Regulating Paid Tax Return Preparers

Since at least 2002, the IRS National Taxpayer Advocate (NTA) has pushed for regulation by the IRS of "unenrolled" preparers (individuals who are not attorneys, CPAs or Enrolled Agents). Several studies have found intentional and unintentional (sloppy and uninformed) errors by tax preparation offices. The IRS held public hearings in 2009 and issued a 57-page report of findings in December 2009 on the need to and how to regulate all paid preparers. Parts of the plan have been rolling out for the past few months and registration is expected to start this month (September 2010)!

The reach is quite broad. Anyone involved in preparing a federal tax return even if they are not the signing preparer will need to register and pay the annual $64.25 fee. If that person is not an attorney, CPA or EA, they will also need to have at least 15 hours of continuing education annually and pass a test relevant to the type of returns they prepare. So, as currently drafted, individuals who are not CPAs or EAs but work for people who are, will need to become a "registered tax return preparer" (RTRP) - a new designation we will likely be seeing after a registered person's name. Such individuals also become subject to rules of conduct established by the Treasury Department (known as "Circular 230").

A big current issue is whether someone working for a CPA, who is already quite regulated by their state licensing board and the IRS and any professional organization they belong to, such as the AICPA, should have to become an RTRP. The AICPA has argued no, but this week, the National Society of Accountants argued yes. The NTA has also argued that anyone involved in preparing the return, even if they don't sign it needs to be registered.

What is the relevance to good tax policy? A few ideas come to mind. First, is the principle of simplicity. If the tax law were not so complicated, we probably would not have so many paid preparers and so many unintentional errors. A complicated tax law also raises the question of just what should be on a test to determine if someone should be preparing another's tax return? Such a test should be open book - does the preparer know how to use appropriate tax research tools to resolve tax issues? Do they have a broad and deep enough knowledge of the tax law to be able to identify the tax issues that need to be researched? When the IRS described the content of the likely tests, they did so by form number rather than by Internal Revenue Code section, thus implying (in my mind), less understanding of the tax law would be necessary to pass the exam. For example, when people study for the CPA exam, they are not studying tax forms, but the tax law itself. [See page 53 of the IRS' 57-page report for the list of tax forms likely to be addressed on the tests.]

Another tax policy consideration is a minimum tax gap. I don't think the proposed system will cause unscrupulous preparers to change behavior. They will just become (or continue to be) paid preparers who do not sign the return so remain off of the IRS radar screen (until the taxpayer gets audited and leads the IRS to the preparer). Recall some of the awful stories that have been in the papers in the past year about the first-time homebuyer credit claimed on paid preparer returns where the taxpayer did not qualify (see my 2/27/10 post). If Congress and the IRS want to reign in these unscrupulous preparers, they really need to (unfortunately) add a very large penalty to the tax system to be imposed on taxpayers who file a return they paid someone to prepare and did not get that preparer to sign the return and list his identifying number. Such a penalty would keep taxpayers away from unscrupulous preparers who would then go out of business.

Also, a complicated tax law, even with testing that many preparers are likely to pass, is still going to result in errors on returns.

I've got an article on this topic in the AICPA Tax Insider (9/9/10) - here. The article has links to IRS information on this topic.

What do you think? Is the proposed regulation system the best way to go? What might be better?