Exposing the Economic Elite: The Challenges of Measuring Wealth at the Top

The World Bank's report "Measuring the Upper Tail of the Income and Wealth Distributions" by Andrew Kerr and Mxolisi Zondi delves into the complexities of measuring top incomes and wealth in low- and middle-income countries. It reviews historical and recent methods, identifies challenges, and proposes solutions to enhance data accuracy. The report emphasizes the importance of better data collection, innovative methodologies, and leveraging administrative data.

Accurately measuring the upper echelons of income and wealth distribution, particularly in low- and middle-income countries (LMICs), is fraught with challenges. This insight stems from the detailed report "Measuring the Upper Tail of the Income and Wealth Distributions" by Andrew Kerr and Mxolisi Zondi, published by the World Bank. The report underscores the difficulties faced in capturing data on the wealthiest segments of society, highlighting both historical and contemporary methodologies, their limitations, and potential solutions.

Historical Foundations and Recent Advances

In 1953, Kuznets and Jenks pioneered the effort to measure top income shares in the United States using tax data, national accounts, and population estimates. Their groundbreaking approach laid the foundation for subsequent studies worldwide. Decades later, Thomas Piketty revitalized interest in this area with his analysis of top income and wealth shares in France, leading to the establishment of the World Top Incomes Database (WTID) and its successor, the World Inequality Database (WID).

Recent literature has built on these foundations, utilizing tax microdata to shed light on the distribution of income and wealth at the top. Piketty and his colleagues have demonstrated the critical role of such data in understanding economic inequality. The WID has provided valuable insights into the share of income held by the top 10% across various regions, revealing significant disparities.

The Hurdles of Accurate Measurement

Several challenges complicate the measurement of top incomes and wealth. High-income individuals often do not participate in surveys, skewing the data—a phenomenon known as unit non-response. Even when they do participate, these individuals may withhold specific income details, leading to item non-response. Measurement error is another issue, where inaccuracies in reported incomes, particularly at the top end, can distort findings. Sparseness in survey data, due to small sample sizes, often misses high-income households altogether.

Responses from less knowledgeable household members, known as proxy respondents, can introduce errors, while data processing mistakes, such as top-coding, can distort the top tail. In many low-income countries, surveys focus on consumption rather than income, leading to a lack of income data. Additionally, surveys may fail to capture any high-income individuals, resulting in a lack of common support and making accurate representation impossible.

Proposed Solutions

To address these challenges, the report outlines both ex-ante and ex-post solutions. Ex-ante solutions involve improving survey design and fieldwork protocols to reduce non-response rates and capture more accurate data from high-income households. Techniques such as oversampling rich households and using unfolding bracket questions in surveys are suggested.

Ex-post solutions include methods to adjust and improve existing data. These involve reweighting survey weights to account for non-response, using statistical methods to predict and fill in missing income data through imputation, inferring income distribution from consumption data, and replacing top incomes in surveys with values from a fitted distribution, such as the Pareto distribution.

Leveraging Administrative Data

Incorporating administrative data, such as tax records, offers another avenue to enhance the accuracy of measuring top incomes. This can be done through reweighting survey weights to match administrative data distributions, substituting survey data with administrative data for the top income brackets, and integrating both reweighting and replacement to refine estimates.

A Path Forward

The report identifies several gaps in the current research landscape, particularly the limited availability of income data in many LMICs and the lack of accessible tax data. It calls for harmonizing existing data, investigating non-response trends, and developing new methodologies to address measurement challenges. Improving access to and the use of tax and other administrative data is also emphasized.

Understanding the upper tail of income and wealth distributions is crucial for addressing economic inequality. The insights from "Measuring the Upper Tail of the Income and Wealth Distributions" provide a roadmap for improving measurement accuracy, highlighting the need for better data collection, innovative methodologies, and increased collaboration between researchers and policymakers.