benefits from trade, there is a minority of people that are hurt very badly from it, and that the policies that are in place really don't address their needs. And it's a little bit -- I think it's a little bit like owning a house where you're not allowed to buy insurance against the possibility it burns down. So let's imagine that that was the rule, we'd all feel very insecure, even though the chance of any one of our houses burning down would be low. The idea that we couldn't have insurance against this catastrophic event would make us all feel insecure. I think that's really the crux of the problem is that more people feel insecure about the prospect of job loss in connection with trade and other economic events than we'll actually experience. But they're right that if the event happens then it will be bad.
Now, the other point I wanted to make is that -- and I tried to -- when I worked with Sebastian on this paper, he wanted me to focus on trade, and I kept saying no, no, no, trade is not the reason why people are losing their jobs, it's technological change. That's a much more important reason. And I want to make a point is that this whole issue of displacement is an old issue here in the United States. At least as far back as I can remember, in the 1950s it came up with regard to people losing their jobs in connection with the 1958 recession and the worry about automation and the notion that automation would eventually eliminate the need for people to work. And what would we do if no one could work? And a lot of the policy at that time was really geared towards re-training people who were losing their jobs because of technological change. And the '50s meatpacking industry is one example. The steel industry was another at that time. And it more recently became connected with trade.
But it's important to keep in mind that people are losing their jobs for a variety of reasons. For the most part, job loss is not a big deal for most people. You go through a period of temporary job loss. You get another job that pays similar to what you've got before. But for a minority of people, the job loss is catastrophic. And it's in the sense that the problem is not losing your job, the problem is not being unemployed. The problem is when you get a new job, it's not going to pay as much as your old one, and that's going to be true for the rest of your life.
And so again, the problem is not for someone who's close to retirement. So again, you wouldn't worry about the effect of trade on people who are, you know, close to retirement. So people in their mid 40s really stand the most to lose from pro-growth economic policies because they're the ones most at risk in these kinds of events.
MALLABY: Can you quantify that a bit more? You've put some numbers in the paper. The problem that you're talking about is people in their 40s who face another 20 years of being in the work force, and the size of the wage hit is what? And how many people in fact suffer this sort of catastrophic wage hit?
LALONDE: It's a little hard to estimate for sure. But I estimate probably about 250,000 people a year are taking a wage hit of more than 20 percent of their pre-displacement earnings. That is, that's the kind of loss they can expect to have, not just when they get re-employed but roughly for the rest of their careers. And so you can imagine being in a position where, you know, you were once earning 40 (thousand dollars) or $50,000 a year, and you've been working for a long time anticipating you would make that much money for the rest of your career and that you would have raises that are corresponding, commensurate with what -- the raises typically people get. And suddenly you find you're working and making, say, $35,000 a year. You're not poor, but nonetheless that's going to change your lifestyle a lot, and that's really the crux of the problem for many people.
Again, this is an issue that doesn't really affect the poor. And a lot of policies that are trying to address this problem really assume it's for people for who are poor. It's not a poverty problem. It's a problem in the middle class. And it's the middle class that are insecure about this issue, not the poor. The poor don't face this issue, because they don't have -- they lose their job a lot, but they don't have wage losses associated with losing a job. So if you're poor and you're working at 6 (dollars) or $7 an hour and you lose your job, your problem is that you don't have income while you're unemployed. When you get another job, you're going to make about what you were before. So this is not a problem with the poor. And so the policies that address this problem should not be thought of as an anti-poverty program but should be thought of and recognizing that it's a problem in the middle class, those earning 50 (thousand dollars), 60 (thousand dollars), $70,000 a year who's going to have a permanent change in their income.
MALLABY: So you're estimating 200(,000) to 250,000 people a year losing their job and suffering this kind of 25 percent hit to their wages. And when you calculate that out, a loss of 25 percent -- i.e., $10,000 a year or something like that -- over 20 years that you would be in the work force thereafter, that's how you get to the idea that this is the equivalent financial hit to your home burning down.
LALONDE: Right. So in the paper, we have a constructed example of a person who is about 40 years old making $40,000 a year. Keep in mind about the people that are most insecure here are people who are good workers. They're not people who need training to teach them how to hold a job or get a job. They're people who have very good work histories. Suddenly, they become displaced from a job, and they can find another job within a reasonable amount of time, but they're making $30,000 for the rest of their life. So they're going to work for another -- (inaudible) -- 20 years or so. Now, that loss is equivalent of losing $165,000. So you could say to that person we'll give you a choice. You can keep your $40,000-a-year job for the rest of your career and retire, but to do that we're going to charge you $164,999. That person should be willing to take that deal. It would be a good deal for them to take that deal. It's the equivalent of losing $165,000.
Now, that number is interesting, because I think at the time I did the calculation, the typical home price was about that amount. So it's like losing your house.
MALLABY: Could have changed by now. (Laughter.)
LALONDE: Could have changed by now, yeah. I mean, this could be -- what -- could change again, yeah, exactly, yeah.
MALLABY: Let's talk a little bit about the policies that other people advocate as a response to worker dislocation and that in the paper you explain why you are not impressed by them. The one that most people cite is training. What do you think about training?
LALONDE: Let me go back for a minute and talk about employment services. We've been pretty good in this country about developing a set of reemployment services, job assistance services, and they are cost effective. There's a lot of evidence -- less targeting toward youth. They do work pretty well and getting people employed faster and providing -- their incomes rise as a result of these services. The problem with those kinds of services is it's a problem for a displaced worker who's a middle-aged worker who's not getting a job. They do get a job. So getting to work faster or slower really doesn't address the issue. So if you tell someone who's 45 years old look we're going to give you help getting a new job, that really shouldn't make them feel much better about what the government's doing to help them deal with this kind of uncertainty in their lives because the problem of being unemployed is not a big problem for them. Their problem is that the new jobs pay a lot less than the old jobs, and they're going to live like that for a long period of time.
So although I think that reemployment services are cost-effective services, and you should certainly make these kind of services available to this population, keep in mind that it doesn't really address the problem or any insecurities that they may have, although they are cost effective.
Now, training -- (inaudible) -- could work better. I mean, training -- the purpose of training is to make people more productive, is to raise their wages, to help them get better jobs. And here's the problem with training is that, first of all, we have no experience with the level of training that these people will need to recoup their earnings losses. The kind of training services that we have largely supported in this country are really what I call low-intensity services. Maybe you're spending $3,000 or so per person, maybe $5,000 or so per person, all right. We have a lot of experience with that. It works very well for women, particularly economically disadvantaged women do well with that kind of training. But we have very little experience of how it does for 45-year-old men, for example. I've studied this. I think one of the few studies I know of is one of mine that I've done with Dan Sullivan and Lou Jacobson. And you know, you can teach old dogs new tricks. They do benefit from training, but the point is that training does not address the kinds of needs they have.
So let me give you an example. Let's say you're going to ask me, how much training would a person need who's making $30,000 today -- is a $30,000-a-year worker, that's what their worth in the labor market -- and you want to make them into a $40,000-a-year worker? You want to increase their earnings permanently by $10,000 a year. Well, according to what we know about the returns to training -- and this an issue that's probably been as studied as any issue in economics -- what the gains are from schooling and training -- that training program would cost about $100,000 -- $100,000 is what you'd have to expect. Could a training program where we spend 3 (thousand dollars) or $4,000 work? No, it won't work. If someone tells you it's going to work, are they right? No, it's not going to happen. The reason why it can't be right is because we know what the returns are to formal schooling. We know what the returns are to college education, and they're not anywhere near that. So the idea that you can take the kind of training programs we currently sponsor and sort of go to a place where we can say we can help you and it's going to help address a lot of your problems is just wrong. It just simply will not happen. It's unrealistic. We have a lot of evidence it won't work. We know it won't work.
Now, let me go another step farther and say that suppose we were going to be more aggressive and put together some meaningful training initiatives for this population. The other problem we have is we really have no experience in how to do it. It's not something we can sort of say okay, we're going to offer you this two years of training, you know, spend all this money on your research training programs. We're going to give you subsidies -- we're going to subsidize two years give you a generous stipend so you're willing to do this. (Inaudible) -- too many people are in their mid 40s, they simply can't take two years off from not working. And that indeed is part of the costs of retraining them is that these are not people who are jobless. These people are going to find a job. So part of the cost of training is the fact that they're going to lose money they could be making while you're retraining them. So they're going to require a stipend to be retrained. You can't just simply say here's -- we're going to subsidize your tuition, go to college for two years or go to vocation training for two years, and this will solve your problems. They're not going to be able to do it.
So it's very expensive, and we don't have much experience doing it. (Inaudible) -- implementing a program like that is still open for a lot of debate and question. And I would question it simply because we don't know how to do it. So I think that's my concern about training is that it really is a false promise at the level which we are offering it. There's no way that it can address the problems of this population. So if someone comes to you and says I've lost 25 percent of my income permanently and I'm going to take a training program that's costing about $5,000 per program, that may raise your earnings somewhere in the area maybe 1,000 (dollars), $2,000 per year for the rest of your life if you're lucky. But that's not going to address the problem that you have of having lost a job.
So the economic insecurity that people feel about losing a job like that is well-founded and well-justified. There's nothing in place to address their problems. And probably the next thing you want to talk to me about is the unemployment insurance. And we think we have unemployment insurance, doesn't that help? And the problem with unemployment insurance is it's structured exactly the wrong way for this population. Unemployment insurance is not meant to address displacement. It's meant to address temporary loss in income. So it provides a little bit of money for six months, sometimes nine months. If you're in part of TAA, it's provides a little longer than that if you stay in training, which is a whole other problem with the way the program is designed. You certainly don't want someone not working in order to get training if they don't have to, all right. So if someone wants training and work at the same time, they should, and you don't want to structure a program like that.
But anyway, the problem with unemployment insurance is that it really addresses the problem -- you only get money if you're unemployed. You don't get money if you're re-employed. And the problem for this population is not being unemployed. Sure, they would rather not be unemployed. But their problem is that when they get new jobs, they don't get anywhere near the income they had before. So unemployment insurance is really not structured for them.
MALLABY: I mean, in round terms, if you have an income of $40,000 a year and you are unemployed, even for a long time like three months, that's a $10,000 hit to you. But the kind of hit you're talking about is a $10,000 hit per year for 20 years.
MALLABY: So it's like one-fifth of the UI is (tacking ?) a problem that's one-twentieth -- 5 percent I mean -- of the --
LALONDE: Exactly, yeah.
MALLABY: And so just to go back on your numbers, in the event that you were to, say, double the existing training promised to a dislocated worker and say we're going to do a really big program, a $10,000 program, and your wages have gone from 50 (thousand dollars) to 35 (thousand dollars), the $10,000 training program on a 10 percent return is going to go from 35 (thousand dollars) to 36 (thousand dollars). Is that the sort of --
LALONDE: Well, that program would be a very generous program by our standards.
MALLABY: That's the point. But even a very generous program gets you --
LALONDE: That may get you -- from the way you've described, it would probably get you maybe $4,000 increase in income.
LALONDE: Expect an average. Some programs do better, some of them do worse. But on average, it's implausible to think it's going to be much higher than that. Otherwise, we would -- I mean, if you thought the programs would be more effective than that, you would think that we should tell our children not to go to school but to take these programs, because the benefits are much larger. So it's hard to believe that these benefits can have a lot bigger effect than, say, a good college education.
MALLABY: Let's take another thing that people say. You know, rather than have wage insurance, why not prevent the jobs from being lost in the first place?
LALONDE: That's a good point and a good question. I mean, I think that the question we have to think about there is if the purpose of the -- if that's the policy we're going to -- if that's the way we're going to try to address this problem that people feel insecure about job loss and try to save the jobs, and there are many programs that do that both here in the states as well as outside the United States. Many of them geared towards trying to retrain the workers to make them more productive so the firm doesn't have to close down.
The difficulty with that is that there's nothing wrong with jobs being created and being destroyed. There's nothing wrong with the fact that sometimes plants close down and people lose their jobs. That's part of how the economy's supposed to work -- that there's a natural -- we have a very dynamic economy in which there's many jobs are created and may jobs are destroyed. About roughly 15 percent of all jobs are created each year, and 15 percent of all jobs are destroyed each year. So even in a year where there's no job growth, there's a tremendous amount of churning. And that's thought to be a very good thing for our labor market in terms of the way it promotes growth. We have a very flexible labor market that allows the labor market and labor resources to be reallocated very quickly. And we have to think of that as a positive thing about the way our economy works.
