Usually, we measure poverty on the basis of objective numbers about income or consumption. Income or consumption levels are put on a continuum from lowest to highest and somewhere along the continuum we put a threshold that indicates the difference between poor and non-poor. For example, the Indian government uses a consumption threshold of 2,400 calories a day in rural areas and 2,100 in urban areas. The World Bank uses an income threshold of one dollar a day (corrected for purchasing power).
There are numerous disadvantages to these objective approaches. One is the inevitably arbitrary positioning of the threshold. One dollar a day, even after correction for purchasing power, means different things to different people in different areas, circumstances, groups etc. Calorie intake also means different things to different people, depending on people’s way of life etc. Moreover, income levels are notoriously difficult to measure (poor people in particular have a lot of informal income, e.g. “income” coming from all sorts of assistance from relatives etc.). Consumption as well is a difficult measure: it doesn’t necessarily have to mean just calorie intake for example. Poverty can mean a lack of non-food consumption. And if you focus on calorie levels after all, you’ll miss the issue of the quality of the food.
Also the third most common approach to poverty measurement suffers from some disadvantages. This approach, also called the multidimensional approach, tries to assess to what extent people suffer from a series of different types of deprivation: do they have access to water, to electricity, are they literate, malnourished etc. Rather than purely quantitative these measurements can be qualitative: a binary yes/no is often enough. Unfortunately, also this measurement system has some drawbacks: it fails to distinguish between deprivation and choice; there’s necessarily a level of arbitrariness in the determination of the “basic needs” or forms of deprivation that are measured; and these needs are often overly general, obscuring some very specific needs for some people in some areas or groups.
That’s why people have been searching for alternative measures of poverty. One such alternative is the use of surveys that ask people about poverty. You could ask people what they believe is “the smallest amount of money a family needs each week to get along in this community”, “what is the level of income below which families can’t make ends meet” etc. That would remove some of the arbitrariness of the cutoff line between poor and non-poor, and putting that decision in the hands of the people rather than the scientists.
Or you could also present people with evocative descriptions of different family situations, of types of families according to their level of income or consumption or according to the type of deprivation. People would then have to decide for every family situation what they believe the standard of living is and which situation can be described as “poverty”. That would specify what poverty means to people. And what it means to people is much more important than what it means to researchers and statisticians.
A disadvantage of this subjective approach is the wellknown effect that people’s income levels affect their judgments about income adequacy. In short, relatively rich people overestimate the level of income inadequacy. A solution to this problem could be to ask only poor respondents about poverty, on the reasonable assumption that poor people are the best experts on poverty. But that’s a circular reasoning: you already think you know what poverty is before you start asking about it. Since you focus only on the poor, you’ve already decided what poverty is.
An advantage of the subjective approach is that researchers don’t have to list basic needs or types of deprivation in order to assess what poverty is; people tell you what poverty is. There’s also no need for researchers to specify regionally or socially undifferentiated and general cutoff levels of income or consumption below which people are considered to be poor.
51.053562
3.720867