Posts Tagged 'global poverty'

The World Bank Responds on Chinese Poverty Forecast

In response to my earlier post on the World Bank’s China poverty forecast I received the following reply from the Bank’s poverty team, which (at their request) I am happy to post here in full. Below their response I offer a few further comments.

Thanks for your interest in the World Bank’s poverty estimates and forecasting. As we often have said, these poverty forecasts are only reliable at the aggregated level, not at the individual country level. That’s why we do not normally release country-level forecasts.

That said, the Bank’s poverty team has no serious concerns about the quality of China’s 2008 survey data. Nor do we want to hide our poverty projection for China in 2015.

The easiest way to answer your questions is to show you our estimates. I am attaching the poverty estimates for China here which has exactly the same format as the GMR 2011.

Please notice that China conducts rural and urban household surveys separately. The national poverty estimates are the population weighted average from rural and urban poverty measures. China’s National Bureau of Statistics will soon release new urban population share from the 2010 census, and the time series of urban population share between 2000 and 2010 will also be updated. So, too, will the World Bank’s poverty estimates for China. But as of now, these are our best estimates.

Your blog said that “the most recent poverty survey for China, which covers the year 2008, has been the subject of considerable rumours in the past. Specifically, there was quite a long delay from when many people thought the results would be released to the public until when they actually were, which was just a few months ago.”

That is simply wrong. The delay in the public release of the World Bank’s poverty numbers had nothing to do with China. Rather, it was due to delays in the access of the data for a number of countries in Africa.

Sincerely,

Shaohua Chen
Senior Statistician
Development Research Group
World Bank

 

First off it’s great to have the World Bank respond on these issues, and especially great to respond by sharing the data. It’s still not particularly clear to me why the China projection was left off of the official publication this year; if the China story is interesting and important enough to merit its own line on the table – as I certainly believe it is – and the Bank has faith in the projection, then surely the 2015 figure should be included in the table, rather than listed as “not available”. But in any case, it’s great to now have the figures.

And indeed I’d argue it’s particularly important to highlight the China figure as it’s seen a rather substantial revision; last year the Bank thought in 2015 there would be just 66 million Chinese living in poverty, and this year that projection has risen to 100 million, a 50 percent increase. That’s an important change, and one that deserves to be discussed. Personally my guess is that it’s far too high, and that several years from now when we have the full data the actual 2015 figure will be considerably lower; time will tell. (Incidentally, it’s possible the 100 million figure is about right as an “expected value” prediction rather than a baseline prediction, in the sense that there are multiple possible equilibria for China’s future, including some unlikely-but-possible ones involving a hard landing, where poverty stops declining altogether or potentially even increases. So there’s a high (baseline) probability there will be fewer than 100 million Chinese living in poverty in 2015, but a small probability there will be much more, which could balance out to something like 100 million. But that’s not how people usually think about poverty projections.)

Finally, one small point of clarification: in the original post I did not mean to suggest that the China data was the reason for the delay of the overall new poverty results released earlier this year, but rather just that there were delays with the China data itself, such that China was not included in the April 2011 Povcal update…

Is the World Bank Deliberately Hiding its China Poverty Forecast?

I want to preface this post by saying that by nature I’m not a conspiracy theorist. But it seems to me that there is something funny going on with the World Bank’s efforts to monitor and forecast poverty in China.

I’ve done some work on poverty forecasting in the past, so I’m always excited to see when the World Bank, the official voice on such matters, puts out new forecasts. Last month the Bank released its Global Monitoring Report 2012, which includes its latest estimates of poverty for the year 2015. Here’s the relevant table:

When I first read this table my eye immediately focused in on the “—“ for China in 2015, which I initially took to mean the Bank believed $1.25 poverty would be effectively eliminated by then. While this would be an optimistic forecast, it actually doesn’t sound that crazy to me; after all, the World Bank’s China office (which, importantly, I believe operates mostly independently of the poverty team) wrote all the way back in March 2009 [PDF] that “extreme poverty, in the sense of not being able to meet the most elementary food and clothing needs, has almost been eliminated in China.”

But when you look at the bottom of the table, you see that “—“ actually means not available, rather than effectively zero. The obvious question, then, is why isn’t it available? Since we’re talking about forecasts, this can’t be a “data availability” issue in the strict sense of the term, because of course none of the data is actually available; these are the Bank’s best guesses at what poverty will be in 2015. So why isn’t there a guess for China?

