Do tax codes hold the key to income inequality and wealth trends?

From Paris to San Francisco, Geneva to New York, talk of taxing the wealthy seems to grow louder by the day on both sides of the Atlantic.

For governments facing gaping budget deficits, taxing top earners can look like a politically palatable solution to fiscal woes. It can also seem like the right thing to do: after all, global inequality has soared in recent decades, with the rich getting richer and the poor getting poorer. Wealth taxes not only bring in extra funds, the argument goes, but may also help create fairer societies.

But those arguments often omit the knock-on effect of new taxes in areas such as investment. In the debate over wealth taxes and income inequality, the devil is in the details – or the tax code.

Our recent research in the Journal of the European Economic Association homes in on earnings from business ownership, the alternative ways this type of income may be taxed and the distortive effect this can have on inequality statistics. This research was conducted together with Wojciech Kopczuk of Columbia University, Annette Alstadsæter of the Skatteforsk Centre for Tax Research and Kjetil Telle of the Norwegian Institute of Public Health.

It considers how high-net-worth individuals generate an outsized portion of their income from ownership in companies. For the top 0.1% of earners, for example, this type of income can account for about 80% of their earnings, compared to 15% at most from wages.

We examined Norway, one of the most egalitarian countries in the world, to understand the impact of corporate income and the tax code on measured income inequality. In 2006, Norway introduced a 28% dividend tax; prior to that, dividends were exempt.

From zero to 28% is a major change. It is also a big incentive to retain earnings within companies, rather than to distribute them as dividends. Not surprisingly, dividend payments hit an all-time high in 2005 in anticipation of the tax hike (something similar seems to have happened recently in France, where a supertax on the rich raised only a fraction of what was expected).

Once the tax reform took effect, companies in Norway began retaining more of their earnings, rather than paying them out as dividends. Those dividends that were distributed were increasingly paid into holding companies, income which was not reported on personal tax returns.

This is where the intricacies of the tax code matter so much. The most common way to tax corporate income is on a realized basis. This is where the income is reported in tax returns and 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 dividends. Only when businesses are run as partnerships do profits – irrespective of whether the profits are paid out or reinvested –show up in personal income tax returns.

If you measure income inequality based solely on income tax returns, ignoring those retained earnings, you find a drop in income and a narrowing of income inequality after the reform. But if you account for the retained earnings that businesses were now keeping in their corporations – a truer and more stable measure – then income inequality does not change much.

We were able to make the comparison because Norway has a comprehensive shareholder register, allowing us to look at what individuals reported in their income tax returns and, separately, at what their share ownership in companies was.

Unfortunately, most countries don’t have corporate shareholder registers like Norway’s. This means our understanding of trends in income inequality, in particular around large reforms, as well as the many changes in dividend taxation, will remain partial.

These findings contribute to a recent global debate around measuring inequality. For decades, the work of three French economists concluding that income inequality in the U.S. was soaring was accepted as gospel.

But using a different methodology, Gerald Auten and David Splinter concluded that post-tax income inequality has remained largely unchanged since the 1960s. For Auten and Splinter, 1986 was a pivotal year in the U.S. because the Tax Reform Act changed some reporting rules for top earners. After 1986, many corporate owners were able to elect a tax regime that taxed their business profits as a partnership. Hence, business income, even that not paid out as dividends, began showing up as income in tax returns.

This means that a potentially large part of the increase in income inequality in 1986 could simply reflect the change in reporting: after 1986, accrued income appears in tax returns, prior to this it didn't.

This is not to say that income inequality isn’t on the rise. But it’s important to bear in mind that it’s exceedingly complex to measure over time or to make cross-country comparisons. It’s normal to rely on individual tax returns as a primary source of data, but it’s also crucial to understand that returns may paint a partial or misleading picture.