The Tax Profs blog has reported on a vigorous debate in the blogosphere over whether the tax code is too complex. Here are my two bits:
Very few people are qualified to use the tax code directly to resolve questions of tax law. I teach graduate students in personal financial planning for a living and empirically, I have discovered that it is a hit or miss proposition for them, and they help clients deal with the tax code as part of their professional duties. The percentage of lawyers who are capable of doing it is small -- as a young attorney in a medium sized law firm I routinely had non-tax expert lawyers come to me for the most basic information. For example, they routinely misunderstood how graduated tax rates work, assuming that moving into a new tax bracket affected the income earned in the prior tax bracket.
Fortunately, very few people use the tax code directly, so its drafting flaws are largely a nuisance for highly skilled tax specialists. The principal means by which ordinary people interface with the tax code are tax forms, instructions to tax forms, commercial guides aimed at lay audiences, and IRS publications, roughly in that order.
Tax complexity is not an equal concern for every taxpayer. Insurance companies, banks, multinational companies, securities brokers, retirement plan sponsors, muncipal bond offerors, and large oil companies, to name just a few examples, do not need a transparently simple tax code. They have highly skilled accountants and lawyers at their disposal who can devote massive amounts of time to understanding complex tax code provisions without unduly inflating the percentage of total revenues spent on tax compliance in large, complex enterprises. In contrast, a single mother who has dropped out of high school, has three kids, and is working at a non-professional job is far less equipped to deal with tax complexity. Thus, I'm far less concerned about complexity in interpreting Code Section 482 (transfer pricing) than I am about complexity in interpreting Code Section 32 (the earned income tax credit).
The number of complex tax provisions which affect your ordinary wage earner is modest and those can be reformed with modest effort -- important progress has been made recently on some of these points, for example, in the tax treatment of the sale of a residence and in the formulas used to compute required minimum distributions from retirement plans.
The number of complex tax provisions which affect small businesses is much greater, but could still be greatly thinned with modest political will (and even more modest distributional effects). For example, there is no reason that choice of entity decisions should be as complex as they are under the existing code (even though the check the box regulations have made vast improvements over the early days of the LLC).
The number of complex tax provisions which affect particular big businesses is vast, but irrelevant.
Complexity is not just, or even primarily, a matter of hard to read tax code sections. A code section whose meaning is exactly spelled out, but consists mostly of cross references, can be puzzled through in 15 minutes at most. Most of the work for tax professionals involves locating and utilizing situations in the code where transactions with similar substance can have significantly different tax implications with modest adjustments in form. These opportunities are largely a product of conceptual problems in how entire areas of the tax law are defined. For example, in my speciality, estate taxation, some of the biggest tax break opportunities come from a failure to formulate a good rule for taxing split interest transactions (QPRTS, GRITs, CRTs) (an incomplete gift treatment would produce more fair results) and from a failure to formulate sensible rules for valuing interests in closely held business entities (e.g. valuing minority interests in companies where family members are likely to cooperate at the same discounts that apply when third parties own minority interests).
One of the main sources of complexity in the tax code is a political unwillingness to be bold. Rather than adjusting tax brackets, we get tax deduction and credit phase outs. Rather than disallowing or limiting abusive deductions and credits, we get the alternative minimum tax. Rather than overhauling the rules for retirement plans generally, we get another half dozen types of plans to worry about.
The process that worked to reduce tax complexity in the 1986 Act needs to be repeated and will have to be similar to fast track treaty rules or the base closure process. A relatively small group of people need to formulate bold changes, with real but limited room for consultation on major issues, which then needs to be given an up or down vote on a package that has something for everyone. As observers noted after 1986, this is not necessarily bad for members of Congress if it happens all at once. Why? Because a member can't be blamed for a lost tax break and because once it has passed, every tax lobbiest in town has to return to you in an effort to get their own tax breaks restored in future years.
The tax code does need to be simplified because its complexity makes alternatives, like a national sales tax, which are bad policy, look artificially attractive.