This section attempts to deal with questions often posed.  We hope this will clarify some issues about our income tax system and illustrate some uses of the data in this report. It also demonstrates how some easy questions need to be clarified and refined.


Why do non-residents pay taxes and what do you know about them?


There are three groups of non-residents: those who lived in the state when they filed their return but did not live here all year, those who lived here part of the year but no longer live here, and those who never lived here. Non-resident is really a misnomer including those who were part year residents. Those who lived here (had a Utah address) when they filed were about 33% of the non-residents. We are not able to distinguish accurately those who no longer lived here from those who never lived here. It is likely, however, those who lived here for only a very short period or who never lived here will have a small portion of their income from Utah.  If we guess that those with less than 20% of their income from Utah are real non-residents, then they make up about 26% of the non-residents. Using 30% as the criteria, they are 32% of the non-residents. (The above data are based on a detailed examination of 2000 returns)



How many taxpayers pay Utah taxes?


This frequently posed question may sound very simple, but it is not. There are a lot of taxes and in one view everyone pays at least the sales tax and probably the fuel tax, as well as many tourists and shareholders. Since this book is about the income tax, we will limit the question to the income tax. Looking at table 1, line 1, in 2001 we had 960,310 returns filed with 882,270 being filed by full year residents (line 11). These figures do not tell the number of taxpayers but only the number of returns. If you have a traditional family with three kids, do you want to count five taxpayers, two taxpayers, counting only adults, or one for the family. Using table 1 again, it would be possible to count two for each married filing joint returns and one for the other returns.


But all these returns do not pay taxes, since deductions, exemptions, or elderly tax exemptions may cause the income that is taxable to be so low no taxes are paid. On table 3 for full-year residents, you will find that 714,016 returns showed a tax liability (before credits).  All full-year residents with liabilities had a mean tax liability of $2,077 and a median of $1,234.


What is the average income in Utah?


The full-year resident part of table 3 shows that the mean adjusted gross income was $43,577 in 2001, but the median, or middle income, was considerably less at $28,912. If you are interested in a typical resident, it may be better to use the median, because very large or atypical amounts will not distort the picture.  We must caution you about several special features of tax data that cause it to be different for other data sources. The figures above are on a per return basis, and are not on a per capita or per household basis. (We do have a partial attempt to present less detailed statistics on a family basis but the calculation has serious shortcomings.)


There are also many other differences in concept from other sources and we will only highlight a couple. One major difference is that tax data will include capital gains, which have been very significant recently, are skewed towards higher income taxpayers, and are erratic from year to year.  We do not know of any other popular source that includes them. Another omission from tax data that is complicated is social security benefits. For most taxpayers they are not included, and for others they are partially included, but are not totally included for anyone. Other pension income is only included if it exceeds any taxpayer contribution.


While Census and other figures will represent the whole population, our data is only based on those who file tax returns, thus low-income residents are under-represented. Since taxes are based on income reported, there tends to be an under-reporting of income on some tax forms. Income that is actually reported is probably more accurate since it must be documented.