Photo credit: Adobe Photostock

Written by Peter A.G. van Bergeijk, Professor of International Economics and Macroeconomics, International Institute of Social Studies, Erasmus University, The Hague, the Netherlands

In the vast landscape of economic analysis, the quest for precision has long been revered as a holy grail. Yet, lurking beneath the surface lies a stark reality – the pervasive presence of measurement errors, often overlooked and underreported. My journey into the depths of economic data accuracy began during as a student when I worked as a junior country risk analyst and was confronted with the unsettling realization that economic statistics, even for developed nations, are far from infallible.

Different sources often presented conflicting numbers for fundamental economic indicators, leaving me puzzled and questioning the reliability of the data I relied upon. The numbers I reported had a tangible impact on critical assessments of country risk, yet their instability and unpredictability left me bewildered. This experience was my initiation into the world of measurement error – a world I had never been formally introduced to during my academic years.

It was during this period that I stumbled upon Oskar Morgenstern’s seminal work, “On the Accuracy of Economic Observations.” Morgenstern’s case studies laid bare the inherent inaccuracies in economic statistics, revealing discrepancies of ten to twenty percent in widely accepted data. His findings resonated deeply with me, particularly his account of gold movements between neighboring countries, where even the most meticulous observers fell short of accuracy.

I was struck by the glaring absence of discourse around measurement error in economic academia. Despite the technological advancements and methodological refinements, the fundamental issue of data accuracy remained largely unaddressed. My conviction in the relevance of Morgenstern’s work persisted, challenging the prevailing notion that modern tools and techniques had rendered his findings obsolete.

In the realm of economics, the pursuit of precision often overshadows the acknowledgment of imperfection. Economists, like the Ancient Greeks, worship at the altar of precision, disregarding the goddess of imperfection. The training of economists ingrains a relentless drive to minimize error, perpetuating a culture of disregard for measurement inaccuracies. However, I contend that ignoring these inaccuracies is not only intellectually dishonest but also undermines the integrity of economic analysis.

My fascination with measurement error grew as I encountered inconsistencies and anomalies in economic data. Rather than viewing these discrepancies as hindrances, I embraced them as opportunities for exploration and discovery. I became a proponent of recognizing and studying measurement error, advocating for its inclusion in economic discourse.

The publication of Jerven’s “Poor Numbers” in 2013 marked a turning point in my journey, reinforcing the global nature of measurement issues in economics. The biases in perceiving measurement problems as exclusive to the Global South became apparent, prompting me to emphasize the universality of inaccuracies in economic data. The need for transparency and accountability in reporting measurement error became increasingly evident, propelling me towards the realization of my long-standing aspiration – to write a book on the subject.

My book seeks to debunk the myth of economic data accuracy and champion a paradigm shift towards transparency. It highlights the pervasive nature of measurement error across economic observations, challenging economists to confront the uncomfortable truth of data imperfection. The scale introduced in the book offers a common language for assessing and reporting data accuracy, empowering users to navigate the murky waters of economic statistics.

The implications of my findings extend beyond academia, permeating into policy-making and research practices. Policy-makers must grapple with the uncertainties inherent in economic observations, recognizing the limitations of relying on flawed data for decision-making. Similarly, researchers must incorporate data sensitivity analyses and replication studies into their methodologies, fostering a culture of scientific rigor and accountability.

Transparency emerges as the cornerstone of addressing measurement error, with standardized reporting formats and accessible change logs proposed as essential tools for enhancing data reliability. By empowering data users to crowdsource error reporting, we can collectively strive towards a more accurate understanding of economic phenomena.

In conclusion, the journey towards economic progress necessitates a fundamental reevaluation of our approach to data accuracy. By embracing imperfection and prioritizing transparency, we can pave the way for more robust research, informed policymaking, and a deeper understanding of the complex dynamics shaping our world. It is time to liberate economics from the shackles of false precision and embark on a journey towards truth and enlightenment.

On the Inaccuracies of Economic Observations
Why and How We Could Do Better

Peter A.G. van Bergeijk, Professor of International Economics and Macroeconomics, International Institute of Social Studies, Erasmus University, The Hague, the Netherlands is available now.

Read the introduction and other free chapters on Elgaronline.