U.S. Department of Labori
Bureau of International Labor Affairs
September 2002

U.S. Department of Labor
Bureau of International Labor Affairs
September 2002

Citation

Materials contained in this publication are in the public domain and may be reproduced without permission of the Federal Government. Source credit is requested.

The views expressed here are solely those of the authors and may not necessarily reflect the official positions or opinions of the U.S. Department of Labor or the European Commission.

Acknowledgement

Conference was sponsored by the Employment and Social Affairs DG, European Commission in Brussels on February 21-22, 2002, under the auspices of the United States - European Union Working Group on Employment and Labor-Related Questions. This group was established in 1995 under the New Transatlantic Agenda.

Contents

Opening remarks:

Keynote addresses:

Presentation

Discussants

Presentation

Discussants

Summary and Comment

I am very pleased to welcome you all here today, especially our guests from the United States. Returning to my home country means a three-and-a-half hour flight each way, so I am particularly aware that travel is a tiring process. Making it all the more important for us to use our time profitably.

There must have been a time when the contrasts in these Trans-Atlantic encounters were rather more dramatic that they are today.

Although the more dichotomous views had largely evaporated by the time of the first G8 Jobs Summit - chaired by President Clinton - nearly a decade ago, which concluded that both the US, and the EU, had things to learn from one another.

On our side, we can learn about the benefits of efficiency and flexibility in the labour market. On your side, you can learn about the benefits of supportive social policies.

Since then, some things have changed, and some things have not.

Among the things that have not changed is the fact that average US living standards are still around 45 per cent higher than in the EU - according to GDP data, adjusted for purchasing power, just as they were 30 years ago. Although we - European as well as Americans - have doubled our living standards during that time, this is not a bad achievement.

But we know the main reasons for our relatively poorer performance. And it is not a lack of productivity. Productivity per hour in Europe is very close to that in the EU.

Our differences in living standards are primarily due to two factors.

• The first is our much lower rates of employment in Europe. Which we see as a bad thing since we are relying on too few people to maintain our living standards.
• The second is the rather longer average annual working hours in the US. About which we are much more ambiguous, since we do rather like our long holidays!

But we are doing things about our employment rate. The level remains way below that of the United States. But it has risen strongly, all the same, in the past 4 to 5 years. It has risen from little more than 60 per cent for most of the 1980s and 1990s to 64 per cent today. And, there is a political commitment to hit a target of 70 per cent before the end of this decade.

For that we can offer thanks to the example of the United States. As well as to some of our own Member States - notably Denmark, Sweden and Austria - who have also achieved employment rates on a par with the United States.

We recognise the challenge we face. But we are pleased to have created some 10 million additional jobs over the past 4 to 5 years, and to have brought unemployment down from close to 11 per cent in the mid-1990s to below 8 per cent today without increasing inflation pressures in any way. This undermines all those carefully concocted theories that sought to prove that Europe was forever condemned to live with high intransigent levels of structural unemployment!

But the greatest significance is not in what we have achieved, but in what we intend to achieve. And how we intend to do it.

Despite all the temporary short-run difficulties our economies face - and we do not under-estimate them - we plan to continue our progress in increasing the employment rate. Through a combination of renewed economic growth, and a continued radical transformation of Europe's labour market and social systems.

The aim is not the destruction of the European model, but its modernisation and revitalisation.

With the emphasis on the continuing reform of national employment policies through the so-called 'open method of co-ordination', in pursuit of common EU objectives, and respecting common EU guidelines. But acting primarily at national level, in line with the specific needs and priorities of the countries concerned.

This dynamic, inter-active, process of mutual support and incentive has spurred all our Member States to take action at a pace that few of them could have thought possible when we began.

The success of this 'method' means that it is now being applied in related policy areas - notably in tackling problems of social exclusion, which are so often linked to employment difficulties at the bottom end of the labour market.

Alongside the reform of national systems and policies, however, we are looking increasingly closely at the operation of the European labour market as a whole.

With the emphasis on increasing skills and mobility, we are bringing the same kind of single market logic to bear on the labour market, that we have employed in working towards a full single market in products, services and financial markets. Recognising that mobility of workers has been the 'poor relation' of EU freedoms.

To this end, we launched - earlier this month - an EU Action Plan on Skills and Mobility, which will be submitted to the Barcelona European Council. Focusing on removing barriers to mobility, investing is skills and education, and providing better information on jobs and mobility generally.

One proposal that has particularly caught the media's imagination concerns not a common EU social security card, as you have in the United States - for that we will have to wait a little longer - but at least a common electronic health card. A practical and symbolic step, which would simplify current paperwork systems, and give proof of the right of Europeans to have medical expenses incurred in another country, reimbursed by their own Member State. This is an important provision, especially for mobile workers.

You will be hearing more about the 25 specific proposals we have made to promote skills and mobility during the rest of the Conference. Like most changes, they take a little time to negotiate and implement. But they will happen. Overall, this programme represents a clear change of vision, and pace, in the European Union.

Just as we get all this underway, of course, we will be working alongside many of the candidate countries, who are looking forward to entry in 2004, although they will not have full labour mobility until several years after that date.

By the time all those transition periods have passed, however, I am sure you will see, in the new enlarged Europe, a labour market which is much more easily recognisable than the one you are used to witnessing, or reading about. But it will be a European labour market, which, I hope, will manage to combine the best of all possible worlds - our traditional standards of social performance, and the new standards of economic and employment performance that we are seeking to achieve.

Thank you.

Director-General Quintin, distinguished guests and colleagues!

This is my very first EU-US event and it’s a real pleasure to be here. On behalf of the U.S. delegation, I want to thank you for hosting this important conference and to say how much we appreciate your warm welcome. Secretary of Labor Elaine Chao asked me to convey her warmest personal regards and her wishes for a successful conference.

We share a common goal: Helping people to achieve their hopes and dreams in the workplace of the 21st century. We are really partners in this pursuit. There is much we can learn from each other about the workforce of the 21st century and the challenges it presents to us.

The profile of employment policy has been raised. It is now routinely seen as a key element to economic success in our increasingly integrated global economy. The benefits that open trade and new technology can provide hinge to a large extent on the skills, mobility and adaptability of our work forces.

I understand that you had a very fruitful discussion in Washington last June before I joined the Department of Labor on this latter issue - work force adaptability - concluding that although labor market adaptability is important to lowering unemployment, we must not lose sight of the importance of flexibility in product and capital markets as well. Plus, the importance of the ‘quality” of the jobs we are creating.

Let me build on the ‘quality of jobs’ issue because it relates to the main theme of our conference - to exchange views on future labor market needs, where it seems clear that knowledge is the driving force. This means understanding that we are part of a new global economy that places a premium on skilled workers, information, and technology.

A few important facts: this is really to just set the stage for the more detailed discussions we will have later as our conference unfolds. These facts support that acquiring skills pays off and there is a growing demand for skills.

• In the United States, the unemployment rate for a high school dropout is four times the rate for a college graduate.
• The earnings gap between high school and college graduates has ballooned to 70 percent.
• Over the next 10 years, 4 out of every 10 new jobs will require a postsecondary vocational or academic degree.

So, the DEMAND for skilled workers is increasing but the SUPPLY of workers is decreasing. The shrinking number of workers is a demographic certainty. For example, in the United States the number of people in the labor force ages 25-34, an age group comprised mostly of persons that work, especially men, is projected to decline by nearly 3-million in just the next 7 years.

If there is a silver lining to the demographic changes, it is that they provide us with an opportunity to reach out to groups whose participation in the labor force has been typically lower. They include -- people with disabilities, older workers, disadvantaged workers and others - that can help take up the slack. We may not have sufficient time to discuss all of these groups over the next day and a half, but they were worth noting because training some of them may require special attention.

If our nations are to remain competitive, and our people are to continue to enjoy the fruits of prosperity, we must build and sustain a high-skill workforce that draws on the talents of our broad and diverse populations. We must recognize this tremendous challenge and commit to finding solutions. We must be open to new approaches and new ideas and share our experiences.

What are we doing in the United States to meet this challenge?

• We are starting by building a solid, basic education foundation.
• We are working more closely with business to identify the needed skills and, where severe shortages exist, we are helping them train their workforces [Workforce Investment Act]
• We are using technology to make it possible for more people with disabilities to work [New Freedom Initiative]
• We are moving to develop more flexible working arrangements for older workers like ‘phased retirement.’
• We are opening our doors wider to immigration [H-1Bs]

The skills and demographic trends we are facing today did not start overnight - and they will not be resolved overnight. We are reaching out widely for solutions, as this conference illustrates, to the business community, the labor community, academic experts, and to other governments.

Thank you!

Good afternoon, and thank you for that warm welcome.

It is a great honor for me to be here today, and compare notes with some of the best and brightest looking at workforce issues worldwide. The nations of the European Union and the United States face so many of the same challenges, and I’m happy to have this chance to share my office’s perspective on the issues we all care about.

First, let me tell you something about the office that I head up.

The Office of the 21st Century Workforce at the U.S. Department of Labor is about as new as the new century. After talking it over with President Bush, Secretary of Labor Elaine Chao announced creation of our office in March 2001 -- in her first major policy address.

As she put it, our mission is to identify the best ways to close the skills gap, respond to labor shortages, and make America’s workplace more responsive to the realities of how people really work in the 21st century.

With a mission like that, you can see why our office crosscuts so much between government and the private sector.

The 21st century workforce is about new technology. It’s about education and training, folks with disabilities, ethnic minorities, women, stay-at-home moms, and older workers.

In short, it’s about just about everybody.

It has been quite a year. The first thing we had to do was get the word out, so we put together a Summit on the 21st Century Workforce last June at a major sports arena in Washington.

President Bush came, as well as so many leaders in economics, education, business and labor - a first-class think-tank for people to trade ideas on what’s happening now and what’s coming up for the workforce.