Now, if you thought that we should try to -- if you thought that was a bad thing, that there's some kind of market failure that we need to be addressing to identify why is there too much churning. My recommendation would be to look at the policies that create too much churning and try to address those policies first before we try to do anything else to try to help particular workers, per se.
But in my view, there's very little evidence there's anything wrong with the amount of churning in our economy. It seems to be an asset to our economy. So you don't want to necessarily save the jobs unless there's some market failure that's creating too much job destruction.
MALLABY: So the point is that an economy that churns in the sense that it's churning the capital stock, it's churning the work practices, the management ideas about how to get more productivity, if you're innovating in terms of management process and in terms of capital stock, you have to also change around the way people are allocated in the work force. And therefore, it's a necessary --
LALONDE: Right. In fact, one of the myths about recessions in this country is that myths are an opportunity for the economy to reallocate its labor force. That actually doesn't happen. Most reallocation in the economy occurs during good times. So there's much more movement across sectors and reallocation of the economy occurs during the good, good periods. So recessions don't have this sort of sense that sometimes I've read about that it's got a cleansing effect on the economy, it sort of cleans things out. It doesn't actually do that. Actually, the time of cleansing in the economy -- a growing economy, again -- people are being reallocated occurs during good periods, not during bad.
MALLABY: Let me ask you something else -- I mean, Matthew Slaughter of the Council who's an adjunct fellow here wrote a piece in Foreign Affairs a few months ago, which got quite a lot of attention. And his response to sort of the same problem that you're looking at, I suppose -- namely, that there's increasing evidence of both. So polling evidence of dissatisfaction with a dynamic economy and globalization and also objective evidence of why people are, you know, right to be dissatisfied in some sense, because the gains from this system have been distributed unevenly. So Slaughter's proposal -- and this is a guy who worked in the Bush White House -- is radical redistribution through the tax system. Whereas your proposal sort of tugs at wage insurance for 250,000 people who are going to use it per year. Can you talk a bit about this difference between a sort of broad-spectrum antibiotic and a very targeted response?
LALONDE: Yes. I think the way I look at this proposal is a little bit like the way I look at the prescription drug insurance plan that Congress passed earlier in this decade in that the temptation when we create policies is to create policies in which a lot of people share in the benefits. And in fact, that's not what an insurance program should be about. So for example, I buy life insurance, and I sure hope I don't have to use it. Or I have house insurance, and I sure hope it doesn't have to be used. And the problem is is that the temptation, though, is to have programs that sort of distribute the gains across a wide variety of people, even though, in some sense, they don't really need it, or it's not appropriate to target it to them. The idea of sort of making the tax code more progressive in such a way that it would help to reduce the losses associated with, say, displacement is a very blunt way to do it. It's a very crude way to try to address the problem.
So what you really want to do is insure people against the risk of losing their jobs and not so much the job-loss period where they don't have a job-loss but against the problem that when they're re-employed they don't make as much money. And there's no way we can tinker with the tax code to address that problem, all right. You have to identify who those people are and sort of target the benefits to those people.
So one politically unattractive part of my proposal would be that in any given year, the benefits are going to go to very few people, because very few people experienced the event. But those who do experience the event are experiencing a very large loss. So what I'm arguing is that we all pay premiums for homeowner's insurance. I would hope that none of us have had to use them recently. But if we did, we would benefit greatly from having that insurance available, and that's the idea behind this proposal is that very few people will get the benefit, but those who get it really do, in some sense, need it, all right. But we would all feel more secure if we were in a position knowing that if something bad happens it's going to be there, just as we all benefit from having homeowner's insurance. It's the same way.
MALLABY: So you bring up homeowner's insurance. I think my homeowner's insurance is with State Farm. I mean, why can't the private sector deliver the wage insurance that you're talking about?
LALONDE: It's a good question, and I've been asked this a lot. And you know, one argument for why we don't have private insurance in the unemployment insurance market is because the government's already there, and it crowds out the private sector.
MALLABY: That's in the UI you mean.
LALONDE: That's just in the UI. But I've already said that UI does not really address the kind of insecurity, the risk that these displaced workers are facing. So a way I look at it is that in some sense there is no competitor out there. The government is not in the market to provide insurance for displaced workers. So I think one of the best evidence that we have that you need government intervention is that the private markets had plenty of time to address this problem. This is not a new problem. It's been here for at least 50 years, and yet no private market's come about to fill the void. So that's just sort of empirical evidence suggesting that well, maybe the (private sector ?) has to come in.
Now, there's also other reasons why, and one of the main problems has to do with the fact that the way you have to structure these plans, any kind of wage insurance plan, is that you have to sort of tie the premium to someone being re-employed. The reason why it's -- when someone loses a job, we as outsiders looking at them, we really don't know what they've lost. We don't know what kind of job they can get. So for some Latin American countries I believe in South America have a -- (inaudible) -- kind of unemployment insurance is that you get a lump sum when you lose your job. So each year you work, more money goes into like an account. And when you lose your job, you get the whole sum. Now, there's some very attractive features of that kind of plan, because what it means is that when you go look for work, you don't lose your benefits. So you have an incentive to go and get a new job as quickly as you can, get a good job as fast as you can. That's a very desirable feature of an unemployment insurance system to have.
But the problem in this case is that you really don't know what a person's lost, all right, until you go out and see what kind of job they can get. So when our house burns down, we have someone who comes out, an insurance adjuster comes out and looks at the house and decides okay, this is what you've lost, this is what we're going to give you, all right. We're not going to give you just any set of money, we're going to give you what we think you lost. Now, when you lose your job, you don't really notice -- you can't go and look at the person and say well, you look like someone who has lost, let's say, $100,000. There's no way to look at that until you see what the person actually does. So you actually have to tie the plan to people's actual behavior.
So you might ask, well, why not insure them against the full loss? Why can (only ?) my insurance pay 50 percent of the loss? The reason for that is because you create disincentives for people to continue looking for work once you insure the full loss. Let's say I want to give you the full difference between your old wage and your new wage as a way of trying to insure against your loss. If you do that, people will stop looking for work. And a lot of job searching in this country actually goes on what I call on-the-job search. There's probably more job movements -- there's much more job movements between job-to-job than there is between non-employment to job. And we know a lot of those movements are actually very productive movements, particularly during the early years of one's career. But the early years that you have a job, it's moving from job to job that's a very productive way not only to increase your own earnings but to increase worker productivity.
