PART II
The decline and recovery of Thai agriculture: causes, responses, prospects and challenges (Continue)

6. Future prospects for Thai agriculture: challenges and constraints

In the past, Thai agriculture has satisfactorily adjusted its production structure in response to the problems of cost-price squeeze. But the world trading environment is changing rapidly. If agriculture is to successfully acquire new comparative advantage, it has to address some major challenges and constraints.

6.1 Future prospects and challenges

The government and most experts on Thai agriculture long ago reached the conclusion that Thailand is rapidly losing its comparative advantage in export of traditional commodities, particularly rice, rubber and cassava. Their diagnosis is based upon Thailand's factor endowments compared to other low-cost competitors. Despite the temporary increase in the export competitiveness due to currency depreciation, the costs of agricultural production have increased quickly, particularly labour and water costs (see discussion below).

Yet the performance of these traditional exports has defied all predictions. Thailand has remained the world's top exporter of rice, rubber, sugar and cassava, with increasing quantities of the first two being exported (Table 6.1).

Table 16. Major exporters of selected agricultural products in 2003

Source: FAO, Key statistics of food and agriculture external trade, and WTO, PC-TAS.

After the AI outbreaks and the consequent ban on imports of fresh Thai chicken, the major exporters stepped up their investment in production of cooked chicken (Poapongsakorn 2004). Yet volumes exported are small by comparison with lost markets for fresh meat. The future for chicken exports looks bleak in the medium term considering the difficulties of gaining effective control over AI, let alone eradication of the virus.

Thailand was knocked down from its position of largest exporter of shrimp a few years ago and its market share has continued to decline. Domestically, intensive shrimp culture has seriously damaged valuable but fragile mangrove areas. Commercial fishing in the mining manner has also affected the population of tiger prawns which are needed for commercial breeding. Meanwhile, countries with still abundant natural resources and cheaper labour, especially Viet Nam and Indonesia, have entered the market.

As exports of tapioca pellets to the European Union market have declined, exporters have diversified towards higher-value modified starch, which has wide ranges of industrial uses. Yet the ratio of high-value starch to pellet and chip production increased only slowly, from 4 percent in the 1980s to 9 percent in 2000. Although the recent increase in the oil price and government policy to promote gasohol consumption will provide another alternative for the cassava farmers,³⁰ an oil price decline in the future will affect the profitability of ethanol production.

³⁰ The government will replace 95 octane premium petrol by gasohol in 2006 and the use of MTBE in the production of 90 petrol will be banned in 2008 (Poapongsakorn et al. 2005).

Thailand also lost its position as the largest exporter of canned pineapple in the late 1990s after the United States imposed antidumping duties on imports from Thailand. The USTR also charged countervailing duty on Thai canned pineapple for a brief period in the past.

If recent trends are a guide, one can conjecture that Thailand will have comparative advantage in safe and high value-added foods, particularly processed foods, and land-intensive products (see discussion below). In addition, Thailand faces other challenges, such as how to promote professional farming and how to cope with a more liberalized trading environment.

The shift from traditional to high quality and safe products

Thailand first developed the new comparative advantage in quality and safe agricultural products for exports and, later, for the domestic market. Contract farming arrangements and new technology have enabled agribusiness exporters to exploit increasing world market demand for quality chicken meat and shrimp, particularly the demand for food with strict maximum residue level (MRL) requirements. As exports account for a large share of production, products sold in the domestic market have the same quality and safety standards as for exports.

This change has not yet happened in the pork market where production is mainly for local consumption. In recent years, as Thai consumers have become more concerned about food safety, the Department of Livestock Development has begun to monitor the use of drugs in livestock farms. The Department of Food and Drugs also regularly conducts tests for chemical residue in meat and other food sold in wet markets. Yet the development of safe meat standards has progressed at a much slower pace than for chicken and shrimp. A few supermarkets have begun to enforce strict safety standards for meat procured by their suppliers, but there are only a handful of modern slaughterhouses that have strict inspection. The main obstacle to the development of safe meat is the slaughterhouse law that still does not allow non-export private companies to operate and own their slaughterhouses. Private operators have to transfer the property rights in the slaughterhouse to the local government. These slaughterhouses do not have hygienic slaughtering processes and do not pay full slaughtering tax, giving them cost advantages that are a crucial barrier to entry of modern slaughterhouses.

The development of quality and safety standards for fresh fruit and vegetables (FFV) was first initiated by the FFV exporters who introduced new technology and cultivation practices to their contract farmers. As a result, FFVs are among Thailand's top 18 agricultural exports. Until recently, domestic consumers did not benefit from these standards, but now every supermarket in Bangkok and large cities has high quality FFV on their shelves. However, our survey found that some suppliers mix “regular” vegetables procured from the wholesale markets with the supplies of safe vegetables from their contract farms. As consumers will not be able to identify the true quality even after consumption, it is likely that consumers are paying prices that are too high. Thailand may not be able to build up core competency in the production of safe foods unless more credible institutions for safety regulation can be established.

Towards a more land-intensive, less water-intensive cropping pattern

The prediction that Thai agriculture will have comparative advantage in land-intensive products (e.g. perennial trees) is based upon the fact that the declining agricultural workforce has caused agricultural land per capita to increase in recent years (Figure 6). It should also be noted that Thai farmers tend to have more land to work with than their counterparts in other Asian countries (see Kaosa-ard and Siamwalla 1998).

Yet land consolidation proceeds at a snail's pace. Government land policy which will determine the future farm size, may be either efficiency-oriented or equity-oriented, in the latter case maintaining small farm size as a safety net. If the former objective is chosen, some legal constraints will have to be relaxed, e.g. on tenancy regulations that are biased against landowners and on limitation of rights to transfer land obtained under land reform programmes.

