Last week I finished some lockdown reading, following my perennial interest in industrial policy and how it impacts development. This time it was the edited 2019 volume How Nations Learn, subtitled Technological Learning, Industrial Policy and Catch-Up.
HNL explores industrial policy in the context of ‘learning’ by governments and firms in ways which can accelerate industrial growth and development. That is, learning by governments in the process of policy making for development and learning by firms in the form of using and adapting already existing technologies to drive productivity growth in the catch-up process.
It is fair to say that most of the chapter authors hold to the idea that industrialisation and the growth of the manufacturing sector in late (relative to today’s advanced countries) developers is a primary driver of learning and of development. In this vein development is seen as more than periods of growth, but as an economic transformation. It is this continuous transformation in economic structure which makes long term growth and broad increases in living standards possible.
The book covers general perspectives on industrial policy and development in today’s global economy, before moving on to examine what are considered industrial and developmental success stories, including Meiji era Japan, post-war Taiwan, South Korea and Singapore and China since the late 1970s, before finishing with what are referred to as ‘late-late’ developers, or countries that have been less successful, in the case of Latin America and parts of Africa, specifically Ethiopia, although the latter has been successful in certain sectors in recent years.
Overall the authors adopt analytical perspectives which focus mainly on technocratic policy successes and failures, and although some chapters draw attention to the importance of historical and political factors, it does not utilize the approach and categories of political economy. This is apart from a focus on the role of the state, which is seen as central to driving learning, both within itself and in firms and sectors, in order to take advantage of developmental ‘backwardness’, which provides an opportunity to catch up with richer countries. Aspects which shape particular contexts, such as power, conflict and class or social groups more generally, are ignored, and politics is given insufficient attention.
Learning by a developmental state and emerging firms is a key emphasis throughout. Also central, to learning in particular, is ‘absorptive capacity’, which makes learning effective, and which requires prior knowledge and experience, setting up a bit of a chicken and egg. There needs to be a passion for learning for development and catch-up on the part of leaders and followers in society and the economy, and an ability to experiment and learn from both policy success and failure. Foreign governments, institutions and firms are a key resource to learn from, again in cases of both success and failure. There is no simple set formula for success, since contexts change and vary over time and space.
This is all very well, and the book makes a very useful and insightful contribution, but for me it needs to go further in differentiating and investigating success and failure in industrial and technology policies. While the authors must be right that historical and political context matter, these need to be examined and integrated with the economic, in the spirit of political economy.
A different kind of economics
Political economy was the name once given to what is now called economics, reflecting a narrowing of vision and method over the years. The classical political economy of Smith, Ricardo and Mill was critiqued by Marx, who incorporated theory, history and empirical analysis into his grand scheme of thought which culminated in his magnum opus Capital.
Marx looked to socialism to replace what he saw as an unjust and dysfunctional capitalism. This concerned many in society who could see the threat to their position from a potentially revolutionary working class, and motivated the abandonment of such controversial concepts as class, conflict and power as political economy was gradually replaced by a more formalistic, ahistorical, asocial, mathematical economics, stressing the harmony of equilibrium and perfect competition, driven by rational maximising individuals. In terms of influence, the social science of political economy has been replaced by economics, the science of rational choice.
But political economy lives on, in marginalised enclaves of heterodoxy, usually but not exclusively in the form of leftist ideas which give more conscious prominence to the role of state intervention and progressive movements more generally in socioeconomic development.
One economist who has kept the flame of political economy burning, and who has written extensively on industrial policy and its role in growth and development is Mushtaq Khan of SOAS.
Khan has written extensively on the role and importance of the dynamic learning processes referred to above, but he goes further and deeper in analysing the causes of both developmental success and failure. He notes that until recently the accumulation of tacit knowledge via learning-by-doing in the form of organisational and technological capabilities has been neglected by development economists. This is so even among heterodox thinkers, who have tended to focus on the accumulation of capital or education and skills, the latter also being a theme in mainstream economics.
