← The CGD Podcast
The CGD Podcast
Whether and How to Measure Inequality Post-2015 – Alex Cobham
00:28:10
Will reduced inequality be included in some form in the post-2015 development goals? Should it be? And, if so, what is the appropriate yardstick? These were the questions explored last week at a CGD conference, Filling the Gap: Inequality Indicators...
Loading summary
Transcript43 lines
- [00:13]
A
Welcome to the Global Prosperity wonkcast. I'm Lawrence MacDonald. I'm pleased to welcome to the studio today my colleague Alex Cobam. He's a senior fellow at the center for Global Development. He happens to be based in our Europe office in London. And Alex has been here, among other things, to participate in a conference we held yesterday on measuring inequality. Alex, welcome to the show.
- [00:35]
B
Thanks a lot.
- [00:37]
A
It was a real opportunity for me to learn. I got to chair the panels. The first was on the ins and outs of inequality measures and the second panel was policy practitioners views on which measures are appropriate. You suggested we do the conference and I wondered if we could start out maybe by you explaining why inequality has suddenly become such a hot topic.
- [01:04]
B
Okay. Well, I'm not sure it's sudden, so perhaps it feels more sudden than it really is. My sense is that there's been a growing consensus. The further we've got down the road with the Millennium Development Goals, the. The more obvious it's been to people that inequality was really lacking. So we've had a set of absolute goals about lifting people out of absolute need, but without any serious inequality component other than the important gender equality aspect. So I think by about 2010, when people started thinking about what would come in 2015, there was the beginnings of a consensus. Now what's happened in the last few years is that's really developed and you've seen in the UN consultation in particular on post2015, a really broad consensus on the importance of inequality, including economic inequality.
- [02:00]
A
Now you were an advisor to that consultation. I don't understand exactly how it ran. Was there a separate consultation on various questions or one big consultation in which inequality kind of percolated up? How did that work?
- [02:13]
B
So there's a global consultation within which there are thematic consultations, including one on inequality. And in effect what you saw was some, I guess, pre populating of potential priority areas in the choice about which would be thematic consultation areas. I don't think there was any controversy over the choice of inequality for that and certainly it was very well engaged with, you know, both organizations more formally but also individuals globally in the online process.
- [02:47]
A
Richard Morgan, who joined us yesterday, was one of several policymakers on the second panel. He's a senior official at unicef. He said they've got the names, I guess, and contact information of about 6,000 people who participated in the global consultation on inequality and they've now established a network where they're staying in touch with these people.
- [03:07]
B
Yeah, and I think that network has the potential to really kind of keep this momentum that came from the consultation building. You know, there's a lot of academic experts and NGOs in that network and very actively so. And what you can see is the thoughts about organizing beyond the Copenhagen conference earlier this year where the consultation report was launched. There will now next year be a Pan African conference in Ghana that I think will be the next shoe dropping, if you like. So this isn't an issue that's going to go away as the discussions go on at the political level about what post2015 looks like. What we'll see is this network and others continuing to raise the issue of inequality. And we'll see where that takes us.
- [03:55]
A
There are those who would say that inequality itself is a value laden term. In preparing to moderate the two panels, I saw a comment on a blog post. It was anonymous. I don't know how widely this view is shared, but the person said if I were refereeing papers for a journal, I would reject anything that had inequality in the title because the proper non loaded term is income distribution. So when the UN sets out and says we're going to have a global consultation on inequality, aren't they sort of predetermining that this is a bad thing and that greater equality would be good? And is that a. I find that comment extraordinary.
- [04:36]
B
I mean, if ever a comment was value laden, it's that one, you know, income distribution is the thing or one of the things within which you might care about inequality. To say that you can talk about distribution as if that covers the fact of inequality. It's, you know, these are separate things, you know, to say.
- [04:58]
A
I should say the next comment on that blog was, this is the dumbest comment I have ever read.