Increasing water scarcity will also influence the choice of crops. Thai farmers will substitute less water-intensive crops (such as high-valued chilli, fruit trees, etc.) for paddy. An appropriately instituted water management policy will accelerate such a restructuring process.

Professional and older farmers

We have argued that with agricultural employment declining as a result of the departure of young people, farming in future will be dominated by young professional farmers who have larger plots of land, a more pronounced commercial attitude and an innovative approach to production methods. A collection of studies by Molle and Thippawal (2003) describing the changing practices of professional farmers, examines introduction of new varieties, organic products for export, production based on the King's new theory and the integrated farming approach. The challenge lies in the role of the government in providing the necessary research and policy framework to accelerate the development of professional farming through, for example, more appropriate extension institutions and certification of safe and organic food. Such public good provision will reduce barriers facing young farmers' entry into modern commercial farm activities as well as the research costs of existing farmers.

As young people leave farming and their parents are ageing, there is also a concern about the future of older farmers. In the near future the government and the community will need to establish care institutions for the aged to complement the eroding system of family security.

Trade challenges

Thailand has not only been an active member of the G20 in pushing for more liberalized global trading rules in the ongoing Doha Round negotiation, it has also negotiated or is negotiating a dozen bilateral FTAs.

Studies show that Thailand will benefit from liberalization of agricultural trade, at both the multilateral (Poapongsakorn et al. 2005) and the bilateral level with Japan (Poapongsakorn 2003a), United States (Poapongsakorn 2003b), China (Sabhasri 2004) and Australia. These studies show that most of the economic benefits will come from market access, followed by the reduction of domestic support and the elimination of export subsidy. Farm households will enjoy higher income gains than non-farm households. But surprisingly, the poor (except the poorest) and the richest farm households will gain more from multilateral liberalization than the middle-income farm households (Poapongsakorn et al. 2005). Market access will be the most beneficial outcome for two main reasons. First, the trade pattern between Thailand and its bilateral partners, China, Japan, the United States and Australia, is complementary because Thailand exports tropical products and imports temperate agricultural products. Second, Thailand still has relatively high agricultural tariffs, hence tariff reduction will give Thai consumers greater choice and lower prices.

Yet there are at least three challenges that Thailand faces in ensuring that full benefits can be captured and those who are negatively affected by the FTAs are able to adjust their production system. The first challenge is to change existing policies that are in contradiction to the agreements. For example the government will have to switch from the current price-support policies towards less market-distorting support measures in the WTO trade negotiation. Market-distorting support will definitely have to be reduced substantially in the current Doha Round.

The second challenge is in encouraging farmers who grow import-competing and temperate zone products to restructure their production. A study on the Thailand-Australia FTA has proposed concrete measures to help Thai dairy farmers to adjust before Australian milk imports will be allowed at zero tariff rates in the next eight years. So far, the government has not provided any promised restructuring budget.

The third challenge is that there are both domestic constraints affecting the supplies of agricultural products and export constraints. Although tariffs on many agricultural products will come down substantially under the FTAs, many forms of non-tariff barriers remain. For example, the Australian Government maintains unrealistic regulations on imported Thai chicken; it must be cooked at a minimum of 160- Celsius (Poapongsakorn et al. 2005). Different regions in China maintain different regulations on imported goods. Moreover, a Thailand-China agreement still exists that allows imports of Thai fruits from only five certified farms. These issues have to be negotiated if Thailand wants to increase its exports.

Domestically, there are also major challenges. For example, an effective programme to eradicate AI will be needed before fresh chicken meat exports can resume.

6.2 Policy implications and constraints

Even without government support, there is an ongoing market-driven process of structural adjustment in response to the cost-price squeeze. But there is also increasing evidence of obstacles and market failure that have seriously affected the restructuring efforts of the farmers and the private sector. Overcoming these obstacles and solving the market failure are major challenges that will determine the future of Thai agriculture. Among the long list of problems, the major issues are research and knowledge, water scarcity, environmental degradation, formation of farmer groups to promote new marketing systems and, last but not least, shifting government policy away from the current market-distorting tendency.

Research and extension

Since R&D is carried out by business, farmers and government, a central policy issue is the delineation of work to be done by each of the three parties. There is an urgent need to reverse the declining public investment in agricultural research because the rate of return to public research in agriculture is very high, at 42 percent (Setboonsarng and Evenson 1991). The role of public research is in the area of genetic improvements, particularly at the biotechnology frontier. Private research and development has been in the area of mechanization. Although the investment is profitable, there are some market failures that call for better protection of intellectual property rights and tax incentives to promote more private research. The plant varieties protection law will also encourage both farmers and business to do complementary R&D for genetic improvement, provided that an appropriate regulatory framework can be established. Last but not least, if agriculture is to be sustainable, the government should provide an incentive system for farmers who have to adapt their own production to the specific resource constraints and opportunities within their own farm. For example, the government can facilitate dissemination of information on the successful activities of other farmers, including farmers in other countries, through financial support to the NGOs and community organizations which can most effectively implement information dissemination and extension services.

Water scarcity

Because of supply shortfalls and surging demand for water, competition for water resources will continue to escalate, resulting in more conflicts among users. As the government can no longer increase the supply of surface water by building more dams and its recent experience suggests that the supply-side approach of water resource management does not result in environmentally sustainable water use, the challenge now is how well the government can move from supply-side policies to demand management using market-based mechanisms that give incentives for more efficient use of water. Examples of measures that are needed include establishment of clear property rights, new institutional mechanisms for community management of water resources, river-basin water management based upon the participatory approach and improvement in government planning and management of water delivery systems.

