It is not only sociologists who care about social mobility. From novels and films that chart the rise and fall of individuals and families to everyday expressions such as “from rags to riches” and “following in father’s footsteps,” interest in social mobility is widespread. And, in recent years, it has risen to the top of the policy agenda. Rates of mobility lower than had been thought and apparently declining rates of mobility have helped to drive these concerns. The solution is often thought to lie in education. Securing a good education is widely seen as the key to improving an individual’s mobility chances, while governments promote reforms of education as a means of equalizing opportunities for mobility among people from different social backgrounds. This book is about the role of education in shaping rates and patterns of intergenerational social mobility among men and women during the twentieth century.
For most of human history, mobility must have been rare. With a relatively simple division of labour, there were few occupations between which people could move. Children of hunter-gatherers became hunter-gatherers themselves; children of peasant labourers grew into peasant labourers. In the vastly more complex societies that have emerged since the Industrial Revolution, mobility has been common, thanks to more differentiated occupational structures and periods of rapid structural change. During the nineteenth and twentieth centuries, in what we now call the developed countries of Europe and North America, there were massive shifts in employment away from agriculture and towards manufacturing and, later, to service and white-collar jobs. These changes greatly reduced the likelihood of people remaining in the same occupation or class as their parents. At much the same time, but particularly in the middle years of the twentieth century, educational systems were expanded, labour markets were reformed, and welfare state provisions were introduced and expanded: all these should have reduced the dependence of class destinations on class origins. Coming to occupy a place in the class structure on the basis of who you were and the connections your family possessed should have given way to selection and allocation on the basis of what you had achieved and could be expected to accomplish. In sociological terms, ascription, as the principle of success or failure, should have been replaced by achievement. We shall investigate whether this was so, and examine the degree to which it happened in different countries and at different times.
The study of social mobility concerns the relationship between the class position a person occupies (their class destination) and the class position in which they were brought up (their class origins): mobility occurs when origins and destinations differ. In keeping with most sociological research on mobility, we distinguish two aspects. Absolute mobility refers to the observed patterns of movement between origins and destinations. The simplest measure of absolute mobility is the proportion of people who are in a destination that differs from their origin: this is the overall mobility rate. Within absolute mobility we can separate upward from downward moves, and we can ask whether these are more common among people from one class origin compared with another.
Relative mobility, or social fluidity, deals with the strength of the relationship between origins and destinations. Measures of relative mobility capture the degree to which a person’s destination depends on their origin. Complete social fluidity—also known as perfect mobility—would hold if destinations were independent of origins. As far as we know, no society has ever come close to this situation, but there is plenty of evidence that countries differ in how strongly the class position of one’s family influences the class one comes to occupy. Sweden is widely regarded as a country with high fluidity, in which origins exert less influence on destinations than they do in countries such as Germany or France. But this does not imply that Sweden must have greater rates of absolute mobility: the two aspects of mobility, absolute and relative, vary independently. And, indeed, there is very little difference between Sweden, Germany, and France in their overall mobility rates.
Our first goal is to document changes in absolute and relative mobility, and we do this by comparing cohorts of people born throughout the twentieth century. But our main aim is to relate changes in mobility to changes in education. From the point of view of society as a whole, a more educated population promotes economic growth and national prosperity. From the individual point of view, education is widely agreed to be the key to getting a good job and securing favourable life chances. In the countries of Europe and North America, the twentieth century was a period of growth in education. In broad terms, at its beginning, the majority of children left school with only primary education; by its end 30 per cent or more of young people were acquiring university degrees. It was also the case (though there is debate about this among sociologists) that the education people attained came to depend less on their social origins. Thus, the twentieth century saw both educational expansion and educational equalization. Our central question is: Were these developments associated with changes in social mobility and social fluidity?
