logo

Has employment growth through labour-intensive industrialisation become history?

Rizwanul Islam | Wednesday, 13 November 2019


Despite steady economic growth for about two decades and an acceleration in growth in recent years, the economy of Bangladesh still exhibits characteristics of dual economies, viz., the existence of a traditional segment where there is surplus labour and of a modern segment where investment can result in growth of output as well as absorption of surplus labour available in the traditional segment. There are models of economic development in dual economies (e.g., those developed by Lewis, and Ranis and Fei, writing in 1954 and 1961 respectively) that describe economic development in countries with surplus labour as a process of economic transformation towards modern sectors and the transfer of surplus labour from traditional sectors to modern sectors until such surplus is exhausted and real wages start rising - a stage that is often referred to as the "Lewis turning point".
Indeed, development in several countries of East Asia (viz., Republic of Korea, Taiwan-China, Hong Kong, and Singapore) during the 1970s and 1980s followed that general pattern. A number of the countries of South East Asia, especially Malaysia and Thailand, and to a lesser extent, Indonesia and the Philippines, were also making good progress in their journey towards the so-called "Lewis turning point" - at least until the time they were hit by the economic crisis during 1997-98.
The model of economic development described above became so influential that it was widely regarded as something to be emulated by other developing countries. Nearly half a century has passed since then; and there are many countries in the developing world which are still characterised by surplus labour and are at various stages of development. They are often advised to pursue open (or export-oriented) economic and trade policies - the expectation being that such strategies would enable them to achieve development, absorb their surplus labour, and reduce poverty.
If one looks at the actual experience, one would see that the countries of South Asia (including Bangladesh) have not been able to attain the same degree of success attained by countries of East and South East Asia (ESEA) in terms of the level or pattern of growth required for absorbing surplus labour and reaching the Lewis turning point. And that was despite their shift from inward looking economic policies towards outward looking and export-oriented policies. Such experiences, in turn, raise the question whether a repeat of the type of employment intensive economic growth through industrialisation achieved by the countries of ESEA is at all possible. This question becomes even more pertinent in the context of two recent strands of discussions.
The first concerns the possibility of premature deindustrialsation while the second is a concern arising out of the availability of new technology that is regarded as the hallmark of the fourth industrial revolution. If deindustrialisation starts before a country's economy has absorbed all its surplus labour, the task of further labour absorption would become correspondingly difficult. Likewise, a premature spread of labour-saving technologies may make it difficult to attain the Lewis turning point.
The purpose of the present article is to address these two issues and examine the difficulties faced by contemporary developing countries of Asia in achieving development and employment through labour-intensive industrialisation.
Labour-intensive industrialisation and development in East and South East Asia: A brief overview
Amongst the late industrialisers, four countries of East and South East Asia (ESEA), viz., Republic of Korea, Hong Kong, Singapore, and the Taiwan province of China (henceforth referred to as Taiwan), are regarded as having succeeded in achieving development through export-oriented industrialisation. However, in terms of possible emulation by other developing countries, the latter three are often considered to be non-representative because of their small size and other special characteristics. Therefore, the present article uses the experience of Korea as a success story and examines how far the other developing countries of Asia have been able to develop along the path trodden by that country.
On growth experience in the region as a whole, one would notice several features. First, Korea and some other countries of the ESEA region achieved high and sustained rate of economic growth over several decades (Korea starting in the 1960s and some others starting from the 1970s). The rate of economic growth (measured by annual compound rate of GDP growth) has been consistently higher in the ESEA countries compared to the South Asian countries.
The second aspect of the growth experience has been the high rate of growth of manufacturing industries, especially in Korea and also in some other ESEA countries. Indeed, elasticity of growth of manufacturing with respect to overall GDP growth was over 2 in Korea in the 1960s - meaning that growth of output in manufacturing was twice that of GDP. This figure dropped to just below 2 (1.9, to be precise) during 1970-80, and to below 1.5 only during the 1980s. In Malaysia also, the elasticity of manufacturing output growth with respect to overall GDP growth has been between 1.5 and 2 for almost three decades (during 1970-1996). Indonesia and Thailand also had similar experience.
On the other hand, growth of manufacturing output in relation to overall GDP has been in the range of 1.3 to 1.4 in India, and lower in Pakistan. In Bangladesh and Sri Lanka, there has been a considerable degree of fluctuation in the growth of manufacturing output compared to GDP growth. On the whole, it is quite clear that not only has overall economic growth been higher in Korea and a few other countries of ESEA (The Philippines being an exception) than in South Asia, the manufacturing sector has been the major (and a more important) driver of growth in the former. And that enabled Korea (and other countries of ESEA in varying degrees) to achieve structural change in their economies of a kind that facilitated labour absorption at a rapid rate.
The third important aspect of economic growth in the countries of ESEA compared to South Asia has been the superior export performance of the former and greater openness of their economies. The share of exports in total GDP has been consistently much higher in the countries of ESEA compared to those in South Asia. Even after the economic reforms introduced in the 1980s and the 1990s and the gradual opening up of these economies, the share of exports in GDP in Bangladesh, India and Pakistan has remained below 25 per cent, whereas it has ranged between 40 and 130 per cent in the countries of ESEA. More important from the point of view of industrialisation, the growth rate of manufacturing exports has been much higher in the countries of ESEA compared to South Asia.
From the point of view of labour-intensive industrialisation, it is important to look at the composition of industrial exports. And in that regard, the experience of Korea in particular is noteworthy. During the 1960s (i.e., the early stages of export-led development) the share of 'light industries' in manufactured exports remained at high levels - ranging from 83 per cent in 1964 to 89 per cent in 1968. Only in the 1970s, this share started declining and went down to 60 per cent in 1978. Between 1960 and 1970, direct employment in exports in Korea increased by more than 18 times, or at an average annual rate of 34 per cent. As a result, the share of export industries in total employment expanded from 5 to 22.5 per cent during the same period. In fact, the degree of capital intensity (measured by capital-labour ratio) of Korean manufacturing declined during the 1960s till about the early 1970s. Capital-intensity of Korean manufacturing and of exports in particular started rising only in the early 1970s.
It should be added here that even during the period of labour-intensive industrialisation in Korea (and other countries of South East Asia), there was growth in labour productivity. The high rate of growth-induced expansion in employment was accompanied by high growth in productivity as well as earnings of workers.
The role of manufacturing in attaining structural change in employment can be illustrated with a few figures. In Republic of Korea, the share of manufacturing in total employment had risen to 23 in 2000 before it started to decline. In Taiwan-China, the peak share of manufacturing in employment was even higher - 32 per cent in 1990. The rise and decline of the share of Malaysia's manufacturing - shown in Figure 1 - illustrates the point very clearly. From less than 16 per cent in 1982, the share went up to over 23 per cent during the late 1990s and started to fall after 2001. By 2016, the share had fallen to less than 17 per cent. In contrast, the corresponding figure for India never exceeded 13 per cent. In Bangladesh, it went up to 16 per cent in 2013 and then declined to 14 per cent in 2016-17 (Figure 2).
Is South Asia facing the possibility of premature deindustrialisation?
The term deindustrialisation is used to describe the decline of the share of manufacturing in the total output and employment of an economy. In the literature on economic growth and structural change, this is considered normal for a growing economy. For example, in the UK and USA, the share of manufacturing employment rose to 35 and 36 per cent respectively before it started to decline. The level of per capita income at which it happened in those countries as well as in some other developed countries of Europe varied between US$9,000 and US$11,000 (at 1990 prices).
In contrast with deindustrialisation experienced by developed countries, some developing countries are facing it at a much lower level of development - a phenomenon that is being referred to as "premature deindustrialisation". While there may be different explanations for this in different countries, the impact, especially on the generation of productive employment through structural transformation of the economy is likely to be negative. It is therefore important to see whether developing countries, especially those with surplus labour and have not yet succeeded in attaining the Lewis turning point are facing the danger of this happening. It is in that context that the question becomes particularly important for Bangladesh and other countries South Asia.
A number of studies (including Rodrik, 2013) point out that there are developing countries where deindustrialisation started much below the levels of per capita income at which this happened in the developed countries. Rodrik (2013) mentions India alongside Brazil and China among countries that have already experienced such premature deindustrialisation. On India, he says:
"Manufacturing employment there peaked at a meager 13% in 2002, and has since trended down."
But data for more recent years do not support Rodrick's contention; in India, the share of manufacturing in total employment did not decline. In fact, between 1993-94 and 2004-05, it increased a bit - from 10.5 per cent to 12 per cent. Furthermore, between 2004-05 and 2011-12, the share increased to 12.9 per cent. The share of manufacturing in India's GDP also shows a small increase during 2000 to 2014 - from 15 to 17 per cent. While these figures do not signal a spectacular increase in the share of manufacturing either in employment or output, they do not warrant the conclusion that India has already experienced premature deindustrialisation.