But it wasn’t just futurist theory.

In everything we do, I try to strike a balance between exploring ideas about the future, while keeping our feet grounded in everyday reality.

With that in mind, the summit was also the scene of a regional job fair featuring more than 130 local employers reaching out to find real workers for real jobs in real time.

Little did we know how practical that exercise would be.

September 11 came as a shock to the whole country and the world - but particularly, of course, to the Washington and New York areas, where the jobless rate suddenly took off as tourists stayed home, travel and transportation industries faced massive layoffs, and slumping sales made economic recession real to thousands of workers.

In our hometown, the department stepped up to the plate in the most practical way we knew how.

On January 10, we took over the Washington Convention Center and hosted the Workforce Recovery Conference - a job and skills fair to help put people back to work. Everything was in one place -a full range of services for workers seeking jobs and companies looking for qualified people. Over 200 exhibitors took part, and over 8,000 jobseekers showed up to check out the opportunities.

Next month, my office is cosponsoring a conference on Women Entrepreneurship in the 21st Century - a two-day event focusing on American women business owners today. We expect about 1,000 women entrepreneurs from across the country to join us.

The time is right. In addition to making up nearly half the U.S. workforce, women have launched 9 million businesses in recent years, employing more than 27 million workers.

By any standard, women-owned enterprises are a driving force in the U.S. economy and they deserve this kind of attention from Washington. I’m very excited about it.

We are also soon publishing the first edition of a magazine called “21,” which will be widely distributed and serve as a clearinghouse for both new ideas and best practices in the workforce.

In many ways, the title of the lead article says it all: “It’s not your Parents Workplace.”

And it’s not.

In the 1960s, the typical American worker was a male, white, full-time, wage earner who generally stayed with one employer and learned most of their skills on the job. Women and minorities were not a large part of the picture - let alone older workers, immigrants and the disabled.

Today, our workplace is transformed. The ethnic, racial, age and gender makeup of the U.S. workforce has been hugely altered due to a fundamentally tight labor market, increased labor force participation, and changes in demographics and immigration.

And workers are on the move like never before. Nobody is staying in one place. By age 34, the average American worker holds nine jobs, which might mean nine skill sets, nine health programs, nine retirement plans, and nine different employment markets around the country.

I know we’re going to talk about mobility elsewhere at this Conference. It’s a great topic and deserves all the attention it gets.

But wherever they go, people relying on old skills, old habits and old ways of working are in trouble. Too many people don’t have the skills today that employers are looking for.

As you know, it’s called a skills gap - the short name for the long distance between what an employer needs and what a prospective employee has to offer. Employers are having a tough time finding qualified job applicants. At the same time, thousands of workers want those jobs, but lack the skills.

The problem isn’t new, but I think it has gotten worse. There has probably always been a skills gap to one degree or another as mankind progressed through the ages.

But they are particularly severe when expanding markets and new technologies meet and greet with such historic impact - which has been true for the last 25 years - and the rate of change is likely to accelerate in the century ahead.

How should we face down the skills gap?

To me, the answer is common sense: Maximize the number of qualified workers in the workforce by --

• One, upgrading education and training for everyone who needs it;
• Two, opening the doors even wider for those who are too often shut out of the market, such as racial minorities and the disabled; and
• Three, making the workplace more attractive for men and women balancing demands of family and career.

At the Office of 21st the Century Workforce, we are pursuing all three.

First of all, closing the skills gap requires better education and training.

It’s no secret that the returns to society from a more educated workforce are positive in any number of ways, in addition to providing the skills that a modern economy depends on.

What are workforce skills, after all? More and more, they are acquired from education.

We in the administration were pleased at the recent passage of the landmark No Child Left Behind Act. We are optimistic that it will improve overall student performance and help close the skills gap that no industrialized country can afford or tolerate.

President Bush's basic policy outlook - insisting on higher standards and greater accountability - will help America’s children acquire the skills in K through 12 that they need for the working world ahead.

The U.S. business community knows this better than anybody, and is investing heavily in its human capital - skilled workers.

Last year, American business spent over $300 billion on training and education for employees - more than the budgets of all institutions of higher education in the United States combined. That commitment is likely to grow - and the business community is not alone in this effort.

It is not widely known, but the U.S. Department of Defense is the largest single training organization in the world - a clear recognition that mindpower is crucial to firepower. In 2000, the U.S. Armed Forces spent a combined total of $14 billion on training millions of members of the Armed Forces.

For our part, the Department of Labor is the largest organization in the U.S. government focused on training in the private sector. Therefore, we are exploring ways for the two Departments to combine forces to ensure that military training resources are more compatible with civilian ones, and do a better of job of making them universally available to promote civilian careers.

We are partnering with DOD in the Advanced Distributed Learning program to develop Internet hardware and software in a standard format, but tailored to individual needs. Instead of the old one-size-fits-all online classes, this program has the potential to reconfigure slices of knowledge, which then directly help people acquire credentials for specific careers. It is personalized learning at its best.

The academic community is also very active exploring new ways to deliver skill training. I recently met with the president’s of 10 community colleges nationwide. Community Colleges are really great places for learning because they’re inexpensive, fast, and focused. We’re collaborating with them to promote e-learning programs where adult learners get homework from the Internet, take classes on the Internet, and talk to teachers on the Internet.

Again, the goal is to make learning available to anybody, anyplace, anytime.

In the lifetime of everyone here, technology has exploded to the point where the United States and the countries of the European Union have moved from economies based on manufacturing to one based more and more on information technology.

It’s the age of the knowledge-based economy.

In the United States today, IT industries directly employ about 8 million people, and are projected to employ over 11 million by 2004. The industry has become the leading engine for growth across all sectors of the economy.

What does this mean for the skills gap? Plenty.

Technology provides the tools to make workers more productive -with all that means for national prosperity. But to use those tools, workers need more and new skills.

In a nutshell, technology contributed to creating the skills gap - but holds the key to closing it. Technology is also making a huge contribution to the more mundane task of even finding a job.

A common frustration among jobseekers in the U.S. has been the difficulty of tracking down quality intelligence on who is hiring and where - and having to go from one place to another to check it out.

At the Department of Labor, the vehicle for bringing together this array of information is our One-Stop Career Centers. These Centers are now in all 50 states - combining services in one central location for jobseekers and employers alike.

Finding jobs and training may never be easy - but it is certainly more convenient than ever.

Again, communications technology is the key. We recently partnered with Monster.com - the private sector’s largest online job bank - to cross-reference all our job listings with theirs. It’s a great example of a public/private partnership maximizing the power of the Internet for job hunters.

All of this is making it easier to reach out and make room for the skills of those people who have been shut out of the workforce in the past. To remedy the skills gap, we need to try even harder to make our workplaces free of discrimination on the basis of race, age, gender, color or disability.

Of course, laws in most of our nations are mandatory on this topic, but the realities of the skills gap mean that employers need to live up to its letter and spirit. As Secretary Chao recently put it, “Opening the doors of opportunity to all used to be a moral imperative. Now, it’s an economic one as well.”

In particular, too many skilled workers with disabilities are unemployed. To address the problem, President Bush recently launched the New Freedom Initiative to help Americans with disabilities buy computers, receive special needs education, and maintain health benefits when returning to work.

At the Department of Labor, we created a new Office of Disability Employment Policy to speak up for the disabled in the business community. The goal is to help disabled Americans enter the workforce, contribute to closing the skills gap, and enjoy better access to the kind of life that the rest of us take for granted.

Narrowing the skills gap also demands more flexibility in the way the workplace itself is set up. Employees want more freedom to choose their own terms of employment - and smart employers are listening.

In a flexible workplace, all kinds of workers - married, single, with or without kids, dual earners - must find jobs that fit their needs - and employers must find employees whose availability and skills fit the job.

In the past decade, so many of us have realized that technology is not only the driving force in the economy, but important in managing our personal lives. Everyone now seems to be interconnected, which is paving the way for more flexible, custom-fit working arrangements.

An office used to be a building filled with employees, a cafeteria and a water cooler. Now, it could be an airplane seat, or a desk in a kitchen with a laptop and a modem. Before long, an office could be nothing more than a cell phone and a Blackberry.

The trick is to make new technologies work for us, rather than the other way around - to use innovations like telecommuting and tele-work to give employees more options and more control over their time, rather than less, and keep them satisfied on the job.

We are entering a brave new world. A global marketplace for goods and services is creating a single, global supply of labor. The interests of working people in all nations are more and more interrelated.

I believe these men and women are at the heart of what we are doing here in Brussels.

I know there will be a lot of talk about various programs and policies to help workers adapt to a changing workforce, but at the end of day, what we really need to offer is hope -

• hope to the unemployed by providing the education and training to help them find work or a whole new career;
• hope to those who lose a job or promotion due to race, gender or religious prejudice;
• hope to the disabled who deserve a shot at the mainstream; and
• hope to anyone overcoming obstacles to joining the workforce.

Certainly, workers themselves must take the initiative to shape up their marketable skills and find the right balance in their professional and personal lives. But government can be a tremendous catalyst to set the stage, show the way and clear the path.

That’s why it’s such an exciting time to be at the Department of Labor. Under President Bush and Secretary Chao, the department is focused as never before on workforce issues.

It’s open season in Washington for new ideas on how to solve immediate problems and put the right tools in the toolbox for success in the years ahead.

I know everyone here might define “success” in different ways, but I hope we can agree on one thing: No matter what path we take to get there, our shared goal for the 21st century workforce is a satisfying job for every man and woman who wants one.

Thanks for listening.

There is an important difference in the comparative performance of the US and EU in terms of employment growth and productivity.

