The Spirit Level: Why Greater Equality Makes Societies Stronger

by Richard Wilkinson

  THE EQUALITY DIET

  It is clear that obesity and overweight are not problems confined to the poor. In the USA, about 12 per cent of the population are poor, but more than 75 per cent are overweight. In the UK, social class differences in women’s obesity can be seen all the way up the social ladder. While obesity affects only 16 per cent of ‘higher managerial and professional’ women, just below them, 20 per cent of lower managerial and professional women are obese. It’s hard to argue in the face of these facts that the obesity epidemic is due to poor nutritional knowledge among the uneducated. In a study of middle-aged British women,137 84 per cent knew they should be eating five fruits and vegetables each day, and another study showed that obese people are better than thinner people at guessing the calorie content of snack foods.138

  Another piece of evidence that it’s relative, not absolute, levels of income that matter for obesity comes from studies in which people are asked to describe subjectively their place in the social hierarchy. Researchers show subjects a diagram of a ladder and tell them that at the top are people with the highest status, and at the bottom people with the lowest status, and then ask them to place an ‘X’ on the ladder to mark their own standing. It has been shown that this measure of subjective social status is linked to an unhealthy pattern of fat distribution139 and to obesity140 – in other words, obesity was more strongly related to people’s subjective sense of their status than to their actual education or income.

  If we can observe that changes in societal income inequality are followed by changes in obesity, this would also be supportive evidence for a causal association. An example of a society that has experienced a rapid increase in inequality is post-reunification Germany. After the fall of the Berlin Wall, inequality increased in the former East Germany,141 and there is evidence from studies following people over time that this social disruption led to increases in the body mass index of children, young adults and mothers.142

  Health and social policies for obesity treatment and prevention tend to focus on the individual; these policies try to educate people about the risks associated with being overweight, and try to coach them into better habits. But these approaches overlook the reasons why people continue to live a sedentary lifestyle and eat an unhealthy diet, how these behaviours give comfort or status, why there is a social gradient in obesity, how depression and stress in pregnancy play a role. Because behaviour changes are easier for people who feel in control and in a good emotional state, lessening the burdens of inequality could make an important contribution towards resolving the epidemic of obesity.

  *BMI = weight in kg/(height in m)2

  *The data on adult obesity within the USA were made available to us by Professor Majid Ezzati from Harvard University School of Public Health. Professor Ezzati bases his calculations of the prevalence of obesity in each state on actual measures of height and weight.

  8

  Educational performance

  Our progress as a nation can be no swifter than our progress in education. The human mind is our fundamental resource.

  John F. Kennedy, Special message to the Congress

  on Education, 20 February 1961

  Across the developed world, and across the political spectrum, everybody agrees about the importance of education. It’s good for society, which needs the contributions and economic productivity – not to mention the tax – of a skilled workforce, and it’s good for individuals. People with more education earn more, are more satisfied with their work and leisure time, are less likely to be unemployed, more likely to be healthy, less likely to be criminals, more likely to volunteer their time and vote in elections.143 In 2006, according to the US Department of Labor, if you had been to high school but didn’t graduate with a diploma, you earned an average of $419 per week. That sum rose to $595 if you had the diploma, up to $1,039 if you’d gone on to college and got a bachelor’s degree, and rose to over $1,200 for an advanced degree.144

  THE HOME ADVANTAGE

  Although good schools make a difference, the biggest influence on educational attainment, how well a child performs in school and later in higher education, is family background. In a report on the future of education in Britain, Melissa Benn and Fiona Millar describe how:

  One of the biggest problems facing British schools is the gap between rich and poor, and the enormous disparity in children’s home backgrounds and the social and cultural capital they bring to the educational table.145,p. 23

  Children do better if their parents have higher incomes and more education themselves, and they do better if they come from homes where they have a place to study, where there are reference books and newspapers, and where education is valued.146 Parental involvement in children’s education is even more important.

  So why, when all developed societies are committed to education and equality of opportunity (at least in theory), do disadvantaged children do less well at school and miss out on the myriad benefits of education, however good the school system? As we shall see, some societies come a lot closer to achieving equality of opportunity than others.

  UNEQUAL ATTAINMENT

  Figure 8.1 shows that international educational scores are closely related to income inequality and Figure 8.2 shows the same relationship for the USA. More unequal countries and more unequal states have worse educational attainment – and these relationships are strong enough for us to be sure that they are not due to chance. Comparable international data on educational achievement come from the Programme for International Student Assessment (PISA), which was set up to administer standardized tests to 15-year-olds in schools in different countries. The programme began in 43 countries in 2000, and assesses children every three years, typically testing between 4,500 and 10,000 children in each country each time; schools are randomly selected. PISA tests 15-year-olds because they are coming to the end of compulsory education in most countries. Each survey gives tests in reading, mathematical and scientific literacy. The goal is to test how well children can apply knowledge and skills.

