What does South East Asia teach us about structural transformation and what would Arthur Lewis have said about it?

Kuala Lumpur, Malaysia

This week, ESRC GPID Director, Andy Sumner, continues the discussion on structural transformation and inclusive growth in South East Asia.

We asked last week what can the experience of South East Asia teach us about structural transformation and inclusive growth and the ‘developer’s dilemma’?

At the outset, it is worth noting that surprises are always possible as the elections in Malaysia show. But what does history tell us?

A quick recap

To recap, developing countries are seeking economic development which requires structural transformation and productivity growth via the intra- and inter-sectoral reallocation of economic activity, notably labour, and which tends to push up inequality, as Kuznets argued. At the same time, developing countries are also seeking some form of broad-based or ‘inclusive growth’ which can be defined in various ways; typically, as who benefits from growth, in what way, and by how much.

Inclusive growth is best achieved, or the inclusivity of growth maximised, with steady or even falling inequality, and certainly not with dramatically rising inequality. In short, the ‘developer’s dilemma’ is a distribution dilemma or tension at the heart of economic development. One region of the world, South East Asia, has achieved both substantial structural transformation as well as broad-based inclusive growth.

In last week’s blog we posed a set of questions: how deep, and of what nature was the structural transformation and inclusive growth in Malaysia, Indonesia and Thailand? How did Malaysia, Indonesia and Thailand deal with the distributional tensions that rapid transformation often entails? And how did Malaysia, Indonesia and Thailand achieve such rapid economic and social transformation in a relatively short space of time?

In other words, what does the experience of those countries imply for theories of economic and social transformation?

This week we consider the first question: how deep, and of what nature was the structural transformation and inclusive growth in Malaysia, Indonesia and Thailand?

At first glance, the structural transformation and inclusive growth in Malaysia, Indonesia and Thailand looks dramatic and impressive. We can consider three waves of transformation which were interrupted abruptly by crises. First, the 1970s to the mid-1980s. Second, the mid-1980s to the mid-1990s. Third, the period since the late 1990s.

The 1970s to the mid-1980s

Over the period there was a decline in the importance of agriculture, albeit gently, and an unambiguous jump in manufacturing as a share of output and employment across Malaysia, Indonesia and Thailand.

The transformation of the export structure from one of agricultural raw materials, and ores and metal exports towards manufactured exports was rapid in Indonesia. The speed of the change in the shares of manufactures in exports was phenomenal.

Consumption inequality was in general very low by developing country standards in Indonesia in 1970, but relatively high in Thailand and Malaysia (the latter of which has data for income inequality). Indonesia and Malaysia’s Ginis, by national estimates, rose very slightly through the 1970s and then fell slightly in the 1980s (Thailand only has a Gini estimate for the end of the period).

Poverty rates at $2.50 per day (2011 PPP) – the average poverty line across all developing countries estimated by Hoy and Sumner – were already below one in ten people by the early 1980s in Malaysia. However, they were higher in Thailand at about 1 in 3 of the population and much higher in Indonesia where almost 9 in 10 people lived below $2.50 per day. At more reasonable poverty lines of $5 per day – the average poverty line for all countries – or $10 per day – a poverty line developed by López-Calva and Ortiz-Juarez (2014) and associated with a low probability of falling back into poverty – even Malaysia and Thailand had notable poverty counts. However, $2.50 poverty was falling in the early 1980s in each of Malaysia, Indonesia and Thailand.

The mid-1980s to mid-1990s

The period of the mid-1980s to the mid-1990s was one of spectacular transformation though ultimately unsustainable, in that it was largely based on exchange rate movements and interest rate differentials leading to the consequential massive relocation of foreign direct investment and finance capital. There were dramatic increases in gross domestic product (GDP) shares to manufacturing in value-added and equally striking increases in shares of employment too. Not surprisingly, value added and employment shares diminished in agriculture accordingly. Employment shares in services increased notably in Thailand and Indonesia in the latter part of the period.

There were very large and rapid increases in the share of manufactures in exports. In Indonesia, the manufacturing share amounted to more than half of exports, and in Thailand and Malaysia this reached close to three-quarters of exports. The proportion of high-tech in those manufactured exports also increased substantially, though from a low base in Indonesia. In Thailand, a quarter of the manufacturing exports were high-tech exports. In Malaysia, almost one half of the manufacturing exports were high tech exports. Agriculture, food, fuel, ore and metal exports all fell as shares consequentially. Import shares shifted too. The share of manufactures in imports rose to around three quarters or more of imports, suggesting that the three countries were importing and assembling manufactured goods for export.

Overall, the Gini was steady in Indonesia until the mid-1990s when it rose slightly. Poverty rates at the lower poverty line of $2.50 per day (2011 PPP) fell across Malaysia, Indonesia and Thailand. The falls in Indonesia and Thailand were dramatic. Rates at the $10 poverty line were much more ‘sticky’ through the late 1980s, though they fell in the early 1990s.

The late 1990s to the present

The period since the late 1990s has been one where value added in agriculture declined slightly across the three countries and employment shares in agriculture continued to fall, rapidly so in Indonesia and Thailand. Value-added shares in manufacturing fell in Indonesia and Malaysia whilst plateauing in Thailand. Employment shares in manufacturing were stagnant and even fell substantially in Malaysia, suggesting that this was a weak period for the manufacturing sector. In contrast, employment and value-added shares rose in services across the three countries, with the exception of a slight decline in value-added shares in services in Thailand.

