The heightened triple pressures—demand contraction, supply shocks, and weakening expectations—facing the Chinese economy have been severely manifested in the employment situation, which both reflects the cyclical unemployment shocks within the macroeconomic downswing and reveals or implies long-term structural employment contradictions. While responding to short-term shocks through macroeconomic policies and active employment policy, it is especially important to prevent short-term factors from exacerbating long-term difficulties.

Since the outbreak of the COVID-19 epidemic in 2020, urban employment in China has suffered from a comparatively long shock over a sustained period. The severity of this employment shock can be appreciated based on three types of unemployment, namely structural unemployment, frictional unemployment (both of which combine to form natural unemployment), and cyclical unemployment. In February 2020, China’s surveyed urban unemployment rate reached a high of 6.2 percent, the highest level since the indicator started being officially recorded in January 2018. It has fallen since then, but assuming China’s natural unemployment rate to be 5 percent, the surveyed urban unemployment rate has been significantly higher than that level in more months since then. In April 2022, China’s surveyed urban unemployment rate increased again to 6.1 percent. Although it then fell back, it remained significantly above the natural rate.

I. Labor market balance and imbalances

In recent years, China’s population figures have been constantly altering people’s perceptions. Data from the seventh census in 2020 revealed for the first time that the total fertility rate was already at the extremely low level of 1.3, and that natural population growth was on a sharp deceleration trend. Data from the National Bureau of Statistics in 2021 showed that the natural population growth rate had fallen to 0.034 percent, and the latest UN projections for 2022 suggest that China’s population will peak this year. As the population reaches its peak and enters negative growth, population aging will also deepen, and the rate of decline in the working-age population will accelerate. It is worth pointing out that this change does not signal any easing of employment contradictions, but instead marks the end of a period of relative ease in China’s employment situation. Based on the fact that the aggregate contradictions in employment have given way to structural contradictions over the past decade or so, one can expect that structural and frictional contradictions in the labor market will become more prominent in the future. This can be understood mainly by observing three change factors.

First, the share of both young workers and older workers in the working age population will increase for a limited period of time. According to UN population projections, the share of the relatively young (16-24 years old) in China’s total working-age (16-64 years old) population is expected to increase from 14.8 percent in 2022 to 17.0 percent in 2035. At the same time, the share of the older population (46-64 years old) will increase from 40.4 percent to 43.1 percent, and the share of the population between 25 and 45 years old will fall from 44.7 percent to 39.9 percent (see Figure 1). Given that the new additions to the labor force lack employment experience and face greater difficulty in matching skills, and the difficulty of adapting the skills of older workers to the needs of the changing industrial structure, the structural and frictional difficulties in employment will increase significantly. From a longer-term perspective, in the context of negative population growth, there will be a further slowdown in growth from new additions to the labor force, especially in the numbers of highly educated graduates of various kinds, leading to a lower rate of improvement in the human capital endowment of the labor stock, thereby increasing the difficulty of labor market matching.

Second, after experiencing the Lewis turning point from unlimited labor supply to a general shortage of labor in 2004 and the peak of the working-age population in 2010, China’s economy underwent a structural adjustment characterized mainly by the substitution of capital for labor, with a significant decline in the share of employment in labor-intensive industries and a significant weakening of employment absorption capacity, and a corresponding increase in the demand for higher-skilled labor for economic growth. Although the share of China’s manufacturing value added in GDP has long since begun to decline, manufacturing value added itself has continued to grow at a relatively fast rate, with a nominal average annual growth rate of 11.5 percent over the period 2004-2019. Employment in urban manufacturing work units grew at an average annual rate of only 1.5 percent over the same period, and has been in negative growth since 2013. If one argues that (national) manufacturing value added and (urban work unit) manufacturing employment are not consistent in terms of the scope of coverage and methods, a complementary proof can be provided by the most representative data on the employment of (local and outgoing) migrant workers within employment outside of urban work units. The share of manufacturing in migrant worker employment has been on a consistent downward trend since 2008, decreasing by 10.1 percentage points over the period 2008-2021.

Finally, one of the main manifestations of the negative macroeconomic effects of the COVID-19 epidemic is its impact on employment, especially informal employment. Both in manufacturing and in services, frontline jobs have seen a wave of layoffs and dismissals. Experience shows that when the macroeconomy undergoes a shock, factors are generated on both the supply and demand sides of employment that make employment suffer permanent losses. On one hand, periods of marked economic downturn are often also times of creative destruction, and industrial restructuring and employment substitution, which are long-term trends, will accelerate and occur earlier under conditions of increased pressure. On the other hand, older workers, especially migrant workers as well as some flexibly employed people, will exit the labor market at that point, and even those who are still looking for work may find it difficult to return to the labor market in the long run. Therefore, a large portion of the jobs lost to the epidemic and its economic impact cannot be expected to be regained.

