The group introduces itself: Heshan, Magda Ivanska (Poland), Gabor Juhasz (Budapest), Lydia Bialek (Poland), Dino Hochyl (Slovakia), Laszlo Ollos, (Slovakia), Ina Breuer, Belinda Cooper.

David Howell: Well, this is an honor to do this here. If there's any problem with my English, let me know.

I wasn't sure how long I had and how much depth to go into; I have some handouts we could work off of, let's start with that. My job here is to talk about labor policy, and what I'm going to talk about more specifically is policy surrounding the earnings crisis in the United States. And I'd like to start with just describing the problem and what's happened in the last decade or decade and a half in the United States; compare it to what's going on in other advanced countries, and then talk a little about alternative explanations for these developments, and then give you my very unconventional story about what's going on.

If you look at Table 1, the picture is really summed up in this little table. You can see by looking at the middle column, 79 to 89, what happened to real earnings for full-time workers during the decade of the eighties. Whereas in the seventies it fell by about 5% for those with less than a high-school education, earnings fell by almost 20% in the eighties. So for the overall period you have almost a 25% decline. It's an absolutely unprecedented decline in living standards for a very large proportion of the American work force. Two-thirds of even young workers have a high-school degree or less; only about one third have a college degree, so these declines are affecting a majority of the work force. You can see that in row 2, for those with 12 years of schooling, the collapse was significant, but not as extreme as for those with less than a high school degree.

The trends for women are similar, except that they're much better. The lowest-skilled women did the worst, and did much worse in the eighties than in the seventies--actually did quite well in the seventies. But you can see that even those with a high school degree had an increase in real earnings, whereas for men there was a huge decline of 11% for those with a high school degree.

So you have two things going on here, at least. One is that it is getting worse over time, and unfortunately the data from 1989-94, 95 continues these trends; secondly, it's worse for the least skilled; and thirdly, if you look by gender, you see very different results for men and women. And that's kind of the story on earnings by educational attainment and gender dimension. The result of this has been an increase in inequality. If you turn to the next page, you'll see one measure of earnings inequality, and that's known as the ninety-ten ratio; you're just looking at the percentile of earners in the ninetieth percentile, and you're comparing that to the tenth. And you can see what happened between 1980 and the mid- to late-80's, say 86-87--huge increase in inequality, really unprecedented in recent US history, and then for men it continues up more gradually, but it has continued. You get the same, even sharper increase for women, but as we saw from the earlier table, the results are very different for women, and this is more like the experience that other advanced countries face; that is, women in the United States were more like all workers in other developed countries in Europe. Why? Because, while inequality grew, it grew because of movement at the top; it wasn't because workers at the bottom collapsed in real terms, which was the case with low-skilled men in the United States. So for women in the United States, and for other advanced countries--Sweden, Germany, France, Canada--any evidence of growing inequality was due to a growth at the top. No other developed country saw a decline in real earnings among low-skilled workers, only the United States, which suggests that our labor market works in a substantially different way than those of other advanced countries, at least for male workers. And I want to get into why that might be in a few minutes. But on earnings and inequality, it's worth noting that only Great Britain had increases of this magnitude. There were some countries that saw slight increases in inequality, nothing as massive as this. So what the United States was able to achieve, at least for male workers, was what no other country was able to achieve during this decade: both collapsing real earnings, and sharply rising earnings inequality.

The conventional explanation for that is that it was due to a shift in demand away from low-skilled workers toward high-skilled workers. The explanation for that shift is usually put in terms of two possible causes. One is trade--that globalization was a major source of increasing competition, and firms outsourced, moved facilities abroad, and that all worked against low-skilled workers. But most economists, especially conventional economists, don't believe that trade pressure could possibly explain the magnitude of this collapse. A contributing factor, but not a primary explanatory factor. What most people turn to is computerization or technological change as being the source of declining demand for low-skilled workers. Well, we're going to talk a lot about whether this makes sense as an argument; let me just emphasize how widespread this conventional view is. From a recent OECD job study, let me give you one quote. The labor market problems on both sides of the Atlantic, according to this jobs study--this was 1994, major effort by OECD, "stemmed from the same root cause: the failure to adapt satisfactorily to change. Management skills, education and training attainments have failed to keep pace with the requirements of a more technologically advanced economy. All countries have experienced the shift of demand away from unskilled jobs toward highly-skilled jobs." Okay. That's quite a blanket statement. Well, conventional economists have taken this position in the United States, leading labor economists, and many of them have worked for the Labor Secretary, Robert Reich, who has done a very good job of hiring the best and the brightest from American universities, and they're very good economists. Well, Reich wrote, in an op-ed piece in the New York Times a year or so ago, that "technological changes have diminished the role of labor, especially unskilled labor, in the modern factory. The most striking change has been the brash arrival of the computer."

As you can see from the next chart, the evidence does suggest that there might be some link. What I've shown here is computer investment measured in constant dollars per full-time equivalent worker. This is mostly computer-related equipment; it also includes office and accounting machinery, but by the early 80s that was a very small part of this aggregate, and the increase is due almost exclusively to the growing use of computers.

Well, when did it begin to increase? If you look in the middle of the graph, it appears that it starts taking off in 1982; for services, in 83 and manufacturing. Well, this isn't a coincidence, since the PC [personal computer] was only introduced in 81-82, and a huge share of this investment is in personal computers and related investment.

So what we have is a dramatic decline in real earnings, growth in inequality, and the timing seems to fit with this growth in computers. Well, let's then start with assessing some of the evidence for this more closely. One drawback of this explanation is that nobody has really been able to measure technological change in a way that's very effective. And so if you can't explain it by anything else, it's kind of the omitted or left-over category, if you're doing a regression analysis in statistics: if you can't measure it, it may be in the residual, so it's sort of a residual explanation, but it fits.

