by Scott Beesley
This article was originally published in the Winter 2002/03 issue of the Expert Witness.
A common problem encountered when attempting to estimate the incomes of the self-employed is that those individuals are able to under-report their revenues and over-report their expenses. Many techniques have been suggested to correct for this bias in reporting, but most such techniques are ad hoc in nature. Also, as most techniques only attempt to adjust expenses (to remove personal expenses from reported business expenses), they are unable to capture the effect of unreported revenue.
Consider the example of a small-scale home renovation contractor. The vast majority of his customers have no use for a receipt, since the work is personal rather than a business expense. When the purchaser is having work done on a rental property, a receipt may be requested, but even then the under-the-table transaction may make both parties better off. (In some cases a rental property will show a loss from mortgage interest, utilities and property taxes alone, so the added “loss” resulting from claiming a renovation expense might be redundant to the landlord.) A second, more general example is the existence of one-off barter deals and bartering systems. If our self-employed renovator can make a deal with another provider to trade $5,000 worth of services, the result is that each under-reports revenue by $5,000. In addition, each probably claims GST input credits on any materials purchased, while not collecting any GST on the sale. Thus there is good reason to look for a “true income” estimator which builds in the possibility of non-reported revenue
Recently, Dr. Herb Schuetze, of the University of Victoria, has developed an objective method for estimating the average amount of under-reporting undertaken by the self-employed. (See Herb Schuetze, “Profiles of Tax Non-Compliance Among the Self-Employed in Canada” Canadian Public Policy, June 2002.)
He uses data regarding family expenditures on food to impute the true income level of self-employed persons. His fundamental assumption is that the fraction of true family income devoted to expenditures on food is the same for the self-employed as for wage earners. A second presumption is that, since the amounts involved are small, persons reporting family expenditures will not worry that their tax evasion will be revealed because their food consumption is inordinately high. Thus, they are assumed to report their food expenditures correctly. By assuming that families of the same size who report the same expenditures on food will have the same incomes, Dr. Schuetze is able to calculate the “true” incomes of self-employed individuals by comparing their food expenditures to the expenditures of wage earners.
For example, assume that most four-person families with after-tax income of $50,000 spend $10,000 on food. If we observe that a family headed by a self-employed individual spends $10,000 on food, it might reasonably be assumed that that family’s income was $50,000 (after-tax). Hence, if that individual had reported only $42,000 of income, Dr. Schuetze would conclude that that individual had under reported his or her income by $8,000.
Employing this general approach, Dr. Schuetze controls for (takes into account) various household characteristics, such as the level of education of the head of the household, the age of any children, the region in which the family lives, and the value of their house if it is owned. The study looks at data from 1969 to 1992, at six points in time.
The results indicate that non-compliance among the self-employed was significant enough to be worthy of further study, and future added attention from the Canada Customs and Revenue Agency (CCRA). The estimates cover families in which at least 30 percent of reported income was generated through self-employment. For those families in which either the husband or wife, but not both, were self-employed, it was estimated that 12 to 21 percent of income was not reported. The figure was significantly lower when both spouses were self-employed (you cannot income-split both ways!).
Another interesting conclusion is that there was no pattern across time and across educational groupings, but there was a significant variation across different occupations. The construction and service industries had the highest degree of non-reporting, whereas product fabricating apparently afforded the least opportunity for evasion. (This is in complete agreement with Economica’s experience in performing loss assessments over the last few years.) The reported income of those in construction had to be multiplied by a factor of 1.46 to estimate true income, implying that a full 31.5 percent of total income (= 0.46/1.46) went unreported. When we consider that that figure is itself based on a sample including households which received as little as 30 percent of their reported income from self-employment, we conclude that for those who are predominantly self-employed, the applicable multiple must be much higher. Indeed, in those cases in which the self-employed person’s reported income is roughly zero, (and we have encountered a surprising number of these) the multiple is infinite!
Finally, we note that Dr. Schuetze grants that his measurement of the income from self-employment is in fact in error (biased downwards) because of exactly the issue being studied! That is, if a family reports $30,000 in earned salary and $10,000 from self-employment, the apparent share from self-employment is 25 percent. At that level they would not have been included in the self-employed pool, for the purpose of his study. (He required at least 30 percent of income to be from self-employment before considering an individual to be self-employed.) If, however, that couple had another $20,000 in hidden income, the true fraction from self-employment is 50 percent. I suspect that the result is that the overall “fraction of income hidden” estimates are, if anything, conservative.
Dr. Schuetze’s paper points out that because of the steady increase in the fraction of the population in self-employment, tax non-compliance is becoming more important over time. He suggests that the results of his analysis may be helpful in identifying occupations and demographic characteristics associated with non-compliance. The article certainly establishes that a marked level of tax non-compliance is not at all unusual among the self-employed in Canada.
Scott Beesley is a consultant with Economica and has a Master of Arts degree (in economics) from the University of British Columbia.
Hanoun Multicervical Rehabilitation Unit (MCU) as an adjunct to traditional forms of assessment and rehabilitation. This leading edge, digital technology provides comprehensive objective and valuable cervical functional diagnostic data. It is used to effectively and objectively assess cervical range of motion and neck strength. Together, these measurements quantify the functional capacity of the neck. This differs from other functional capacity evaluations, in that this technology allows us to specifically measure the functional ability of the injured neck. Once functional deficits have been identified and diagnosed, the system can then be used to efficiently and accurately rehabilitate the injured area(s).