- [05:04]
B
That doesn't surprise me and is more reassuring, I guess. Look, I think inequality has been a politically laden term and you can see where, you know where that comment is coming from. But I don't think that reflects certainly in the development debate where we are today. And I think you can see within, within high income countries like the UK and the us, inequality is a much more widely discussed political topic. But within the development debate, I think it's very clear. It's stronger than that even. It's clear that there is a sense now that perhaps wasn't there in 2000, that inequality is part of the challenges of development. It's not something that may or may not be a problem. Instrumentally, it is much more at the heart of what we think about as development. So I don't think there is the same sense of if you like the old left and right debate when you use the term inequality, at least within development.
- [06:08]
A
One of the things that was surprising to me and I think even for people who track it quite closely, we had keynote remarks, opening remarks from Rebecca Greenspan, who's the deputy head of the UN Development Program. And she told us about the toing and fro ing within the un, the process, the unresolved tension between a new version of Development Goals and Sustainable Development Goals. And a lot of it was sort of bureaucratic ins and outs. But she did point out that while the Global Consultation had come out very strongly in favor of having some kind of goal or measure of inequality in the post2015 goals, that the High Level panel had then said the focus should be on equality of opportunity rather than outcomes. And that's something maybe worth unpacking a little bit. But I think the most important thing is that there's now the, as I understand, correct me if I'm wrong, the Secretary General has done a report that has integrated the findings of the consultation and the High Level Panel, and now it goes to the UN General Assembly. So the. We're hardly at the end. We're maybe at the end of the beginning of what is going to be a protracted process of deciding whether or not the world should focus more on inequality as a goal or target or indicator. And what indicator should be used. Is that a fair summary of where we stand?
- [07:32]
B
I think there's one thing to add to that. What seems to be the case, you know, most of the discussion now is within the open Working Group, which is country representatives bringing together, if you like, the discourses around post2015 and the sustainable Development Goals. That's where the action is politically and where the decisions will ultimately be framed. What we're hearing is that within those discussions, the High Level panel report is almost invisible. So people use it to support bits and pieces they like from time to time, but it's really almost not present, apparently, other than what I consider to be one of the two most important pieces of the, the panel report. And that's this, this emphasis on disaggregated data, on making sure that in particular saying that no target can be considered met unless it is also met separately for the most marginalized group in terms of income, in terms of ethnolinguistic groups, in terms of gender or people living with disability and so on.
- [08:40]
A
This would apply to things such as education, various health care goals, access to fresh water, a whole range of issues. And one of the things that came out of the discussion yesterday, it seemed to me was, well, people who had hoped there would be a focus on inequality were disappointed that the high level panel seemed to be saying let's focus on opportunity. But then they came back around and said no goal will be met unless it's meant for the most disadvantaged groups. So it sort of comes back in through a back door, if you will, that in fact at least concern about the bottom of the distribution is going to be very much a part of it. There were those like Paul o' Brien who said, really the problem is not focusing on the bottom, the deprivation in the bottom of the distribution, it's the accumulation of wealth at the very tippy top.
- [09:26]
B
Well, I think that's perhaps in a sense two extreme positions where we might think of being in the middle. What the high level panel has done is say that in their view there should only be goals that are about meeting absolute needs, but that we have to recognize that that has to be done for the people who are most marginalized. People at the wrong end of inequality have to have their basic needs met. Now that's in a sense the weakest way of dealing with inequality because it continues to focus on absolute needs. What Paul o' Brien was talking about, focusing on people at the very top is interesting, and perhaps we wouldn't want to go as far as some of Paul's thinking, but the idea of having a focus on the level of inequality independent of whether people's absolute needs are met, because we recognize that that inequality is and can be bad for societies, bad for health outcomes, bad for economic growth, bad for the prospects of conflict. That's quite a powerful idea. And that, I think, is where the division really arises between people who think there should be a target or a goal on economic inequality and people who don't, who see that as, if you like, dangerously political.