If you look at figure 2, it has a traditional demand and supply diagram. If you take an introductory economics course, in the first chapter this is what you'll see, and how economists think is through this graph. It's very powerful in certain respects, but I'm going to argue that it can be very limiting. What you have is a demand curve sloping down, so if you reduce the wage there'll be a greater demand for workers. On the other hand, the supply curve is an offer curve; workers are willing to supply more labor for a higher wage.

There's been no evidence that supply-side changes can explain this wage collapse. There's been an increase in immigration, and that affects certain cities, certain populations in certain cities. More and more people have at least a high school degree. But the studies that have been done on supply-side effects, there's no evidence that it can explain the wage collapse. So the conventional view is that it's a demand-side problem, something has happened on the demand side, meaning employers have changed what they need from the workforce. And the story is a shift in the curve; it's a shift from D1 to D2. If you read from the wage side of the graph, at any given wage, employers will demand less employment; instead of demanding E1 employment, it would be where the D2 line crosses the W1 line. If you read up, they are only willing to hire E1 employment at wages well below W2--where that line from E1 intersects the D2 line. So something's happened on the demand side. And if you buy the argument that there's been a downward shift in demand, we should see two things when we look at the data. One is that wages have fallen from W1 to W2, and you can see that on the graph. We've gone through that data, and yes, wages have fallen. But you should also find a decline in low-skill employment, from E1 to E2. Well, that's a fundamental question: do we see a decline in low-skill employment? That's the first question, and I'm going to spend a fair amount of time on that. But now that I've introduced it, let me go back to some other conundrums, and we'll come back to this issue of employment.

If the problem is a shift in demand, and we think that's due to globalization, on the one hand, or even more fundamentally perhaps, new technology, computerization, then you have to ask the question, why did US workers experience a collapse that was not experienced by Canada, right across the border; in Great Britain, which is very similar, very decentralized economy, labor markets, just like the US in many respects; not in Germany, not in France, not in Sweden; certainly not in Japan; but those countries are all introducing new technologies, and those are all players in the world market. What was unique about the U.S.? You can see its uniqueness in the data, in these two tables I've included, these long bar charts. This is from Richard Freeman's paper that he just wrote, he is a very prominent Harvard labor economist. And these show just how far out there the United States is.

EM: What does this 65 represent?

DH: It means that workers who are in this 90th percentile earn 5.65 times what the 10th percentile worker makes. In Canada, which is right across the border, it's 4 times; it's only 3.4 times in France, and you go right down the list. Interestingly, Switzerland is very low, and yet it's a very decentralized capitalist economy.

Question: From what year is the data?

DH: I'm not sure; he didn't label it. It's within the last few years; I'd put early 90s, to be safe. That's a good question . . . Well, the 90-10 percentile could be an increase at the top, or a decrease at the bottom, or both. The next graph gives you a little more information. While we know that the top increased--corporate executives used to make 40 times what a worker on the floor makes, and now it's 220 times or some extraordinary number like that--the top 1% has skyrocketed in the United States, they did very well in the eighties and early nineties. But what this second chart shows is that, relative to the median, the fiftieth percentile; if you ignore this winner-take-all syndrome that's taken place in the United States, where the people at the top--where Steffi Graf, to make the analogy to tennis, makes all the winnings, and none of the other women make any of the winnings--everything goes to the winners, and even though you're only slightly less good, there's nothing left in the pot for those only slightly less good. But there's more to the story than that, because you can see on this graph, the US is way below; the lowest paid workers are paid much less than the median worker, compared to other countries. Again, Canada is closest to us, which is another interesting story that I can't get into here.

So that's a second conundrum. In our explanation of what's driving this enormously important collapse--I mean, it's very likely that this could drive the U.S. presidential election, and therefore affect -- to try and get it into your area of interest -- affect world relations and U.S. foreign policy in the future. Buchanan's rise as a serious contender was fueled in large part because he has tapped in to the frustrations of a majority of the American work force--white men, in particular, who are non-supervisory workers.

So we need to have an explanation that accounts for why the U.S. is different, and we need some evidence that technology, if that's the main explanation, is a good explanation.

If it's the case that it's a decline in demand for low-skilled workers that's driving this whole issue, and we say that Europe is a separate issue, let's just focus on the United States and we won't worry about why these things didn't occur in Europe. Well, we should be able to see a shift in the skill mix of employment. It's very straightforward. If you have the graph that I just showed you, we have a decline in employment here, and this is low-skill workers; and we have another graph that shows high-skill workers, and we have some kind of supply curve, and we have their demand curve here, but then because of new technologies or whatever, we have an increase, well then we've got an increase in employment. So the ratio between the high skill workers and the low skill workers should be growing. Now, a paper was published in the Quarterly Journal of Economics, one of the top five journals in the country, by three very prominent people, and they addressed this issue; and they defined skill by the ratio of non-production to production workers--assuming that non-production workers are skilled, and production workers are unskilled. Now that's really a gross generalization; they acknowledged that, but argued that when all is said and done, it's still good enough for our purposes. So even though there are unskilled administrative support people who just do Xeroxing, and there are very skilled craft workers, they wash out. So we're going to go along with this assumption very temporarily that this is a good measure. We should see an increase in the ratio of high skilled to low skilled workers if this demand shift story has any robustness, has any explanatory power.

