There are lies, damned lies, and, the saying goes, statistics. And then there are statistical claims made by Polly Toynbee.

Toynbee embraces the government’s plan to publish figures on the “well-being” or “happiness” of the nation, but mocks this as an own goal, because “most answers to the happiness conundrum point leftwards, 180 degree opposite to the Cameron/Osborne direction.”

According to Toynbee, “every model shows that the most unequal societies are the least happy. […] Even the rich in unequal countries are less happy than the best off in more equal countries. You see it in their miserable politics, the hatred, the fear, the exaggerated dislike of the poor”.

One of the sources which Toynbee quotes is, of course, The Spirit Level. This seems a safe bet because the authors, Pickett and Wilkinson, claim that almost every good (bad) thing in the world is negatively (positively) correlated with income inequality. But even a safe bet can go wrong, and for Polly Toynbee this one has: happiness is the one variable for which even the Spirit Level authors do not repeat their usual claim. There is no graph anywhere in the Spirit Level on the relationship between self-reported happiness and income inequality. The authors have good reason to omit this because, as Christopher Snowdon shows, there is no relationship whatsoever between inequality and happiness. When looking at cross-country data, self-reported happiness does have a correlate, but Pickett and Wilkinson do their best to obscure it because it is the very thing they want to drastically curtail: overall prosperity.

So on a macro-level, happiness data provide scant support for a left-wing agenda: they are correlated to prosperity and unrelated to inequality. Mind you, this does not mean that free-market liberals should use these flimsy data to support their case. Some happiness studies show that in Western societies the happiness of women has not improved in the decades since the late 1960s, the time in which women achieved equal rights and economic opportunities. Would anyone conclude that these achievements should be reversed again, because of their “failure” to raise women’s happiness? Some of these studies show that education can decrease happiness (probably by raising aspirations). Would anyone conclude that education should be discouraged through taxation, like tobacco and alcohol? Some studies show that having children reduces happiness. Would anyone see this as a plea for the introduction of a “child tax”? Oh, and let’s not even get started on the relationship between ethnic diversity and self-reported happiness in a neighbourhood.

So what on earth makes Polly Toynbee think that happiness indices would automatically take us towards her cherished agenda? The reason is that Happiness Economics has long been the hobby-horse of people pushing an agenda of curbing consumption, stalling economic growth and redistributing income. In the early days of happiness economics, only rudimentary data was available, and as Toynbee puts it, “all the evidence showed that a doubled GDP in 30 years hadn’t made people in developed countries any happier.” What she omits is that since then, more datasets from larger samples have become available, refuting and reversing most of the initial conclusions. The opponents of prosperity responded by burying their heads in the sand and pretending this was not happening.

And this is exactly the danger with using happiness indices to derive real-world policy conclusions. People will look for those pieces of “evidence” which support the type of policies they would have promoted anyway. And they will find them.