Bangladesh witnessed a rise in the share of manufacturing in both GDP and employment. During 2000 to 2014, the rise in the share of GDP was small. But the rise in the share of employment during 1999-2000 to 2016-17 was substantial - from 9.5 per cent to 14.4 per cent. Of course, the country has been able to attain this mainly through high growth of one labour-intensive export-oriented industry, viz., ready-made garments. And employment growth in that industry has stagnated in recent years -raising fears about where to find jobs.
Further light on the question of deindustrialisation can be thrown by data on growth of employment in the manufacturing sector (Table 1). A couple of points may be noted on the basis of this data. First, in all the four countries for which data are available for two periods, there has been a decline in the rate of employment growth in the sector. Second, the decline has been particularly sharp in the case of India and Pakistan. And during the 2000s, the rate of growth of employment in both those countries has been very low - 1.5 and 1.0 per cent respectively. In Bangladesh also, there has been a decline, although the extent has not been so much. But even in this case, the decline has been quite sharp in recent years - to 2.33 per cent between 2015-16 and 2016-17 compared to the annual average of 5.12 during the previous five years.
What can be said to conclude this discussion on the danger of premature deindustrialisation in the countries of South Asia? If the shares in GDP and employment are used as indicators, India and Bangladesh clearly have not entered the phase of deindustrialisation. This, however, should not be taken to imply that these countries are firmly positioned on the path of industrialisation, and manufacturing will be able to perform the role of absorbing the surplus labour in those countries. The share of the sector in total employment is still way below that attained by the peak share of 23 per cent attained by both Republic of Korea and Malaysia (both in 2000). In India, the share rose from 10.5 per cent to 12.9 per cent over a 15-year period. In Bangladesh, the rise was from 9.5 per cent to 14.4 per cent over a 17-year period. Whether the figures in these countries will ever cross the 20 per cent mark, and if so, when, remain open questions. Given the pace and pattern of industrialisation in the countries of South Asia, even if they are not showing signs of deindustrialisation in a technical sense (i.e., decline in the share of manufacturing in GDP and employment), they may remain prone to this possibility. And the danger is becoming increasingly real with the availability of labour-saving technologies associated with the fourth industrial revolution - an issue to which we now turn.
Does labour-saving technology pose a challenge?

In the history of evolution of human society, technological progress has been a continuous process, and such progress has been associated with automation of various degrees and kinds. That, in turn, has had significant implications for employment and the world of work. If the issue of development is looked at in a longer-term perspective, it would be necessary to take this into account and see how the employment situation in developing countries may be influenced by technological changes that are likely to take place.
The world is currently witnessing the fourth industrial revolution, the basic characteristics of which include the use of robots, artificial intelligence, nano technology, and biotechnology. A common perception in that respect is that this is going to threaten employment of human beings. Robots are making inroads even in countries that still characterised by the existence of surplus labour (e.g., Bangladesh); and they could be used in various sectors including manufacturing industries. If the use of such labour-saving technology becomes a reality even before millions of underemployed workers find good jobs characterised by high productivity and incomes, there will be serious problems of mass unemployment and underemployment. But how realistic would it be to paint such a scenario for the future - even if one considers a period of several decades?
The first and immediate impact may be the loss of jobs in some sectors and specific operations as machines may indeed replace some human jobs. But in addition to this immediate impact, technological progress leads to changes that may have a positive effect on employment. For example, one positive impact is often a rise in productivity leading to a decline in prices and a rise in the demand for products. For example, think of how many more pieces of dresses per capita an average household consumes today compared to half a century ago. Likewise, compared to a low proportion of households having telephone sets, how many mobile sets are used per household today compared to the previous generation? Such changes in consumption patterns have positive impact on both output and employment.
Second, technology replaces certain tasks rather than complete occupations. Of course, new jobs that are created are likely to require different types and levels of education and skills compared to the jobs that may have been lost.
Third, automation, by raising the productivity of workers, creates a necessary condition for wage increases. Moreover, by reducing the drudgery of manual jobs, machines may lead to improvement in the quality of jobs. What is also important to note is that only in a small proportion of occupations, jobs are completely automated. Machines often work together with human beings - thus creating positive complementarity and raising productivity.