Well, the way the authors presented these data, was to say, well, if you look at 1979 the non-production share was about 28%-29%, and they're just looking at total manufacturing, which is that middle line. And then they say, well look at 88 or 89--there was a big increase. And that's true. But if you go back and actually plot it out, which I did, and then I broke it into durable and non-durable, you see two things at least. One is that all the change took place in 80, 81 and 82, okay? Three years. Those were two recessions that were almost back-to-back; 1980 and 1982 were two back-to-back recessions. It also came on the heels of a lot of inflation in the seventies and very bad trade, the value of the dollar was disastrous for manufacturing firms in the U.S. So there was inflation, there was the value of the dollar, and there was a serious recession, all combining in the 80-82 recession. Well, that led to a collapse in production worker employment. Now, what employers were able to do was to keep that very low level of production worker employment that they reached in 82, and they held it that way for the rest of the decade.

Now, there are three problems with this for the demand shift explanation. First, if it's either computerization or trade, whatever caused the demand shift, the question is why did it curve between 80 and 82, and not from 83 on? Particularly given that computers, as we saw, didn't enter the scene in any significant way until 1983. Here we're showing that all the change in employment towards advanced, more highly-skilled workers is taking place before 83.

Secondly, you can see that what's driving total manufacturing is durable manufacturing. Well, durable manufacturing is where the trade effects are most severe, that's what Americans trade most heavily, in manufacturing.

And third, there was another recession in 90-91. If employers restructure just in recession years, we should see a restructuring in the 90-91 recession, but you don't see any change in this chart. One of my next projects is to extend this to 1994, but I don't think we're going to see much change.

So that struck me as undermining the demand shift argument. There's another way to try to improve upon this, and that is to break up non-production and production worker employment, and to try to get a better measure of skills. So as you see on the next page, what I and a dissertation student of mine did was to distinguish white-collar from blue-collar workers, and within each category, distinguish high skill and low skill workers. So we've got four categories now, and high and low skill are just defined by very conventional occupations; high skill white collar would be professional, technical and managerial, low skill would be administrative support. High skill blue collar are craft workers, low skill are operators and laborers. We did this for manufacturing and services as aggregates, and then we chose five industry groups that were fairly computer-intensive. This table shows the essential stability of the skill mix. Let's look at the low skill blue collar share; in manufacturing, those are mostly male, and those are the workers who got really hit. Well, if we look after 81, you'll see that there's a decline from 45 to 42 from 78 to 80; that's what you saw on the chart we just went through. But after 82 it's very stable--at 42, then it drops to 40, then it goes down to 39, then it's back up to 40; certainly no massive changes in the 80s.

If there's any trend the timing of which seems linked to computerization, it's in column 2, which is low-skill white collar. Look at the change from 82--12.6; it steadily declines to 10.6 by 1990. Now that's something that's more sensible, in my view; that's mostly women, and computers are mostly in offices. So if you expect an effect of computers on workers, you would expect them to hit there, and that's in fact what the data show. The problem is that those workers saw an increase in real earnings. The problem of the wage collapse is with males in services and manufacturing. Intuitively we know that there aren't many computers on the workplace floor; a recent survey by Harry Holtzer, who's at Michigan, found that in the early 90s only 5% of laborers use a computer on the job or deal with computer based information systems. If that's the case in the early 90s, how do you explain the collapse in real earnings of workers with a computer-based explanation in the eighties, if only 5% in the early nineties use them? We did some graphs then, and you can see, the next one shows--let's stick to manufacturing just to be consistent--I've put computer investment there, and that's on the left hand side, the left axis, running from below a 100 dollars for full time worker invested to almost 700. Well, again you see a big increase after 83; what happened to the low skilled blue collar share in manufacturing? very stable. Here we're just plotting what was on that table. But you see an observable decline in low-skill white collar.

We see the same thing in services, that's figure 3B. Interestingly a huge growth in computers from 83 to 86-87; after 88, it seems to taper off a bit. But what we notice pretty dramatically--well, it looks dramatic because of the way I've done the axis on the right side of the graph; it's very easy to mislead readers by how you design the axis. It actually only goes from 19 to 23%,which isn't a huge change, but on the other hand changes in the skill mix at this aggregate level over decade-long periods--over say even the post war period, from 45 until now--have been very incremental, very small changes. So a decline in the low skill white collar from 21, almost 22% to 20 or under is a fairly large change, given the historical experience.

Well, that's some change, the low-skill blue collar blind shows virtually no change. Indeed, the last years, 88-90, there was an increase in the share. And when we go into the five industries, three service industries and two manufacturing industries, we see the same results. This was enough for me; I'm convinced. This is very different from your standard explanation, which would expect to see a collapse in the low skilled portion of the workforce.

If it's not a shift in demand, what is the problem? That I show on the last page, on diagram three. My argument is that technology does matter, but it doesn't matter so much in the workplace, or at least it hasn't so far, as in the context within the workplace is set. That is, technology has enabled communications, a new way to communicate, and it's facilitated transportation, and that in turn has facilitated globalization. It means that my co-author in Paris, I can talk to him instantaneously, and I can send this paper, attach it to my e-mail message and send it to him. And firms can do the same thing, and they're more footloose now than they used to be, for that reason.

So technological advances in communication and transportation have played a role, and foreign economic development has played a role. Throughout the postwar period, Western Europe and Japan have converged on the US in productivity levels and in growth and in living standards. It's increasingly a competitive threat. But if that's the case, it should have been gradual, from the 50s to the 60s to the 70s to the 80s. But it wasn't. We see a slight decline starting after the OPEC crisis in 73, but where the real collapse takes place is after 79, after the second oil crisis. So what explains the distinctiveness of the US? And my argument is that you have to look at the left-hand side of the graph. And this is what Americans, who are very individualistic and don't see things in ideological, kind of group terms, and particularly economists, have a very hard time recognizing.

US public policy, and underlying US public policy is ideology. And that ideological foundation tends to be compared to European nations--very individualistic, very laissez-faire; you know, it's the the pioneer stuff. Well, while there's no question that the US is on that end of the spectrum, it's also true that that fluctuates; there are long cycles in ideology. You could see in the 1920s a shift away from a recognition that government was important for the environment--Theodore Roosevelt's emphasis on the parks in the West; anti-trust legislation in the 1890s and just prior to World War I. There was a great deal of government activism from 1890 to 1920 in the US. Populist sentiment against banks and big industries was kind of had a heyday. Well, the 1920s reverted to a very laissez-faire approach; the Roaring Twenties, where competitive solutions to problems were accepted. We saw the same thing happen in the 1950s under Eisenhower. And I'm arguing--and I'm not the only one, Arthur Schlesinger, the very famous historian; Hirschman, a very important economist, Princeton, have argued the same thing--that starting in the late 70s you see another of these thirty-year swings, from the 20s to the 50s to the 80s. Began really under Carter; President Carter emphasized de-regulation and the limits of the government's ability to solve all problems. And that's what he was elected on. Carter was elected, though a Democrat, on an anti-government platform. Well that just began the process; Reagan took it and ran with it. It was much more extreme under Reagan. He won by a landslide, and did something that even the most anti-union private sector employers rarely did; and that was to replace strikers in a labor dispute with permanent workers. He used permanent replacements in the strike. There had never been any law against that, but there was a social norm, kind of agreed upon, and acceptance that that's just not acceptable behavior--and there aren't many of those; in the US basically anything goes in labor disputes. But that was one, and I don't understand it, but employers generally live by a rule not to permanently replace strikers.

EM: It was air traffic controllers?

DH: Thank you, yes, it was PATCO, the air traffic controllers, and I think it was 1981.

Douglas Fraser, who was president of the United Auto Workers in the late 70s, I have a quote from him in 1978 saying, what we're seeing now is an assault on workers by corporations. That was 1978, so it began before Reagan's 1981 behavior, but I think the PATCO strike was very important, because in the midst of this assault, in the midst of a very serious recession where American employers are trying to restructure to maintain their competitiveness, Reagan sends a message to private-sector employers that it's okay, not only okay but you SHOULD replace strikers, you should take an extremely hostile stance towards workers, you've got to put workers in their place. And that message went out to the private sector, and lo and behold, the demands for wage concessions skyrocketed from 82 through 85, absolutely skyrocketed; unprecedented levels of anti-union campaigns, and demands for wage concessions. And the use of replacements in strikes. Well, this led to an even larger collapse in unions, which continued a long run trend, and it scared workers; and now, by the mid- to late-eighties, you see hardly any strikes. We're still setting records each year with how few strikes there are. Just in the last year, there were a couple of unions won their strikes, but they're the exceptions.

At the same time, there were other federal policies that facilitated a restructuring of wages. Probably the most important was a collapse in the minimum wage. Under Reagan, the Congress did not approve an increase in the legal minimum. Well, from 79 through the early 80s were very high inflation years, so you saw a huge collapse in the legal minimum. And it was only in 1989 or 90, I guess it was 90, that there was an increase, but it's still way below historic levels. Well, that allowed the whole secondary labor market, all the bad jobs, to decline in real terms and still be legal. We're talking about a 25-30% decline in the real minimum wage. Now that's just about the percentage decline in the [tape turns over].

There were deregulation, there were a whole bunch of federal policies that facilitated this assault on workers. But what I have in the middle of the graph are low-road employer strategies. And what I'm trying to get at here is that on the left hand side, pro-market, anti-labor government policies combined in an environment of ideological shift towards laissez-faire, led to these low-road employer strategies, led to confrontation with workers, and this was a decade in which workers were very weak, and the result is at the bottom of the page: declining wages and growing inequality. And this is distinguished from the European and Japanese paradigm, in which globalization of trade and production read to high-road employer strategies, or at least not low-road employer strategies. And the reason was--to be symmetrical, I should have another panel on the right side of this graph showing that the ideological shifts either didn't take place or, it's well known that a lot of socialist parties lost in the eighties, but the right wing in Europe is so much to the left of the middle of the road in the United States that it didn't have the kind of impact on the social policy regarding the labor market that our government policies did. There's a social contract in Europe, I think, though this is getting way out of my expertise, that doesn't find drastic cuts in wages and dramatic growth in job insecurity the way to solve the problem. And I think we see that in the French . . .

EM: What about Britain and the Thatcher approach?

DH: Well, Thatcher's whole thing was privatization. But interestingly you don't see any decline in real wages, or in the minimum wage, the legislated minimum doesn't decline in Great Britain.

EM: There's just an increase in unemployment.

