Food consumption expenditure structure in Thailand 1998: the case of vegetables

In the past 40 years, Thailand successfully developed her agricultural based economy towards rapid industrialization. At the same time, real income increased, and consumption expenditure shares shifted from food to non-food items - consistent with the famous Engel's Law. Changing consumption behaviour due to pronounced urbanisation enhanced this movement. However, within food expenditure, the share of vegetable and fruit consumption increased at the expense of basic foodstuff like cereals and rice. In the light of theses changes in food consumption, and taking account of the enforced challenge to produce "safe" and "healthy", vegetables in Thailand caused by the intensive use of chemicals in agriculture and horticulture, the Faculty of Horticulture at Hannover University launched a joint research project on environmentally friendly production systems, including economic issues. Within the scope of possibilities and constraints of marketing environmentally friendly produced vegetables, we carried out an analysis of the prevailing consumption expenditure structure of vegetables in Thailand in order to generate information on the relative importance of different kinds of vegetables so far not available. The results of this market research can contribute to the selection of promising research projects, and may ease production decision-making. The Statistical Data Bank and Information Dissemination Division made available the Socio-Economic Survey Data of 1998, consisting of 23,549 Thai households, to elaborate the structural aspects of vegetable consumption including regional as well as socio-economic characteristics. The descriptive analysis revealed persuasive relationships between incomes, regional, and socio-demographic factors and household vegetable expenditure structure. Finally, we tried to quantify the presumed causal structure point towards the right expected economic impact of income, regional and social affiliation of the household by applying single equation models using Ordinary Least Squares (OLS) to the data. The generated estimates are reasonable from an economic point of view. However, albeit these encouraging aspects, however, the statistical significance is not sufficiently good in all cases.