The Postindustrial Welfare State

Gøsta Esping-Andersen’s The Three Worlds of Welfare Capitalism is among the most influential works in the study of welfare states. Rather than conceiving of welfare and industrial policy on a single state-market axis, Three Worlds develops a typology to situate welfare states within broad and complex historical trajectories. In Esping-Andersen’s framework, modern capitalist states with systems of social provision developed along three general paths. Social democratic regimes like those found in Scandinavia emerged through a political coalition between industrial and agricultural workers, and are characterized by universal benefit schemes. By contrast, conservative regimes in Germany and France were born of coalitions between the left and the Church, and are characterized by fairly generous welfare provisions whose distribution is dependent on traditional family structures. Finally, liberal regimes like Britain and Ireland are ones in which the labor movement was unable to form meaningful political coalitions. The mark of these states are their limited, means-tested benefits typically available to only the very poor.

Three Worlds is emblematic of the “power resource theory” tradition developed by Esping-Andersen and colleagues like Walter Korpi. Unlike their counterparts in the Varieties of Capitalism school, who tend to view social safety nets as the product of high value-added economies in which employers aim to foster skill development among workers, power resource theorists hold that welfare systems are primarily explainable as the product of labor’s ability to organize against profit maximizing firms. Central to this view of welfare state development is the concept of decommodification, a qualitative and quantitative measure of the degree to which basic human needs are protected from market fluctuations. Just as central is the Polanyian notion of double movement—the dialectical process through which workers organize against the market to decommodify their labor.

In recent years, Esping-Andersen’s work has examined how institutional and political conditions have determined state responses to deindustrialization, reconceptualized class structures, and, in particular, analyzed how these changes have shaped demographics and family structure in welfare states. We spoke with Esping-Andersen last fall.

An Interview with Gøsta Esping-Andersen

Maya Adereth: Can we start by discussing your theoretical influences?

Gøsta Esping-Andersen: By the 1970s, when I began researching in the social sciences, the emerging stratification patterns in Western Europe and the US were becoming very clear. In my view, they demanded a new conceptual framework that would allow for stratified class composition and alternative social dynamics. Many friends, like Erik Olin Wright, were working within a Marxist framework. I had read quite a lot of Marx, but I was looking for something to help think through these emerging patterns.

This is where Karl Polanyi came in. The first person who suggested I read Polanyi was, as I recall, Fred Block. When I started reading it, I was very impressed with the subtlety and dynamism of his account of capitalist development—particular his idea of double movement, and the ongoing resistance to the naked market domination of society. I grew more and more interested in the movement against commodification on the side of labor, and I also grew interested in the idea that free markets are planned and developed through the power of states. This was my entry point into thinking about market failures, and my first focus was on education. A market driven education system would produce huge deficits in skill development and production because the cost of education would exclude 80% of the population. This could only be rectified with social policies.

While looking at the development of distinct social policies in various welfare states, I came upon the idea that the counter-movement is shaped by social context. In particular, it’s shaped by the interests and coalitions formed by different classes from the 1800s until the postwar period, and still to this day. This is where Three Worlds began, but the project itself was empirical rather than theoretical. At the time, existing literature conceptualized welfare states through a single indicator: social spending as a percentage of GDP. It had nothing to say about the structural differences between societies. As I compiled data on how these social security systems were developed and how they worked, I came to understand the distinct decommodifying logics which guided the system in each instance. This was rooted in distinct histories of class development.

What was unique about Scandinavia was the fact that the emerging working class faced similar life risks as peasants, and they consequently found common ground in fighting for a welfare state. Employer-based pension schemes made little sense in an industrial-rural coalition, because it would exclude rural workers. If you look back at discussions in Scandinavia, and Sweden in particular, during the very early decades of the 20th century, working class movements and the Social Democrats were in fact moving towards a contributory pension scheme. The Universal People’s Pension was adopted instead only due to the importance of this rural-industrial coalition. This logic of universalism which is characteristic of Scandinavia hasn’t really been instituted anywhere else. In places like the UK, you saw the development of minimal social assistance, similar to that of the US.

ma: Jumping forward a bit, the liberalization and breakdown of different welfare models is sometimes attributed to the loosening of labor markets with the introduction of women into the laborforce. What do you make of the relationship between these two factors?

gea: I don’t believe there is a causal relationship between the rise of women’s employment and welfare state deterioration. The changing role of women has promoted a shift of focus towards programs concerned with care work.

I try to theorize these dynamics through multiple equilibria modeling. To me, the most interesting thing to understand is the transition from the traditional role of women as primarily dedicated to household production, to fully integrated members of the labor force. The traditional equilibrium was self-perpetuating as long as it was normatively dominant or hegemonic. What fractured this equilibrium? It had to have been a series of exogenous factors, from birth control to household technologies.

This transition was most common among educated women. Across societies, it’s either the highly educated women, or the very poor women who work. So the question is, what were the dynamics once these exogenous factors promoted entry into the market? In the transition from the old equilibrium to this new one, we see once again widely diverging welfare state responses. For the labor supply of women to converge with that of men, the welfare state requires childcare, healthcare, accessible education, and so on. Women’s participation in the labor market is very closely linked to welfare provision.

This is why different societies exhibit polarized patterns of female employment. But we should be very careful not to be deterministic here. It is not certain that new equilibria will ever really consolidate. The US is the best example of that so far. It was the avant-garde in terms of the feminist movement for women to join the workforce, but it stopped along the way. The empirical evidence is very clear: about 25% of American women still remain normatively dedicated to the household and not to work.

ma: One thing that struck me about your data is that the US fertility rate declined sharply, but then recovered at a faster rate than Spain and Italy. Why is that?

gea: During the first phase of women’s entrance into the labor market, all countries experience declining fertility. But both the US and Scandinavia recover. By contrast, continental Europe—Italy, Spain, Germany—didn’t recover.

One of the key variables is the availability of affordable and high quality childcare. I’ve done a lot of work on that aspect. If it is not high-quality, mothers won’t want to use it. That’s why in countries like Spain, where you do have 30-40% coverage of low quality childcare, you don’t see a recovery. Whereas Scandinavia, in particular Denmark, is a key example of high quality early childcare. That’s why you have universal female participation throughout life. Part-time work in Scandinavia has become a totally new phenomenon compared to what it was and what it still is in Holland, Germany, or Britain. Part time work in Denmark is now primarily sought to transition between childcare and employment. So you have about 25% part-time workers in Denmark, but a large part of that is simply women who are transitioning back into the labor market for a couple of months.

That is not the case in Holland or in Britain. There, if women are in the labor market, their primary role is still with one foot in the labor market and one foot in the family. That seems to become a stable alternative that reproduces itself from cohort to cohort, certainly in Germany, certainly in the Netherlands, and certainly also in Britain. It may also be that among the competing alternative equilibria of women’s new rules, you have a part-timer model. That then begs the question, how might that rupture? Maybe high quality universal childcare will make a difference.

Germany is probably going to be our key test case. Germany, after decades, has finally decided to mount, relatively rapidly, relatively high quality childcare for all women. In the next five years, we will be able to use Germany as a test case: does universal access to childcare move women from an equilibrium of part timer-ism into the full time role?

One thing I would like to argue explicitly is that families are not, as Lesthaeghe and his followers argue, disappearing. They emphasize the importance of postmodern values and individualism as a feature of postindustrialism. If we look at trends in the US and Europe, let’s say from the 70s up to the end of last century, all the indicators supported this thesis. Virtually all family demographers or family sociologists bought into the second demographic transition model of family dynamics.

But I strongly believe this is wrong, and I demonstrate that people’s partnership and child preferences have not changed much in the last four generations. Namely, the two child norm has not changed. There are some countries where it has declined a bit, Germany being one. But otherwise, it has been incredibly stable. We do see the emergence of a new kind of family. Families in the 21st century will not look like those of the 1950s. But they will probably become equally stable once we settle into the new equilibrium. I’ve done a lot of work with time use data. In Denmark, men’s contribution to total housework is about 43% on average. In Sweden, the women’s contribution to unpaid work is only about 20% higher than men’s. In Spain, it’s 100% percent more than men’s. That’s where the major difference is: in the distribution of that work, not necessarily in the composition of the family.

ma: How has increasing women’s labor market participation changed your categorical understanding of welfare states? Is the Three Worlds model still intact?

gea: Today we see movements in all kinds of directions. In Scandinavia, conservative parties are moving towards a Bismarckian logic by adding an insurance component to the flat rate universal benefit system. In some conservative welfare states, you have a movement in the directly of universalism. Germany’s childcare policy is one example of that. Obviously, that also has a lot to do with the changing coalitions that are driving social policy debates and development.

ma: To elaborate on that, I wonder what you think about the transition to so-called service economies and the political coalitions they generate. What are the fault lines along which welfare politics takes place today?

gea: I think education is clearly the most visible fault line. It’s also the most powerful engine of political polarization, partially due to the emergence of the dual earner couple. What we’ve seen is that highly educated and high earning couples marry one another: two doctors for example, or an architect and an engineer. On the other end, you have more traditional families who also have lower incomes. Labor market and education disparities multiply. This is probably the most dangerous source of polarization in our societies today.