DH: There's an increase, particularly in the North, in unemployment, and that gets to an argument that OECD has made, which is that these trends affect all advanced countries, but the US has this decentralized labor market, Europe has more institutional protections, so in that respect they agree with my argument, and because of those institutional protections you don't see the wage adjustments, you see an employment adjustment. And at this OECD conference in Copenhagen that I was at a year ago November, the argument of most of the ministers and economists there from Europe was that they needed to follow the US path; they needed to decentralize, to increase flexibility--flexibility was the word, more flexibility. I don't think they had ever experienced what it's like to work at two minimum wage jobs, working 16 hours a day, as increasingly American workers are doing. You know, it's very easy for ministers and economists to talk about wage flexibility, but the kind of devastation that's affected individuals, families and communities in the United States is not something I think most Europeans would like to see. The reality is that yes, unemployment increased, but unemployment compensation in Europe is better than our low-wage jobs. So if you were making the choice, where would you rather bring up your kids? That's the basic question I have. Instead of having 6% unemployment here, you have anywhere from 8 to 12%; in Spain it's 20, that's a special case. But 12 is a crisis. It's high, there's no question about it, but it's concentrated in certain areas, among certain parts of the work force, and those people are given social support that we don't give people here. It's also been pointed out by several people in this country that our unemployment rate is much lower than it should be; in fact in real terms it's higher, it's closer to the European, because we have much higher rates of part-time employment than in Europe, and that's involuntary part-time employment, and we have a very large prison population; 25% of black males in urban areas between 20 and 35 are either in jail or on probation. In some cities, it's 40%.. You can make an argument, and Richard Freeman has--and I use him because he's very respected, you can take him more seriously!--that you should add the prison population to our unemployment rate, and that would notch it up a percentage or two, and you should add the involuntary part-time, and if you do these kinds of institutional adjustments, you get something that's closer. But it's true that we've adjusted by offering a lot of low-wage, bad jobs. The question is is that the right way to go, either from the point of workers and families and communities, in the short run, or I would argue in the long run, is that what we should be pushing for economic policy? Because if employers have access to a very low-wage pool, there's no incentive to invest in new technology, in new capital equipment. What it means is that we are adopting in this country an essentially third world development strategy. WE either outsource to Mexico or other Caribbean areas or East Asia . . .

EM: What does it mean, outsource?

DH: It means instead of producing everything in Detroit, you take all the parts and have it produced in Mexico, and then have the parts shipped back and assembled in Detroit. But you can also outsource to low-wage areas in the US. A good example of what I'm arguing is happening is when Pratt and Whitney, a very established country in Hartford that made jet engines and exports them all around the world--I think co-produces with Rolls Royce to produce engines as well--requiring very high skilled blue collar workers, left Hartford and went North to Maine and South to Georgia. What happened there is you get no unions in those two places, you have lower wages, you have more job insecurity, but the skills haven't changed; there's been no shift in demand for skills, the same workers are doing the same work. So that I think is the metaphor, the example that I would give for the difference in a political economy explanation for a wage collapse, and the conventional neo-classical explanation which emphasizes a shift in demand.

So I think I've spoken more than enough, and would be happy to clarify anything or take your questions.

Heshan: I'd just like to add that I think on top of the strength of employers, there's something else also particularly with the corporate sector, where to work for a large corporation, the public relations, if you like, replaces the idea of a union. Watching a film where they talk about the building of the Boeing 777, where they talk about Pratt and Whitney and Boeing, and at the end of this process they hire Disney to bring in all the workers to look at the new plane, and they have nice fancy lights, and the whole thing about various benefits you can get by being employed by large corporations and things like that. Everything is dependent on your employment, and you have to not say much when you don't make enough. I think that's another social trend that's going on that really emphasizes the atomization of American workers.

DH: I think that's a good point. There's a contradiction in this, especially if you believe that skill requirements are growing, but at the same time you're offering these workers lower wages and less job security. What investment do workers have to train? Or what investment do they have in behaving well on the job? It raises questions about motivating the workforce. And especially among what we call subordinate primary, or good blue collar jobs. The trade off was you give me job security, and growing real wages, and I will commit myself to the firm and won't ask any questions about how you organize production. But now the corporations have pulled back from their part of the deal, and it may be that these are strategies that they're trying to use: we'll lower your real wages and no longer give you job security, but we'll let you see movies and go to Walt Disney and stuff. We try to use public relations--or simply the threat of job loss--to keep people in line.

EM: John Eatwell had argued, there was a little seminar a week ago in which a lot of things you were saying today overlapped. He was discussing hidden unemployment. But what I was struck by was the argument that you also brought here, that the actual situation--the relatively low unemployment here, and relatively high unemployment in some of the western economies--are very closely related to the disappearance not only of job security, but of the safety net; there are no unemployment benefits that people receive in America, compared with Germany, for example, are ridiculous, so people are much more willing to accept much lower paid jobs, and therefore contribute further to the growing supply of low paid workers, not necessarily low skilled workers, instead of looking somewhere else; they accept something for the time being, and very frequently, as you pointed out, two low-paid jobs. And the whole picture of the job market in America is not true, it is really far from reflecting the real situation; there is a hidden unemployment in that willingness of people to accept wage concessions or to accept jobs that Germans or French would not accept, because they would rather go on unemployment and look for something better, or retrain themselves in the meantime. Looking at the lack of safety net and actually unemployment benefits specifically in America could explain quite a bit about the bad situation on the labor market. I asked Eatwell during that discussion whether in America a large group of people available, the large pool--you also mentioned immigrants coming and supplying more and more, at a time when they don't know the language, and don't know much, there is this incredible competition; and of course he agreed, that's one of the reasons, but it can't be all explained by that, and of course the immigration policies in America are much looser than in Germany or England, for that matter.

DH: There are those differences, but it's interesting: a recent study by David Card and some of his colleagues at Princeton compared France and Canada and the United States to explore what I call the OECD hypothesis, which is that everyone faces the same problem, but Europe's labor market is more centralized, so that the effect has been to increase unemployment there and the effect for us has been more employment but lower wages. And they find no support for that; when they look at the pattern of employment changes in France and Canada and the US, by level of schooling, and I believe by occupation, they see employment patterns changing identically; there's no significant difference in the way each country distributes its workers. Changes in the employment population ratio by education group is the same, you see the same trend in France as you do in the United States. So those results suggest that . . . the skill mismatch story suggests that we should have had a smaller decline in the low-skill labor market than Europe, because we're able to reduce wages. Whereas in Europe . . .I'm drawing the supply curve the more elastic way . . .in Europe you get a larger decline and a smaller wage collapse, because institutions keep the wages from declining, so there's more unemployment; whereas in the US it should be, the employment decline should be much smaller and the wage decline should be much larger for low skilled workers.

But you don't see, all their differing fancy statistical techniques for looking at differences in employment across these three nations, they see no differences. That is, employers are responding to new technologies and global markets in much the same ways in each; and yet in the United States, that's coinciding with collapsing real wages at the bottom. So again, it suggests an institutional explanation and not a purely supply and demand explanation.

Question [Lydia]: You mentioned that in cities 25% and sometimes 40% of black men are in jail or on probation. Should I assume that this number are unemployed?

DH: That's a good question. For those who are in prison, and that may be 10 or 15%, they are not counted; in the unemployment rate, the institutional population is excluded. For those on probation who are not in jail, they are very unlikely to be employed. Probably the worst thing you can have, other than being black, in this country, is to have a criminal record, and if you have both it's absolutely deadly.

Lydia: So my next question is, over the age of 35, because you said it's between 20 and 35, is it the same over the age of 35?

DH: No, it's much less. It's really concentrated in youth.

EM: But what happens to those people who are 40, who five years ago were 35, and were in jail or on probation and they never worked?

Lydia [?]: If there's a link between having a criminal record and chance for employment, then what about the percentage over 35?

DH: Well okay, there's a moving thing through time here. Black men have always had much higher rates of involvement with the criminal justice system in a negative way, but this skyrocketed in the mid to late 70s, and it's associated with crack cocaine. The higher level than anyone else may have to do with racism, may have to do with poverty combined with racism and high drug use, but there was a huge increase in the late 70s and early 80s. Probably some of it has to do with the decline of cities and the worsening economic status of a lot of cities. But what also happened was a shift in the drug of choice, and crack has been particularly . . . police do not arrest very much people who use cocaine, but they go after people who use crack. Whether that's a good social policy is another question. But it turns out that middle-class whites use cocaine, and minorities use crack. And so a huge percentage of the black men in prison who are teenagers or in their twenties are in for, it's drug related, and more specifically crack related. So what you're going to see after this catastrophe of the 80s affecting this black population is in the 90s, as these people age and get out of jail, there's going to be no employment for them, that's true. Private sector employers won't hire them.

Lydia [?]: So you think there will be no big difference between this percentage in this group between 20 and 35, and older?

DH: I guess that would be safe to say, yeah. It'll be close to that percentage or higher than it has been in the past. The unemployment rate of adults in the 35 to 55 age group, the core working years, will be much higher in the late 90s than they were in the late 80s.

EM: The population over 35, the rate of survival among young blacks is pretty low, so as horrible as it sounds . . .

Lydia: And on the other hand I understand that it can be observed now, or as time goes by, in the future.

DH: Yeah. Although here I think I cite statistics to show that the unemployment rate for young black men did not change between 79 and 89. So unemployment statistics and labor force participation statistics and employment rate statistics are three different ways to measure employment status; all show that between the late 70s and the late 80s, very little changed, which isn't consistent with the demand-shift explanation. Where you get the worsening employment situation for blacks, especially black men, was in the seventies, in the aftermath of the riots of the sixties, it was really the middle years of the seventies where you saw a huge collapse in employment for blacks.

Laszlo?: Did these changes have an impact on voters activity?

DH: Which election?

L: All.

DH: You're pushing my limits here. You have to get David Plotke to answer that . . .

EM: The Bush campaign was about taxes, it wasn't about employment, for example.

L: Is there growing voter passivity connected to this change, or not?

DH: There are two issues here though. I think there is a growing passivity, but that doesn't have economic roots. That has roots in TV. Probably the best piece of social science research I've read in years is in a recent American Prospect by Robert Putnam, I think, who's a very prominent Harvard sociologist, on trying to explain civic involvement. And what he finds is after those born in 1930--in fact, Charles Tilly, who's on the faculty here, is quoted in this article in response to having read the first draft of this paper, as having said, we're the last suckers. Charles Tilly was born in 1927, I believe. And they were the last generation to be really involved civically. And how do you define that? By reading newspapers, by voting, by bowling, by doing any activity that involves a group. Playing cards on a regular basis. In the US, Elks Clubs and Veterans Clubs and that kind of thing. Any kind of social club behavior. It falls off with each decade. For each decade, those born in the twenties are very different from the thirties, etc. By the time people are born in the sixties, and they're coming of age in the eighties, they're completely atomistic, not involved in anything. This is based on survey data that he has. And the collapse is staggering. And he goes through every possible explanation: woman's work, changes in the economy--he goes through five or six standard explanations, and none of them hold up. And he says, the main culprit seems to be TV; when people have a free moment, you sit down and watch--by yourself--TV. This is way off from where we started, but if you're interested in the demise of civic behavior, and the manifestation of that in voting, I think his work is very interesting.

The other dimension of that is, for those who do vote, how is the economic scene played out, how has it affected them. As I said, I think in this current election it's playing a big role in Buchanan's election. In 1992, I think Clinton won on the basis of the middle class . . .

EM: Channeling economy to reach larger group of people, and he said we have to stop with the trickle down economy, meaning the rich will make a lot and a little will go down and create new jobs . . .

DH: And if you look at the 1994 Congressional elections, where the Democrats got destroyed, there was a huge shift for older white men who moved away from Clinton and voted for the right. It turns out that the 90-91 recession was the first recession in American history that affected mainly older white men. I'm not an economic determinist, but . . .