Based on China’s stage of development and the patterns that pertain to that stage, the basic trend of economic growth will be a gradual deceleration, and thus the demand for labor will tend to decrease. At the same time, a deceleration in economic growth due to a decline in the potential growth rate is also the result of the aging of the population, and the reduction of the working-age population in particular. This endogeneity determines that labor supply and demand are balanced in the aggregate, and that overall, China no longer faces total employment pressure. However, China will also be in the most dramatic period of development in terms of the progress and application of technology, and of structural change in industry. In the course of the new technological revolution and the industrial revolution it is leading, an asymmetry between the destruction of old jobs and the creation of new jobs will inevitably arise. That asymmetry will be expressed most prominently where workers who lose their jobs do not have the human capital required for newly created jobs, leading to a mismatch between the supply and demand of skills. Looking at the nature of the new technological revolution, the speed of change is so fast that traditional mechanisms for human capital cultivation are hard-pressed to match it, and quantitatively speaking, job creation often fails to fully offset job destruction. This will be manifested as structural unemployment, or the phenomenon of withdrawal from the labor market due to chronic employment difficulties arising from skill shortages.

In the period since 2010, in which the contradiction in aggregate employment has given way to structural difficulties, China’s labor market has also exhibited a coexistence of job destruction and job creation. From the perspective of employment flows, job destruction is no doubt an increasingly important factor in the decline of net job creation (the difference between job creation and job destruction), in addition to the negative growth of the working-age population. Based on this logical relationship, although there is no corresponding data on job creation and job destruction in the existing employment statistics, it is still possible to make a less rigorous data comparison and roughly observe the quantitative relationship between the two. In Figure 2, the “new urban jobs” reported in official statistics is used as a proxy for new job creation. This indicator only records the number of new jobs created and does not subtract the number of jobs lost, so it is not a concept of net job creation. Consequently, new market players or start-ups, which emerge in huge numbers every day, are the main source of such new jobs or job creation. In addition, the “net gain” in employment shown in the graph is the net gain in jobs at the end of the year, obtained by subtracting the previous year’s total employment from the total employment at the end of the year, which can be seen as the net gain in overall employment.

Under conditions of absolute reductions in the working-age population, the reason why there is still a net increase of urban employment every year is mainly because the rural labor force is still being transferred on a certain scale to non-agricultural industries every year, and that is increasingly covered by urban employment statistics. From official data, we can obtain the number of jobs “created” and the “net gain” in the number of jobs, and the difference between them is the number of jobs “destroyed,” which is expressed as layoffs, unemployment, etc. The difference between the two is the number of “destroyed” jobs, which is represented by layoffs, unemployment and other reasons for withdrawal from the labor market. In 2019, for example, the officially reported number of new urban jobs was 11.86 million, while the net increase in jobs reflected in the records of the statistical system was 10.22 million, meaning that 1.64 million jobs were destroyed that year. It is worth noting that the number of jobs “destroyed” is likely to be grossly underestimated because the “net gain” figure includes factors such as changes in statistical coverage and methods, for example the gradual inclusion of informal employment such as migrant employment in urban employment statistics.

The concepts and data used in the discussion of job creation and job destruction here are in the sense of flows, and do not involve the number of unemployed or the unemployment rate as point-in-time data, i.e., the registered urban unemployment rate (number of unemployed) and the surveyed urban unemployment rate published annually by the National Bureau of Statistics are not taken into account. Admittedly, unemployment should be considered an important element in job creation and job destruction. However, the analysis here suffers in reality from inconsistent data coverage and methods. For example, the registered urban unemployment rate for 2019 and 2020 was 3.6 percent and 4.2 percent, respectively, and the number of registered unemployed was 9.45 million and 11.6 million, respectively. If we work backwards according to the numbers and rates, we can see that the denominator for calculating the registered unemployment rate, which is the so-called “sum of the number of employed persons and the number of registered unemployed persons at the end of the period,” was 262.5 million and 276.19 million, respectively, which is only 58.0 percent and 59.7 percent of the number of urban jobs in those years. That is to say, on one hand, the scope of registered unemployment statistics is much lower than the actual urban economically active population, and as for the surveyed urban unemployment rate, only the ratio but never the number has been published so far; on the other hand, the stock data on which calculations of the net increase in urban employment are based is not directly related to it. Therefore, when considering flow concepts such as job destruction and job creation, the number of unemployed can be taken as a constant and assumed to be relatively stable.

