2. Some theory

As discussed above, all political leaders or elite need sources of revenue in order to maintain their position of power. This revenue can come from a number of sources. Natural resources (oil, natural, gas, etc.) can via exports generate a stream of revenue that a state can very easily tap into through state ownership or via agreements with private producers. Foreign aid can also serve as a source of revenue not requiring a lot of effort upon the part of the state in order to tax this source. In this paper, these sources of revenue will not be discussed, instead the focus will be on tax revenue.

Levying taxes requires the state to gather information on income, wealth, and economic activity in general. In addition, an institutional structure must be constructed to allow for the collection of the revenue. In many circumstances, taxes will only be paid if services are provided in return for the tax revenue and this is labeled as reciprocity. The willingness to pay a tax is often related to the quality and quantity of services received in return for that payment (Levi, Reference Levi1988; Moore, Reference Moore2001).

In poor countries, the collection of revenue runs into several problems and these are clearly discussed in Gordon and Li (Reference Gordon and Li2005). First, the ability to provide quality services in return for revenue is often quite limited. This stems from a lack of material resources as well as limited human capital. In addition, the ability to construct the institutions necessary to collect information and then use that information to collect revenue and administer penalties for non-payment is also quite limited. The result is then that many producing firms can avoid paying taxes by shifting their operations to the informal sector. In the latter, one does not have to pay taxes, but one gets no formal sector services either. Alternatively, firms can choose to operate in the formal sector and receive services, but then one must also pay taxes. The services of the formal sector include financial and marketing services, access to some forms of public infrastructure, etc.

In less-developed countries (as compared with developed countries), a larger share of economic activity occurs in the informal sector. Firms have to choose whether the benefits that accrue from using formal sector services outweigh the taxes that may possibly be levied on them. In many poor countries, the quality of services provided by the state is low and the probability of being caught and prosecuted for non-payment is also quite low. Thus, a large proportion of the firms involved choose to operate in the informal sector. In addition, the state must concern itself with the fact that any increase in taxes will likely drive more economic activity into the informal sector. Thus, in most less-developed countries, tax rates are much lower than those found in higher income countries (Gordon and Li, Reference Gordon and Li2005).

Thus, developing countries face a fundamental dilemma in terms of raising revenue. According to Gordon and Li (Reference Gordon and Li2005), this has led the political elite in many of these countries to pursue policies that reallocate resources from the hard-to-tax informal sector to the easier-to-tax formal sector. In order to make sense of this argument, one must first explain the characteristics of firms that are likely to operate in the two sectors. The formal sector is likely to include firms that are large and capital intensive in nature. This is due to the fact that these firms are much more dependent on the financial services provided by the formal sector. In addition, firms that are highly dependent on infrastructure, ports, communication network, etc., are also likely to be members of the formal sector. Alternatively, firms that are more labor intensive and less dependent on marketing, port facilities, modern communication systems, and operate in the interior of the country are more likely to operate in the informal sector. Much agricultural production generally falls into the informal sector as well as some proto-manufacturing as well.

Thus, the political elite face real difficulties raising revenue. However, if policy can be used to transfer resources out of the hard-to-tax and into the easy-to-tax sector, then revenue streams can be generated. What policies are likely to be used to achieve this goal? First, trade taxes are likely to play an important role. The most critical will be tariffs on imported goods that would compete with domestic, formal sector firms. Thus, tariffs will be used to protect capital (both human and physical) intensive firms that are significantly dependent on formal sector services. This raises the domestic price of formal relative to informal sector production. Resources will then flow out of the informal and into the formal sector where they can then be more easily taxed.

A second set of policies involves exchange rate regulation. Such regulations will likely overvalue the domestic currency-making production for export more difficult, but enhance the profitability of the production of import-substituting goods. This will provide additional protection to formal sector firms whose revenue can be more easily taxed. In addition, the state's ability to control foreign exchange allows for such resources to be funneled to formal sector firms who can then be taxed.

These policies can be combined with financial market regulations. These regulations allow the political elite to direct financial flows away from the informal sector and toward formal sector firms. Competition via direct foreign investment can also be limited through regulation of such flows.

Of course, resistance to paying tax by the formal sector firms may limit the extent to which the state can extract revenue. Indeed this has often led to expansion of state ownership of firms operating in the formal sector. By this means a proportion of the profits of state-owned firms can be directly extracted by the ruling political elite. Indeed, this was the case in China, where during the Mao and even post-Mao eras, state-owned firms made up a significant share of the output of the formal sector in certain industries. Interestingly, this was also the case in early Taiwanese development. Up until the early 1960s, most of Taiwan's larger firms were under state control. ‘The government's basic policy toward the ownership of industry was to retain direct control of utilities; natural monopoly industries that were considered vital to the country's defense (such as the railroad and highway buses); and industries held for revenue purposes . . .’ (Mao and Schive, Reference Mao, Schive and Mellor1995, p. 43).