So you could argue, for example, some of the growth in productivity occurs simply because there is movement, people who change jobs, and most of the job changing that occurs goes from job to job, not from unemployment to job. So that's why you don't want to insure the full risk, because you want to give people incentives to continue looking for work while they're employed. So I get a job that pays, say, half of what I had before, and I'm going to cover half that loss, so I'm still off by 25 percent or so. Then I still have an incentive to go and find even a better job, all right. Where if you cover the full loss -- someone would say well, I'm just as well off as I was before. I kind of like this job. It's convenient for me to be close to home or whatever. I'm not going to look for work, all right. And that's the kind of activity you want to prevent because again, job movements are very productive in this country. The economy gains, industry gains from people moving from job to job.
MALLABY: So you're saying that the idea is that to prevent the moral hazard of disincentivizing people from finding another job, you don't insure the full wage loss, you insure 50 percent of it.
MALLABY: Perhaps you could just run down the other features of your proposal. How long would you be covered for? How do you distinguish between somebody who's being fired for doing the job badly versus somebody who's the, you know, deserving victim of a technological change? Can you just sort of lay out what you're actually proposing?
LALONDE: Well, I would certainly use the same criteria used on unemployment insurance. That someone who's fired for cause would not be eligible for the benefit.
MALLABY: How would you distinguish, though?
LALONDE: Well, the same way. If someone files a claim, there could potentially be a hearing and it would resolve the same way. So someone who's fired because -- the typical ways people get fired in this country I believe are still absenteeism, theft and substance abuse. Someone who's fired for those reasons would be not eligible. And someone who's fired for doing a bad job, I would tend to make them eligible only because people rarely do get fired for doing a bad job. That's not the typical way of people who get fired. And the other point is that the issue about whether someone's doing a bad job is not so -- it's very hard to document whether it's true or not. So I think you'd want to include that. As long as the person would have been eligible for unemployment insurance, they should be able to afford wage insurance. And you could use the same mechanisms to determine that.
MALLABY: And they would be covered for how long, how many years?
LALONDE: Well, I think that's the policy question. There's two problems I think with existing wage insurance proposals. One is that they only go for two years. And so you get wage insurance for two years, and then it's cut off. And that's I think a huge problem, because it doesn't really -- for someone who's 45 years old, that's almost meaningless. It doesn't really address the problem that you're going to spend the rest of your life working at much lower earnings than you expected. And so from your standpoint, your labor union or your congressman should still be fighting to keep your job. So it needs to on many more years than that -- you know, potentially 20 years if you like -- but again, that would depend on how much policymakers wanted to spend.
The other problem with existing proposals is that they tend to cap the payments of people earning less than $50,000 a year. And the problem with that, again, is what I said earlier, is this is a problem the middle class faces, not the poor. So a lot of people are going to be 50 (thousand dollars), 60 (thousand dollars), $70,000 would certainly consider themselves middle-class workers, and sometimes they stand the most to lose. A poor worker really doesn't stand much to lose from displacement. That's very important to understand. Poor workers do not lose from displacement the way middle-class workers do, because they have nothing to lose. So if you're making, again, 6 (dollars) or $7 an hour, your reemployment earnings are going to be pretty close to what they were before. So displacement is not a problem. So if you're going to say we're going to cut -- I think one of the Hamilton Project's proposals was that $15 an hour is a cutoff. Well, that's -- again, that doesn't really address the problem. That may be helpful to the working poor and the poor, but it does nothing to address the problem for the middle class where most --
MALLABY: The Hamilton Project would come back and defend itself in a second --
LALONDE: They should. They should, yes. They should. They should. So I think that you have to sort of allow the income to be much, much higher than that, which will increase the cost of the proposal quite a bit. But again, you have to recognize who's really at risk here, and the poor are not at risk.
MALLABY: Let me just ask one last thing before turning it over to everybody else, which is so you want to insure people for longer than other proposals. You want to raise the salary cap on eligibility, so it's going to cost a lot.
LALONDE: It will.
MALLABY: How much will it cost? How will you pay for it?
LALONDE: Okay, so my rough ballpark figure suggests for every two years you increase the length of time you're willing to provide this benefit to people, it's going to increase the cost by 3 (billion dollars) or $4 billion per -- every two years. So you're talking -- my proposal probably is going to be in the $15 billion range to be realistic. Now, all I'm saying is -- I'm not saying that this is what Congress should do. But I'm saying if Congress wants to do something to meaningfully address the insecurity of middle-class workers, this is the kind of policy they would have to do.
MALLABY: Fifteen billion a year.
LALONDE: Fifteen billion a year, somewhere in that neighborhood. And that would include tax credits for health care as well, which is also a big part of the problem.
MALLABY: Well, Congress is currently considering what to do about extending trade adjustment assistance. The cost of simply rolling it over is more like 1 billion (dollars) a year. And you're saying 15 (billion dollars). I think that shows the difference between what's regarded as practical in this town and what, in your proposal, is still modest versus the losses that these people would take. So then you're saying four years of coverage, right?
MALLABY: So you've got somebody who's going to lose wages for 20 years, and they're going to be covered for four years of that. And even that costs 15 times what Congress is currently contemplating.
LALONDE: Well, no. My -- if you're going to talk about something that goes sort of would cover the full lives if you're talking $15 billion. In four years, it's probably going to cost 8 (billion dollars). Because again, the reason why is because fewer and fewer people are going to need the benefit as time goes on.
MALLABY: I see.
MALLABY: Okay. Well, let's see if people got questions. I see one over there.
Can you please identify yourself first?
QUESTIONER: (Off mike.)
MALLABY: Maybe you could use the microphone.
QUESTIONER: Thank you. Two points really. First, I wonder if you've looked at the data set that would exist of what seems, at least to me, to be a subset of wage insurance, which is disability insurance, especially provided by the private sector. Because you know, this is really not something that most people take advantage of. We have, you know, some form of government-provided disability insurance through the Social Security system. That does allow people to go back to work to some extent if they need to. In the private sector, people really don't get it unless it's employer provided. Even high-wage professionals who would seem to have lots of costs and no trust fund are really not always inclined to get this, even though the cost is very low. And you know, if they were disabled and couldn't work, their lifetime losses would reach into probably seven or eight figures very, very easily. So I just would ask about that. Because I'm interested in your point that there has been no market response. I don't know that there is a real market. That's the first question.