For many years, notwithstanding its popularity as a topic of sociological research, intergenerational mobility was not a subject of widespread public concern. But over the past twenty years it has risen close to the top of the policy agenda, at least in the UK and US. In the UK the policy concern has been driven by the belief that rates of mobility are declining, to the disadvantage of children from poorer backgrounds whose path to a better future is obstructed. In the US the concern about mobility has been prompted by the finding that, contrary to popular belief, rather than the US being one of the most intergenerationally mobile societies in the developed world it is, in fact, among the least mobile (when measures of intergenerational income or earnings mobility are used) as well as being one of the most unequal.1 These concerns were summarised by US president Barack Obama, speaking in 2013:
The problem is that alongside increased inequality, we’ve seen diminished levels of upward mobility in recent years. A child born in the top 20 percent has about a 2-in-3 chance of staying at or near the top. A child born into the bottom 20 percent has a less than 1-in-20 shot at making it to the top. He’s 10 times likelier to stay where he is. In fact, statistics show not only that our levels of income inequality rank near countries like Jamaica and Argentina, but that it is harder today for a child born here in America to improve her station in life than it is for children in most of our wealthy allies—countries like Canada or Germany or France. They have greater mobility than we do, not less.2
In the UK and the US the promotion of greater social mobility has entered the manifestos of political parties.3 In 2010, the then British prime minister Gordon Brown said: “Social mobility will be our theme for the coming election and the coming parliamentary term. Social mobility will be our focus, not instead of social justice, but because social mobility is modern social justice.”4
A favoured mechanism for addressing these concerns is education, and in both the US and the UK various policies have been proposed with the avowed aim of equalizing the chances of intergenerational mobility. But whether they will succeed is not clear, not least because studies have come to conflicting conclusions. While many authors report evidence that changes in education affect intergenerational mobility (for example, Blanden, Gregg, and Machin 2005; Causa and Johansson 2010; Mayer and Lopoo 2008) others are sceptical (Goldthorpe 2007). Research on the impact of the raising of the school leaving age in England and Wales in 1972 showed that, although it led to an increase in the average number of years of schooling completed, it had no discernible effect on intergenerational mobility (Buscha and Sturgis 2015). A US study (Rauscher 2016), focusing on the introduction of compulsory schooling laws in the US in the nineteenth century, also failed to find a positive impact on mobility. On the other hand, Betthäuser (2017) found that increasing the school leaving age in Germany promoted greater intergenerational mobility.
If we want to understand what drives rates of intergenerational mobility and what determines how strongly a person’s mobility chances are tied to their social background, examining what has happened in the past is a good place to start. And that is what we do in this book. The countries we deal with are sufficiently similar to make comparisons between them sensible, but they are also sufficiently different for us to be able to gain some idea of how variations in the timing and extent of educational change might have had differential impacts on mobility. Considered together they should allow us to draw some conclusions about if, and how, educational developments that took place during the twentieth century were related to subsequent changes in mobility, and in doing so may help to inform us about how much education can do to promote greater social fluidity.
Sociological studies of trends in social fluidity have reached conflicting conclusions. The major cleavage is between those authors who find no trend over time in the association between class destinations and class origins and those who see a steady reduction in the degree to which a person’s own class position depends on the class position of his or her parents. The latter is associated with modernisation theory: as societies develop, the forces of competition drive institutions to steadily become more meritocratic (see, for example, Ganzeboom, Luijkx, and Treiman 1989 and Treiman 1970).6 The proponents of the rival view—sometimes called “trendless fluctuation”—claim that the modernisation argument neglects the degree to which those in advantaged positions can secure similarly advantaged positions for their children, despite the forces of modernisation (Erikson and Goldthorpe 1992; Goldthorpe 2000, chapter 11).
The great majority of previous studies of trends, no matter which side of the debate they support, have taken a period approach. This means that they have drawn comparisons of the mobility of the whole population at different points in time. In contrast, we compare the mobility of people depending on when they were born; in other words, we compare birth cohorts. For the most part, the cohorts we use identify people born in the first quarter of the century, and then people born in ten-year intervals up to the mid-1970s. Because we deal with men and women aged between 35 and 70 (the exact range differs slightly between countries) and because, for most countries, the latest data we have come from surveys undertaken in the first decade of the twenty-first century, we cannot observe mobility among people born after about 1975.
