Saturday, April 19, 2008

What should we make of survey results showing no increase in happiness as income rises?

The issue I want to address is whether survey results showing little or no increase in happiness as incomes rise in high income countries mean that further economic growth is not worth having in those countries. I will assume that the observation of no increase in happiness as incomes rise cannot be attributed to factors such as changes in income distribution. (For a reference discussing this possibility with regard to the U.S. see here.)

First we should clarify the relationship between increases in income and increase in happiness in economic terms. It seems to me that in terms of economic theory what a person would be doing in comparing his happiness at different times would be analogous to calculating the change in value of his/her assets by discounting the future flows of income they are expected to produce. In computing his happiness level this person would ignore the past and just consider the expected future flows of goods of all kinds and the satisfaction that he/she and his/her family will derive from them.

How relevant are the happiness and satisfaction concepts measured in surveys to this concept of happiness as the discounted present value of the future? If surveys can be viewed as akin to barometers which measure how people are feeling about the direction of events affecting their well-being, they have some relevance. But if incomes just keep rising as expected, people will not necessarily record higher levels of satisfaction.

The economic concept of happiness as the discounted present value of the future seems somewhat at odds with prevailing psychological theory relating to subjective well-being – although a reconciliation may be possible. The set-point model, which is pivotal in the field of subjective well-being (SWB), suggests that while people may initially react strongly to events affecting them, their SWB subsequently returns to a stable level (set point) determined by their personality predispositions.

In its original form the set-point model implied that no matter what benefits an increased flow of goods and services may provide to individuals in helping them to achieve their goals, it will not result in any long-term improvement in measured SWB. This view has been modified in the light of evidence that set points can change as a result of a variety of factors, including widowhood and unemployment. A recent study by Frank Fujita and Ed Diener using panel data for Germany covering 17 years of life satisfaction judgements showed that almost 9 percent of respondents changed 3 or more scale units (on a 10 point scale) from the first five year average to the final five year average. Nevertheless, the majority of respondents showed long term stability in life satisfaction between the beginning and end of this period ( ‘Life satisfaction set point: stability and change’, Journal of Personality and Social Psychology, 88 (1), 2005).

It seems to me that these changes in set-points may correspond to the kinds of changes a forward looking person might calculate in his/her happiness stocks when life does not turn out as expected. For example, people who suffer large losses on the share market often tend to become more pessimistic about the future and discount future earnings more heavily as result - and this pessimism may also be reflected in a decline in their happiness assessments. Similarly a period of unemployment may cause a person to be more pessimistic in valuing his/her human capital and this will be reflected in a decline in happiness assessments.

To sum up, it does not seem to me that there are strong grounds in terms of set-point theory to expect avowed happiness to continue to increase in high income countries in which economic growth has come to be expected as a normal part of life. However, the discounted present value of the future could be expected to continue to rise under those circumstances. In other words, a future in which incomes continued to rise could be expected to be valued more highly than one in which incomes were stationary.

Those who argue that economic growth is not worth having when it has little or no effect on set point levels might change their view if they considered whether they would also be prepared to argue against other things, for example marriage, on the same grounds.

No comments: