Definitional and measurement difficulties aside, it's clear that mental health is a serious social problem. According to the UK's Mental Health Policy Group in the UK 'one in six of all people suffer from depression or chronic anxiety, which affects one in three of all families'.
Mental health is of course difficult to quantify - to put it mildly. So how can government aim for improvement? It can readily measure and increase spending on treatments like psychological therapy, as the Group advocates. But how are we to know whether such spending is cost-effective, or even effective? Quantifying well being, or as economists put it 'social welfare' is a broader question than mental health and I have no easy answers. The World Bank and United Nations publish tables of social indicators of development, encompassing such variables as literacy, water supply and sanitation, and natural resources. These are obvious priorities for developing countries, and they do happen to be easy to quantify. But even in these countries mental health is a hugely important issue: should it be ignored just because it's difficult to measure?
One approach could be to take small randomised samples of a population, and measure their behaviour and responses to specific questions or psychological tests. This is the approach taken by crime surveys, which are thought to be more reliable indicators of criminal activity than numbers of crimes reported to the police. These surveys simply poll a sample of people and ask whether and how they have been affected by crime.
A Social Policy Bond regime should probably target mental health explicitly, but so too should our current political system. Like the physical and social environment, with both of which it's inextricably bound up, psychological wellbeing is in danger of being allowed to deteriorate by default, because nobody got round to quantifying it until the effects of its degradation were too catastrophic to ignore.