The fact is there is a guess for China – there has to be – the Bank just won’t tell us what it is. And not only will they not explicitly tell us what it is, but they’ve gone out of their way to ensure we can’t calculate it ourselves.

We can be sure that there is a China estimate simply by noting that there’s an estimate for the East Asia and Pacific region as a whole; China accounts for about two thirds of the region’s population, so obviously it’d be impossible to guess how many poor people there’d be in the region without having a pretty good guess as to how many poor people there’d be in China.

Even more intriguingly, however, look at the bottom two rows of the table; the Bank gives a 2015 estimate for “World”, but the figure for “World excluding China” is once again “not available”. Note that logically this makes very little sense; if the figure for China were truly “not available”, then the Bank should be able to estimate “World excluding China” but not “World”, not the other way around. So why is the figure for “World excluding China” “not available”? Is it because if it were available we’d be able to work backwards and calculate the Bank’s 2015 forecast for China, which for some reason it doesn’t want to reveal?

One face-saving explanation would be if somehow the Bank’s model for producing these results truly only produced regional data, i.e. if the regional figures didn’t represent aggregates of national data (or aggregates of the big countries plus some residuals), in which case there wouldn’t be any “China” figure to show. But I’m sceptical of this for a number of reasons. In order of increasing conspiracy-ness:

One, it would just be a strange way to forecast poverty. While you maybe wouldn’t include every country in the world when you want to be able to add up to a global aggregate, whatever form of model you’re using can surely handle more than the six regions the Bank divides the world into, and the more fine-grained you get the better. And you’d certainly want to have specific data for China and India, as these two countries drive the global picture (and both have some controversial issues concerning their poverty counts, so it’s important to be able to speak about them specifically). And the source says ‘World Bank staff calculations from PovcalNet database’, and the PovcalNet database definitely builds regional aggregates by summing national data.

Two, in earlier editions of the Global Monitoring Report there’s always been a figure for China in 2015. Here’s the relevant table from the 2011 report:

So last year the projection method clearly allowed the Bank to forecast 2015 poverty in China; why not this year?

Three, and this is where we get to the most conspiratorial part – and admittedly most speculative, but of course what is a blog for if not wild speculation: the most recent poverty survey for China, which covers the year 2008, has been the subject of considerable rumours in the past. Specifically, there was quite a long delay from when many people thought the results would be released to the public until when they actually were, which was just a few months ago. Rumours from within the Bank suggested that the Chinese representatives at the Bank were being very secretive with the data, did not want to grant many people access to the raw data, and were deliberately holding up its public release. I don’t think I’ve ever seen anything written on this, so take it with a grain of salt, but I know a number of people who care about poverty data who spent a long time waiting to learn the results of the last China survey…

What are we to make of all this? My guess is there are two possible stories going on. One is that the poverty team within the Bank just doesn’t have enough faith in its China data to be willing to publish a specific forecast for the country – perhaps because of some of the well-known problems with the country’s PPP exchange rate estimate, or perhaps because of some issues with the 2008 survey, or something else. On the one hand this makes some sense – when you’re aggregating national forecasts into regional forecasts you have some room for errors to cancel each other out, and it’s easier to have more confidence in the broader picture than in the narrow one. But on the other hand, China’s a huge country; if we don’t have confidence in the China national data, then why would we have any confidence in the East Asia and Pacific regional data? And, perhaps more to the point, why in the 2011 GMR could the Bank feel confident enough to put out a 2015 forecast, but a year later it doesn’t?

The second possible story is more sinister: that there is some deliberate effort to keep Chinese poverty data and estimates from being released to the public. Did the poverty team within the Bank produce a figure for China – they almost certainly did – but someone else within the Bank decided this wasn’t the “right” number, and so didn’t want it published? Was there originally a number on the China line, but somewhere during the editing process it was crossed out? I know that sounds kind of crazy, but poverty data can be easily politicized – just ask India. I’d hope the Bank would be able to keep these political issues to a minimum, but it is of course not immune to political pressures, both internal and external.