  Figure 8.1 Maths and literacy scores of 15-year-olds are lower in more unequal countries.148–149

  Figure 8.2 Maths and literacy scores of eighth-graders are lower in more unequal US states.

  For consistency with the data available for the US, we combine national average scores for reading and maths only and plot them against income inequality (Figure 8.1). However, if scientific literacy scores are added in it makes little difference to the results. No data were available for the UK from PISA 2003, as too few schools agreed to take part in the survey to meet the PISA standards. The same strong international relationship with income inequality has been shown for adult literacy scores as well, using data from the International Adult Literacy Survey.147

  To examine the same relationship among the fifty states of the USA, we combined maths and reading performance scores for eighth-graders (aged around 14 years old) from the US Department of Education, National Center for Education Statistics for 2003 (Figure 8.2). The scores are significantly lower in states with wider income differences.

  As a further test, we looked at the proportion of children dropping out of high school in the USA. As Figure 8.3 shows, children are much more likely to drop out of school in more unequal states. The states with the lowest drop-out rates are Alaska, Wyoming, Utah, Minnesota and New Hampshire, with drop-out rates around 12 per cent. In three states, Mississippi, Louisiana and Kentucky, more than a quarter of children drop out of high school with no educational qualifications.

  Figure 8.3 More children drop out of high school in more unequal US states.

  You might think that this striking association is due to absolute poverty – that kids drop out of high school more frequently in poor states, so that they can start earning sooner and contribute to the family budget. And it is true that high school drop-out rates are higher in poor states, but poverty and inequality have independent effects. Poverty does not explain the inequality effect. No state has a poverty rate higher th
an 17 per cent but drop-out rates are above 20 per cent in sixteen states and dropping out is not confined to the poor.

  STANDARDS OF PERFORMANCE

  It is often assumed that the desire to raise national standards of performance in fields such as education is quite separate from the desire to reduce educational inequalities within a society. But the truth may be almost the opposite of this. It looks as if the achievement of higher national standards of educational performance may actually depend on reducing the social gradient in educational achievement in each country. Douglas Willms, professor of education at the University of New Brunswick, Canada, has provided striking illustrations of this.150 In Figure 8.4 we show the relation between adult literacy scores from the International Adult Literacy Survey and their parents’ level of education – in Finland, Belgium, the UK and the USA.

  This figure suggests that even if your parents are well educated – and so presumably of high social status – the country you live in makes some difference to your educational success. But for those lower down the social scale with less well-educated parents, it makes a very much larger difference. An important point to note, looking at these four countries, is the steepness of the social gradient – steepest in the USA and the UK where inequality is high, flatter in Finland and Belgium, which are more equal. It is also clear that an important influence on the average literacy scores – on national levels of achievement – in each of these countries is the steepness of the social gradient. The USA and UK will have low average scores, pulled down across the social gradient.

  Figure 8.4 Literacy scores in relation to parents’ education in four countries (data source: International Adult Literacy Survey).

  Willms has demonstrated that the pattern we’ve shown in Figure 8.4 holds more widely – internationally among twelve developed countries, as well as among Canadian provinces and the states of the USA.151 As well as the tendency towards divergence – larger differences at the bottom of the social gradient than at the top – he says ‘there is a strong inverse relationship between average proficiency levels and the slope of the socioeconomic gradients’.

  Epidemiologist Arjumand Siddiqi and colleagues have also looked at social gradients in reading literacy in 15-year-olds, using data from PISA 2000.152 They found that countries with a long history of welfare state provision did better and, like Willms, report that countries with higher average scores have smaller social differences in reading literacy. Finland and Sweden have high average reading scores and low levels of inequality in reading scores; Greece and Portugal have low average scores and a high degree of inequality in reading literacy. Siddiqi and colleagues do, however, note some exceptions to this general pattern. New Zealand and the UK have high average reading scores, but a high degree of social inequality in reading literacy. On the other hand Norway combines a rather mediocre average score with very little socio-economic inequality in reading literacy. One explanation offered by these researchers is that New Zealand and the UK have a greater proportion of children who should sit the tests, but do not, because they have dropped out, or are truants.