Food safety

If Thailand is to become “the kitchen of the world” - one of three competitiveness strategies of the present government³¹ - a major stepping stone will be developing “a kitchen for the Thai”, according to an executive of a leading FFV exporter. As discussed in Part 5, many FFV suppliers in domestic supermarkets are currently supplying lemons because of the problem of asymmetric information between consumers and producers. Thailand does not yet have effective consumer-protection groups and an independent consumer-protection organization, as stipulated in the 1997 constitution, has not been established.

³¹ The other two strategies are “Detroit of Asia” for the automotive industry and “regional fashion centre” for the textile and clothing industries.

Another challenge is the incentive that farmers receive to produce safe FFV. Casual observation through interviews with farmers suggests that farmers producing safe FFV for local supermarkets receive only a few baht more than in the regular market. Yet to increase yield, they have to apply much more care than those who use chemical pesticides and fertilizer. Normally the yield of chemical-free production is much lower than that of regular production. Most of the economic rent from safe FFV goes to the brand owners and the supermarkets (which also have their own house brands). Part of the rent may go to the farmers as a reward for their skills and return to high quality lands. Otherwise the farmers will receive only normal profits. One challenge is to develop the reputation of these farmers as a group. In addition to increasing farmer bargaining power, such groups are in a better position to develop reputation (or brand name) and trust among consumers. These activities are a long-term collective investment in local public goods. But there may also be a free-rider problem. The main challenge is to find appropriate institutions and cooperation with agribusiness companies to strengthen farmer groups and provide farmers with adequate incentives to produce safe food.

Environmental degradation

The fishery sector has suffered negative growth rates of value added and TFP in recent years, mainly because the expansion of fishery activities, both marine fishery and shrimp cultivation, has been made possible by excessively exploiting open access resources. Overexploitation of common property resources not only affects sustainable growth, but also impoverishes thousands of small fisherfolk whose livelihoods critically depend on natural resources.

Thai agriculture has used increasing amounts of pesticide, insecticide and herbicide. Total chemical inputs more than doubled from 29 697 tonnes in 1993 to 63 881 tonnes in 2002. As a result, agriculture has become highly polluted. According to a survey by the Department of Pollution Control, it is the third largest source of water pollution, behind communities and factories.³² Although there is no information on soil pollution, anecdotal evidence suggests that excessive use of chemical inputs affects land quality and production yield. There are a few studies showing that Thai farmers overuse pesticide and insecticide (Jungbluth 1997; Poapongsakorn et al. 1998). The reasons for overutilization are: (1) the prices of insecticide and pesticide do not reflect the true social cost since imports are duty free and there is no pollution tax; (2) there is widespread uses of illegal drugs; (3) path dependence, that is, in the past one government agency trained farmers to use pesticide and insecticide in a cocktail dose. In addition, asymmetric information allows farmers to profit from the sale of unsafe products.

³² It is estimated that the wastewater discharged by communities contains 425 tonnes per day of organic matter, industrial wastewater adds another 97.9 tonnes and agriculture another 86 tonnes (Pollution Control Department 2005).

Unfortunately, there is neither adequate information nor studies on the effect of pollution on the food chain. There is only a little evidence that improper use of chemical inputs imposes unnecessarily high health costs on Thai farmers (Jungbluth 1997; Grandstaff 1992; data on occupational pesticide poisoning cases from the Department of Epidemiology). Some older studies found that fruit and vegetables tend to have residue problems. The Division of Toxic Substances found that 37 percent of vegetables were contaminated with organophosphorous insecticide residues, and 20 percent of kale and 10 percent of cowpea showed residues exceeding the MRL in 1994. One study also found that 73 percent of tangerine samples were contaminated with pesticide residues, while 10 percent of the sample had residues exceeding the MRL (Palakool et al. 1995). Casual evidence on environmental impact on the food chain is that most exporters of safe FFV have to carry out careful preliminary risk assessments of alternative locations for contract farming.

If Thailand is to achieve the goal of sustainable growth of agriculture and to be the world kitchen of safe foods, its main challenges are to institutionalize effective sustainable management of natural resources and to create good incentive systems as well as economic instruments that will lead to optimal uses of natural resources and chemical inputs. Coastal resources also need to be more effectively managed and monitored to reduce economic and social conflicts. One of the immediate needs is systematic data on the socio-economic impact of pollution.

Reducing market-distorting support measures

As per capita income has increased, the policy regime for Thai agriculture has shifted from one that penalized farmers to one that, as in developed countries, subsidizes them. Currently the major means of agricultural support in Thailand, measured in terms of budget allocation, is price support, which not only highly distorts resource allocation but will also negatively affect the competitiveness of agricultural exports in the future. As there is not much political will to reduce such price support, the only feasible alternative is that the Thai trade negotiators should support the United States and G11's proposal to sharply reduce highly distorted domestic support.

7. Lessons

As Thai agriculture has long been subject to a high degree of trade openness (trade accounted for 150 percent of agricultural GDP in 2003),³³ its production is strongly outward oriented. Except for the high tariff walls and increasing support, sound macroeconomic policy has allowed the market mechanism to perform its function with a relatively low degree of intervention. As a result, Thai farmers, who have been constantly subject to a high degree of world price volatility, have had to learn to adapt readily and flexibly to the changing environment. The Thai Government has facilitated their adjustments by providing the necessary enabling environment and public goods as well as tackling critical bottlenecks and market failure.

³³ The agriculture sector has a higher degree of trade openness than the whole economy, which has the trade/GDP ratio of 110 percent.

7.1 The dynamism of Thai agriculture and the role of prices

There is voluminous literature, including this study, which documents the dynamism and flexibility of Thai agriculture (e.g. Siamwalla et al. 1991; Poapongsakorn et al. 1995). Perhaps, the most important evidence is the shift resulting from the cost-price squeeze of the mid-1980s when the dominance of upland crops diminished in favour of higher value products. Field surveys have also shown that when farmers have to switch from one crop to another they will attempt to maintain as much as possible the flexibility to switch back and forth (Poapongsakorn et al. 1995).