One might wonder whether a land tax might serve as a viable alternative to the process of revenue extraction discussed above. Most current less-developed nations have found it very difficult, if not impossible, to establish such a tax. However, historically it has played a role. For example, in Taiwan, such a tax was imposed by the Japanese during the former's colonial period. At independence, the Taiwanese continued to use the tax. However, the bulk of tax revenue extracted from agriculture came through manipulation of the barter rate between rice and fertilizer and the purchase of rice at below market prices. The total hidden tax extracted (shifting resources from the rural sector to urban-based manufacturing) was greater than Taiwan's total income tax ‘in the entire period before 1963 and was more than twice the farm land tax before 1961, except in 1954’ (Mao and Schive, Reference Mao, Schive and Mellor1995, pp. 37–38).

The set of policies discussed above has often been labeled collectively as import substitution industrialization. The political foundations for this approach to economic development are quite straightforward. The political elite's coalition is made up mainly of these firms who are operating in the formal sector, in particular the large, capital-intensive firms. They are critical to providing the income stream necessary for the political survival of the governing elite. In return, the political elite engage in policies which protect and subsidize this sector.

The economic coalition supporting the political elite is unlikely to support policies which allocate resources to other sectors of the economy (agriculture, labor-intensive manufacturing, etc.). Such policies would be opposed for two reasons. Such a change in policy would threaten the formal sector firms with economic losses. In addition, their relative political power would also suffer and it is the political power that is crucial in maintaining or unifying the political coalition. Even when reforms are forced upon the ruling elite by external forces (World Bank, IMF, etc.), such reforms are often not strongly enforced; or the effects of such reforms are often offset by other activities, often involving significant amounts of corruption. For example, privatizing state-owned firms has often been a policy promoted by agencies interested in reform, but often privatization has only meant the transfer of assets at bargain basement prices to members of the ruling political coalition. Reductions in tariffs have often taken place, but often the effects of such reductions have been limited by rules and regulations imposed on imports of specific goods. The reforms are often half-hearted because the structure of the ruling coalition has not changed and thus the transfer of resources to coalition members often continues even after reform.

The economic impact of such economic policies based on the formal sector coalition varies through time. Initially rapid growth is achieved as the resources are transferred into formal sector production. As production in this sector expands, it displaces imported products and thus there is room for rapid expansion. However, once the domestic market is saturated, then growth slows. Further rapid growth depends upon the ability of formal sector firms to be able to compete in international markets. Given that these formal sector firms are large and capital intensive (both in physical and human capital terms), it is very difficult for these firms to develop comparative advantages in these sorts of goods. For rapid export growth, it is necessary that there be an economic restructuring with resources flowing into other, more labor-intensive sectors (agricultural, manufacturing, and services) which by their very nature make it easier to develop comparative advantages (given the labor-abundant nature of most developing countries).

However, for this economic restructuring to occur, there must be a political restructuring as well. That is, the governing political elite must restructure the political coalition to include the newly arising and evolving firms or, if they are already included, to enhance their relative power in the coalition. The restructuring of the political coalition will then allow for a restructuring of the economy, with resources flowing into those sectors that are likely to generate rapid export growth. But, here is the problem. The powerful capital-intensive producers in the existing coalition will vigorously resist any political restructuring which would result in economic restructuring. Initially, the firms in the new economic sectors that would benefit from economic restructuring will be small in both their economic and political influence. Thus, such a political and economic restructuring is likely to be stymied by the power of large, capital-intensive firms in the current political coalition.

If the governing elite have long time horizons, then the economy could be restructured and as the economic activity shifts, the economic influence of traditional members of the coalition would decline relative to the new, export-oriented firms, and political restructuring could occur later. However, the time horizon for most political elite is quite short (discount rate is high), thus they cannot withstand the political opposition to economic restructuring.

So, how is this dilemma to be overcome? Obviously a number of nations have been able to restructure their economies and thus follow a rapid growth path (China, Vietnam, Taiwan). To understand this process, it is necessary to introduce the concept of political legitimacy. Institutions are conceived as the rules, both written and unwritten, which govern social, economic, and political interaction (North, Reference North1990). Why do people obey certain sets of rules (explicit and/or implicit)? One obvious reason is that they fear punishment. Most institutional structures come with an enforcement mechanism. The punishment can come in a variety of forms. One may lose access to certain economic or business opportunities or lose economic assets. A powerful force is the loss of social and/or political status within the community. However, the enforcement of sets of rules via detection and punishment is a very expensive process in terms of the usage of society's resources and therefore the extent to which conformity can be assured is quite limited. If this is the only reason that people obey rules then one would expect greater violation the further away a community is from the center of political power. Thus, political legitimacy would likely fade the further from centralized authority one moves.

There is a second reason that people obey the rules and that stems from the fact that they view them as being legitimate. By legitimate, I mean that people obey the rules because they think the rules are appropriate or acceptable or just. Thus, they will obey or submit to those institutional structures whether or not the enforcement mechanisms are effective or not. One can extend this notion to economic policies as well. Economic policies are proposals for instituting new rules as to how resources within the economy are to be allocated or reallocated. People will follow these policies because they fear punishment or they may go along with these policies because they see them as being legitimate, acceptable, and appropriate.

