Covid-19 and the problem with ‘happiness’ research

Today is supposedly Blue Monday. Sixteen years ago, a travel agency published a press release claiming that the third Monday in January is the most depressing day of the year. The idea is superficially plausible. It’s mid-January. It’s cold. You’re skint after Christmas. You’re back at work after the weekend. There are worse candidates for the most miserable day of the year. But as a scientific claim, it was swiftly debunked and the academic responsible for it has since disowned it. It lives on as a way of filling space in newspapers and is probably most famous for being untrue. If it weren’t for the ‘I think you’ll find it’s a bit more complicated than that’ crowd popping up once a year to tell us that’s it’s pseudo-science, Blue Monday might have been forgotten by now.

The original research, such as it was, did not attempt to measure happiness, but there is a whole field of happiness economics that tries to do just that. The methodology is straightforward. People are simply asked in surveys how happy they are on a scale of 1 to 10. Variations on the question include ‘how satisfied with your life are you nowadays?’ and ‘to what extent do you feel the things you do in your life are worthwhile?’

It has been suggested that happiness statistics should replace Gross Domestic Product as the primary guide to policy. Every now and again someone writes an article or delivers a speech saying that governments should stop treating GDP as the only thing that matters and instead focus on wellbeing. Leaving aside the fact that neither the government nor anybody else thinks GDP is the only thing that matters, one of the problems with deriving policy from happiness stats is that they don’t respond to policies very well. The most famous happiness study was produced in 1974 by Richard Easterlin who showed that although rich people tended to be happier than poor people in the USA, overall happiness scores had been virtually static since the second world war, despite average incomes rising significantly. This became known as the Easterlin Paradox and it has been cited by those who claim that there is no further benefit from economic growth ever since.

Easterlin’s data have been questioned and there is clear evidence from other countries that happiness scores rise over time, but it is true that they have not risen very much. The Office for National Statistics has been running a happiness survey since 2012 and it shows that the number of people who were ‘happy’ had risen by 2019, perhaps because the economy recovered from the recession. But the increase was fairly marginal. The average happiness score went from 7.3 out of 10 in 2012 to 7.5 out of 10 in 2019.

If you look at most developed countries – and many undeveloped ones – the happiness score sticks rigidly between seven and eight out of ten and moves only slightly, if at all, over time. The question about happiness economics should therefore not be ‘why doesn’t economic growth affect happiness?’, but ‘why doesn’t anything affect happiness?’ A measure that is so unresponsive to a changing world isn’t much use as a guide to day-to-day policy. When the philosopher Jamie Whyte wrote about the happiness literature in 2013, he predicted that happiness scores in the UK would never go below seven and never go above eight. I agreed with him, but a black swan event last year proved us both wrong.

Covid-19 and the associated lockdowns had a measurable impact on wellbeing, as you might expect, with the average happiness score dipping to 6.4 in late March 2020 before rising during lockdown and returning to pre-lockdown levels in the summer. It dropped again in September and was 6.8 in December. It is unlikely to be going up much at the moment.

The average life satisfaction score was 7.3 in February and hovered around 7 until September when it fell again, briefly hitting 6.5 during the November lockdown. There was a more modest change in the number of people who felt their life was worthwhile, with scores dropping from 7.6 in February to 7.3 in December.

This all shows that wellbeing surveys are not useless. They reflect the mood of the country to some extent. It would have been very odd if happiness and life satisfaction scores had not fallen in a year that saw a pandemic, stay-at-home orders and the deepest recession in 300 years. 2020 was a rare year in which you could eyeball the wellbeing statistics and see the impact of events on the chart. But even in this extreme situation, average wellbeing scores only fell by a few decimal points. It seems that it would take an almost unimaginable catastrophe to move the dial much further.

What chance do we have of seeing the impact of public policy in more normal times when it takes the triple whammy of disease, isolation and recession to make a small but measurable difference to wellbeing scores? It is interesting to see how the pandemic has affected self-reported wellbeing to the nearest decimal point, but the surveys only confirm what common sense already told us. We knew that lockdowns made people less happy without having to turn to findings from social science.

And that is the problem with wellbeing research in a nutshell. It rarely gives us insights we couldn’t work out for ourselves. Happiness studies show that some things, such as employment, education and income, enhance happiness at a personal level, but this is fairly obvious and governments were focusing on these issues long before the first happiness survey was filled out. Other factors which create happiness, such as marriage, children, a belief in God and friendship, are beyond the government’s reach.

The lessons from the happiness literature for politicians are therefore banal or irrelevant. They suggest that the things that make people happy are either things the government is already working on (however unsuccessfully) or things the government can’t do anything about. If it’s a choice between focusing on GDP or focusing on wellbeing scores, GDP wins every time.