Output per worker, 2000

Many factors can contribute to productivity improvements, including high levels of enterprise and innovation, high levels of domestic competition, high levels of capital investment, better management and a more highly skilled workforce. There is a well-established relationship between improvements in skills and increased productivity. Growth theory suggests that human capital is one of the prime determinants of labor productivity. Human capital is increased both by formal education and training and by learning-through-doing. Skilled workers can often adapt faster and more effectively to change and may be better at implementing new investments and innovation. They can thus increase the ability of a firm to update its practices and products and the rate demanded by rapidly changing markets, making the economy more flexible and more productive over the longer run.

Percentage of workforce by qualification level

The impact of skills on productivity also works partly through effects on capital investment levels. Evidence suggests that having highly skilled workers helps firms gain the full rewards of new investment and thus increases the likelihood that investment will occur. Research examining the causes of international productivity performance suggests that differing levels of skills play an important role. For example, estimates have been presented which suggest that between half and all of the UK productivity gap with Germany can be explained by skills differences. Plants with high productivity tend to have a higher proportion of skilled workers in their workforces. Differences in physical and human capital can explain around 60 per cent of the productivity gap between domestically owned firms and US-owned firms in the UK, and nearly all of the gap with other foreign-owned firms.

Research at the worker level finds that holding a tertiary qualification (e.g. a degree) in the UK increases productivity by between 30 and 100 per cent when compared to a worker with no qualifications. Holding an intermediate qualification at NVQ level 3 results in higher productivity of between 8 and 50 per cent when compared to productivity of an unqualified worker.

Turning to international comparisons of skills, although the UK has a large number of workers with high skills, it falls behind France and Germany in terms of intermediate skills, with a higher proportion of UK workers having low skills. More than a third of UK workers have low skills, compared to less than a fifth in Germany. It is difficult to be certain, but the UK may lag significantly behind the US as well as European countries in terms of immediate skills. It is important to recognise that the US data includes hardly any adult/workplace training due to the lack of recognised qualification structures, whereas UK data includes much of this training, as many people acquire nationally certified qualifications. Around 9 per cent of US workers cite long-term workplace training as their main source of skills, with a further 27 per cent citing moderate workplace training. This suggests that US skills might be substantially understated by the conventional data. Despite debates over some of the international evidence, some clear messages emerge from the data:

• The number of people in the UK with high skills compares well with international levels (although the UK still trails the world leader - the US);
• The level of intermediate skills in the UK is low, especially compared with Germany, other European countries and, under some interpretations, the US;
• There are a large number of people in the UK workforce with low skills. UK performance in terms of basic skills is very poor, as highlighted by the Moser Report. The International Adult Literacy Survey, for instance, found that over 20 per cent of the adult UK population lack functional literacy. Only Ireland and Poland of the thirteen countries surveyed fared worse.

The evidence also shows that the UK has fewer workers with vocational qualifications than its European competitors, and often provides young people with a less diverse academic education than does the US. There were important lessons to be learnt on both sides of the Atlantic from the data on skills and productivity. It is necessary to look behind the numbers, which oversimplify the differences between the European and US labor markets and look in detail at successes and failures on both sides.

In its discussion of purpose, the abstract for this conference notes that ‘knowledge’ will be the ‘driving force’ shaping labor market needs for the 21st Century Economy. To create a focus for our discussion, the abstract further notes the importance of understanding the roles of “education, training, and skills as well as the need to promote occupational and geographical mobility of workers.”

This paper explores three issues related to the purpose of this conference using evidence from U.S. labor markets. The first examines the statistical evidence for United States’ labor markets that underlies the basic theme of this conference - namely, has the well-documented trend toward hiring more-skilled workers in U.S. labor markets that began in the early 1980s continued as the economy entered the 21st Century?

Second, most discussions of the demand for skills in the labor market use educational attainment as a proxy for skill levels. While I do not depart from this tradition, I attempt to refine the discussion by examining the variability of the returns to educational attainment within each educational attainment category. I use a group decomposition approach to examine the role of gender, occupation, and potential experience as proxies for describing the nature of this heterogeneity within each educational attainment group. After controlling for gender, occupation, and potential experience, I find that the increased inequality in earnings that occurred between 1992 and 2000 was accompanied by an increased variability of earnings within each of the educational attainment groups. The human capital variables that are traditionally used to measure the skill/earnings relationship do not effectively capture the changes that occurred over the last decade.

After examining the recent trends in the returns to educational attainment and the variability of those returns within each educational category, I then turn to an examination of the prospects for the future using recently published Bureau of Labor Statistics’ projections for occupational employment demand over the 2000-2010 period.

To lay the groundwork for these three themes, I first provide an overview of the concept of skills as it is used in this paper and provide a description of the data used to explore each of the three sections that follow:

The focus on education as an empirical proxy for skill levels, and the measurement of the returns to skills on the basis of the relative earnings of different educational attainment groups, reflects both the explanatory power of education in human-capital earnings models, as well as the limitations of existing U.S. cross-sectional data sets in measuring other dimensions or proxies for skill acquisition. For example, the data set used in this paper, the quarter-sample monthly earnings file from the Current Population Survey (CPS), does not collect routinely any measures of job training (formal or informal).

In contrast, there are some very significant longitudinal data sets, most notably the 1979 and the 1997 National Longitudinal Surveys of Youth (NLSY79 and NLSY97, respectively) that contain extensive information on both training and education. In addition, both surveys contain measures of skill levels through a series of tests known as the Armed Services Vocational Aptitude Battery. The NLSY97, an annual longitudinal survey that began in 1997, contains individuals who were 12-16 years of age on December 31, 1996. This survey is also collecting high-school transcript information (and intends to collect college-transcript information). This survey has tremendous potential to add to our understanding of the role of education, training, and skills on labor market returns as members of this cohort complete their education and enter the labor market.

For this study, however, I use the earnings file from the U.S. Current Population Survey, a nationally representative data set that provides a more complete picture of U.S. labor markets at the beginning of the 21st Century. The earnings file from the CPS is based on interviews with one-quarter of the monthly cross section (2 of the 8 rotation groups in the sample) and provides rich information on the demographic and socio-economic characteristics of individuals ages 16 and older in the U.S. population. While there is a relative loss in the level of detail in terms of information on training and other measures of skills, there remains an analytical benefit to the breadth of the CPS in terms of its information on the educational- and occupational-structure of earnings for the entire working population.

To focus on the workforce of the 21st Century, I examine full-time wage and salary workers, using usual weekly earnings as a measure of the returns to their human capital in the labor market. Throughout the paper, I compare 1992 to 2000, a period that roughly covers the latest U.S. recovery. I chose 1992 as the starting point because the CPS began collecting educational attainment data on degrees received (instead of years of school completed) in 1992.

The proposition that the demand for more-skilled workers is a ‘driving’ force underlying U.S. labor markets is perhaps without controversy. The extensive literature on trends in U.S. earnings inequality and earnings differentials between educational attainment groups since the early 1980s is well documented. However, to focus our discussion, it is useful to review the evidence as it pertains to the most recent period of economic expansion in the U.S. economy, and to document the extent to which this recent period has (or has not) continued the increase in the relative demand for more-skilled workers.

Table 1 provides median weekly earnings for full-time wage and salary workers in 2000 CPI-U dollars by educational attainment. Not surprisingly, those with less than a high school diploma lost ground in real terms over the 1992 to 2000 period. High school graduates experienced a gain of 3.5% in real earnings, while those with a bachelor degree and those with more advanced degrees experienced much more rapid gains (6.5% and 5.4%, respectively).

Most previous analyses of earnings inequality have used years of schooling completed as the proxy for educational attainment. The availability of data starting in 1992 on degrees received gives us insight into the potential impact of the acquisition of skills from attending community college. Those with some college, but without a degree saw a gain in real earnings of 3.1%. However, those with associate degrees, either on a vocational- or a traditional-educational track, experienced no real-earnings gains over the period as measured by real median weekly earnings.

The same trends emerge (as expected) by comparing changes in relative earnings differentials between 1992 and 2000. Table 2 provides the evidence. In terms of educational attainment, using those with a high school diploma as the base group, those with a bachelor or an advanced degree gained substantially, while the differential between those with an associate degree and a high school diploma did not change over the period. Those with some college, but without a degree maintained their relative earnings differential over high school graduates.

The evidence on real earnings serves as a proxy for the workings of the labor market. Although educational attainment is but one aspect of the forces that shape the demand and supply of workers, an examination of this worker characteristic by itself is very telling. In table 3, I provide the levels and percentage changes in the supply of workers by educational attainment

category from 1992 to 2000. One approach to inferring that there has been a continued increase in the relative demand for more-skilled workers is to observe rising relative real wages for higher education groups even in the face of increases in the supply of these workers that outpace supply increases for less-educated groups.

Consider the group holding bachelor degrees. Between 1992 and 2000, this group experienced a 6.5% increase in real wages, and the relative earnings differential between those holding a bachelor degree and those with a high school diploma increased from 1.61 to 1.66. At the same time, the employment of those with a bachelor degree increased by 33.6% compared to the modest 5% gain among high school graduates. Ceteris paribus, such a relative supply shift should lead to a decline in relative wages between these groups being compared, but the rising differential suggests an increase in the relative demand for these higher-skilled workers that outpaced the movements in relative supply. A comparison of those with some college (but without a formal associate degree) with high school graduates shows the same trend.

Viewed against changes in the supply of workers, the nearly flat trend in real earnings of those with an associate degree from Table 2 takes on a different perspective. The substantial increase in the employment of those with an associate degree in the educational track (52%) largely has been met by a sufficient demand response to maintain the value of real wages over the period. The same holds true for those with a vocational associate degree.

The preponderance of evidence suggests that the returns to skill as proxied by levels of educational attainment increasingly favor those with more rather than less educational attainment. However, what these data do not provide is any insight into the variability of earnings across- and within-educational attainment groups.

