Fifteen years after the HIRE Act, the official audits are in. The world's most ambitious tax transparency machine found the compliant, burdened the accidental, and left a three-million-person gap it is only now turning toward.
FATCA was signed on 18 March 2010 as the revenue-raising title of the HIRE Act, with a simple premise: make every financial institution on earth report its American clients, and hidden offshore wealth would have nowhere left to live. Fifteen years later, the machine exists, at a scale no tax administration had ever attempted, and the question worth asking is no longer whether FATCA works in principle but what it has actually produced. On that question, unusually, we do not have to speculate: the United States audits itself, publicly, and the reports of the Treasury Inspector General for Tax Administration and the Government Accountability Office, together with the State Department's renunciation statistics, form a data record most commentary never assembles.
This article assembles it. The numbers tell a double story that both FATCA's defenders and its critics tend to tell only half of: the data flows, massively and automatically, and discovery is real; and yet the enforcement machine behind the data ran years late, the burden landed disproportionately on the already compliant and the accidentally American, and a gap measured in millions of potential non-filers persists, documented by the government's own auditors.
Part I recalls what was built. Part II presents the data, with sources. Part III says what we think the data means. Part IV draws the consequences for anyone with a French-American financial life, which is where this desk lives.
I. The machine as built
A. Three layers and a stick
FATCA's architecture has three layers. Foreign financial institutions identify their US account holders and report them annually, on Form 8966, in most of the world through intergovernmental agreements: France signed a Model 1 IGA in 2013, which is why the reporting runs from French banks through the French administration to the IRS, and why the discovery letter arrives on bank letterhead. Individuals duplicate the disclosure on Form 8938, stacked on top of the older FBAR regime. And the stick that made the world comply is a 30% withholding on US-source payments to non-participating institutions, a sanction severe enough that no significant bank chose to test it.
B. The promise
The fiscal promise was quantified at birth: the Joint Committee on Taxation scored FATCA at approximately 8.7 billion dollars of additional revenue over the following decade, roughly 800 million dollars a year (JCT, JCX-5-10, 23 Feb. 2010). The compliance promise was grander: with institutional reporting universal, individual concealment would become structurally impossible, and the data would feed an enforcement program matching every reported account to a filed return. Fifteen years on, both promises can be checked against the government's own audits.
II. What the data says
A. The matching gap: TIGTA's numbers
The central audit is TIGTA's report of 7 April 2022, Additional Actions Are Needed to Address Non-Filing and Non-Reporting Compliance Under FATCA (report 2022-30-019). Its findings deserve quotation by number. The roughly 15 million Forms 8966 received for 2016 through 2019 contained 4,976,521 unique taxpayer identification numbers; over the same period, the IRS received 1,827,556 Forms 8938, leaving more than 3.1 million taxpayers identified by their banks who filed no corresponding individual disclosure. A majority of the bank reports, 56%, carried missing or invalid identification numbers, and IRS attempts to resolve the discrepancies failed 61% of the time. The auditors identified over 330,000 taxpayers with reported accounts above 50,000 dollars and no Form 8938, representing some 3.3 billion dollars in penalties that could have been, and had not been, assessed. And the frame around these numbers: having spent almost 380 million dollars implementing FATCA, the IRS abandoned its original comprehensive compliance roadmap for two narrow campaigns that, at audit date, had assessed no penalties.
B. The burden ledger: GAO's findings
The GAO's April 2019 report (GAO-19-180) audited the other side of the ledger: what the machine costs the people it watches. It documented that data quality and management issues limited the IRS's ability to use what it collects; that foreign institutions closed existing accounts of US persons or refused to open new ones rather than bear the compliance load, a phenomenon a Democrats Abroad survey had estimated at roughly a million closed accounts as early as 2014; and that the duplication between Form 8938 and the FBAR imposes an overlapping burden Congress should eliminate. The GAO's own follow-up page records that, as of February 2026, Congress has not acted on that recommendation. The report also connected the machine to its most visible human output: annual State Department approvals of citizenship renunciations rose from 1,601 in 2011 to 4,449 in 2016, an increase of nearly 178%.
C. The exits and the base: renunciations against population
The renunciation series extends the GAO's curve. Through the 2000s, fewer than 500 Americans a year gave up citizenship; after FATCA, the annual figures ran in the thousands, peaking above 6,700 in 2020, and Federal Register data analyzed in September 2026 shows the latest twelve months at their highest level since that record. Set against the base, the ratio is the finding: the State Department estimates roughly 9 million US citizens abroad, while the Form 8938 filings and the IRS's own matching suggest that only a fraction of the identified population files. The renouncers, who must certify tax compliance to exit cleanly, as we detailed in our article on accidental Americans, are drawn disproportionately from the compliant; the gap is made of everyone else.
D. The money: projection against collection
Against the 8.7 billion dollar projection, no official figure measures what FATCA's matching machine itself has collected, which is a finding in its own right. What the IRS has reported collecting, more than 10 billion dollars in tax, penalties and interest, came from its voluntary disclosure and streamlined programs: from taxpayers walking in, prodded by the machine's shadow, rather than from the machine's own enforcement. The National Taxpayer Advocate observed as early as 2013 that FATCA's compliance costs, borne by institutions and individuals worldwide, equal or exceed its projected revenue. The honest accounting, on the government's own numbers, is that FATCA's fiscal yield is real but indirect: it monetizes through the fear of discovery, not through discovery itself.