The second question is -- or maybe it's going to be more of a comment. I'm very interested in this solution, even though I'm not 100 percent convinced that it's a social good that people be provided a guaranteed income based upon their job at age, you know, 42, which is the age I am. I wonder if you could just respond to the question of the structural wage problem between the United States and not middle-income countries but other high-wage countries, which is really a product of the fact that employers have to provide health insurance or feel they do in a lot of these firms where people are losing their jobs when they're middle aged. You know, in other economies similar to ours, these are not costs that employers face. These are government-provided services. And so it's always going to be easier -- it's easier to manufacture anything in Ireland, where they have highly skilled workers, than it is here. It just costs less. People didn't pay for their education to go to Trinity University. They're not paying for their health care. So I mean, to me, this is the problem. The people coming behind those displaced persons, people who are half a generation younger than myself, are never going to realize those same incomes over time. They're not going to get those same jobs 10 or 15 years from now. So that's my opening comment.
MALLABY: So there's a question of generational equity there.
QUESTIONER: Yes, generational equity.
MALLABY: And there's a question of whether people would actually take up their insurance if it was offered to them. Do you want to respond to that?
LALONDE: Yeah. I think there's several good questions there. You know, you could sort of make the same argument about life insurance. If you sort of look at how much life insurance there is today, people do underinsure their lives or behave in a way that suggests they do -- maybe they don't care about their spouses and children. That may be one way of thinking about it. But it seems like they underinsure their lives. Disability insurance is the same way, although I do think it's -- I buy it myself on the private market. I think it's expensive. I can see why it may deter some people who don't think it's ever going to happen to them, I guess.
But the other thing is that with disability insurance, I think it's a little bit easier to address the adverse selection and moral hazard problems there because you can sort of say here's a set of disabilities we can cover. We can verify in fact if you have these disabilities. Now obviously, there's problems with it. We know that from the Social Security disability system that there's problems. We know, for example, people go on disability and get disability during bad economic times more than they do good economic times. So in some sense, it's used a little bit as a way of addressing economic downturns or uncertainty. That's just the way it works. So there's uncertainty there.
But you can imagine it's even greater in this case verifying what exactly a person's lost when they lose a job. And the problem is -- I think the real way to think about the problem is that people are very different. So every steel worker who loses his job is not the same. Each has different capacities, given to other kinds of jobs, and they vary quite a bit. And the problem is that as an adjuster -- imagine you're an insurance adjuster, and you're trying to figure out which steel worker has more potential to make it in another career than another. It's very difficult to do until you actually observe them over a period of time. So the nice thing about wage insurance is that it sort of allows that adjustment to change over time as the person evolves and acquires new skills, becomes more successful. But I think --
MALLABY: But your question was about whether they would even come and ask for it. Is that right?
QUESTIONER: (Off mike.)
LALONDE: Yeah. Well, I think that's -- obviously, it's going to depend on the cost. But I think the real question I think is not really -- I think there's an efficiency issue here from the economy's point of view. And it's not really the traditional one that economists talk about with a market failure, and if you address the market failure people will behave more productively. Here, I think, the arguments are a little different is that one is what I think -- think of the political economy argument -- is that people legitimately fear losing their jobs, and they're willing to do a lot to prevent those jobs from being lost. And so if we think about one of the impediments to more open trade or more open trade is that that's certainly an issue that is a legitimate issue which many people in organized labor are raising. They're right. Many of the people who they represent are going to be made worse off by trade for sure, and we don't really have a mechanism for dealing with that.
So the way in which -- if we're going to improve the world trading system, there are tremendous gains to still be had by improving the world trading system. That's one way we can think about trying to push that agenda further. So that's what I call a political economy argument.
The other argument is that whether or not we have -- it's more of a normative argument; that is, we have policies that make people worse off. Environmental policies potentially have the same thing. So, if you have a policy that's going to restrict the kind of coal you're going to burn or you're going to affect whether you use ethanol or not, which is a big deal in the Midwest, that's going to affect adversely some people and not adversely other people. Or if you're going to protect the spotted owl. The spotted owl is my old one; we used to have a spotted owl argument in Washington State; it used to be a big deal. And that was a question of whether you protect the creature against the jobs of people who would normally cut down their trees. So that's another example of government policy having adverse effect on people. And the question is, when government policy adversely affects people, should government be in a position to try to address the damage that they do to people? That's more of a normative argument. And whether you think it's right or wrong depends on your point of view.
Whereas I think the political economy argument I gave earlier about the fact that gee, we've got this problem that no one seems to want to expand trade anymore. And that the problem here is that people legitimately think that they're going to be hurt by it, that there would be gains by trying to make a grand deal, in some sense, by trying to expand trade by more by providing some reasonable assurance that people would be protected from these losses in the event that they occur.
MALLABY: And there was one question over here.
QUESTIONER: Thank you. Fred Tipson with Microsoft.
I'm surprised that your insecurity scenario is so linked to just the wage level. I mean, as most of us know, the insecurity that most of us would feel in losing our jobs is losing health insurance and losing pensions or retirement benefits. I mean, if I were losing my job, I'd be willing to take a $15,000 cut if I could have good health insurance and some retirement plan. By avoiding those, you seem to be avoiding the public policy environment in which wage insurance would be offered, which has to do with, you know, reasonable, universal health care, reasonably cost health care and some way of porting your pension or porting your retirement benefits and -- not to mention the inadequacy of the work force training system that we have.
So an expensive wage insurance program seems to avoid dealing with what are the more fundamental issues around job insecurity and job -- and changing, within the course of your career, from job to job.
MALLABY: So you're saying that wage insurance is all very well, but health insurance is more important -- portability there.
LALONDE: Well, I think it's still the case that if you think about a person's total compensation, particularly if you're looking not at the very high end of the pay scale but in the middle class, a large portion of the compensation comes in the form of wages than in the value of health insurance and pensions. Having said that, most proposals, and mine, too, would recommend that some tax credit be provided to allow people to buy health insurance in the private market. That is definitely one of the sources of insecurity. And obviously, it would be better for people in this position if there were sort of a universal health insurance scheme that allowed them to have health insurance under any scenario. I don't disagree with that. I think that's a different debate which I didn't want to get into.
