China's 25 Years of Rapid Growth
China's economic development has been remarkable. Over the past several decades, China has achieved rapid growth, raised living standards dramatically, and eradicated extreme poverty. These are remarkable accomplishments in absolute terms and even more so relative to other economies during this period. How did China achieve this? And, how can China continue to grow rapidly?
China's experience is sometimes labeled a growth miracle. This is wrong. China's growth is not a miracle. Rather, it is the direct result of China's reform and opening up. From 1980 to 2023, China's average annual growth rate was 9.0 percent. This is much faster than any other major economy. Focusing on just the last 25 years (1999-2023), the period this article focuses on, China grew by 8.3 percent per year on average. If China's growth was not a miracle–in the sense that the growth simply fell from heaven–then what was the source?
The paper proceeds as follows. First, it looks at China's experience during this period relative to other countries, with a focus on total factor productivity (TFP) growth (or productivity growth for short). By most metrics, China's outperformance in this period was among the best in the world. Based on these findings, it will then look at what will determine China's growth going forward. The key determinant will again be productivity growth. Thus, it will dig deeper into how to generate productivity growth, emphasizing not only innovation but also the crucial–and often under-appreciated–role of resource allocation. It will conclude with a summary.
The main conclusion is that future growth will depend on reforms. The classic investment disclaimer applies to economies as well: Past performance does not guarantee future success. Just as reform and opening up drove China's remarkable growth to date, continuing reform and opening up will determine China's future growth. Nonetheless, the evidence is clear that China has significant potential for robust growth in the coming decade. Realizing this potential will depend, for the most part, on successfully lifting productivity growth. Thus, it is encouraging to see the continued focus in official communications on the importance of boosting total factor productivity. Likewise, the emphasis on ensuring the market plays the decisive role in the allocation of resources is also encouraging–as improved allocation of resources will be one of the key ways to lift productivity.
China's Growth
Over the previous quarter century, China has been the fastest-growing (GDP per capita) major economy. This article primarily uses data from the Penn World Tables 11.0 (Feenstra et al., 2015). The analysis uses the 75 largest economies, as measured by the Purchasing Power Parity (PPP) GDP in 2023, the last year included in the dataset. This keeps the charts manageable and the sample economies more comparable. The analysis also generally starts in 1999, the year that China's PPP GDP surpassed US$4,000. In figure 1, China's per capita GDP growth is the highest in the sample, averaging a bit under 8 percent. The line in the figure shows the fitted value of GDP growth per capita based on an economy's initial income level. Specifically, the initial per capita income relative to the USA in 1999. The downward slope captures the idea that lower income economies tend to grow faster on average as they catch up–or converge–to richer economies. This idea that poorer economies will grow faster than richer ones is referred to as unconditional (or absolute) convergence. The downward slope–which is statistically significant at conventional levels–is consistent with unconditional convergence. The horizontal axis uses a natural log scale, but the numbers on the axis are converted back to the percent of USA GDP per capita in 1999. For example, in 1999, income per capita in China was 8.0 percent of that in the USA.
Controlling for initial income levels, China's performance is even more impressive. As shown in figure 2, China's per capita growth was 4.3 percentage points higher than suggested by unconditional convergence. To understand this figure, it is similar to figure 1 but rotated counter-clockwise such that the downward-sloping line is now the zero line. Thus, it shows the distance between the fitted value of growth per capita based on initial income and the actual value. This difference (technically the residual in the unconditional convergence regression) can be thought of as outperformance controlling for initial income. The second-best economy, Ireland, outperformed by 2.8 percentage points. So China's outperformance was some 1.5 percentage points higher than that of the next closest economy.
Growth Accounting
How did China grow so fast? Growth accounting provides a framework to decompose GDP growth into the contribution from: (1) total factor productivity; (2) capital; and (3) labor. The Technical Annex provides the details of these calculations. The intuition is as follows. Think of the economy as a factory. A factory can increase output by acquiring more machinery and equipment (capital), adding more workers (labor), or using labor and capital better. This "use-it-better" is total factor productivity (or productivity for short).
Total factor productivity growth is the key determinant of living standards over time. It is by using capital and labor better that GDP (or equivalently income) can continue to rise. There are limits to how much capital can increase. Moreover, since capital depreciates each year, as the capital stock grows as a share of GDP, more investment is needed just to replace depreciating capital and maintain the stock. Labor supply, meanwhile, is limited by demographics. The quality of workers can also improve–such as through education–and the measure of labor used in this analysis includes quality improvements (so increases in education boost the labor input).
China's growth has slowed in the past 15 years. The year-by-year decomposition of China's growth into productivity, capital, and labor is shown in figure 3 The annual data are bouncy, making it hard to discern patterns. To better see trends, figure 4 shows the 10-year moving average for each component and unstacks them to make it easier to read. The slowdown in China's GDP growth is more apparent in this chart. The 10-year average fell from a peak of 10.1 percent in 2011 to 5.8 percent by 2023.
