How does economic development happen? After World War II, many development economists rose to prominence, such as Paul Rosenstein-Rodan (the big push), Arthur Lewis (the dual-sector model), Walter Rostow (the linear stages of growth) and Albert Hirschman (unbalanced growth and linkages). Given the continued importance of industrial policy, it is particularly worthwhile to revisit the idea of forward and backward linkages 鈥� one of the central tenets of development thinking pioneered by Hirschman.

In Hirschman鈥檚 The Strategy of Economic Development (1958), he proposes an unbalanced growth theory, emphasising specific industries which have particularly strong linkages with the rest of the economy. He argues that a developing country can grow from prioritizing strategic sectors even with a relatively small set of resources to utilize.
Hirschman uses the concepts of complementarity and external economies to support his views on unbalanced growth. Complementarity is an interdependence among industries in the production process and the external economy is the subsequent growth of other industries originating from growth of a given industry. The relationship among domestic industries is known as 鈥榓 domestic linkage鈥� (sometimes just referred to as linkage). Broadly speaking, there are two kinds of linkages: backward and forward. On the one hand, a backward (input-provision) linkage of a given industry is strong when its growth stimulates the production/use of other upstream industries. An increase in these inputs is thus required to sustain the production process. For example, a rising demand in cars leads to an increase in the demand for automotive parts (e.g., wheels, seats, and thermal). Consequently, when backward linkages are strong, growth fuels the rest of the economy. On the other hand, forward linkage effects occur when the output of an industry becomes the input for other industries. For example, an increased amount of rubber can lead to an increase in the production of goods that use rubber as inputs, such as tyres and gloves. However, growth driven by linkages also hinge upon several factors such as institutions (think of rules that encourage supply management among firms without burdensome red tape) and services (think of infrastructure needed to support several stages of production).

Backward and forward linkages are used to identify 鈥榢ey sectors鈥� to focus on and invest in (Rasmussen, 1956; Hazari, 1970). The key sectors are considered more capable of contributing to growth through their spread effects, compared to low-linkage industries (Yotopoulos and Nugent, 1973). The concept of key sectors was widely used in identifying sector priority in the import-substitution industrialization (ISI) era of the 1950s and 1960s. The objective of ISI was to achieve economic growth through developing domestic capabilities of an economy to produce manufactured goods that used to be imported, so trade barriers had to be erected for that purpose.

The existence of strong linkages was for example used as a criterion to target certain industrial sectors in South Korea. South Korea pursued the enacted in the 1970s. During that time, were selected: (a) iron and steel, (b) non-ferrous metal; (c) machinery; (d) shipbuilding; (e) electrical appliances and electronics; and (f) petrochemicals. Even though this big push in South Korea was temporary, (2017) finds that its effects are lasting long in targeted industries relative to other manufacturing sectors.

The focus on linkages: Still fruitful today?
In the current era of economic globalization, it is questionable whether the same targeted policy approach of promoting high domestic industrial linkage is applicable and meaningful as policy guidance.

The twentieth century witnessed a radical change in international trade as the production of single goods spans various countries that are part of the global value chain. In other words, thanks to lower costs of communication and coordination, there has been an international separation of factories (Baldwin, 2016). A classic example of this is the production of the Ford Escort whose components come from many countries, including the fan belt from Denmark, the radiator and heater hoses from Austria, the fuel tank from Germany, and the glass from Canada (Dicken, 1986). Another example is how Boeing Dreamliner 787 is manufactured across countries (see figure below).

Source:

The break-up of production processes into vertically separated stages carried out in several countries has become a key structural phenomenon of present-day global trade (Krugman, 1995; Jones and Kierzkowski, 2001; Athukorala, 2005; Feenstra, 2009). This phenomenon, which is referred to as 鈥榞lobal production sharing,” (GPS) opens up opportunities for countries to specialize in different segments of the production process depending on their relative cost advantage and the other economic fundamentals relevant for cross-border spread of production processes. An important facet of GPS is that it expands the choice of development strategies to pursue industrialization.

Without GPS, it would be necessary for developing countries to be proficient in all components of production in order to compete in the global market. However, GPS allows developing countries to join the production networks by focusing only on some tasks. If they can competitively export these components, they this can allow them to grow and produce for a big global market.

Promoting specific industries which have strong domestic linkages can be difficult in several ways under the era of global production sharing. On the forward linkage side, a country can produce a tiny piece of part of hard disk drive and export it to be assembled in another country. This practice naturally sets a limit on domestic forward linkages because this production does not lead to an establishment of further industries in the same country. On the backward linkage side, a final assembly activity in a given country can rely entirely on imported intermediate inputs. This process does not require any domestically produced inputs resulting in low domestic backward linkages. Still, both forms of GPS can create jobs.

Support for sizable domestic linkages can pose another limitation when a country starts to produce more capital-intensive goods due to a rising proportion of capital-intensive intermediates. An example is when the use of synthetic fiber, that is more capital-intensive, was substituted for natural fiber, which requires more labor in its production process. This resulted in a rising capital-labor ratio in many industries, for example, textile, footwear, and electrical equipment. Accordingly, an increased use of domestically produced capital-intensive intermediates in labor-intensive industries can harm employment despite its linkages. It has been considered a curse of developing countries (Little, 1981).

Industrialisation-led development has long been a key strategy for developing countries. Yet, an emphasis on domestic linkages, an interconnectedness among domestic industries, may no longer be relevant. A focus on linkages, which is used as a powerful argument to support active industrial policy, appears to be increasingly difficult in contemporary times, when countries are encouraged to integrate into a highly globalized and fragmented trading system.

References
Athukorala, P 2005, 鈥楶roduct fragmentation and trade patterns in East Asia鈥�, Asian Economic Papers, vol. 4, no. 3, pp. 1-27.

Baldwin, R 2016, The Great Convergence, Belknap Press, Cambridge, MA.
Dicken, P 1986, Global Shift: Industrial Change in a Turbulent World. Harper & Row, London.

Feenstra, R C 2009, Offshoring in the Global Economy: Microeconomic Structure and Macroeconomic Implications, MIT Press, Cambridge, MA.

Hazar, B R 1970, 鈥楨mpirical identification of key sectors in the Indian economy鈥�, Review of Economics and Statistics, vol. 52, no. 3, pp. 301-305.

Hirschman, A O, 1958, The Strategy of Economic Development. Yale University Press, New Heaven, CT.

Jones, R W & Kierzkowski, H 2001, 鈥楢 framework for fragmentation鈥�, in Fragmentation: New Production Patterns in the World Economy (eds Arndt, S W & Kierzkowski, H), Oxford University Press, New York.

Krugman, P 1995, 鈥楪rowing world trade: Causes and consequences鈥�, Brooking Papers on Economic Activity 1: Macroeconomics, 25th Anniversary Issue, pp. 327-377.

Lane, N 2017, 鈥楳anufacturing revolutions: The role of industrial policy in South Korea鈥檚 industrialisation鈥�, Retrieved from

Little, I M D 1981, 鈥楾he experience and causes of rapid labor-intensive development in Korea, Taiwan Province, Hong Kong, and Singapore and the possibilities of emulation鈥�, In: Export-led industrialization & development (ed Lee E.), pp. 23-45. Geneva: International Labor Organisation.

Newfarmer, R S, Page, J & Tarp, F (2018), 鈥業ndustries without smokestacks: Industrialization in Africa reconsidered鈥�, Oxford University Press, Oxford.
Rasmussen, P N 1956, Studies in Intersectoral Relations. North-Holland Publishing Company, Amsterdam.

Yotopoulos, P A & Nugent, J B 1973, A balanced-growth version of the linkage hypothesis: A test. Quarterly Journal of Economics, vol. 87, no. 2, pp. 157-71.

Wannaphong Durongkaverojis a PhD candidate at Arndt-Corden Department of Economics, Crawford School of Public Policy, College of Asia and Pacific, The Australian National University (e-mail address: wannaphong.durongkaveroj@anu.edu.au). Photo by .

As within-country inequality is on the rise worldwide, considering how people actually perceive inequality in their societies and how they respond to it is a question worth asking. In 1973 Albert Otto Hirschman proposed an explanation of changing tolerance for inequality associated with different 鈥榮tages鈥� of the development process. In this post I鈥檒l revisit Hirschman鈥檚 theory and link it to emerging studies of how inequality is perceived in China. The Chinese people generally seem to be satisfied with rising inequality, yet it is unclear how long this tolerance will last.

Development inevitably creates both winners and losers. This generates, at least in the early stage, an economic disparity in society. In 1973 Hirschman published 鈥淭he Changing Tolerance for Income Inequality in the Course of Economic Development鈥�. In it he argues that in the early stage of development, when income inequality among different classes, sectors, and regions tends to rise, society鈥檚 tolerance for this inequality will be substantial. The reason people can tolerate this inequality is that they hope and expect that this disparity will fall at one day or another. However, if this disparity does not narrow, there comes a point when people will no longer endure it. Hirschman uses the example of a two-lane tunnel to explain how people respond (the so-called 鈥淭unnel Effect鈥�). Suppose you are in a traffic jam, you are in the left lane and cannot move. Then, you can see that the cars in the right lane start to move gradually. You feel good because you know that good things will come. Start thinking about the beach? Even though you can鈥檛 move, you feel better because you expect to move soon. The initial gratification is the tunnel effect. However, suppose that it is only the right lane that can move and yours never starts moving. You feel furious and unfairly treated and want to do something to correct this injustice. Your expectation has not been met.

Hirschman also explains the determinants of this effect. He states that the tunnel effect will be strong in cases where the group that does not advance (the left-lane car) is able to empathize with the group that advances. In this case, class matters. If people from your class are a part of the growth process, you will be more likely to expect some positive spillover effects, such as new jobs or other opportunities. However, if the group of people involved in the growth processes are from a different class, ethnic, and religious group, the tunnel effect may not be realized.

In a homogenous society where resources are owned domestically, tolerance for income inequality tends to be high because there is no language, ethnic, or other systematic barrier that keeps the people lagging behind from empathizing with those benefiting from the economic growth. However, Hirschman suggests that the greater the tolerance, the greater is the scope for the reversal that comes once the tunnel effect wears off, if inequality is not corrected in time. National homogeneity is defined in terms of static characteristics, for example, unity of race, language, and religion. However, another effective homogenizing agent is an intense historical experience, for example, a war or political revolution, if shared by all members. Thus, the historical experience of country can also reveal something about the likely strength and duration of the tunnel effect.

Studies of inequality in China, whose economy has transformed dramatically over the past decades, are now emerging. Despite the difficulty in analyzing an evolution of Chinese inequality over time due to the limitation of data availability in household surveys, research by Milanovic (2016) and Alvaredo (2017) suggests that although income inequality has been rising, it is likely to reach a plateau and eventually start to decline, corresponding with a Kuznet鈥檚 curve. Kanbur et al. (2017), Wang et al. (2014) and Han et al. (2016) actually suggest that income inequality started decreasing after 2010. However, the ownership of capital is still heavily concentrated among urban households, similar to many other developing countries (Chi 2012).

There are two elements of Hirschman鈥檚 theory that allow for empirical testing in this context. The first is to examine whether the poor still have a positive attitude towards inequality (and development). The second is to examine whether the poor start being more unsatisfied with their condition over time as inequality increases or fails to decline. Although no long-term household survey exists that can give us definitive answers, there are several studies that provide us with indicative information.

Knight and Gunatilaka (2010) find that rural households tend to report higher subjective well-being than do the richer urban households. In fact, several studies find a positive relationship between the Gini coefficient and happiness in a Chinese context (Knight et al. 2009, Jiang et al. 2012), but it is worth noting that these results may not be generalized across China (indeed, Wu and Li 2013 finds that local income inequality significantly reduces the level of happiness). For example, Cheung (2016) finds that inequality is associated with higher life satisfaction in rural areas, but not in urban areas (drawing on a large sample of 30,255 Chinese respondents). This result is consistent with Whyte and Im (2014), who use Chinese national surveys from 2004 and 2009 to examine people鈥檚 attitudes to income inequality. Their key questions are whether the rise in income gaps lead to rising popular anger about current inequality patterns and whether the social contours of attitudes toward current inequalities shifted over the five years between surveys. They find that 2009 respondents were significantly more likely to view current inequality as fair, despite the increases in the income gap in this period. It also suggests that 2009 respondents did not rise in anger against the rich and successful, but that they rather expressed a stronger desire for government-provided social safety nets. Another example is from Wang et al. (2015), who investigate an inverted-U shaped association between the Gini coefficient and individual happiness. They find that a certain level of income inequality is beneficial to individual happiness, while an excessive income inequality yields the opposite effect. Inequality can encourage a strong motivation to work hard and obtain more wealth, especially among the middle and lower class income groups. However, when income inequality goes beyond a personal threshold of tolerance, this can lead to anxiety and reduce confidence in future well-being. Thus, based on the studies mentioned above, it is similar to the Hirschman鈥檚 key quote 鈥淎s long as the tunnel effect lasts, everybody feels better off, both those who have become richer and those who have not鈥�.

Figure 1: China鈥檚 Gini Index Since 1980

Source: Solt (2016)

Note: Graphs are derived from the Standardized World Income Inequality Database (SWIID). Gini Index of inequality in household disposable income uses post-tax and post-transfer income. Gini Index of inequality in household market income uses pre-tax and pre-transfer income.

The figure above shows that income inequality using the Gini Index (market income)[1] has increased since the 1980s and started to flatten out in the 2010s. This raises the challenging question of how the Chinese respond to a slight decrease in income inequality. However, there is no recent study to my knowledge on how people respond to a lower inequality at this time. For the prediction based on Hirschman鈥檚 proposition, there would not be a discontentment towards inequality among the Chinese community after this period because income inequality is expected to decline.

However, one of the missing explanations of Hirschman鈥檚 argument is that it does not inform us how long the left-behind people can tolerate for income inequality. In the Chinese case, a negative feeling can occur in the 1990s or the early 2000s after a decade of high growth with increasing inequality. A study by Smyth and Qian (2008), using a large-scale survey in 31 Chinese cities with 9,284 respondents in 2002, find that people who think that the income distribution is unequal report lower levels of happiness, especially among poor respondent. This may signal a change from gratification to indignation. At some point, people could become unhappy about the inequality. However, as mentioned there are other studies that suggest that many Chinese people are still optimistic about inequality. These contradictions suggest that there is a need for further studies on this topic.

The world today has dramatically changed since Hirschman theorized about the tunnel effect 45 years ago, as societies are more connected and highly interdependent through international economic linkages. These changes may call for a new set of frameworks for analyzing how people perceive economic disparity.

Reference

Alvaredo, F, Chancel, L, Piketty, T, Saez, E, & Zucman, G 2017, 鈥楪lobal inequality dynamics: new finding from WID.world鈥�, National Bureau of Economic Research Working Paper 23119.

Cheung, F 2016, 鈥楥an income inequality be associated with positive outcomes? hope mediates the positive inequality-happiness link in rural China鈥�, Social Psychological and Personality Science, vol.7, no. 4, pp. 320-330.

Chi, W 2012, 鈥楥apital income and income inequality: evidence from urban China鈥�, Journal of Comparative Economics, vol. 40, pp. 228-239.

Kanbur, R, Wang, Y & Zhang, X 2017, 鈥楾he great Chinese inequality turnaround鈥�, BOFIT Discussion Papers 6/2017. BOFIT Institute for Economies in Transition, Bank of Finland.

Knight, J, Song, L & Gunatilaka, R 2009, 鈥楽ubjective well-being and its determinants in rural China. China Economic Review, vol. 20, pp. 635-649.

Knight, J & Gunatilaka, R 2010, 鈥楾he rural-urban divide in China: income but not happiness?鈥�, Journal of Development Studies, vol. 46, pp. 506-534.

Hirschman, A & Rothschild, M 1973, 鈥楾he changing tolerance for income inequality in the course of economic development鈥�, The Quarterly Journal of Economics, vol. 87, no. 4, pp. 544-566.

Milanovic, B 2016, 鈥楪lobal inequality: a new approach for the age of globalization鈥�. Belknap Press, Cambridge, MA.

Smyth, R & Qian, X 2008, 鈥業nequality and happiness in urban China鈥�, Economics Bulletin, vol. 4, No. 23, pp. 1-10.

Solt, F. 2016, 鈥楾he standardized world income inequality database鈥�, Social Security Quarterly, vol. 97. SWIID Version 6.2, March 2018.

Wang, C, Wan, G & Yang, D 2014, 鈥業ncome inequality in the people鈥檚 republic of China: trends, determinants, and proposed remedies鈥�, Journal of Economic Surveys, vol. 28, no. 4, pp. 686-708.

Wang, P, Pan, W, & Luo, Z 2015, 鈥楾he impact of income inequality on individual happiness: evidence from China鈥�, Social Indicators Research, 121, 413-435.

Whyte, M K & Im, D-K 2014 鈥業s the social volcano still dormant? trends in Chinese attitudes towards inequality, Social Science Research, vol. 48, pp. 62-76.

Wu, X & Li, J 2013 鈥楨conomic growth, income inequality and subjective well-being: evidence from China鈥�, PSC Research Report 13-796. Population Studies Center.

Wannaphong Durongkaveroj is a PhD candidate at Arndt-Corden Department of Economics, Crawford School of Public Policy, College of Asia and Pacific, The Australian National University (e-mail address: wannaphong.durongkaveroj@anu.edu.au).

[1] According to Solt (2016), the SWIID covers several income inequality data set. User can see several types of inequality based on different types of income, for example, net-income inequality (disposable income) and market-income inequality (pre-tax and pre-transfer income). The author suggests that net-income inequality series is likely to be the best suited to researcher鈥檚 definition of inequality because market-income inequality cannot be interpreted as a pre-government inequality as it also includes non-redistributive policies, for example, public education and job-training programs, which can affect income distribution. However, the use among these two types of inequality depends mainly on researcher鈥檚 question.