Now, the pension part is a good point. Now, there are different ways to address this problem. One is that you could incorporate into the system a system where your Social Security fund would be added to as part of the wage insurance scheme. Now, of course, that doesn't address the case where people have private pensions they've lost and certainly a loss. But remember, as you move further down in the income distribution, the importance of private pensions as a source of income when you retire gets less and less. Now, for this group, there will still be some, and my proposal does not address that.
And indeed, another thing my proposal does not really address, and I don't really mention this in the report, is that there's actually another loss associated with losing your job. It's not the fact that your new job doesn't pay as much as your old job. And so the way I'm tying the payment is to what you earned before you lost your job. In fact, that old job if you hadn't lost it would probably have wages that would have grown. So the numbers that you see in my report, which are based on Labor Department figures -- the statistics actually understate what the losses are. And I have a paper that I wrote with Dan Sullivan years ago that tried to address how big that is. It's bigger. So in some sense, I'm still understating what the losses are. And you're absolutely right, pensions should be included, health care should be included, the whole package. But keep in mind that for most groups of workers, even in the middle class, the largest share of their compensation comes in the form of salary and wages.
MALLABY: One of the interesting things about bring health into the debate, though, is that you quickly notice that what the public policy is prepared to spend on health in terms of tax subsidies and so forth is very big, which makes, you know -- if you compare Bob's proposal to what's currently spent on trade adjustment assistance, it looks radically large. But still, compared to what we're spending on health subsidies, it's very small.
Let's go here and then to Paul and then Karen. Do you want to go first -- yeah?
QUESTIONER: Oh, okay. I think -- Mac Destler from the University of Maryland, and also I have an association with the place across the street, the Peterson Institute for International Economics.
I find basically your analysis to be very attractive. And while I recognize that $15 billion or even $8 billion is not all that politically workable in the short run, I do think those are at least the range of numbers we should be thinking of. In fact, in my trade politics book, I argued for 20 (billion dollars) as a target for all programs. I guess my question is, should we really be thinking, though, of wage insurance as THE one way of addressing this issue? Or should we really have more of a multiple track strategy and recognizing the limitations of government-sponsored re-training, trying to make that better, at least for a set of people?
And the other question I wanted to raise is a colleague of mine Catherine Mann who's talked about giving tax credits to firms for the re-training of people who they employ for the first time, presumably displaced workers, and the argument being that the firms will know what they want to train them for and that the firm-based training empirically has a much better track record than public training programs. I wonder what you think of proposals like that. And I realize there's some problems in terms of how much of the actual credit will in fact lead to additional training as opposed to simply fund training that it would have done anyway, et cetera. But I wonder if you could comment on that type of proposal.
LALONDE: Okay. There's a couple of things, and one thing you reminded me of was a question of Sebastian's I didn't answer and that was, how do you pay for all this? And the way I've calculated it, and I think it's standard -- I think, even the Brookings calculations, too -- is it costs about 2 (dollars) to $3 per month premium on everyone's pay -- you call it a tax, you call it an insurance premium, whatever you like -- to finance about two years of wage insurance, all right.
LALONDE: Everyone's, everyone's. You me, everybody, even though, in some sense, we're not going to face -- we don't face -- I'm a tenured professor at the University of Chicago. I don't face this risk at all, all right. But I would have to pay it, too, all right. But there's other ways to pay it, too. I mean, I would argue that TAA should be scrapped. That the idea of targeting this funding for people who are only losing their job because of trade, I think, has done a lot of damage. One, it creates the impression that job loss is due to trade. And I think that's been really damaging in our debate.
QUESTIONER: Would you abolish training or just abolish the separate TAA --
LALONDE: You know, I'm even uncomfortable with training for this group, sort of government-targeted training for this group, because I don't understand why you should earmark the money just for training. So actually, there's more diversity in our proposals than it appears, because what I'm saying is you can give people this premium, this money, and if they want to spend some of it on training themselves they can; they go down to the local community college and take some courses and get trained. So there actually is training built in to the proposal. It's just that the worker has this right to do it rather than sort of using money that's earmarked especially for training. I don't understand the logic for earmarking the money for training for this group for whom we have no evidence that it really works, all right. So maybe someday we'll find, through a few demonstration programs, we'll identify some reasons why government should earmark money, especially for training. But I think what I've said is saying instead of taking the money and saying you can only use this money for training, you take the money. If you want to use some for training yourself you can. We have community colleges that are heavily subsidized already and provide very good courses and, in many cases, provide good training possibilities.
QUESTIONER: Okay. What about subsidies for company-based training?
LALONDE: Here's the problem with that, and I think we've had some experience with this, is how do you prevent it from being rather being a training a subsidy to being an employment or wage subsidy? So what happens with the poor when we've done this is that effectively what it becomes is a way of saying instead of paying the person a minimum wage, what you're really paying them is one-half the minimum wage, because the government makes up the rest. It really is just an employment subsidy, not a training subsidy. And so you have to think about how you're going to actually make sure training gets done other than all you're providing is some kind of on-the-job experience that may provide some informal set of skills. But that's proved to be a problem with on-the-job training programs.
We've experimented a lot with the economics advantage is that it really becomes more of an employment subsidy where I'm going to subsidize the person's wage for a year -- let's say 40 percent or 50 percent -- and during that period the sort of implicit agreement is you're going to then give them some training. But it's very difficult actually to monitor the firm to make sure they do that. And the incentive for the firm, of course, is to take advantage of the fact that I don't have to pay this person $6 an hour. I can basically pay them $3 an hour, because the government's giving me the other 3 (dollars). And that's the problem with those proposals. And they become bureaucratically expensive as well to try to enforce. So I don't really see that as necessarily being a way out of the problem.
QUESTIONER: Thank you. I was interested in what insight you can give us about the types of beneficiaries, the types of industries where these would-be beneficiaries come. I'm conscious of a comment that Hillary Clinton made the other day in response to a question I asked her about information technology and computer science and the churn. And she gave a very broad, sweeping, very interesting, comprehensive answer. She also expressed her concern about the engineer who I think probably meets your age group that you can imagine has done all the education, become an engineer but who is now busy training his replacement who will be outsourced. So I'm wondering how many of these 250,000 are in particular kinds of industries which might help inform how we better go about dealing with this churn.
LALONDE: Well, historically -- and the only reason why I'm against something like TAA is because historically there's been a lot of diversity in what are these industries people come from.