The increase in consumer demand for healthy and safe food in recent years has been quickly responded to by farmers. The rapid increase in the supply of safe vegetables is amazing, given that five years ago one could hardly find a safe vegetable, even in the modern high-end supermarkets. The pace of farmers' response has also been evidenced by the rapid fall in the prices of some popular fruits just a few years after the surge in their demand, e.g. the Kieu Sawoei mango and Morn Thong durian in the 1990s (see the production expansion and price trend of durian in Poapongsakorn et al. 2005).

Absence of price distortion both allows farmers to reap benefits and compels adaptation to new competitive conditions. For example, the liberalization of rice export taxes, particularly the rice premium in 1986 has allowed farmers and exporters to reap the full benefits of world market prices. In response to the declining long-term trend of these prices, on the other hand, Thai rice exporters have successfully improved the quality of Thai rice.³⁴ Farmers in the Lower Central Plains grow five crops of rice in two years because the opportunity cost of their irrigated land is very high. To be more competitive in the world market, broiler growers have increased farm size to exploit economies of scale (Poapongsakorn et al. 2003).

³⁴ The f.o.b. prices of Thai rice are higher than those of Viet Nam.

Farmers have also actively adjusted their production processes in response to higher production costs. They have heavily invested in mechanization to cope with higher wage costs and in water pumps to overcome the scarcity of surface water by tapping underground water.

Last but not least, to cope with the volatility of agricultural prices and production, Thai farmers have not only diversified their farm output, but also engaged in more non-farm activities. In responding to agricultural risks, farmers have developed alternative modes of work: Many have become casual farmers, earning more cash income from non-farm activities, while others have become professional farmers. Some of the latter have entered contract farming; others have adopted integrated, self-sufficient farms operating on less commercial lines. The variety of farming systems implies that Thai agriculture is flexible and can readily adjust to external shocks, provided that the price system is permitted to play its role of information provision. One important problem, however, is outside the farmers' control and that is the declining trend of world agricultural prices.

7.2 Lessons from agricultural trade and macroeconomic policies

Despite increasing support policies that highly distort the agriculture sector, Thai governments - both past and present - have adopted a wide range of policies and institutions that improve the allocative functions of the market. These policies can be broadly classified into three sets, namely, trade and macroeconomic policies, policies of providing infrastructure and public goods and services, and policies to correct for market failure and bottlenecks. Here we discuss the first of these.

Since about 1980 agriculture has no longer been the leading sector. Its share in production and exports has been eclipsed by manufacturing. Despite a more favourable policy stance and the recent commodity boom, which helped to revive agricultural growth, the prospects for Thai agriculture may not be very promising.

From an economic point of view this is, of course, nothing to worry about. As a country's comparative advantage shifts, we would expect some sectors to increase their shares and others to decline. The key point is whether or not government policy settings allow relative domestic prices to reflect relative international prices. One of the most important turning points in the modern history of Thai agriculture is the shift in 1986 from trade and exchange rate policies that penalized farmers to a neutral policy stance. Since then macroeconomic policies have favoured the agriculture sector. Further, when Thailand began to lose its comparative advantage in agriculture, economic policies shifted from an agricultural to a manufacturing focus, or from low-end agriculture to high-value agriculture. Therefore, much economic policy, with the exception of the price-support policies, has facilitated the shifts in agricultural production systems.

Indeed, we wish to make the strong claim that in some sense, the “good” macroeconomic policies have been strong enough to overcome the “bad” sectoral policies and thus allowed agriculture to grow in accordance with the comparative advantage that the country had prior to 1980.

There are a number of reasons why good macroeconomic policies may be essential to the good performance of the Thai agriculture sector, some of which are specific to Thailand, while others may have broader relevance.

The feature which is probably unique to Thailand, compared to other Asian countries, is the situation of land surplus that prevailed in Thailand prior to 1980, and has been emerging rapidly since the late 1990s. In such a situation, farmers have far more freedom to choose which crops to grow, though this ability to choose leads to effective growth only if the price signals are not distorted. Given that exports provided the vent for surplus land, having a non-distortionary exchange rate regime is essential. The recent sharp increases in the price-support budget for some crops, may largely offset the positive effect of sound macroeconomic policy, but this issue needs more research.

For most countries, however, the central argument is that most agricultural products are, with very few exceptions, tradable products. While commodity-specific distortions will lead to a reallocation of resources among commodities, with probably little effect on agricultural output as a whole, exchange rates can affect agriculture, vis à vis other aggregative sectors. If we are interested therefore in agriculture sector performance, the macroeconomic regime which sustains a particular exchange rate is something we cannot afford to ignore.

7.3 Infrastructure and public goods

Like the East Asian governments, the Thai Government has always invested heavily in infrastructure. The rapid growth of agriculture through land expansion in 1960 to 1980 was made possible by the construction of highways, the rural road network and irrigation dams. Investment in education and research allowed the farmers to improve their productivity. In the 1980s and 1990s the government stepped up rural electrification in almost all villages (Poapongsakorn et al. 1995). In the late 1990s when Thai agricultural and food exports experienced SPS problems, the Thai Government quickly came to the rescue by providing laboratory tests and necessary services to improve the food safety of exports. In the case of shrimp exports, the government and the private sector have jointly invested in expensive equipment and laboratories needed to run tests on the MRL of nitrofuran. Moreover, the government has actively cooperated with the exporters in bringing cases of unfair trade practices by its trading partners to the WTO panels. Those cases include the United States' import ban on Thai shrimps, the European Union's export subsidy of sugar, etc.

Finally, the government has also actively gathered market intelligence for Thai agricultural exports. Such provision of public goods and services has greatly benefited both the exporters and farmers, large and small.

