Thursday, February 05, 2015

About those tax "cost" estimates

A client recently had a question about the current issue of Estate Planning Report:

I’m not sure what the dollar amounts mean in the second paragraph??  Tax cost the first year is $239 million and the ten-year cost is only $384 million.
These are the numbers that the Joint Committee on Taxation submitted as the lost revenue for allowing the tax-free rollover of funds from an IRA to a charity.   Here’s my source. You may recall that in December Congress reinstated the tax-free charitable rollover for a single year, and it has now expired again. They scored it as losing $239 million in the first year and from $12 to $19 million every year after that.  It’s line A8 of the table.

JCT does not explain their methodology or assumptions for specific line items. Frankly, the numbers make no sense to me.  I take it that they assumed the restoration was for one year only, that the charitable rollover would be repealed for years 2 – 10, so the later years cost far less. But then why don’t they cost zero?  How do they lose $19 million in year 10 for donations made in year 1?

But even worse, let’s unpack the numbers a bit to see what they mean.  By definition, the donations are coming from those over 70 1/2 and can’t exceed $100,000.  For the sake of round numbers, let’s assume that their tax rate is 23.9%.  To lose $239 million in one year, you have to assume that $1 billion would have otherwise been included in the income of these retirees, presumably as required minimum distributions.  To get to that number you need to have 10,000 retirees each make a maximum $100,000 charitable rollover contribution—that seems absurdly high to me.  Alternatively, 100,000 taxpayers could donate $10,000 each, but that still seems equally unlikely.  I doubt that there are that many IRAs large enough to sustain such large donations, especially just in the last two weeks of the tax year when such rollovers were allowed. Remember, only those over age 70 1/2 are even eligible.  Presumably only those with IRAs worth $1 million or more would consider such a large donation, and GAO reported there are only about 600,000 IRAs that large in the country.

But still worse than that, they also must have assumed that the affected seniors would not have exercised their right to simply take the RMD, give it to charity, and claim the deduction in the usual way!  That seems like the most unlikely assumption of all.  That’s why I concluded the paragraph with the notion that the JCT seems to be assuming that charitable gifts won’t be made at all in the absence of this provision.

Am I missing something?


JLM said...

Thanks for taking the time to analyze this silly tax cost projection. (Smart librarians put projections of future revenue gain or loss on the fiction shelves.)

Your estimated top tax rate for donors seems low. I should think people who can afford to give away $100,000 from their IRAs – on top of their routine charitable giving – would be in brackets of 33% and up. Most would have a marginal rate of almost 40%.

Jim Gust said...

I agree that my tax rate is too low, I picked it to avoid launching my calculator. But if we double my tax rate, we would simply halve the other numbers, and they still are way too big to achieve in the last two weeks of the year.

This isn't the first time the JCT projections have seemed outlandish to me. Makes me start to wonder about the legitimacy of the statistics of current costs of "tax expenditures."