Poor health does not preclude a happy life, though it does increase the odds against it

BMJ 2012; 345 doi: (Published 25 July 2012) Cite this as: BMJ 2012;345:e5073
  1. Nigel Hawkes
  1. 1London

The first results of the programme to measure national wellbeing in the United Kingdom show that 40% of people who rate their own health as bad or very bad nevertheless also report medium to high levels of satisfaction with life.

The statisticians from the Office for National Statistics (ONS), which is responsible for the programme,1 were surprised that ill health did not have a greater effect on satisfaction, but there is no denying that good health makes a difference: 80% of people whose health was good or very good reported medium to high levels of life satisfaction, twice the proportion of those whose health was poor.

So far, the nascent programme does not yet allow any deeper explanation of causes, except for a few obvious ones, such as a clear link between unemployment and low life satisfaction.

It remains unclear, for example, why people of black Caribbean or African origin profess the lowest level of life satisfaction of any ethnic group (6.7 of a possible 10, compared with a score of 7.4 for white people), why the people of Rutland and of Bath and north east Somerset are so indecently satisfied with life (with more than 85% in both areas scoring high on this measure), or why Londoners endure chart topping levels of anxiety.

The programme aimed to develop subjective measures of wellbeing, to be read in conjunction with traditional objective measures such as life expectancy and employment status. It does not seek, said Glenn Everett, its director, to define a single happiness index.

“The measures will supplement and not supplant older measures,” he said at a London press conference to launch the first year’s findings. “And the distributions, not just the means, are important.”

A total of 150 000 people aged over 16 were interviewed as part of the Integrated Household Survey. This sample was large enough to enable breakdown by region, ethnicity, employment status, and other variables without unacceptable loss of statistical significance.

Respondents were asked four questions in the year between April 2011 and March 2012: how satisfied they were with their lives nowadays; to what extent did they feel that the things they did in their lives were worthwhile; how happy they felt yesterday; and how anxious they felt yesterday.

Three quarters (75.9%) rated their life satisfaction at 7 or more out of 10, while 80% gave the same rating for the worthwhile question. Just over one in 10 (10.9%) said that they were unhappy yesterday, with a score of less than 5 out of 10, while 21.8% scored their anxiety as high (more than five out of 10). So twice as many people are anxious as are unhappy.

Women were more likely than men to profess themselves satisfied and to rate their own activities worthwhile but were also more likely to be anxious. Arabs were the most anxious ethnic group; and married or cohabiting people were the most satisfied.

When analysed by age, life satisfaction follows the well established U shaped curve, higher in the years up to 39, lower in middle age, and then rising again in the 60s and 70s before a final decline in the 80s.2 The same trend was more emphatic for the “happy yesterday” question and, in the inverse sense, the “anxious yesterday” question. A quarter of 50-54 year olds were very anxious yesterday, for example, compared with 17% of 16-18 year olds and 19% of 75-79 year olds.

Everett said that the statistics were experimental and that their final form would depend on input from outsiders and further analysis by ONS. He agreed that it was important, once the final form was determined, to avoid frequent future changes so that a time series can be built on a consistent method and set of questions.

ONS is encouraging businesses to use the same set of questions in staff surveys, so that the happiness of their workforce can be compared with the national benchmark.

To the extent that any long term comparisons are possible—as is the case with objective measures such as income and some subjective measures that have been examined before—people do not seem to be sinking into misery. For example, average rating of satisfaction with family life rose from 7.9 in 2003 to 8.2 in 2007, according to the European Foundation for the Improvement of Living and Working Conditions; and those who agreed strongly that they belonged to their neighbourhood rose from 71.6% in 2003-4 to 73.8% in 2008-9, according to the British household panel survey. But these two measures pre-dated the ONS’s new initiative—and the recession.

Richard Layard, an economist from the London School of Economics, a Labour peer, and a founder of the organisation Action for Happiness, took a gloomier view. Across every region, profession, age group, and ethnic group a considerable number of people are living very unhappy lives, he said.

Somewhere between 21% and 27% of people in every UK region have life satisfaction scores described by the ONS as “very low” or “low.” Across the UK as a whole the findings indicate that around 15 million people fall into these two categories.

Layard said, “Our priority as a nation should be to help improve the lives of those people who are struggling the most. These data show that significant numbers of people are anxious and unhappy regardless of their occupation, ethnic background, or where they live. We need a much wider concept of deprivation than is usually used in public debate. A person is deprived not only if they lack income but if they lack the other means to enjoy life.”


Cite this as: BMJ 2012;345:e5073


  • The ONS website includes maps of local happiness measures, special articles on health and housing, and the first results of the subjective measures (