Tackling Non-Response Bias in Surveys of Wealthy Households in Low- and Middle-Income Countries


Picture this: a survey targeting the wealthy or aiming to be representative of the entire population, but rich respondents can’t be accessed, refuse to participate, or hold back sensitive information. Unit non-response and item non-response become a gated wall, hindering our view of the global picture, and on wealthy individuals.

Capturing the perspectives of wealthy populations in research is not only crucial for our understanding of income and wealth inequality but also to shed light on diverse areas such as energy demand, entrepreneurship, consumer behaviour, housing markets, urban services, education, and career trajectories. Comprehensive data on wealthy households is key to unlocking a multitude of insights and informing evidence-based policies across a wide range of domains.

In our recent publication “Addressing Non-response Bias in Surveys of Wealthy Households in Low- and Middle-Income Countries: Strategies and Implementation” in the Journal of Development Studies, we offer a comprehensive view on the hurdles that researchers face when surveying wealthy households. We specifically delved deep into the causes and consequences of unit non-response and item non-response, which can sometimes puzzle researchers.

Surveys targeting wealthy households come with their fair share of challenges, from gaining physical access, refusals to participate, or disclosing specific information. Thus, unit and item non-response pose a formidable threat to data quality. When individuals do not take a survey or key survey items go unanswered, valuable information is lost, shrinking our sample size, and potentially skewing our analysis. This becomes a real concern when there are systematic differences between those who respond and those who do not. Researchers must then think outside the box and explore alternative data sources and methods to capture the full picture.

By developing well-designed field plans (including specific interview training, advance communication, expanding survey hours, call backs, short interviews, etc.), we can bridge the gap between field researchers and wealthy respondents. But the battle against non-response bias doesn’t end with data collection. Reweighing and replacing can correct for biases post-data collection.

By tackling non-response bias head-on, researchers can pave the way for a deeper understanding of income inequality, consumer trends, entrepreneurship, and a myriad of other fascinating subjects. Because when it comes to research, every piece of the puzzle matters.

Authors: Dr Johanna Choumert Nkolo, Director of Research, and Gabriela Santana Tavera and Prakhar Saxena

Work With Us

Get In Touch

Interested in impactful research and sustainable development? Drop us a line to get started.