The Tale of Two Asian Tigers Part 2

5 Underlying Policy Differences In this section we describe a number of specific government policies that likely played an important role in explaining differences in outcomes reviewed in previous sections. 5.1 Hybrid Rice Development We saw the large difference in rice yields between the two countries, and that part of this could possibly be accounted by higher application of fertilizers in China. Another important cause was the faster progress made by China in developing new hybrid varieties of rice. Yu (2012) estimates that 74% of the rise in rice yields between 1980-2009 was accounted by adoption of superior rice varieties, while changes in climate and in application of nitrogen fertilizers accounted for 4.4% and 9.3% respectively. Song et al (2014) estimate the role of improved rice varieties to have been 62% which is lower than Yu’s estimate but still accounts for the bulk of the rise in yields. Hybrid rice varieties owe their origin to scientific discoveries made in China during the mid-1960s, particularly by Yuan Longping widely recognized as the ‘father of hybrid rice’ who continued to be active in R&D for the next forty years. In contrast to ‘inbred’ HYV rice that was being adopted in India and many other Asian countries during the Green Revolution, hybrid rice involves ‘heterosis’ or cross-breeding of two different varieties, involving genetic transfer of male sterility genes from wild rice to create a male sterile line for cross-pollination. Though it involves some genetic manipulation, hybrid rice differs from GMO rice because it relies on a natural pollination process rather than being created in a laboratory via gene splicing. In the first phase, the research advances led to 43 different varieties of hybrid rice in China, which began to be planted from the mid-1970s onwards, and spread rapidly throughout rice growing regions, accounting for 50% of total area planted by 2009 (Khush 2013). These varieties were associated with a yield advantage varying between 10-35% over traditional in-bred rice (Cheng et al. 2007a; Huang et al. 2016; Khush, 2013; Peng et al., 2008; Wang et al. 2014; Xie et al. 2013). A subsequent ‘super-hybrid’ rice project was initiated in 1996. Between 1998 and 2005, 34 commercially released super hybrid rice varieties were grown on a total area of 13.5 million ha and produced an additional 6.7 million tonnes of rice in China (Cheng et al., 2007b). These hybrid varieties produced a grain yield of 12 t/ha in on-farm demonstration fields, 8 — 15% higher than the hybrid check varieties (Peng et al., 2008). The hybrid rice program was launched in India only in 1989, releasing 65 hybrid varieties by 2013. However, the area devoted to hybrid rice in India was only 4% (Singh et al 2015), and most of this was in states outside WB (Prasad et al 2012). Low adoption in India has been attributed to higher seed costs (as every year new seeds have to be purchased instead of recycling from the previous crop), fertilizer requirements, greater susceptibility to disease and pests, and problems with low market price owing to problems with cooking quality (Nirmala et al., 2012; Viraktamath et al., 2012; Siddiq and Prasad, 2012; Singh et al., 2015). Not much is known about how these problems were overcome in China. The public sector played a more important role in R&D efforts for hybrid rice in China. 31 of the 43 varieties released in China were developed by the public sector (including SOEs, universities and state research centers), while another 5 involved joint public-private ventures. In contrast, 40 out of the 65 hybrid varieties in India were developed by private companies. Viraktamath et al (2012) describes how some of the problems associated with slow development in India were related to inadequate public sector investment in research, diffusion of technology and seed production (90% of which happens to be produced by private seed corporations). Karunakaran (2013) estimates 2008 total spending on agricultural R&D to be $3.4 billion (0.5% of GDP) in China, $2.3 billion (0.4% of GDP) in India, as compared to an average of 0.56% of GDP across all developing countries, 1.8% in Brazil, 3.6% in Australia and 4.7% in Japan. 5.2 Land Acquisition Policies The process of urbanization and industrialization requires a reallocation of key factors out of the agricultural sector into industrial and service sectors. One of the most difficult reallocations in practice concerns land which is converted from agricultural use to factories, roads, office buildings and homes for urban dwellers. In countries like China and India dominated by smallholder agriculture, building even a single factory requires acquiring land from thousands of small farmers. Under laissez faire a prospective factory builder would have to negotiate simultaneously with all these farmers, which is a forbidding exercise in coordination besides providing individual farmers to hold up the entire project by asking for a sale price that greatly exceeds their true reservation value. The problem is akin to the free rider problem associated with any public project. Even in market societies, therefore, this market failure creates a rationale for rules of eminent domain which empower the government to