7.4 Market failure

In addition to public goods provision, the government has played an important role in tackling the problem of market failure. One of the most important market failures is the imperfect credit market. Unlike most of the agricultural banks which always experience serious problems of non-repayment, the BAAC introduced group guarantee as a means to lend short-term production credits to farmers. This method of loan guarantee has resulted in a high rate of debt repayment, which in turn, ensures the financial sustainability of the BAAC. As a result, the BAAC has become the most dependable institution for most farm households. In addition, the government's effort to issue land titles to farmers has enabled them to obtain more medium- and long-term credits to upgrade or restructure their production. In 1999, about 53 percent of the BAAC's credit disbursements were medium- and long-term credits.

The Bank of Thailand's backing credit for agricultural exports also helped to reduce exporters' risks and transaction costs in new markets.

There is a problem of asymmetric information in the market for safe food. Without government intervention, it is difficult for consumers to trust the products and thus the market for safe food may not exist. Both the Department of Agriculture and the Food and Drug Commission have issued safety certificates to producers who meet the regulation standards. As a result, the market for safe foods was created in a short time.

During the AI outbreak in 2004 the government culled more than 30 million birds and established a strict disease control system. While these measures are necessary if outbreaks are to be rapidly controlled, they impose a huge cost on the farmers. The government for the first time decided to share these costs by providing partial compensation.

7.5 Pro-poor growth strategy

Despite the fact that agricultural employment is very large and that poverty is concentrated in agriculture, Thailand has successfully reduced poverty incidence to a level that already surpasses the Millennium Development Goals. This study has found evidence that Thailand's growth strategy is pro-poor. From 1960 to 1985 poverty reduction was made possible by rapid agricultural growth. After that, it was through non-agricultural growth. A good transportation and communication system and a promigration policy have enabled poor farm households to earn extra income from non-agricultural employment and receive remittances from their children who have emigrated to the cities. Finally, the successful negotiations in the Doha Round to reduce domestic and export subsidies and to provide more market access will help reduce poverty in food-exporting developing countries through their positive impact on world agricultural prices.

Appendix 1

Total factor productivity growth in Thai agriculture

To understand the transformation process of Thai agriculture, this study investigates the pattern of changes in agriculture sector during 1980 to 2003. During the two decades of the 1960s and 1970s, agriculture made a relatively large contribution to Thai economy and was considered the engine of industrial growth. Over the period, the economy grew at an average annual rate of 7.07 percent whereas the agriculture sector grew at 4.61 percent.

However, the prospect for Thai agriculture was bleak. The importance of agriculture to the economy diminished in favour of manufacturing and service industries over time. The share of agricultural value added in the GDP peaked at 37.88 percent in 1951, but declined gradually to 20.20 percent in 1980, even though agricultural production more than trebled from 50.5 to 184.6 billion baht in 1988 constant prices. Over the same period, the manufacturing share in the GDP rose from 13.57 to 23.10 percent. With the rapid growth of Thailand's industrial sector, there has occurred a further drop in the share of agricultural value added in the GDP, to 13.55 percent in 1990. In addition to the economic development based on foreign investment-driven export-oriented industrialization, there were at least three main causes of the declining contribution of agriculture in the 1980s: 1) closure of the land frontier, 2) worldwide agricultural depression (Siamwalla 1991; 1996) and 3) low-income elasticity of demand for agricultural products.

Historically, Thailand's land-abundant economy allowed agricultural growth to be sustained through the expansion of land area under cultivation and enabled agriculture to continue to absorb large amounts of labour. The land area per agricultural worker was 0.76 ha in 1961, rose to 1.15 ha in 1978 and dropped to 0.93 ha in 1987 (Siamwalla 1991). Before 1990 the cultivated area grew substantially as agriculture expanded into previously forested areas. Total production of major and second rice crops has increased rapidly during the past 23 years. Statistics from the Office of Agricultural Economics indicate that major rice production increased more than 50 percent from 17 million tonnes in 1980 to 27 million tonnes in 2002, partly due to expansion of the cultivated area. In addition, average yields increased over the same period, possibly reflecting increased productivity through the introduction of new technologies, labour quality and a major expansion of irrigation.

However, Thailand's former advantage in the supply of land and low-cost labour has diminished, causing negative effects in its agriculture. As the comparative advantage of agriculture declined and the policy framework was biased towards industry, the adjustment process encouraged labour to migrate to urban areas.

In the early 1980s, agriculture absorbed 69 percent of the Thai labour force, but in more recent years it has fallen below 45 percent. Full-time farmers are even fewer, as most farmers have taken up a second occupation. Since the 1980s, agriculture no longer takes up most of the incremental increase in the labour force. Specifically, the female workforce in agriculture during the wet season declined. A large pool of trainable and inexpensive labourers began to move out of agriculture into more remunerative activities arising from the boom in industry, where the average real wage was much more attractive. Thus, the employment rate in agriculture declined consistently overtime.

In addition, agriculture encounters another problem caused by the shift of terms of trade against agriculture. This could be partially explained by the Engel effect as there is a low income-elasticity of demand for agricultural products. When income rises a smaller than proportional amount of the increase is spent on agricultural products. Slow growth in the demand for its products causes the sector to grow more slowly than the rest of the economy. Therefore, if production in world agriculture increases, then it is likely that the terms of trade for agriculture in these markets will decline as well.

Two decades of these changes during 1980 to 2003 brought certain changes in the structure of input utilization in agricultural production. Hereunder, we examine the consequences of these changes and the extent to which various input factors caused the output to change.

