Perpetuum mobile?

Social mobility does not work as most people imagine, and it's hard to do much about it
November 20, 2005

Social mobility is again the focus of intense interest in political circles. There has been much concern, especially on the centre-left, that mobility appears to have slowed or even reversed, despite eight years of a Labour government committed to redistribution, increased investment in state education and so on.

This concern is understandable, but there is much confusion surrounding mobility rates and their relation to wider socioeconomic trends, which it is worth trying to sort out.

First, we need to understand that there are two quite distinct types of intergenerational mobility. Sociologists label these as "class structural change" and "fluidity" within the class structure. To illustrate the difference between the two, we should consider the analogy first proposed by the economist Joseph Schumpeter.



For many years, an oddly designed residential hotel remained in its original, roughly pyramidal form, with the rooms improving from the bottom floor to the top. Over this period, "mobility" could only occur by individuals changing rooms; and, since the hotel was always full, such mobility had to be quite symmetrical: for every resident who moved up to a better room, another had to move down to a worse room. The extent of such movement is an expression of "fluidity" effects.

However, after a time the hotelier decided to upgrade his establishment, and did so by reducing the number of rooms on the bottom floor while expanding at the top. As a result of this development, therefore, some "upward" mobility could occur in an asymmetrical way: some residents could move up without any having to move down.

"Schumpeter's hotel" illustrates the single most important finding of mobility research: the effects of "class structural change" are far more important than fluidity effects in regard to either changes in observed rates of class mobility over time within particular national societies, or differences in observed rates of class mobility among national societies.

The big story regarding class mobility in Britain over most of the 20th century is that rates of upward mobility steadily increased—and did so essentially as a result of class structural change. This was a result of the proportionate growth of the professional and managerial salariat and the corresponding decline of the manual working class. There was, quite simply, "more room at the top." Fluidity effects played virtually no part.

Increases in fluidity have been observed in some other modern societies, but still only of a relatively slight and episodic kind. Moreover, there are substantial similarities in the pattern of fluidity across societies. This finding undermines claims of national exceptionalism—most notably, perhaps, the claim that the US has a distinctively fluid or "open" class structure. It does not. In the light of the available research, we can produce broad groupings of national societies in terms of fluidity (see table). Fluidity within class structure

High - Australia, Israel, Japan, Norway, Poland, Sweden Intermediate - France, Great Britain, Hungary, Netherlands, US Low - Germany, Ireland, Italy

So what are the wider implications of these findings for public policy?

First of all, they underline the importance of clarity in policy objectives. Is our aim simply to increase actually observed rates of mobility? Or is it to increase the degree of fluidity? When politicians talk about social mobility, they often seem confused on this point. If the concern is with increasing actually observed—and upward—mobility, then the focus should be on increasing the rate of class structural change: increasing, in other words, the proportion of relatively high-grade jobs and decreasing the proportion of low-grade ones. It means "upgrading the hotel," as we saw in Britain, and many other countries, in the 20th century.

However, it is not easy for governments to ensure that class structural change does in fact take this form. Economic growth may not be enough. For under certain conditions, growth can have the effect of increasing the numbers in less advantaged class positions, just as much as it does for more advantaged positions. In Britain since the 1990s, for example, largely as a consequence of de-industrialisation, a certain "hollowing out" of the class structure seems to have occurred. The professional and managerial salariat has continued to expand, though more slowly than before; intermediate-grade jobs often associated with the manufacturing sector have fallen off sharply, but we have also seen a growth in low-grade jobs, especially in the service sector, with low pay, high insecurity and minimal prospects for advancement.

Changes in the labour market are reducing the objective opportunities for movement from lower grade to higher grade employment. As a result of this pattern of class structural change, upward mobility rates in Britain are tending to level out.

If, on the other hand, the aim is to increase fluidity, then structural change becomes less relevant. The focus has to be on reducing the inherent "stickiness" between the class positions of parents and those of their children. Again, this is far from easy to do.

To bring about greater fluidity was an explicit objective of governments in several communist societies in the years after the second word war. But, even with the advantages of complete control over educational selection and a command economy, the increases in fluidity that these governments managed to achieve were only quite modest and proved difficult to sustain.

Much of the explanation of why it is so difficult to increase fluidity lies in the fact that parents in more advantaged class positions can use the superior resources that they possess—economic, social and cultural—to protect their children against the threat of downward movement.

This points to the general conclusion that if we want to increase fluidity within the class structure—to create greater equality of opportunity—we will first need to create a greater equality of condition. For it is from inequalities of condition that the inherent "stickiness" between the class positions of parents and children derives. The Swedish experience of increasing fluidity from the 1940s through to at least the 1980s was associated with the development of a distinctive welfare state and form of political economy that markedly reduced class inequalities in incomes and living standards. And conversely, the latest evidence from the US and Britain suggests that, following on the rising economic inequality of recent decades, social fluidity is now showing a downturn.

However, there is a further difficulty with increasing fluidity—even for a government committed to increasing equality of opportunity. This is the fact that the role of education in mobility may be less important than is often supposed. In so far as an increase in mobility results from class structural change, then education is almost by definition of no great significance, (although it may help to select those from less advantaged backgrounds who move up into the new higher class positions). In Britain, for example, upward mobility in the 1940s, 1950s and 1960s owed little to educational change. Later in the century, education did become of greater importance as a channel of mobility; but this is still not the same thing as being a cause of mobility. For education actually to increase mobility, it has to increase fluidity, and for it to do this, three things are necessary.

First, the effect of individuals' class origins on their educational attainment must weaken; second, the effect of individuals' educational attainment on their eventual class destinations must strengthen; and third, no offsetting strengthening should occur in the "direct" effect of origins on destinations: the path that does not run through education.

However, even leaving aside this last issue, the available evidence on the first two paths is not encouraging for those who suppose that educational expansion and reform are the key to the creation of a more open form of society.

On the one hand, the effects of class origins on educational attainment are highly resistant to change. In some societies—Sweden and France, for example—these effects do appear to have weakened somewhat overall. But in others—Britain, for example—such change, if it has occurred at all, is very slight and patchy.

On the other hand, what research over recent decades shows is that the effect of education on class destinations is, if anything, tending to decline. Why is this? First, as educational qualifications of all kinds become more widespread, they in turn become less distinctive—and carry less information for employers. Second, despite all the talk of "knowledge economies," in recent decades the main growth in employment at all levels has been in services, and especially in personal service jobs which cannot be replaced by technology. And in such jobs, cognitive and technical abilities of the kind that formal qualifications best warrant may be of less importance to employers—above some threshold—than a whole range of other "lifestyle" attributes that are as much a product of social background as of ability and effort.

Social mobility is a complex matter, although one which sociologists have made real progress in understanding. The results of their work indicate that policies that aim to increase mobility will not be easy either to formulate or apply.