China is not alone in having unreliable growth data

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

When China’s premier, Li Qiang, announced at Davos on Tuesday that the country’s economy grew at an estimated 5.2 per cent in 2023 it caught economists off guard. Many had projected a figure close to Beijing’s 5 per cent growth target — but markets had not expected the data to come until Wednesday, as per the official release schedule. It is fair to say that this is not the only example of China’s hazy approach to reporting national statistics.

Economists now largely consider Beijing’s official economic data to be only a reference point. While China puts significant resources into its National Bureau of Statistics, trust in its output has been dented by deteriorating transparency. The number of economic indicators made available by the agency has dropped significantly since President Xi Jinping became leader in 2013. At the same time, forecasters have become more sceptical over China’s official gross domestic product numbers. The country’s slowdown in trend growth over the past decade has increased the spotlight on its system of GDP targets.

It is of course possible that China’s economy has done a consistently accurate job of hitting those targets. But limited transparency, including on its statistical methodologies, does not engender trust. Economists have meanwhile developed their own GDP estimates using directly observable statistics, such as night-light density, or data collected by surveys and international agencies. Staggeringly, Rhodium Group thinks China’s growth last year may have been as low as 1.5 per cent. Capital Economics’s “China Activity Proxy” suggests that Beijing has been overestimating its output notably since the start of 2022. Its modelling also implies faster growth in 2023 as official data may have lowballed the extent of Covid-19 disruption the year before.

China is not the only country with unreliable growth statistics. India’s are spotty too. This is one of the fastest-growing major economies, and the most populated. Yet World Economics ranks the quality of the country’s GDP and population data around 90th place globally. A lot of India’s economic data is based on outdated surveys and information. Its existing numbers are also too frequently revised. While official estimates place annual average GDP growth between 2011-12 and 2016-17 at about 7 per cent, a Harvard study suggested it could be closer to 4.5 per cent.

A lack of transparency, subpar methodologies and the scope for data manipulation are common in much of the so-called developing world. Indeed, researchers lament the dearth of credible economic data across Africa. Large informal sectors and limited resources mean measurement is not simple either. Even advanced economies have data problems; Britain’s labour market survey suffers from a potentially distorting low response rate.

Yet as the economic and demographic centre of gravity shifts away from the west it is even more important that developing world data becomes less opaque. Domestic policymakers will struggle to tackle and identify barriers to economic growth without decent numbers. Foggy statistics add risk for investors, and can raise the cost of capital for developing nations. For China and India, in particular, trust in their data is important to maintain and attract the interest of multinationals and capital markets. Ambiguity is not good for companies.

After lamenting the global “trust deficit” in his Davos speech, Premier Li should ensure the world’s second-largest economy greatly improves the transparency of its own economic statistics and methodologies. For its part, India needs to invest in updating its data systems. International agencies and the private sector should also step up their support — better technology, training and data sharing can help raise data standards. After all, given how important good data is for effective government, investment and business, as emerging economies rise so must the reliability of their statistics.