II. Long-term rising trend in the natural unemployment rate

Under conditions of innovation-driven economic growth, both the asymmetry between job creation and job destruction and the mismatch between the supply and demand of skills in the labor market tend to aggravate structural and frictional employment difficulties, thereby driving the formation of a long-term rising trend in the natural unemployment rate. In reality, the natural unemployment rate in China’s urban areas is undergoing a process of gradual increase, and from international experience, together with the change in China’s economic development stage and the new characteristics of the labor market, the natural unemployment rate can be expected to increase further in the post-epidemic period. This judgment comes from a consideration of multiple factors. Below are listed some of the key facts that serve as the basis. In addition, trends in the natural unemployment rate are discussed, using it as a benchmark to further observe the urban unemployment rate trend and its nature (see Figure 3).

First, the results of estimation in existing studies support the judgment that the natural rate of unemployment has increased and will increase further. Theoretically and methodologically, the natural rate of unemployment, i.e., the “non-accelerating inflation rate of unemployment” (NAIRU), can be estimated by filtering out the effects of economic cyclical changes on employment. Using data from 1987-2009, Du Yang and Lu Yang estimate the natural rate of unemployment in China to be between 4.05 percent and 4.10 percent, and find that it is on an upward trend. Zeng Xiangquan and Yu Yong estimate that the natural rate of unemployment in China has fluctuated in the range of 4.8 percent to 5.6 percent since 2002. Referring to these studies and based on some other observations, in Figure 4, two levels of the natural unemployment rate, 4.1 percent and 5.1 percent, are used as benchmarks for before and after the turn of the century, respectively, to observe how the actual unemployment rate has fluctuated around them. It can also be seen in Figure 4 that the real unemployment rate has been more frequently and significantly higher than the natural rate of unemployment since the outbreak of the COVID-19 epidemic. If this situation continues, the natural rate of unemployment will take another step upward.

Second, there is tendency for employment groups that are vulnerable to structural and frictional factors, such as migrant workers, new graduates of various types, and older workers, to increase their shares of the total employment population. This labor market characteristic is the seedbed, as it were, of structural and frictional unemployment phenomena. Vulnerable groups, characterized by informal employment, are also prone to become victims of natural unemployment. Observing the diversity and heterogeneity of China’s population and worker groups, and the resulting easily differentiated characteristics, provides a micro-analytical basis for accurately understanding structural and frictional unemployment or employment difficulties. In 2021, 34.9 percent of China’s population did not live in the same place as their household registration, with the mobile population accounting for 26.6 percentage points of this total. Out of the total labor force of 780 million, some 750 million were actually employed, of which 76.4 percent were in non-farm employment and 61.6 percent were employed in urban areas. If within urban employment we take the people engaged in temporary jobs, employed people dispatched by labor companies, and those employed by market entities such as individual entrepreneurs as representative of flexible employment or informal employment, the corresponding proportion exceeds 30 percent. The total number of migrant workers in non-agricultural employment was 290 million, of which 121 million were employed in their own towns and cities and 172 million were migrant workers (133 million migrant workers were still living in cities and towns at the end of the year).

Finally, both theory and experience show that after each recession or crisis, the structural problems in the labor market tend to intensify and the natural rate of unemployment increases. Among cyclical economic changes, recessions usually imply the destruction of a part of production capacity and the elimination of corresponding enterprises from the market. Accordingly, new start-ups often play a greater role in economic recovery and hence in the recovery of production capacity. Compared to the eliminated enterprises, new start-ups represent a better production function, and thus the macroeconomic process from recession to recovery is also a reorganization of production factors, that is, a process of creative destruction in the Schumpeterian sense. Therefore, compared with the pre-recession period, the demand of enterprises for workers’ skills is significantly higher after economic recovery, which inevitably leads to some workers having difficulty adapting. The recurrence of so-called “jobless recoveries” in the U.S. recession and recovery cycles from 1992 to 2008 is a manifestation of increased natural rates of unemployment. In general, China’s future economic growth will be increasingly innovation-driven. The creative destruction mechanism will play a greater role, and the frequency of job-changing by workers will increase significantly.