One important dimension to understanding these trends in the returns to educational attainment is to examine how these returns vary by occupation. An examination of both the employment composition of and the returns to educational attainment within occupational groups provides useful insight into the labor market demand for skills.

There are a variety of approaches that can be adopted to examine how the returns to educational attainment vary by occupation. Two alternative approaches are adopted here. The first examines differences in the patterns of median earnings that emerge across classes of jobs where each class is a collection of occupations that are grouped together based on similarities in the percentages of occupational employment found in each educational attainment category.

While very instructive, this method does not provide a statistical summary measure of the degree to which the variability in the distribution of earnings can be explained by differences between and within groups of individuals based on their individual occupations and levels of educational attainment. The second method provides such a measure by using the Theil decomposition technique for estimating the contribution of between- and within-

group variations in earnings to explaining total variation, where each individual is assigned to a group based on their educational attainment and choice of occupation (as well as by gender and years of potential labor market experience).

The occupational classification used in both methods is based on the detailed occupational classification scheme from the CPS. There are 45-detailed occupations that are shown in table 4. These occupations have been grouped according to a 5-tier schema based on the distribution of employment in three different educational categories: high school diploma or less, some college (including associate degree), and a bachelor degree or higher. An

occupation is classified as a predominantly (high-school diploma or less) occupation if more than 60% of employment is in this occupational category.

Similarly, an occupation is defined as (some college) if 60% or more of the employment in the occupation has an education attainment level of some college (with or without an associate degree). An occupation will be defined as (bachelor degree or more) if 60% of the employment in the occupation consists of individuals with a bachelor degree or an advanced degree. There are three mixture occupations that are classified according to the educational attainment categories having at least 20% and no more than 60% of the employment in an occupation. The three mixture occupations are the (bachelor degree or more/some college) mixture occupations, the (some college/high school diploma or less) occupations, and the (bachelor degree or more/some college/high school diploma or less) occupations.

This division, while based on reasonable criteria, is arbitrary nonetheless. Some jobs are easily classified in terms of a college path, such as the group of occupations at the beginning of table 4. The jobs at the bottom of Table 4--classified as (high school diploma or less) occupations-are based on having over 60% of employment in the occupation being in the high school diploma or less educational attainment category. These occupations are ones associated with on-the-job training and few formal educational requirements beyond secondary school.

However, the large number of mixture occupations, having significant percentages of employment from each of the educational attainment categories, are much more difficult to describe as ‘requiring a college education’ or ‘only requiring a high school diploma’. This difficulty is testimony to the variety of pathways that individuals can take to obtain employment in a variety of occupations. Computer-equipment operators are a prime example that comes to mind, with the assortment of training courses, community college, and 4-year colleges that provide background training in this occupation.

Table 4 shows a rising trend in median earnings across the classes of occupations as the average level of educational attainment increases for each class of occupations. The group of occupations classified in the (high school diploma or less) category has median earnings of $426.11 as compared to the median earnings of $831.34 for jobs classified as (bachelor degree or more) occupations. There is a significant jump between the jobs classified as (high school diploma or less) or (high school diploma or less/some college) occupations and the jobs classified as (bachelor degree or more), (bachelor degree or more/some college), or (bachelor degree or more/some college/high school diploma or less) occupations. This trend seems to follow the general observation that there is a continuing higher relative demand for workers with higher levels of skills.

However, this raises the natural question of how does a member of an educational attainment group fare when he/she is in an occupation that is classified as having an average educational attainment category that is higher than his or her own level of education? For example, the median weekly earnings for those with an educational attainment of high school diploma or less were $436.34 in 2000. A member of this group who is working in an occupation classified as (bachelor degree or more) earned $550.68 in 2000 (see table 5). In contrast, a college graduate who is working in an occupation classified as (high school diploma or less) earned $556.26 in 2000 as compared to the median earnings for all college graduates of $869.73. As well, a worker in the some-college educational group earned $697.52 in occupations classified as (college) as compared to the $560.61 median earnings for all members of the some-college educational attainment group.

Table 5 also shows the heterogeneity present in the returns to mixture occupations. A worker with an educational attainment of high school diploma or less who is in a (bachelor degree or more/some college/high school diploma or less) occupation earns 34% more than the median earnings for all workers with a high school diploma or less. In fact, the mixture occupations (bachelor degree or more/some college) and (bachelor degree or more/some college/high school diploma or less) provide median earnings opportunities for members of each educational attainment group that outpace the median earnings for all of their educational counterparts.

The general lesson to be drawn from this analysis is that while the general trend favouring more skilled workers continues, there is enormous heterogeneity in the returns to both educational attainment and choice of occupation. This exercise is instructive, but it does not provide a summary statistical measure of the degree to which we can explain the variability of earnings in terms of differences between groups that are classified or grouped on the basis of both their educational attainment and their occupation.

In this section I examine the degree to which the variability in the earnings of full-time wage and salary workers can be explained by knowledge of each individual’s level of educational attainment and detailed occupation. I also control for differences attributable to gender and years of potential labor market experience. I use the Theil inequality statistic, which allows decomposition of the total variability of earnings into two summary terms-one that measures the sum of differences between each group mean and the overall mean of earnings (in log form), and the second that measures the total variation within each group-where the groups are defined so as to be mutually exclusive and exhaustive.

Table 6 provides the results. The grouping used includes the 45 detailed occupations
discussed above. Each of these groups is further divided by gender, 4 potential experience groups, and 7 educational attainment groups. Theil decomposition statistics are calculated for 1992 and for 2000. In 1992, these detailed groups helped explain 49% of the total variability in earnings in terms of the contribution of the sum of differences between each group and the overall mean. From 1992 to 2000, the overall Theil statistic increased suggesting that overall earnings inequality increased over the period. At the same time, between-group differences now only explain 47% of total earnings variability. In other words, as overall inequality increased, so too did total within-group variability.

In the context of examining the demand for skills for the 21st Century workforce, the importance of this finding is to add a note of caution to our interpretation of the general results that there has been a continuing shift in the relative demand for more skilled workers. While this trend is clear, it is also the case that there remains a substantial amount that we cannot explain about the variability of earnings. Often termed unmeasured heterogeneity, the variation in earnings within each educational attainment/occupation/gender/potential experience category means that receiving a degree and choosing the ‘right’ occupation is not a complete guarantee of success in the labor market.

In truth, some of the unmeasured heterogeneity in these results could be explained by a fuller specification of the human capital variables in the Theil definition of groups. However, the general finding in the literature is that there is no set of currently measurable characteristics that will increase the explanatory power of between-group differences to a level that is much higher than what has been found in this exercise. There are some potentially key measures that are not captured by the CPS, such as training or an independent test-based assessment of skills. However, there will always be a large degree of unexplained variation within groups that remains owing to characteristics such as motivation, as well as the process of self selection by which individuals choose different education and occupational career paths.

Every two years, the U.S. Bureau of Labor Statistics publishes ten-year long-term projections of the U.S. economy, including projections of occupational employment. The latest round of projections was released in November 2001 for the 2000-2010 period. The material in this section of this paper is based on these results and on the analysis that appeared in the November 2001 Monthly Labor Review article on occupational employment projections. Some of the material is taken directly from the article without direct attribution.

As mentioned previously, the occupational employment projections are largely based on data from the Occupational Employment Statistics (OES) Survey, an establishment survey with 400,000 sampled units per year. The OES does not collect information on individual characteristics such as educational attainment, gender, or potential experience (among others) as in the CPS. The OES data are supplemented by data from the CPS in order to include estimates for private household workers and the self-employed. Due to sample size differences, the level of occupational detail in the OES is far greater than in the CPS. Any discussion of the demand for skills in the context of the underlying educational attainment distribution of occupations requires that comparisons between the OES and the CPS surveys be at a more aggregate level of detail in the CPS (for example, the 45 detailed occupations considered in earlier sections of this paper).

Under usual circumstances it is possible to use the projected occupational employment trends based on OES data to examine the changing demand for skills as measured by educational attainment using the more aggregated CPS occupational codes. However, the OES is currently on the 1998 Standard Occupational Classification System, and the CPS will not convert to this system until 2003. For the purposes of this discussion, I will present projected trends as they relate to the OES-based projections data, and relate them in a heuristic manner to the demand for more skilled workers.

Total employment is projected by BLS to increase by 22.2 million jobs from 2000 to 2010, an
increase of about 15 percent. Table 7 shows how these projections are distributed based on the major occupational categories.

The two largest major occupational groups, professional and related occupations and service occupations, are projected to increase the fastest and to add the most jobs of any occupational categories over the period. In general, these two categories are at opposite ends of the earnings and educational attainment spectrum, and are expected to account for more than half of the employment growth from 2000 to 2010.

The three slowest growing occupations are expected to be office and administrative support; production; and farming, fishing and forestry occupations. Each of these occupations is expected to grow by less than 10 percent as compared to the overall expected growth of 15 percent. Administrative support occupations correspond to the category of “Other administrative support occupations, including clerical”, which in table 5 is classified as a ‘some college/high school’ occupation. The other two occupations have corresponding categories found in the ‘high school’ occupation list.

Employment in transportation and material moving occupations, corresponding to a similar category in table 5 in the (high school diploma or less) occupation list, is projected to grow as fast as overall employment. Occupations that will grow slightly less than the average 15 percent include management, business and financial occupations (bachelor degree or more/some college); construction and extraction occupations (high school diploma or less); sales and related (bachelor degree or more/some college/high school diploma or less) and (some college/high school diploma or less); and installation, repair and maintenance occupations (some college/high school diploma or less) and (high school diploma or less) occupations.

Professional and related occupations. Employment in professional and related occupations is projected to grow faster and add more workers (7 million) than any other major occupational group. Three industries, business services; education-public and private; and health services, will account for about 60% of this growth (data on industry composition of occupational employment growth not shown). The remainder of services, which includes legal, engineering and management, and social services, will account for another 20% of the growth.

Almost 75 percent of the job growth for professional and related occupations is projected for three particular detailed occupational subgroups - computer and mathematical occupations; health care and technical occupations; and education, training, and library occupations. Computer and mathematical occupations are projected to add the most jobs, 2 million, and grow the fastest among the 8 detailed professional and related occupation groups. Overall, three out of five new jobs for computer and mathematical occupations are projected to be in the rapidly growing business services industries-primarily in computer and data processing services.

Health care practitioners and technical occupations are projected to add 1.6 million jobs over the 2000-2010 period, with 75 percent of this job growth in the health services industry. Education, training and library occupations are also projected to add 1.6 million jobs. Four of five new jobs in this occupational category are projected for public and private educational services. Among these occupations, social services will have the fastest growth from its 2000 level at 36 percent (not shown in table).

Legal occupations are projected to add 216,000 jobs. Paralegals and legal assistants are expected to grow the fastest, while lawyers should add the most jobs.

Architecture and engineering jobs are projected to increase by 325,000, or 12.5 percent, the slowest growth among the eight professional and related occupations. Life, physical and social scientists are expected to add 223,000 jobs, or 19.1 percent growth.

Service occupations. They are expected to add 5 million jobs over the 2000-2010 period, the second largest numerical gain of the major occupational groups. More than one half of the growth in service occupations is projected to be in the services industry-primarily in the health, business and social services. About 30 percent of new service occupation jobs will be in retail trade, primarily in eating and drinking establishments.

Within the 5 major subgroups making up service occupations, food preparation and serving related occupations are projected to add the most jobs, about 1.6 million. Healthcare support occupations are projected to add 1.1 million jobs, while protective services are projected to add over 800,000 jobs. This latter estimate is certainly an underestimate, given the changes that have occurred since the tragic events of September 11, 2001.

Personal care and service occupations are projected to add over 850,000 jobs, while building and grounds cleaning and maintenance occupations are projected to add about 780,000 jobs.

The remaining major occupational groups can be seen in tables 10-16. Employment in management, business and financial occupations is projected to increase by 2.1 million, but only grow by 13.6 percent. Nearly 25% of employment gains will be in business services industries. About 354,000 new jobs are projected for the rapidly growing engineering and management services and 238,000 for finance, insurance, and real estate.

Employment in office and administrative support occupations is projected to increase by 2.2 million, but only grow by 9.1 percent. In nearly all industries, employment of workers in these occupations is projected to grow slower than overall employment due to continued trends in office automation and electronic communications.

Farming, fishing, and forestry occupations are projected to grow by only 50,000 over the 2000-2010 period. BLS projects that self-employed individuals in these occupations will decline by 9.1 percent.

The construction and extraction occupations are projected to add 989,000 jobs, a 13.3 percent increase. Not surprisingly, a high percentage (60%) of new jobs are projected for the construction industry, but the fastest growth of jobs in these occupations will be in business services (58%), which should account for 10% of new jobs. A decline of 18,000 is projected for the mining industry-mainly for extraction workers.

Installation, maintenance, and repair occupations are projected to add 662,000 new jobs with only an 11.4% percentage change. Approximately 1 new job in 5 in these occupations are projected for the automotive repair, services, and parking industry. Most growth, however, is projected to occur over a wide range of industries.

Production occupations are projected to add about 750,000 jobs, with only a 5.7% increase. In 2000, nearly 3 of 4 production workers were employed in manufacturing industries, but only a quarter of new jobs in these occupations are projected for this industry. More than half of new jobs are projected for the business services industry.

Transportation and material moving occupations are projected to add 1.5 million jobs, with a percentage change of 15.2, about the same as the overall economy. More than 1 of 3 new jobs in this occupational category will be in the transportation and public utilities industry, while 30% of new jobs will be in the business services industry. Water transportation occupations are projected to grow slowly, while railroad occupations will continue their long-term decline.

This article has examined three issues related to the theme of this conference. First, I find that the strong relative preference in U.S. labor markets for higher-skilled workers continued over the most recent recovery and expansionary period in the U.S. economy. Second, while U.S. labor markets continue to differentially reward, on average, those with greater levels of educational attainment, there is substantial and increasing heterogeneity in these returns. One significant measurable aspect that underlies the variability of earnings is the distribution of occupations and educational attainment. These variables, along with gender and years of potential experience, explained 49% of the total variation in earnings in 1992 and 47% in 2000. There remains substantial variability in individual earnings, however, that we cannot explain, a variability that has grown--as has overall earnings inequality--over the 1992-2000 period.

Finally, the BLS projections for occupational employment for the 2000-2010 period were examined. The largest- and fastest-growing occupations are projected to be in the professional and related group, occupations that tend to have higher average levels of educational attainment. At the same time, the second- largest and also second-fastest growing occupations are projected to be in the services group, which tend to have lower average levels of educational attainment.

In the end, the evidence points to an economy that is increasingly skill based, with a widening gap between different skill groups. At the same time, the widening variability in earnings suggests that beyond one’s decision on educational attainment and choice of occupation, there are other significant, difficult to measure characteristics--such as motivation--that are key to determining one’s actual labor market success.

Michael Horrigan’s presentation regarding the demand for educated workers in a knowledge-based society was developed and presented in a highly effective manner. It is clear that he is extremely gifted in his profession, and it is my intent to respond from the perspective of a high-technology employer rather than analyze his methodology and findings.

Prior to assuming my current role with Motorola, I served as a leader in strategizing and developing our employee benefits programs. Part of that role required that I review the reports of our health care vendor partners; in doing so, I would ask three questions:

• What…do the data show?
• So what…difference does it make, are the trends, is the relevance, etc.?
• What now…is the action we should be taking?

Mr. Horrigan has already articulated what the data show, and I will not repeat that.

So what difference does it make to Motorola? Motorola has experienced a dearth of highly qualified software engineers. First, at one time, we have had up to 3,000 open engineer positions that we could not fill. Second, we have been one of the largest users of H-1B visas in the United States. Third, we have found it necessary to make increased use of the flexible work force that Shelley Hymes discussed in her opening remarks.

What now?

First, our education programs need to be enhanced. In the United States we have recently passed the “No Child Left Behind” legislation, which is a huge step in improving our education system. It will increase the emphasis on ensuring that each child in the United States is provided a quality education with standards established, and teachers and school districts held accountable for those standards in an atmosphere of what we call a “six sigma quality” focus on continual improvement in quality. Schools will be recognized for their achievements and face the consequences if they don’t improve to meet the standards. There will be an increased emphasis on teachers teaching in the subject areas in which they have attained an academic major, and steps will be taken to especially increase the number of math and science teachers.

Second, Barbara Clark, one of Motorola’s leaders in integrating education and the work force has advised me that the following skills must be emphasized and improved if we are to be successful as employers:

• Communication skills. The ability to read, write and listen. This is especially important in a the diverse work force of companies such as Motorola
• Computation skills. The ability to learn to think. Employees must be able to disaggregate in their thought processes, and to think logically.
• Critical thinking and problem solving. My wife, who is a third grade teacher, informs me that this is emphasized in early grade school in her district.
• Individual and teamwork skills. A person must learn to take individual responsibility to work as a team member. Individuals will continue to be asked to work as team members in the future.
• Knowledge about a specific job.
• Systems thinking skills. The ability to think about entire systems in which one works…such as accounting systems, government systems, educations systems, health care systems, etc., and the interdependence of them.
• Technological skill. The ability to understand technology and its continued role in our society.
• Resource management. The ability to manage money, time, people and material.

Third, we must have a cross-fertilization of cultures and thoughts. With a diverse work force, we will have strengthened education and skill sets. By allowing, and encouraging, the movement of people from culture to culture, country to country, we will better serve our publics. Motorola must have such diversification in our thinking processes and education if we are to compete in a global economy.

Fourth, and this may sound “too far out”; we must change our mindset about work. Too often, work is seen as drudgery. “Thank goodness, it’s Friday!” Work is often seen as something to be minimized.

In reality, however, work can be one of the most rewarding aspects of our lives. If one is provided excellence in education, enters a vocation in which she is “gifted” and has a “passion” for what she does, and is provided with competent leadership, work can be very fulfilling. So, we must find ways to match education, skill sets, gifts, abilities, and passion to available positions to maximize the opportunities for a successful work force. When we do that, people will want to continue their learning and enhancing their skills throughout their lives.

Finally, I would merely close by quoting one of Michael Horrigan’s summary statements. “Beyond one’s decision on educational attainment and choice of occupation, there are other significant, difficult to measure characteristics, such as motivation, that are key to determining one’s actual labor market success.”

We concur. We have found that there are other intangibles such as “passion” or “fire in the gut” that are just as key to success as the school from which one graduates or one’s grade point average.

Mr. Jerry Zellhoefer, European Representative, AFL-CIO

In introducing this intervention, it would be useful to recall the Director General’s opening comments on the need for strong social partners to help face the challenges before us and the Deputy Under Secretary’s remarks that, in dealing with this subject, it is important to reach out to the business, labor and academic communities.

With that in mind, it would be useful to reflect on some points made by the Trade Union Advisory Committee (TUAC) to an OECD International Conference on Adult Learning Policy, last December in Seoul, Korea. In recent years, trade unions, realizing the problems that a lack of skills and competencies mean for their members, now seek to foster wider access to learning. They have continually argued the case that there is a need for individual entitlements to access foundation levels of learning skills, and for employers to be given obligations and incentives to provide such opportunities.