Productivity, capital, and labor are each contributing less. The contribution of labor has been steadily falling. This reflects demographics, with China's working age population having already peaked. Increases in the quality of labor (education) are partly offsetting the impact of a shrinking workforce. Labor's contribution, however, is essentially zero for the last 10 years. In contrast, labor contributed about 2 percentage points a year to growth in 1990-99. Capital has been the main contributor to growth. Its contribution to growth peaked at 5.3 percentage points in 2012 (meaning 2003-12 average). This reflects both the takeoff in investment following China's SOE reform (late 1990s) and accession to the WTO (early 2000s). Capital's contribution, however, has since been steadily falling and by 2023 was down to 3.5 percentage points.
The productivity story is more complicated. Productivity is calculated as a residual and tends to be volatile. It was especially low during the Asian Financial Crisis (1998-99) and dipped again during the pandemic. Focusing on the more recent period, productivity's contribution to growth peaked at 4 percent of GDP in 2009 (2000-2009 average), which largely reflects the dividends from reforms (such as to the SOE sector) and opening up (WTO accession). Since then it has fallen significantly, and averaged 2.2 percentage points in 2023 (meaning 2014-2023). Even before the pandemic, the average had fallen to 2.3 percentage points (2010-2019).
Global Perspective
However, declining total factor productivity growth is a global phenomenon. Prior to the pandemic (Adler et al., 2017) and after (International Monetary Fund. Research Department, 2024), researchers at the International Monetary Fund (IMF) have been flagging the global decline in productivity. In particular, that productivity in most economies fell noticeably after the global financial crisis (GFC) in 2008. This provides important context for the slowing of China's productivity and begs the question of how China fared relative to other economies.
In the post-GFC period, China's productivity grew faster than any other major economy. Despite the considerable slowing noted above, it was still much higher than that of others and the only one that averaged over 2 percent productivity growth in the post-GFC period. This is shown in figure 5, which compares the major 75 economies' productivity growth in 1999-2007 (horizontal axis) with 2011-2019 (vertical axis).
Productivity in most economies slowed. This is demonstrated by the majority of the points being below the 45-degree line. Some economies are in the lower left quadrant, which means their average productivity growth was negative in both periods (for example Brazil and Mexico). In the pre-GFC period, China had the fourth fastest productivity growth, trailing only three former Soviet Republics. One caveat, though, is that several fast-growing economies had to be dropped from the sample due to missing data (such as Vietnam). Using data from 2011-2023 (instead of 2011-2019) on the vertical axis does not qualitatively change the results.
Future Growth
How fast China can grow will largely depend on China's success in generating productivity growth. Many studies have estimated China's growth potential. Sasaki et al. (2021) summarized this research and added their own estimates. Muir et al. (2024) also looked at potential growth and emphasized how it depends on structural reforms. In a reform scenario, potential GDP growth could reach 4.6 percent in 2030 and 3.6 percent in 2040. This article will not repeat these exercises, but instead will take a qualitative approach.
Growth Accounting
Returning to the simple model of the economy, future output depends on growth in labor, capital, and productivity. Following International Monetary Fund. Research Department (2024), it makes sense to consider labor and capital first. The upshot is that there are gains to be had in these areas, but the potential boost to GDP is much smaller than from productivity. Globally, International Monetary Fund. Research Department (2024) estimated modest gains to global growth from increasing labor supply through labor force participation (0.16 percentage points increase in medium-term global growth). It also mentions other policy options such as improving labor quality through training and education. In China, such measures can help offset the drag of a declining workforce due to an aging population. For example, recent reforms to the retirement age could increase labor force participation of older people and thus boost labor's contribution to GDP growth. For capital, globally, reforms to increase saving and investment can increase the growth of the capital stock and future GDP. However, these are less relevant for China, which already has a high capital stock–and high saving and investment rates–suggesting limited potential for capital accumulation to be an engine of growth going forward.
The biggest gains to growth, for China and the world, are to be had through increasing productivity. Productivity gains can, in turn, be divided into two types: (1) innovation, such as new technology; and (2) better allocation of resources. Resource allocation can improve productivity by moving capital and labor from low to high productivity activities. This could include, for example, having poor performing firms close and their labor and capital move to more productive firms in the same sector. It also includes movement of resources across sectors, such as from farming (which is low productivity) to manufacturing and services (higher productivity).
