The Tale of Two Asian Tigers

The Tale of Two Asian Tigers: Comparing Development in Selected Provinces of China and India since 1980 1 Introduction The fast rates of economic growth of China and India over the past four decades are among the highest recorded in the history of mankind. They have been accompanied by increasing urbanization and industrialization, sharp reductions in poverty, illiteracy and improvements in many indicators of human development. It is notable that these achievements were broadly comparable despite marked differences in history, culture and political systems. Compared to India, China has a long history of a centralized and authoritarian state, and relatively low ethnic and religious fragmentation. Chinese society and politics continues to be dominated by the Chinese Communist Party (CCP), while control of the Indian government at the central and regional level is divided and frequently contested among rival political parties. However, the transition to fast growth in both countries was accompanied by a move towards a free market economy progressively integrated with the rest of the world through flows of trade and investment. While there have been some debates regarding the accuracy of statistical data commonly used, it is broadly accepted that of the two countries, China has achieved faster rates of growth, urbanization, industrialization and poverty reduction. In this lecture I shall start with a detailed review of the key facts. I will then argue that part of the difference can be accounted by underlying differences in the role played by their respective states in the development process, which in turn stem from differences in underlying governance 1 institutions. Given the vast heterogeneity within the two countries, I shall focus on specific provinces in the two countries which were broadly comparable in the 1970s in terms of population size, cropping patterns, occupational structure and standard of living: West Bengal (Figure 1) in India, and three provinces Anhui, Hunan and Jiangxi (Figure 2) in China. West Bengal occupies a middle rank among Indian states in per capita income, with levels and changes in economic indicators close to the Indian average. The Chinese provinces were chosen for their similarity to West Bengal in population, cropping patterns and per capita income in the late 1970s, besides achievng growth similar to the Chinese average. There are also some similarities in political context: for much of the period since the late 1970s, the state government of West Bengal was controlled by the Left Front (LF), a coalition of Left parties led by the Marxist wing of the Communist Party of India (CPI-M) whose ideology was the closest among major Indian political parties to that of the CCP in China. The LF was elected to power in the 1977 state elections, roughly the same time that Deng Xiaoping emerged as the leader of the CCP. In both cases, these changes in political control were accompanied by a large land reform, a program of devolution of authority over economic decision making to local governments, and initiation of a marketbased strategy of industrial growth. The Left Front held an absolute majority in the state legislature all the way from 1977 until 2011, and was dominated by the CPI-M which was a relatively disciplined party with a top-down hierarchy within the state. Of course there was an important difference in political context: the Left Front was exposed to political competition and ultimately lost the 2011 election to the TMC — an issue we shall return to later. Section 2 presents indices of income, consumption, nightlights, poverty, inequality and various human development indicators in the selected provinces from the 1970s until 2015. We then review developments in the agricultural sector in Section 3, with land reforms boosting smallholder agriculture and increases in rice yields in the 1980s in both countries. The data also reveals a sustained difference in rice yields between the two countries. Section 4 reviews structural transformation and industrial growth performance of the two economies. The Chinese provinces achieved higher growth of urbanization, industrial pro- 2 Figure 1: West Bengal: Selected Province in India Figure 2: Selected Provinces in China 3 duction and manufacturing wages, particularly after 2000. While there are potentially many different explanations for their comparative performance, I subsequently discuss the role of three areas of state policy are likely to have been important: R&D in hybrid rice, land acquisition for urbanization, and urban governance. Section 5 provides further details of each of these policy areas, based on available facts and the respective institutional environments in which these policies were formulated and implemented. This discussion suggests that the distinctive political institutions of the two countries had substantive implications for these policy differences. Besides highlighting these, our review of related research literature in these areas indicate a number of significant gaps in current knowledge, thus suggesting interesting directions for future research. Section 6 concludes with a summary, and reflection on broader implications for existing views on determinants of structural transformation in current macro-development literature. 