LALONDE: It's not something limited to durable-goods manufacturing workers who are losing their jobs because of trade. More workers in fact don't fit that stereotype than do. And so I think that from the workers' perspective it really doesn't make any difference whether you lose you job because of trade or because a hurricane came and wiped out your city or whatever the reason was. You face all sorts of uncertainty, and once your job is destroyed it's destroyed. And so that's the kind of -- and again, the thing I think is bad about TAA is it sort of makes its link between job loss and trade. And people think this is the reason why people are losing their jobs because of trade, the way immigration has sort of been muddied for the same reason.
LALONDE: And so that it's better to make people understand the reason why people lose their job is because of technological change. And one of the advisers when I was writing this proposal is from the Communication Workers of America. And he pointed out for their workers, trade has nothing to do with their job loss. It's all technological change. And they're in the situation where it's been bad for the last 20 years. It's even going to get worse in the next 20 years where how do you manage a union in an environment where it's going to face sweeping technological changes. And there's many industries that are like that. But I think that what you'll see over time is just tremendous diversity in the industries people come from, and it changes.
QUESTIONER: Is it shifting? Can you say that it's more services than manufacturing --
QUESTIONER: -- and therefore, we ought to be thinking in a different mindset?
LALONDE: Certainly, if you compare the early '80s. I mean -- well, '82 when the surveys first started being done to now, there's been a shift. But even within that period, there's changes over time.
QUESTIONER: Oh, sure.
LALONDE: And I think that the point is it's a very dynamic economy with lots of changes in different parts of the economy. And it's going to be difficult to think of a program that's going to be able to target one sector. Because by the time you get that program in place, it'll change quickly enough, and you'll have to target another sector. So I think it's better to have a program which sort of treats all workers the same in terms of what sectors they're coming from, realizing that this is the way our economy works. It's a very dynamic economy. We have a tremendous amount of job creation and job destruction each year -- a tremendous amount of job creation. Most jobs -- I think it's, what, 30 percent of all jobs in any given year are either brand new jobs or been destroyed.
MALLABY: So we've got just under 14 minutes and four people on the list here.
Karen Johnson is next.
QUESTIONER: Thank you. Karen Johnson of Federal Reserve.
Let me start by saying that, you know, I think the basic, if you will, the political economy side of what you're talking about makes a lot of sense. And economics is filled with conclusions that say that we could all be made better off and the winners could accommodate the losers. But of course, we never do.
LALONDE: We don't. Well, we try, but we don't.
QUESTIONER: And this proposal could be thought of as an answer to that, that you're trying to find a way where the -- (inaudible) -- a society which gains compensates the losers. And therefore, everybody wants to move forward. But oftentimes, when we think about that, it's really some kind of adjustment cost we're talking about. And that if we tell the story of, say, a technologically changed world or something or even the hurricane world, it's that some people bear the cost of relocating or the search costs or what have you, and nobody compensates them for that. And that seems to be something you're taking exception to, because you want something that's very long lasting, and you want something that doesn't have subset clauses and doesn't have a lot of tests about we have to do X or Y or Z in order to get the money and so forth.
So I still can't get my head around the idea. Why is it, what's going on in the process, that this person once had the capacity to earn $40,000 -- using the number you used -- but now can only earn 30 (thousand dollars)? How do I understand that? Isn't it important that I understand why that's happened before I figure out what the policy ought to be to remedy it? One possibility that would have some efficiency legitimacy to it would be some kind of specific human capital. But you don't seem to be arguing that, because you talk about training not being the answer to the problem. So giving them so new, specific human capital isn't going to solve the problem.
So is it economic rent? He just sort of lucked out. He got in a union, they raised the wages. Well, I don't want to subsidize economic rent. So don't you have to tell us why this person could once earn $40,000 and now could only earn 30 (thousand dollars) in a big, flexible, well-working labor market?
LALONDE: Okay, so let me go after the question about the union premium, because this is something -- people have asked me a lot about this over the years saying we don't want to subsidize that. And I would say, in some sense, from a practical standpoint -- again, the argument I'm making is not really an economic argument as opposed to an economy one -- and that is, that's the person that sometimes you're going to have to pay off to get the policy change. So I don't care what the reason is that they had this premium. And I would argue you probably don't want to give people who have had a windfall. That is, if you notice in the proposal, I sort of say you probably want to focus on people who've accumulated a fair amount of tenure on the job before you would provide this wage insurance. So if you've had a job for a year and you lose it, you would not want to treat them the same way as someone who's had a job for, say, five years and lost it for whom the gains and -- generally, the loss are much bigger anyway.
But the problem is that you may not feel good about providing this insurance to someone who's a member of a union and therefore earns a wage premium. But again, in some sense, they're part of the problem that you're facing. So whatever the reason is for the higher income you got, you're going to have to address the fact that they didn't have incentive to insure it. Now, what's the reason for this? Well, a lot of the reason is that people's skills do become obsolete; that there's a sufficient amount of firm-specific skills, as you call it, which you can't re-train. So general training's not going to solve that problem. General training's going to provide general skills which are easily transferable from place to place. And so if it really were all skills are mostly general skills, eventually this problem should, in some sense, should solve itself. It should all be an adjustment cost problem. And it's not an adjustment cost problem.
You know, Dan Sullivan and I have followed people now for quite a few years in Pennsylvania who lost their job in the early '80s, and they're still behind. And Dan in fact has a fascinating paper he's just written looking at the -- he went and got the death records of people who were losing their jobs and keeping their jobs in the early '80s in Pennsylvania and found that people who are in this position are dying sooner, apparently for a variety of reasons. So there's long-term effects here. So this is not just simply an adjustment issue. It's the fact that for some people this has been a permanent effect on their lives.
If it were an adjustment issue, then this is going to be cheap. It's really a case of people go from New Orleans to Houston. They eventually settle down. After a couple of years, they're back to where they were before. There's no wage insurance anymore. And the nice thing about these wage insurance proposals -- not just mine but all of them -- is that they're flexible. So once you sort of get yourself back to where you were, there's no more subsidy, no more wage insurance that you're going to receive.
MALLABY: I want to go to Thea Lee because she's from the AFL-CIO, and it's a good time to bring her in.