Labour and capital are two important inputs that are combined to make outputs. For Thailand as an agriculture-based economy, land is another input factor that helps to drive the growth of agricultural outputs, so it is regarded as one of the important factors in our analysis. Generally, we decompose growth that comes from using more inputs and from making better use of inputs in agriculture and its subsectors based on the growth-accounting framework (Appendix 2). The output efficiency of factors or total factor productivity (TFP) is important in economic growth in the sense that technical change compensates for the diminishing return to factors.

1. Measurement of output and inputs

An obstacle to an attempt to study the productivity of the agricultural subsectors in Thailand is unavailability of relevant and accurate data. The major sources of published agricultural data are from the National Economic and Social Development Board (NESDB), the Office of Agricultural Economics, Department of Fishery and Department of Livestock Development. Although the national income source does disaggregate value added of agricultural output into many subsectors, other sources for input factors do not. Therefore, we must rely on other sources like agricultural census, marine fishery census and economic data of livestock that allow the construction of a time series of input use.

1.1 Output

The output measure for each sector used in this study is the value added produced in a given period or GDP at 1988 prices as the measure of real income. We use the growth rate of GDP for each sector in measuring TFP. The series comes from the NESDB. However, an adjustment has to be made to the series of agricultural GDP because the NESDB changed the way of reporting the value added from simple agricultural processing by defining it to be a contribution from the non-agricultural sector. Thus, we have to adjust either by simply excluding them from the 21-year series of agricultural GDP or by extrapolating for another two years. We decided to make the extrapolation, instead of omitting some information from the series, with the estimation based on logarithmic trend regression.

Appendix Table 1 shows the growth rate and share of GDP at constant prices for the whole economy, agriculture sector and agricultural subsectors. Agriculture was still predominant, contributing approximately 20 percent to the GDP during the early 1980s. Its growth was at an average rate of 3.43 percent during 1980 to 2003. After the economic recession of 1981 to 1982, growth was achieved mainly in the non-agricultural sectors. Manufacturing output led the recovery. Although the economy grew at double digit rates during 1988 to 1990, agriculture could not keep pace and even experienced negative growth in 1990. Its share of GDP shrank to 11.43 percent in 2003 from 20.20 percent in 1980, mainly due to lower outputs from the crop sector.

Structural changes in agrarian production also took place. The output of livestock more than doubled and fishery increased more than three-fold. Thailand's value added generated from agriculture was estimated at 395 billion baht in 2003, of which the fishing industry contributed 51 billion or about 13.01 percent of total agricultural GDP, an increase from the average of 9.6 percent during 1980 to 1985. The main reasons for higher contribution to GDP by the fishery sector in recent years have been its relatively important contribution to export earnings and ability to attract more employment owing to higher average wage payments than other agricultural subsectors.

1.2 Factor inputs

Labour

To measure labour input, we employ both quantity and quality of labour input data from the Labour Force Survey. Quantity data include total employment in each agricultural subsector and average weekly hours of work. For labour quality, we focus on age, gender and education of the labour force, assuming that employees of different age groups, gender and educational attainment have different marginal productivity.

In order to compare our results with those of Tinakorn and Sussangkarn (1996; 1998), we follow their categorization of age and education groups. Specifically, we group the labour into five age groups - 15–19, 20–29, 30–39, 40–49 and over 50. However, in contrast to Tinakorn and Sussangkarn, the total employment in this study does not include the 11–14-year-age group, because since 2001 the National Statistical Office (NSO) has changed the definition in its labour force survey to include only persons at least 15 years old.

We then categorize the labour into five educational groups including: 1) those with no formal education and with elementary education or lower, 2) those with upper and lower secondary education, 3) those with vocational education, 4) those with university education or high level technical vocational education and 5) those with teacher training.

As our sample period covers 1980 to 2003, the study uses the second round survey for 1980 to 1983, as it is the only round available and the third round survey for 1984 to 2003.

The labour force of Thailand was estimated at 21.62 million in 1980, of which some 15.13 million employed persons, or about 70 percent of the total, were employed in agriculture (Appendix Table 2). Poapongsakorn et al. (1998) document that rapid economic development since the 1980s increased aggregate demand for labour in Thailand and led to an increasing competition for labour that threatened the labour market in agriculture. They also show that in the drop in the number of employed persons in agriculture, the decline was especially large for the 15–24 age group. One important reason for the decline in agricultural employment is due to intersectoral migration to sectors with higher real wages.

Appendix Table 2 shows a declining share of agricultural employment in total employment since 1980, from 70 percent in 1980 to less than 45 percent in 2003. Consistent with the declining trend in share, average weekly working hours of agricultural workers, especially in the crop sector, fell from 60 hours per worker in 1980 to about 44 hours in 2003 (Appendix Table 3). This happened probably because most farmers have a second occupation.

In addition, there might be an intrasectoral migration from crops to livestock and fisheries. The proportion of the Thai labour force in these two sectors has increased in recent years. The average growth of total employment in these two sectors during 1980 to 2003, was 11.4 percent and 5.79 percent respectively, compared with the crop sector's unimpressive negative average rate. Intrasectoral migration could be due to government intervention that aims to replace production of lower-value products with higher-value products.

Land

Although land is usually considered as a fixed factor of production in industrialized countries, it is more appropriate to account for land as one of the important factor inputs in the Thai agriculture sector. Land abundance had allowed Thailand to have a strong comparative advantage in agriculture. However, land was no longer in plentiful supply after 1980, the area per agricultural worker declining at an annual rate of about 1 percent (Siamwalla 1991).

As our focus is on three different agricultural subsectors, the land utilization for each subsector is defined differently. For the crop sector, employ the total area under cultivation for rice and other crops such as maize, cassava, sugar cane, soybean, mung bean, groundnut, cotton, sorghum, kenaf, coconut, kapok, sesame seed, tobacco, coffee, pineapple, oil palm, etc. For fisheries, the surface area for fish and shrimp culture is used. Finally, pasture and individually owned areas for livestock are used as an input factor.