2 Comparative Development: The Key Facts 2.1 Choice of Provinces WB occupies middle rank among Indian states on most measures of development. Its growth experience is also representative of the overall all-India performance. Figure 3 plots per capita GDP in West Bengal (WB hereafter) and India respectively between 1980-2014. WB’s income starts below the all-India average but catches up by 2005 after which the two figures are very close. Both WB and India show a marked acceleration in growth after 2005. Moreover, urban and rural per capita consumer expenditure were remarkably similar both in levels and changes (Figure 4). Figure 5 plots poverty head count ratios (with thresholds based on household nutrition standards) and Gini coefficient of household consumption distribution. WB started with higher poverty (70%) than the all-India rate (55%) during the mid-1970s, but declined faster enabling WB poverty to converge to the all-India average rate of slightly under 30% by 1998. Consumption inequality in WB was the same (.3) as India in 1972, with a similar (slight) downward trend thereafter. Overall, both WB and India experienced a marked acceleration in GDP per capita, large drop in poverty and 4 Figure 3: Per capita GDP: India and West Bengal Figure 4: Per capita consumption: India and West Bengal slight decline in inequality. The three provinces of China were chosen on the basis of two criteria: representativeness of China’s growth experience, and similarity to WB with regard to population and cropping patterns in the early 1980s. With regard to the former criterion, Figure 6 shows these provinces occupied a below median rank within China with regard to 1981 GDP per capita, and in this respect were closer to WB compared with the median Chinese province. Figure 7 shows they were similar to the rest of China in levels of rural consumption, while lagging behind in urban consumption. With regard to growth rates of GDP per capita and consumption, the experience of these provinces were representative of China as a whole. Turning to the second criterion, Table 1 shows rice was the dominant crop in WB, Hunan and Jiangxi accounting for approximately 60% of cropped area, whereas wheat was the 5 Figure 5: Poverty and Inequality: India and West Bengal Figure 6: Per capita GDP: China and Selected Provinces main crop in Anhui. Anhui, Hunan and WB had populations near or slightly above 50 million in 1981, while Jiangxi’s population was 33 million. 2.2 Trends in Living Standards, Inequality and Human Development I start by comparing trends in key indicators of development between WB and the Chinese provinces. Figure 8 shows GDP per capita levels were approximately 50% higher throughout in the Chinese provinces. Table 2 confirms that their per capita GDP growth experiences were similar: annual growth rates over 1981-2015 were identical (6.05%) in WB and Jiangxi, slightly higher (6.3%) in Hunan and slightly lower in Anhui (5.8%). In both countries growth rates nearly doubled after 2001. Consumption per capita however shows a different picture. Figure 9 presents corresponding comparisons of per capita consumption expenditure between WB and the two 6 Figure 7: Per capita Consumption: China and Selected Provinces Table 1: West Bengal and Chinese provinces: Comparison in 1980s Year West Bengal Anhui Hunan Jiangxi GDP per capita 1981 832 1225 1395 1306 National Rank (GDP p.c.) 1981 7/15 26/31 21/31 22/31 Population 1981 55 50 53 33 Share of Rice in Cropped Area 1985 63 26 58 60 Notes: GDP per capita in 2017 US dollars after PPP adjustment. Unit of population is million. For West Bengal, GDP per capita is taken from the Bureau of Applied Economics & Statistics, Govt of WB; GDP per capita rank is taken from CSO; population and population density are taken from Census; rice share is taken from Directorate of Agriculture, GoWB. For Anhui, Hunan, and Jiangxi, GDP per capita from China Compendium of Statistics 1949-2012; population from provincial yearbooks; rice share from Ministry of Agriculture of the PRC (2009). 7 Figure 8: GDP per capita: WB and Chinese Provinces Chinese provinces for which we have the data (for Hunan it is missing for all but the last few years). It is remarkable that consumption levels were almost the same in WB and the Chinese provinces in 1978, when the reform process was initiated. However, consumption grew faster in the Chinese provinces, resulting in a significant divergence: in 2012 living standards were more than twice as high in China. This is partly accounted by a relative stagnation in WB until the late 1990s. In both countries we see an acceleration in growth from 2000 onwards. After 1999, Table 2 shows both rural and urban consumption grew considerably faster in the rice growing province of Jiangxi compared with WB. The discrepancy in the growth comparisons between output and consumption data is somewhat puzzling, but the two sets of data are collected in different ways and many earlier authors have noted similar discrepancies (e.g., Deaton (?), Ravallion (?)). Other possible explanations may lie in differences in saving behavior or in inequality, though these seem somewhat unlikely. Next, trends in poverty and inequality are compared in Figure 10. Here data is not available at the province level within China, so we compare WB with China as a whole. The poverty line in China was 2300 yuan per year at 2010 prices, which translated into $2.12 per day in 2015 prices after PPP adjustment. In India, the PPP equivalent threshold 8 Table 2: Growth Rate Comparisons (a) GDP per capita 1981-2015 1981-2001 2001-2015 WB 6.04 4.18 8.74 Anhui 5.79 4.39 7.81 Hunan 6.32 4.45 9.04 Jiangxi 6.05 3.97 9.10 (b) urban consumption expenditure per capita 1980-2011 1980-1999 1999-2011 WB 3.38 1.97 5.28 Anhui 3.67 2.58 5.34 Jiangxi 4.02 2.25 6.89 (c) rural consumption expenditure per capita 1980-2011 1980-1999 1999-2011 WB 2.39 0.96 4.32 Anhui 3.18 -0.96 6.39 Jiangxi 3.17 1.52 5.83 Notes: GDP per capita (in 2017 US PPP adjusted dollars). Anhui’s growth rate is calculated for 1990-2011, 1990-1999 in the first 2 columns of Table (b) and (c) due to data limitations. For WB, GDP per capita source: Bureau of Applied Economics & Statistics, Government of West Bengal; consumption source: NSSO,Planning Commission. Data source for Chinese provinces: China Compendium of Statistics 1949-2008, provincial statistical yearbooks. 9 Figure 9: Consumption per capita: WB and Chinese Provinces declined from $2.26 in 1973 to $1.51 in 2004. These differences make it difficult to compare poverty across the two countries. The lowering of the poverty line over time in India implies that trend comparisons are also difficult. Based on their respective poverty lines, poverty declined faster in China, while India also achieved a substantial reduction. The falling poverty line however implies the downward trend in India is larger than would have been with a stationary poverty line. Hence the evidence suggests a more significant reduction in poverty in China. On the other hand inequality of consumption nearly doubled in China, while it declined slightly in both WB and India. With regard to levels of various human development indicators, the Chinese provinces consistently outperform WB. WB has substantially higher birth and death rates at the beginning, but these had converged to Chinese levels by 2014. Life expectancy rose in both countries, with a faster rise in WB which helped narrow the gap with China. Infant mortality and illiteracy in both countries fell markedly, and the gaps in these did not narrow. WB initially had a substantially lower female/male gender ratio at birth, but it rose faster, eventually surpassing all three Chinese provinces by 2010. In summary, the two countries achieved similar growth in output over the entire period 1980-2015, while consumption grew faster in China. In both countries growth accelerated markedly after 2000. Standards of living diverged progressively, both in rural and urban 10 Figure 10: Poverty and Inequality Trends: WB and Chinese Provinces areas. China also achieved faster reduction in poverty but experienced a rise in inequality unlike India. Human development levels were consistently higher in China, while WB achieved greater improvements in some dimensions (mortality, fertility and gender ratio) and similar improvements in others (health and literacy). 11 Figure 11: Human Development Comparisons: WB and Chinese Provinces 3 Agricultural Performance Starting from 1978-79 both WB and China embarked on ambitious programs of land reform and agricultural development. The transformation in China is described bu Huang (2018). 1978 Chinese agriculture was organized in communes, and characterized by compulsory grain procurement by the government to sustain the industrial sector, which resulted in massive rural poverty. Following 1978, the reform created a new Household Responsibility System (HRS) where land was leased to individual households who were empowered to produce on their own. Grain procurement levels were lowered, with households allowed to retain residual output above the required delivery target to the government. Procurement prices were also raised. Agricultural markets came into being where households were allowed to sell. These reforms led to significant growth in agricultural output and rural household consumption. McMillan et al (1989) report a 61% increase in output between 1978-84, partly explained by increased agricultural labor force, capital and other material inputs, and TFP growth (which varied between 4-10% per year). They estimate 78% of the associated gains in productivity could be attributed to the changes in farmer effort incentives (measured by share in own-output which increased by a factor of three times), 12 after controlling for changes in output prices. In WB the new LF government also initiated a large program of land reform and devolution of local agricultural development to elected local governments (described further in Banerjee et al (2002) and Bardhan and Mookherjee (2010, 2011)). The land reform consisted of two components: distribution of titles (pattas) to small plots of land acquired in the past from households owning more than the mandated land ceilings, and registration of tenant farmers providing them security of tenure and a minimum crop share (Operation Barga). Between 1978-98, the patta distribution program expanded from 1.4 to 5.4% of cultivable area, with coverage of rural households expanding from 5 to 15%. The corresponding coverage of Operation Barga expanded from 2.4 to 6.1% of cultivable area, and 3 to 4% of rural households. The two programs combined thus affected nearly 10% of area and 20% of rural households by the late 1990s. Bardhan and Mookherjee (2011, Table 5) estimate value added per farm doubled between 1982 and 1995, accounted partly by a 70% rise in cropped area, a five times rise of value added per acre in rice cultivation (owing to a rise in rice area under high yielding varieties (HYV) from 6 to 67%). However, the program in WB was accompanied by a reform in the delivery mechanism of agricultural inputs and complementary rural infrastructure. From 1978 onwards, a three tier system of directly elected local governments was created; the state government shifted responsibility to these elected local officials (away from state bureaucrats in line ministries) for selecting beneficiaries of subsidized agricultural ‘minikits’ containing seeds and fertilizers, subsidized credit, employment programs and construction of local infrastructure (roads and irrigation projects). Bardhan and Mookherjee (2011, Table 11) provide evidence that most of the rise in farm value added per acre between 1982-95 was accounted by the delivery of minikits and credit by local governments, while the contribution of the land reform program was negligible and restricted to the pre-1985 period. However the land reforms contributed in other ways: by inducing expansions in cropped area (Bardhan and Mookherjee 2011, Table 12) and stimulating private investment in minor and medium irrigation (Bardhan, Mookherjee and Kumar 2012). We now examine differences between WB and the selected Chinese provinces with regard to different components of agricultural performance. Figure 12 shows absence of any 13 Figure 12: Cropped Areas: WB and Chinese Provinces Figure 13: Share of Rice in Cropped Area: WB and Chinese Provinces significant changes in cropped area in the Chinese provinces, in contrast to a modest upward trend in WB. Figures 13, 14 show no change in cropping patterns in either country. Figure 15 shows a very marked difference in rice yields, which are at least two times as high in China. Both countries display some growth in rice yields, but the trends are parallel. This indicates a significantly superior level of productivity in Chinese agriculture. To probe possible underlying differences in rice yields, Figures 16, 17 compare fertlilizer and irrigation inputs. Fertilizer application was throughout lower in WB and grew at the same rate as in the Chinese provinces, while irrigation grew faster in WB. Hence differences in fertilizer may account for some of the difference in rice yields. In the next section we shall argue that differences in R&D played an important role. 14 Figure 14: Share of Cash Crops in Cropped Area: WB and Chinese Provinces Figure 15: Rice Yield: WB and Chinese Provinces Figure 16: Fertilizer Application: WB and Chinese Provinces 15 Figure 17: Irrigation: WB and Chinese Provinces 4 Structural Transformation: Urbanization and Industrialization In this section we compare trends in structural transformation and industrial growth. Figure 18 shows a faster drop in the share of the primary sector both in GDP and employment compared with WB. The difference in growth in urbanization is particularly striking: in China the urban share of population rose from 10% to 50% between 1970-2014, compared to an increase from 25 to 31% in WB. Urbanization rates are not comparable across the two countries owing to different definitions, with India using a more restrictive definition that excludes a large number of ‘census towns’ that are classified as urban in the Census but not recognized by the state government. However, trends in urbanization are likely to be more comparable across the two countries, though it is possible that these measured trends are an underestimate in India if ‘census towns’ keep growing Given the large differences in productivity and consumption standards between rural and urban areas, an increase in urbanization tends to raise per capita output and consumption. This is why growth in urbanization tends to play an important role in driving growth in developing countries. Indeed, given the striking difference in growth of urbanization between the two countries, one may have expected China to achieved a substantially higher aggregate growth rate than India. Part of this puzzle may perhaps be explained by underestimation of growth of urbanization in India on account of ignoring growth of ‘census 16 Figure 18: Structural Transformation and Urbanization: WB and Chinese Provinces towns’. It is also possible that the gap between rural and urban productivity in India is larger than in China, consistent with evidence on rural-urban wage gaps: 47% in India in 2007 as against 10% in China in 2006 (Munshi and Rosenzweig (2016)). Urbanization tends to be driven by industrial growth associated with rise in manufacturing and services. Figure 19 shows trends in the number of industrial enterprises. Panel (a) pertains to registered factories in WB and above-scale industrial enterprises in China as a whole and the three provinces, while panel (b) presents data for all industrial enterprises both registered and unregistered (from the Economic Censuses) in WB and China respectively. In both panels we see a faster increase in the number of enterprises in China, particularly after the late 1990s. Figure 20 presents data on employment in registered or above scale factories after 1999, which shows a similar pattern. And Figure 21 shows earnings of factory workers in China and in Anhui grew faster than in WB after the 1990s. In all these respects, the industrial sector was considerably more dynamic in China. Admittedly the choice of WB understates industrial progress compared to the rest of India, as shown in Amirapu and Subramanium (2015).3 However there is no doubt that China achieved much greater structural transformation and growth in manufacturing than India. 3 See their Figures VI-VIII which shows a decline in output and employment share of registered manu- facturing in WB in contrast to a flat profile for India as a whole. 17 Figure 19: Number of Industrial Enterprises: WB and Chinese Provinces 18 Figure 20: Factory Employment: WB and Chinese Provinces Figure 21: Earnings of Factory Workers: WB and Chinese Provinces 19