Available in PDF format  



Gösta Esping-Andersen

In recent decades, the economies, societies and demography of Europe have undergone revolutionary change. Through it all, our welfare states have remained pretty much the same. But there is mounting pressure for a major overhaul of social policy, primarily for two reasons: the impact of societal transformation has been unevenly distributed; and citizens are facing new kinds of risks and needs. The great challenge now is to design a welfare-state architecture which is appropriate to post-industrial society. It is a challenge that can be analyzed under three main headings: women, class and chronos (i.e. relating to time).

Women and the gender divide

As many feminists have pointed out, the architecture of the post-World War II welfare state was biased in favour of men. This is hardly surprising, given that men dominated the socio-economic order of that period, and large numbers of women withdrew from the labour market. The typical production worker was a young male with modest education who enjoyed a reasonable level of job security, belonged to a labour union, and was well-paid.

Post-war welfare capitalism was a working-class society in which households enjoyed increasing affluence as a result of full employment and steady wage growth. It was sustained by marital stability, high fertility, and the contributions of traditional housewives. The resulting dependence of family members on the male breadwinner led to a policy with three main components: job security for the breadwinner, assistance to large families, and better pension guarantees.

In post-industrial society, all aspects of life are changing, and women are in the vanguard of that process. The role of women in society is being transformed, with the following implications:
  • demographic change, including fewer births, ageing populations, new family patterns and increasing rates of divorce
  • new household and family structures, with an increase in single persons, lone parents, dual-career couples and family instability
  • the growing dominance of a service economy in which families ”out-source” their needs; this in turn affects job-skill requirements (social skills become more important, for example).

As can be seen in Table 1, the post-industrial labour market is overwhelmingly dominated by women, just as the old industrial economy was and still is a male preserve. It is also worth noting that job-skill requirements are becoming more demanding. These trends raise a number of questions: Is post-industrial society prepared for the fact that its driving force is provided by women? What does this entail? Will the expanding role of women lead to good political solutions in post-industrial society?

Table 1. Women create post-industrial society

Formula for change

Figure 1 presents a formalized model of post-industrial society’s basic dynamics. The first expression (a - 1) is essentially a rendition of Engel’s Law: As we become wealthier, our demands shift from basic needs to services and non-essentials.

The second expression (added by myself) argues that the dynamics of the service sector depend on income distribution. A highly unequal distribution (e.g. of Latin American type) means that a tiny elite consumes the greater portion of services, which are provided by masses who are too impoverished to share in them. The United States may be heading in this direction. At the opposite extreme, a highly egalitarian society like that of Sweden may suppress low-end services because wages are so highly compressed; without large income differentials, there are relatively few maids, busboys, etc.

The third expression is popularly referred to as the ”Baumol cost disease”. Since the productivity of many services increases more slowly than in manufacturing, those services will be priced out of the market. In Stockholm, for example, there are virtually no cleaning establishments; in New York, there is one on every street corner.

The final expression ( D ) relates to the revolution in the roles of women. Regardless of incomes and prices, the new household needs services because the traditional housewife is disappearing. As she does so, families have to ”externalize” their service needs.

Double catalysts

As the roles of women change, so do labour-market dynamics. Industrial employment increased because households demanded refrigerators. Post-industrial jobs are on the rise because households now require services from outside providers. The less that families provide their own services, the more jobs in the labour market, primarily for women. Thus, women are double catalysts of change.

The limits of the industrial model are related to globalization: Workers in the Third World produce commodities such as cars and stereos more cheaply than workers in the developed world. But the barriers to post-industrial change are less global and more ”domestic”, because they are related to households’ service needs and purchasing power.

To ensure dynamic post-industrial development, it is necessary to rethink existing systems for the distribution of society’s resources. One alternative that would encourage growth of personal and social services implies a deregulated labour market. This is the low-wage model of the U.S. and England. But it also implies great inequality between households.

Double dilemma

The Anglo-American model poses a double dilemma. First, a large segment of the population has too little purchasing power. Second, income inequalities are easily reinforced and can result in a polarization of society. Also, marital homogamy is the norm, which means that powerful men marry women of similar social status, resulting in huge household incomes.

Poorly educated couples with low incomes represent the other extreme. Their situation may be further aggravated by costs associated with the female partner’s employment, such as day care. Such costs comprise an ”implicit tax” which, as a proportion of income, can be very high. Consequently, a deregulated post-industrial labour market could result in sharper class divisions.

An egalitarian, redistributive system like the Nordic model reduces the cost of caring services by means of subsidizes. This encourages female employment, and is especially important for low-income households. The ancillary benefits include less dramatic class divisions, and a superior capacity to deal with the ”chronos problem” (see below).

