Thai household's rice consumption and its demand elasticity.

I. Introduction

Economic growth and urbanization have inevitably affected household food consumption patterns, particularly for rice, which is a staple food in almost all Asian countries. Indonesia's per capita rice consumption in both urban and rural areas has been decreasing over the last two decades. The share of rice in household expenditure has been decreasing as well because of a higher per capita income (Sudaryanto et al. 2002). Huang and Bouis (1996) found that the per capita rice consumption in Taiwan declined; while that of meat, fish, and fruits increased. In Thailand, the growing economy has stimulated the consumers to adjust their consumption behaviour towards an increasing demand for luxury food items, particularly meat and horticulture products, and with less rice (Patamasiriwat and Poldee 1990; Agribusiness Research Unit 1997).

Patterns in rice consumption can be assessed by using information on price and income elasticities of demand. The demand elasticities are normally measured from a single-equation demand function with price and income variables, using time-series data. Wong (1978) found the price and income elasticities of rice consumption in Thailand for the short run to be -0.42 and 0.09, respectively, and for the long run to be -0.47 and 0.10, respectively. By allowing the estimation of income elasticities to vary with level of income, Ito et al. (1989) found a negative income elasticity of rice consumption in Thailand. An alternative approach for the estimation of food demand is to employ the theory of demand as a guideline for mathematically specified models of consumer choice under the imposition of constraints on demand parameters (Deaton and Muellbauer 1980). A separability concept is assumed for this technique in order to partition the choice of a food group from other goods in the household budget constraint. Analysis of the share of the household food budget under the basic model of an Almost Ideal Demand System (AIDS) framework can provide the expenditure and price elasticities. Huang and Bouis (1996) employed this technique in analysing the parameters of food consumption in Taiwan. They found the expenditure and price elasticities of rice consumption in Taiwan to be 0.17 and -0.61, respectively.

In Thailand, the estimation of demand elasticity for rice consumption has been limited to the estimation of the Engel curve or expenditure elasticity using cross-sectional data of the households' expenses on food products (see Jansai 1996 and Prasertsung 2004 for instance). In this paper, using a more updated data set of cross-sectional data from the socio-economic survey (SES) of Thai households in 2002 allows for the estimation of both expenditure and price elasticities of rice demand for Thai household consumption. The analysis is based on the framework of a model of consumer behaviour introduced by Deaton (1988). The availability of data also allows exploration of consumption patterns in response to the differentiation of rice quality, subject to different income levels.

In the next section, food expenditure shares and rice consumption patterns are discussed. The analytical framework is then presented, followed by the empirical results. The final section contains the conclusion and recommendations.

II. Food Expenditure and Rice Consumption Patterns

Food is an important component in household expenditure. In Thailand, the monthly expenditure of an average Thai household is 14,361 baht. (1) Around 34 per cent of this is spent on food (Table 1). This is disaggregated into five food items, i.e., rice and cereal; meat and fish; milk, cooking oil and sugar; vegetables and fruits; and "others". Excluding the "others", the share of meat and fish is the largest, followed by milk, cooking oil and sugar, and then by fruits and vegetables. Rice and cereals have the least share. Households in communities with different development status differ in their rice expenditure patterns; households in the villages tend to spend a larger share on rice than those in the urban and suburban areas. In contrast, households in urban areas spend a higher proportion on "other" items, which include food consumed outside the home and beverages and alcohol, compared with those in the villages.