Loading...
[Global Corruption Report 2007] Identifying reticent respondents in Romanian corruption surveys
Azfar, Omar Murrell, Peter
Azfar, Omar
Murrell, Peter
Author(s)
Author(s) (Additional)
Illustrator(s)
Producer(s)
Contributor(s)
Contributor(s) (Other)
Editor(s)
Advisor(s)
Contact(s)
Data Collector(s)
Keywords
Collections
Files
Loading...
n0521700702c22_p352-353.pdf
Adobe PDF, 41 KB
Research Projects
Organizational Units
Journal Issue
Online Access
Abstract
"The accurate measurement of corruption has become an issue of critical importance to both researchers and policy makers. One increasingly prevalent method of collecting data on corruption is surveys. But corruption is a highly sensitive topic; many survey respondents are ‘reticent’ and may prefer to give incomplete or non-truthful responses. The following summarises how reticent respondents can be identified and removed from the data, thus increasing the accuracy of survey-based estimates of corruption.2 The IRIS Center of the University of Maryland conducted a survey of 514 private sector firms in Romania. The survey design included innovative modules that could be used to identify respondents who were not giving candid answers to sensitive questions. The general purpose of the survey was to understand the degree and nature of corruption in registering, licensing and inspecting businesses in Romania. The questionnaire therefore focused on business interactions with two government entities: the One-Stop Shops, charged with administering the registration (and frequent re-registration) of businesses; and the inspections and authorisations departments of the local branches of the health ministry. The former was chosen because all firms have to deal with the One-Stop Shops and the latter because health licensing is one of the most intrusive and administratively burdensome requirements that Romanian businesses face. The randomised response method used to identify reticent respondents is as follows. Each respondent was asked to read a sensitive question and to toss a coin. An example of the question asked might be: ‘Have you ever paid less in personal taxes than you should have under the law?’ The respondent was then asked to say yes if either the coin came up heads or if he or she had indeed committed the act (which in this case would mean that he or she had paid less in personal taxes than was legally required). The procedure was repeated another six times with a different coin toss and a different sensitive question. None of these seven sensitive questions were about corruption. If the respondent said no seven times in a row, he or she is classified as reticent, because it is very unlikely that the respondent would have tossed seven tails in a row. All other combinations are classified as ‘possibly candid’ as this group contains both candid respondents as well as some reticent respondents."(pg 352)
Note(s)
Topic
Type
Book chapter
Date
2007
Identifier
ISBN
9780521700702
DOI
Copyright/License
With permission of the license/copyright holder