How is legitimacy constructed? Moore (Reference Moore2001) argues that the process of reciprocity is an important factor in this process. Economic development requires sacrifice (savings) from certain sectors of the economy to allow for the reallocation of resources. Since most developing countries are initially dominated by agriculture, it is this sector that becomes the aim of economic policies which seek to reallocate resources. The pure extraction of resources alone from this sector does not result in rural producers viewing these policies as being legitimate. Instead, most producers will view such policies and the state that promotes them as being illegitimate. In fact, the typical set of policies followed by the political elite so as to generate adequate revenue flows which were discussed earlier in this section are generally discriminatory with respect to agricultural production. For example, tariffs and other trade taxes end up extracting resources from the agricultural sector by increasing the price of purchased inputs, etc., for this sector and depressing the relative price of what this sector sells. Thus, the terms of trade are turned against agriculture. This is unlikely to result in rural producers seeing these policies as being legitimate.

Legitimacy comes from reciprocity. That is, resources are reallocated from agriculture, but in return services are provided to that sector as well. The net flow of resources may be out of that sector, but it comes as the result of a flow of services and resources into agriculture first with a following or resulting outflow that is larger. This can only come about if there is significant productivity growth in the rural sector as a result of the gross inflow of resources. Thus, rural producers find their productivity and incomes growing dramatically. Under these sorts of conditions, the rural producers are likely to view policies aimed at reallocation as being legitimate in nature. It is the reciprocity that is the key.

The inflow of resources into agriculture must create a technologically driven growth in productivity since this is the only way that reallocation can occur while also raising the rural living standard. The import substitution policies followed in the past by the political elite (aimed at enriching its supporting political coalition) actually create a technological backlog in the agricultural sector. Because these policies extract resources from agriculture without providing services in return, the ability and willingness of farmers to invest in new technology is dramatically reduced. In turn, these policies slow the structural change that is supposed to accompany economic development. That is, economic development is usually accompanied by a reduction in the proportion of the workforce employed in agriculture relative to manufacturing. However, Dennis and İşcan (Reference Dennis and İşcan2011) find that ‘policies that have discriminated against agriculture have likely reduced the very surpluses they wished to mobilize, and may even have significantly retarded economic growth and the pace of structural change’ (p. 902).

While the application of new technology in agriculture is slowed as a result of these policies, the research into the development of new agricultural technologies has continued to occur (the technological frontier has continued to shift outward). As a result, a significant backlog of agricultural technology is likely to have occurred as the result of policies aimed at discriminating against the agricultural sector. Rakotoarisoa (Reference Rakotoarisoa2011) used a panel data set of 33 rice producing countries between 1961 and 2002 to determine whether policy distortions have led to a rice productivity gap between production in high- and low-income countries. The author finds that poor countries’ rice productivity (where agriculture is discriminated against) has fallen further behind that in richer countries (a productivity gap). Thus, if poorer countries reduced the discrimination against rice producers, dramatic increases in productivity could be obtained.

One might argue that such gaps exist not just in agriculture, but also in manufacturing in poor countries. However, the technological gap in agriculture is much easier to close than that in manufacturing. In poor countries that are predominately agricultural, farmers are quite skilled in agricultural production and understand how local markets for inputs and outputs works. Thus, if an appropriate (to the local ecological conditions) technology is made available, producers will be able to adapt to new production conditions (especially if policies discriminating against agriculture are weakened). This is not so with manufacturing. Managing the development and application of new technologies in this sector is much more difficult to achieve. In addition, it is policy which has created the barrier to new technology in agriculture, thus removing the barrier will dramatically enhance productivity.

If such a productivity gap exists, then it provides a mechanism whereby the political elite can achieve a restructuring of their political coalition. A dramatic increase in agricultural productivity will create an opportunity through which the rural producers will begin to view the policies and institution of the ruling political elite as being legitimate. Yes, resources will flow out of the agricultural sector so as to provide for the expansion of alternative sectors, but this will occur within an environment in which agricultural productivity is growing rapidly. This reciprocal relationship leads to the rural population viewing government policy as being politically legitimate and this greatly strengthens the political position of the political elite. Having the support of the rural population allows the latter to restructure the ruling coalition by reducing the power and influence of the large, capital-intensive firms in the formal sector. New partners can be incorporated into the governing coalition and given greater influence over political decision-making. This would include new firms which are likely to be labor intensive in nature and ultimately export oriented. The formal sector is expanded to incorporate these new interests as the quantity and qualities of services provided by the state increase and improve.

Policy in this case can be much more flexible in nature, thus leading to a restructuring of the economy based upon a restructuring of the political coalition upon which the political elite are based. The reforms are thus likely to be much more effective since the economic structure of the economy will come to reflect the political structure of the ruling coalition. The political elite are able to achieve this by using the political weight of the rural population to leverage changes in the ruling coalition.

It should be emphasized that the analysis presented above is not based on the idea that the rural population will be incorporated into the ruling coalition. The political elite seek to keep the ruling coalition as narrow as possible, thus reducing the number of threats to their continued dominance of the political structure. Thus, political incorporation of the rural sector is not likely to occur. Instead, the political weight of the rural sectors’ support is utilized by the political elite to alter the structure of the political coalition. It is the threat of potential unrest that strengthens the governing elite.

