Academic journal article Journal of Business Strategies

A New State-Level Panel of Annual Inequality Measures over the Period 1916-2005

Academic journal article Journal of Business Strategies

A New State-Level Panel of Annual Inequality Measures over the Period 1916-2005

Article excerpt

ABSTRACT

This paper introduces a new panel of annual state-level income inequality measures over the ninety year period 1916-2005. Among many of the states inequality followed a U-shaped pattern over the past century, peaking both before the Great Depression and again at the time of the new millennium. The new panel reveals significant state-level variations, both before the year 1945, and regionally. While Northeastern states are strongly correlated with aggregate U.S. trends, we find many of the Western states have little overall correlation over the past century. The availability of this new panel may prove useful to empirical researchers interested in all aspects of income inequality, particularly given the panel's unusually large number of both time-series and cross-sectional observations

INTRODUCTION

As the threat of war loomed, the 16th Amendment to the U.S. Constitution was ratified on February 3rd 1913, giving the U.S. Congress the authority to levy a federal income tax. (1) Congress followed by adopting a 1% tax on incomes of more than $3,000, with a surtax of 6% on incomes of more than $500,000. Since 1916, the Internal Revenue Service (IRS) has published income and tax statistics based on information reported on the Federal tax returns filed by U.S. individual taxpayers. These annual IRS publications provide unique and comprehensive access to the incomes of Americans over the past century.

The primary innovation of this paper is to use IRS income tax filing data to construct a comprehensive state-level panel of annual income inequality measures (the panel may be obtained online at www.shsu.edu/eco_mwf/inequality.html). Although IRS income data has several important limitations, including the censoring of individuals below a threshold level of income, it has the unique feature of being available annually for each state since the year 1916. Current empirical research on income inequality has benefited primarily from the construction of two prior income inequality data sets: the international panel of Deininger and Squire (1996), and the U.S time-series data of Piketty and Saez (2003). Deininger and Squire (1996) offer inequality measures for a wide panel of nations with several time-series observations for each nation beginning in the year 1960. These time-series observations are spaced over multiple decades, with an average of six observations per nation in their high-quality subset of the panel. (2) Piketty and Saez (2003), on the other hand, construct a high-frequency U.S. time-series data set. Unlike the large-N small-T panel of Deininger and Squire, the Piketty and Saez data contains up to 85 annual observations for the U.S. covering the period 1913-1998.

This paper seeks to contribute to the literature on income inequality by providing a third data alternative: a panel which covers an under-exploited unit of observation, U.S. states, and that is large in both cross-sections and time-series observations. While a panel of U.S. states is more homogenous than most cross-national panels, it still retains a useful degree of heterogeneity derived from each state's unique political/institutional history, and regional heritage. Moreover, a moderate amount of cross-sectional heterogeneity would appear to be a useful econometric feature, as the overwhelming cross-sectional heterogeneity in the international panel of Deininger and Squire (see Li, Squire, and Zou, 1998) has led to some econometric misuse, a point of emphasis in Quah (2001) and Partridge (2005).

Our new state-level panel shows that many states followed a distinctive U-shaped pattern over the past century, with inequality peaking both before the Great Depression and again at the time of the new millennium. This trend is consistent with overall U.S. trends (see Piketty and Saez, 2003), though we do uncover sizable state-level variability over time. This variability is particularly large before the year 1945, and appears to reemerge during recent years. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.