appropriate private property from multiple owners for the sake of building (or allowing the building of) public projects whose aggregate social benefits outweigh the costs. Such powers of eminent domain wherever exercised require expropriated famers to be compensated, but it is very hard to design acquisition systems which adequately compensate all farmers since the latter are privately informed about their true reservation price. Usually acquisition laws require farmers to be compensated at least at the market value of the property (though compensating at the market value would still be inadequate for those farmers for whom the personal valuation of their land exceeds its market value). But in practice these laws are often not followed, either because problems in assessing the true market value of land (owing to thinness in land markets, or problems assessing the true quality of the land), or because the government intentionally pays less compensation compared to the market value. This can then give rise to farmer distress leading to adverse inequality impacts and possibly political protests. Both China and India have been subject to these problems in recent decades. Between 1998-2005, the constructed area of Chinese cities grew from 214 sq km to 325 sq km. Cao et al (2008) estimate that this process caused 2.5-3 million farmers to be dispossessed every year, constituting the largest single source of farmer protests (involving approximately 385,000 farmers in 2006 alone). Liu et al (2014) estimate a total of 40 million farmers had been dispossessed by 2006, and estimated this number would grow to 100 million by 2020. Cao et al 2008 explain that the protests were triggered by compensations paid by local governments that were below market value by a huge margin: they were 7 to 10 times lower than the prices charged to private developers, with the difference retained by the local governments. This bolstered the finances of the local governments, and thereby their capacity to finance large urban infrastructure projects (besides also possibly generating high opportunities for corruption). Effectively this meant that the farmers (rather than the developers) ended up financing urban infrastructure (rather than the actual beneficiaries of the infrastructure). As we explain in the next section, the problem was compounded by competition across different city officials to attract private investment, induced by policies of the central government where local mayors and province governors’ promotion prospects depended on how well they fared in boosting growth in their city relative to growth in other parts of China. The central government has also sensitive to the problems faced by farmers and sought in various ways to limit the extent to which these problems arose with suitable laws governing compensation. However, Cao et al 2008 report from a 16 city survey that 50% of the land acquired violated these laws. Farmer perceptions of unfairness were compounded by the lack of any systematic procedure for determining actual compensations that were based on negotiations and discretion of local government officials. Ding (2013) argues that the process resulted in possible over-investment in manufacturing, manifested 30% of real estate consisting of industrial properties as compared to less than 10% in Hong Kong, Seoul and Paris. West Bengal has also been subject to serious political problems arising from land acquisition via eminent domain. Indeed, there were two prominent cases of such acquisition in 2006 and 2007 in Singur and Nandigram respectively where the Left Front government acquired land from a large number of farmers under the aegis of the 1894 Land Acquisition Act, in order to allow prominent industrialists to build a car factory and chemical hub respectively. These acquisitions led to a large number of farmer protests, eventually creating a law and order problem resulting police violence against the protesters. The news of these incidents were widely circulated in the media, generated widespread criticism from civil society groups, intelligentsia and activists. The protests led to cancellation of both projects. The opposition party TMC capitalized on these events in its electoral campaign to unseat the Left Front, eventually succeeding in defeating the Left Front and wresting an absolute majority in the 2011 state legislative elections. While these events were not the only cause of the loss of power of the Left Front after 35 consecutive years of incumbency, they likely did contribute to it (Bardhan et al 2014). Interestingly, compared to the Chinese experience, the problem in WB was less related to systematic under-compensation of farmers whose lands were acquired. Ghatak et al (2014) in their detailed study of lands acquired in Singur show that on average the actual compensation paid by the government was slightly above the legally required compensation of 130% market value (even when the market value used in their estimates were self-reported by the farmers). The problem instead lay elsewhere: in the dispersion of actual compensations around the average. Owing to outdated land records, the government misclassified land quality, resulting in overcompensation of some farmers and under-compensation of others. It