The analysis in this sector is thought to apply to both democratic and autocratic political systems. Under both sorts of systems, the political elite are unable to govern by themselves, they need to construct political coalitions to provide a political foundation for their regime. ‘Like democratically elected rulers, authoritarian rulers are sensitive to the preferences of those whose loyalty or acquiescence is needed to maintain power, and they act strategically to build a supporting coalition’ (Gallagher and Hanson, Reference Gallagher and Hanson2009). According to Gallagher and Hanson (Reference Gallagher and Hanson2009), one of the significant differences between these two types of regimes is that dictators possess sticks as well as carrots in their arsenal of tools. They can provide rewards to followers, but they can also use political repression and violence to maintain their position.

But one should be careful here in making this distinction. Many democracies function within socio-political environments characterized by patron–client systems of politics. In these situations, political patrons utilize ‘carrots and sticks’ to mobilize political support during elections so as to maintain the position of the political elite. Thus, democracy, as practiced in much of the world, also represents a mixture of policy aimed at promoting the strength of the governing coalition as well as repression used to weaken opponents to the regime. The point of the argument here is that whether democratic or authoritarian in nature, it is the nature of the political coalition that determines the direction of economic policy.

A political elite whose coalition is dominated by a formal sector made up of large, capital-intensive firms will not have the autonomy to construct a growth coalition. However, if a technological backlog exists in the agricultural sector, then policies that allow this gap to be closed will lead to dramatic productivity growth in the rural sector, and this sector will regard the political elite as being legitimate. This in turn will provide the political power for the political elite to reconstruct the political coalition so as to support economic reform promoting the expansion of smaller, labor-intensive manufacturing and services. In those situations in which a technological backlog does not exist, it is much more difficult to construct a growth coalition.

Of course, all models have limitations and are valid only in certain conditions and situations. The analysis above presumes that the benefits of agricultural productivity growth are widespread throughout rural society. Thus, rapid agricultural growth in a situation of significant inequality of land ownership is unlikely to generate political legitimacy to as great an extent as in situations in which land ownership is more equitably distributed. The Philippines seems to be a good example of this. With much of the benefits of agricultural productivity growth going to a narrow minority it was very difficult for the ruling coalition in the Philippines to generate the political legitimacy necessary to shift the structure of the ruling coalition so as to provide the political foundation for export-oriented growth. A shift to such an economic strategy without generating political legitimacy would likely have disrupted the existing ruling coalition, thus reducing the time horizon of this group. This shorter time horizon makes economic reform very difficult. One could find numerous similar situations in much of Central America where coffee and other primary commodities serve as the mainstays for export revenue and GDP growth. This can be contrasted with the experience of Taiwan, China, and Vietnam. In these three countries, rapid agricultural growth benefited a significant proportion of the rural population. The political legitimacy generated allowed a restructuring of the ruling coalition that allowed the pursuit of a more export-oriented growth. The legitimacy gained allowed the time horizon of the restructured ruling coalition to lengthen (the experiences of these countries will be discussed in more detail in the next section of the paper).

In addition, those countries possessing strong comparative advantages in either commercialized agricultural exports or labor-intensive manufactured are also likely to evolve in different ways. In the former, a backlog of technology in agriculture may have the potential to lead to dramatic increases in exports (and thus revenue via export taxation) and may provide an incentive to the ruling elite to restructure the ruling coalition so as to include export producers and thus open the economy quickly. However, such a scenario is unlikely to result in rapid growth in manufacturing.

Alternatively, if a country's potential comparative advantage in labor-intensive manufacturing is strong enough, then the political elite may be able to quickly open the economy and generate rapid growth in labor-intensive exports, which would provide new opportunities for extracting revenue. This latter possibility, though, seems unlikely since some of the most labor-abundant countries have had trouble rapidly expanding labor-intensive exports without a prior period of learning often promoted by government policy (expenditures on infrastructure, subsidies, etc. all requiring new sources of revenue). For example, both Taiwan and South Korea (both labor-abundant countries) engaged in periods of import substitution industrialization prior to turning outward. In addition, as colonies of Japan prior to World War II both countries received significant investments aimed at creating firms whose production would substitute for imports.

3. Some examples

This section will illustrate the ideas developed above by referring to the experiences of Taiwan, China, and Vietnam. They have all experienced rapid economic growth in the recent past. In all three the agricultural sector dominated the economies at the time of significant economic reform. The growth processes of all three began with rapid agricultural growth. It will be argued that agriculture was not just economically critical, but was important in the politics of development.