III. Potential for fuller and higher-quality employment

While economics textbooks define full employment as a situation in which the unemployment rate is at a level that “does not cause accelerating inflation,” this is not always the case in reality. On one hand, the so-called “natural unemployment” in macroeconomics or labor economics is not “natural.” That is to say, “natural” states of unemployment, and even some withdrawal from the labor market, can be avoided. On the other hand, the natural rate of unemployment is not always statistically certain and precise, and mechanical estimation under a priori assumptions may not reflect a complex and changing reality, and may even sometimes cause distortion of reality and thus misguided policies. In general, while an overly low natural rate of unemployment may send a distorted signal of insufficient incentives to invest in human capital, when the natural rate of unemployment is overly high but is not estimated, and it therefore does not attract due policy attention, the result may be underutilization of labor resources, affecting both the improvement of people’s livelihood and causing a loss of resource allocation efficiency.

Therefore, in the case of labor market shocks, macroeconomic policies need to respond to cyclical unemployment phenomena in a timely manner. At the same time, economic and social policies should be used in an integrated manner, striving to eliminate the structural and frictional factors that prevent full employment, so as to prevent the consequences of short-term shocks from becoming chronic and normalized. Under normal macroeconomic conditions where there is no cyclical unemployment phenomenon, reducing the natural rate of unemployment from all angles should be the main objective of active employment policy placed at the macro policy level. In China’s economic reality, there are still factors that have the effect of raising the natural rate of unemployment, that is, of creating “unnatural” natural unemployment. Identifying these factors and analyzing them can help to identify the right entry points for promoting labor market development and implementing active employment policy, thereby promoting the goal of fuller and higher-quality employment.

First, employment for some groups of people does not conform to the essential meaning of “full employment.” Whether it is the surplus labor still remaining in agriculture, or people employed in low-productivity industries and the informal sector, they do not fall within the unemployment category according to survey coverage and methods. In China’s economic reality, however, there are still significant productivity differences between agricultural and non-agricultural industries, between various industries, and between different enterprises in the same industry. This situation is to a large extent caused by barriers to the flow of production factors, and in particular by structural and frictional factors in the labor market. As long as these barriers persist, the allocation of labor resources cannot reach an optimal state and full employment cannot be truly achieved. This fact tells us that active employment policy needs to be implemented in concert with industrial policy and competition policy in order to smooth the channels for the flow of production factors, especially labor.

Second, not all groups with employment intentions can be successfully employed through the usual procedures such as “job searching” and “training.” In reality, there are always some workers who cannot be employed by the labor market alone or even by the conventional support policies. There are two typical manifestations of this failure to achieve employment intentions. The first phenomenon is the so-called Discouraged Worker Effect, which refers to the situation where a person is unemployed for a long time and fails to find a job, so that he or she loses confidence and stops actively looking for a job. In the statistics, this is classified as “withdrawal from the labor market” rather than unemployment. Considering that the “withdrawal” of this group is not a willing choice, the U.S. Bureau of Labor Statistics also calculates corresponding unemployment indicators, such as U-6, as a supplement to the conventionally defined unemployment rate indicator (U-3). The second phenomenon is that some workers with employment intentions face greater barriers to employment than others, including high family caregiving burdens, chronic illnesses, disabilities, and advanced age. Both of these phenomena require the government to take greater supportive measures for specific groups and their special vulnerabilities.

Third, not all jobs can be created by relying on the market. Due to their natural properties or changes at particular times, some economic activities have strong externalities, where social returns are greater than private returns. Spontaneous market mechanisms are not sufficient to reflect the real demand for them or to create satisfactory returns on investment and compensation for workers, thus often leading to their under-supply. Examples include: socialization of household chores through childcare, which eases burdens on families and helps stabilize the fertility rate; action on resources, the environment, and climate change that contribute to sustainable development; public health provision to improve health levels; community services to improve the quality of life; the care, assistance, and respect for the elderly that are necessary to cope with aging; and public service jobs created to stabilize employment in the event of shocks. Obviously, the creation of such jobs relies on a certain degree of policy intervention to provide additional incentives.

Finally, the heterogeneity of the workforce creates complexity in labor market matching. The new additions to the labor force graduating from schools of all levels and kinds are often plagued by structural and frictional unemployment, resulting in a “paradox” wherein the higher the education level, the lower the degree of matching. While this certainly reveals the employment difficulties that university graduates face, it tends to be misleading as to directions for educational development. On one hand, as education levels rise, human capital specialization increases and generality decreases, or the singularity of skills increases, the range of employment choices narrows accordingly and matching becomes more difficult. On the other hand, the structural upgrading of industries also places higher requirements on human capital, and the direct signals are demands for specialized skills. In order to understand these two regular phenomena taken together, one must perceive that, under conditions of constant technological change and industry structural change, demands for skills are not only ever-changing but also rapidly changing. This creates a market failure, as both job seekers and employers are unable to accurately identify and judge labor market demands on an individual level. It is clear that only governments can coordinate education, industry, and employment to break this labor market matching paradox.

