IESE Insight
How taxation muddies the waters of income inequality measures
Tax reforms can cause large shifts in reported income while underlying inequality remains stable.
By Martin Jacob
Is income inequality widening or not? In a simmering debate over how to measure earnings trends, the tax code has become a defining factor behind conflicting results.
On one side of the debate are three French economists — Thomas Piketty, Emmanuel Saez and Gabriel Zucman — whose seminal work showed that income inequality in the U.S. has surged since the 1980s. The findings gave foundation to more than a decade of academic research and public policy on ways to narrow the widening gap.
On the other side are Americans Gerald Auten and David Splinter, who, using a different methodology, found that income inequality in the U.S. has barely budged since the 1960s. Auten and Splinter focus on a 1986 tax reform, which, among other things, changed the reporting rules for certain types of business income. Because of the new rules, income that previously wasn’t reported in individual tax returns was now there — making the wealthy seem wealthier.
While discussions of this kind tend to fester in academia, this one has received wide coverage in recent years because of its wide-reaching social and policy implications. Proposals to hike taxes on the wealthy in many countries, for example, are predicated on the idea that the rich are richer than ever before.
We step gingerly into this transatlantic debate with a case study from Europe that also highlights how the tax treatment of business income may skew broad measurements of income trends. The research, with coauthors Wojciech Kopczuk, Annette Alstadsæter and Kjetil Telle, was published in the Journal of the European Economic Association.
How Norway’s tax reform hit top earnings
Our research focuses on Norway, for two reasons: In 2006 the country began taxing income from dividends at 28%, whereas previously those earnings had been exempt. Additionally, Norway has a comprehensive shareholder register, which gave us detailed information on who owns shares in companies, independent of tax filings.
As could be expected, dividend payouts hit a record high in 2005 ahead of the tax hike. In the years before, firms paid high dividends and owners then decided to reinvest the money or to spend on consumption.
After 2006, companies began retaining most of their earnings rather than paying them out in dividends; the dividends that were distributed were increasingly paid into holding companies rather than to individuals.
Here’s where we get into the nitty-gritty of the tax code. The most common way to tax corporate income is on a realized basis — meaning it’s reported in tax returns and thus taxed when dividends are paid or shares are sold. Alternatively, taxing on an accrued basis requires taxpayers to report earnings from businesses regardless of whether the company retains earnings or distributes profits. This is how partnerships are taxed.
If you ignore retained earnings and dividends paid into holding companies, and only examine individual tax returns, you find that income dropped and inequality narrowed after the reform.
But were top earners genuinely poorer? If you account for the retained earnings, incomes and income inequality were largely stable.
Difficulties of measuring income inequality
The complexities of tracking income inequality over time are myriad, especially when tax codes everywhere undergo almost constant revision. Researchers must also make consequential methodological choices on how to account for employer-provided benefits and government welfare spending, among other issues.
It’s not surprising that most measures rely on individual tax returns as a primary source of data, but it’s important to remember that they reflect the tax code at a given point in time. Unfortunately, very few countries have the sort of shareholder register that Norway has.
The Nordic experience reveals the challenges of tax reform that reduces inequality on paper — but perhaps not in reality. While tax overhauls can produce large, sudden shifts in reported earnings, that’s not the same as a long-term, underlying trend in income levels.
ALSO OF INTEREST:
