Written by Ryan McGuine //
The conventional path for economic development has been to use industrialization — the process of increasing industry’s relative contribution to total GDP and employment, at the expense of those of agriculture and services — as a means to enable rapid convergence. “Convergence” refers to a catch-up effect whereby countries that have low levels of income achieve higher rates of economic growth than countries with high levels of income. The Solow Model suggests that convergence should occur any time there is a difference in incomes between countries, since decreasing returns in high-income countries implies that a higher return to investments is possible in low-income countries. This should cause capital to flow from high- to low-income countries.
There is much empirical data on convergence, but the proper interpretation of it has long been contested. In 1992 Robert Barro and Xavier Sala-i-Martin found no evidence of unconditional convergence, but some evidence of convergence between countries with similar investment rates and worker skills; in 2013 Dani Rodrik suggested that convergence on the country level is the exception rather than the rule, but that unconditional convergence does occur within the manufacturing sector; and in 2016 Sutirtha Roy, Martin Kessler, and Arvind Subramanian claimed the previous 20-30 years have been a “golden era of convergence.” With any question based on large, cross-country analyses, the truth usually lies somewhere between extremes. Deitrich Vollrath points out that convergence is a reliable force, albeit a weak and slow-acting one, so it is probably the case that convergence is constantly nudging low-income economies forward, and can be accelerated by good policy.
Industrialization has played such an important role in the historical development experiences of many countries for a number of reasons. Compared to agriculture and services, industrial processes tend to be easier to adopt from one country to another (the products tend to be easier to trade between countries as well), tend to have a greater capacity to absorb unskilled workers than agriculture or services, and tend to have higher rates of productivity growth — the main driver of overall economic growth.
In addition to economic benefits, industrialization has tended to foster social structures that yield the spread of bourgeois liberal values. The economic growth generated by industrialization has driven the creation of corporate interest groups which advocate for infrastructure construction and investments in human capital by the public sector, and invest in further human capital for employees. This economic growth has also enabled the widespread accumulation of wealth that facilitates the development of a middle class of citizens with the time and incentive to participate in democracy according to class solidarity. Acknowledging industrialization’s historical importance, the UN chose “Industry, Innovation, and Infrastructure” as the 9th Sustainable Development Goal.
Over time, advanced economies have a natural tendency to “deindustrialize,” whereby the relative contribution of industry to output and employment begins to decline while the relative contributions of services increases. As a country reaches higher levels of income, it loses its comparative advantage in industry to countries that have large numbers of low-skilled workers, so the amount of industrial goods imported from abroad increases. Additionally, since industry has a relatively higher rate of productivity growth compared to agriculture and services, it displaces workers (particularly unskilled workers) more quickly than those sectors.
The trend of advanced economies experiencing structural shifts away from industry is well-documented. More recently, though, Mr Rodrik has brought attention to a pattern of “premature deindustrialization.” Premature deindustrialization refers to a pattern of countries experiencing shifts away from industry relative to other sectors in terms of output and employment, but at lower levels of personal income, as shown in the plots below.
Source: Dani Rodrik (2015), Premature Deindustrialization
Source: Dani Rodrik (2015), Premature Deindustrialization
On a global scale, there has been no net deindustrialization in terms of either output or employment, only spatial relocation — that is, industrial production and jobs move elsewhere. However, it remains true that the level of personal income at which deindustrialization begins to take place has become lower over time. The drivers of premature deindustrialization must be different than those of traditional deindustrialization, since the shifting comparative advantage driving deindustrialization in advanced economies should drive industrialization in developing economies. This has not been observed, though. The (much more than) million-dollar question, then, is whether industrialization via labor-intensive sectors remains a realistic pathway for developing countries to achieve rapid convergence. There are good reasons to think it is becoming more difficult.
One reason is that the automation of low-skill manufacturing jobs threatens the industrial sector’s ability to absorb workers from lower productivity sectors, making it less likely that an emerging economy’s workforce will be well-suited to the sector. Further, as developing countries that have long maintained relatively closed economies begin opening up to trade, they are faced with the major challenge of competing against countries that have become incredibly good at manufacturing. China and other East Asian countries have developed large, powerful industrial complexes that are able to flood the rest of the world with cheap consumer goods. Typically, wages tend to rise with productivity, but Asia seems to defy the historically-observed drag of deindustrialization.
Nonetheless, in the book Beating the Odds: Jump-Starting Developing Countries, Justin Yifu Lin and Celestin Monga contend that industrialization remains the best option for today’s developing countries. Despite the rise of mechanized production, they argue, demand for handmade labels will remain high, and some industries like garments and footwear will remain labor-intensive by nature. Despite Asia’s impressive industrial complexes, they argue, wages will always be lower where productivity is lower given time (remember: America’s employment share of manufacturing increased from the 1870s and peaked in the 1940s, whereas the process for China only just began in the 1970s). In short, today’s emerging economies can rest assured that they will be able to find and exploit latent comparative advantages in industry.
Here again, the truth is most likely somewhere between extremes. Suffice it to say that even if industrialization remains a possible road to convergence, success is no longer as straightforward as it was once considered. Therefore, the spectacularly rapid convergence achieved by countries in East Asia, which experienced annual economic growth of nearly 4-5% for decades, may be unavailable to much of Latin America and Africa.
What will future economic growth look like? In a Center for Global Development podcast episode, World Bank Group President Jim Yong Kim suggests the need for a “Plan B for Development,” but offers no concrete alternative to industrialization. During an interview with Tyler Cowen, Professor Jeffrey Sachs proposes that what is crucial is that developing countries carve out a niche in global markets and figure out how to export it, not necessarily what they export. For example, it is possible to achieve economic growth by exporting services and expertise, natural resources, or agricultural products rather than manufactured goods. India has been successful at exporting services related to information technology, and Chile has made a name for itself exporting its mining expertise. Commodities have long been a staple for many developing countries as well. However, Mr Sachs’ dismissal of differences in exported products is misguided.
While the service sector does exhibit higher productivity growth than agriculture, the growth is less dramatic and diminishes quicker than that of industry. As such, production develops around a diffuse set of enterprises, preventing development of common interests and the incentive to organize politically. This can increase the likelihood that workers will vote according to personal and ethnic identities, rather than class interests, enabling elites to rig the game in their favor. Additionally, reliance on natural resource exports for growth is notoriously challenging, leading to the notion of a “resource curse” — that resource possession may actually harm development prospects. Dependence on commodities can both distort economies structurally, and lead to political maneuvering between elites to secure the associated wealth. All of this is to say that exporting products other than industrial goods is unlikely to provide the same widespread economic benefits, and thus political benefits, that exporting industrial goods has.
Accurately predicting the economic future and offering a new, grand development paradigm has proven repeatedly to be a fool’s errand. Instead, focus should be on fostering conditions that offer an economy the best chance to effectively respond to any external circumstances. Countries can still achieve sustained growth through the slow, challenging work of accumulating human capital and improving institutions — here, the importance of continual marginal progress cannot be overstated. This of course requires political stability and thoughtful policy. In the face of the apparent decline of democracy around the world, then, it is as important as ever to continue upholding Amartya Sen’s claim that the expansion of freedom is both the primary end and the principal means of development.