IV. Conclusions and policy implications

International experience shows that when there is a shock to economic growth, if the policy response is not timely and appropriate, the shock will leave a “scar” that affects subsequent development, producing a so-called “hysteresis effect” that makes the post-recovery norm in economic growth less favorable than before. Similarly, if the response to cyclical unemployment is not timely, comprehensive, and appropriate, or if it relies only on macroeconomic stimulus without simultaneously addressing the structural and frictional contradictions in the labor market, it may leave a “scar” on the employment issue, such that the post-recovery labor market operates with a higher natural rate of unemployment. Although the fundamentals of China’s economy have not changed due to the temporary macroeconomic downturn, the labor market landscape may indeed change as the population enters an era of negative growth and economic growth faces new challenges.

This change can be recognized from both the supply side and the demand side. First, the rapid decline in the working-age population will bring greater supply-side constraints, and the input-driven growth model will become increasingly hard to sustain. Shifting the engine of growth to an innovation-driven track requires creating an environment of creative destruction and increasing productivity through the selection of winners and elimination of losers. Correspondingly, employers’ demands for workers’ skills will increase significantly, the difficulty of skill-matching in the labor market will increase, and much higher precision will be required in the targeting of active employment policy and public employment services. Second, a declining and more deeply aging population means that demand-side factors, especially consumption, are increasingly becoming normalized constraints on economic growth. Since the employment status of workers affects household consumption and aggregate demand through its impact on people’s incomes, active employment policy needs to have a demand-side macro perspective. For example, the surveyed unemployment rate of the urban population aged 16-24 in June 2022 was as high as 19.3 percent, which was 4.29 times higher than the surveyed unemployment rate of the population aged 25-59 (4.5 percent) in the same period. Since the consumption of the former age group is significantly higher than that of the latter, as the sample survey data show that the arithmetic mean of consumption expenditure of the urban population aged 16-24 is 1.45 times higher than that of the population aged 25-59, the high youth unemployment rate resulting from the macroeconomic cycle will become a serious drag on macroeconomic recovery.

Therefore, there is still a lot of room for policy and institutional operations to deal with the current unemployment problem, which is of a short-term shock nature, especially with regard to managing cyclical unemployment, which needs to focus on preventing labor market shocks from lasting too long and damaging specific groups of workers too much, thereby causing cyclical and transitory factors to turn into long-term and structural factors. As the epidemic is gradually brought under control, production and consumption activities will return to their regular tracks, economic growth potential can be released as soon as possible, and cyclical unemployment is expected to gradually disappear. However, since unemployment is not determined by cyclical factors alone, active employment policy should devote greater efforts toward eliminating structural and frictional difficulties and reducing the natural rate of unemployment. Economic development is a combination of medium- and long-term growth and cyclical changes, subject to the constraints of growth potential on the supply side and the effects of cyclical fluctuations on the demand side. Accordingly, active employment policy faces multiple tasks at the same time, including promoting long-term employment, managing natural unemployment and helping with employment difficulties under normal conditions, managing cyclical unemployment when shocks are encountered, and constructing labor market institutions.

Placing priority policies for employment at the macro policy level requires that these tasks be organically combined, making the policy measures stronger and more effective. First, job creation should correspond to long-term economic growth. In order to expand the number of new jobs, a sustained flow of new market players should be maintained by improving the business environment. To address the coexistence of job destruction and job creation, and in response to the mismatch between the demand for human capital and its cultivation, as well as the resulting asynchrony between employment expansion and economic growth, the government should pay more attention to promoting the effective linking of education, employment, and training. Second, the function of the labor market system should be improved and enriched. Social mechanisms for labor rights and labor-management relations determined by labor laws and regulations should be established and regulated, so that they can fully reflect the special nature of the labor factor, of which people are the vehicles, and always ensure the protection of workers in creative destruction. Third, to address structural and frictional factors in the labor market, public employment services should focus on special support for the young, elderly, and informal employment groups, so as to reduce the natural rate of unemployment. Finally, macroeconomic policies should respond to cyclical unemployment by using monetary policy tools to create an accommodative monetary environment and unconventional internal and external demand to drive growth back to the potential growth rate, and also by being more adept at using fiscal policy tools to stabilize incomes and consumption through social insurance, family subsidies, and job creation.