Imagine an industry had, through its own recklessness, caused over 10,000 deaths in the UK. Or imagine that a government minister had, by his own stupidity, caused hundreds of deaths. There would surely be demands for severe punishment.

But in a sense, this is just what has happened, with pitifully little outcry. This is because, in utilitarian terms, the misery caused by mass unemployment is equivalent to very many deaths.

Let's start from the ONS's survey on the question: "overall, how satisfied are you with your life nowadays?" On a 0-10 scale, the unemployed report an average of 6.47 whilst the employed report an average of 7.53, If we assume the causality runs from employment status to well-being (so that being grumpy doesn't cause folk to lose their jobs), this means that unemployment costs the individual 1.06 "utils." On the same scale, married people score 7.72 whilst the widowed score 7.26 - so the death of a spouse costs 0.46 "utils."

Now, this doesn't mean unemployment is worse than the death of a partner. A person's death represents a loss to his other family and friends. Let's assume (maybe generously) that 20 others feel as bad as the widow. This gives us a total bereavement loss from death of 21 x 0.46 = 9.66 points.

In utilitarian terms, then, the death of a loved one is 9.1 times as bad as unemployment. But this isn't the end of the story. Death is permanent but unemployment isn't. Let's therefore weight the two events by duration.

Latest labour market flows data (pdf) show that during Q4 2012, 595,000 of 2.48m people moved from unemployment to work, which implies an average unemployment duration of just over a year. For simplicity, let's call it a year. How much longer does bereavement last?

Not long! Andrew Clark and Yannis Georgellis have estimated that there is "complete habituation" over time to widowhood; it takes an average of one year for men and three for women to return to their pre-bereavement level of well-being. Let's split the difference and assume two years of breavement. Let's assume it takes as long for family and friends to adjust. This gives us a total bereavement cost of 21 x 2 x -0.46 = 19.3 "utils".

To this, we must add the fact that death represents a loss of well-being to the deceased. Let's call this a drop 7.72 utils, the average for a married person. And let's assume, arbitrarily, that had our subject not died when he did, he's have enjoyed another 10 years of this utility. This gives us a total utility cost of death of 96.5 utils.

This gives us a utility cost of death of 91 times that of unemployment.With unemployment now 944,000 higher that at the end of 2007, this implies that the welfare cost of recession is equivalent to the deaths of over 10,000 people.And if you think Osborne's misjudged fiscal policies have added more than around 24,000 to unemployment - and I suspect his critics would put the number far higher - then he is, in utilitarian terms, worse than Britain's worst serial killer.

Now, this calculation isn't intended to be remotely precise; it's a Fermi estimate. How should we respond to it? There are, at least, three possibilities:

1. It shows that there's something fishy about happiness research.

2. It shows that utilitarianism is wrong. Causing someone's unemployment is qualitatively different from causing someone's death - especially if you intend one but not the other - and you just cannot compare them on the same scale.

3. It shows that unemployment is indeed a social evil of huge magnitude, a fact which we don't realise because the jobless are socially and politically marginalized.

Personally, I'm not sure how to respond.