To account for improvement in the quality of land, we also adjust the land input with access to irrigation. As Siamwalla (1991) pointed out that most of the irrigation structures were designed for rice production, we should expect a positive impact of irrigation on the yield of crop outputs. In the estimation of TFP growth, we therefore adjust the index of total cultivated area used in the crop sector by multiplying with the index of proportion of irrigated area.

Appendix Table 4 shows that the irrigated area increased gradually from about three million ha in 1980 to 5.24 million ha in 2003, with an average annual growth rate of 2.46 percent. The area accounts for more than 28 percent of total cultivated land for rice and all other crops. Based on the Agricultural Census in 1993 and 1998, about 72 to 74 percent of irrigated land is used in rice production while field and permanent crops account for nearly 26 percent. Irrigation provides a better quality of land and allows multiple crops per year. However, expansion of the irrigated area is constrained by the availability of capital investment in effective irrigation systems.

The total public and private area for livestock tended to increase until 1996 whereafter it declined. However, the decline does not come from the reduction in pasture, but in the individually owned areas, as surveyed by the Department of Livestock Development; in 1997, the privately owned beef cattle area fell by nearly one-third compared with the previous year.

Capital

In a country that moves from the use of labour-intensive to capital-intensive methods, the capital inputs' contribution to the growth of output should be increasing. For the overall agriculture sector and the whole economy, the data on accumulated capital stock from the NESDB is presented in Appendix Table 5. Although the net capital stock in agriculture increased year after year, its share in the total capital stock declined. Conversely, the net capital stock of the manufacturing sector has increased tremendously.

The capital stock data of the NESDB do not show how much capital is accumulated in each agricultural subsector. We therefore use alternative sources of information to estimate capital inputs used in the crop, fishery and livestock sectors.

For crops, we collected the number of units of agricultural equipment - tractors, water pumps and threshing machines - held by each farm household from six different Agriculture Censuses and InterCensuses and adjusted with numbers reported by the Office of Agricultural Economics. Appendix Table 6 confirms the evidence of mechanization in cultivation: the average growth rates of two-wheel tractors, big tractors, water pumps and threshers used in the sector are relatively high. As the unit values of equipment for the whole series is not available, we indirectly use the values of farm property per farm in 1996 in measuring the values of agricultural equipment during 1980 to 2003.

For fisheries, we take into account the estimated investment value of fishing assets (hull, engine and gear) for the marine fishery and the investment cost in culture units and equipment for freshwater fishery. We also use the number of unregistered fishing boats based on the Marine Fisheries Census to adjust the total number of boats used in marine capture fisheries reported by the Department of Fishery, because the latter source does not include unregistered small-scale vessels that make up a significant proportion of vessels in Thailand (Appendix Table 7). The investment costs of freshwater fisheries are not available. However, we use the depreciation of culture units and equipment to compute the initial capital investment.

For livestock, we recognize that the inventory of breeding animals is a major capital input to livestock production. The inventories of major productive livestock like breeding swine, milking cows, non-dairy cattle, laying hens and buffalo are included in the analysis. The investment costs of housing units and equipment per unit for each type of livestock is calculated based on average depreciation and expected useful life.

2. Estimates of factor income shares

In calculating the factor income shares, we compute the cost of inputs as a portion of sectoral GDP at factor cost. The derived income shares are retained for the end points due to data limitation, but are smoothed by a three-year moving average for other points of time to reduce the short-term effects of business fluctuations before they can be used to calculate TFP. However, the difficulty is in the calculation of the cost of inputs for land, labour and capital.

For the factor income share with respect to labour inputs, we adjust average yearly wages of all private workers from the Labour Force Survey with the wage payment from the 1995 and 2001 Social Accounting Matrix (SAM) constructed by Sussangkarn (2001). In this way we take care of the earnings of own account workers. In addition, as qualitative change of labour inputs may have some influential effects on labour productivity, we take the composition of age, sex and education attainment into account and adjust the labour inputs as explained in Tinakorn and Sussangkarn (1996). Specifically, the composition of workers by age, sex and education is computed each year by multiplying the percentage of workers in each age-sex-education group by the corresponding average wage differential index and summing the products.

For the income share of land, the rent for subsectors (crops, fisheries and livestock) other than all sectors and the agriculture sector is estimated based on average rent per hectare, which is imputed from the actual rent in the National Income Account. However, this method imposes the assumption that the rent per hectare paid by a farmer is the same irrespective of the type of land utilization. After computing both of the income shares of labour and land inputs, we can calculate the factor income share of capital as a residual.

3. TFP growth of the Thai economy

Using equation 3 and the available data, we compute the contribution of each type of input and of technological change to growth. Appendix Table 8 summarizes the TFP results for each major sector based on the results presented in Appendix Tables 10–16.

From the estimation, the average annual growth rate of the overall Thai economy during 1981 to 2003 is lower than the average growth rate during 1981 to 1995, perhaps due to the adverse effects of the financial crisis in 1997. For the whole period, the output grew at an average rate of 6.07 percent, of which 85.18 percent (77.51 percent) is due to the contribution of factor inputs with total employment (total working hours). Among the three inputs, the main contribution to economic growth comes from the capital stock, at 52.18 percent. Labour inputs make the second largest contribution to the growth. Total labour quality adjusted for employment shows a larger contribution of 32.55 percent when compared with total working hours adjusted for labour quality that accounts for 24.88 percent of GDP growth. The contribution from land is trivial, less than 0.5 percent for all sectors. Adjusting labour quality for total employment (total working hours), this renders the average growth of TFP during 1981 to 2003 at about 14.82 percent (22.49 percent) of GDP growth. However, the causes of the financial crisis had some impact on the growth rate of the whole economy as well as the growth rate of TFP. It is evident that TFP started to decline in 1996 until 1998, before it increased again in 1999.

