Media Briefings

Inequality Does Not Reduce Growth: Evidence From China

  • Published Date: December 2011

High or rising inequality in villages in rural China during the reform years since the late 1980s cannot be blamed for slower economic development. That is the central finding of research by Professors Dwayne Benjamin and Loren Brandt of the University of Toronto and John Giles, a Senior Economist at the World Bank.

'Concern over the rapid rise of inequality in China may be entirely legitimate’, they conclude in a study published in the December 2011 issue of the Economic Journal. ‘But it should be driven primarily by the inequity itself rather than its potential impact on growth. Inequality may be vicious, but there is no circle.'

The researchers conduct a thought experiment in which two children grow up in different villages in rural China. In the first village, household incomes are relatively equal: rich families make about double the income of poor families. In the second village, average income is the same, but inequality is higher: rich households earn three times as much as poor ones.

Does it matter in which village the child grows up? Would families be more prosperous in the low-inequality village?Many economic theories suggest that income inequality is a double curse, in which a possibly unfair initial distribution of income is compounded by slower future growth.

Analysing household survey data for China from 1987 to 2002, this study find that inequality early in the country’s reform period was associated with slower growth, but that this adverse relationship fades over time. Moreover, while village inequality rose significantly, it had no long-term impact on the evolution of household incomes.

Why might inequality reduce growth? There are two sets of explanation. In the first set, credit markets are poorly developed and households rely on their own resources to finance income-earning activities. Poor households are therefore at a significant disadvantage despite the fact that they have the greatest growth potential. High inequality villages have more of these poor, credit-constrained households and average income growth is lower.

The second set of explanations focuses on the political consequences of inequality. Less equal communities make collective choices that reduce the growth potential of all households. For villages in China at the outset of reforms, local policy could have pronounced long-term effects, especially given the limited geographical mobility of individuals. Villages were relatively isolated from markets and had considerable autonomy over local taxation and expenditure.

In these circumstances, high inequality villages could adopt policies that benefitted a lucky few – for example, spending less on schools or investing less in local roads and other public goods that help households branch out of farming. Inequality would then be a ‘marker’ for poor growth prospects, though disentangling the relationship between inequality and the bad policies themselves would be difficult.

To estimate the effect of inequality on growth, these researchers use a longitudinal survey that tracked over 3,000 households from about 100 villages between 1987 and 2002. The use of household-level data has significant advantages for distinguishing between competing explanations for the relationship between growth and inequality. Most importantly, it is possible to control for initial household income and directly evaluate the potential role of imperfect credit markets. It is also possible to rule out the role of aggregation in generating a spurious relationship.

The study finds that households in high inequality villages experienced lower growth for the first 10 years, but that by the late 1990s, the adverse effect attenuated. Growth rates were generally very high and villages integrated with the broader Chinese economy, so the effect of villages’ initial conditions faded over time.

The researchers draw several lessons from this experience. First, concern over the rapid rise of inequality in China may be entirely legitimate, but should be driven primarily by the inequity itself or the underlying factors that generate inequality. There is no evidence that the level of inequality or changes in inequality can be blamed for slower economic development. Inequality may be vicious, but there is no circle.

Second, by using household-level data, the researchers address a number of technical issues that plague the usual attempts to interpret and relate growth to historical differences in inequality across countries.

And third, throughout the paper, the researchers confront by way of confession the most difficult problem of all: it is almost impossible to find a real world counterpart to the original thought experiment. High-inequality villages (or countries) are never 'otherwise identical' to low-inequality villages (or countries). Isolating the pure effect of inequality on growth, as if variations were generated by Robin Hood as social scientist, will always be difficult.


Notes for editors: 'Did Higher Inequality Impede Growth in Rural China?' by Dwayne Benjamin, Loren Brandt and John Giles is published in the December 2011 issue of the Economic Journal.

Dwayne Benjamin and Loren Brandt are at the University of Toronto. John Giles is at the World Bank.

For further information: contact Dwayne Benjamin on +1 (416) 978-6130 (email:; John Giles via email:; or Romesh Vaitilingam on +44-7768-661095 (email: