It's an unlikely match, I know, but I have a friend who is a Jehovah's Witness. One day, after overcoming a certain amount of embarrassment on both sides, he asked whether he could try to persuade me to let Jesus into my life. I promised him a fair hearing. Some of what he said made sense, but the story fell apart for me when he claimed that in biblical times "people were a lot more moral than they are today". I argued that half the Old Testament appears to be a record of divinely inspired genocide, as God's people sought to exterminate the other tribes they encountered. "Ah yes," said my friend, "but there was a lot less fornication."
This was the point at which I understood that people of the same neighbourhood can entertain very different conceptions of morality. It is a theme on which the psychologist Jonathan Haidt expands, fascinatingly and persuasively, in his book The Righteous Mind. And it is the theme on which he stumbles, stupidly and disastrously, when seeking to apply his findings to politics, as he did in the Guardian last week, and as he has done to great effect within the Democratic party.
Drawing on a wealth of experimental evidence, Haidt argues that we tend to make moral decisions on the basis of intuition, rather than strategic reasoning. We then use our capacity for reason to find justifications for the decisions we have already made. "Our moral thinking," he says, "is much more like a politician searching for votes than a scientist searching for truth."
Our intuitions are shaped by, and help to bind, the groups or tribes to which we belong. The moral codes of progressives in the west are built, Haidt says, on just three foundations: the pursuit of care rather than harm, of liberty rather than oppression, and of fairness rather than cheating.
Conservative politicians, by contrast, have "a broader variety of ways to connect with voters", as their moral narrative is built on these foundations plus three more: loyalty/betrayal, authority/subversion and sanctity/degradation. "Most Americans", he tells us, "don't want to live in a nation based primarily on caring."
Rather than voting on economic issues, working-class people have been "voting for their moral interests". He argues that "when people fear the collapse of their society, they want order and national greatness, not a more nurturing government". This helps to explain, he says, why "working-class people vote conservative, as most do in the US".
Haidt's analysis has been taken up enthusiastically on both sides of the Atlantic. But his admirers appear to have missed something. While the psychological findings he presents are well-attested and thoroughly referenced, he offers not a shred of evidence to support his political contentions, either in the article or in his book. His claims are unsourced, unsubstantiated and plain wrong.
As Larry Bartels, professor of political science at Vanderbilt, Nashville, points out, the political views of white working-class voters in the US "have remained virtually unchanged over the past 30 years". Voting for the Democrats by those on low incomes has in fact increased. Political decisions in this class are still shaped overwhelmingly by economics. On what Haidt calls "moral" values, there is "no evidence of any shift" in this group. It is only among more affluent voters that the Democrats have lost support. "Economic status has become more important, not less important, in structuring the presidential voting behaviour."
The real issue is surely turnout. In the US it has been low for a long time: between 50% and 60% for presidential elections and 30% to 45% for mid-term congressionals since the second world war. In the UK it has slipped dramatically, from 84% in 1950 to 65% in 2010. An analysis by the Institute for Public Policy Research shows that the collapse has occurred largely among younger and poorer people. "Older people and richer or better educated people … are now much more influential at the ballot box".
The major reason, the institute says, is the "'low-stakes' character of recent elections": the major parties "fought on quite similar platforms". The biggest decline in recent political history – from 1997 to 2001 – lends weight to this contention. In 1997 the young and the poor believed they faced a real political and economic choice. By 2001, Blair had moved Labour so far to the right that there was scarcely a choice to be made.
If Haidt and his admirers were right, the correct strategy would be for Labour, the Democrats and other once progressive parties to swing even further to the right, triangulate even more furiously, and – by seeking to satisfy an apparent appetite for loyalty, authority and sanctity – to join the opposing tribe. But if the real problem is not that working-class voters have switched their voting preferences but that they are not voting at all because there's too little at stake, the correct political prescription is to do the opposite: to swing further to the left and to emphasise not "order and national greatness" but care and economic justice.
Haidt's unsupported assertions suggest that he, too, is using reasoning to justify his intuitions. I am sure he is right when he claims that we all have this tendency. But we might have expected him, of all people, to try to think like "a scientist searching for truth".