Big data in China

Most of the accolades recently bestowed on Nobel-winning economist Angus Deaton have focused on his work in India. And it’s true that by pioneering ways to improve the quality and precision of economic data, Deaton revolutionized our understanding of South Asia’s grinding poverty and how to reduce it. Yet Deaton’s work may have even greater relevance to China.

No major economy could benefit more from Deaton’s focus on seeking out more and better data. Take China’s most recent GDP numbers, released last month. Some economists had hopes that the data might be more accurate than usual, after the Chinese government subscribed to a stricter standard employed by the International Monetary Fund. In fact, the claim that the economy was growing at a 6.9 percent annual rate struck most experts as suspiciously high.

Contradictions in the numbers were obvious to anyone with an elementary grasp of mathematics. The government reported robust 23.8 percent growth in telecommunications services through July. Yet looking at more specific components - as Deaton would prescribe - reveals dramatic discrepancies. Instead of growing, the volume of fixed-line calls and length of mobile calls both declined significantly. Mobile and broadband subscriptions did increase, but only by a modest 2.5 percent and 4.5 percent, respectively - hardly enough to produce almost 24 percent aggregate growth.

Similarly, comparing official and unofficial statistics casts serious doubt on the former. The government claimed hotels and catering expanded almost 9 percent in the first three quarters of 2015. Yet private measures - which inspire greater confidence, given that they’re drawn from bank card payments and reported revenue of listed restaurant groups - show double-digit declines. The figures for hotel revenue and occupancy are flat.

The job of a modern policy maker is hard under the best of circumstances. In trying to wean the economy off an addiction to investment and debt, Chinese officials are struggling with one of the most complex and dangerous restructurings in modern economic history. As former Premier Wen Jiabao once said, “Any small problem multiplied by 1.3 billion will end up being a very big problem.”

Bad information makes the task even more herculean. The poverty and unreliability of Chinese data forced current Premier Li Keqiang to devise his now famous index to measure GDP using just three variables: electricity consumption, rail freight traffic and bank loans. Chinese leaders especially need to know the truth about employment figures, given potential for layoffs to spur unrest. Yet official unemployment has hovered around 4.1 percent for more than a decade, despite changing economic cycles. The government has dispatched more than 15,000 statisticians throughout China armed with tablet computers to test how accurate those numbers are.

Historically, Chinese technocrats have often been reluctant converts to high-quality, regular and transparent data. In some cases, outside pressure has forced changes, as when U.S. Embassy air quality readings made a mockery of the government’s assurances about Beijing’s skies. At other times, ordinary Chinese have rejected implausible official numbers, as with a national home price index that became such a joke that the government had to abandon it in 2010.

What’s changed is that the range and quantity of information available to private investors, businesses and the public has grown exponentially. Payment network Union Pay has leveraged its vast reach into the Chinese consumer’s wallet to produce data on spending habits by sector. Alibaba has exploited its position as the dominant online retailer to start a financial-services arm, using customers’ purchasing history and demographic data to judge their creditworthiness. CEO Jack Ma recently predicted that data would become one of the world’s most valuable commodities, “like water, electricity and oil.”

Despite their reputation as passive citizens, Chinese are now voracious consumers of good information. Their distrust of government rhetoric has fueled a bustling market in privately produced data, such as the unofficial monthly purchasing-manager index produced by the Chinese financial media group Caixin. Soufun, a real estate company, has tapped its internal data to help build a private real estate price index that’s now used next to the official measures.

This explosion in data presents the government with an excruciating choice. On the one hand, losing its monopoly over information means that the regime can no longer dictate the narrative about China for domestic and international consumption. Investors and the public now have better means to grade the leadership’s performance.

This can complicate policy-making: Contradictory data helped prompt investors to sour on the Chinese economy over the summer, forcing officials to scramble to halt a market rout. Vested interests, many of them influential Communist Party officials, resist the release of statistics that could reveal their privileged position.

On the other hand, challenging those vested interests is - or should be - a central element of the government’s reform efforts. If China really wants to tilt the economy away from manufacturing and build a thriving service sector, it’s going to have to allow companies rapid and easy access to ideas, data and information. Deaton has said that “democracies require good data.” What Chinese leaders need to accept is that healthy economies do, too.

*The author is an associate professor of business and economics at the HSBC Business School in Shenzhen and author of “Sovereign Wealth Funds: The New Intersection of Money and Power.”

by Christopher Balding