We find that China’s potential growth in GDP per capita is substantially underestimated if the level of GDP per capita is employed as the convergence indicator as done in previous studies (e.g., Barro, 2015 and 2016). Using data on China’s position in the global value chain (GVC) prior to 2010, we predict that the country’s GDP per capita could have grown at 7%–8% annually between 2010 and 2015, which is closer to the actually growth rates of 7.8%, than predicted by the previous estimation of around 4%, which was based on GDP per capita.
China’s GDP growth rate decelerated to only 6% in the third quarter of 2019, raising a new round of concerns about China’s growth potential. This is similar to the scenario in 2012 when China’s GDP growth rate fell below 8%, which was the first time that happened since the 1990s. Barro (2016) predicts that China’s growth rate will soon fall from 8% to 3%–4%. In contrast, Justine Lin, a professor of Economics at Beijing University and former chief economist of the World Bank, believes that China’s GDP has the potential to grow at a rate of 8% until at least 2028. As the world’s second largest economy, China contributes to about 30% of global growth. An accurate estimation of China’s GDP growth potential is not only crucial for the country, but also critically important for the global economy.In a recent paper, we argue that standard empirical growth analysis underestimates China’s GDP growth potential. In previous studies, the growth rate of per capita real GDP for a cross-section of economies between years t and t + τ is regressed on a convergence potential indicator and a list of control variables at time t. The convergence potential indicator tends to measure the distance of a country’s technology from the world technology frontier and per capita real GDP (in logarithms) (or the difference from that of the U.S.) is commonly used as the proxy for this indicator in empirical studies (e.g., Sala-i-Martin, 1996 and Barro, 2015). We argue that per capita real GDP is a misleading indicator for China.
The asterisks in Figure 1 represent data regarding China. Both China’s per capita GDP and GVC positions have moved up significantly from 1997 to 2010. In particular, China’s per capita GDP has successfully moved from low-income countries in our sample to mid-income countries. However, the GVC position of China remains relatively low among the countries in our sample. The discrepancy between the relative positions of China’s per capita GDP and GVC position index induces the difference between our results and Barro’s (2016). The GVC position index suggests that China’s GDP growth potential may remain high, while the convergence indicator measured by per capita GDP predicts a much lower growth potential.
Our proposed GVC index (along with a set of control variables commonly used in the literature) performs better than per capita real GDP in forecasting China’s GDP growth (see Note 3). The root mean squared error (RMSE) is 11.7% smaller in the GVC modelwith both time and country fixed effects than that in the corresponding Barro model. The difference is over 40% if we compare our GVC model with Barro’s favorable model(the model with only the time fixed effects).Specifically, when using data from before 2010, we predict a much higher growth rate for China after the 2008 global financial crisis than Barro (2016). For instance, Barro’s favorite model predicts China’s 2010 growth rate to be 5.52%, while our estimate with the GVC index is 9.44%, which is much closer to the actual growth rate of 9.77% (Table 1). According to our out-of-sample prediction, China’s per capita GDP growth should have declined to 7.7% in the years between 2011and 2015, in sharp contrast to Barro’s pessimistic forecast of 3%–4%.
Many researchers believe that China’s high growth is unsustainable, which is understandable from a historical perspective. No large economy in history grew at China’s rate when its GDP per capita was at a similar level. When a country moves close to the global technology frontier (GTF), its growth inevitably slows as innovations happen at a much slower pace than technology adoption.
However, we argue that GDP per capital is a misleading indicator for China’s distance to the GTF. For instance, people have longer working hours and fewer vacation days in China than in most other countries. The data of GDP per capita that does not appropriately adjust for these factors may underestimate China’s distance to the GTF and its growth potential. Using an arguably better measure, the GVC position index, we estimate that China’s GDP growth potential remain above 7% between 2010 and 2015. It is reasonable to believe that the growth potential will remain between 6% and 7% after 2020. China needs to conduct structural reforms to fully utilize its potential rather than pessimistically believing that its economy can only grow at less than 6% each year. The pessimism and the policies based on such pessimism can be self-fulfilling, which would not only drag down China’s GDP growth but also threaten global growth prospects.
(Dazhong Cheng is a Professor of Economics at Fudan University; Jian Wang is an Associate Professor of Economics at The Chinese University of Hong Kong, Shenzhen; Zhiguo Xiao is an Associate Professor of Statistics at Fudan University.)