Trouble in the Making? The Future of Manufacturing-led Development
Trouble in the Making? The Future of Manufacturing-led Development
This timely book addresses the challenge of premature industrialisation laid down by Dani Rodrik, the idea that developing countries are reaching ‘peak manufacturing’ at ever lower levels of income; and also examines the future challenges associated with artificial intelligence, automation, and other aspects of the so-called 4th industrial revolution. Manufacturing has been the principal engine of development and poverty reduction. But will that continue? Maybe not:
‘Changing technologies and shifting globalisation patterns call the feasibility of manufacturing-led development strategies into question. Trade is slowing. Global Value Chains remain concentrated among a relatively small number of countries. The Internet of Things, advanced robotics, and 3-D printing are shifting the criteria that make locations attractive for production and are threatening significant disruptions in employment, particularly for low-skilled labour‘. (Pg 1).
It’s OK, though. Don’t panic. The answer to the over-arching question, we are told, is that manufacturing has potential, provided the right sectors are chosen and the right policies followed. The policy model is described as ‘3-Cs’: competitiveness, capabilities, and connectedeness. Some countries are there already. Many others are not, and need to act.
I’ll have quite a few things to say about the argument, including on inter-sectoral linkages, climate change, and the approach to policy. I end with some thoughts on progressive industrial policy. But it would be churlish not to begin by saying that the topic is well-chosen and resonates with current debates. My own shelves are groaning with works on how new technology will disrupt the world economy, and with others on how to manage structural transformation. References later, but see also Taming Cerberus for discussion of how I think technical change, climate change and globalisation turn out to present very similar problems which will not be solved by either neo-liberal or populist approaches. Mary Hallward-Driemeier and Gaurav Nayyar have done us all a service by laying out so clearly an argument we can examine.
What is the argument?
Part 1 reminds us why manufacturing has been so important to development, and reviews the changing manufacturing landscape. It provides an analysis of the pro-development character of different manufacturing sectors.
Part 2 deals with technology, globalisation and the future of manufacturing-led development. It reviews the changes associated with the 4th industrial revolution and assesses the impact on manufacturing. It also deals with the relationship between manufacturing and services, emphasing what it calls the ‘servicification’ of manufacturing, viz the importance of things like design and marketing to manufacturing success.
Part 3 presents the 3-Cs framework.
The key contribution of Part 1 lies in the identification of five variables that are indicative of the pro-development characteristics of different manufacturing sectors. These are:
- Scope to employ unskilled workers;
- Sector’s share of labour in the overall economy;
- Labour productivity;
- Tradedness; and
- Scope for innovation and diffusion.
Some sectors shine in this analysis, some do not (Figure 1): textiles, transport equipment and electronics at the top of the class, oil and most ‘commodity-based regional processing’, including food and beverages, closer to the bottom.
Part 1 also has twelve ‘stylised facts’ about changes in manufacturing (Figure 2). For example, ‘the manufacturing shares of both total value added and employment are peaking at lower levels and at lower levels of per capita income than in the past’; and ‘few lower-income countries outside of Asia have a revealed comparative advantage in anything but labour-intensive tradeables or commodity-based regional processing—although not all have passed even these thresholds’.
Part 2 offers a primer on new technology, and then turns to the implications. The key messages are that the bar is rising for establishing or maintaining competitiveness in global manufacturing, especially in sectors such as electronics, computers, optical instruments, pharmaceutical products and transportation (Pg 112ff). However, ‘doomsday scenarios about technological unemployment are overblown’ (Pg 140) and there remains potential, for example in commodity-based processing, in low–quality goods for regional trade, and where unit labour costs are especially low. Service sector contributions to manufacturing will also grow in importance.
Finally, Part 3 presents the 3-Cs framework. Figure 3 provides a summary.
Does the argument stack up ?
Now, of course, the issues at stake in this analysis are easily recognised. As I said, the shelves are groaning. On technology: Brynjofsson and McAfee, Ryan Avent, Tim Harford, Daniel Franklin, Yuval Noah Hariri, Klaus Schwab, and Herweijer et al among many others. Andrew Norton has usefully reviewed the development implications. On structural transformation: Baldwin, McMillan, Rodrik and Sepuldeva, Lin and Monga, Jacobs and Mazzucato and the new report from AGRA on the business of agriculture. ODI’s work on Supporting Economic Transformation is a key resource. And these are just the things I have been reading in the past few months.
So, how does Hallward-Driemeier/Nayyar stack up?
The steps in the argument
The steps in the argument work well; defining pro-development characteristics; working through stylised facts; identifying the key elements of change; thinking through the implications; and developing a policy framework. There are many useful pieces of original analysis, many visual presentations, some good case studies, and a coherent policy framework. The book has been edited well, for example with summaries of key messages. Personally, I would have preferred more cases and fewer calculations, but policy-makers as well as researchers will find this a useful volume.
However, and of course there are some howevers, some parts of the analysis work better than others.
The treatment of linkages
First, the treatment of linkages in the report is such that obvious synergies between sectors are missed. This may distort the identification of pro-development sectors. An obvious example of such linkages is between agriculture and food manufacturing, the former not forming part of the analysis and the latter dismissed as merely of regional interest.
For those of us brought up on Hirschman, Rosenstein-Rodan and Singer, the treatment of linkages, especially backward linkages, is a disappointing lacuna. In fact, linkages are mentioned only 16 times in the book, frequently in terms of the link between manufacturing and services. This is a methodological problem, caused by the way manufacturing is broken down into 16 separate and unrelated sub-sectors using a two-digit classification.
I wonder. Would it have been possible to combine agriculture and processing, in order to examine the pro-development potential of the entire agro-industrial complex? Even better, would it have been possible to be selective within agriculture, for example to look at higher-productivity enterprises? My hunch is that the ‘agro-industrial sector’ would have ranked more highly than food manufacturing on its own, certainly in terms of labour employed and possibly also some of the other pro-development standards.
One other change would have affected the ranking, which is that the book uses an odd metric for ‘scope for innovation and diffusion’, viz the ratio of R and D to value added in US data (Pg 14). This is not a measure of technology applied by firms, much of which will have come from outside, or will be embedded in intermediate products. Really, what is needed is a measure of current and potential technical change. In agriculture and food, this is significant: popular topics in coverage of these sectors include the use of robots in farming and distribution, IT-intensive distribution systems, and of course a large number of productivity-enhancing biological innovations, like genetic engineering. Here’s an example: lighting companies invest in combinations of differently-coloured LED lighting patterns to boost the growth or change the taste of herbs and salad crops grown under hydroponic systems. R and D in the lighting sector; technical change in the agro-food system.
Technical change of this kind might be a mixed blessing for the contribution of the system to long-term employment, if not long term growth. Food production, manufacturing and distribution in countries like the UK has become progressively more high-tech and less labour-intensive. It also does not follow, as many have pointed out, that having a comparative advantage in e.g. the production of cocoa automatically means that a country has a comparative advantage in producing chocolate – hence Switzerland as a major chocolate producer. But, surely, linkages are worth debate?
The treatment of climate change
A second beef is the treatment of climate change, which is mentioned in passing as a big topic not discussed. You would like to know how many mentions there are of climate change? I counted six, none of which is substantive. That is not good enough, because action on climate change is driving and will drive deep changes in the character of manufacturing: the size of markets, location of production, technology used, price and profitability.
A small part of this is related to adaptation, which will affect where production might take place (special economic zones on the coast might be less attractive, for example, if storm surges increase in size and number; or water scarcity or temperature might constrain factory development).
A larger part is related to mitigation, with investments in renewable energy, grid infrastructure, energy efficiency, transport, and city design all affecting the cost structure and employment potential of producers in different places. What, for example, would be the impact of a carbon price of $US 100/t on the profitability of airfreighted manufactured exports?
But the biggest part in the impact of climate change is what we in CDKN called transformation, meaning the economic geography of changes associated with climate change. China developed a solar industry originally to service demand in Europe, especially Germany. Bolivia sits on large reserves of lithium, needed for batteries, and could industrialise on the back of that resource. In terms of sectors and countries, never mind people, there will be winners and losers around the world on a scale which may well dwarf the impact of robots or 3-D printers. At least that issue should be addressed in a report examining the future of manufacturing.
Two more issues spin off from the climate change dimension.
First, what are the resource implications of continued, rapid growth in manufacturing, not just CO2, but all resources, including water? The UNEP Emissions Gap Report, the 2017 edition of which is just out, shows how much of a gap still has to be closed to meet the Paris targets on climate change. Any strategy for manufacturing-led development must necessarily put green growth at the heart of the analysis. For Hallward-Driemeier and Nayyar, there is other World Bank work available, including a policy paper on Green Growth, and other material available via the Green Growth Knowledge Platform. A key theme is correcting market failure, for example by pricing externalities (including by reducing subsidies to fossil fuels).
The speed of change
Second, climate change illustrates another issue not much discussed in the book, which is the speed of change. The price of solar power has fallen by over 90% in very few years, changing completely, for example, the economics of coal production. I searched the book for ‘hurry up’ – no hits. ‘Speed’ did better, with a couple of references to a new factory built by Adidas to accelerate shoe production, and one reference to the speed of light. But it needs to be appreciated that industrial revolutions happen really, really fast. They always have. The invention of a chemical purple dye destroyed the Indian indigo industry in just a few years. The container destroyed traditional docker communities with equal speed. I often say that all the clusters we studied at school in the UK in the 1950s and 1960s have now pretty well gone (carpets in Kidderminster, lace in Nottingham . . .). New technologies are diffusing even faster, as the well-known graph in Figure 4 illustrates.
How will the manufacturing sectors of developing countries handle creative disruption on a large scale and at rapid pace? Will it be enough to say, with Hausmann, that complex capabilities gained in one sector can easily be transferred to another? The answer to that question must surely be No. Ask the many skilled workers whose livelihoods disappeared and who voted for Trump or for Brexit.
The policy framework
This brings us to policy, and to the 3-C framework: competitiveness, capabilities and connectedness. I hope the authors will forgive me if I say that their framework smacks of the Washington Consensus: too narrow in its remit, without due attention to the difficulty of managing disruption and dislocation, and without sufficient attention either to state-led support for technical innovation, what Mariana Mazzucato calls The Entrepreneurial State.
To illustrate, compare the action items suggested in Figure 3 with those in a selection of industrial and related social policy frameworks in the Appendix: the KPMG Readiness to Change Index (invented at ODI, by the way, and originally called the Capability Index): the WEF Competitiveness Framework; also from the WEF, the Inclusive Growth and Development Framework; the UK Industrial Strategy; the approach to mitigating adjustment costs proposed jointly by the World Bank, the IMFInternational Monetary Fund and the WTO; and, last but not least, the approach to industrial policy set out by ODI’s programme, Supporting Economic Transformation, for example in Kenya. All these, in different ways, stress both the role of the state, and the importance of a social and political dimension to industrial policy. Agency matters too. I have been re-reading The Making of the English Working Class, about how this was handled the first time round: unions played a key part in improving the condition of workers. Corporate social responsibility and the accountability of the private sector also needs to be factored into the mix.
If the task is to find a path between neoliberalism and populism (see Taming Cerberus), then industrial policy is nested within the challenge of rethinking capitalism. Michael Jacobs and Mariana Mazzucato have brought together narratives about low growth, stagnant living standards, rising inequality, climate change and environmental risk, in their book ‘Rethinking Capitalism’. They argue that three key insights underpin the necessary new approach. First, ‘it is not helpful to think of markets as pre-existing, abstract institutions which economic actors ‘enter’ to do business; . . . they are better understood as the outcomes of interactions between economic actors and institutions, both public and private. Second, ‘investments in technological and organisational innovation, both public and private, . . . are the driving force behind economic growth and development’. And, third, it is important to recognise the ‘role of the public sector in the innovation process. . . The creation of economic value is a collective process . . . No business today can operate without the fundamental services provided by the state.’ These three insights, they argue,
‘have profound implications for how we think about economic policy-making. Public policies are not ‘interventions’ in the economy, as if markets existed independently of the public institutions and social and environmental conditions in which they are embedded, The role of policy is not one simply of ‘correcting the failures of otherwise free markets, It is rather to help create and shape markets . . .’.
Of course, it is necessary to be practical, and not to fall into the trap, familiar from earlier debates about governance, of insisting on long lists of reforms, all to be implemented simultaneously. Merilee Grindle pioneered the idea of ‘good enough governance’. Lin and Lin and Monga have adopted a similar approach with respect to economic governance: looking for ‘latent comparative advantage’ and adopting the approach of ‘ambitious pragmatism’, solving one problem at a time to boost industry. They call this the Growth Identification and Facilitation Framework, contrasting it with earlier and more comprehensive approaches to Growth Diagnostics, associated with Hausmann and Rodrik. Looking for latent comparative advantage sounds plausible – provided the scale and depth of future disruption are taken into account.
Where does this leave us?
The key take-aways of the book are really quite pessimistic (Pgs 213 ff). The context within which a strategy of manufacturing-led development can be assessed and implemented is changing rapidly, and not necessarily in a good way for low and middle income countries. Industrial policy (as in the 3-Cs) can make a difference, providing the basic ‘horizontal’ infrastructure (social and economic), as well as the more targeted ‘vertical’ sector-specific interventions – but policy makers should brace themselves for high costs, and should expect rising inequality. There may be some opportunities for leap-frogging (as in mobile phones rather than land lines, or distributed grids rather than centralised power distribution), and for new contributions to Global Value Chains, but, again, these are likely to require high skill levels and greater flexibility.
This leaves me thinking that the world needs a new, progressive narrative about industrial policy for the poorest countries – and, indeed, for richer ones like the UK, as for example in the book edited by Chuka Unnuma or IPPR’s Commission on Economic Justice.
Green growth as an opportunity. Urbanisation creating new, local markets. Regional integration. Exploitation of linkages between agriculture and non-agriculture. Active industrial policy, pragmatic but ambitious. Strong Government leadership, including for innovation. Political mobilisation. Ruthless targeting of vested interests and elimination of regulatory capture. High standards of accountability for corporate social responsibility. Social policy to underpin transition and protect losers. A relentless focus on implementation and results. And an emphasis on speed.
This is not yet a 10-point plan. But it could become one.
Frameworks for industrial policy
- KPMG Readiness to Change Index
- WEF Competitiveness Index
- WEF Inclusive Growth Framework
- The UK Industrial Strategy
- Mitigating Adjustment Costs (WB/IMF/WTO)
- A Plan to create 300,000 jobs and double manufacturing in five years in Kenya