Trade unions negotiate training agreements with employers, raise their members’ awareness of learning, mentor their members and help to broker the provision of education and training with colleges and universities. An important issue on the trade union agenda is how to prevent employers from creating low-end jobs instead of relatively well-paid, skilled jobs. This is at the heart of the discussion of “high road” versus “low road”. The high road is characterized by the expansion of relatively well-paid, skilled jobs, particularly in the middle of the employment structure and the low road by the expansion of low-skill, low-wage jobs.

What may help close off the low road approach is the fact that organizational innovation and the skilling of workers are increasingly considered as key factors in company competitiveness and continuing training has become seen as an area in which the interests of employers and workers can converge. Although it is now widely recognized that lifelong learning has a central role to play in the knowledge economy, there has been slow progress so far in making it really happen.

For the US and the EU, the experience in most OECD countries contain some useful lessons, including training participation is significantly higher for those in employment than for the unemployed. In almost all countries, less educated workers are significantly less likely to be trained. Companies that are unionized train more than non-union companies. Workers on temporary contracts and in part-time jobs - most of whom are women - are significantly less likely to receive training.

In general, findings also suggest that schooling and training are complementary, therefore policies to strengthen schooling can also encourage further training. Against this background is the point that employers and unions should negotiate agreements including with levels of government and providers of education that make participation in lifelong learning feasible in practice. According to a recent study by the International Labor Office (ILO), the influence of bargaining is greatest in those countries where the continuing training system is based on agreements between employers’ organizations and trade unions, which have a high degree of joint responsibility in the regulation and management of training and are strongly linked to sectoral collective bargaining.

In the United States, the AFL-CIO participated along with employers, educators and the government in developing a “Blueprint for Improving Workers’ Ability to Thrive in the New Economy”. The 1999 report titled “Skills for a New Century: A Blueprint for Lifelong Learning” underscored that “the challenge was too large and too complex to be met by any single stakeholder. Partnerships must include students, workers, employers, unions, educators and government and occur at all levels.”

The 1999 Blueprint set out four key workforce learning goals: delivering education, training and learning that are tied to high standards, lead to useful credentials and meet labor market needs. Secondly, improving access to financial resources for lifelong learning for all Americans, including those in low-wage jobs. Thirdly, promoting learning at a time and place and in a manner that meets workers’ needs and interests, i.e. family friendly learning opportunities and, fourthly, increasing awareness and motivation to participate in education, training and learning.

Another focal point in recent years, especially for the OECD countries is the Internet in the “new” and other segments of the economy. The Internet is like a huge wave poised to crash against our shores. It is not only a major cultural force; it has emerged as a powerful engine of economic change. Employers, employees and, yes, labor unions, will need to learn how to ride this wave, or risk being submerged by it in the coming decades.

Unions, rather than opposing the Internet, are embracing it as a means to help working families obtain a greater voice in the global economy. They employ the Internet to enhance membership communications, do outreach, coordinate internal activities, conduct research, mobilize for political and organizing campaigns, and deliver training. The AFL-CIO, most major labor unions, and numerous local labor bodies operating their own web-sites. In December 1999, the AFL-CIO launched its “Workingfamilies.com” initiative, offering low-cost Internet services to active and retired union members, potentially reaching 17 million households.

The growth in information technology (IT) jobs and worker earnings reflect a tightening of IT labor markets. IT industry leaders have responded in two ways. First, claiming a severe “shortage” of IT workers, they successfully lobbied Congress to raise the cap on H1-B visas, which regulates the number of foreign, high-skilled workers U.S. firms are allowed to employ, on limited-time basis. Second, many IT firms employ large numbers of skilled U.S. workers on a contingent basis, many working alongside and doing work comparable to that of permanent employees on a year-round basis (“perma-temps”). For example, of Microsoft’s approximately 6,000 “temporary” workers - roughly 1/5 of its Seattle workforce - more than 20 percent are “perma-temps.”

While both approaches supposedly are aimed at filling short-term skill shortages, neither addresses the long-term labor force needs of the Internet economy. More important, they appear designed to apply downward pressure on labor cost. For large numbers of workers, IT employers are freed from providing pension, health insurance, stock options, and other benefits offered to regular employees. Not only is this unfair to U.S. workers, it sends a distorting signal to labor markets. It also creates disincentives for industry investment in attracting, recruiting, educating and training a domestic pool of IT workers.

In this highly volatile environment, governments and businesses are mostly concerned about obtaining an adequate supply of skilled workers. Unions and their members are worried about maintaining secure incomes, getting and keeping current the skills needed to get good jobs, developing career mobility, obtaining health care and family benefits, and providing for retirement. All therefore have a stake in revamping the U.S.’s poorly managed “pipeline” for skilled workers.

We need substantial public and private investments to help workers make the transition to the new economy. Along with income supports and job search assistance, we need new training initiatives to equip workers with the skills they need to obtain the newly created IT jobs. Unions have taken the lead in such efforts. In 1996, the Communication Workers of America (CWA), representing 630,000 workers in the IT-producing industries, joined AT&T, Lucent Technologies and the International Brotherhood of Electrical Workers in forming the “The Alliance”, a joint training trust which annually trains over 23,000 employees. CWA has also joined with Cisco Systems to provide hands-on-training for IT workers in sites around the nation.

We need to replicate the successful efforts of unions and employers, often government aided, to form regional and sectoral partnerships aimed at addressing workforce needs while promoting business objectives. in several different industry sectors.

Government, businesses, labor and communities must join together to help underrepresented groups - especially in low income and rural communities - overcome entry barriers into the IT labor markets. The AFL-CIO strongly supports government policies to ensure fair, affordable universal access to information technology and services. But bridging the “digital divide” between information “haves” and “have nots” requires more than putting computers and the Internet into schools and libraries, or even into people’s homes. It is equally essential that we give adults the skills they need to use information technology and compete for good jobs in the Internet economy. Achieving this goal, in the end, may be our most difficult and important challenge.

The AFL-CIO also has concerns about the future of low-wage employment. New technologies and innovations have and will continue to transform the economy and reshape jobs and the workplace. But this is only a partial view of the future. Receiving less publicity is the point that whenever major innovations take hold, income gains typically go to the better-educated workers who can quickly master the new techniques and technologies.

The truth is that low-wage employment is not going away in the 21st Century and that many of the ordinary, humdrum, low-skill, low-paid activities of the past will continue on as before in the years to come. In some places, low-wage employment may actually be leading gains in employment.

There were roughly 33.7 million low-wage jobs in the U.S. in 1998, roughly one in every four. These are jobs in occupations with median annual earnings that pay $ 16,800 or less a year. They are above the minimum wage, but just at the level it would take to lift a 4-person family out of poverty. By the year 2008, estimates are that there might be as many as 38.5 million such jobs, a 14% increase.

Most of the occupations are in the service sector, beginning with retail workers and cashiers. There is some agricultural, forestry and fishing work represented by farm workers, other related jobs and some production workers engaged in meat, poultry and fish cutting and trimming and cannery workers. A number of low-wage jobs are involved in the textile and garment industries, including such jobs as working at sewing machines and pressing machine operators and tenders.

About 14% of American workers are union members. The only fast-growing but low-wage occupations with higher union membership are teacher assistants and hand packers and packagers. Nursing aides, orderlies and attendants have 12% union membership, guards 13%. The other fast growing occupations have less union representation, most under 5%. The benefits of union membership will not be extended to many workers who need improved pay, benefits and working conditions unless union density is increased in these select occupations.

The importance of union membership for low-income workers can be seen by comparing the weekly earnings of union versus nonunion workers. Union cashiers, for example, may earn half again as much as nonunion cashiers. Union janitors may earn 60+ percent more than nonunion people doing the same work.

The above does not include the higher benefit levels that union members would probably have under a union contract such as, importantly, health insurance and provision for pension for example.

But are we really ready for the global, high-tech challenges of the new century? And will our new economy, which has been so slow to generate real economic gains for most of us, finally come around to meeting the needs of all working families? Such questions may best be answered by those who hold our nation’s future in their hands - the young people who are just now embarking on their work lives. And so we asked them, in a nationwide survey of workers between the ages of 18 and 34 commissioned by the AFL-CIO.

These young workers are willing to do their part to meet the changing demands of today’s economy. They are optimistic and confident, committed to working hard and upgrading their skills. They know lifetime jobs are a thing of the past, but they are hopeful that what the new economy lacks in security it will make up for in opportunities for advancement.

Unfortunately, young workers do not see employers meeting their end of the deal. Instead of fulfilling the prized American promise of upward mobility, they see employers failing to share the rewards for success with all who helped to create it. Corporations show too much concern for the bottom line and not enough for employees, they say, and employers fall short when it comes to sharing profits and investing in workers.

This helps explain why young workers are not particularly impressed with the nation’s current economic performance. They think the media has lost sight of what makes for a good economy. The true measures, they say, are not the stock market or corporate earnings. Rather, they are living standards and the availability of good jobs-and from where young workers are standing, it looks as though the new economy is creating mainly low-paying jobs that do not offer the health care and retirement security they need.

Young workers believe education is a key tool for survival in the new economy. But for all the talk about the importance of knowledge in the global economy, higher education remains largely unobtainable: Fully three-quarters of young workers today do not have college degrees. And young workers see fundamental inequities that education alone will not solve. They think it’s time for new rules to hold corporations to a higher standard in the way they treat their employees. They support policy changes to protect good jobs and help them meet the needs of their families, such as requiring employers to provide basic health and pension benefits, strengthening equal pay laws for women, expanding the Family and Medical Leave Act.

Increasingly, they see another way to make the new economy more fairly distribute the wealth, and that is by joining together with their co-workers in unions. Today, 54 percent of young workers say they would vote to form a union, up from 47 percent just three years ago. And today’s unions are changing and mobilizing to meet the new challenges, fighting for higher pay and rewards for good work, standing up for equal pay for women and people of color and improving access to education and health care and retirement security for all working families.

If working families are to thrive and prosper in the new economy, one of our priorities as a nation must be to restore the basic American freedom to choose to form a union.

Building career ladders is also a key strategy. Unions historically have spearheaded or supported programs that help workers advance their skills and careers through apprenticeship programs, community-based skills centers and creation of career ladders. Creating a career ladder also requires an employer willing to generate jobs with advancement potential and to think about developing the company’s workforce. Yet, many companies have downsized their labor force and outsourced work that once provided steppingstones to more advanced jobs, and employers are ill-prepared to offer nontraditional ladders to the growing number of women and minority workers. Some industries, such as printing and metalworking, are dominated by small firms with few internal career advancement opportunities and without clear career structures, workers may find changing jobs among firms makes it difficult to advance.

At the same time, workers face many barriers when seeking to continue their education-including parenting responsibilities, financial costs, child care arrangements, lack of transportation and scarcity of time. Without paid time off and financial support for advanced education, workers may well find the next rung of the career ladder out of reach.

Unfortunately, a reduction in funding for important training programs is planned. Despite the sharp increase in unemployment and economic hardship over the last year, the Administration’s budget makes significant cuts in critical worker training programs. Overall discretionary spending for these programs will be cut by 9%.

Additionally, the Administration has not effectively addressed the management of state and local block grants, which has led to an estimated $ 1.3 billion in carry-in into FY 2003. This carry-in will be used in place of increased funding for these programs, with the result being that those already denied service must wait alongside the newly unemployed for the meager resources the budget offers. Some states have questioned the accuracy of the Administration’s carry-in estimates, which could mean even fewer resources available to help workers during the next fiscal year.

Adult training currently funded at $ 950 million, the Administration budget would cut 5% or $ 50 million below 2002 appropriated levels. The result of this cut could be to deny services to at least 18.000 new participants, including many who have left welfare and are seeking help to enter or stay in the labor market.

The question of labour mobility is currently considered within the EU from three points of view. First, there is a desire, on political grounds, to have the European population taking full possession of their territory, in order to favour the formation of an European identity. Second, based simply on law, any European citizen should have equal access to jobs anywhere in the Union without being discriminated. In this respect, a full portability of his political and social security rights should be guaranteed.

Whatever important, and even self-sufficient, these first two arguments are, this paper will concentrate on the third argument: the labour market efficiency. First, we need to remember that “market efficiency” applied to the “labour market” cannot be defined, if at all, in the same way as for commodity markets or capital markets. When applying it to labour, we should limit ourselves to consider that efficiency is just about all the measures, at whatever level, that one might think of for “improving the situation”, i.e. reducing the bottlenecks or, for example, maximising the revenue or minimising the cost - depending on the point of view. This totally open content derives from the fact that the “market” itself is metaphorical when applied to “labour”. Therefore, we will only be looking at some specific aspects, namely geographical and job mobility on the external side of the labour market, as opposed to the internal side of the labour market, i.e. within the organisations and the work process.

The concern with geographical mobility within the EU, in the recent years, has emerged from the increasing awareness of the regional imbalances in the frame of the demographic downturn of the labour force. At one end, Denmark as a whole has not only a very high employment rate (76 % in 2000 for the 15-64 age group) that leaves hardly any unused labour capacity, but also a stable working age population - index 101.4 in 2008-2010 (100 in 2000), projected down to index 100.6 in 2020. Therefore, Denmark can hardly afford any growth in overall employment over the next two decades. At the other end, we have for example Ireland, where on one hand the employment rate is starting from a lower level (64.5 % in 2000) and where on the other hand the size of the working age population is projected to grow further (index 111 in 2010, index 115 in 2020). Therefore, Ireland, with a 1 % growth p.a. in employment, would reach an employment rate of 72.1 % by 2010. But Ireland, notwithstanding the most favourable demography in EU15, could not do much better: with 2 % p.a. growth in employment over the next decade, it would push the Irish employment rate above 80 %, i.e. above the reasonable limits. With 1 % and 2 % p.a. growth in overall employment, the EU15 would reach in 2010 an employment rate of 68 and 75 % respectively, starting from 63 % in the year 2000. Just with 1 %, it would reach the limit before 2020. The USA, starting in 2000 with a (civil) employment rate slightly above 70 %, would reach the
 
limit by 2025. Japan, facing a 30 % decrease of its working age population over the next half century, should also face a continuous decrease of employment. (See chart 1 and 2). This tells fairly enough how the European labour market faces prospects that are even tighter than they are in the US. This soon forthcoming labour scarcity in Europe will be adding pressure to improve drastically the productivity of a scarce resource. In the USA, the labour availability would be higher, unless the southern immigration was systematically used to widen the margins.

Maps 1-2-3 show the picture at regional level for EU15 and for the EU-pre-accession countries, with the same scenarios of 1 % and 2 % p.a. growth in overall employment. On one hand, in the Nordic Member States, in UK and in Portugal, 1 % over the present decade would be sufficient to determine a clear labour scarcity, because they are starting from high employment rates. On the other hand, southern Germany and northern Italy are starting from lower employment rates, but are to face dramatic decreases in the working age population. Therefore, even a moderate 1 % p.a. growth in employment would be enough to generalise the labour scarcity. In the 2 % p.a. scenario for the next decade, only (most of) French regions and north-eastern Spanish regions would keep in the area of 72-76 % employment rates, close to the maximum level, and only southern Spain and southern Italy would lag below a 65 % level in employment rates.

Maps 1-2-3

It tells also where the reserve pool is in the EU - not to mention qualification and ageing characteristics: out of an estimated “unused labour capacity” close to 20 millions in 2010, in the scenario of 1% p.a. growth in employment and considering a

maximum employment rate close of 75 %, one eighth would be in the most western regions of Germany, one quarter in France (north, Paris region and south-east), a third quarter in southern Italy, and the last quarter in (mostly southern) Spain. Another 6 million, under the same 1 % scenario, would be available in 8 pre-accession countries (Estonia, Czech Republic, Hungary, Latvia, Poland, Slovenia and Slovakia), of which 2/3rd in Poland, plus another 1.5 million in Bulgaria and Romania, besides additional reserves that would emerge from sectoral re-allocation.

The geographical mobility, in terms of “moving house”, seems to lie in EU15 at half the level in the USA, with 7.2 % of people involved in the year against 16.1 % (see Chart 3). Within EU15, a wide variety appears. The extended family area, i.e. the Mediterranean countries plus Ireland, show much lower figures compared to the Nordic states, that are much closer to the US figure.

Looking at the moving distance, there is a common pattern: slightly above half of the moves are local: 52 % of moving Europeans move in the same town, city or village, and 56 % of moving Americans do so in the same county (Chart 4). A second dimension that should be taken into account is related to the age structure: young people are notoriously the most mobile. Although based on different methodologies, the age profile illustrated by Chart 5 suggest that the young mobility is rather more driven by studying motives in the age group 15-19 for the EU than it is the case for the USA, where the drop in mobility rates becomes significant only above the age of 30. For the EU15, it is reasonable to consider that with the fast decreasing number of young people would rather depress the overall mobility rate.

When considering the European diversity, it is obvious that the mobility patterns are of different kinds. Three models seem relevant:

a. The Nordic model has high mobility rates. It has certainly much to do with the early leave of young people from the parental home, as is common with traditional matrilineal system that is prevailing in these countries. Chart 6 shows that the ages at which half of the young people leave lies below 20 or close to 20 in Finland and Denmark, with little differences between males and females. Chart 7 shows how mobility concentrates much more on young people in Finland and Sweden, compared to the EU average. Close to half of the overall mobility seems to involve the capital regions. A link should also be made with the individual-based social rights.

b. The “simple nuclear family” model prevails in the Netherlands, UK and France, with relatively high levels of mobility. A link should be made with the family-derived system of social rights. In the case of UK and certainly France, the state centralism and some degree of urban macro-cephalic is certainly adding to the mobility involving the capital region. The region “Ile-de-France”, for example, makes one fifth of the French population, but involves two fifths of the national inter-regional mobility.

c. A third group of countries shows lower figures of mobility. On one hand, we have the cases of Germany and Belgium are certainly to be related to their typical polycentric occupation of the territory, combined with family-derived social rights and relatively closed labour markets with more lifelong-like jobs. Austria is certainly a more specific case, combining macro-cephalic of the capital region, strong tracks of matrilineal behaviour keeping the youngster later at the parental home, and the highest share of lifelong-like jobs that make the people less mobile. On the other, there is the group currently referred as the “extended family area” that should rather be named the “prolonged family area”. It comprises most typically Italy, Spain, Greece, and also Portugal and Ireland. The remarkable thing is that it is confirming with time its characteristics: in Italy for example, the median age at which half the young men have left the parental home has shifted, between 1992 and 1999, from 28.2 years to 29.7, and from 25.1 to 27 years for young women. Levels and changes are hardly lower in Spain and Greece. And this, combined with the traditional localism, and a job system that owes much to the family network, id obviously discouraging mobility. Furthermore, this model reveals more and more its perverse effects, linked to the increasing privatisation of the cost of childhood versus the public mutualisation of the cost of the aged. By all means, the internal mobility in Italy represents certainly the strongest example of resistance to economic theory. In no other country are the regional disparities so high, here between north and south, and more irrelevant the argument that would make differential in unemployment feed the migratory moves. Instead, we have the northern regions showing an employment level close to 60 %, or the EU15 average, while the southern regions, according to Eurostat, go on showing employment rates in the area of 40%.

Whether these self-confirming behaviours are to define some cultural dimension is an open question. It would by all means be interesting to have a closer look at the extent to which these “cultural behaviours” could still be tracked in the USA depending on the national lineage.

On the whole, if we face the difficult task of estimating the global level of inter-Member States mobility, in terms of flow, we must remain very cautious. The data given by the Council of Europe, by the Labour Force Survey and by the Euro barometer are far from appearing similar, in terms of definition and of reliability, and henceforth in terms of levels attained. Chart 8 compares the three results, but the conclusion must remain vague: the number of EU15 nationals that are moving from one Member State to another, on an annual basis, should not exceed by much a quarter of a million people. With half of the EU15-nationals in other Member States having a job, this would mean that the intra-EU inter-Member State mobility would imply something between 1 and 2 per 1.000 jobholders. The very clear point in this is that this level is very far from reaching the critical mass that would imply any significant contribution to the labour market efficiency. Even if any policy would bring it to five times its current level, it would still involve below 1 in every 100 jobs.

There is little doubt that the mobility within the 15 Member States has gone down significantly between on one hand the post-war decades and on the other the last two decades.

Originally fed by agricultural reorganisation and national imbalances, the flow from the southern countries to the north has by now been reversed, with ageing southern migrants returning in their homeland. The number of Italians and Spaniards in France decreased by close to two thirds and three quarters respectively, which can be attributed only to a very limited extent, considering EU-nationals, to naturalisations. The number of Portuguese was peaking around the early 1980s, but they spread since then in other countries. The decrease in Germany involves only the Spaniards, whose number decreased by half, while the Italians and Greek show hardly any change, with the Portuguese peaking over the recent years.

As for the short-term changes, in reaction to the economic cycle, the main point is that they are little documented, at least for the EU. For the USA, the data from the US Census show hardly any change over the 1990s, with a share of movers declining from 17 % in the early 1990s to 16 % in the late 1990s: not enough to speculate on any slowdown of residential mobility in phases of strong economic growth. For the EU, the regional 5-year shift in the age structure, compared to the national averages, suggests that there was a decline of inter-regional mobility between the early 1990s and the late 1990s, in a clearer manner for the 25-

34 age group than for the 15-24 age group. But inasmuch the basic statistical material is reliable, this decline appears strongest in Spain, and still significant in Greece. But on the whole, the evidence is too shady to establish a clear link from improved economic situation to decreasing mobility.

Maps 4-5-6 show the number of people aged 20-29 per 100 persons aged 55-64, France, 2000-2010-2020. The source is INSEE 2001 Projections Démographiques (Scénario central), class limits: Blue to red 207-151, 150-121, 120-106, 105-96, 95-81, 80-65, 65 and below.

It is nevertheless strongly tempting to suppose that the far-reaching demographic changes over the next decades will rather depress the geographical mobility. The point is that the number of incomers aged 15-24 per 100 persons aged 55-64 leaving the working age will fall dramatically, from a EU15 average of 113 in 2000 to close 80 in the year 2020, before, seemingly, a further degradation. In the USA, the same ration would go down from 156 in 2000 to 96 in 2020, before restoring above 100 in the decades after. Some EU Member States would show an exceptionally fast degradation over the next two decades, like Spain, joining Germany, Austria, Italy and Belgium with ratios below 80. Only Denmark, Ireland, Portugal and Sweden would preserve a replacement ratio close to parity. The question is then of course the following: will the young people not be stuck down while they would be given much wider opportunity? And would it not depress the overall geographical mobility further and further down.

The case of France is interesting; showing how the past trends in mobility, when integrated into the latest demographic projections as they are by the INSEE, would produce strongly perverse effects. Map 4-5-6 shows how these projections end up in 2020 with the sole region of

Lyon and Grenoble, showing replacement Paris, plus the departments of Toulouse, levels above parity. Well over half of the departments would by then show a ratio below 8 incomers for every 10 persons of leaving age. Can we imagine that things could do that far without provoking some preventive slowdown in mobility? And if things were going that far, the problem would be passed over from DG Employment to DG Regulation.

The share of moving people declaring they did so for job-related reasons is similar in the EU and in the USA: 15.2 % and 17 % respectively. This bring the people moving for job-related reasons, at whatever short or long distance, to the equivalent of 2 % of the Labour force in EU15 (approx. 3.5 million out of 170 millions), as against 4 % of the Labour force in the USA (5.6 millions out of 140 millions).

Beyond the statement of such values, it is not easy to comment the extent to which they contribute more or less to improving the labour market efficiency. At this stage, we should only refer to the difference, and take notice, once more, of the “European diversity”. Chart 10 shows how the reasons given for moving do interfere in the 15 Member States. It suggests, again, that the Nordic distribution is similar to the US pattern, while in the “prolonged family” area, they are the least significant. Maybe it might be noticed that the “close labour markets” in Austria and Germany would lead to a higher share of job-driven mobility to compensate, but the argument would not mean much for France, and on the whole, it remains rather unclear. And we are certainly close to the limits of sample reliability and probabilistic laws, as we are whenever we are chasing rare events.

The US Census data provide some more detailed insight on the characteristics of the people who moved in the last year. Globally, it must be noted that the share of people moving is rather decreasing with the improving social status. Chart 11 illustrates this by choosing the household income. Looking at the shares per educational attainment is confirming this. Also, unemployed show higher share of moving people (25.2 %) than employed people (17.6 %). Expectedly, people living in owner-occupied units show much lower shares (9.1%) than people living in renter-occupied units (32.5 %). Considering occupational groups, we find the highest shares for “handlers/cleaners/­laborers” and for “private household service”. All this suggests also that some forms of nomadic behaviour do persist for the least favoured class. A fine analysis of Euro barometer indicates a similar link, but very much weakened, just at the limit of statistical reliability.

At the same time, some signs do appear, for example when looking at the educational attainment, that mobility rates can increase again, although slightly, with the highest levels of educational attainments, at least for graduates in their thirties, as illustrated by Chart 12. This is probably the area where the behaviours correspond best to the paradigm of mobility as part of some social ascending strategy.

This could be connected to some extent to the situation of EU-nationals that are expatriate in other Member States: according to the LFS 2000, the rate of expatriation to other Member States is higher for people with high educational attainment from the northern countries, as opposed for example to expatriates from Portugal, Greece or Spain, where the people with lower or medium educational attainment are - or were - more “on the move” than people with high educational levels. To put it in a rough wording, Swedish expatriated in other Member States (or Americans in the EU) are managers, and expatriate Portuguese are construction workers. (See Table 2).

Commuting to hold your job in another region appears certainly as a substitute for residential mobility, but the statistical recognition of it depends a lot on the territorial division and on the transport facilities. Map 7 and 8 show for some data for the EU. For the seven member States for which the LFS2000 gives data, as shown on Maps 7-8, the total number of inter-regional commuters would amount to close to 6 million. Obviously, a NUTS2 region like Brussels, with over half of the jobs held by commuters from other NUTS2 regions, is a most extreme example, and Belgium ranks first by far with close to 1/5th of commuters in the national labour force. Austria comes second, with 1/8th of commuters - the influence of Vienna. Germany comes third, with 8 %. At the other extreme, Spain with its vast areas of low-density population is discouraging commuting, that involves hardly 1 % of the national Labour force. The Netherlands show an example of fast progression in commuting, multiplied by close to 2 between the early 1980 and the late 1990s, while residential mobility remained at low levels. In UK, the recent division of Greater London into two new NUTS 2 regions would certainly add more millions to the number of formal commuters.

Maps 7-8 show commuting between NUTS2 regions in EU - 2000. Trans-border commuting needs a special mention, and certainly considering the work that the EURES network is doing. It involves, when including the commuters going to Switzerland, close to 0.25 % of the EU15 Labour force. But Switzerland is taking 40 % of the total number, and an additional 1/5th, if not 1/4er according to the latest data, is going to Luxembourg, where the commuters are now making well over one third of the total number of jobs, besides one third of the resident population being non-nationals. But Luxembourg is certainly a very special case. (See Table 3).

There is little doubt that the total number of commuters, in the EU15, is a multiple of the number of people migrating from one region to another for work-related reasons.

Any data on occupational mobility appears absolutely overwhelming compared to geographical mobility, and certainly if one isolates the small part of residential mobility that is due to job-related reasons. We are no more discussing about 2% (EU15) or 4% (USA) of the Labour force, but of shares of much wider magnitude. The immediate lesson from this is that the fluidity, or say the potential for re-allocation of the workforce, relies incomparably more on occupational mobility than on geographical mobility.

Chart 13 tells that 18 % of the EU employees held their job for less than one year, against 27 % of US employees. At the other extreme, 40 % of EU employees had their job for 10 years or more, against 26 % of US employees.

The global profile is sketching a dual labour market, with on one hand short-tenure workers, and on the other stable workers, having a job tenure above 3 years. It also suggests that this dual character is more pronounced in the US compared to EU. Noteworthy, this EU-US gap remains for every age group, as illustrated in Chart 14. In the EU, half of the employees in the 35-44 age group reach a 10+-year tenure, while there are only 31 % reaching this in the same age group in the USA. Also, when considering the ageing workers (55-64 age group), we find 18 % of American employees, against 12 % of European employees, with a job-tenure shorter than 3 years. This means also that re-entry into employment is certainly easier for Americans in their fifties than it is for Europeans.

Consequently, the median job-tenure of employees lies in the EU15 at twice the value of the USA: 7.1 years against 3.5 years. But one very striking feature appears when comparing the median tenure depending on age and on the educational attainment. In the USA, any higher educational attainment provides no advantage in terms of job security, while in the EU15, this link appears very clearly: the difference between the median tenure of people with high educational attainments, in the 55-64 age group, against people with low attainments is 6 years (23 against 17 years), while it does not even reach one year in the US (10.9 against 10 years). (See chart 15). And if this were, speculatively, to crosscheck some organisational characteristic, it would be tempting to say that baronies are much better locked in the EU compared to the US. Alternatively, it could be argued, from the point of view of the employee, that investing in higher education brings about better job security.

The second striking thing is, again, the European diversity. But there is one case where the profile is almost the copy of the US profile, and that is UK (cha