Vietnamese domestic policy: economic strategy and the “shift to services”

Full Text

Introduction

This paper has two aims. First, I introduce recent important Vietnamese economic policy changes. These are to do with the desired pattern of structural change. I interpret the policy changes as shifting away from a focus upon ‘factories and industry’ and towards ‘services’. Second, I place these changes into two wider contexts: first, the twin facts that global averages in developing countries have shown both servicisation, and the faster the growth the more servicisation, and second that mainstream analysis and policy advice has continued with support for, and interest in, industrialisation; and second, some ideas for possible implications for the country in the future, and how its recent past might be explained.

The policy issue: the pattern of economic change

Both before and after đổi mới, the Vietnamese Communist Party’s slogan for the country’s development was ‘hiện đại hóa, công nghiệp hóa’ – modernisation and industrialisation. This was re-affirmed in the early 1990s with approval from donors such as the World Bank. Industrialisation had a clear meaning: the share of total economic activity (GDP) coming from agriculture, forestry, and fishing, dominant in the largely rural society of the period just after reunification in 1975–1976, should and would fall, and instead workers and capital would increasingly be concentrated in industry. This image matched the basic vision of Vietnamese society as being the alliance between farm and factory, the ‘liên minh công nông’. At the time, this was widely seen as being the natural path of economic change. It had normative implications as something that was progressive and should happen, and policies, public and private decisions were expected to bring it about.

Earlier Vietnamese Communist economic data based upon Soviet methodology measured what was called ‘Total Social Product’ and ‘(Produced) National Income’[Fforde & Paine 1987: Tables 19–24]. This methodology tended to somewhat ignore services but offered measures of the share of industry in Produced National Income, rising (for the Democratic Republic of Vietnam – North Vietnam) from 18.6% to 27.9% between 1960 and 1975 [Ibid.: Table 21). With this data also went stress upon industrial physical capacity, which was part of the ‘trope’ of ‘modernisation and industrialisation’ and was also related to capacity to produce military materiel, which at that time, and supporting Soviet and American victories against the Axis powers in WWII, relied in part upon capacity to out-produce Axis capacity to produce tanks, aircraft etc. [Ellis 1990]. VCP commitment to industrialisation as a defining economic change after 1975–1976 and after the end of the Cold War is therefore easy to understand. Replacement of Soviet statistical methodology by the National Income Accounting system developed in the later 1930s and early 1940s [Kuznets 1942], in some ways using very similar methods, allowed for this process of structural transformation to be measured by changes in the share of GDP going to industry: the share of rewards to labour and capital occurring in industry should, and it was expected would, rise with successful economic development. This would involve also satisfying the perceived need to create a domestic armaments industry (as in North Korea), and to supply the industrial goods needed by the population for consumption, and construction of buildings as part of the expected urbanization.

However, whilst the speed of economic change and development after the end of the Cold War and the associated reductions in poverty were very great, the country did not industrialise; rather, since GDP rose fast and the share of rewards to labour and capital in agriculture, forestry and fisheries fell, without the share of industry rising, this meant that, as the GDP data showed, growth was occurring in services sectors. To quote from an early study [Fforde 2016: 9]:

“Whilst some 15 million jobs were created between 2000 and 2014, manufacturing and mining only generated about 3.9 million jobs” and “… the service GDP share rose from 38 per cent in 1992 to 43 per cent in 2013. In addition, whilst the share of the broad category ‘industry’ over the same period rose from 23 per cent to 29 per cent, this growth was largely due to increased mining output. In 2013 mining – included in the industry statistical definition – was 12 per cent of GDP yet below 5 per cent in the early 1990s; this means that the non-mining ‘industry’ share of GDP fell from around 18 per cent in the early 1990s to around 17 per cent in 2013 (GSO 1993: Table 18 and GSO 2014: Tables 66 and 68). If one cares not to believe this data then the point to bear in mind is that it is the public face of quantified economic change [Fforde 2016: 12].

 The original trope in the slogan refers clearly to factories; increased Vietnamese production of crude oil was not seen as a sign of qualitative economic development.

This issue, of the distinction between measures of ‘industry,’ and measures of ‘manufacturing’ output signals a need for care. In another paper [Fforde 2021]. I look at the Vietnamese data and statistical issues in greater detail, but three main points are usefully stressed:

First, GDP measures, not actual ‘output’ but, like the earlier Soviet data, levels of economic activity defined as rewards to labour and capital (the latter had conceptually to be finessed in Soviet definitions): that is, to ‘value added.’ Calling this ‘output’ is misleading. It equals the difference, in a sector, between total sales and payments for inputs that are neither labour nor capital. For Kuznets, this was what was ‘real,’ not statistics that result from attempts to take out price changes to produce ‘volume’ estimates. These payments for labour and capital then result either in savings or demand for goods and services. In economies where most GDP is generated in services sectors, a higher share of this demand will be for services, by value, than for industrial goods – higher than when industrialisation, rather than servicisation, was experienced by countries with fast economic growth.

Second, globally, since the end of the Cold War the observable pattern of structural change in developing countries has been like that in Vietnam, on average [Fforde 2018] servicisation. These are averages, but these show that Vietnam, a faster growing-country, has servicised more, and the slower growing countries have been servicising less. This means that GDP/worker in services has tended to be higher than in other sectors, for that is simple arithmetic.

Vietnam started its period of fast growth with GDP/worker in manufacturing much higher than the national average, around twice as much [Fforde 2022, Table 3]. However, by 2018 it was about 10% below the national average. Workers moving into services sub-sectors on average were generating more GDP/worker than in manufacturing. Further, whilst it is true that in recent years FDI into manufacturing in Vietnam has been large, and the value of manufactures exports has grown fast, the economic activity involved has generated little incomes for Vietnamese workers and capital. The data shows this. Total GVA in the sector for 2018 (using 25,000 VN Dong to the USD as an exchange rate) was around USD 35 bn. This compares with total manufacturing exports in that year of USD 208 bn on an SITC basis [NGTK 2018: 612)¹, or USD 227 bn using the GSO’s definition [Ibid.: 611]. Clearly, these export values included rather little Vietnamese value-added (below 15%, as some of the sector’s output was domestically consumed). This shows the usefulness of the GDP data in showing the relative value (in terms of generation of wages and rewards to capital) of economic activities.

Third, detailed examination of the services sector shows that it is unlike industry in a key aspect: there is no sub-sector that can play the role conceptually of ‘manufacturing’ as a leading, superior sub-sector that others must ‘serve’. Two sub-sectors (financial services and real estate) report very high levels of GDP/worker, but they generate low levels of employment and the ratios of their GDP/worker to the national average vary greatly year-to-year, suggesting (as do other considerations) that they are areas of rent-seeking and rent-acquisition. Services, then, as part of a servicisation strategy, appears conceptually as requiring a focus upon current values of rewards to labour and capital (what Kuznets considered the ‘real’ data) and requires a more holistic approach as there is no ‘super sub-sector’ to match manufacturing in its claims to be a growth driver: there is no obvious hierarchy of sub-sectors. Rather than being, as many seem to think, a place where there is a plethora of badly paid petty producers on the margins of the modern, formal economy, services sub-sectors are areas of relatively well-paid workers and owners of capital.

This, perhaps oddly, fits with a linguistic issue. Why do Vietnamese statistics abstain from refering to services sectors with the common term ‘phục vụ’ and instead use the term ‘dịch vụ’? In my reading, the latter term avoids the sense if inferior subordinate status that might be associated with the former. I have asked Vietnamese colleagues but remain intrigued. My dictionaries translate both as ‘to serve’.

The policy issue: orthodoxy talks about ‘premature deindustrialization’

The previous section would suggest to any researcher that there should be an established literature on servicisation. But this is not the case. The term appears not at all in the World Bank’s World Development Review, and there are a total of 36 mentions in Abstracts of articles in the EBCSO database (https://www.ebsco.com/products/research-databases#) from 1961 to August 2021, compared with 10,977 references to industrialisation.

In the famous Economic Miracle study of the World Bank in the early 1990s [World Bank 1993], the recipients of praise were NIEs – Newly Industrialising Countries (such as Malaysia and Thailand – p.1). World Bank advice to Vietnam followed, obediently, this analysis [World Bank 1995, see also Fforde 2016 and 2018].

Such views were echoed by a wide range of experts, such as Rodrik, who deployed the telling phrase ‘premature de-industrialisation’ [Rodrik 2015].

The gathering support from literature: heterodoxy emerges

The shared doctrine of the necessity and value of industrialisation as a development strategy has never been unchallenged. For Vietnam, an early report [UNDP 2005] pushed for servicisation as a strategy, but this seems to have been simple shelved. Fforde 2016 was a paper for a conference on ’40 years of doi moi’, and the commissioner, a Vietnamese at ISEAS, requested consideration of the slogan of modernisation and industrialisation. Subsequently, Nguyen HC 2018 is a peer-reviewed academic paper looking at structural change, but then Nguyen Dinh Chuc & Ta Phuoc Duong 2019 overtly (but politely) challenge the slogan by titling their paper ‘the shift to services’. As the next section shows, Party public thinking had already shifted at the 2018 Congress. But was still deploying the slogan into areas such as education policy in 2017 [Công nghiệp hóa, hiện đại hóa: 19.08.2017].

Reasons of space preclude detailed examination of the non-Vietnamese literature. Fforde 2018 was loudly critical of the lack of interest in servicisation in contrast to the observed reality discussed above. Helbe & Shepherd 2018, an ADB report, talk positively about ‘leveraging services for development’, and contrast with Haraguchi 2017 who avoids the issue by focussing on the fact that servicisation as a % of total global developing country output has not increased, but this is largely due to the massive industrialization numbers in China. China has probably been servicising strongly since the start of the current decade, but Chinese economists report that the Chinese Communist Party does not have a thought-through conceptualisation of servicisation strategy [Li et al 2018]. Mention can also be made of various other studies that seem increasingly to grapple with the realities and nature of servicisation (for example Hoekman & te Velde 2017). This poses a wide range of theoretical problems, as discussed in Fforde 2021, such as the shortcomings of measurements of constant price sectoral GDP (see also Arrow 1974 and early work he cites): the available measures do not measure what standard economic theory deploying production functions needs, which is estimates of changes in the real (volumes) of factor inputs.

Against this fascinating background, we find interesting changes in Vietnamese strategy.

The new policies

The VCP, in 2018, produced a major change of emphasis. In a recent official exposition of this new direction. Nguyễn Quang Thuấn (in 2021 the author is cited as ‘Deputy Head of the (Party’s) Central Theoretical Council’), with reference to the documents of the 2018 13th Party Congress, states:² “Regarding the development of the services sector, the Documents of the 13th Congress stressed promoting development of the services sector in a modern direction, securing a growth rate faster than the productive sectors and faster than the growth of the national economy … The Documents emphasised “A strong development of services … especially those with high value-added …” [Nguyễn Quang Thuấn 2021: 4].

This was then concretised in the 2021 (State) policy [Premier 2021] which is a national development strategy that moves away from industrialisation, seeking to raise the share of services sectors in GDP.³ These changes in strategic emphasis, combined with the reality of servicisation during the Miracle, have generated an increasingly vivid discussion in Vietnam. Directive # 531, as I read it, considers that the main policy issue is to ensure that factor markets work well and so continue to seek out areas of relatively high returns to labour and capital; as this happens to be services, then such processes should be encouraged. I return to this in the next section.

It is important to realise how unusual and path-breaking the 2018–2021 Vietnamese policy shift is, in a global context. However, as services issues are researched and, in a new ‘moment’, become part of the cognitive aspects of economic growth, this may point to Vietnam’s emergence as one of the first fast-growing developing economies to openly accept servicisation and, through an understanding of it, improve growth quality. Since servicisation is now the global average – and has been since the end of the Cold War (see below) – this suggests that we may well be seeing a further development of Vietnam’s position as a source of useful ideas and practices for others.⁴ Research metrics still show a formidable bias towards industrialisation and lack of interest in servicisation [Fforde 2018]. Vietnamese contributions are striking in the speed with which they are increasing and developing.

A Google search on terms such as ‘national services strategy’ turns up rather little. Perhaps the Vietnamese are unique in this area – further research is needed. This might include examination of the situation of countries such as Thailand and Malaysian, praised as Newly Industrialising Countries in the early 1990s just after the end of the Cold War.

Conclusion

As already mentioned, ‘industrialisation’ has had high status as a hoped-for reality and a purpose. Much data suggests that this has proven to be troubled by facts, and by the persistence of a striking lack of academic and expert interest (though this may be changing).

It is possible that the experience, for Vietnam, may have intriguing implications.

First, does it matter than ‘the Party got it wrong’ when the experienced reality was one of rapid growth? On the one hand, this leads us to ask questions about the role in Vietnam of national development strategy as a source of hoped-for political support for the Party. Here perhaps the rather rapid emergence of a new strategy is indicative. Of what, more research may tell us. Loss of face does not seem to matter that much. As a leading retired economist remarked to me in 2019 (having long been, obediently, an advocate of industrialisation, when presented with the facts) – ‘dynamic flexibility is the thing’ (linh hoạt là chính). This, to my mind, reinforces Kuznets’ view that it is actual values of rewards to labour and capital, which generate savings and demand for goods and services, that are ‘real’, not attempts to use GDP data to talk about ‘actual output’. What matters is what is earned and what it can be spent on [Fforde 2021].

Second, it was not just public officials who bought into the industrialisation narrative, but also citizens, who had good reason to believe in it as the most prestigious donors (the World Bank) and experts (Rodrik, Ohno) were its advocates. Clearly one can expect much rethinking, and there is evidence that this is ongoing. Perhaps one basic lesson is that it is well-functioning markets for labour and capital, powerful enough to respond to economic fundamentals rather than slogans, that require thought. And of course, if previously officials could ‘get up’ public investment projects by saying that the massive investments required by manufacturing were profitable, and that environmental damage would be part of rapid growth that would create the resources to clean up later, this is now less likely. Again, this ‘turn’ is a very fruitful area for research.

Third, and more fundamentally, perhaps, and also intriguingly, ‘servicisation’, in that it is not well understood, but clearly positive in its creation of economic growth, pushes to reposition national leadership in ways that are less mechanically scientific, and more to do with creating suitable environments, where the population can then decide what to do, as the national labour market does.

Vietnam is rarely dull.

¹ This refers to the Vietnamese General Statistical Office’s Niên giám thống kê (Statistical Yearbook) available in downloadable form from their site, starting with NGTK 2005, and before then in hard copy.

² Author’s informal translation.

³ # 531 has 18 pages and was quickly echoed by related directives from provinces and cities, available for free on the massive website www.thuvienphapluat.vn . Like # 539, from central government, this pattern deserves detailed analysis, but I avoid this here for reasons for space. I am personally intrigued by the responses of cities like Đà nẵng.

⁴ It is striking how methods of economic analysis and practical experience with the commercialization of a planned economy that were worked out by the Vietnamese in the early 1980s seem to have been somewhat off the radar screens of officials working for the then major aid donor, the Soviet Union [Fforde & Mazyrin 2018]. See also [Prostiakov 1998].

About the authors

Adam Fforde

Author for correspondence.
Email: adam@aduki.com.au

Adjunct Professor, Victoria Institute of Strategic Economic Studies, Honorary Professorial Fellow

References