The funny thing is, the Bank could have rather easily avoided this issue by simply eliminating the China-specific rows from the table; it almost seems as though they’re going out of their way to say “we have an estimate for China but we’re not going to tell you what it is”. Just what is going on here? Maybe there’s a good, logical explanation for all of it – and I hope there is – but right now I don’t see it…

UPDATE: See a response from the World Bank here.

Foxconn Factories, Sweatshops, and the Revealed Preferences of Chinese Workers

So it turns out that the much-lauded This American Life story from a few months back about working conditions in the Foxconn factories in China which produce Apple products was factually flawed. While there weren’t outright lies, exactly, there was certainly some coloring/blending of the truth, enough so that TAL is retracting it, a first for the program. This is of course embarrassing for TAL, but I’m also somewhat sympathetic to the defense put forward by Mike Daisey, who developed the piece: the story was originally created for his Off-Broadway one-man show, which obviously has much lower standards for factual accuracy than journalism, and he never particularly claimed to be a “journalist” anyways.

In any case, I actually think there are much bigger issues with the story other than some factual fibs, which have to do with the general approach of how it is framed (some of which Matt Yglesias touches on here). It’s of course true that working in one of these factories isn’t the best of all possible worlds. But if we’re going to be bemoaning the terrible conditions in Chinese manufacturing, we need to be clear about in comparison to what. When most Westerners hear about the hard lives of factory workers in China, I think they implicitly are comparing it to the employment opportunities and circumstances in their own societies. And they, quite understandably, think they’d never want that kind of job. But that of course is the reason why these jobs aren’t in America but are in China, a much much poorer country, where the other opportunities available for workers aren’t any better, and are often noticeably worse.

Chinese factories offer long hours, low (by Western standards) pay, and tough physical conditions. And yet there’s still huge demand from Chinese laborers wanting to move from rural to urban areas to seek these kinds of jobs. Indeed, when a new Foxconn factory opened in January, thousands of people lined up for a chance at a job. If we take the idea of revealed preferences seriously, it’s clear that Chinese workers think these jobs are better than the alternatives available to them, which for many of them is back-breaking work on low productivity, subsistence farms.*

To come back to last week’s discussion on misplaced Western advocacy efforts, I’ve long been at best ambivalent about anti-sweatshop campaigns, which were huge a decade-plus ago, then mostly died out, and now are somewhat coming back in the form of these Foxconn and other similar stories. It’s possible, under certain targeted circumstances, for Western pressure and the leverage of Western markets to be used to improve labor conditions in developing countries – the best example of this is probably the US-Cambodia textiles agreement, which allowed Cambodia the opportunity to increase its quota share of the US textile market by voluntarily increasing its labor standards. But to the extent that broader anti-sweatshop campaigns served simply to lower demand for these goods in Western markets and hence put some of these factories out of business, they probably did more harm than good. It’s easy to complain about workers being “exploited by the global economy”, but the fact is, in the absence of opportunities presented by the global economy, these workers could just as easily be exploited by the local economy, probably to much worse – if less visible – effect. There’s a reason so many millions of Chinese laborers want to get off the farm and into factory jobs, even when the conditions in these jobs are shockingly bad to Western societies. Advocacy efforts aiming to help the world’s poor shouldn’t be trying to stop them.

* It’s worth noting that this argument doesn’t apply in circumstances where the jobs don’t actually represent revealed, informed preferences, i.e. if workers are forced to work against their will, are cheated or lied to by their employers, or are too young to make an informed decision about where and whether to work. Thus the case for a basic set of universal labor laws proscribing practices such as slavery and child labor.

Washington Post Op-Ed: The Secret Success of Falling Poverty

My colleague Laurence Chandy and I have an op-ed in today’s Washington Post, which is based on some ongoing research we’ve been doing on updating global poverty data. (Our main findings are here.)

The central point we’re making is that our sense of current global poverty is dramatically out of date, since the World Bank hasn’t updates its global poverty figures since 2005.  In the past six years developing countries have enjoyed remarkable growth, especially those countries home to the highest number of poor people, such as India, China, and Ethiopia.

On our estimates close to 500 million people have escaped poverty since 2005, and the Millennium Development Goal of halving global poverty has already been achieved.

Go read the whole thing.



Follow

Get every new post delivered to your Inbox.

%d bloggers like this: