Purpose
The purpose of this paper is to test whether household preferences satisfy the assumption of base-independence, to examine the effects of household income on equivalence scales and thereby food consumption economies of scale and to examine how far conventional poverty rates require adjustment when scale economies in food consumption are taken into consideration.
Design/methodology/approach
To achieve these aims, the authors use a Pendakur (1999) adaptation of the test of base-independence, and income dependent Engel (1895) equivalence scales.
Findings
In Sri Lanka, the hypothesis of base-independence is rejected: the equivalence scales increase with household income both at the national and the sectoral level, that is urban, rural and estate sectors. This suggests that low-income households enjoy greater scale economies. After adjusting for scale economies, urban, rural and estate poverty headcount ratios decline by 3.2, 8.8 and 13.7, respectively, while at the national level the decline is about 8.3.
Research limitations/implications
The results are based on the assumption that all of the adults in the households have identical tastes, irrespective of their gender and age. Furthermore, the survey data exclude three districts in the northern province of Sri Lanka due to resettlement activities took place after the civil war.
Practical implications
Higher scale economies among the poor imply that poverty among low-income households is overstated when using traditional measures of poverty rates.
Originality/value
The novelty of this paper is that it provides insights on the effect of income on food consumption economies of scale and implications of this phenomenon on poverty estimates in the context of a developing country like Sri Lanka.
Assessing health gradient with different equivalence scales for household income – A sensitivity analysis