Measuring the unmeasured: Aggregating, anchoring, and adjusting to estimate small business performance

Journal Article
Published on 1 June 2021

Working paper available through PEDL. Published article available here.

Abstract

Business sales and profits for small firms are notoriously difficult to measure, with multiple sources of measurement and recall error. This paper introduces a new survey methodology that combines data triangulation with dynamic adjustment to arrive at plausible estimates of these performance outcomes. Business data across three emerging markets show this aggregating, anchoring, and adjusting (AAA) method consistently outperforms the traditional self-reported measure of sales, while the improvement in profits is more nuanced. Consequently, the paper offers a “light” version that harnesses the gains of AAA in estimating sales to prime and elicit an aided recall of profits, which correlates highly with AAA profits and reduces survey time and cost. The gains in reliability and accuracy result in much higher statistical power of detecting impacts and estimating required sample sizes in field studies.

Authors

Stephen Anderson

Stanford University

Christy Lazicky

IDinsight

Bilal Zia

World Bank