The missing piece of the social mobility puzzle

This is the first of a pair of posts about a project I’ve been working on looking at downward social mobility. In this first post I’m going to talk about why we should care about downward mobility, and what might be missing from our current understanding of it. You can find the second post here.

Opportunity hoarding

When a politician or media commentator talks about improving social mobility, they might say something like ‘we want to make it easier for people from less advantaged backgrounds to succeed’. They usually don’t say ‘we also want to make it easier for children from privileged backgrounds to fail’.

But you can’t have one without the other. Room at the top is finite. Especially now that the post-industrial expansion of ‘professional’ jobs has largely ground to a halt. So in order to increase the probability that children from poorer backgrounds will rise up the socio-economic ladder, we also have to increase the risk that the children of the rich will fall down it.

This is something that high socio-economic status parents very much do not want to happen to their own children. And because of their social and economic resources, their ability to prevent it is considerable. For example, they can enrol their children in private schools or pay for outside tuition. They can move near the best state schools. They can tap up their friends and colleagues for internship or job opportunities. The friction created by these ‘opportunity hoarding’  strategies helps keep the scions of advantaged families from slipping down the socio-economic spectrum.

Enumerating the advantages accruing to children from privileged backgrounds can make it seem like downward mobility is basically impossible – that if you are from an advantaged family, you can’t really help but succeed. But what does the data say?

A 2020 report from the Social Mobility Commission, using data from the UK Labour Force Survey (LFS), found that, from 2014-2018, around 35% of men and 41% of women from ‘professional’ backgrounds had been downwardly mobile. These figures collapse together ‘higher’ professional  and ‘lower’ professional origins (see below for an explanation of this distinction). If we disentangle these the same data, we see that around 60% of men and 75% of women from higher professional and managerial backgrounds are in a lower occupational class than their parents. (I should note here that both sets of figures only include people aged 30 to 59 – so they do not include people who are just starting out in their careers).

This may sound surprising, and totally contrary to everything we’ve been talking about so far. But to understand exactly how surprising it is (or isn’t) we need to take a step back and think about the structure of the labour market. The table below shows a rough breakdown of the proportion of working people in each occupational class category. These figures are derived from the 2019 LFS (the last year of the survey before the pandemic), and again we are talking about working people aged 30-59:

Higher managerial, administrative & professional (Class 1) Examples: Lawyers, Doctors, Senior managers & executives	18%
Lower managerial, administrative & professional (Class 2) Examples: Nurses, Teachers (primary and secondary), Writers, IT technicians, Many ‘office’ workers	25%
Intermediate (Class 3) Examples: Paramedics, Police officers, Graphic designers	17%
Small employers & own account workers (Class 4) Examples: Shopkeepers, Landscape gardeners, Taxi drivers, Farmers	7%
Lower supervisory & technical (Class 5) Examples: Plumbers, Chefs, Train drivers, Mechanics	7%
Semi-routine (Class 6) Examples: Receptionists, Care workers, Market research interviewers, Factory machine operators	15%
Routine (Class 7) Examples: Hairdressers, Lorry drivers, Labourers, Cleaners	12%

NS-SEC classes and examples

Imagine a society with perfect social mobility – where people move freely into any occupational class, regardless of their background. In this society, around 18% of people from ‘higher professional’ backgrounds would end up in higher professional jobs themselves – meaning 82% would be ‘downwardly mobile’. Britain in the real world departs quite substantially from this ideal. 40% of men from the highest class backgrounds end up as higher professionals themselves, compared with only 10% of men from ‘routine’ backgrounds. It is also worth noting that, of the men and women from advantaged backgrounds who are downwardly mobile, fully half have fallen only one rung – down to ‘lower professional’. In social mobility parlance, this is known as ‘short-range downward mobility’. Only 30% of people from privileged backgrounds have fallen further down than this second rung.

But even taking all of this into account, these figures suggest a somewhat healthier level of downward mobility than might be expected from the ‘opportunity hoarding’ narrative. But what we have ignored so far is what these occupational class categories are actually capturing about people’s position in society, and, crucially, what they are missing.