III. What the data means
A. Discovery is real; enforcement ran late
Two propositions, both true, must be held together. The data pipeline works: five million identified taxpayers is a number without precedent in tax history, and any French bank's American client should assume, as a matter of fact rather than prudence, that their account data sits in an IRS database. And the enforcement lagged by a decade: the audits show an administration that collected the haystack and only recently began building the magnets, with TIGTA's recommendations, accepted in large part by the IRS, pointing precisely at the 3.1 million mismatches and the unassessed penalty pool. A machine that has run fifteen years without systematic individual enforcement, and whose auditors have now mapped exactly where the enforcement should go, is not a machine that will idle forever.
B. The burden fell on the wrong population
The second meaning is distributional, and it is the one our practice sees weekly. The architecture was aimed at concealed wealth; its weight landed on ordinary residents abroad, on savings products that were never hidden, and on accidental Americans whose banking access it damaged, a dynamic the State Department itself acknowledged this year when slashing the renunciation fee. This is the pattern we described among the ten structural mismatches: an information regime decoupled from tax due punishes paperwork failure, and paperwork failure concentrates among the unadvised, not the fraudulent. The data confirms the structure.
C. The asymmetry stands
Third meaning: nothing in fifteen years has made the system symmetrical. The United States still stands outside the Common Reporting Standard, the duplication GAO flagged in 2019 remains unlegislated as of 2026, and the reform proposals that would exempt same-country accounts or harmonize the forms accumulate in annual reports without passage. The practical planning assumption must therefore be stability: the machine as described, with its gap and its momentum, is the machine for the foreseeable future.
IV. Consequences for the France-US taxpayer
A. Never plan on the gap
The tempting inference from 3.1 million untouched non-filers is that non-filing is safe. It is the wrong inference, three times over. The gap is mapped: the IRS holds the list, and the audits exist precisely because the agency has agreed to work it. The gap is asymmetric: it shelters the unidentified, while anyone with a TIN-matched French account sits in the smallest, most actionable segment. And the gap's repair instruments are discretionary: the streamlined procedures that convert fifteen years of this history into a penalty-free landing, which we detailed in our eligibility analysis, are administrative creations, closed to anyone once an examination opens and revocable for everyone whenever the IRS decides the amnesty era is over. Betting on the gap means betting that a documented, budgeted enforcement backlog will never be worked, against auditors publicly insisting that it be.
B. Use the window the data describes
Read as strategy rather than scorekeeping, the fifteen-year record describes a window: discovery universal, enforcement arriving, repair still cheap. Our practice draws the conclusion mechanically. Every French-American file begins from the assumption that the IRS already holds the account data, because it does; non-willful histories go through streamlined while it exists; renunciation candidates exit through the dedicated procedures with the certification the data era demands; and nobody, ever, is advised to be the test case for whether the 3.3 billion dollar penalty pool gets assessed. Fifteen years of data reduce to one sentence of advice: the machine knows, the machine is slow, and the only rational use of its slowness is to arrive compliant before it arrives at you.
Conclusion
Fifteen years in, FATCA's record is neither triumph nor failure but something more instructive: a discovery machine of unprecedented reach, an enforcement program a decade behind it, a burden borne by the compliant and the accidental, a revenue stream flowing through voluntary disclosure rather than through the machine itself, and a gap of three million identified non-filers that the government's own auditors have now mapped for action. Every number in that sentence carries an official citation.
For the French-American taxpayer, the data resolves the only question that matters: not whether you are visible, you are, but what to do with the interval between visibility and enforcement. The interval is the opportunity. It is spent either arriving at compliance on your own terms, at today's near-zero penalty cost, or waiting to learn what the backlog's first systematic campaign looks like. Fifteen years of official data say the first option has an expiry date and the second has a price. We know which side of that trade our clients should be on.
Frequently asked questions
Does the IRS really receive my French bank account data automatically?
Yes. France operates under a Model 1 intergovernmental agreement signed in 2013: French institutions identify their US account holders and the data flows annually through the French administration to the IRS. The scale is documented by the government's own auditors, roughly 15 million institutional reports covering nearly 5 million unique taxpayers for 2016 through 2019 alone. The correct planning assumption is that your data is already in the system.
If millions of identified taxpayers were never pursued, why should I regularize?
Because the gap is mapped, not forgotten. The 2022 TIGTA audit identified the mismatches, quantified 3.3 billion dollars in assessable penalties, and obtained IRS commitments to act on most of its recommendations. Meanwhile the repair routes, streamlined procedures at zero penalty for non-willful taxpayers, are discretionary programs that close the day an examination opens and can be withdrawn for everyone at any time. The rational use of enforcement lag is to regularize inside it, not to test its duration.
Has FATCA actually raised the money it promised?
Not in the way promised. The 2010 projection was about 8.7 billion dollars over a decade; no official figure measures collections from FATCA matching itself, and the 10-plus billion dollars the IRS reports collecting came through its voluntary disclosure and streamlined programs, taxpayers coming forward under the machine's shadow. Auditors and the National Taxpayer Advocate have observed that worldwide compliance costs equal or exceed the projected revenue. FATCA's yield is real but indirect: it works by inducing compliance, not by catching evaders one by one.
Are renunciations really driven by FATCA?
The correlation is documented, the causation partial. Renunciations ran below 500 a year through the 2000s, rose through FATCA's implementation to 4,449 by 2016 (a 178% increase from 2011, per State Department data cited by GAO), peaked above 6,700 in 2020, and remain near record levels in the latest twelve months of Federal Register data. The GAO links the rise in part to banking and compliance burdens, and the State Department's 2026 fee reduction acknowledged those difficulties. Renunciation remains a minority exit from a 9-million-person population, and it requires tax compliance to execute cleanly.