Thursday, 28 February 2013

Does income equality make us happier?

I've been interested in the central claims of The Spirit Level for quite a while now.  And for those who have been in a bunker since 2009, its authors - the epidemiologists Richard Wilkinson and Kate Pickett - argue that the level of income inequality in a society determines a wide range of health and social problems.  Reducing income inequality, then, is a proposed cure for a whole batch of social ills.

I think in general The Spirit Level has increased the standard of debate around inequality.  No longer do we think of inequality in crass economic terms: will a bit more inequality make everyone a little bit richer?  Now we think about the social costs of inequality.  And part of this is how inequality makes us feel.

One of the key arguments of the pro-equality lobby is that high levels of inequality make us feel bad.  We might feel an injustice that others get so much more than us.  We might feel uncomfortable in places where people are different to us.  We might feel shame that we can't afford what others so easily can.

Turning this on its head, the opposite should also be true.  We should feel better when other people - especially those around us - are similar to us.  The world should appear more just, we should feel more comfortable and less ashamed of our social status.  But to what extent is this true?  Does higher equality make us happier?

Fortunately, data from the (now defunct) Citizenship Survey allow us to explore this further.  There is an interesting question from 2010 that asks people whether their friends earn a similar amount to them:

Subsequently, we can split people's answers into two categories: a) people who have friends who all have similar incomes to them (1) and b) people who don't (2-4).  In practice, 27% of the survey's respondents fall into category (a): what we might call the 'equal' group.  When we examine these groups against their average level of life satisfaction, we can see if there are any systematic differences.

In the Citizenship Survey, life satisfaction (once reversed) is measured on a scale of 1 - 5: where 1 equals 'very dissatisfied' and 5 equals 'very satisfied'.  The average score for the whole sample in the survey (11,063 people) is 4.1.  So where do the 'equal' and 'unequal' groups fit in?

As it turns out, the 'equal' group - those people whose friends have similar incomes to them - have an average life satisfaction score of 4.2.  The 'unequal' group - those who tend to have friends with different incomes to them (whether higher or lower) - have an average score of 4.1.

This is a relatively small difference, but it could be an important one.  And when we look at relationships like this we want to know whether such a difference is due to other factors.  It could turn out, for example, that having friends with similar incomes is correlated with other things we know increase life satisfaction: religion, where you live, age or having children.  So the next step is to assess the effect of equality on life satisfaction using multiple regression techniques that control for a whole range of variables that we also think are important for happiness.  In my model, this includes labour market status, education level, age, gender, region, income, religion and ethnicity.

Once we've done this, the results still suggest that having a strong level of equality within your group of friends is linked to a higher level of life satisfaction.  The effect is not huge (but it is statistically significant).  It is about a third of the size of the effect that marriage has on life satisfaction and a quarter of the size that having a job has.  If you want to be happy then, perhaps the old adage is true.

And don't make friends with bankers.

Thursday, 7 February 2013

Is workfare really demoralising?

Workfare is a much disputed term.  In the strictest sense, it refers exclusively to work experience type programmes that unemployed people are required to take part in as a condition of social security: 'work for benefit'.  At a broader level, critics use workfare as a catch-all term to describe a wide range of welfare-to-work measures.

Whichever definition we use, the campaign against conditionality in the benefits system - spearheaded by groups such as Boycott Workfare - has come alive in the past year.  The catalyst was the public and media furore in February 2012 that accompanied the 'revelations' around Tesco's involvement in a scheme.

The various campaigns against 'workfare' often make certain assumptions.  In particular, that being placed on a workfare scheme is degrading and demoralising.  Critics often use human stories about the effect that compulsory referral has had on an unemployed person's morale and wellbeing.  The argument is that workfare makes the unemployed feel even worse than they do already.

There is no doubting the integrity of some of the claimants who have publicised their workfare journeys and the harmful experiences they've had.  However, to what extent are we exposed to a broad range of workfare stories?  Or, alternatively, are we merely hearing the voices of a small, vocal, well-educated and articulate minority of workfare participants?  Could workfare have different effects for different groups of people?

Data I'm looking at for my doctoral work support this view: that how workfare makes us feel - whether it makes us happier or unhappier - depends upon how well educated we are.  The data come from the Annual Population Survey 2011-12.  To clear this up, let's first look at the average level of 'life satisfaction' for selected groups within the labour market.  The higher the score, the higher the satisfaction.

Average life satisfaction score by labour market status

Employed : 7.52
Self-Employed: 7.53
Unemployed: 6.37
Welfare-to-Work: 6.55

As is clear, those in work - whether employed or self-employed - have significantly higher life satisfaction than those who are either unemployed or on welfare-to-work schemes.  This is not surprising; lots of research has shown that being in work is good for your general wellbeing and being out of work isn't.  Welfare-to-work participants have slightly higher life satisfaction than the unemployed, but not by too much.

However, let's next look at the welfare-to-work group more closely and separate it out by highest level of education.  Doing so will tell us whether the relationship between workfare and life satisfaction is similar for all participants irrespective of how educated they are.

Average life satisfaction score for welfare-to-work participants, by highest qualification level

Higher Education: 5.55
A-Levels: 6.30
GCSE: 6.66
Other Qualification: 6.71
No Qualifications: 7.54

The results are stark and the pattern is clear.  For people on welfare-to-work programmes who have been through higher education, life satisfaction is very low.  Much lower in fact than it is for the average unemployed person.  For workfare participants with degrees, then, the campaigners look right: workfare is demoralising.

However, workfare isn't as demoralising for other groups.  In particular, for workfare participants with no qualifications, life satisfaction is relatively high.  It's even around the same level, for example, as a person in paid work.  This strongly suggests that the effect of workfare on a person's level of life satisfaction is heavily dependent upon their education level.  

This evidence suggests that the stories we hear about welfare-to-work are unrepresentative of all participants. And in a digital age, when we increasingly consume information about policy through the internet, this is a cause for concern.  Online policy stories are often snapshots delivered by particular groups: the well-educated, the articulate, the confident.  Campaigning for policy change based upon the experiences of such vocal groups is a dangerous pursuit; especially for those who claim to defend the disadvantaged.