El rol de los ingresos en los modelos de elección discreta. Su importancia en los proyectos de evaluación
Research on discrete choice models based on microeconomic principles has developed a framework for detecting the income effect to estimate demands or even measure well-being with special emphasis on the transport sector. However, these theoretical investigations have not been used frequently in the...
Main Author: | |
---|---|
Format: | Article |
Language: | English |
Published: |
Universidad de San Martín de Porres
2020-07-01
|
Series: | Revista de Análisis Económico y Financiero |
Subjects: | |
Online Access: | http://contabilidadyeconomiausmp.edu.pe/OJS2020/index.php/RAEF/article/view/26 |
Summary: | Research on discrete choice models based on microeconomic principles has developed a framework for detecting the income effect to estimate demands or even measure well-being with special emphasis on the transport sector. However, these theoretical investigations have not been used frequently in the practice of evaluation projects. The aggregation of benefits across individuals requires sophisticated tools to avoid misspecification of income in the models. The latter can potentially distort the results and unintentionally lead to non-optimal decisions that become critical in large investments.
This document seeks to highlight the importance of considering the effect of income in evaluation projects. For this purpose, based on discrete choice theory, several models have been tested using multinomial logit models. The collected data set comes from an urban transport survey of Metropolitan Lima in 2004. The models have tried to detect the presence of the income effect, thus assuming an income level. It has been found that the people surveyed in the study can be classified statistically into two representative income groups. Then, following Jara-Díaz and Videla (1989), he shows that the marginal utility of income decreases with income and these are statistically different when calculated between groups. |
---|---|
ISSN: | 2617-9989 2663-8908 |