- [10:43]
A
It occurs to me that perhaps the starkest contrast in what was very lively panel yesterday, in the first panel, besides you and Nancy Birdsall and our colleague Nora Listig, we had Martin Ervalian, former head of the research department at the World bank, and then notably Paul o' Brien from Oxfam and James Foster, who is a guru in all kinds of measurement and presented for me quite technical piece about the ins and outs of various ways to measure inequality. But then he said something quite provocative, which is if you're going to worry about inequality per se, then you have to be in favor of a policy solution that involves burning the money of the rich. What did he mean by that?
- [11:25]
B
Well, you know Yeah, I thought James's presentation was really fascinating. And what he talked through was, in a sense, measures which are about the mean level of income, let's say, or about the relative performance of different parts of the distribution. And he talks about, if you like, strong inequality measures. So a measure like the Palma that I would support, which is the ratio of incomes of the top 10% to the bottom 40%, his point was that a measure like that, you can make better not only by increasing the incomes of the bottom 40%, but actually simply by reducing the incomes of the top 10%. In other words, if you burn the cash of the rich, you make this indicator perform better. I think, although that's a legitimate concern to have. History tells us that there are very few situations in which the incomes of the rich fall, and therefore the likelihood of policies being enacted that prompt precisely that is relatively small.
- [12:31]
A
I mean, short of a communist revolution, the best one might hope for would be taxing the rich and having some redistribution. Nobody is going to seriously propose we improve the inequality indicator by destroying the wealth of the rich, maybe, you know, taking some of it and redistributing it. Yes, but indeed, it's not a sensible policy proposal short of, you know, short of communist revolution, which I don't think is in the cards.
- [12:58]
B
No. And politically, you know, we know the top 10% have most of the power, so they're not likely in most countries most of the time to end up in a. In a, In a burning their cash type scenario. I think, you know, having a target on that type of indicator would. Would encourage you very strongly towards measures to raise the incomes of the bottom 40%, which, you know, as a relative share of the national total, which is, you know, what I would say we want.
- [13:24]
A
Going to take a short break. When we come back, I want to talk to you about some of the measurement proposals that were put forward at the conference. Nancy Birdsall's proposal for greater focus on the median, which went first because it was the simplest, at least in my mind, and then you contended that, no, yours was even more simple. And that's the one described in your newly released CGD working paper with Andy Sumner. It's all about the. The Palma measure of income inequality. We will be back in a bit. Welcome back to the Global Prosperity wonkast. I'm Lawrence McDonald. My guest is Alex Koban. We're talking about a conference that CGD organized yesterday on the measurement of inequality. The first panel put forward some alternative measures. As I mentioned before the break, the one that Struck me simplest was Nancy Birdsel's idea for greater focus on the median. That is the amount that is earned by the person who's exactly in the middle of the income distribution. So 50% of individuals or households earn more and 50% of individuals or households have less. Alex, what's not to love about the median? I love the simplicity of that.
- [15:04]
B
No, I like the median a great deal and not only because Nancy is at the top end of the power distribution around here. The median captures all the difference between the median and the mean captures very neatly the extent to which income is loaded towards the top end. So in that sense it is a very powerful measure. And I recognise that within the US context in particular, the median income is often reported as a measure and it's, you know, it has some political traction in a way that perhaps isn't so common elsewhere, but I think is worth thinking about. So I'd have no objections to that.
- [15:47]
A
Okay. But the palma. Now make the case for the palma. You said it's the ratio of the income of the top 10% to the bottom 40% and I understand that there is a remarkable consistency that the middle 50%, that is the people who are below the top 10 and above the top 40 over time tend to have the same share of income. Is this across societies and across time?
- [16:09]
B
Yeah. So it's, you know, this finding by Gabrielle Palmer is, in the further exploration that Andy and I have done is remarkably resilient. What we find is in countries at very different income levels, but also within given countries over time, even as their income levels change quite, quite substantially, that middle 50% from decile 5 to decile 9 has a remarkably consistent share of national income. Also we find, using some.
- [16:41]
A
And how much is that?
- [16:43]
B
Well, it's round about 50%, tends to be slightly higher, but about 50%. So it's about half of income goes to that half of the population. And what that leaves you is a situation where in effect, most of the politics of distribution are, are about the tails. It's what happens between the top 10% and the bottom 40%. So using the work that Nora Lustig and others have done on the impact of taxes and transfers in Latin America, what we also find is that that middle 50 share barely moves throughout each stage of taxation and transfers. It's basically the extent to which there's a redistribution from the top 10% to the bottom 40, or in some cases not very much at all, but that's where the movement is. So across countries, across time and through stages of taxation and Transfers. It's the top 10% and the bottom 40%. So that's why we focused on this particular Kuznets ratio and called it the Palmer, because we recognize that finding central to it. What we've then explored, however, is the relationship with the Gini, which tends to be most commonly used. We find that relationship is very close.
- [18:00]
A
The Gini coefficient being a commonly used measure of inequality, mostly used by economists and really totally unintelligible, as far as I can tell, to ordinary people.
- [18:10]
B
Yeah, I mean, well, the Gini has a reasonably clear, intuitive explanation. We did discuss this a bit yesterday, but. But it does require you to sit down with somebody and say, okay, this is. And then go into a process of explanation which is, I think, you know, it's just a stage further than you have with the Palmer. It's this ratio.
- [18:29]
A
For those who are really junkies about this, you can watch the video and there's a wonderful piece where I say the genie doesn't seem very intuitive. And Martin Revalian says it's totally intuitive and proceeds to explain it in a way that left me feeling quite ashamed that it wasn't intuitive to me, except that then you came to my defense and pointed out the fact that it had taken him three minutes, or maybe it wasn't three, but it took some time to explain something, which I'm sure from Arten is totally intuitive, but I mean, I travel in a fairly well educated crowd. I would say everybody that I hang out with has got a college degree. And if I said to them, what is the Gini coefficient? Not only would they not be able to explain it, they wouldn't have a clue what it was. They wouldn't know that it was a measure of inequality. They would think it had something to do with magical spirits inside a bottle. They would have no clue.
- [19:16]
B
I mean, there may be an issue there in a sense that inequality has become, perhaps because it's at different times, it's been politically charged, it's become an area of work, I think, that's dominated by technical discussion in a way which actually takes it away from the understanding of people who aren't engaged specifically in that technical discussion. And that's one of the things we were thinking about when we came up with the Palmer. If you want a policy measure of inequality on which policymakers can be held accountable. So for the post2015 framework, that's certainly the type of measure I think you'd want. If it's a measure that needs to be explained to somebody before they can understand, then what has happened from one year to the next to see whether their government is doing what they would like them to do. The chances of that being effective for accountability are pretty small.
- [20:10]
A
Yeah. In fact, my comment in, you know, insertion as the moderator was among the problems with inequality is that those who are particularly disadvantaged have much worse education and very little ability to understand analytical questions. So if we're going to have a debate that involves them intimately and we're going to choose a measure that they can't understand, that's a problem.
- [20:35]
B
I mean, I think that's right. I mean, whether we are not wishing to cast aspersions on the intellectual capacity of large numbers of people, but at least, you know, the background necessary to understand the Lorenz curve and therefore what the genie is, is, I would say, non trivial.
- [20:51]
A
Well, we've got lan Pritchett's work in his forthcoming book showing that, you know, most children who graduate, or many, many children in the developing world who graduate from primary school can't read a simple paragraph or do simple sums. I mean, it's going to have to be pretty simple for them to understand what we're talking about. But I think they could understand the median and they might get the Palma. They are not going to get the genie.
- [21:12]
B
Yeah, no, I think that's fair. And the question is really two parts. You know, one is, do we want to measure inequality and measure it better than we're currently doing? And the second is do we want measures that are not only technically reasonable but also useful for accountability? And I think that's, you know, that's where the rubber hits the road, as you people say, on this side of the water. And that's, you know, where I think the Palmer, but also the median, you know, have a role to play. I think measures that people can grasp quickly, which are intuitively clear, are the type of measures we need for inequality in post2015.
