Income Splitting Among the Self-Employed

Which system of taxation on the self employed evokes the greatest tax distortions?

I) An income-splitting system where individuals may appeal to tax non-compliance; or,

II) A system of individual taxation where individuals may simply not report income.

In an interesting paper forthcoming in The Canadian Journal of Economics, University of Victoria professor Herbert J. Schuetze estimates (with a few caveats) that in Canada’s income-splitting system “…approximately one half of a billion dollars in taxes were avoided in 1998 by more than 90,000 businesses.”

An abstract from “Income Splitting Among the Self-Employed:”

Whether the individual or the household should be the unit of taxation is a long-running debate in the economics literature. One potentially important cost associated with a switch to individual taxation, which has been overlooked in this debate, is the impact of such a move on tax non-compliance. In particular, under individual taxation with progressive marginal tax rates households in which the distribution of income among household members is unequal benefit from attributing income from the higher to thelower income household member. The absence of a third party reporting income enables self-employed households to "split" income among family members to reduce income tax liabilities. Using the Canadian experience as a case study this paper sheds light on the magnitude and nature of this activity by developing a unique estimator of the incidence of illegal income splitting among couples. These estimates provide evidence that the occurrence of income splitting is likely non-trivial and suggest that the costs associated with this activity are potentially significant.

Schuetze draws a comparison between the U.S. and Canada. “The current US tax code, which primarily treats the household as the unit of observation, is such that the distribution of income within the household does not affect household tax liabilities.”

His findings:

…the raw reported employment rates of wives of self-employed men in Canada are significantly higher than those of their US counterparts. …. No significant difference in employment rates are observed among husbands…. I find strong supporting evidence that the employment differential found among wives is indicative of Canadian self-employed men attributing income to their wives.

He concludes:

…the pattern of income splitting found above suggest that countries that have high marginal tax rates, unequal wages across men and women and high rates of self-employment may find a system of joint taxation optimal, while for those with low incentives for income splitting individual taxation may be more appropriate.

Perhaps as equally interesting as his findings (at least to me) is the methodology he uses. He attempts to identify hidden behaviours, taking a lesson from Duggan and Levitt (corruption in wrestling) and Jacob and Levitt (cheating among teachers), among others.