QUESTIONER: Speaking of union premiums -- Thea Lee with the AFL-CIO -- (laughter) -- I guess I want to start with the political economy problem that you posed. And for us, the problem with wages insurance -- there's a lot of problems with wage insurance the way it's been proposed -- and the way you propose it, in particular, is extremely problematic, because you've proposed funding it by basically scrapping TAA and diverting funds from unemployment insurance. And I think actually the question that was just raised is an important one. What is the reason why there's this huge downward shift in income, growing inequality, eroding wages for the majority of American workers? And I think that wage insurance is a completely inadequate solution to the problem of eroding wages. That you need to go back to the source, whether it is the way we've structured our trade policies, the way we've structured our industrial policies or the tax policies, that you have to go to the source of the inequality because you can't afford to fix the kind of growing wage inequality we're seeing in this economy with any kind of program such as this.
The one that you've proposed, you talk about 10 years. We all know we're not going to be able to fund 10 years of wage insurance. The programs that are on the table -- usually two years, as you mentioned. And so you have a two-year solution for a permanent problem. And what happens to the workers at the end of two years? We have a very, very little bit of data on wage insurance at this point, because the Labor Department's done a bad job with the pilot program that's in place. But most of the evidence says that after two years, workers basically take a pay cut. They're in jobs like in donut shops and in Targets. And so they're not gaining skills on the job because we know that wage insurance is a subsidy only to low-wage employers. It's targeted at low-wage employers, which seems to me perverse that you're actually rewarding low-wage employers for paying low wages.
And you talked about the problem with --
LALONDE: No such plan doesn't advocate that at all, all right. So I'm critical of current plans that exactly do that, because I don't think there should be a special cap that does that. I think it's wrong, you know.
QUESTIONER: But then it gets more and more expensive if you take the cap away.
QUESTIONER: But you talked about the problem with the employment subsidy. I think wage insurance is particularly an employment subsidy but to exactly the employers for exactly the wrong reasons. And it doesn't develop the skills that workers need. Workers do need new skills. And the training programs -- you know, you're right, $3,000 probably isn't enough. Then I think one of the other issues we've had around globalization and trade and so on is if there really are hundreds of billions of dollars of gains from trade liberalization out there waiting to be held, then why are we talking about one just $15 billion worth of a program to remediate those? We ought to be willing to spend hundreds of billions of dollars by taxing the winners to pay off the losers or by changing the policies at an earlier stage.
And I think -- you know, what we would say in terms of on-the-job training is that there are some effective programs out there. Most of them are labor-management programs where labor and management work together, where the employer can identify the skills that they're going to need, and the unions can be part of ensuring that workers are getting the benefits and that they're not being taken advantage of and so on. Those programs, apprenticeship programs have been tremendously useful. And certainly in areas like energy, you know, development of renewable energy sources, infrastructure, there's all sorts of areas where there could be tremendous job creation. We do need some skills in the economy. Your proposal, I think, does exactly the wrong thing which is it pays people to take low-wage jobs they wouldn't otherwise take.
And I think the one other piece I wanted to say in terms of the evidence that we have, which is very sketchy, on how wage insurance has worked is that it has in the past -- this the Upjohn study in Canada -- that most of the benefits came from workers with a certain level of skills displacing workers with lower skills. So that basically, for the employer, you get a more-skilled worker than you're paying for, right; let's say more experience and more skills. And who loses out? It's the people at the bottom. So you have an equity problem in addition.
MALLABY: Go for it. (Laughter.)
LALONDE: Sure, yeah. No, I don't disagree with you that the amount of money that's being spent on this issue is trivial. And I certainly agree that $20 billion seems to me like a reasonable amount of money to start addressing the issue of worker displacement. But again, I think the question we're talking about here is much narrower than the one that you're defining -- why is there all this growing wage inequality -- which really has to do with -- the primary reason for it is quite simple, and that is that it has to do with the fact that returns-to-skill have gotten a lot bigger. And for whatever reason, supply of people willing to -- acquiring those skills has not kept up. That's the problem.
I mean, we talked about this earlier about the fact that, you know, why aren't boys going to college anymore? Why is there this big gap in the way in which boys and girls go to college? A lot of that has to do with the fact -- one of the puzzling things for many of us who have studied labor markets is why haven't people picked up on the fact that returns to college are really high, and yet the enrollments in college have not kept up. And I think that's an enormous problem. But I think it's the education system is going to -- work force training is not going to address that problem. It has to be addressed through the education system. There's no way we're going to -- I mean, I think Jim Heckman years ago did a calculation, and he asked, how much money would you have to spend to return the earnings distribution which has become much more unequal to what it was in 1979? And it would cost an investment of something like $1.5 trillion. That's a huge investment. Now, the only -- we spend money like that. I think it's something like three or four years of the total investment we make in primary and secondary education in this country. So the thing is it's a different problem addressing inequality. And most of that change is not occurring because of job loss, it's occurring because of slower-growing wages on the job.
So people are experiencing inequality, not experiencing because of the fact they're losing their job and they go from a good job to a bad job. And again, with the poor, for example, this not a problem with the poor at all. They lose jobs a lot, and they go from one bad job to another, all right. So they don't have any wage loss associated with getting a new job. The middle class is a somewhat different situation where they have two ways in which inequality has been growing for them. One is that their wage increases don't keep up, all right, which the primary way real wages have been falling, but the other would be occasionally they lose their job and it's an enormous hit so they can't get a job that pays anywhere near the old job.
But I think the point is that these job losses reflect some kind of -- whether it's the union premium or some kind of specific human capital which, again, you can't fix the retraining. Retraining can provide people with general skills but it can't provide people with specific human capital. In other words, specific human capital is linked to the employer. And I know Dan and I did a lot of work on this finding, that even if you get a job back in your old industry, you're still going to experience pretty large, permanent losses. So it's not just getting people back into their own industry that will help. This is a problem that reflects not just a short-term adjustment but a long-term loss and marketability, if you like, in the labor market.
So my argument would be that what you're talking about is a different problem. This is not the same problem. And I think you're absolutely right. If we want to address it, it's going to require an enormous investment in education because that's the only mechanism we have to do it. Work force training is not going to address that problem. We're not even close to having the kind of systems in place to address that kind of issue.
MALLABY: Well, with apologies to two or three people who on my list, we've run out of time, but we have. So I think my first duty is to finish punctually.
Thanks a lot, Bob, for coming here. Thanks for explaining it. (Applause.) Maybe those who didn't get to ask questions can come and see Bob. He'll be here for a few minutes.