What’s happening to inequality in China? Update from a visit to Beijing

June 7, 2016

     By Duncan Green     

China gini 2003-14Spent a fascinating few days in Beijing last week, at the invitation of Oxfam Hong Kong. The main topic was inequality, including a big seminar with lots of academics (NGOs are very research-based in China – it was a graphtastic, PhD-rich week). Here are some of the headlines:

Income Inequality in China is changing fast. According to the National Bureau of Statistics, the Gini index peaked in 2008 and has fallen back since (see graph – and note it only starts in 2003, so omits the big rise in the gini before that date). But before we declare victory, there are three big caveats:

  1. Measurement: the index is based on the NBS’ annual survey of 140,000 households. The problem with that, as Thomas Piketty pointed out in his LSE talk, is that it misses the very wealthy households, and so underestimates overall inequality. According to the top Chinese inequality guru, Li Shi (who’s is speaking in London this afternoon, at a King’s College event), tax data is also unreliable (even if the government was willing to release it) as fewer and fewer people report their real income.
  2. Urban v Rural: confusingly, income inequality is actually rising within urban areas, and within rural areas, even as it falls overall. This apparently contradictory effect is because the gap between urban and rural areas has been shrinking, as the government has given priority to pushing industry, jobs and public spending inland. It’s a bit like the global picture, where the Global Gini is falling because of the rise of China even as many/most countries experience rising within-country inequality.

This means that China is undergoing a shift from horizontal inequality (between groups, in this case regions) to vertical inequality (between individuals). As Frances Stewart has shown, these two kinds of inequalities typically produce different social effects – horizontal inequality often leads to conflicts, civil wars between powerful and excluded groups etc, whereas vertical inequality is more likely to trigger social breakdown as the poor start taking redistribution into their own hands. Given the Chinese government’s concern with stability and social cohesion, its increasing focus on inequality makes a lot of sense.

  1. China wealth deciles 2002-10Wealth inequality is rising incredibly fast, largely driven by China’s massive housing bubble. The bar chart shows the extraordinary rise in the assets of the richest 10% of Chinese over just the 8 years to 2010 (and the bubble has continued since then). Thanks to rising rents and capital gains when people sell their houses, the bubble is also responsible for about a third of the rise of urban income inequality, according to Li Shi. One Piketty-esque solution would be to introduce property or inheritance taxes but that is likely to run into opposition from the people with the big houses, and in any case could burst the bubble, precipitating a property crash with disastrous wider effects on the economy.

What to do? The government plans to continue to extend the social security system within rural areas, which has expanded massively in recent years. It also says it will reform the hukou system, which restricts migration and condemns migrant workers from rural areas into second-class citizens in the cities. I spent an afternoon with an Oxfam partner, the Beijing Migrant Workers’ Home to get a better feel for the hukou system. Set up by Sun Heng, a migrant street musician, the Centre has set up a cinema, theatre and even a museum, as well as building a school for migrant workers’ kids who are excluded from the public system. Enough text, here are a couple of short video clips with my impressions, and some views from the Home’s Lü Tu.

Finally, a rather interesting idea came up which I said I would ask around about. Inequality data is much more unreliable and patchy than income data. Could Big Data help? Has anyone thought about how to use Big Data either to fill in the gaps in household surveys (mentions of BMWs or champagne on social media as a proxy for high end consumption? Only half joking – you get the idea), or to do something in realtime on rising and falling inequality in different regions? Would love to hear what’s going on, as would the researchers in Beijing, because while the government may be tight-lipped, there is lots of access to data from China’s hyper-vibrant social media users.