Since 2008 IRS offshore crusades added 100,000 ordinary new filers who don’t owe any U.S. tax

Via TaxProf Blog, we learn that the IRS has released its annual report on Individual Income Tax Returns, 2012. Page 9 has statistics on Foreign Earned Income Exclusion usage, from which we can calculate the average amount of the FEIE per return:

	2008	2009	2010	2011	2012
Number of returns w/FEIE	371,885	396,405	415,519	445,276	475,386
		+6.59%	+4.82%	+7.16%	+6.76%
Total amount of excluded income (constant 1990 US$ million)	13,899	14,907	15,482	16,305	16,866
		+7.25%	+3.86%	+5.31%	+3.44%
Average FEIE per return (constant 1990 US$)	37,374	37,604	37,258	36,618	35,478
		+0.62%	-0.92%	-1.72%	-3.11%

Here’s another way of looking at these numbers. Assume that all of the FEIE users in 2008 continued to be FEIE users in 2012 (or equivalently, that they were replaced by similar filers or married-filing-jointly couples), and that over that period they suffered the same drop of 0.96% in earned income (in constant 1990 US$) that all U.S. returns demonstrated on average during that period. What would the average FEIE for each of the 100,000 marginal new filers have to be in order to fit with the above figures?
Answer (again, in constant 1990 US$): US$29,709, almost exactly the same as the average earned income across all U.S tax returns.

Few “wealthy expats” among new filers or non-filers
The average amount of FEIE has always been higher than the earned income on the average U.S. return. Homeland pundits often take this as confirmation of the stereotype of “wealthy expats getting great tax breaks”. However, the more likely explanation is that well-paid folks often have tax advice to match their wages (e.g. thanks to tax assistance provided as part of corporate international assignment packages) through which they learn of their U.S. filing requirements and hence show up in FEIE statistics, while people of more average means — whether expats in less high-flying jobs, or ordinary employees living in the countries they think of as home — don’t show up in FEIE statistics because they don’t know they have to file.
Of course, the total number of FEIE users remains less than a tenth of the estimated number of “U.S. Persons” outside of the U.S., and the number of FBAR filers is not much higher, either. This suggests that the IRS’ shock-and-awe OVDP-style publicity about “offshore tax evaders!” is doing a piss-poor job at making people aware of their filing requirements.
Naturally, the ranks of millions of non-filing “U.S. Persons” outside of the U.S. will include some financially successful “accidental Americans” from whom the IRS could extract significant taxes, if it can find them and get the cooperation of their governments. But in general, the increase in FEIE filers in the past 5 years demonstrates that average new filers are likely to be pretty average, and won’t owe the U.S. any taxes on their wages.
Alternative ways to extract money
The IRS might be able to get some money out of these poor folks deemed to be “U.S. Persons” by taxing unrealised gains in their retirement plans (which would be tax-free if they were U.S. plans). However, an average wage earner’s gains in such a plan are likely to be less than their U.S. personal exemption & standard deduction, meaning they wouldn’t owe any U.S. tax on those items either.
In other words, the only way the U.S. government can extract money from average “U.S. Persons” outside of the U.S. is to assess wildly disproportionate fines for their failure to report those plans as “offshore trusts”, or to go ahead with that rumoured hike that could take renunciation fees from their current level of $450 (already expensive compared to other first-world countries) to US$ 1,500 or beyond.
The Joint Committee on Taxation ignores the existence of the FTC and computes the “cost” of the FEIE as the tax that would be charged if all of the FEIE’d income were instead U.S. taxable income. They released their latest estimates earlier this month:
https://www.jct.gov/publications.html?func=startdown&id=4663
Their projections, of course, assume that the future growth rate in the number of FEIE users will resemble the past growth rate. But with FATCA, there’s probably going to be a huge one-time jump in the number of filers that their projections don’t take into account: people who learn about their filing requirement for the first time. When those people start filing, the JTC will claim that the “cost” of the FEIE is exploding, both because there’s more people taking it and because they think that future usage will continue to grow at a similar rate.
E.g. their last estimate of the “cost” was about $40 billion total for 2014–18. But if the number of filers is 4x what they expect, then they’ll claim the “cost” is $160 billion over those five years. By my eyeballing, that would make it one of the top 20 “tax expenditures”, beating out (for example) the $250k capital gains tax exemption for the sale of a primary residence. And of course it would be a much more tempting target for repeal than any of those other top 20, since those benefit constituencies that Congresscritters understand and generally sympathise with (e.g. highly-leveraged real estate investors, AKA “homeowners”).