Shawn’s post on eliminating poverty got me thinking, and judging by the number of comments I’m not the only one.
on that post seems to have attracted outsized interest on that thread, and given the number of people talking about what I said I thought it would be helpful to clarify my point, as well as giving everyone a centralized point to debate what I was saying.
Shawn’s point, as I understand it, is that he feels too many on the left treat poverty as a diversity group, a population or subculture that should be treated like any other marginalized group. Sean’s rebuttal is that poverty itself is the real problem, not that poor people are badly treated. He feels the left should focus more on lifting the poor out of poverty than trying to make poverty a little easier to bear.
Let me be clear that I like the sentiment behind this. Having a low income sucks, and lifting up people’s incomes is a good thing. Some ethical questions are complicated, but this is not one of them. Sean and I might disagree on how we achieve this, but if the goal is “make the poor better off”, than sign me up.
My point of disagreement is on the specific idea of “getting rid of poverty”. My objection to that phrase isn’t ideological, but rather technical. At it’s core, what I do for a living is measure things. Because of this, I’m very sensitive to vaguely-worded goals. And this isn’t just a personal tick, if you don’t specify your goals properly, you have little to no chance of actually achieving them. The reason your government is still in Iraq is because it didn’t bother to specify its objectives in entering Iraq clearly enough, and now anything short of Iraq becoming a European country will look like a defeat, so they have to stay in to ward off final judgment. Make no mistake – vagueness kills, and the more important a policy goal is the more you need to be sure you have defined it properly, for without definition there can be no success.
It is not mean pedantry that leads me to criticize the idea of “getting rid of poverty”, but rather the concern that unless we’re clear about what we mean by poverty.
There are basically three poverty measures at large in the world today, and they each have their problems:
- The first is the UN’s extreme poverty measure, which amounts to about USD 2.25 per person per day. This is calculated on a basic subsistence-level standard of living, which makes it a nice concrete target, but far too low to be of use for a developed country. This is the trouble with fixed standard-of-living poverty measures, they become obsolete, necessitating moving the line. Every time poverty is eliminated, more people are defines as poor.
- The second is the US’s poverty line, which works out a minimal budget for food, and triples it. This one is also concrete, but it only looks at one budget item (if a large one for low-income people), crucially it doesn’t account for the cost of housing, which varies hugely in different parts of the US. This makes it not particularly useful as a poverty line, because it undercounts poverty in New York and over-counts it in rural areas.
- The third measure is used by most OECD countries, which is to set the line at 60% of median income. This is a bad measure because it has some utterly perverse outcomes. For one thing making everyone richer only reduces poverty if the bottom 50% get richer faster than the top 50%. Increasing incomes should not make your poverty measure worse. Now some of you may feel, that the gap between the rich and poor is important, and you are free to feel that way, but income inequality should be measured with an actual inequality metric (like the Gini Index) and not a poverty measure. The reason the 60% of median line is a bad measure of inequality is that it only looks at the bottom half of the income distribution. A disappearing middle class would actually cause poverty to drop under this measure because the gap between the middle and bottom of the income distribution would shrink. It’s entirely possible that, using a 60% of median poverty line, a smaller share of the population was poor in Medieval Europe than in modern Europe. That’s a fatal problem as far as I’m concerned.
I think some people took my objection to poverty definitions as an attempt to sidestep the issue of poverty by deploying rhetorical chaff “We can’t fix poverty, because who knows what poverty is anyway?”. That’s not what I’m trying to say here, my goal is to clarify, not obfuscate. To that end, let me outline how I would go about measuring poverty:
- Conduct wide-scale social research outlining what goods and services people consider essential to a decent life. Get beneath platitudes like “good healthcare” and “decent housing”. What kind of healthcare? What constitutes decent housing?
- This research allows you to create a Poverty Basket – a set of goods and services that society feels are essential to live a decent life. In practice, this basket will be different in different parts of the country, but I think that nuance is a plus.
- Track the price of the Poverty Basket above over time, weighted by their share of consumption for low-income people. This is less work than you may think, because your statistical agencies are already tracking prices and consumption of goods and services to calculate CPI. This will just use the same data.
- Periodically (say every 10 years or so), revise your Poverty Basket. But, and this part is important, label your poverty line with which basket is being used, and calculate a few of the older poverty lines too. That way you can see how poverty (2013) compares to poverty (2003) and poverty (1993).
This method follows Adam Smith’s definition of poverty: that poor people are those who cannot afford the necessities of life, or things society considers it to be indecent to be without. In The Wealth of Nations he uses leather shoes and linen shirts as an example, these things are not necessary to life (he contrasts contemporary England with ancient Rome and Greece, as well as other of contemporary Europe in this regard), but nonetheless most English people of his day would consider someone who didn’t have those things to be poor. But not only does it allow for the standard of poverty to change over time, it does so in a transparent way, so you know how much of the change in poverty rates is due to changing standards and how much is changing incomes.
But for any non-trivial poverty measure, the standard will change as incomes rise, and that makes any talk of eliminating poverty problematic. Rather than aiming at a moving target, I suggest focusing on the reality behind the poverty measure – the suffering that poverty causes, and the lack of opportunity it implies. These are worthy goals to pursue, so at even if the poor are always with us, they will at least have a better life.