Laszlo? : You mentioned this [] tendencies. For example, they are probably in central Europe, second and a half world . . . But in our countries there is an interesting phenomenon; if there is growing voter passivity for economic reasons, this means a growing number of voters for new populists who could gain legitimacy to change the political system from democracy to some kind of political system.

EM: They are from Slovakia, and the landscape there is dominated by former Communists, turned populist, turned nationalist, pulling on both the prior experiences in highly centralized political life, and on the other hand throwing on people promises of security and safety. So this is anchored in post-communism of a new version. It does not sound much different from Buchanan.

DH: Isn't the same thing happening in Russia?

EM: The people around Zhirinovsky are trying to appeal to nostalgia after Communism, and kind of reworked it into nationalism. But I understand, the question here is what one should do in places where labor market did not exist because there was disguised employment, everyone supposedly had jobs, that old story that we pretend we are working and they pretend they are paying us. But the situation now is different, some of the economies are being partly decentralized, and where to look for a model? America is one of the models, and of course all those countries are looking for that model, because Jeffrey Sachs is a fellow traveler of that model!! And what else is available. And what to do in order not to fall into that kind of a hole that you just presented, in which there is a large group of people who fell out of the resources. What are the policy implications for those who are trying to study this model as one that is successful?

DH: But what is success here? You can't possibly call what has happened in the United States in the past 15 years successful, where you see a majority of the population seeing declining real earnings--over half the population. That's not normal in an economy that's not in recession, that's growing, you see people with declining real earnings. And that's not a temporary thing, that's something that's been going on since the early 70s. I would say that this is not a model to look towards. I don't know what the model is. But the problem with a decentralized focus in the labor market . . . Decentralization is fine in the market for oranges, or the market for pencils. But people have to eat and breathe, and they make up the economy and they make up the social life of the community. You can't have a democracy when people are starving, or when people aren't educated. And you need to get people to be educated and to bother voting, they need to have bread in their stomachs. And they do not have bread in their stomachs with a decentralized labor market, in my view. You can go right back to Adam Smith, or Marshall, all the great economists upon which neo-classical economics is built, and you will see that they argue for government intervention. Because a labor market that is purely decentralized will bid the wage down below the subsistence level. It will always happen. If there are no social norms to keep it from happening, it will happen. Now Malthus thought that well, enough people will die off, and there will be a shortage of people, and the next generation will be better off. That's one way we could go [laughter]. But I don't think in the current day . . . we're going to get a fascist reaction way before that happens. That's not an option.

EM: How is it possible to have a not decentralized labor market which has all those pitfalls in a decentralized economy?

DH: You have labor standards, as you have in most of Western Europe. They trade on the international market, firms compete against one another in France, but wages are set collectively. In Austria, 98% of all workers, that is everyone, managers, professionals, their wages are set collectively. In every European nation, it's over 80%.

EM: What does that mean, set collectively?

DH: That means your wages are set not as an individual, but as part of a group. It's either you're part of the civil service, so you work for the government and your whole occupation is set. A colleague of mine works at a university in Paris. His wages aren't set by his boss. His wages are set by a national scheme. He has to take a test . . .I don't know what your situation is in Eastern Europe, but in France there's a national test, and if he passes the test he gets placed at a university, and there's a salary rate that he is locked into. When I came to the New School, the wage I cut was between me and the dean. That's a decentralized labor market; everywhere else wages are collectively determined. If we really need someone, we're desperate, we'll pay that person way more than people who have more seniority to get him here. Europe doesn't work that way, and it shouldn't work that way. I think there's other ways to obtain a vital economy than to follow a low wage race to the bottom, essentially. Because in the end it doesn't benefit anyone. It's like cities competing, or States competing. New York is unique because we're on a border with two other States, so New Jersey agrees to reduce or eliminate taxes or subsidize firms that leave New York. But the only winner in that game is the firm; New Jersey's not going to get any taxes, and New York loses employment. What's required is federal legislation that makes that not possible, that you can't offer those inducements to move. Firms should move not based on taxes, but on the basis of real economic considerations. Any basic textbook in public finance will tell you taxes are inefficient if they affect the behavior of the agents, individuals or firms. Well this is the most extreme example, where States compete with one another. That's another example where there can be too much competition. The competitive market can work too well, and not benefit the economy in the short run, because of the pain it exacts, or in the long run, because of the incentive for firms to go what I call the Third World route.

Finally, Japan has a lot of regulations on its labor market, and it's doing real well in manufacturing. It faces a lot of global competition, it has new technologies that it's introducing more rapidly probably than any other economy, but it does not have a freewheeling decentralized labor market. It's not necessary, I'd argue, for economic success; we can see that in Korea, Singapore, and Japan.

Question [Gabor]: What do you think is the possibility to introduce such a policy? Because as far as I see, there is a short-term contradiction between short-term competitiveness and competitiveness in the long run. I think your argument is quite similar to the ideology of the Swedish welfare state. They placed emphasis on lowering wage inequalities, and behind the ideology was the idea that it would create an incentive for employers to employ high-skilled workers, because they are more productive than low skilled. But you know the wage difference was quite small. But right now if you would set some standards, raise the minimum wage to influence employers to employ or to invest some money to create let's say jobs for high skilled workers, what would happen on the one hand with low skilled workers. I'm not sure in the United States it is possible to introduce such a policy, where firms have such a voice . . .