Taiwan, as well as Korea, was a colony of Japan prior to World War II. Japan's objective was to use its colonies as a source of rice and other agricultural products (in particular sugar). Because Taiwan was a rice-based economy, Japan attempted to transfer much of its own agricultural technology to Taiwan in order to raise productivity there. Japan formally annexed Taiwan in 1895 and in 1897 began an extensive land survey to determine land values for both taxation purposes as well as to secure property rights in land. In addition, a system of research and experiment stations was established with the aim of adapting seed varieties that had been successfully utilized in Japan, to Taiwanese experience. In order to disseminate this new technology the colonial government created farmers’ associations, small agricultural units, and agricultural cooperatives. The cooperatives worked to supply credit to farmers while the associations served as coordinators and suppliers of information and inputs (Myers and Ching, Reference Myers and Ching1964). In addition, the Japanese also invested in infrastructure, such as roads, railroads, and harbors. As a result of these activities, an institutional foundation for a highly productive agricultural sector had been laid. Initially, this of course was aimed at benefiting the Japanese mainland, but with independence it provided a basis for rapid agricultural growth.

With the end of World War II, conflict broke out in mainland China between the communists and nationalists (KMT) in which the latter were defeated and retreated to Taiwan. A new government was established and much of the existing industrial structure (left by the Japanese) was taken over by KMT refugees. ‘During the initial period of state consolidation and industrial take-off, the KMT made great efforts to create and sustain this new transplanted class of urban industrialists, mostly of mainlander origin’ (Davis, Reference Davis2004, p. 215).

Alternatively, the indigenous Taiwanese dominated the agricultural sector. This sector was characterized by small, family operated farms which intensively farmed the land. The sector, if the reader will remember, had been the target of significant rural investment by the Japanese and thus had the infrastructure necessary to sustain rapid productivity growth.

Initially the KMT-led state followed an inward-oriented strategy in which tariffs and other policies were used to protect the large, capital-intensive firms of the formal sector which were controlled by mainland immigrants. However, the size of the domestic Taiwanese market was very limited. In addition, aid provided by the United States in the 1950s was being rapidly reduced. This seems to have been a critical juncture in the economic development of Taiwan. The ruling elite's political coalition was crucially dependent upon the large, capital-intensive firms in the formal sector. ‘There is little doubt that many of the KMT-affiliated industrialists were able to nurture their ISI (import substitution industrialization) manufacturing activities so successfully precisely because the Taiwanese state protected them and their firms’ (Davis, Reference Davis2004). However, the degree to which the policies would generate growth was limited. Somehow the KMT elite had to shift the composition of the ruling coalition so as to include labor-intensive, manufacturing firms operating on a smaller scale and perhaps owned and operated by indigenous Taiwanese. This would allow the KMT elite to shift the economic strategy toward a more outward orientation via a reallocation of resources.

The analysis presented in the previous section implies that if a technological backlog exists in the agricultural sector, then it is possible that the political elite can gain political legitimacy in the countryside by investing in the countryside so as to create a burst of agricultural growth. The Japanese had provided the necessary preconditions for such a burst in agricultural growth. In addition, significant land reform also played an important role. With the urging and support of the United States, the new Taiwanese state moved toward the administration of such a reform. However, one should not think of this as reform being forced upon the political elite by the United States. In 1947, there was a serious uprising by the indigenous Taiwanese seeking greater political autonomy (the February 28th Incident). Given the experience of the KMT political elite on the mainland of China, there was agreement that land reform must be carried out. This was the critical juncture at which agricultural reforms become feasible. As a result of the reform, according to Gold (Reference Gold1986), land cultivated by owner cultivators rose from 50.5% of the total in 1949 to 75% in 1953. Tenant cultivated land fell from 41% to 16% during the same period.

With land ownership also came significant investment by the Taiwanese state in the agricultural sector. These investments expanded and improved the system of irrigation and spurred the development and application of new seeds and the use of commercial fertilizers. Indeed, agricultural growth was quite rapid. In Taiwan, agricultural growth was 4.7% and 4.2% for the periods 1950 to 1959 and 1959 to 1969 (Hayami et al., Reference Hayami, Ruttan and Southworth1979). It should be pointed out that as a result of this rapid productivity growth, significant amounts of resource were transferred from agriculture via government policy. However, even with this outflow the standard of living in the rural areas improved dramatically. One should note that at no time did agricultural growth rates exceed those rates of growth in some sectors of manufacturing, even during the early period of import substitution. However, the improvement in rural standards of living earned legitimacy for the ruling elite, allowing for a restructuring of economic policy toward a more outward-oriented growth strategy.

As a result, the KMT elite-controlled state came to be viewed as politically legitimate by the rural population. The rural population was not directly incorporated into the ruling coalition, but the political elite used their legitimacy to leverage a restructuring of the ruling coalition. The power of mainland immigrants who controlled larger, more capital-intensive firms in the formal sector was reduced. Now policy could also be restructured to promote the growth of smaller, more labor-intensive, export-oriented firms.

The patterns of expansion in Taiwan seem to reflect the analysis presented. Import substitution was used to protect urban industrial production, shifting resources out of agriculture. From 1949 to 1954, this caused traditional industries such as textiles, paper and paper products, and leather products to expand by 54.9%, 11%, and 23%, respectively. At the same time, much more capital-intensive economic activities such as basic metals, metal products, and electrical machinery grew at 36.8%, 31.9%, and 49.8%. In addition, the effects of the policies discussed above increased the capital intensity of all of these lines of production. However, by 1955–1962 growth in the entire import substitution sector, including all of the industries discussed above, declined dramatically. This was the exhaustion of the domestic market that often occurred with import substitution. Thus, the growth in potential tax revenues from the protected sectors of manufacturing slowed dramatically. With the turn to outward-oriented growth, some of the older traditional industries performed well (textiles), but new industries also experienced export success (electrical machinery, appliances and supplies, etc.; Ho, Reference Ho1978). Thus, potential tax revenue from these outward-oriented firms began to grow rapidly.