- [21:50]
A
And while there was a lot of difference yesterday about which measures, there did seem to be a remarkable consensus that greater measurement and attention, including some sort of institutionalization of those measures that would be collecting the numbers and reporting them over time, would be useful. In fact, one of the participants in the policy panel who I've neglected to mention, Andrew Berg from the imf, came out towards the end and said, you know, he thought it made Sense to pick three measures. The share of income held by the top 10%, the median, and I can't remember the Gini or the Palma, but, you know, just pick three reported over time. And I think it was Andy who pointed out that for some countries this data is going to be lacking, but if there's a international commitment to collect and gather it, that creates pressure domestically to say how come other countries are reporting the share of income held by the top 10% and we're not?
- [22:47]
B
Yeah, no, exactly. And I think that that's critical. As you say, there was really a pretty broad consensus on the importance of measuring and monitoring inequality, whether or not there should be a goal or a target. And I think that's, you know, what I hope we at CGD and others can build on, you know, and start bringing that together. The data revolution that the high level panel have proposed is potentially very useful to that. If they're.
- [23:15]
A
What does that mean, the data revolution?
- [23:16]
B
Well, so what they've. What they've. The data revolution is what they've called the process that would be necessary to get the kind of disaggregation in order to have these separate criteria they've talked about. So to make sure that any given target is met for the poorest 20% or the poorest ethnolinguistic group, basically that means every year you're going to need the type of household surveys that allow you to see that that in turn is going to generate an enormous amount of additional information about the income distribution and all sorts of other things. If there is a serious commitment and the funding is put in place to do that, even if we don't have inequality goals and targets within the framework itself that will generate exactly this kind of information consistently of a high enough quality that we can do comparable statistics over time and across countries.
- [24:08]
A
I think Paul o' Brien would say all very well and good to have this army of household surveys, but you also need to disclose the tax returns and probably something you would be sympathetic to with your work on illicit financial flows, the bank account information, because the concentration of income in the top 1% is something that is not going to come out in these household surveys.
- [24:28]
B
Absolutely. I mean, the work that Piketty and Saez and Atkinson have done showing how different the top of the income distribution looks when you have tax return data demonstrates how important it is to have that in addition to household surveys. But the work that we've done on illicit financial flows tends to support a view that what's actually in the tax returns of the top 1% is itself an enormous understatement of the real income position. We have to be thinking very seriously about how we get much better data on that part of the distribution.
- [25:00]
A
Well, Alex, I think that we're out of time. I want to thank you for joining me and say that I look forward to continued CGD work on this. We did say at the conference yesterday that we've begun some exploratory conversations within CGD about the role that we might play in this. One idea that Nancy put forward is that we might, if you will, monitor the monitors, that is to track what the various international organizations are themselves doing. We had very good representation yesterday, United nations, imf, World Bank, Inter American Development bank on the policy panel and it seems to be very much alive issue within all of those institutions. Alex, any parting thoughts?
- [25:43]
B
Well, just this. In the panel presenting the Palmer, I referred to the tyranny of the genie. And this is the sense that we've got ourselves in a position where we think, at least at a technical level, that the Gini is basically the best measure of inequality. That's forced us to use data that's simply inappropriate to generate genies because we think that's what we need. It's also led to the exclusion of people from the discussion about inequality. The idea that we at CGD and others might get much more engaged in tracking the measures of inequality, but particularly with a focus on policy accountability, makes me feel very excited, very optimistic. The fact that, as you say, all these institutions were represented yesterday in this conversation. Thinking about these issues, it's possible, you know, the data revolution the high level panel talked about could actually be a real revolution in how we see and understand inequality. And that would be enormously powerful whether there's a goal or a target or not.
- [26:47]
A
I like the phrase the tyranny of the genie because it lends itself well to a cartoon punning on the idea of the spirit in the bottle. This has been the Global Prosperity Wonk cast from the center for Global Development. My guest today is Alex Cobham. We've been talking about various measurements, inequality, and particularly about a conference that we organized here at the center for Global Development yesterday. You can find the Wonkast online on itunes and on Stitcher. Just search for wonkcast or CGD and subscribe to hear a new interview every week. Until next time, I'm Lawrence MacDonald. Thanks for listening.