The composition of exports shifted substantially though in differing directions. Shares in manufacturing exports fell notably, with large declines in Indonesia and Malaysia though less so in Thailand. Furthermore, the share of high tech in those manufacturing exports fell substantially. At the same time, fuel exports and exports of ores and metals rose, reflecting the commodity boom of the 2000s.

Growth was more inclusive in Malaysia and Thailand than in Indonesia in the sense that the Gini fell in Malaysia and Thailand, whereas it rose in Indonesia. Poverty rates at $2.50 per day fell across the countries to the point that it is plausible to say that absolute poverty at $2.50 per day has been eradicated in Malaysia and Thailand. In Indonesia, absolute poverty still affects one in four of the population, but this fell from two-thirds of the population in the late 1990s.

As noted above, at first glance, the structural transformation and inclusive growth in Malaysia, Indonesia and Thailand looks dramatic and impressive. However, at other levels, it has not been as ‘deep’ as it may first seem, and structural barriers remain on the journey from low- to high-income country status.

It is true to say that labour, GDP, and exports transited out of low productivity agriculture and absolute poverty fell dramatically. Despite this, deeper questions remain, given that production and exports in the high value-added sectors of manufacturing are dominated by foreign investments, and thus upgrading to higher value-added points in global value chains remains challenging. Further, although absolute poverty fell to very low levels, poverty at higher lines, associated with a permanent escape from absolute poverty such as $10 per day, remains substantial and the employment intensity of growth has weakened, adding to social pressures. Furthermore, a dualism remains in employment to a considerable extent in Indonesia and Thailand. A third of the labour force in both Indonesia and Thailand remains in agriculture and over a tenth in even higher income per capita Malaysia. Informal employment in non-agriculture also remains large – three-quarters of non-agriculture employment in Indonesia, more than forty per cent in Thailand, and a tenth of non-agriculture employment is accounted for by the informal sector in Malaysia.

What would W. Arthur Lewis have said if he were alive today?

We can turn to pioneering development economist W. Arthur Lewis to assess this recent economic history. What might he have said? Lewis might have been surprised by the persistence of dualism. He might have argued that surplus labour remains in the subsistence sectors and the turning point has yet to be reached. Lewis would also have likely noted that underlying the empirical experiences of developing countries is heterogeneity in the various forms of structural transformation. He might have been surprised that the very notion of industrialisation has shifted from infant industry-building to global value chain access via foreign direct investment or domestic suppliers, and he might highlight the consequences of international ownership or control. He might have noted how the assumption of labour transfer from traditional or rural sectors to modern or urban sectors was just one route for structural transformation. In fact, it is one of ten potentially progressive modes of sectoral structural transformation (each of which can reverse, making twenty modes of structural transformation), one or more of which may be the dominant driver of productivity gains, and each with specific distribution dynamics of the labour transition.

He would have probably been surprised that developing countries could go from agrarian/rural in the 1970s to industrial in the 1990s and begin to deindustrialise in the 2000s, all in the space of thirty to forty or so years. One substantial surprise to Lewis would be the emergence of premature deindustrialisation, though he did argue that dependency on a handful of agricultural or primary exports, whose value fell relative to what is needed to be imported, could evolve into a new dependency. This would be where a handful of manufacturing exports, which are falling in value relative to what countries need to import, take the place of primary exports: in short, the replacement of primary commodity dependency related to a small number of commodities with manufacturing dependency based on a small number of manufactures and a difficulty in upgrading.

The idea that the process of structural transformation could go into reverse at a lower level of development than previously observed would have no doubt raised an eyebrow. That said, Malaysia, Indonesia and Thailand did achieve structural transformation in various senses through a determined set of state-directed interventions and public policy to counterbalance the forces of rising inequality that Lewis, Kuznets and others identified structural transformation could unleash.

How did South East Asia do that? Next week, we pose the question: how did South East Asia deal with those distributional tensions that rapid change often entails?

Andy Sumner is a Reader in International Development in the Department of International Development, King’s College London. He is Director of the ESRC Global Poverty & Inequality Dynamics (GPID) Research Network.

Related reading

Briefing paper 12 Andy Sumner Is the Lewis Model of Economic Development still Relevant to Developing Countries? 20/04/2018
Briefing paper 13 Andy Sumner Trickle Along not Trickle Down? Inequality and the Lewis Model of Economic Development 20/04/2018
Country Note 1 Arief Yusuf and Andy Sumner Poverty, Inequality, and Structural Change in Indonesia 29/09/2017
Country Note 2 Kunal Sen Poverty, Inequality, and Structural Change in India 29/09/2017
Country Note 3 Sergio Firpo and Renan Pieri Structural Change and Wage Inequality: the Brazilian Case 29/09/2017
Country Note 4 Morné Oosthuizen Poverty and Inequality in South Africa 29/09/2017
Briefing paper 4 Cinar Baymul and Kunal Sen What Do We Know about the Relationship between Structural Transformation, Inequality and Poverty? 29/09/2017