Izvorni znanstveni članak
https://doi.org/10.1515/zireb-2016-0006
Forecasting of Households Consumption Expenditure with Nonparametric Regression: The Case of Turkey
Noyan Aydin
; Faculty of Economics and Administrative Sciences, Department of Econometrics Kutahya, Turkey
Taner Akmercan
; Faculty of Economics and Administrative Sciences, Department of Econometrics Kutahya, Turkey
Sažetak
The relationship between household income and expenditure is important for understanding how the shape of the economic dynamics of the households. In this study, the relationship between household consumption expenditure and household disposable income were analyzed by Locally Weighted Scatterplot Smoothing Regression which is a nonparametric method using R programming. This study aimed to determine relationship between variables directly, unlike making any assumptions are commonly used as in the conventional parametric regression. According to the findings, effect on expenditure with increasing of income and household size together increased rapidly at first, and then speed of increase decreased. This increase can be explained by having greater compulsory consumption expenditure relatively in small households. Besides, expenditure is relatively higher in middle and high income levels according to low income level. However, the change in expenditure is limited in middle and is the most limited in high income levels when household size changes.
Ključne riječi
Household Consumption Expenditure; nonparametric regression; LOESS
Hrčak ID:
168637
URI
Datum izdavanja:
15.11.2016.
Posjeta: 1.096 *