|
Activity Number:
|
371
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2004 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Business and Economics Statistics Section
|
| Abstract - #300610 |
|
Title:
|
The Effects of Imputation on Income Distribution as Estimated by Means of Dagum Models
|
|
Author(s):
|
Andrea Regoli*+ and Rosalia Castellano and Claudio Quintano
|
|
Companies:
|
Parthenope University of Naples and Parthenope University of Naples and Parthenope University of Naples
|
|
Address:
|
Department of Statistics and Mathematics, Napoli, 80133, Italy
|
|
Keywords:
|
missing data ; parametric income distribution models ; inequality indices
|
|
Abstract:
|
The problem of missing data is a crucial point for the quality of the statistical information. In the presence of item nonresponses, one can choose to carry out the analysis only on the valid cases or to fill up the nonresponses through imputation methods in order to reconstruct the complete dataset. We have studied the effects of the imputation of missing income values on the main income descriptive statistics and the measures of association with income. Our interest now lies in the evaluation of the consequences that different imputation procedures may have on the whole income distribution, with special attention to the inequality measurements. Starting with the estimation of the Dagum income distribution model, we want to analyze to what extent a given missing income treatment can modify the parameters of the model. Moreover, since each parameter has its own economic interpretation, we try to interpret the meaning of such changes. The study is carried out on the income data from the Survey of Household Income and Wealth by the Bank of Italy. The missing responses are artificially generated through a simulation technique.
|
