What is Poverty? (6): Absolute or Relative Deprivation?

Is poverty a lack of basic resources, or instead the unequal distribution of resources? Is it the absolute income or wealth of people that matters, or the fact that other people are richer and can afford more luxuries? Intuitively, one would go with the former of those options: people are poor when they are starving or homeless or when they lack some other basic necessity. People can have enough of all basic necessities and still be a lot worse of than some group of ultra-rich. One the other hand, what counts as a basic necessity is not always obvious, and people may form their ideas about necessities in light of the lifestyle of the average member of their society at the current moment in history.

This is another way of expressing the difference between absolute and relative poverty. In the US, it’s common to defend and use an absolute definition of poverty (as does the World Bank), whereas in Europe the focus is on relative poverty. The difference is an important one, because the use of one or the other definition of poverty determines who counts as poor or not. Hence, it also determines who gets government assistance.

Now, something strange is going on here. Intuitively most people favor an absolute definition of poverty – that’s what my own intuition and an unscientific sample of friends tells me –  and yet, if you ask people what one needs to get by in life, the amounts they give you are far above commonly used absolute poverty thresholds. In fact, these amounts are closer to median income. And as median income rises, the amounts supposedly necessary in order to get by also rise. This tells us that people actually use a relative notion of poverty. And this is true even for the country that is supposedly most naturally in favor of an absolute notion of poverty, namely the US.

I made a similar point here. More posts in this series are here.

What is Poverty? (4): Does the Concept of Poverty Collapse Under the Weight of Historical Comparisons?

Many of the people who are considered poor in developed countries have a higher living standard than the average middle class citizens of some centuries ago. If we bracket the minority of the extremely poor in developed countries (the homeless for example), poverty today seems to be a relatively comfortable position to be in, once you see it in a historical perspective.

The same is true for people in poor countries. In 1820, average income per person was low everywhere in the world: about $500 in China and South Asia, and about $1000-$1500 in Europe (1993 US$ PPP). In developing countries today the range is between $1000 and $3100 (the world average is about $6000, the US has more than $40,000). So, the poor of today are equally well off or even better off than the average world citizen 200 years ago. 75% of the world’s population lived on less than $1 a day in 1820. Today, almost no one does in the West. In China it’s less then 20%, in South Asia 40%, in Africa half. Globally, it’s less than a quarter. Historically, almost everyone was poor; today it’s a minority.

So it seems almost futile to talk about poverty today. What is defined as poverty now was the normal way of life not so long ago. However, if that’s the way you want to go, the concept of poverty evaporates. You’ll always find someone who’s worse off. You just need to go sufficiently far back in time (or move in space) to find people who are more deprived and who make the current poor (or the local poor) seem relatively well off. The baseline is then the caveman and everyone else isn’t really “poor”.

Hence, if you want to keep talking about poverty, you can’t engage in historical comparisons. Does that mean that poverty can only be measured against the current average standard of living? That poverty is a percentage of current median income? In that case, there will always be poverty and the fight against it is a Sisyphean task. I’m not entirely convinced of the usefulness of the concept of relative poverty – that you should compare people’s living standards to society’s average standard (where poverty becomes basically income inequality) – and the historical rather than spatial version of relative poverty reinforces my doubts. However, I know that people commonly see poverty as a relative thing and that they may feel deprived because they compare themselves to their living compatriots and not only because they are below a certain absolute level of income, consumption or capabilities. Conversely, the middle classes of some centuries ago, even if they had the same standard of living as some of today’s poor, felt good about themselves because they looked at the poor of their time and felt that they had done comparatively well.

Still, relative poverty is not the only solution to the problem of historical comparisons. Poverty can be measured relative to average historical or current standards of living, but can also be measured by comparing consumption, income or capabilities to a commonly accepted absolute minimum level (for example a minimum amount of calorie intake).

In the latter case, it’s not important how rich the rich really are, or what the median income is, or how poor the poor were centuries ago. It’s important to know what are people’s basic needs, how much they cost, and how many people currently can’t buy the stuff to fulfill their basic needs. Of course, these basic needs can’t always be determined scientifically (as in the case of calorie intake) and some level of arbitrariness is unavoidable. A lot depends on the capabilities we believe are necessary in order to have a minimally decent life, and that’s controversial.