In earlier work, to which many of the authors represented here also contributed, we adopted the period perspective, comparing social mobility in twelve European countries in the 1970s, 1980s, and 1990s. But we now believe that there is a compelling reason to prefer the birth cohort approach. Simply, change in social fluidity is a cohort phenomenon. As Müller and Pollak (2004: 96) explain:
Educational reforms, educational expansion or changing competition on the labour market among groups with different qualification levels will affect mainly cohorts which are in school, pursuing higher education or making the transition from school to work, when the respective changes take place. In contrast, these effects may remain largely without consequences . . . for cohorts which had already settled in the labour market. Similarly, dramatic historical events, like World War II . . . may have different impacts on the social opportunities of different cohorts and particularly affect members of cohorts which are in a susceptible stage of their life course.
Müller and Pollak (2004) show that period changes in fluidity in Germany in the last decades of the twentieth century can be explained as the result of cohort replacement: that is, older, less fluid cohorts exiting the labour market and being replaced by younger, more socially fluid cohorts. Using data on cohorts born between 1912 and 1974, Breen and Jonsson (2007) find the same result for Sweden, leading them to propose that “changes in fluidity are normally and mainly—though not exclusively—driven by cohort-related, rather than period-related, factors” (2007: 1777). And they go on to show that educational change, in the form of both expansion and equalization, drove the change in social fluidity over birth cohorts of Swedish men and women.
If we believe that educational change causes, or is, at any rate, associated with, changes in mobility and fluidity, it must be the case that education has changed. And while there is no debate about the expansion of educational provision and the increase in overall attainment over the twentieth century, sociologists have disagreed about whether there has been a trend towards greater equalization. The point at issue here is whether, and to what extent, the association between class origins and educational attainment weakened. To put it another way: did educational fluidity increase or not?
In the late 1990s, most sociologists would have answered no. A number of analyses of single countries and a major comparative study (Shavit and Blossfeld 1993) covering thirteen developed countries gave support to the thesis of “persistent inequality.” This is the view that, during the twentieth century, despite dramatic educational expansion (and in contrast to the marked reduction in gender differences in attainment), there remained “a persistently high degree of class inequality of educational attainment that can change only under rather exceptional circumstances” (Breen et al. 2009: 1476). But since then beliefs have shifted, driven both by the publication of single-country studies that overturned previous findings about a lack of change (for example, Shavit and Westerbeek 1998 who deal with the Italian case) and a large cross-national study (Breen et al. 2009, 2010) that found strong evidence that the association between class origins and educational attainment had declined over birth cohorts born in the twentieth century in Sweden, the Netherlands, Britain, Germany, and France (and somewhat weaker evidence for declines in Italy, Ireland, and Poland).
1. But when mobility is measured in class terms, the US does not seem to be a particularly rigid society (Beller and Hout 2006a; Breen, Mood and Jonsson 2016).
3. Mobility has been a political issue in other countries too, but not to the same extent as in the US and UK. In November 2005, the German chancellor Angela Merkel, in her first government policy declaration, stated, “In this country, origins must not be allowed to determine young people’s future” (Herkunft darf in diesem Land nicht die Zukunft der jungen Menschen bestimmen). Before his election as the French president in 2017, Emmanuel Macron was quoted as saying, “We need to invent a new growth model. To be fair and sustainable, it must be environmentally friendly and increase social mobility.” (https://www.ft.com/content/3691a448-fa1d-11e6–9516-2d969e0d3b65).
5. Breen and Jonsson (2005) provide an exhaustive review of work on social mobility and educational inequality. See also Torche (2015).
6. For example, Ganzeboom and De Graaf (1984) predicted that, based on trends between 1954 and 1977, the Netherlands would reach perfect mobility by 2023 (see also Ganzeboom and Luijkx 2004: 345).