The IRS doesn't love it.
The courts have never agreed.

Where the IRS drew its line.

On December 9, 2022 (released December 16), the IRS Office of Chief Counsel issued Generic Legal Advice Memorandum AM 2022-007, "Payment of legal fees to a third party." It used a hypothetical: the defense insurer sends the law firm's fee to a third party, the third party invests it in a portfolio, the deferred balance grows with investment returns, and the firm can borrow against it. Chief Counsel concluded the firm must include the fee in income in the year the money goes to the third party.

The memo rests on the assignment-of-income doctrine, the economic-benefit doctrine, and §83, and in the alternative on §409A (enacted in 2004, after Childs). It says Childs only decided §83 and constructive receipt, and distinguishes it: in Childs the defendant's insurers stayed on the hook for the payments and owned the annuities. It also notes the IRS never issued an acquiescence or nonacquiescence in Childs.

A memo from one lawyer to another. Not a regulation. Not a ruling.

To understand why AM 2022-007 didn't end attorney fee structuring, you have to understand where Chief Counsel memos sit in the IRS hierarchy:

• Treasury Regulations: issued under §7805 notice-and-comment authority. Binding on IRS and taxpayers; since Loper Bright (2024) courts no longer defer to agency readings under Chevron.
• Revenue Rulings: published positions of the IRS Commissioner. Binding on the IRS, can be relied on by taxpayers in good faith.
• Revenue Procedures: published administrative procedures. Similar weight to Rev. Ruls.
• Notices and Announcements: official IRS guidance, varying weight.
• Private Letter Rulings (PLR): taxpayer-specific. Binding on the IRS as to that taxpayer only; non-precedential.
• Chief Counsel advice and generic legal advice memos (CCA, GLAM): legal advice from IRS counsel to IRS personnel. Not precedent. Released publicly.

AM 2022-007 sits at the bottom of that ladder. It tells you how IRS counsel reads the law on its own hypothetical facts. It does not change the law. Childs, by contrast, is a precedential Tax Court opinion. Be clear about its limits too: the Eleventh Circuit affirmance was unpublished, so it is not binding circuit precedent, and the IRS has never acquiesced.

A memo is not a ruling, a regulation, or a court loss.

As of this writing, the memo has not been followed by:

• Audit letters challenging structured attorney fees on the constructive-receipt theory.
• Tax Court petitions filed by attorneys whose deferrals were disallowed.
• A Revenue Ruling formalizing the memo's position.
• Regulations under §451 addressing structured fees specifically.

We are not aware of a reported court decision agreeing with the memo's analysis, a Revenue Ruling, or a regulation. That can change, which is why the structure should track the Childs facts: payments fixed before the fee is earned, no investment account, no loan against the deferred fee, and an obligation that stays unfunded and unsecured as to you.

The real modern question isn't Childs. It's §409A.

Section 409A was enacted in 2004, a decade after Childs, and governs "non-qualified deferred compensation" arrangements. AM 2022-007 argues that if a fee deferral is unfunded, it is a nonqualified deferred compensation plan subject to §409A. The penalty for falling inside §409A and failing its rules is severe: immediate inclusion of all deferred amounts in income, plus a 20% additional tax, plus interest. So the design question on every modern attorney-fee structure is: is this arrangement non-qualified deferred compensation, and if so, does it meet §409A's requirements?

The dominant practitioner view is that a properly designed attorney-fee structure falls outside §409A because:

• The "service recipient" is the defendant or defense carrier, not the law firm. The plaintiff's attorney is not a "service provider" to the firm in the §409A sense for purposes of the contingency fee.
• The structured arrangement is set up before the lawyer has a vested right to a lump sum, so there's no "deferral" of an already-earned amount in the §409A sense.
• Even where §409A might arguably apply, the arrangement is structured to comply (fixed payment schedule established before service is complete, no acceleration, no subsequent deferrals).

This is fact-specific. A solo lawyer with a contingency case operating through a single-member PLLC has a different §409A picture than a fee award structured into a 50-attorney firm's partner draws. The point is: the live debate is here, not on constructive receipt. Childs handles the constructive-receipt question. §409A is the modern overlay that good structures plan around.

Low. Not zero. Manageable.

In our experience and the published practitioner literature: the audit rate on properly designed structured attorney fees is low. When audits do come, they tend to focus on documentation (was the assignment really pre-settlement? are the assignment company papers in order?) rather than the underlying legal theory. We are not aware of a reported decision since Childs disallowing a properly documented structure.

What materially raises audit risk:

• Structures designed after the settlement is already signed (the "retroactive" defect).
• Assignment papers signed weeks after the settlement closing.
• Self-dealing: the lawyer is also a beneficial owner of the assignment company.
• Payment streams that look like wages to the lawyer's own firm rather than annuity payments from a regulated carrier.

All of these are documentation and design failures, not problems with the underlying legal theory. The cure is to do it right.