I also understand that social norms evolve and that basic needs can change over time. Several centuries ago a microwave and a cellphone were obviously not a basic need; now you will be considered poor if you lack these tools. In a pre-modern agrarian society, you would have been considered poor only when you were on the brink of starvation. You didn’t need technological tools, child care, education etc. in order to have a minimally decent life, because no one had those things and your functioning in the economy didn’t require them. Today, if you don’t have them, you’ll feel excluded, less than normal, weird, “trash” and in certain cases you’ll end up deeper in poverty because you’ll have a hard time finding a job if you don’t have a car, a cell phone or child care.

Also, why shouldn’t we become more ambitious over time? Should we be content if we’re able to avoid only the worst kind of deprivation? Or should we try to continually improve many different capabilities? The latter is I think a sign of civilization and progress. That doesn’t mean we should scatter our attention and forget to focus on the worst deprivation. It only means we shouldn’t stop after we’ve dealt with the worst. And we haven’t dealt with the worst simply because the percentages of those worst off have been coming down (see the numbers cited above). Indeed, a smaller share of the world’s population suffers from low income than some time ago. But because of population growth – which is a good thing resulting from higher life expectancy rates – the total number of people with low income is now higher. And total numbers also count, just as much as percentages. As Thomas Pogge has argued, the Holocaust wouldn’t be any less horrible if it turned out that the number of people killed was a smaller percentage of the world’s population than initially thought.

The Ethics of Human Rights (38): Should People Be Equal or Should They Have Enough?

There are people who believe income inequality is a major problem – and I’m one of them – and there are others who say that the real problem isn’t a relational one but rather one of absolute means. Harry Frankfurt for example argues that it’s not important whether a person has less than another regardless of how much either of them has. What is important is whether people have enough of what they need for a decent human life.

Sufficientarianism

This so-called sufficientarian approach – as opposed to the egalitarian one – is supposedly not comparative or relational but humanitarian. It focuses on the alleviation of absolute suffering and deprivation instead of relative inequalities. Rather than diminishing the distance between the worst off and the best off, it wants to improve the situation of the worst off. The latter goal can be the result of the former, but doesn’t have to be. Or it can promote the former but doesn’t have to. For example, imagine a society where incomes are highly unequal but where none of the people at the wrong side of this inequality are below a threshold value of wellbeing (the threshold determines the difference between suffering and non-suffering). So, according to the sufficientarian approach, there’s no need to diminish inequality in such a society. There’s no need to do anything, in fact. Conversely, you can have a society – not so imaginary perhaps – with low levels of inequality but almost all of the people live below the threshold. Tinkering with inequality will not do much good in that society. What you have to do is raise the living standard of almost all the population.

Income inequality for sufficientarians is relevant only to the extent that the wealth of those who are better off is a useful means to alleviate the suffering of the worst off. Diminishing inequality isn’t a goal in itself, and inequality doesn’t do any harm in itself.

It’s an appealing view, and I have been tempted by it myself. Even if you believe, as I do, that inequality can be harmful no matter what the income levels of the worst off are (harmful to democracy for instance) and that more equal societies almost always do better, you may still agree that the most urgent priority is the suffering of those who are worst off. Income inequality should then only be tackled afterwards. Anyway, tackling that first priority is a good step on the way to more equality; helping the worst off will reduce inequality almost automatically I would believe.

However, appealing though it may be, is sufficientarianism really all that much different from egalitarianism? As soon as you talk about the “worst off”, you have already engaged in comparative and relational analysis, by necessity. Another problem with sufficientarianism is the setting of the threshold: that is bound to be somewhat arbitrary. Of two people in very similar situations only one will receive help. You may say that cut offs are always inevitable, and perhaps that’s true, but sufficientarianism makes them painfully obvious to those concerned. People just above the threshold are told that they don’t matter, even if their neighbors who are just below matter a great deal. And finally, basic needs change over time, hence also the meaning of “suffering”. Will sufficientarians keep the threshold fixed, or allow it to rise over time? In the latter case, the difference between their approach and that of egalitarians is again rather small.