Thus, if the overall effects of technology and automation on development and employment are considered in a longer-term perspective, it would be difficult to come to a clear conclusion as to what scenario is going to unfold. Even from a short-term point of view, very little is known on what is happening on the ground in Bangladesh.
There is, of course, some evidence to indicate that automation is indeed making inroads into manufacturing industries, including in labour intensive industries like the ready-made garment industry of Bangladesh. And that could be one important reason for the recently observed decline in the growth of employment in the sector. For example, a report in the Wall Street Journal (22 February 2018) mentions that their reporter found the use of machines made in Germany and Japan which require comparatively less labour than conventional machines. Not only that, some factory owners told the reporter that this trend is likely to continue. Likewise, in a report published in 2016, the International Labour Organisation mentioned that labour use in the RMG industry of a number of Asian countries is going to decline due to automation.
Whether such practices, i.e., the use of labour-saving technology, would be adopted predominantly by larger firms or whether it would be widespread is something that cannot be predicted with confidence.
Concluding observations
It is by now well-known that success in fully absorbing surplus labour through labour-intensive industrialisation has remained limited to a small number of developing Asian countries. How does one explain this? The oft-quoted quip that the laggards had been pursuing inward-looking economic policies that constrained their growth prospects is no longer as valid as it was during the late 20th century. Most countries of developing Asia have now adopted more outward-looking strategies of development and have introduced market-oriented economic reforms. Some have also been able to narrow down the differences in the rate of economic growth. Among the countries of South Asia, India has achieved growth rates that are comparable to those achieved by the successful countries. Bangladesh has also attained growth on a sustained basis and has accelerated it. And yet, success in achieving the kind of industrialisation and labour absorption seen in the pioneer countries has been elusive.
Not only has been the growth rate of manufacturing relative to GDP growth lower in the countries of South Asia, they are also facing the danger of premature deindustrialisation. Although they have not experienced clear declines in indicators like the share of manufacturing in GDP and total employment, the rate of growth of manufacturing employment has fallen in most countries. With the share of employment stagnating at much lower levels than the peak shares attained historically by successful countries like Korea and Malaysia and growth rates of employment falling sharply, it does not look like that such success can be repeated.
In addition to lower rates of output growth and lower (and declining) employment intensity of growth, the countries of South Asia are facing the prospect of automation in manufacturing that is associated with the fourth industrial revolution. Although the danger of employment declining due to automation is sometimes overblown, if the manufacturing sector does indeed embrace automation on a large scale, the ability of the sector to generate employment may decline further.
If emulation of success achieved by the front-runners was already difficult in the twentieth century, it must have become much more so in recent years. The global context for trade and industrialisation has changed substantially in various respects. First, competition based on cheap labour has become much more intense with the emergence of new late starters like Bangladesh, Cambodia, Laos, and Vietnam. In addition, there are large countries like China and India who not only have the advantage of cheap labour but also fulfill one of the basic preconditions for success, namely, a prior degree of industrialisation. Given this scenario, the quest for productivity growth has resulted in a greater focus on technology (and hence, capital) than was the case with front-runners like Korea and Taiwan. Instead of declining capital-labour ratio (Korea in the sixties) and the use of lower level technology including second hand machines (e.g. in Taiwan), the recent experience has been one of capital deepening and continuous efforts to be at the frontier of technology.
The second respect in which the global context has changed is the functioning of labour markets. What has been referred to as "superior labour market performance" in East and South East Asia basically refers to a situation in which there was labour repression either in the form of wage repression (euphemistically referred to as "wage restraint") or rights repression and absence of union power. In the current global context, both seem to have become more difficult - given greater union power and international pressure for compliance. If labour market flexibility implies labour repression of either kind, it is neither desirable nor feasible.
The overall policy environment may get further tilted in favour of the use of capital in situations where measures introduced in the name of creating incentives for investment result in artificially lowering the price of capital. In the absence of any such incentives for the use of labour, the relative factor price gets distorted and results in excessive use of capital -- either in the form of investments in capital-intensive products or technologies. Given the changed international context mentioned above and the kind of policy environment prevailing in the economies of late starters in labour-intensive industrialisation, it is not surprising to see a much lesser degree of success in such countries than in the front runners.

Dr. Rizwanul Islam, an economist, is former Special Adviser, Employment Sector, International Labour Office, Geneva. This article is based on a chapter of his recently published book titled Full and Productive Employment in Developing Countries: Towards the Sustainable Development
Goals (Routledge, London, 2019).
[email protected]