Allocation of Resources
In contrast to innovation, the role of resource allocation in boosting productivity tends to be under-appreciated. However, both globally and in China, it is where some of the largest gains to growth are to be found. Globally, International Monetary Fund. Research Department (2024) estimated that structural reforms to improve the allocation of resources (what the paper calls allocative efficiency) could lift medium-term global growth by 1.2 percentage points. China-specific research also argues there is huge potential for resource allocation to increase GDP. International Monetary Fund. Asia and Pacific Department (2024) suggests that better allocation of resources, especially within the service (tertiary) sector, could lift productivity considerably (and emphasized that innovation in the service sector was strong but overall productivity was being held back by weak allocative efficiency).
Another way to illustrate the gains from resource allocation is to focus on labor productivity. Labor productivity is defined as output (GDP) per worker. Better technology and more equipment (capital) would each boost labor productivity. Viewing China's experience through the lens of labor productivity illustrates the gains from resource allocation.
China's rapid growth over the past 25 years was fueled in large part by a better allocation of resources. Specifically, the flow of labor from lower productivity agriculture (primary sector) to higher productivity manufacturing (secondary) and services (tertiary). China's reform of state-owned enterprises in the late 1990s freed up labor resources, while further reforms and opening up created conditions for a boom in private sector manufacturing–which was the source of virtually all net job creation in China in the 2000s (Lardy, 2014). Each worker that moves from farm-to-factory results in a considerable increase in labor productivity. Why? Because labor productivity in manufacturing is much higher than in agriculture.
Figure 6 illustrates how the flow of labor from farm-to-factory propelled China's growth. Following Zhu et al. (2019), the left panel shows the level of labor productivity in China's three sectors. During the 2000s, taking 2007 as an example, a worker moving from farm-to-factory (primary to secondary sector or primary to tertiary) produced on average nearly five times more output. Even today, labor productivity in manufacturing (and services) is about 3.5 times higher than agriculture. This has two significant implications. First, unlike farm to factory, the next stage of development for an economy typically involves the growth of the service sector. Thus a flow of labor from factory to florist. This factory-to-florist (secondary to tertiary sector) labor flow, unlike farm to factory, does not by itself lift growth as labor productivity is similar in the two sectors. This puts more of a burden on reallocation of resources within a sector to drive gains. Second, it does not matter much whether labor is flowing from the farm to a factory or a florist: Both yield similar productivity gains. The key is to continue the flow of labor out of the primary sector, with little difference for growth whether that worker goes into the manufacturing (secondary) or service (tertiary) sector.
Thus, continuing the flow of labor out of agriculture is an important way to boost growth. The middle panel of figure 6 shows that this flow of labor has declined considerably. For example, from 2003-08 the share of labor in the primary sector declined by 9.5 percentage points. Whereas in the most recent 5-year period, the share of labor in the primary sector declined only 2.5 percentage points. Thus, the pace of labor reallocation has slowed considerably. While some of this is natural as the share of the primary sector employment in the economy shrinks, there is nonetheless room for much more labor to move out of agriculture. For example, advanced economies' share of labor in agriculture is typically well below 5 percent. In contrast, China had more than 20 percent of labor in the primary sector in 2024. In fairness, other large emerging markets have similar or even larger shares. But, the point stands, accelerating the flow of labor out of farming can provide a considerable lift to labor productivity growth for many more years.
Reallocation of labor out of farming has been a large driver of growth. At its peak, the 5-year average contribution to GDP growth from labor reallocation was 3.6 percentage points. This is shown in the far right panel of figure 6. In the most recent data, it was contributing only 0.5 a percentage point to growth. Thus, a decline in the pace of labor flowing out of the primary sector by itself can account for 3 percentage points of the decline in China's growth. Put differently, accelerating the flow of labor out of the primary sector has the potential to considerably lift productivity (both labor and total factor productivity) for many years to come. This is a prime illustration of how improving the allocation of resources lifts productivity and in a way that is distinct from innovation.
Reallocation of labor out of farming is just one example of how improving the allocation of resources lifts growth. As highlighted above, any movement of capital or labor from lower to higher productivity activities will lift economy-wide productivity. This type of resource allocation is what markets excel at and explains why it is so important to give markets the decisive role in the allocation of resources. This, of course, is precisely what China's policymakers have been emphasizing as a top priority. Successful implementation will provide a considerable boost to productivity.
Conclusion
China has experienced remarkable economic development over the past quarter century. This growth, however, is not a miracle. Rather, it is the direct consequence of reform and opening-up policies. Likewise, achieving robust growth going forward will depend not on miracles but on continued reform and opening up policies. During this period, China has outperformed most economies in terms of per capita growth and productivity gains. To continue to do so, total factor productivity will have to continue to outperform. This is possible. The recipe for sustained growth is not a secret. It depends on lifting total factor productivity through a combination of innovation and resource allocation. The gains from improved allocation of resources, moreover, are often under-appreciated yet are where some of the biggest gains have been and could still be found. This is true for China and also true for the world.
Steven Barnett is Professor of Practice, HKU Business School, the University of Hong Kong