However, the shift in policy did not occur without a struggle. A major debate within the ruling coalition took place from 1954 to 1958 between those who wished to continue the import substitution approach and those who wanted to promote export expansion. The latter was argued by K. Y. Yin who was the Minister of Economic Affairs and the former was argued by P. Y. Hsu who was the Minister of Finance. Hsu believed that reform would threaten the main source of government revenue (trade taxes) and create economic instability. Eventually, the economic reformers were able to achieve a reshuffling of the cabinet in which Hsu was replaced as Finance Minister in 1958. It is hard to contemplate this occurring if the agricultural sector had been stagnant and land reform had not occurred (Kuo and Myers, Reference Kuo and Myers2012).

It appears that a similar process unfolded in the People's Republic of China. After World War II, communist rule was established throughout the Chinese mainland. Private ownership of the means of production was replaced by state ownership and control. All industries became state owned and all land, after initially being redistributed to farmers, became state controlled via communal or collective farms. The new political elite faced the same problem that ruling elites everywhere face. They needed revenue and initially lacked the means to collect it. Initially the bulk of the population and production occurred in the agricultural sector. The ruling elite combined the collectivization of agricultural production with state monopsonization of the purchase of agricultural output.

Private marketing of agricultural output was forbidden and collective farms were required to supply quotas to the state at artificially low prices. The states were also the monopolistic suppliers of critical inputs to agricultural production and then raised these prices (Lin and Liu, Reference Lin and Liu2007). Resources captured in this way by the state were used to keep the costs of production for large, urban, state-owned firms low. This greatly enhanced their profitability and the state was then able to extract revenue directly from these firms. The non-tax revenue from these firms accounted for 57% of total government revenue in 1970 (Lin, Reference Lin2009). This was the mechanism by which resources were transferred from the difficult-to-tax agricultural sector to the easier-to-tax large state-owned firms. In addition, trade with the rest of the world was highly restricted and this reduced or eliminated competition, further enhancing the profitability of state-owned firms. Again, allowing for additional revenue to be extracted by the state.

These kinds of policies greatly inhibited the growth process in the rural areas. Specifically, farm gross output grew at an annual rate of a little more than 2% from 1957 to 1978, barely keeping up with population growth. The driving force in this growth was increased input usage, not productivity. This did not mean that technical change or the potential for rapid technological change was not increasing over time. Irrigated areas grew at 2.41% per year and this is very important since irrigated land is the key to the application of modern agricultural technology (Maddison, Reference Maddison2007). Research in agriculture continued and semi-dwarf seed varieties were developed in 1964 in China before rice varieties of the International Rice Research Institute were released in the rest of Asia. However, the incentive system created by collectivized agriculture and the discriminatory policies aimed at agriculture limited the spread and usage of new technology. ‘But the weak link between effort and remuneration on the Chinese collective, and the growing disaffection of the population as a result of both the famines of the early 1960s and the repeated campaigns of the subsequent decades probably ensured that the rural economy was operating inside its production possibility curve’ (Bramall, Reference Bramall2000, p. 249).

The ruling political coalition at this time was dominated by the bureaucracy and leaders of the state-owned, urban-based firms. They benefited by the transferring of resources from agriculture and in return provided the revenue for the political elite to maintain power. However, this internal-oriented system of economic organization stifled the development of the Chinese economy. Effective economic reform was critically dependent on altering the distribution of influence and the composition of the ruling coalition. The death of Mao provided an opportunity for such reform and the technological backlog in the agricultural sector provided the means.

Reform began surreptitiously in the rural areas. Some villages began to experiment with relaxing many of the constraints associated with collective farming. Families were allowed to utilize plots of land for their own use, deciding what to produce and how and were allowed to keep a substantial portion of the production. This was the beginning of the household responsibility system. In this process, collective/communal farms were dismantled and production brigades and teams were replaced by township and village governance structures. Land was allocated to individual households for extended leases. Each household had to first meet the production quotas required by the state at prices set by the state. However, residual output was controlled by the household which could consume it or sell it at local and regional markets which began to develop. Also, prices offered for the quota output were dramatically increased by the state (Bramall, Reference Bramall2000).

It should be emphasized here that these reforms were not initiated by the central state. However, Deng Xiaoping was astute enough to allow experimentation and then to spur the spread of those experiments that seemed to be economically successful. The impact was dramatic and significant. The total agricultural growth and average annual rate of growth from 1978 to 1984 were 42.23% and 6.05%, respectively (Lin et al., Reference Lin, Cai and Li2003). From 1978 to 1987, labor productivity rose by 5% annually and total factor productivity grew at 4.6% per year (Maddison, Reference Maddison2007). This rapid growth resulted in rapid income growth for farm families. From 1980 to 1984, rural per capita income in real terms rose by 14.5% per year. Thus, there was a sudden burst of economic growth that occurred in the rural areas.

The political elite in China led by Deng sought to use this newfound economic growth to leverage changes in the ruling coalition. Full understanding of this process is not possible given the secrecy within which political and economic reforms were shrouded. However, it seems that the political elite did not attempt a frontal assault on the foundations of socialism and, in particular, on the political coalition upon which it was built. Instead, reform leaders shifted the composition of the ruling coalition to reflect those who gained as a result of the new reforms (Hanson and Gallagher, Reference Hanson and Gallagher2012).

Rural growth dramatically increased the demand for labor-intensive, simple manufacturing goods. Township and village enterprises expanded production in response to the dramatic growth in demand. These enterprises were not privately owned, neither were they owned by the central state. Instead, they were controlled by local levels of government and paid taxes to the local, government jurisdiction. The central state and the political elite used economic decentralization initiatives to create new political coalition partners among provincial officials. These lower level officials provided the base from which the influence of conservative coalition members in the center could be offset (Shink, Reference Shink1993). A major part of this process involved a major fiscal reform in 1980. The basic thrust of the reform was that a lower level regional government contracts with the upper levels of government concerning the amount (or share) of tax and profit revenue to be remitted for the next several years and the lower level of government was allowed to keep the rest (Montinola et al., Reference Montinola, Qian and Weingast1995).

It was the last part of this agreement that was so important. If the lower level government should earn revenue in excess of the contracted amount, it could keep it. This provided a strong tie between the resources available to the local government unit and local economic prosperity. Local officials will have an incentive to foster local prosperity. This created a virtuous circle where economic reforms generated local and regional political supporters which allowed additional reforms to be carried out. Thus, the political influence of the bureaucratic structure controlling the large urban-based state-owned firms declined.

The ability of the political elite to restructure the governing political coalition and thus restructure the economy was dependent upon the political legitimacy with which the rural sectors viewed policymaking by the state. It was the dramatic agricultural growth that provided this legitimacy. If the initial reforms in the rural sector had failed to dramatically increase productivity, then there would not have been a political foundation upon which to restructure the economy. For example, if there had been no technological backlog in agriculture, the initial rural reforms would not have dramatically altered life in the countryside. Without the political legitimacy generated by rapid growth, how would a restructuring of the political coalition have taken place?

The final example of the role of the rural sector in generating political legitimacy is Vietnam. From independence from France until the mid-1970s, North and South Vietnam followed very different economic, social, and political paths. The North was a socialist state, in which firms in urban areas were state owned and agriculture was collectivized with land controlled by the state. The South's economy was market based with private ownership of both capital and land. With the withdrawal of the United States, South Vietnam quickly fell under the control of the North and the country was reunified. In the South this meant that firms now were owned by the state and the agricultural sector was collectivized. The end result of this process was that the economy of Vietnam did not function very well. In 1950, per capita income in Vietnam was 150% of that in China, 78% of that in Indonesia, 85% of that in South Korea, and 80% of that in Thailand. By 1980, these percentages had become respectively 71%, 40%, 18%, and 30% (Klump, Reference Klump, Besley and Cord2007). By the early 1980s, inflation was becoming a major problem with budget and trade deficits increasing in intensity. The bulk of the population was classified as being poor. The urban sector was dominated by state-owned firms whose bureaucracies and management formed a powerful part of the ruling political coalition.

This was a critical juncture similar to that of China and even, to some extent, Taiwan. The political elite needed revenue and resources to survive. The easy-to-tax sector was that characterized by urban, capital-intensive firms owned or controlled by the state. The difficult-to-tax sector was the rural area. Thus, policy was aimed at transferring resources and revenue from the hard-to-tax sector to the easy-to-tax sector and then directed into the hands of the political elite. The key to restructuring the economies of these countries was dependent on restructuring the political coalition. It has been argued that for China and Taiwan the key to this involved generating political legitimacy in the rural areas via rapid agricultural growth (technological backlog). This was also the case in Vietnam. It should be pointed out that the demonstration effect of Chinese experience may have been important in the first steps toward rural reform.

The work by Athukorala et al. (Reference Athukorala, Huong, Thanh, Anderson and Martin2007) provides empirical support for the idea that in the 1980s the policy regime in Vietnam was characterized by significant bias against the agricultural sector. Pricing policies, tariffs, and exchange rate policies were used to transfer resources out of agriculture. This discrimination against agriculture also dramatically reduced the gains to farmers of innovation. During the 1960s, 1970s, and 1980s, there was little growth in yields and no significant expansion in area planted (Ut and Kajsa, Reference Ut and Kajsa2006).

This did not mean that technical change was not occurring. In the mid-to-late 1970s, new higher yielding varieties were being developed that were very responsive to irrigation and the application of commercial fertilizers as well as being resistant to pests and diseases. However, these technological advances were having little impact on actual innovation since anti-agricultural policies limited the actual application of the technology.

