Understanding global poverty: What does being poor mean?

Global poverty is a complex issue that has been the subject of extensive debate and research. This blog explores the concept of being poor from a global perspective, focusing on the International Poverty Line (IPL) and the ongoing discussions about how best to measure poverty worldwide.
 

Defining Global Poverty

People who live on less than $2.15 per person per day (2017 PPPs) are considered extremely poor. In practice, the local currency amount providing a purchasing power “equivalent” to these $2.15 is computed using PPP price indices. This approach defines the International Poverty Line (IPL), which aims to match the cost of living in different countries.
 

Criticisms of the International Poverty Line

Criticisms quickly arose when the IPL was first developed by Martin Ravallion. More than two decades ago, Reddy and Pogge suggested a different way to define poverty. They argued that we should look at whether people can afford basic needs like food and housing. For example, can they get 2000 calories of food per day and have a warm place to live?
 

Objective vs. Welfaristic Poverty Lines

There are two main ways to measure poverty: "objective" and "welfaristic." Objective poverty lines look at whether people can afford a fixed list of basic needs. Welfaristic poverty lines look at whether people are worse off than a certain standard of living. These two methods can give different results. This is illustrated in Figure 1, where the poverty threshold is assumed equal to px in the reference country with prices p. The local currency amount in another country with prices p’ is equal to p’x according to “objective” price-adjustments, but equal to p’x’ according to “welfaristic” price-adjustments. Observe that price-adjustments affect what it means to be poor. A person is objectively poor if she cannot afford the fixed list x. In contrast, she is welfare poor if she is worse-off than when consuming x.
 

Figure 1: Objective and Welfaristic poverty lines rely on different price-adjustments. 

^{Note: Budget sets are red, and the indifference curve is blue.}
 

Ongoing Global Poverty Debates

The debate about how to measure poverty continues. In a 2017 paper, Bob Allen proposed a fixed list of basic needs to define global poverty. In defense of the IPL, Ravallion and Ferreira argued that his list is arbitrary.
 

Recent Research and Findings

In the hope of shedding some light on these debates, I recently published a paper comparing these two types of poverty lines. The paper considers three properties: fairness, respecting preferences, and having a fixed list of basic needs. The third property, called “Objective Anchorage”, is new and requires the existence of a fixed list of achievements that provides an objective meaning to the line, independently of prices and preferences. In Figure 1, if x was an objective anchor for the poverty line, then any person whose achievements are in the South-West quadrant is poor and non-poor if they are in the North-East quadrant.

Implications of Different Poverty Lines

The paper shows that welfaristic poverty lines work well if everyone has the same preferences. But if people have different preferences, objective poverty lines might be better.

Illustrating the Limitations of the IPL

This limitation of the IPL is illustrated in Figure 2. The IPL’s welfare level is drawn in dashed blue. In country A, the IPL is implicitly based on 2000 calories of food and 3 square meters of housing ( z*_A ). In country B, it is implicitly based on 1700 calories and 4 square meters ( z*_B ). This means that someone who gets 1600 calories and 3.5 square meters  ( x_a ).might be considered poor in country B but not in country A. This shows that z*_B is not an objective anchor for the IPL (because x_a  << z*_B ). In fact, no list of achievements can serve as an objective anchor for the IPL. This result formalizes the criticism raised by Allen. This also highlights a limitation of the recommendation by the Atkinson Commission to produce basic needs interpretation of the IPL. To be sure, the problem has a conceptual cause, it is not the result of PPPs being imperfectly constructed in practice. 
 

Figure 2: The IPL cannot consistently reflect a fixed list of achievements under heterogeneous preferences.

^{Note: Budget sets are red, the indifference curve of individual a is blue and the welfare level defining the IPL is dashed blue.}

Anchoring the IPL in Preference-Specific Lists

The paper also shows that, although the IPL cannot be anchored in a fixed list of basic needs, it can be anchored in preference-specific lists of basic needs. This sheds some light on the arbitrariness criticism raised against objective lines. Indeed, anchoring the poverty line to a fixed list of basic needs is less appealing if this list is perceived as arbitrary and thus depends on preferences.

Conclusion

More research is needed. Welfarist poverty lines work well when everyone has similar preferences. Is it approximatively the case at low incomes? If not, objective poverty lines might be better. But then how can we find a consensual list of basic needs?  Recent studies on the cost of healthy diets are a good start, but we also need to look at other basic needs.