Fighting an Upskill Battle? Public Policy and the Robot Reserve Army

Ticket machine at a Taiwanese High Speed Rail station

This week, we continue a special series on the future of economic development, work, and wages in developing countries.

Last week we considered the potential employment impacts of labour saving technology in the developing world. In this post, we take a look at public policy responses to automation.

A political pandora’s box

Automation, we have argued, has the potential for major shifts in employment and labour markets with wide-ranging economic consequences – but it may be prove to be equally politically disruptive.

Macroeconomic and labour market dynamics determine the quality, quantity, and distribution of citizens’ employment opportunities and thus of their wages, and living standards. Such socioeconomic factors in turn are known to have a profound bearing on sentiments of (in)security, relative deprivation, and societal equity which have repeatedly been shown to influence political preferences and political outcomes. There is a large body of literature providing evidence for a causal relationship of this sort (see e.g. studies on the impact of economic factors on voting, on welfare preferences or on political values).

The wider interest in the role of work and (un)employment as underpinnings of political agency goes back to early empirical social research, and even to the classical social theories of Karl Marx and Max Weber. As technological change influences labour market dynamics, an important field of research is the examination of ‘modernization losers’ as political catalysts: specifically, via so-called “technological anxiety” and resistance to innovation; the relationship of economic inequality, and political polarization and extremism; and the political implications of deindustrialization. Automation is opening a political pandora’s box.

Coping or containment?

Major political implications imply public policy responses. One can characterize policy responses to automation as follows: first, there is a class of policies that try to attenuate or reverse the automation trend (‘containment strategies’). Among those, there are “quasi-Luddite” measures such as taxes and regulation that make domestic automation more (or even prohibitively) costly. Countries could also follow a strategy of what one could call “robot-substituting industrialization” where they impose tariffs on inputs/imports with non-human-produced contents.

The problem with such strategies is that protectionism of labour is difficult to implement in an open economy. Luddite policies tend to be in conflict with integration into a globalized competitive market, as they assume that the economy can somehow be insulated from competition with automated production elsewhere. The mirror image of making automation costlier would be to reduce the costs of labour, e.g. by reducing income taxes or social insurance contributions, by reducing minimum wages, or costly labor regulations. The question is how economically desirable and politically feasible such strategies are.

Second, there is a class of “coping strategies” for the trend toward automation. The most prominent one is to develop the skills of the labor force and (re)train workers. A widespread policy recommendation is to invest in skills and thus move the labour force away from automatable routine tasks. The problem with this approach is that (i) it is by no means clear what skills will be automation-resistant for a sufficient time to make the skills investment worthwhile and (ii) whether upskilling is at all realistic given the pace of technological innovation. Even granting considerable lags in technology adoption, competition with currently available and commercialised technology increasingly seems to require a tertiary education which is still very rare throughout the developing world. Given that even advanced industrialized countries are struggling to keep their labour forces competitive, the success of a skills development strategy alone remains questionable.

A second coping strategy would be to provide economic transition support as well as safety nets, unemployment insurance, or wage subsidies. This approach addresses the distributional skew which automation may create. However, such transfers presuppose the existence of a productive sector in the first place, from which profits can be siphoned off for redistribution. In the absence of the existence of such a sector, there may be a case for the provision of international aid to support basic income guarantees or automation adjustment assistance overseas.

Towards postindustrialisation?

In many countries, one could say that the coping strategy adopted so far has been to invest in currently labor-intensive sectors such as infrastructure and construction. A—risky but potentially inevitable—long-term coping strategy for developing countries would be to anticipate automation trends and to try to (further) develop a productive post-industrial sector. If industrialization begins to look increasingly unattractive due to reshoring of hitherto outsourced production in value chains, countries would be well advised not to invest in the costly creation of manufacturing clusters but rather in the growth of a long-term automation-resistant (service) sector. Such a sector could, for example, involve the social, education and health-care sectors, and some forms of tourism, and infrastructure construction which are generally considered resilient despite increasing service automation.

The problem with such an approach is that highly productive and tradeable services are skills-intensive, and non-tradeable services (such as social care, personal services, etc.) are not (yet) highly value-adding, may not be sufficiently scalable, and may generally be too heterogenous to be targeted by post-industrial policies, in a similar way that industrial policies targeted the emergence of industrial clusters. Looking ahead, ‘late developers’ are thus are facing novel and complex policy challenges.

We’ll wrap up the series of blogs next week with some conclusions.

Lukas Schlogl is a Research Associate with the ESRC GPID Research Network at King’s College London. He works on structural change, digital transformation, and political behavior in developing countries.

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.