Khan argues that, typically, firms in late developers will not be competitive in producing very much, if anything, even with their low wage costs, due to the lack of tacit knowledge in production, which is needed to ensure sufficient levels of productivity and product quality at the global market price. A period of learning in order to accumulate new capabilities and catch-up with the industry leaders is required. The accumulation of tacit knowledge is achieved in the workplace through actually learning how to combine and use new technologies, skills and organisational routines in the production process, and is different from formal learning in the classroom. Given the high level of risk involved in this process, finance from what is typically a weakly developed private sector may often not be forthcoming, so the state will need to step in.
Industrial policies need to provide incentives to ensure sufficient effort on the part of managers and workers in the learning process to limit the catch-up period and give emerging infant firms the chance to become competitive, so that they can sell on global markets and compete effectively with other global producers. Policies should enable them to increase productivity more rapidly than they would be able to in the absence of intervention, in order for the relevant firms to become globally competitive in terms of both cost and quality.
Khan argues that what he calls ‘learning rents’ are important in the catch-up process. These are the excess incomes that can be generated by policy and accrue to emerging firms which both provide the incentive to learn and the resources to invest in expanding production.
The mainstream neoclassical theory of rents and rent-seeking is typically associated with negative outcomes and a waste of resources. It holds to an ideal of the presence of zero rents in a general equilibrium with perfect competition as a (unrealistic) benchmark.
The idea of learning rents is part of a broader set of definitions of rents, which can have positive effects on economic growth and development in particular contexts. Even if learning rents encourage forms of rent-seeking such as lobbying or corrupt activities, the net overall economic effect can still be positive.
Governance capabilities on the part of the state have also typically been neglected as a determinant of incentives for the necessary high levels of effort in catch-up learning processes in late developers, even in the literature typical of HNL.
What Khan calls the political settlement, defined as the distribution of power between organisations in a society, also plays a vital role in determining the outcomes of policy and institutional change.
The successful East Asian development case studies, such as Japan, Taiwan and South Korea, were perhaps unique in that their states were relatively autonomous from relevant parts of the private sector in the early stages of catch-up and were able to intervene so as to credibly manage learning rents which were conditional on improved firm performance. If the relevant firms did not catch up over a particular period, support was often withdrawn and reallocated. This reduced the likelihood of large social welfare losses and induced a high level of effort for learning in particular industrial sectors.
During the postwar period, countries in South Asia, Latin America and Sub-Saharan Africa also attempted industrial policies in the form of support for infant industries, but with less success. Growth spurts were ultimately followed by stagnation and crisis, partly because states were less able to make policy support conditional on improved performance. Picking industrial winners became sustaining losers in many cases, even with the best of intentions on the part of policymakers.
It is this contrast between success and failure in different contexts that needs explaining in a way that goes beyond appeals to good and bad policies.
Khan argues that the degree of effort to engage in learning processes in emerging firms is determined by four factors: financing instruments, such as loans or subsidies; governance agencies, such as development banks or fiscal agencies; firm structure, including size, initial capabilities and political connections; and the political settlement. These are interdependent and interact in a non-linear way, making any analysis potentially complex.
These key variables influencing effort levels for learning require a careful analysis of context before embarking on policy changes. This will typically involve acquiring knowledge of historical, political, social and economic factors in a particular country case study.
The political settlement will be the hardest factor to change and the most risky as well, as it could result in highly unpredictable and unstable outcomes in any society.
The financing instruments and governance agencies are typically the more immediate policy variables and the easiest to change, so if possible policymakers should address these first.
To take a theoretical example, if large firms in a particular country have established political connections and the state is less able to discipline them according to performance and limit rent-seeking as it provides financial support, it may be better for industrial policy to focus support on smaller, less well connected firms, which may then lead to better policy outcomes.
In sum, more attention needs to be paid to factors such as power and organisation in the process of learning for developmental transformations. Industrial and technology policies are experiencing something of a renaissance when it comes to growth and development theory and practice, not least in late developers, but also in advanced countries. But perhaps those who have the most power are often less than willing to publicly acknowledge and share it with others. Even considering the concepts of power, conflict and class can make many in the mainstream uncomfortable. This makes the assimilation of more radical ideas into mainstream thinking difficult to achieve.