DH: I agree. The political feasibility of moving sharply away form a decentralized labor market isn't great, I agree with you, but that doesn't mean it isn't the right way to go. Again, in the U.S. it's particularly difficult because there's an ideology that's particularly individualistic, and you've got to counter that. It would require, not protectionism, but some kind of managed trade; it would require regulations on the ability of firms to move within the United States in search of low wage labor. These are not complicated; there are lots of countries that have these, including Japan. You can't just close a plant in Japan and move to another part of Japan, much less Thailand, without informing the government and without some period of time. There are regulations. We don't have those. That doesn't mean I think it's feasible politically to move there quickly. But with enough pain, and the kind of response that we see to Buchanan's very protectionist campaign, I think you're going to see a shift. Even in Clinton's State of the Union speech, you saw him center stage argue for a rise in the minimum wage; he could never have done that two years ago. Reich has been pushing a minimum wage since he came in, and it's only in this last year that the political situation was such that he felt comfortable arguing for an increase in the minimum wage.

The really difficult situation is where you have borders with very low-wage regions. Where you have WEST Germany-East Germany or East Germany-Poland or the US and Mexico, where it's fairly easy to move and you have huge wage differentials, how do you limit the search for the race to the bottom, taking advantage of low-skilled workers. There have to be international agreements; we need international agreements on the exploitation of child labor. they exist, but they're not enforced.

Gabor: But from our perspective it's maybe not the best, because we would make restrictions on yourself on our workforce for investors. It's an argument for the exploitation of these countries.

DH: Well, like trade, it's something that needs to be managed. It's not either or, either complete decentralization or complete controls. It's got to be management of the process. so before a firm can move into a low-wage region, there has to be a year's notice. There has to be a certain recognition of the effects on the community in the developed area. And there has to be a limit to the exploitation in the underdeveloped area; that is, you can't use children, you have to pay a certain wage, you know -- that seems sensible to me, but the political reality of it is something I leave to David Plotke and the rest of you.

EM: Could you comment on the phenomenon of social dumping? Moving factories to Asia, to those countries in Asia that do not have a social safety net, security for the workers, which drives the prices of the products down, and therefore Germany, or France, or the United States cannot compete with these countries.

DH: Well that's what we're talking about. Social dumping is sort of the end result of this process of decentralization. And this works wonderfully if you're an upper income consumer. This whole system is designed for the top ten percent, the only people who are seeing increasing real earnings. Why is their real earnings increasing? More income is going to the upper ten percent, and particularly the top one percent; but on top of that, prices are declining, so they're reaping the benefits of these declining prices that are based on Thai workers, children working in Bangla Desh, Mexican workers, . . . It's all based on labor.

Belinda: I'm not sure it's fair to say it only benefits the upper income levels. Cheap clothing, that sort of thing, is beneficial for every income level.

DH: That's true, but those who have the lower incomes are earning less every year. So prices are going down, but in real terms their income is going down independently of the price decreases. So for example an apparel worker in a sweatshop in Chinatown will be earning 20 or 25% less now than 10 or 15 years ago. So yes, some parts of their food and clothing budget is less, but their earnings are also less. What's really amazing is that at the top, earnings, money income has increased, but the costs of living have fallen, so it's a windfall. It's the definition of a consumer society. The access to that well-being varies by income class. And that has increased over time, it's gotten worse.

Belinda: What are the implications of the fact that some workers, like the clothing workers union, are starting to be interested in how workers are treated overseas, like the GAP campaign, what's going to happen with that?

EM: What's the Gap campaign?

Belinda: The Gap [clothing company] was subcontracting in Latin America, and had sweatshops and was using child labor, so the clothing workers union, UNITE, started a boycott campaign in this country against the Gap, and the Gap did agree to improve conditions to some extent, and that's an interesting development, I think.

DH: Well, there are several strategies that unions have adopted, and that's one of them. If you're in the old-line manufacturing, you need to go that route, or there's no way to compete, because firms are moving overseas. There's no way that an American worker can compete with a 12-year-old or a 10-year-old Bangladeshi child.

Question [Gabor]: It's not necessary to be a child, they could be adults . . .

DH: But they don't employ adults, they employ children. So the first step would be to legislate internationally against the most egregious cases of exploitation, which is children. The next step would be to keep, there's clearly a surplus pool of labor in Bangla Desh, and in Mexico, so the next step is to insist on labor standards going beyond that, which is that they have the right to join unions, they have the right to collectively organize, and they have the right to a minimum wage that sets a floor on the standard of living, so that these demand and supply curves, to the extent that they work, don't bid wages down to below a level that people can subsist on--which is what, as long as you have a surplus pool, will always happen. Either you need a shortage of workers, or you need social norms, or you need government regulation, one of the three, to keep wages from being bid down.

EM: Legislate internationally--how does one do it?

DH: Well we do it now, it's just not very effectively implemented, enforced. Through the UN. There are laws against child labor, which the U.S. has signed. But we simply don't do anything about it, we just say yes, we agree it's a bad thing.

Question: But the Bangladeshi government government is primarily responsible for using child labor.

EM: This is the fault of the Bangladeshi government. Or is this the fault of American corporations?

Question: You do not have the force to implement this. AS far as I understand it, right now there is this debate with China about piracy, and it is quite clear that the firms of the United States and the United States government has an interest in forcing China to ban piracy. But I can't see where is the interest of American firms to make pressure on the government to make a boycott agianst products from Bangla Desh which were produced by an American factory.

DH: It has worked in Guatemala. The union has organized enough political action through churches and through its own contacts. A lot of these companies are very threatened by public relations. The Gap--I have a ten year old, and the Gap now is not cool. In their school, the older kids have organized against Gap. But on top of trying to stay up with the fashions, the last thing they need is to be known as an exploiter of children. That's not a good image. And so the fear of that drove their agreements not to use these contractors, who are outrageous. It's not the Gap that was doing it; they were contracting out, outsourcing, to these suppliers who would set themselves up in villages and basically go unregulated.

It's the intersection between politics and economics in this issue that make it so interesting.