The above resulted in the development of a technology backlog in agriculture. In the late 1980s and the early 1990s, a period of reform known as Doi Moi (renovation) began. The first of these reforms was directed at the agricultural sector. Price controls were gradually removed and farm households were given control over farming operations via control over the land (collectives were slowly dismantled). Farm households were allowed to sell surplus production in private markets and state quotas were reduced (Glewwe, Reference Glewwe, Glewwe, Agrawal and Dollar2004). Linh (Reference Linh2009) used Malmquist productivity indexes to analyze total factor productivity growth (technical change) in 60 Vietnamese provinces over the period 1985 to 2000. He found that total factor productivity grew on average 3.37% per year. Thus, a significant backlog of technology was successfully applied during this time period.

The dramatic growth in the agricultural sector provided the ruling political elite with the legitimacy they needed to alter the political coalition which dominated politics in Vietnam. Malesky argues that the Central Committee of the Vietnamese Communist Party (CCOM) was fundamentally divided into two groups. There was a conservative coalition made up of managers of state-owned enterprises, representatives of the military, and representatives of the provinces where a substantial amount of the provinces output was produced in state-owned enterprises. The second group was made up of reformers composed of officials and representatives of the provinces deriving little benefit from state-owned enterprises (SOE). The rapid burst of agricultural growth (stemming from the closing of the technological gap in agriculture) strengthened the position of the latter group. They used this influence to redraw the boundaries of existing provinces and create new provinces which were not dominated by state-owned firms. ‘Because the non-state provinces were less dependent on SOE for revenue, they were more willing to vote for progress on economic reform’ (Malesky, Reference Malesky2009, p. 134).

Malesky (Reference Malesky2009) estimates that, in 1991, the conservative coalition made up about 58% of the Central Committee. However, over time, the non-state-denominated provinces increased the number of seats held from 20% to 33% and the reformer's coalition had by 2006 control over 52% of the seats.

The initial reforms in agriculture were followed by a series of other reforms in macroeconomics and foreign trade. Central government spending was reduced and tax reform was undertaken. In the late 1980s, this involved actually establishing a tax system involving a variety of taxes. In the 1990s, a second stage of tax reform was undertaken which established a value-added tax, reformed the income tax, and sought reductions in trade taxes (Quang and Dung, Reference Quang and Dung1997). In terms of trade issues, the exchange rate was unified and a significant devaluation was undertaken. Barriers to exports and imports were lifted and foreign investment faced a much more hospitable environment (Glewwe, Reference Glewwe, Glewwe, Agrawal and Dollar2004).

The rapid growth that ensued was initially dominated by rapid growth in the agricultural sector. However, by the late 1990s growth in industry became the prominent driver of rapid economic growth (Glewwe, Reference Glewwe, Glewwe, Agrawal and Dollar2004). The rapid growth was accompanied by rapid reductions in poverty. The poverty rate declined from 75% of the population in 1984 to 58% in 1993 to 37% in 1998. Thus, the growth that occurred had a dramatic impact on those who were initially poor when the reforms began (Glewwe, Reference Glewwe, Glewwe, Agrawal and Dollar2004).

Taking stock of the Vietnamese experience, it seems similar to that of Taiwan and China. A technology backlog in agriculture allowed a burst of agricultural growth to drive more rapid overall growth. It was the legitimacy that this earned for the reforming political elite that allowed them to restructure the ruling political coalition so as to provide the basis for further reform. If the technological gap had not existed, then the initial rural reforms would not have led to rapid growth and it seems highly unlikely that the reforming political elite could have engineered any additional reform. So the sequence of events in all three countries was rural reform, agricultural productivity growth, restructuring of the ruling coalition, then a shift to export-oriented growth.

4. Summary and conclusion

Many scholars have argued that agriculture plays a number of key roles in the process of economic development. Generally, it has been argued that agriculture provides the food to feed the urban sector, the savings to finance the expansion of manufacturing, the demand for the output of the expanding manufacturing sector, the labor to work in the new factories, and the exports necessary to pay for critical imports (Johnston and Mellor, Reference Johnston and Mellor1961). This paper too has argued that agriculture, under certain conditions, can play a critical role in the process of economic development.

All ruling political elites create ruling coalitions which allow them to govern. In order to construct and maintain their political power, they need revenue and resources. This poses a particular problem since major parts of many developing countries’ economies are very difficult to tax. Thus, economic policies are generally aimed at transferring revenue and taxes from hard-to-tax to easy-to-tax sectors where revenue can be extracted. The easier-to-tax sector is usually represented by urban-based, capital-intensive firms in the formal sector whereas firms in the informal sector (including large parts of the rural sector) are much more difficult to tax. Thus, the ruling coalition is likely to be dominated by large formal sector firms and they are likely to oppose economic reform aimed at restructuring the economy. Thus, economic reform will not likely occur without a restructuring of the ruling political coalitions.

Such a restructuring is not likely to occur without the ruling political elite being able to generate political legitimacy. In a country dominated by agriculture, legitimacy can be generated by generating rapid growth in the agricultural sector and this is likely if a technological gap has developed in agriculture. Gaining legitimacy allows the political elite to leverage a change in the ruling coalition and this provides the basis for further economic reform. The experiences of Taiwan, China, and Vietnam were used to illustrate this process.