Column: China-India: a matter of record

Written by Surjit S Bhalla | Updated: Aug 7 2013, 10:38am hrs
This article attempts to document the record on poverty reduction in both China and India. That China has grown substantially faster than India is a matter of record and great pride for China. That the welfare of the poor in China has also improved at a much faster rate than India is also conventional wisdom as articulated by various scholars at international institutions like the UN, World Bank, Asian Development Bank, etc, and these findings have been supported and endorsed by Nobel laureates like Amartya Sen, e.g. Why India Trails China, NYT, June 19, 2013.

There are several household surveys of consumption in India and such surveys are led by the official National Sample Surveys (NSS) conducted every few years. These data are publicly available. Unfortunately, while everyone applauds Chinas GDP growth rate, no one, with the possible exception of the World Bank, has access to household unit level data. In other words, we have to take the official government view on household income and consumption growth, etc. Given the very large transparent accumulation of foreign reserves of more than $3.5 trillion, there is little doubt that GDP level, and GDP growth, in China is extraordinarily high. The story about how much trickle down there has been in China is documented below.

There was an important non-government household survey conducted jointly by Texas A&M University, USA, and the Southwestern University of Finance and Economics in Chengdu, China, in 2011. The survey results were widely reported in the international press in December 2012. One of the results of the survey was that the Gini index of income inequality in China was a high 0.61, much higher than government reported results of a Gini of 0.48. Unlike official Chinese household data, the explicit stated policy of the survey organisers was that researchers worldwide could access these data. Somewhat strangely, repeated applications to the website have not met with success. The results for China reported below are obtained from the World Bank site,

The Indian data are based on the NSS consumer and expenditure surveys for the years 1999-2000 and 2011-12 and for China for 1999 and 2010. The table reports the head count ratio of poverty for the World Banks poverty line for developing countries like China and India, i.e. PPP $1.25 per capita per day, which also just happens to be Indias official (Tendulkar) poverty line. For India, two estimates of poverty are presented (the survey details for the China surveys are not known). The two estimates are the standard official measure of consumption (what is called the mixed recall period and referred to as Type I) and the mixed recall period with perishable items like vegetables, fruits, meat, etc, measured with a recall period of 7 days. (Referred to as Type II; the recall refers to memory recall of items purchased over the last 7, 30 or 365 days.)

Chinas record: Spectacular growth in per capita GDP, per capita consumption, and poverty decline. From a poverty level of 35.6% of the population in 1999, China reduced it to 11.6% in 2010, for an average pace of decline of 2.2 percentage points a year. Both survey and national accounts (NA) consumption growth in China were near identical during this decade, at 6.7% and 7.3% per annum, respectively, i.e. the survey measured consumption growth was higher! If the efficiency of growth is (crudely) defined as the decline in head count ratio per unit of consumption growth, then Chinas efficiency for the decade 1999 to 2010 was around minus 0.3, i.e. for each 10% growth, poverty in China was reduced by 3 percentage points.

Another term for efficiency could be inclusiveness of growth. The poverty level for India in 1999-2000 and 2011-12 were 42.9% and 21.9%, respectively. This gives an average pace of decline of 1.8 percentage points per annum, somewhat below the 2.2% average recorded in China. But note the low level of household consumption growth observed in Indiaonly 2.8% per annum. This yields an efficiency-inclusive index of 0.6, twice the level observed in China! Also note that survey consumption growth in India (2.8% per year) is considerably below national accounts growth (4.8% per year).

If nothing else is changed but just the collection of household data on consumption, India shows a pace of poverty decline almost identical to China2.1 percentage points a year for Type 2 recall data compared to 2.2 percentage points for China. This yields an efficiency index for India which is one and a half times higher than China, i.e. 10% growth in average consumption reduces poverty in China by 0.3 percentage points, in India by 0.74 percentage points.

Another way of looking at the respective growth rates is by looking at the survey capture of consumption, i.e. the ratio of survey consumption to national accounts consumption. For China, this capture averaged a high 71% 1999-2004 and 74% 2004-2010. This survey capture for China is among the highest in the world. The survey capture in India averaged 53% in the early 2000s and 51% 2004-2011, and 46% in the recently concluded 2011-12 survey. In striking contrast to China, this survey capture level for India is among the lowest ever recorded for developing countries.

The low survey estimate for mean consumption in India and high survey estimate for China has strong implications for our assessment of poverty levels in the two countries. If India had the Chinese average of survey capture in 1999 and 2011 (about 75% rather than 50%), it would mean that per capita consumption levels in India were 50% higher than those observed, which means that poverty levels in India in 2011-12 would be substantially below Chinas in 2010, at any reasonable range of poverty lines, e.g PPP $1.25, 2 or 2.5 per person per day.

Another measure of the efficiency of growth is the change in inequality. Using the Gini as one measure of inequality (the Gini is equal to 0 if everybody has the same income and 1 if 1 person has all the income) it is observed that inequality in China increased marginally (6%) between 1999 and 2010. In India, nominal Gini increased by 12.5%; however, the real Gini (consumption adjusted by rural urban price differences) increased by only 6%! Indeed, if one goes back to 1983, NSS data suggests that real inequality in India has stayed constant for nearly 30 years (Gini of 0.303 in 1983 and 0.317 in 2011/12), whereas World Bank estimates that inequality in China has increased by more than 50%and much more if the banned China CFS survey are considered accurate.

No matter how one slices the data, it is the case that growth in India has been substantially more inclusive and pro-poor than Chinaand that the bottom one-third of the population have had a higher standard of living than China, despite the substantially higher GDP per capita level (and growth) in China. So before the international organisations, and/or scholars like Amartya Sen lecture India about the example of China, it would help if they were less political, and less ideologically biased, and it would help even more if they looked at the data before leaping to unsubstantiated conclusions.

Surjit S Bhalla is chairman of Oxus Investments, an emerging market advisory firm, and a senior advisor to Blufin, a leading financial information company. He can be followed on Twitter, @surjitbhalla