A strategy of deregulation would be inexpensive for the state, but may become very costly in post-industrial society. The egalitarian strategy necessitates heavy taxation. But public expenditures should be defined as investments that can be recovered through the dynamics of rising female productivity, due mainly to the cumulative lifetime earnings of women.

Not everyone agrees. Sherwin Rosen, for example, has argued that Sweden’s ”women-friendly” policies are uneconomical because the productivity of mothers with small children is less than the cost of the subsidies they receive. But this is an overly static view. My own research has shown that, within a dynamic perspective of lifetime earnings, the subsidies received by mothers of young children are more than amply recovered as a result of their superior lifetime earnings and tax payments.

Independent women

The ”cost” of children varies in relation to the revolution in women’s roles. It has been suggested that modern women forsake children because they do not want them. This is not true. Surveys indicate that young Swedish women (and men) want an average of 2.4 children, as do young Italians and Spaniards for the most part. But they want children on different terms than those which previously applied. The post-war welfare state assumed that families would ”internalize” the cost of children, i.e. that they would provide the necessary services, themselves. But that is no longer the case.

The current situation is very similar to that described by Alva and Gunnar Myrdal in their writings on the fertility-work dilemma of women in the 1930s. Since working-class women were at that time compelled to work, the priority for the Myrdals was ”to ensure that working women are also able to have children”.

Subsequently, during the ”golden age” of post-war capitalism, husbands did very well financially, so that their wives were able to abandon industrial labour for the role of housewife and mother to several children. As a consequence, the new challenge was to ”ensure that mothers are also be able to work”. In the emerging post-industrial society, we see a return to the original problem described by the Myrdals: Women are now demanding economic independence and find it difficult to fit children into their lives.

Costs and benefits

There are costs and benefits associated with children, both private and societal. The private cost consists primarily of losses to the mother of wages and career opportunities, a penalty that affects total household welfare. The wage penalty is higher for the first than the for the second child, then becomes very high for the third. In most countries, the penalty is lower for highly-educated women, because their connection to the labour market is weaker. In Sweden, however, the reverse is true.

The fact that female employment rates are so high in the Nordic countries is due partly to wage equality, which provides a strong stimulus for poorly educated women to work. Conversely, a low-wage strategy might have the opposite effect, thus reducing the labour supply of less educated women.

The private cost could be reduced or eliminated with a ”woman-friendly” package of child allowances, generous leave schemes, and affordable day care.
The private benefit is simply that the desire for children is fulfilled. A clear example of inadequate family welfare is provided by Italy, where women have only half the desired number of children. The negative welfare effects of the 1990s’ economic crisis in Sweden are reflected in the fact that fertility dropped to two-thirds of the desired level during that time.

Essential investment

With regard to the costs of children for society as a whole, the post-industrial era is revolutionary. Not only are more children needed to bridge the coming demographic age imbalance and contribute to future economic vitality, the future post-industrial society also needs strong, resourceful and productive adults more urgently than ever. Since nurturing such adults is a function of childhood welfare, society can invest in the well-being of the elderly while reducing the risk of future class polarization by assuming the costs of children.

As a rule of thumb, childhood poverty results in two years’ less education, thirty percent less lifetime earnings, and twice the risk of unemployment. In addition, poverty in childhood tends to be passed on from one generation to the next.

No society, not even Sweden, can afford to cut back on investments in children. Unfortunately, this is what occurred during the 1990s-- due not so much to cuts in public expenditures, but rather to a failure to stay abreast of changes taking place. The principal losers during the economic crisis were immigrants and families with children, especially those headed by lone mothers. Favourable post-industrial development will probably require greater investments in children than at present.

A child- and women-friendly policy is essential to positive welfare development in the future. A sound post-industrial welfare policy is necessary, not only for the sake of women, but for the sake of society as a whole.

Chronos and the generational divide

Many fear that we are heading toward a clash of generations, that chronos will become the most serious new social cleavage. But the current debate on ageing and pension reform obscures key issues. In a nutshell, a good post-industrial retirement policy begins with children.

The invention of retirement for all and the relative generosity of post-war pension systems derived their logic from the situation of the elderly in the 1950s and 1960s. These were age cohorts of widespread misfortune and dire need. Their lives had spanned periods of inflation, depression, war, unemployment, and low earnings. Having accumulated little wealth and meagre pensions, they retired into poverty.

The experience of those retiring today has been quite different. They came of age and built careers during the golden age of full employment, growth in real wages, improved health, and generous pay-as-you-go pension systems. Their asset accumulation is substantial, corresponding to 3-4 times the value of annual pension income; and their retirement income is usually more than adequate, averaging 80-90 percent of median income throughout Europe. There is no doubt that pension levels in many countries have overshot the original goals; and in some countries, child welfare is being crowded out by fiscal commitments to the elderly.