  EDUCATIONAL WELFARE

  Siddiqi and colleagues emphasize that high reading scores and low social inequalities in reading literacy are found in nations ‘marked by stronger welfare provisions’. This is a point we will return to in Chapter 12, when we look at public spending on education in relation to income inequality. But how else might income inequality affect educational outcomes?

  One important connection is likely to be through the impact of inequality on the quality of family life and relationships. Social inequalities in early childhood development are entrenched long before the start of formal education. We know a lot now about the importance of the early years for later development – learning begins at birth and the first few years of life are a critical period for brain development. This early learning can be enhanced or inhibited by the environment in which a child grows up. A nationwide study in the UK found that, by the age of three years, children from disadvantaged backgrounds were already educationally up to a year behind children from more privileged homes.153

  Essential for early learning is a stimulating social environment. Babies and young children need to be in caring, responsive environments. They need to be talked to, loved and interacted with. They need opportunities to play, talk and explore their world, and they need to be encouraged within safe limits, rather than restricted in their activities or punished. All of these things are harder for parents and other care-givers to provide when they are poor, or stressed, or unsupported.

  In Chapter 4 we described how the general quality of social relationships is lower in more unequal societies, and in Chapters 5 and 6 we showed how inequality is linked to poor physical and mental health and more substance abuse. It’s not a great leap then to think how life in a more hierarchical, mistrustful society might affect intimate, domestic relationships and family life. Domestic conflict and violence, parental mental illness, poverty of time and resources will all combine to affect child development. The results of these stresses can perhaps be seen in an analysis by economists Robert Frank and Adam Levine, of Cornell University. They showed that in the United States, counties that had the largest increases in income inequality were the same counties that experienced the largest rises in divorce rates.154 Children living in low-income families experience more family conflict and disruption and are more likely to witness or experience violence, as well as to be living in more crowded, noisy and substandard housing155 – the quality of the home environment is directly related to income.156 The way parents behave in response to relative poverty mediates its impact on children – there is evidence that some families are resilient to such problems, while others react with more punitive and unresponsive parenting, even to the extent of becoming neglectful or abusive.157–158 It is important, once again, to note that difficulties in family relationships and parenting are not confined to the poor. Sociologist Annette Lareau describes how parenting differs between middle-class, working-class and poor families in America: there are key differences in the organization of daily life, the use of language, and the degree to which families are socially connected.159 We have found that within the UK Millennium Cohort Study, a large survey of children born in 2000 and 2001, even mothers in the second from the top social class group are more likely to report feeling incompetent as a parent or having a poor relationship with their children, compared to those in the topmost group.

  Societies can do a lot to ameliorate the stresses on families and to support early childhood development. From the very start of life, some societies do more than others to promote a secure attachment between mother and infant through the provision of paid maternity leave for mothers who work. Using data on the duration of paid maternity leave, provided by the Clearinghouse on International Developments in Child, Youth and Family Policies at Columbia University, we found that more equal countries provided longer periods of paid maternity leave.

  Sweden provides parental leave (which can be divided between mothers and fathers) with 80 per cent wage replacement until the child is 18 months old; a further three months can be taken at a flat rate of pay, and then another three months of unpaid leave on top of that. Norway gives parents (again either mother or father) a year of leave at 80 per cent wage replacement, or forty-two weeks at 100 per cent. In contrast, the USA and Australia provide no statutory entitlement to paid leave – in Australia parents can have a year of unpaid leave, in the USA, twelve weeks.

  As well as allowing parental leave, societies can improve the quality of early childhood through the provision of family allowances and tax benefits, social housing, health care, programmes to promote work/life balance, enforcing child support payments and, perhaps most importantly, through the provision of high-quality early childhood education. Early childhood education programmes can foster physical and cognitive development, as well as social and emotional development.160–162 They can alter the long-term trajectories of children’s live
s, and cost-benefit analyses show that they are high-yield investments. In experiments, disadvantaged children who have received high-quality early childhood education are less likely to need remedial education, less likely to become involved in crime, and they earn more as adults.160 All of this adds up to a substantial return on government investments in such programmes.

  UNEQUAL LEARNING OPPORTUNITIES

  So far we have described ways in which greater inequality may affect children’s development through its impact on family life and relationships. But there is also evidence of more direct effects of inequality on children’s cognitive abilities and learning.

  In 2004, World Bank economists Karla Hoff and Priyanka Pandey reported the results of a remarkable experiment.163 They took 321 high-caste and 321 low-caste 11 to 12-year-old boys from scattered rural villages in India, and set them the task of solving mazes. First, the boys did the puzzles without being aware of each other’s caste. Under this condition the low-caste boys did just as well with the mazes as the high-caste boys, indeed slightly better.