As for the overall economy for the whole period, all the factor inputs including the land factor have a major contribution to the growth of outputs in agriculture. Accounting for the adjustment in labour quality for total employment and total working hours, the percentage contributions of all factor inputs are 63.91 and 47.48 percent respectively. The capital accumulation in the agriculture sector makes a significant contribution to the value added originating from agriculture while the labour inputs play a less important role in the growth of the sector. The land factor adjusted with the irrigation index shows a higher contribution to the growth because agriculture makes use of most of the land area in cultivation. In the earlier period of 1981 to 1995, agricultural output grew at an average rate of 3.68 percent and the percentage of contribution of all factors to the growth in the agriculture sector is quite similar to that of the whole period. However, one important point is that when we account for the total working hours as labour input, its contribution to the growth is positive at 5.90 percent for 1981 to 1995 compared with -7.26 percent for the whole period. This is mainly due to the decline in total working hours of employed persons during the postfinancial crisis period. Specifically, average weekly hours of employed persons in the agriculture sector declined from 59.79 hours in 1980 to 44.21 hours in 2003. We can observe the decline in the crop and fishery sectors, but an increase in the livestock sector.

Compared with the findings in Tinakorn and Sussangkarn (1998) using a similar methodology, average calculated TFP growth in our study is much higher for the same period of 1981 to 1995. The four main sources of difference are as follows. Firstly, the definition of employment used in this study includes only those of at least 15 years of age while their definition includes people of 11 years or above. This could make total working hours and wage payment on average in this study higher. Secondly, the adjustment ratio on annual wage based on average wage from the SAM is also different because we use total annual wage from the 1995 and 2001 SAM data. Thirdly, we define land areas for agriculture to include not only planted areas for major crops but also areas for fish and shrimp culture and for livestock. Finally, our rent figures are from land use while their figures cover the rent from land and buildings.

Different from the study of Tinakorn and Sussangkarn (1998) and this study, Shintani (2002) using input-output data for Thai agriculture, constructs a series of related data for 1950 to 1997 and estimates the TFP index from the Divisia production index and the Divisia total input index. Shintani's estimated results show that the TFP decreases or stagnates until the 1970s, but increases afterwards. During 1981 to 1995, the author insists that the production growth of agriculture is largely explained by the flow of labour input and the flow of land input instead of stock-based series.

Based on the results from various studies, we can observe that the rates of TFP growth differ widely, with its contribution to agricultural growth ranging from 5.99 percent to as high as 52.52 percent. However, this depends largely on the definition of variables in use, period, data set and methodology.

To understand more about the growth and the changing pattern of Thai agriculture, we further investigate the development of agricultural subsectors. We include crop, livestock and fishery sectors, but do not include the forestry sector as its accounts for only a small percentage of agricultural GDP.

Applying the same methodology described above, we measure the TFP growth for each subsector separately. One pattern that can be observed for all the three subsectors is that capital stock makes a major contribution to its own sector's output growth. However, caution is needed when interpreting the TFP growth in the fishery sector. Fishery can be principally divided into freshwater and marine fishery and these two subsectors have different natures. We therefore estimate TFP separately for each of them. Looking at the overall picture of the fishery sector, the contribution of capital in fishery is relatively high, probably because we use many unregistered fishing units in the analysis. The TFP growth in this sector is thus negative on average. Analysing the subsectors separately, we find that, between 1980 and 2000, the average TFP growth of the marine fishery and its contribution to output growth are higher when compared with those of freshwater fishery. One possible explanation is that those who are engaged in marine fishery might enter into an agreement with neighbouring countries to capture fish in their territorial sea area. However, the average TFP growth for marine fishery activities has been declining in recent years. Between 1981 and 1995, technical change averaged 2.59 percent annually compared to an annual average rate of-5.6 percent between 1996 and 2000. This primarily could be because of overexploitation leading to the depletion of marine fishery resources.

For the crop sector, labour does not make any positive contribution to the sector's growth, perhaps because of the impact of mechanization. Nevertheless, we may still underestimate the growth of its capital stock because we do not have a series of the value of equipment and machinery. We at best assume that the value of equipment and machinery is constant throughout the period and is imputed from the value of farm machinery per farm in 1995 to 1996.

Aside from the crop sector, the contribution of TFP in the livestock and fishery sectors is very small for livestock and even negative for the fishery sector. On the contrary, the contribution of factor inputs appears to be very high. Capital investment and quality adjusted labour factors account for the largest proportion of output growth while we fail to find much contribution from land. This is not surprising because the livestock and fishery sectors do not use much land area compared to the cultivation of crops.

In sum, the rise in TFP in agriculture, particularly the crop sector, in the postcrisis period reflects improved varieties of crops and changes in output composition. One pattern that could be found for all agricultural subsectors is that capital accumulation plays an important role in the output growth of its associated sector. However, labour input does not show a consistent pattern. The higher average wage rate in the livestock and fishery sectors tends to attract migration of employed persons into the sectors, leading to higher contributions.

Appendix Table 1. Growth rate and share of GDP at 1988 prices by major sector

* Including Simple Agricultural Processing.

Source: National Economic and Social Development Boa

Appendix Table 2. Total employment with at least 15 years of age by major sector

* Adjusted for foreign employees.

Sources: LFS (Round 2) for 1980–1983, LFS (Round 3) for 1984–2003, Department of Fishery.

Appendix Table 3. Average weekly working hours by major sector

Sources: LFS (Round 2) for 1980–1983, LFS (Round 3) for 1984–2003, Department of Fishery.

Appendix Table 4. Land used in agriculture

Unit: Hectare

Sources:   Office of Agricultural Economics.
Department of Livestock Development.
Department of Fishery.