What does Variation in Survey Design Reveal about the Nature of Measurement Errors in Household Consumption?
| Published date | 01 June 2015 |
| Date | 01 June 2015 |
| Author | Kathleen Beegle,John Gibson,Jed Friedman,Joachim De Weerdt |
| DOI | http://doi.org/10.1111/obes.12066 |
466
©2014 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 77, 3 (2015) 0305–9049
doi: 10.1111/obes.12066
What does Variation in Survey Design Reveal about
the Nature of Measurement Errors in Household
Consumption?*
John Gibson,†Kathleen Beegle,‡Joachim De Weerdt§ and
Jed Friedman‡
†Department of Economics, University of Waikato,Private Bag 3105, Hamilton 3240, NZ
(e-mail: jkgibson@waikato.ac.nz)
‡Development Research Group, The World Bank, 1818 H Street, NW, Washington DC
20433, USA (e-mail: kbeegle@worldbank.org, jfriedman@worldbank.org)
§Economic Development Initiatives (EDI), Bukoba, Kagera Region, United Republic of
Tanzania (e-mail: j.deweerdt@edi-africa.com)
Abstract
We randomly assigned eight different consumption surveys to obtain evidence on the
nature of measurement errors in estimates of household consumption. Regressions using
data from more error-prone designs are compared with results from a ‘gold standard’
survey.Measurement er rors appear to have a mean-reverting negative correlation with true
consumption, especially for food and especially for rural households.
I. Introduction
Household consumption surveys are crucial for measuring living standards, especially in
poor countries where labour and income surveys cover neither the full population nor
all activity. The complexity of consumption surveys makes measurement error likely but
validation studies to reveal the nature of errors are mainly for labour market settings, using
either administrative data (Bound and Krueger, 1991) or firm records (Pischke, 1995).
Also, such studies are limited to rich countries, including for the one type of consumption
– health expenditure – for which validation has been attempted (Cohen and Carlson, 1994).
This note reports on a survey experiment in Tanzania that may inform practitioners
about the nature of measurement error in developing country consumption data. Eight
different survey designs were randomly assigned to 4,032 households (three households
per design in 168 sampling units), covering urban and rural areas. One resource intensive
design – individually kept diaries with dailysuper vision for 14 days– approximates a ‘gold
*This work was supported by the World Bank Research Committee. We would like to thank two anonymous
referees, the editor and seminar participants at Hitotsubashi and Monash for useful comments. All views are those
of the authors and do not reflect the views of the World Bank or its member countries.
JEL Classification numbers: C21, C81, D12.
To continue reading
Request your trial