Prioritarianism

Some of these problems are sidestepped by a similar view called prioritarianism, made famous by Temkin and Parfit: benefiting people is more important the worse off they are. No need for a threshold here. When having to choose between two policies, you always take the one that is best for the worst off, whatever their level of well-being.

Benefits to the worse off matter more than benefits to the relatively better off. A benefit has greater moral value the worse the situation of the individual to whom it accrues. If we have some benefit to distribute, and this benefit has a value of x (no matter how we define “value”), it’s better to give this to the worst off than to anyone else. Strict utilitarianism, as opposed to prioritarianism, doesn’t care about who gets the benefit of x, because who gets it doesn’t change overall well-being. However, utilitarianism does take into account the possibility of the diminishing marginal utility of something: lots of money for a rich person isn’t as useful as the same amount of money for a poor person. But when comparing two people who would benefit just as much from such an amount of money, utilitarianism – as opposed to prioritarianism – doesn’t care who gets it; either person, the better off or the worse off, can get it. Prioritarianism doesn’t merely say that the worse off person should get it, but also says – contrary again to utilitarianism – that we should benefit the worse off even if that means diminishing total well-being; e.g. we can harm the interests of the better off if that means improving the well-being of the worse off:

Imagine choosing between two outcomes: In outcome 1, Jim’s well-being level is 110 (blissful); Pam’s is -73 (hellish); overall well-being is 37. In outcome 2, Jim’s well-being level is 23; Pam’s well-being level is 13; overall well-being is 36. Prioritarians would say that outcome 2 is better or more desirable than outcome 1 despite being lower than outcome 1 in terms of overall well-being. Bringing Pam up by 86 is weightier than bringing Jim down by 87. If we could move from a society described by outcome 1 to one described by outcome 2, we ought to. (source)

So prioritarianism avoids some of the counterintuitive implications of strict utilitarianism. And it also avoids the equally counterintuitive implications of strict egalitarianism. The latter may demand bringing everyone down to the level of the worst off while benefiting no one. Prioritarianism on the contrary does not propose a move toward more equality if that doesn’t benefit the worse off. And finally, it avoids some of the practical problems of sufficientarianism, while maintaining the appeal of the sufficientarian focus on the absolute deprivation of the worst off.

Measuring Poverty (10): Multidimensional Poverty

Poverty can be many different things. It can be different things to different people in different countries or circumstances. It can mean one thing for people in Africa and another for people in the favelas in Rio, and still another for those in the inner-cities in the U.S. It’s probably different for men, women and children. It can be absolute deprivation or relative poverty (i.e. inequality). It can be insufficient income or insufficient consumption. It can be a lack of one thing or another. For some people it means inadequate healthcare, for others it means insufficient water. It can be physical suffering or the stress inherent in insecurity. It can be malnutrition or a lack of self-esteem. It can be illiteracy or child mortality. Etc.

Most poverty measurement systems try to keep it simple. The most common systems just measure income. Poverty is then insufficient income (typically below $1-a-day, corrected for purchasing power; this measures the number of people incapable of buying a basic basket of commodities). That makes sense, because without sufficient income, you’re likely to experience child mortality, illiteracy, malnutrition, inequality, water shortages, stress, insecurity and all the other nasty things that come with poverty.

However, it is important to know those details of poverty. Two people who both have an income of less than one dollar a day, may experience very different consequences: one may be deprived in lots of areas, the other one maybe in just a few. One may lack good health, may be starving and may be illiterate. The other one may just be illiterate. If we want to help people, it’s important to know what the exact nature of their problem is. Which we don’t if we just focus on how much their income is.

That is why some researchers at the Oxford Poverty and Human Development Initiative at the University of Oxford have tried to come up with a so-called Multidimensional Poverty Index (MPI).

