(This essay was published in Hong Kong Economic Journal on 11 May, 2016.)
Service industries have grown rapidly in Hong Kong as a result of the opening of China, the globalization of economic activity, and the rapid growth of the Asia-Pacific region. From 1980 to 2015, the real GDP share of the service sector grew from 75.0% to 92.3% (see Figure 1). Meanwhile, the manufacturing of goods declined from 13.8% to 1.5%.
In the early-1990s, many regarded this transformation as an economic catastrophe that should be averted. They worried that deindustrialization or hollowing out would retard productivity and economic growth, increase the proportion of low quality service jobs, and worsen income distribution among households and workers.
They called upon government to intervene to arrest the process through aggressive industrial, innovation, and technology policies. Such voices never disappeared completely, and on the surface the concerns about hollowing-out and its consequences seem to have been vindicated by subsequent developments.
Twenty years ago, I argued that fear of deindustrialization was unfounded pessimism (at the Symposium on Services Promotion, Hong Kong into the 21st Century: The Servicing Economy, 12 March 1996). I pointed out that the most important feature of service sector growth at the time was that most of it was concentrated in intermediate producer services and not final consumption services.
Intermediate services provide input to the production of other goods and services. An expanding cross-border manufacturing base was the driver of producer services that in turn supported the base in the Pearl River Delta.
Distinguishing between final and intermediate services is not a trivial matter with existing statistical data because the critical difference between the two is not in the type of service, but its use. For example, hotel accommodation service is an intermediate producer service if an executive on a business trip purchases it, but is a final consumption service if the same executive is on a vacation. Similarly a meal purchased at the hotel’s restaurant by an executive on a business trip would be an intermediate producer service if it is part of his work, but not if he is meeting with an old personal friend from high school days.
To measure intermediate producer services correctly, I first divided output into two sectors: the goods-producing sector and the service-producing sector. The latter is further subdivided into producer, consumption, and government services sectors. Producer services are measured as a residual category after deducting consumption services and government services from total services. This measure of producer services is a correct approximation of all those services that are used as intermediate inputs.
Using this taxonomy, we can see that almost all the growth in services is in producer services. The value-added share of producer services increased from 31.2% to 47.5% over the period 1980-2015. Consumption services are all those services that are used in final consumption and are obtained from the expenditure-based estimates of GDP. The growth of consumption services is much more limited; the value-added share increased from 34.8% to 38.2%. Government services, which are the same as those from the production-based estimates of GDP, declined from 8.9% to 6.6%.
Producer services grew rapidly in the period before 2005, but began to stagnate afterwards. The growth of the producer service sector is closely linked to the expansion of manufacturing activities of Hong Kong firms across the border in south China.
Producer services are the vehicle by which new technology is introduced into the goods and services production process. New technology can lead to lower production costs, the development of improved and new products, and new and more efficient methods for distributing goods and delivering services. These are service inputs that respond to greater demand for goods and services that, in turn, contribute to and change the mix of demand for goods and services. They account for the enormous growth of labor productivity in the production of goods and services.
Bankers, accountants, engineers, scientists, marketing, communications and transportation professionals, whose output is used by firms, generate a large proportion of all high value-added producer services. The production of education and medical services are also inputs into the creation and maintenance of society’s stock of human capital. Without this expenditure, the productivity of any population would quickly decline.
In free market economies, the satisfaction of consumers is continuously increased through better and cheaper products. Entrepreneurial firms draw heavily on highly-educated workers. They also develop and apply new technological and scientific knowledge. They sell advice and deliver creative and innovative services to other firms. These firms in the producer services industries are the main vehicles through which the human and knowledge capital of society is produced and introduced into the production process. Evidence of the higher educational attainments of producer services sector workers when compared with those in the manufacturing sector supports this view.
Human capital is enhanced not only through investments in education and on-the-job training, but also through investment in health care, nutrition, soft skills, and the moral and ethical standards that make for honest, reliable workers and create a willingness to work hard. Knowledge capital is all the scientific and engineering knowledge that permits the design of more efficient machinery and products. It also includes that knowledge that makes for the efficient organization of firms and society in general. Therefore, even if the measured rates of productivity growth in these producer services industries are not high, they are the reason why firms can have high rates of productivity growth.
Economists have described the role of producer services in enhancing productivity in many different ways. Economists of the Austrian School described it as increasing “roundaboutness” with more stages of production and employing an increasing number of intermediate inputs. Adam Smith described it as specialization through the market-driven division of labor. Today the process of market specialization is described variously as “dehiving,” “unbundling,” “vertical disintegration,” or “contracting out.”
The producer services sector in Hong Kong is a highly heterogeneous industry. The rapid expansion of this sector reflects a dynamic market process at work, where workers are reallocated from low-productivity manufacturing jobs to high-productivity service jobs. Fears of low productivity following deindustrialization were unfounded because the rise of our producer services sector supported the growth of the manufacturing base across the border in the Chinese mainland.
This was the situation in Hong Kong, at least until 2005. But since then the share of producer services has stagnated atbetween 47% and 48%. The slowing growth of producer services is the primary reason why real GDP growth has also slowed (see Figure 2).
Although the economy picked up after the Asian Financial Crisis ended, it has remained lackluster. There is much talk that the economic base is too narrow. Finance, property, trade and logistics have been the primary engines of growth, but globalization is slowing, and the future of international trade and investment flows is cloudy. Tourism is not a high valued-added service export although it has helped to keep unemployment rates low.
Recent hope is focused on cultural and creative industries, medical services, education services, innovation and technology industries, testing and certification services, and environmental industries. Government has been repeatedly called upon to provide incentives to support these industries. But all these sectors depend heavily on the human capital of the population having the appropriate skills.
Unfortunately investment in tertiary education only started to increase after economic globalization and China’s opening had already started. The rate of expansion was also very rapid. University entrance places funded by the University Grants Committee increased rapidly from 2,000 places per annum in 1983 to over 14,000 per annum in 1994, but have since flattened out.
The result has been to increase the number of tertiary-educated graduates in the population aged 20-29 from 100,000 in 1981 to 326,000 in 2001 and 555,000 in 2011 (see Figure 3). The number of graduates in the population aged 30-39 increased from 61,000 in 1981 to 297,000 in 2001 and 464,000 in 2011, and is projected to rise to 678,000 in 2021.
The growing number of tertiary graduates could in principle provide the human capital stock necessary for nurturing a broader economic base. The lack of expertise in new industries is a handicap that could be surmounted given time. However, the Asian Financial Crisis struck at a most unfortunate time and probably postponed the discovery of new industries by a decade or longer.
The uncertain environment of the Asian Financial Crisis and the lack of opportunities discouraged the earlier generation of tertiary graduates from taking risks. Subsequent graduates have to date probably been too inexperienced to venture into new industries. But in the coming decade there is a good prospect that some of them will do so.
The large numbers of tertiary graduates in their age cohort is also depressing their salaries in the labor market and this will be an additional incentive for some of them to go entrepreneurial. There are more and more tertiary graduates aged in their thirties and it seems inevitable that some of them will start new successful businesses. I will not be surprised if the next wave of producer services emerges from this generation.
The slow economic growth and depressed wages that new tertiary graduates have faced over the past decade is a consequence of the very rapid increase of university places during a short period from 1983 to 1994 and the timing of the onset of the Asian Financial Crisis. Deindustrialization is merely a consequence not the cause.