Estimating Obesity Rates in the Presence of Measurement Error. IZA Discussion Paper No. 7288


O'Neill, Donal and Sweetman, Olive (2013) Estimating Obesity Rates in the Presence of Measurement Error. IZA Discussion Paper No. 7288. Discussion Paper. IZA (Institute for the Study of Labor), Bonn.

[img] Download (287kB)


Share your research

Twitter Facebook LinkedIn GooglePlus Email more...



Add this article to your Mendeley library


Abstract

Reliable measures of obesity are essential in order to develop effective policies to tackle the costs of obesity. In this paper we examine what, if anything, we can learn about obesity rates using self-reported BMI once we allow for possible measurement error. We combine selfreported data on BMI with estimated misclassification rates obtained from auxiliary data to derive upper and lower bounds for the population obesity rate for ten European countries. For men it is possible to obtain meaningful comparisons across countries even after accounting for measurement error. In particular the self-reported data identifies a set of low obesity countries consisting of Denmark, Ireland, Italy, Greece and Portugal and a set of high obesity countries consisting of Spain and Finland. However, it is more difficult to rank countries by female obesity rates. Meaningful rankings only emerge when the misclassification rate is bounded at a level that is much lower than that observed in auxiliary data. A similar limit on misclassification rates is also needed before we can begin to observe meaningful gender differences in obesity rates within countries.

Item Type: Monograph (Discussion Paper)
Keywords: obesity; measurement error; bounds;
Academic Unit: Faculty of Social Sciences > Economics, Finance and Accounting
Item ID: 4339
Depositing User: Olive Sweetman
Date Deposited: 07 May 2013 15:27
Publisher: IZA (Institute for the Study of Labor)
URI:

Repository Staff Only(login required)

View Item Item control page

Document Downloads

More statistics for this item...