The categories I have been referencing come from the National Statistics Socio-Economic Classification scheme (NS-SEC). The NS-SEC, and schemes like it, are an attempt to organise many thousands of different jobs into clumps, so that researchers and policymakers can get a better understanding of patterns and trends. The NS-SEC specifically groups jobs together according to the conditions of their employment – with the main focus being on the amount of discretion people typically have over their work. So at the bottom, we have people with occupations in the ‘routine’ and ‘semi-routine’ classes (Class 6 and 7) who generally do very narrowly regulated tasks (e.g. cleaning or assembling products on a factory line). By contrast, people with occupations in Class 1 are typically paid a salary to perform a more general role – such as ‘Operations Director’ or ‘University Professor’ – within which they have a large amount of autonomy. Class 1 also includes large employers who purchase and direct the work of others.

The NS-SEC dominates social mobility research in the UK. This is partly for good scientific reasons – the NS-SEC classes are (and are intended to be) highly correlated with things like pay, job security, and opportunities for progression. However, it is also because occupational class is relatively easy to measure in large surveys. You just need to ask a few questions about someone’s job, and then use a well-defined set of rules to assign them to a particular class.

So there is nothing wrong with using occupational class to study social mobility. The risk comes when we slip into the perception that the measure is the concept – that our statistics on occupational class mobility are giving us a holistic idea of ‘social mobility’ in general. (This is a very easy thing to do with any measure in social science).

The very narrow criteria on which NS-SEC is based clearly do not capture everything about an occupation’s position in society. Even if we also consider income, we are still missing something crucial from the picture. Imagine an elite parent – let’s say they are a retired lawyer. Their 40-year-old daughter works as a manager in a bar. Their friend – a retired surgeon, has a daughter the same age who is a classical violinist. Which of these parents is most likely to think their daughter has been ‘downwardly mobile’? Who is likely to proudly tell their friends what their daughter does for a living, and who may be just a touch embarrassed?

According to each daughter’s NS-SEC class, there is no difference between these two situations – ‘musician’ and ‘bar manager’ are both NS-SEC Class 2 occupations. And as far as income goes, it is entirely possible for even a relatively successful musician to earn less than a bar manager.

The missing ingredient here is social status (or prestige if you prefer). This is the esteem in which your occupation is held by wider society. If you tell someone what you do, are they impressed? Do they look up to you or down on you? This is obviously correlated with income – higher income jobs tend to be more prestigious. But there are plenty of jobs which attract much higher social esteem than their pay would suggest (nurse, teacher, writer, priest), and vice-versa (plumber, long-distance lorry driver, office manager). Social status is also correlated with, but distinct from, NS-SEC class. For example, almost all jobs in media and the arts (highly sought-after, highly prestigious) are in NS-SEC Class 2. Few would argue that IT manager was a more prestigious job than theatre director, but the former is Class 1 while the latter is Class 2.

When set against the obvious importance of income, or the detailed, technical edifice of NS-SEC, ‘status’ can seem a bit…soft, and potentially superfluous. But status is arguably more vital and salient in most people’s lives than either occupational class or even income.

Research has shown that people carry around in their heads a remarkably consistent prestige ranking of occupations. The prestige of someone’s job is therefore our readiest source of information about other people’s position in the social pecking order – and their best source of information about ours. Our social status also determines, to a much greater extent than income or occupational class, who we associate with – who we befriend and marry. In fact what most people think of when they think of ‘social classes’ are arguably ‘status groups’. Imagine a typical ‘middle-class’ (in the British sense) dinner party. What jobs would you expect to find there? What jobs would you be surprised to see? You’ll find that the list is much more closely tied to prestige than it is to income or NS-SEC class. If we ignore status, we are potentially missing a huge part of the story of downward mobility and opportunity hoarding. Downward mobility in terms of occupational class may be somewhat alive and well in Britain. But are those who are falling down the class ladder really also falling down the status ladder? Or are they mostly ending up in coveted, prestigious jobs that just happen to fall into a different NS-SEC class? Tune in for the next post to find out…