Global Poverty Research Group

Modelling and Measuring Subjective Well-Being


Background to the research

While economists have traditionally eschewed the use of subjective data, there is a recent explosion of interest in the use of subjective well-being or ‘happiness’ data by economists and other social scientists. An increasing number of analysts have come to believe that since happiness is generally considered the ultimate goal of life, economics is or should be about individual happiness. In his Lionel Robbins lectures at the LSE in March 2003, the labour economist Richard Layard stated: “The scientific study of happiness is only just beginning. It should become a central topic in social science”.

Insights from happiness research have far-reaching implications for government policy. Much economic research on such issues as income determination, provision of public goods and services, environmental quality, economic inequality and economic growth is motivated by a concern to raise people’s welfare. The availability of measures of subjective well-being now means that their effect on welfare can be directly assessed. For instance, it is possible to measure whether an increase in unemployment can be compensated, in terms of welfare, by the resultant decline in inflation. Subjective well-being data can also be used to improve our understanding of how best to alleviate poverty.

Research currently in progress

Against this backdrop, a new direction of research motivated by the GPRG programme concerns the relationship between the concept of well-being poverty and other concepts of poverty. Geeta Kingdon and John Knight have been working on a paper titled “Well-being poverty versus income poverty and capabilities poverty?”

Research questions being posed

The conventional approach of economists to the measurement of poverty in poor countries is to use measures of income or consumption. This has been challenged by those who favour broader criteria for poverty. These include the fulfilment of ‘basic needs’, the ‘capabilities’ to be and to do things of intrinsic worth, and safety from insecurity and vulnerability. To what extent are these different concepts measurable, to what extent are they competing and to what extent complementary, and is it possible for them to be accommodated within an encompassing framework?

Economic research on ‘subjective well-being’ (or ‘happiness’, or ‘satisfaction with life’) is sparse and recent but growing rapidly. It is apparent from the literature that there are two important gaps. First, reflecting the availability of data, there is little research on poor countries. Second, within any country, there is little research on the relationship between well-being and the notion of poverty. This research is intended to help fill these gaps.

Any attempt to define poverty involves a value judgement as to what constitutes a good or bad quality of life. We argue that an approach which examines the individual’s own perception of well-being is less imperfect, or more quantifiable, or both, as a guide to forming that value judgement than are the other potential approaches. We develop a methodology for using subjective well-being as the criterion for poverty, and illustrate its use by reference to a South African data set which contains information on reported well-being as well as much socio-economic information on the individual, the household and the community.

We argue that it is possible to view subjective well-being as an encompassing concept, which permits us to quantify the relevance and importance of the other approaches and of their component variables. The estimated well-being functions for South Africa contain some variables corresponding to the income approach, some to the basic needs (or physical functioning) approach, some to the relative (or social functioning) approach, and some to the security approach. Thus, our methodology effectively provides weights of the relative importance of these various components of well-being poverty.

Some preliminary findings

The importance of our variables representing relative position in society (the ‘relevant others’ being others in one’s locality or of one’s race) provides grounds for viewing poverty partly as a relative concept.

We find that average income of the cluster (a small local grouping of households) enters the happiness function positively and significantly. We argue that this can be interpreted as suggesting a small, tight community containing elements of altruism, mutual insurance, or the existence of a ‘social wage’ (or some combination of these positive aspects of a locality). In contrast we find a negative impact of average income of the (larger) district. This finding suggests a role for envy, rivalry or feelings of relative deprivation.

We find that different factors affect the subjective well-being of the poor and the non-poor. We find absolute income to be an important determinant of the well-being of the poor but relative income (defined in terms of race) to be important for the non-poor.

Ongoing research

Several issues require closer attention. Further research will focus on obtaining multiple measures of happiness, rather than relying on a single global happiness question. Better measures of happiness can generate greater confidence in the use of the subjective well-being approach to poverty. Research will also attempt to measure the size of habituation effects, which can tell us how much of any given increase in income or other resources actually translates into increased happiness. These questions will be addressed with the World Values Survey data as well as new data being collected in China under the GPRG project.

Some of our explanations for observed associations included altruism as well as envy/rivalry. These explanations can be tested by using ultimatum and public goods games in the same communities where the subjective well-being data are collected. Future research will attempt to do this.

Reseachers to contact for this project

Geeta Kingdon and John Knight