At the risk of exaggerating, I believe that welfare states have always managed to adopt the wrong pension policy. Benefits were too miserly for the disadvantaged generation between the two world wars. To compensate for this, governments introduced pensions for the fortunate post-WW II cohorts which were probably too generous. Now, some governments are planning to cut back on pensions, even though coming generations may be as disadvantaged as their ancestors during the 1920s and ’30s.

Broad consensus

Excess income among the elderly can result in perverse distribution outcomes. Pensioners with excess income pass it down to their children and grandchildren within the private sphere. Yet, it is based largely on societal programmes and benefits: In pay-as-you-go systems, the major portion of pensioners’ income is furnished collectively by working adults.

With a retirement policy based on investments in children, the first priority would be to guarantees equity and adequate welfare to future pensioners. I believe there is now a broad consensus that such a policy implies the need to:
  • maximize employment rates among the active population, including women and immigrants
  • delay retirement by abolishing the compulsory age and providing incentives; this would be equitable and efficient, given the rapid and significant improvements in health and education for every new retirement cohort
  • invest in today’s children so that they are better equipped to finance the coming pension burden with their productivity
  • maintain strong pension guarantees in the future.

The two last points require some elaboration. Today’s pensioners may be well-off because they benefited from favourable conditions during most of their lives. But it is very uncertain as to whether today’s children will enjoy equally favourable condi-tions during the 50 years ahead The more we invest in the capacities of today’s children, the better they will be able to sustain tomorrow’s elderly population, and the more likely that they will also enjoy ”good lives” during the retirement years.

What the current pension debate obscures is that the fate of cohorts is linked to the conditions of specific periods. Resolving the problem of today’s excess pension income by scaling down the level of benefits may have a terrible boomerang effect in the future, if the life experience of those who retire in 2040 or 2050 resembles the experience of those who retired in 1950. That is likely to be the case if we do not begin our pension reform by investing in children,

In short, post-industrial pension policy should place a greater emphasis on the needs of today’s children than on the problems of today’s elderly.

Cognitive skills and life chances

European social democracy has abandoned its mission as a working class movement. It also risks leaving behind its previous commitment to universal coverage as, increasingly, it caters to more specific interests. It has, for example, been in the vanguard of initiatives on behalf of women and the aged.

But with a suitable post-industrial welfare architecture, neither gender nor generational divisions need result in polarization or winner-loser conflicts. However, other forces may produce such effects. Social democracy will be very ill-prepared for the coming decades if it complacently assumes that the problem of class has been eliminated for all eternity.

Social scientists have sketched a variety of future scenarios with regard to class. Daniel Bell has presented an optimistic version in which a knowledge-based economy simply annuls class, because rewards are based on skills and talent. Others believe that such an emphasis would result in a polarized society with a nucleus of winners and a mass of low-wage proletarians. Still others, for example Van Parijs, fear the development of an excluded, unemployed mass in a society of privileged insiders and disenfranchised outsiders.

But all agree on one fundamental point, namely that education and cognitive skills will largely determine life chances and form the primary basis of social inequality. This suggests four probable features of post-industrial society.

First, service jobs in general will require a degree of skill. But post-industrial households will create a large category of low-level jobs; an educated guess is that this will involve 15 percent of all jobs, considerably less than those in conventional blue-collar occupations. The challenge here is not to prevent the growth of this sector, or to find a suitable way to supplement low wages. The real issue for class equality is how to prevent long-term entrapment in the ”Macjobs economy”.

Risk of social exclusion

Second, education and cognitive skills will increasingly be the keys to favourable life chances. This applies not only to the level of skills, but also to the kinds of skills that are required. Social skills and ”social-cultural capital” will gain in importance (grunting steelworkers are hard to recycle in the labour market). In particular, it is now fairly evident that life chances will depend on a strong foundation of cognitive skills. Highly developed but limited specialization will not be sufficient to survive the dynamics of change: A broad base of cognitive skills will be necessary.

Depending on how investments in cognitive skills are made, it is possible to envision two rough post-industrial scenarios: islands of knowledge in a sea of ignorance (the U.S. trajectory); or a shallow lake with few ripples. The first may produce marvellous progress in some regards, but it will likely come to an abrupt halt if the masses are not able to consume the marvels. In the long run, the second alternative will win in terms of efficiency and human welfare.

Third, even with full employment, it is not unthinkable that post-industrial society will include a number of unemployed ”surplus” citizens. The competitive demands of the new order will punish the completely resourceless. In the jargon of the European Union, there is a problem of social exclusion, and it is clearly manifested in the division between ”work-poor” and ”work-rich” households.

But the excluded are not homogenous. It is necessary to distinguish between ”soft” and ”hard” categories. Workless households comprise 8-13 percent of the European total; the proportion is lowest in Scandinavia, highest in Belgium, Britain and France. Somewhere around half consist of poorly educated single males or are headed by individuals with chronic illness. This ”hard core” is not likely to be helped by activation programmes. Those in the ”soft” category are not totally excluded; they are often headed by lone mothers, and can be helped by activation programmes.

It is symptomatic that, in Great Britain, work-poor households have not declined at the same pace as declining unemployment. It will therefore be necessary to maintain basic welfare guarantees in the future, just as today. The size of the so-called surplus population can be limited to the hard core if greater investments are made in children. Inadequate schooling is the factor that is most clearly associated with socio-economic exclusion

Fundamental requirement

The fourth likely feature of post-industrial society is that homogamy (i.e. marriage between socio-economic equals) will tend to reinforce inequalities between households, as more and more women enter the labour market, But this tendency can be counteracted by wage policy and/or transfer payments. Investing in the employment capabilities of poor mothers may be the key. Still, homogamy may decline due to rising educational differences between women and men; a possible consequence could be that increasing numbers of women will have no real marriage market.

The severity of future class cleavages will depend very much on the transmission of welfare from one generation to the next. A policy based on educating the entire population will never work unless it can guarantee that everybody has access to an equal education.

The most fundamental requirement is that children not grow up in poverty; the evidence is exceedingly clear on this point. Furthermore, this would not require massive investments. Since Nordic poverty rates are already close to zero, the additional cost would be minute-- for Sweden, around 1/10,000 of one percent of Gross Domestic Product. At the other extreme, the spending requirements would be more substantial-- for the United States, over 3/10 of one percent of GDP. But in every case, the cost of eliminating child poverty would be well under one percent (see Table 2).

Table 2
Cost of completely eliminating child poverty,
as percentage of GDP and social expenditures


*Thousands of households-with-children below poverty level,
defined as fifty percent of median disposable income (adjusted).
Sources: LIS Databases and OECD National Accounts, 1990-97

A related requirement is to reduce the impact of parents’ resources on children’s life chances. This requires far more broad-based efforts to redistribute societal resources so that, for example, all parents have books, read with their kids, and have adequate social capital.

The Nordic countries are genuine success stories in these respects, as evidenced by comparative research on literacy levels and social mobility. Whereas 15 - 20 percent of U.S. citizens fall into the lowest category of cognitive skills (they are essentially dysfunctional), only 4 - 5 percent do so in Denmark and Sweden. The high level of literacy among Nordic youth can be clearly traced to the reduction of inequalities among their parents (see Table 3).

Table 3. Impact of parents’ education on
children’s literacy (regression coefficients)

Effect of parental education level on children's cognitive ability.
The higher the number, the stronger the "inheritance" effect.
The first two columns compare the impact on offspring who
are now under 25 years old with those who are now over 25.
A plus change ( + ) means that the parental impact on the
younger group is greater; a minus change ( – )
means that the impact has declined.

”Folkhem” for the future

The title of this seminar is ”The Nordic Alternative” and, appropriately enough, it has been arranged by the Olof Palme International Center. I associate both the theme and the sponsor with a social democratic vision of the future. Alas, the kind of overarching vision that Nordic Social Democrats had in the days of Gustav Möller, Ernst Wigforss and, more recently, Olof Palme, has lately been obscured, in Scandinavia and everywhere else.

Although a mere academic, I have learned two important lessons from my studies. One is that it is relatively easy to be a social architect like Gustav Möller in a society that is well-defined, as in the case of the youthful, male-dominated working class industrial society of his time. It is more difficult within the context of the just emerging and somewhat diffuse post-industrial society of today.

The second important lesson is that social democratic policy succeeds best when its solutions to the problems of some particular group benefit all other groups, as well. That was the genius of the ”people’s home” (folkhem) of Möller and his associates. It is also why pension policy should start with children, and why women-friendly policies should be people-friendly policies, in general.

The great challenge for Nordic social democracy today is how to construct a folkhem for the post-industrial society of tomorrow.

* * *

Gösta Esping-Andersen is currently Professor of Sociology at the University of Pompeu Fabra in Barcelona. Widely regarded as one of the world’s foremost authorities on the modern welfare state, his published works include: Politics against Markets; Three Worlds of Welfare Capitalism; Welfare States in Transition; and Social Foundations of Post-industrial Economies. In addition to his teaching and research, he has been involved in a variety of policy-related activities with the World Bank, the OECD, the United Nations and the European Union.