| Home | Sitemap | Contact Us |  
top_img
Hum. Ecol. Res Search

CLOSE

Journal of the Korean Home Economics Association 1990;28(4):87-101.
논문편 : 가계분석에 있어서 Engel curve 의 함수형태에 관한 연구
배연수
영남대학교 상경대학 경제학과
Study on Forms of Engel Curves in the Analysis of Household Budgets
배연수
ABSTRACT
This study was undertaken to test to fit forms of the Engel curves to data. The comparisons were confined to the linear, semi-logarithmic and double-logarithmic forms. Data from the 1970-1987 Urban Household Economy Survey were used to estimate the Engel curves. The twelve categories of consumption expenditure were considered for investigation. Parameters of the Engel curves were derived from OLS and TSLS. In this paper the size of the family was used as the deflater. The results could be summarized as follows: 1. Comparing with the R2 of three foms, it could be concluded that, the linear form generally gave a better fit to data than the other forms did. Only for housing and clothing and foot wear, did the semi-logarithmic form give a better fit. Only for meals outside the home, fuel, light and water charges, and miscellaneous, did the double-logarithmic form give a better fit. 2. Comping with the income elasticities based on the alternative forms, it could be concluded that the differences between the estimates were since each form made different assumption as to the way in which elasticity varied. In general, the semi-logarithmic form gave the highest estimate and double-logarithmic form did the lowest estimate. The difference between semi-logarithmic and the other forms were greater than the those of linear and double-logarithmic form. 3. It was found that the income elasticity varied with the difinition of income used as an explanatory variable in Engel curves.
Editorial Office
The Korean Home Economics Association
TEL : +82-2-561-6416, +82-2-561-6446    FAX : +82-2-562-2999    
E-mail : khea6416@daum.net
About |  Browse Articles |  Current Issue |  For Authors and Reviewers
Copyright © 2014 The Korean Home Economics Association.                 Developed in M2PI