was the latter group that constituted the bulk of the protesters. For the latter farmers, detailed survey evidence of their earnings from different sources show that they were indeed left worse off. Besides direct losses of the owners, there were significant indirect costs borne by other groups: tenant farmers who were leasing in land that was acquired were no longer able to cultivate those lands, and landless agricultural workers lost employment opportunities. Neither of these two groups were entitled to any compensation under the 1894 Law. Similar problems of farmer, civil society and media protests have arisen in other parts of India as well. Misra (2020) provides evidence that a large fraction (approximately 40%) of proposed SEZ projects in India did not receive approval, and that a political economy model of the cohesiveness of farmer coalitions (which is assumed to be decreasing in land inequality, religious and social fragmentation) can explain the pattern of lack of approvals across different Indian states. As the demise of the Left Front in WB showed, the fact that incumbent governments in Indian states are subject to electoral competition means that the scope for land acquisition is more limited in India compared to China. The WB experience and the protests against SEZs in other states eventually led the Indian Parliament to supercede the antiquated 1894 Land Acquisition Act and pass a new law (LARR) in 2011 governing land acquisitions which imposed several lengthy regulatory checks. These contributed to a further slowing of approvals after 2011. It is therefore quite plausible that the greater political resistance to land acquisition arising from pressure exerted by farmer groups and related civil society activists were partly responsible for the slower speed of structural transformation in India. 5.3 Urban Governance I turn now to differences in patterns of urban governance between the two countries. While systematic research on this topic is lacking, a number of different case studies, reports and examination of municipal government budgets indicate wide differences in the extent of financial autonomy, responsibilities devolved as well as accountability of urban government officials. This implies corresponding differences in the organizational capacity and motivation of urban governments to build infrastructure and attract private investment. The importance of infrastructure is suggested by Table 3 which compares firm owners’ responses to questions regarding the most significant obstacle to their doing business in different locations in WB, Hefei in Anhui province and the average across 25 cities in China.4 While these indicate relative importance of different obstacles in different regions, and incorporate a combination of demand and supply factors rather than the latter alone, it is notable that access to finance, informal sector practices, and tax rates seem to be the most important problems in both WB and China (rather than problems emphasized in much of the academic literature such as labor regulations, corruption, access to land or courts). However, they differ significantly on the importance of one dimension: electricity which was listed more often in WB (13.4%) than in China (4.8%) or Hefei (0.4%). Infrastructure of course involves a larger range of services than just electricity supply. The 2011 Urban Infrastructure Report submitted by a High Powered Expert Committee (HPEC) appointed by the Ministry of Urban Development of the Indian government reported severe deficiencies in Indian cities relative to international norms with respect to drinking water, sewerage, solid waste management, roads, street lights and affordable housing. Indian cities have huge slum populations with severely deficient services. For instance, municipal water supply is available to Indian households for an average of 1-6 hours per day, compared to 24 hours in Chinese cities. Required improvements were listed in municipal finances (tax reforms and transfers from upper level governments), accounting systems, human resources and land acquisition processes (which accounted for 70% of delays in 4 The World Bank surveys are conducted across states in India and cities in China. Hefei is the only one city from the selected three provinces in China covered by the survey. So we compare WB with Hefei and the average across Chinese cities. Table 3: Firm Survey Responses on Obstacles to Private Investment Percent of Firms For Whom Most Significant Obstacle was: WB China (average) Hefei Access to Finance 24.7 22.4 13.9 Informal Sector Practices 19.0 19.6 42.5 Tax Rates 17.2 15.1 4.2 Electricity 13.4 4.8 0.4 Labor Regulations 3.9 1.9 0 Corruption 3.3 1.2 0 Licensing and Permits 3.3 0.2 3.9 Access to Land 1.9 5.6 23.9 Courts 0.1 2 0 Source: World Bank Firm Surveys, WB 2014, China (average of 25 cities) 2012. infrastructure projects in India in 2008). It also highlighted deficiencies in organizational structure and accountability systems in municipal governments, an issue we return to below. Differences in infrastructure reflect differences in spending on infrastructure. Per capita municipal expenditures in India in 2010 were estimated at $50 per year by a McKinsey Report (cited in Ahluwalia et al 2019), compared with $362 in China, $508 in South Africa and $1772 in the UK. Corresponding capital expenditures were $17 in India, $116 in China, $127 in S Africa and $391 in the UK. Based on data reported in Ahluwalia et al (2019), municipal government expenditures in WB constituted 0.8% of state GDP in 2013. This contrasts with 4.08% in China, 2.87% in Anhui, 2.01% in Jiangxi and 2.94% in Hunan (based on data from China City Statistics Yearbooks).5 Song (2013) estimates urban infrastructure investments in China rose from 0.5% of GDP in 1980 to over 3% in 2003, 5 It could be argued that these estimates are not comparable across China and India because local governments in China bear most of the responsibility for education expenditures, whilst this is a state government responsibility in India. However, even if we subtract spending on education from the spending estimates in China, we obtain urban fiscal spending to be 3.58% of GDP in China, 2.52% in Anhui, 1.79% in Jiangxi and 2.65% in Hunan, which is higher than the corresponding proportions in WB by a factor of at least three times. and thereby declined slightly to 2.5% in 2008. Part of the reason for low spending on municipal services in India is low tax raising capacity of municipal governments. In 2013 grants from upper level tiers of government constituted 73% of total income of municipal governments in WB. Property and land taxes accounted for only 13% of their revenues, with ratepayers constituting 6.6% of the population, and collections amounting to 26% of assessed demands, reflecting low collection capacity (West Bengal Municipal Statistics 2013).6 Clearly, WB municipal governments are restricted in their fiscal autonomy. This has been further undermined in recent years after the passage of the new goods and service tax (GST) on an all-India level, which seeks to harmonize taxes throughout the country by eliminating devolution of octroi and other local taxes from local governments. In contrast, in China the role of central budget grants in local urban finances fell from 40% in 1980 to near 0 in 2000 (Song 2013). After 2000, three sources accounted for approximately one third each of total spending of urban bodies: domestic loans, the local government budget and self-financing mechanisms involving extra-budgetary revenues such as land transfer fees. Song also provides detailed econometric evidence of the role of infrastructure (roads and transport) in stimulating the conversion of land to urban development and thereby raising land prices. Resulting increases in land transfer fees then constituted a form of self-financing in China that seem to have no equivalent in WB. In India, borrowing constitutes a small fraction (between 2-3%) of municipal budgets. The municipal bond market which developed in India in the late 1990s (allowing Kolkata municipal bodies to supplement their financing in this manner) contracted sharply from 2005 onwards. Banerji et al (2014) explain this by growing creditworthiness concerns on the part of bond investors regarding the capacity of municipal governments to generate own revenue, low productivity of infrastructure investments resulting from low organizational efficiency, and about whether elected governments would honor repayment commitments. The allocation of grants to municipal bodies by the WB state government do not appear to follow any transparent process, so it is hard to understand how they are determined, and 6 Admittedly WB is an outlier among Indian states in this regard, as the grants constitute 40% of municipal income for all Indian municipalities on average. whether they generate incentives to municipal governments to raise local taxes. Moreover, accounting practices are lax, as the Comptroller and Auditor General (2015) report on WB mentioned that most urban local bodies failed to present accounts on time, and exhibited inadequate internal control mechanisms (quoted in Ahluwalia et al 2019). There is also a growing literature on selection and incentives of local leaders in China resulting from a top-down structure within the CCP generating a system of career mobility via promotion incentives. Maskin et al (2000) initiated this literature highlighting a key difference between Soviet style ‘U-form’ governments based on line ministries and Chinese style ‘M-form” decentralization to local governments, resembling different forms of quasicorporate organization. Much of the related literature has dealt with this mechanism as it applied to leaders of provincial governments in China whose promotion prospects to the Central Politburo depended on competition based on measures of growth performance of their respective provinces (Li and Zhou (2005), Xu (2011) and Jia, Kudamutsu and Seim (2015)). Papers by Zhang (2015) show a similar phenomenon applied also to career mobility patterns of city mayors, using data from leaders of 312 Chinese cities between 1994-2010. Chen, Li and Yu (2018) show that performance evaluation systems for city mayors were modified in 2005 to give greater weight to reduction in sulphur dioxide emissions, which subsequently resulted in a reduction in emissions as well as local GDP growth rates. Hence the evidence suggests an explicitly designed top-down incentive mechanisms for city mayors to pursue goals set by the CCP. The existence of such disciplinary mechanisms reflect the unique political hierarchy within the CCP. While no comparable studies exist of selection and career incentives of city leaders in WB or India, it seems unlikely to be as high powered as in the Chinese system. 6 Summary and Concluding Observations In summary, both WB and the comparable Chinese provinces experienced similar growth in GDP per capita between 1980-2014, while WB experienced slightly slower growth in consumption per capita, and a substantially slower structural transformation from an agrarian to an urban industrial economy. Living standards were higher in China to start with, with a similar gap in proportional terms but widening one in absolute terms. While data comparability problems make it hard to compare the extent of poverty reduction it appears that poverty dropped faster in China. On the other hand, consumption inequality rose markedly in China, but fell in WB. Indices of human development were throughout higher in the Chinese provinces, though on some dimensions the gap narrowed and in others they remained the same. Growth rates in agriculture were similar; rice yields in the rice growing provinces were throughout almost double that of WB, most likely explained by greater investments and success in R&D and diffusion of hybrid rice in China. Urbanization and industrial growth in the Chinese provinces outstripped WB by a wide margin. While there are many possible explanations for the faster structural transformation in China, we discussed two that stemmed from its distinctive political economy. The first was a faster pace of land acquisition from farmers for purposes of conversion to factories, real estate and construction of related urban infrastructure. Expropriated farmers were substantially under-compensated in China relative to market value, unlike WB. This under-compensation provided a substantial fraction of local governments financing of urban infrastructure spending. In both countries the process of land acquisition gave rise to farmer protests, but the stronger exposure of incumbent governments in India to electoral competition and pressure from farmer groups, media and civil society caused many SEZ projects in India to be reversed or not receive approval. Democratic institutions in India therefore likely imposed stronger curbs on the structural transformation process. The second difference was in the nature of urban governance. There was greater devolution of authority and responsibilities to Chinese city governments, accompanied by accountability of appointed city mayors to upper echelons of the CCP induced by a withinparty promotion mechanism. Accordingly city governments had greater financial resources, autonomy and incentives to pursue goals of growth by investing in city infrastructure in order to attract private investment. Municipal governments in India (and WB in particular) were devolved less authority and finances by upper level governments, had low tax collections and fewer instruments of financing, organizational efficiency was low as responsibility for infrastructure investment was divided among municipal, state, central governments and parastatal corporations, oversight mechanisms were weak, and overall the Indian state de- voted less resources on urban infrastructure. By enhancing the organizational effectiveness of city governments, the Chinese state therefore played a more positive role in the process of urbanization and industrialization. I conclude with some observations on the macro-development literature on structural transformation in the light of our descriptive account of comparative development in the selected Indian and Chinese provinces. This literature focuses on the substantial gaps in labor productivity (Gollin et al 201?) and wages (Munshi and Rosenzweig 2016) between urban and rural areas that are unlikely to be explained by human capital differences, thus raising the question why these gaps persist rather than get eroded by labor migration that generates faster structural transformation. Part of the answer may lie in policy induced restrictions on rural-urban migration. However such restrictions were indeed in place in China in the form of its hukou registration system; there were no such policy-based migration restrictions in India, so this seems to be unlikely to account for the larger rural-urban wage gaps in India. Some recent literature has explored the role of differences in informal insurance between rural and urban areas in India (Munshi and Rosenzweig 2016), uncertainty, risk aversion and borrowing constraints faced by rural households in neighboring Bangladesh (Bryan et al 2011) and the possible role of non-pecuniary differences in urban and rural standards of living (Lagakos et al 2020). Such non-pecuniary differences may arise from poor amenities in urban slums resulting from low investments in urban infrastructure, besides weaker social networks in urban areas that limit access to consumption insurance and information. Differences in urban governance between China and India may thus partly explain explain the slower rate of labor migration in India, besides lowering the growth of private investment in urban areas that lowered growth of employment opportunities.