The new international geography of deindustrialisation
The problem with Donald Trump’s trade narrative is that it ignores the simple fact that deindustrialisation in Latin America and sub-Saharan Africa has been more severe than it has been in the US and Europe.
Trump seems intent on starting a global trade war. His recent imposition of tariffs on steel and aluminium is consistent with his rambling inaugural address about “American carnage.” Trump’s narrative of industrial decline is informed by the common sense notion that developing countries have experienced economic growth at the expense of American jobs and industry. This narrative has been deeply rooted in American politics, popular media, and scholarship since the late-1970s. Furthermore, it fuels populist politics in European countries that have also witnessed declining manufacturing employment such as the UK, France, and Italy.
Deindustrialisation in developing countries
Deindustrialisation in developing countries differs in important ways from the pattern observed in the OECD. Dani Rodrik refers to it as “premature” because it takes place prior to an increase in productivity in the service sector and/or wage increases. Rather than an endogenous evolution of the economy from manufacturing to services, premature deindustrialisation represents stalled development. It is largely driven by external shocks and events that local policy makers are ill-equipped to address. Structural adjustment policies imposed by international financial institutions forced developing countries to liberalise their markets, and manufacturers in Latin America and sub-Saharan Africa have struggled to compete in a liberalised global marketplace.
Why has this happened?
To understand why, it is necessary to come to terms with the geography of offshoring and the establishment of durable global value chains in the 1990s. Although the OECD offshored a significant amount of manufacturing from 1990-2010, it was concentrated in a handful of countries and “only six developing nations…saw their share of world manufacturing rise by more than three-tenths of one percentage point since 1990.” Multinational corporations relocated production facilities to these developing countries due to the availability of cheap labour and their business-friendly regulatory regimes. The result has been the formation of global value chains that combine high-tech production techniques and know-how from the OECD, with low-wage labour in certain developing countries. This high-tech/low-wage combination is more competitive than high-tech/high-wage production in the OECD, as well as low-tech/low-wage production in developing countries.
The situation in the US
Economists have taken note of the impacts of the rapid shift in the global geography of production, and most notably, the dramatic expansion of China’s industrial output. In the US the impact of Chinese imports has unfolded unevenly, but regions whose industrial specialisation is in direct competition with Chinese imports have unsurprisingly registered increased unemployment. David Autor and colleagues refer to this as the “China shock,” yet their research is limited to the US where production tends to be of the high-tech/high-wage variety.
The situation in Latin American and sub-Saharan Africa
Deindustrialisation has been even more severe in Latin America and sub-Saharan Africa whose domestic industries are no match for the global value chains that combined OECD technology and know-how with cheap labour in East Asia. Trade data allows us to infer the impacts of East Asia’s rapid incorporation into global value chains on these regions. A 2015 report from UNCTAD shows that trade among developing countries increased ~15% annually from 1990-2010, and by 2010 more than 80% of trade between developing countries originated in East Asia.
Although manufacturing remains an important avenue of development, manufacturers in many countries cannot integrate with, or compete against, global value chains. And unlike the pattern observed in the US and Europe, deindustrialisation in developing countries has not been accompanied by growth in the service sector.
The new international geography of deindustrialisation has implications for politics in the OECD and for developing countries. First, deindustrialisation in the OECD is associated with urban and regional decline, and it conjures up images of cities like Detroit, Sheffield, and Charleroi. The backlash against globalisation is evident in these “places that don’t matter,” but the rhetoric employed by right-wing populists in the OECD is undermined by the recognition that deindustrialisation is more severe in many developing countries.
Second, if voters in these regions respond to deindustrialisation like their OECD counterparts then the 21st century could witness a dramatic political shift to the right. To take one example, Brazil has experienced significant deindustrialisation as its economy has undergone a process of “primarisation.” Its trade deficit with China has grown dramatically in the 21st century, and importantly, its exports to China are denominated by an increasingly large share of primary products. This is significant because cities with a significant amount of manufacturing – or “production cities” – tend to exhibit higher levels of social welfare and less inequality than cities whose economy is based on resource extraction. Thus, in addition to increased unemployment, deindustrialization and the primarisation of the Brazilian economy may presage a crisis of social welfare in its cities. Indeed, the recent ouster of democratically elected Dilma Rousseff signals an elite backlash, which may culminate in the election of Jair Bolsonaro who dreams of re-establishing a version of Brazil’s military dictatorship.
In order to halt the spread of right-wing populism in the medium-term, social welfare programmes must be strengthened in localised labour markets that have experienced significant declines in manufacturing employment. The task is much more difficult in the long-term, because countries will likely have to establish development models that deliver equitable growth without a significant increase in manufacturing output.
 Richard Baldwin refers to these as the “industrialising six,” and they are China, India, South Korea, Indonesia, Thailand, and Poland.