• Robots are rapidly becoming a key part of manufacturing in developed and developing economies
  • Two economists find that robot diffusion can increase the quality of manufactured exports.
  • Sophisticated or more basic robot applications are associated with quality gains in developed and developing economies, respectively.

An increasing number of media articles highlight how rich countries are rapidly automating and that developing countries may lose out – endangering the future of manufacturing-led development. Much of the literature is drawn from robot adoption in rich countries and focused on the labour market impacts (Acemoglu and Restrepo 2017, Adachi et al. 2021, Deng et al. 2021). We know surprisingly little about robot diffusion in developing countries and the impact on trade. Recent evidence on automation and reshoring is at best mixed and finds that automation in rich countries may actually increase imports sourced from poorer countries (Artuc et al. 2018, 2019, Stapleton and Webb 2020).

In this column we explore a new channel – whether automation leads to export quality upgrading – in both developing and developed economies. Robots perform tasks repeatedly to the same high level of accuracy, so are commonly used in the assembly of small electronic components, precision welding of car parts, or cutting of metals. Some types of advanced robots are able to operate within extremely accurate tolerances – for instance, those with lasers can cut to within 10 micrometres (0.01 millimetres) – or are equipped with sensors that allow the machines themselves to identify product defects. Robots may therefore reduce human error in production, increasing product quality.

To assess this, in a new paper (DeStefano and Timmis 2021), we combine data on robot diffusion and export quality that span a broad sample of countries (28 developed and 31 developing economies) over the period from 2000 to 2015. Export quality is derived from detailed HS 10-digit US import data (following Khandelwal et al. 2013) and combined with data on robot use at the country-industry level. The extensive country coverage in our data allows us to distinguish effects in developed and developing economies, and the richness of the trade data enables us to examine which types of products are being upgraded within each country.

AI, machine learning, technology

How is the World Economic Forum ensuring that artificial intelligence is developed to benefit all stakeholders?

Artificial intelligence (AI) is impacting all aspects of society — homes, businesses, schools and even public spaces. But as the technology rapidly advances, multistakeholder collaboration is required to optimize accountability, transparency, privacy and impartiality.

The World Economic Forum's Platform for Shaping the Future of Technology Governance: Artificial Intelligence and Machine Learning is bringing together diverse perspectives to drive innovation and create trust.

  • One area of work that is well-positioned to take advantage of AI is Human Resources — including hiring, retaining talent, training, benefits and employee satisfaction. The Forum has created a toolkit Human-Centred Artificial Intelligence for Human Resources to promote positive and ethical human-centred use of AI for organizations, workers and society.
  • Children and young people today grow up in an increasingly digital age in which technology pervades every aspect of their lives. From robotic toys and social media to the classroom and home, AI is part of life. By developing AI standards for children, the Forum is working with a range of stakeholders to create actionable guidelines to educate, empower and protect children and youth in the age of AI.
  • The potential dangers of AI could also impact wider society. To mitigate the risks, the Forum is bringing together over 100 companies, governments, civil society organizations and academic institutions in the Global AI Action Alliance to accelerate the adoption of responsible AI in the global public interest.
  • AI is one of the most important technologies for business. To ensure C-suite executives understand its possibilities and risks, the Forum created the Empowering AI Leadership: AI C-Suite Toolkit, which provides practical tools to help them comprehend AI’s impact on their roles and make informed decisions on AI strategy, projects and implementations.
  • Shaping the way AI is integrated into procurement processes in the public sector will help define best practice which can be applied throughout the private sector. The Forum has created a set of recommendations designed to encourage wide adoption, which will evolve with insights from a range of trials.
  • The Centre for the Fourth Industrial Revolution Rwanda worked with the Ministry of Information, Communication Technology and Innovation to promote the adoption of new technologies in the country, driving innovation on data policy and AI – particularly in healthcare.

Contact us for more information on how to get involved.

Robot diffusion in global manufacturing

We find that robots are becoming an integral component of global manufacturing – in developing countries too. Today there are roughly 2.7 million robots in operation around the world, with the number of new robots installed each year more than tripling over the last decade (IFR 2020). While most of the evidence so far has focused on automation in rich countries, the increasing automation of manufacturing tasks is occurring not only in developed economies, but also in developing economies (see Figure 1). Robot use in developing countries has increased nearly 10-fold over 2000-2015. While developed economies remain more intensive users of robotics, developing countries are catching up.

Robot diffusion in the top five developed and developing economies
Robot diffusion has increased in the past decade in many countries.
Image: VoxEU

Robot diffusion among foreign export customers strongly predicts automation adoption at home. To address endogeneity concerns, we instrument robot adoption of the home country-industry using robot diffusion among their foreign customers located in other world regions, using their initial foreign trade network defined in 2000. Our motivation for this instrument comes from the extensive evidence of technology spillovers from foreign multinational customers to their overseas suppliers (Javorcik 2004). We find that automation cascades from foreign customers to their suppliers.

Robot adoption and export quality upgrading

We find that robot diffusion leads to increases in the quality of exported products. Across all countries, a 10% increase in robot stock results in a 1.2% increase in quality (see Figure 2). The strongest quality gains accrue to developing economies, where a 10% increase in robot stock leads to a 2.7% gain in quality. While quality gains are still achieved through robot use in developed economies, the size of the effect is considerably less (roughly a 0.4% gain in quality from a 10% increase in robot adoption). One reason why we find a smaller effect of robots on quality in developed economies over the entire sample period is that more advanced countries were early adopters, and thus quality gains were realised early on.

We also find some evidence of differences in quality upgrading by type of robots. Adoption of more basic robots leads to quality upgrading in developing economies, whereas some sophisticated robot applications lead to stronger quality upgrading in advanced economies. For instance, in terms of cutting machines, we find mechanical cutting robots matter more for quality upgrading in emerging economies and laser cutting robots more for high-income economies.

Effect of robot adoption on export quality: All countries, developing and developed
Developing countries have the most effect of robot adoption on export quality.
Image: VoxEU

Differential gains depending on initial quality

Within each country, robots lead to quality upgrading of initially low-quality exports, furthest from the quality frontier. This is true of both advanced and developing economies. For example, products that are initially 10% further from the quality frontier achieve a 1% faster increase in quality from a given growth in robot stock in developed economies, and a 1.3% faster increase in quality in developing countries. Since developing economies in general tend to produce lower-quality exports, this reconciles the stronger aggregate quality upgrading we observe for developing countries in Figure 2.

The fact that robots matter more for initially lower-quality goods in both developed and developing countries suggests that robots are important for economic development. Quality improvements in production and exports help determine economic development (Hidalgo et al. 2007, Fontagne et al. 2008). Robot adoption appears to enable catch-up in export quality, given that the gains are stronger for initially poorer quality exports. For developing countries, robot diffusion may therefore help them to overcome quality standard that firms face looking to participate in and climb the ladders of global value chains.


We find that robots lead to export quality upgrading, but the gains are not spread evenly. Much of the quality upgrading is obtained by developing countries. Within countries, we find that robots lead to upgrading for initially low-quality products – a result found for both developed and developing countries. Since developing countries tend to export lower-quality goods, the gains from robots are larger for these types of countries. Developing countries therefore have greater potential for quality catch-up through automating their production.

Our results have several important implications for policy. First, technology always creates winners and losers. While automation poses risks, it also presents new opportunities for developing countries through the ability to upgrade production and leverage the benefits of global value chains. Second, trade openness can help technologies like robots diffuse through supply chains and across borders. Increased protectionism may stymie the potential for international technology diffusion and constrain the ability of firms in developing countries to upgrade production processes, move into higher value-added activities, and produce the high-quality products increasingly demanded by consumers globally. Finally, there is no ‘one size fits all’ recipe for policy targeting technology adoption. Firms and countries adopt different types of production technologies depending upon their factor endowments and incentives to do so. Thus, encouraging only the most advanced technologies is unlikely to be appropriate, especially for developing economies.