The index seeks to build up a picture of the prevalence of poverty based on the fraction of households who lack certain basic things. Some of these are material. Does a family home have a dirt or dung floor? Does it lack a decent toilet? Must members of the household travel more than 30 minutes on foot to get clean water to drink? Do they live without electricity? Others relate to education, such as whether any school-age children are not enrolled or whether nobody in the family has finished primary school. Still others concern health, such as whether any member of a household is malnourished. A household is counted as poor if it is deprived on over 30% of the ten indicators used. Researchers can then calculate the percentage of people in each country who are “multidimensionally poor”. (source)

Such a multidimensional approach has the advantage of identifying which specific aspect(s) of poverty is/are most common in certain areas or among certain groups of people. It shows how people are poor, and what contributes most to poverty in a specific place and among a specific group. This will obviously greatly enhance response capacity. Rather than just trying to generally increase income, we can target our efforts more specifically: in one area or among one group of people we know that we should focus on nutrition; elsewhere we know that we should focus on literacy for instance. The MPI also shows us how different aspects of poverty overlap: for example, how many people who are illiterate also have health problems?

If 30% of people are malnourished and 30% of children are out of school, it would be useful to know if these deprivations affect the same families or different ones. (source)

The approach also helps us to distinguish between deprivation and choice. People may actually prefer mud floors to concrete floors in some places, and don’t consider having a mud floor as a form of deprivation. It also helps to identify the depth of poverty: deprivation along a wide spectrum of indicators means that poverty is deeper.

Unsurprisingly, the results of the MPI are substantially different from traditional poverty measurements:

Also the totals are different:

About 1.7 billion people in the countries covered – a third of their entire population – live in multidimensional poverty, according to the MPI. This exceeds the 1.3 billion people, in those same countries, estimated to live on $1.25 a day or less, the more commonly accepted measure of ‘extreme’ poverty. (source)

One of the disadvantages of this new approach is the weighting of the different measures: there’s inevitably some arbitrariness involved. Is the death of a child equivalent to having a dirt floor? Worse? How much worse? More criticism of the MPI is here.

There’s a really cool interactive map of the MPI here.

What is Poverty? (3): Vulnerability

Definitional discussions about poverty have convinced me that there are actually different types of poverty. I don’t think that all types are equally urgent policy problems, although they’re all worthy of attention (personally, I think poverty as absolute material deprivation is the one to focus most attention on, rather than relative poverty, poverty as a mental harm etc.).

One type I haven’t discussed a lot is poverty as vulnerability. This isn’t actual poverty in the sense of existing destitution; it’s rather the presence of a high level of risk of poverty, a high level of insecurity or a high probability of becoming poor. Indeed, it’s fair to say that poverty isn’t merely current insufficiency of income or consumption, but also the absence of stable and predictable income or consumption.If you can eat today you’re not poor according to some measures (other measures would correctly include more than just food). But what if there’s a good chance you can’t eat tomorrow? Wouldn’t it be correct to call someone living with such a high risk a person suffering from poverty? People who have enough to eat and who have shelter, but who would starve if they faced unexpected costs or events, such as a health crisis, a flood, a drought, unemployment etc. should be considered poor.

Of course, you might think we’re all living under such risks. Even the wealthiest among us can’t be sure not be become poor tomorrow. Hence we’re all vulnerable, but some are more vulnerable than others. The issue is then how to measure vulnerability and risk. The risk is higher for some than for others, and the consequences when the risk events occur are tougher for some than for others, but how can we know and measure this? We can look at resources and savings for instance. Some people, and some people in some countries, are better armed to deal with risks. They can insure themselves, or their government insures them (unemployment insurance for instance). They may be able to smooth over these events: sell some assets, take a loan… Other people can’t insure themselves, or they live in a country that doesn’t provide public insurance, or they can’t smooth without jeopardizing their future wealth.

This vulnerability is not just a risk for the future; it creates problems here and now. When the risk is perceived – correctly or not – as being very high, then it produces fear, stress and feelings of insecurity. That’s not deprivation or poverty but it sure isn’t pleasant. Those feelings can also be self-fulfilling: people may take irrational precaution measures, counterproductive family planning decisions etc. So poverty as vulnerability is a real problem. Perhaps not as urgent as absolute destitution, but not without importance.