The Equal Access to Justice Act, 28 U.S.C. § 2412, is the principal vehicle for shifting attorney's fees and costs from a plaintiff to the United States in immigration mandamus cases. This is a detailed analysis of the doctrine — the four prerequisites for recovery, the post-Buckhannon "prevailing party" problem that defeats most casually drafted mandamus settlements, the substantially-justified defense, the statutory hourly cap and special-factor enhancement, the pro se limitation under Kay v. Ehrler, and the procedural mechanics of a fee petition.
The Equal Access to Justice Act was enacted in 1980 to deter the federal government from imposing unjustified burdens on private parties by forcing them to litigate in cases where the agency's position was not substantially justified. Congress's animating concern was that ordinary citizens — and small businesses — could not afford to vindicate their rights against a federal agency with effectively unlimited litigation resources. EAJA shifts that economic asymmetry by making the United States pay the prevailing party's reasonable attorney's fees and expenses in certain civil cases.
In immigration mandamus litigation, EAJA is theoretically a perfect fit. The mandamus plaintiff is, almost by definition, an ordinary individual who has been forced into federal court because an agency would not perform a non-discretionary duty. The government's pre-litigation position — refusing or failing to adjudicate — is often impossible to justify. And the underlying case is generally short, factually clean, and procedurally simple, meaning fee awards are modest and manageable.
In practice, however, EAJA recovery in mandamus litigation is the exception, not the rule. The reason is almost entirely doctrinal: the Supreme Court's 2001 decision in Buckhannon Board & Care Home, Inc. v. West Virginia Department of Health & Human Resources, 532 U.S. 598, abolished the "catalyst theory" of prevailing-party status. Because the typical immigration mandamus action ends not with a judgment but with USCIS simply adjudicating the underlying application — mooting the case — the plaintiff frequently obtains the relief sought without ever becoming a "prevailing party" in the statutory sense. The result is that the plaintiff effectively wins the war but cannot collect fees for fighting it.
This article explains the EAJA framework in detail, with particular attention to the structural traps that defeat fee recovery in mandamus actions and the few drafting and procedural choices that preserve it.
The Equal Access to Justice Act, codified at 28 U.S.C. § 2412 and at 5 U.S.C. § 504 (the parallel administrative-proceedings provision), creates two distinct fee-shifting mechanisms relevant to federal court litigation. Both run against the United States. Both are best understood as carefully bounded exceptions to the American Rule that each party bears its own attorney's fees absent a specific statutory or contractual basis for shifting.
Subsection (a) authorizes the recovery of "costs" against the United States in any civil action brought by or against the federal government, on the same basis as would apply between private parties. "Costs" here are the routine taxable costs enumerated in 28 U.S.C. § 1920 — filing fees, printing, witness fees, copying — not attorney's fees. § 2412(a) recovery is the more permissive of the two avenues and does not require a "prevailing party" showing in the same demanding sense as § 2412(d). It also does not require that the government's position have been substantially unjustified.
Subsection (b) waives the United States' sovereign immunity for fee awards "to the same extent that any other party would be liable under the common law or under the terms of any statute which specifically provides for such an award." This subsection encompasses the common-law exceptions to the American Rule — most relevantly, the bad-faith exception — and applies them against the United States. In practice, § 2412(b) is rarely invoked in mandamus actions, because proving the government acted in objective bad faith is far harder than the § 2412(d) "not substantially justified" standard.
Subsection (d)(1)(A) is the operative provision in nearly every mandamus EAJA case. The text reads:
Four elements appear in that sentence and govern every § 2412(d) analysis:
Subsection (d)(2) defines the operative terms: "fees and other expenses" includes reasonable expenses of expert witnesses and "the reasonable cost of any study, analysis, engineering report, test, or project which is found by the court to be necessary for the preparation of the party's case," together with reasonable attorney's fees. The same subsection imposes the now-famous statutory rate cap: attorney's fees may not be awarded "in excess of $125 per hour unless the court determines that an increase in the cost of living or a special factor, such as the limited availability of qualified attorneys for the proceedings involved, justifies a higher fee." Section 5 of this article deals with that cap and how courts apply it.
Subsection (d)(1)(B) specifies what the fee applicant must file and within what deadline. The applicant must, within thirty days of final judgment, submit to the court a fee application that:
The thirty-day clock is jurisdictional in some circuits and treated as a non-jurisdictional but mandatory deadline in others, but in either case it is rigid. Scarborough v. Principi, 541 U.S. 401 (2004), holds that the substantive allegation that the government's position was not substantially justified can be amended after the thirty-day period, so long as a timely application was filed; but the deadline for filing some application is not extendable absent equitable tolling.
A § 2412(d) fee application succeeds only if the court is persuaded that each of four conditions is met. The first three appear in the statutory text; the fourth (the eligibility cap) is in § 2412(d)(2)(B). Take them in order.
No element of EAJA does more practical work in mandamus litigation than the prevailing-party requirement, and no Supreme Court decision more decisively shapes its application than Buckhannon.
Buckhannon involved the fee-shifting provisions of the Fair Housing Amendments Act and the Americans with Disabilities Act, which use the phrase "prevailing party." Plaintiffs sued for declaratory and injunctive relief; before the case was decided, the State changed the law that had been challenged, mooting the case. The plaintiffs argued they were prevailing parties because the lawsuit had been the "catalyst" for the legislative change. The Supreme Court rejected the catalyst theory. It held that "prevailing party" requires "a material alteration of the legal relationship of the parties" effected by a "judicial imprimatur on the change" — typically a judgment on the merits or a consent decree.
Although Buckhannon construed FHAA and ADA fee provisions, every federal circuit has applied its reasoning to EAJA. The result is that a mandamus plaintiff whose case is mooted by USCIS adjudicating the underlying application — the most common outcome — is generally not a prevailing party for EAJA purposes, no matter how clearly the lawsuit caused the adjudication.
Under Buckhannon, the following typically suffice to confer prevailing-party status:
Under Buckhannon, the following typically do not suffice:
The structural difficulty for mandamus plaintiffs is now obvious. The remedy sought in a mandamus action is, by design, an order requiring the agency to perform a duty. The agency almost always responds by performing the duty, before the court has any occasion to order it. The case is then dismissed as moot. Under Buckhannon, the plaintiff has not "prevailed" in the statutory sense, even though the plaintiff has functionally won the lawsuit.
This is sometimes called the "Pyrrhic mandamus" problem. The plaintiff achieves the practical relief sought — adjudication — but cannot recover fees because the relief was not delivered through court order. Two responses are available, and both require deliberate drafting at the settlement stage:
The bottom line: drafting the settlement matters more than litigating the merits when EAJA recovery is on the table.
Assume the plaintiff clears the prevailing-party threshold. The next battle is over whether the government's position was substantially justified. The burden here is on the government, not the plaintiff. The standard is set by Pierce v. Underwood.
Justice Scalia, writing for the Court, defined "substantially justified" to mean "justified in substance or in the main — that is, justified to a degree that could satisfy a reasonable person." The Court rejected the more demanding "justified to a high degree" formulation. A position can be substantially justified even if it was wrong; the question is whether it had a reasonable basis in law and fact. The standard is more demanding than "frivolous" but less demanding than "preponderance" or "more likely than not."
Crucially, the Court also held that "position of the United States" — the focus of the substantial-justification inquiry — encompasses both the agency's litigation position and the underlying conduct that gave rise to the litigation. 28 U.S.C. § 2412(d)(2)(D) codifies this rule.
In a mandamus case, the underlying conduct is the agency's failure to adjudicate. The litigation position is whatever the government argued in court — typically a motion to dismiss asserting that the delay is reasonable under TRAC, that the case is not ripe, or that the case is moot. Both must be substantially justified for the government to defeat fee recovery.
The government will commonly defend its pre-litigation conduct by pointing to:
Courts have varied substantially in their receptivity to these defenses. The dominant pattern is this: where the pre-suit delay is significantly outside published processing times, where the plaintiff exhausted informal remedies before suing, and where the government can point to no individualized reason for the delay specific to this applicant's case, courts tend to find the agency's conduct not substantially justified. Where the government can identify a specific, individualized factor — a pending RFE, an open investigation, a security-related hold — courts tend to find the conduct justified, even if the overall delay was long.
Even where the pre-suit conduct survives substantial-justification review, the litigation position may not. If the government files a motion to dismiss arguing the case is moot — and a court denies it — the government has lost on its litigation position, and that loss is some evidence (though not conclusive) that the position was not substantially justified. Conversely, if the government concedes early, withdraws disputed arguments, or adjudicates the application without contesting the merits, the litigation-position prong is harder for the plaintiff to win.
An important procedural point: once a court determines that the government's position was not substantially justified for purposes of the underlying litigation, no separate substantial-justification finding is required for fees incurred in litigating the fee petition itself. In other words, fees-on-fees are available without a second round of substantial-justification analysis. This matters because the fee petition itself can be expensive to litigate, and circuits have generally followed Jean to permit recovery of those fees.
Section 2412(d)(2)(A)(ii) sets the EAJA hourly rate at $125 per hour, "unless the court determines that an increase in the cost of living or a special factor, such as the limited availability of qualified attorneys for the proceedings involved, justifies a higher fee." This cap, set by Congress in 1996, would yield an obviously inadequate rate in modern practice if read literally. The cost-of-living adjustment exists precisely to remedy that.
Federal circuits have universally allowed CPI-based adjustment of the $125 baseline. The mechanics vary by circuit:
As an order-of-magnitude approximation, the CPI-adjusted EAJA rate has been running in the range of $230 to $240 per hour in recent years and continues to rise gradually with inflation. The actual rate applied in any given fee petition should be calculated using current CPI data and the controlling circuit's preferred methodology. Most courts will accept a methodology of (i) using the U.S. Department of Labor's published CPI tables, (ii) applying the rate in effect during the period the work was performed, and (iii) requesting the specific calculated rate by month.
Subsection (d)(2)(A)(ii) also permits a rate above the statutory cap (and above the CPI-adjusted rate) for "a special factor, such as the limited availability of qualified attorneys for the proceedings involved." Pierce v. Underwood read this provision narrowly:
"The 'special factor' formulation suggests Congress thought that there would be an occasional case where the rate is to exceed the statutory cap. We do not think Congress meant to give automatic rate increases under EAJA based on cost of living alone. … Examples of 'distinctive knowledge or specialized skill' needful for the litigation in question might be an identifiable practice specialty such as patent law, or knowledge of foreign law or language." Pierce v. Underwood, 487 U.S. 552, 572 (1988)
In immigration mandamus practice, special-factor enhancement above the CPI-adjusted rate has been awarded inconsistently. Some courts have credited specialized immigration knowledge as a special factor; others have not, on the ground that a non-specialist attorney could perform the work. The trend has been against enhancement absent a particularized showing that the case required expertise beyond general civil-litigation competence. Standard immigration delay claims, however complex they feel to the practitioner, generally have not satisfied the special-factor test.
The narrowest reading of the special-factor doctrine asks whether the attorney was selected because of unique skills bearing on the issue actually contested in the litigation — not the attorney's overall expertise, but the specific competencies that the case demanded. Mandamus delay claims rarely contest fact-specific immigration questions; they usually involve general administrative-law and TRAC-factor analysis that any reasonably competent federal litigator could perform. That analytical posture has made it difficult to win special-factor enhancement.
The conservative approach in a mandamus fee petition is to (i) request the CPI-adjusted rate for the period each block of work was performed, (ii) include the underlying CPI table as an exhibit, and (iii) reserve any special-factor argument for cases involving truly distinctive expertise — for example, foreign-language documentary review, scientific or medical issues bound up in the underlying application, or specialized post-conviction expertise where criminal-immigration issues are present.
Subsection (d)(2)(B) restricts EAJA's benefits to parties whose net worth falls within statutory limits, reflecting the statute's animating concern with redressing the litigation asymmetry between ordinary citizens and the federal government.
An individual must have a net worth not exceeding $2,000,000 at the time the civil action was filed. The vast majority of immigration mandamus plaintiffs satisfy this requirement, often by a wide margin.
A business or other organizational applicant must have a net worth not exceeding $7,000,000 and not more than 500 employees at the time the action was filed. These limits matter most in EB-2/EB-3 PERM mandamus actions filed by employers, in I-140 cases brought in an employer's name, and occasionally in EB-5 regional center disputes.
Where a mandamus complaint names multiple plaintiffs — for example, a principal applicant and a derivative beneficiary — the eligibility analysis is performed individually for each plaintiff. A high-net-worth co-plaintiff does not defeat fee recovery for the qualifying plaintiff, although it complicates the apportionment of fees.
The applicant bears the burden of proving eligibility. The standard practice is to submit a sworn declaration of net worth as of the date the civil action was filed, identifying assets and liabilities with sufficient particularity to allow the court to verify that the cap is satisfied. Detailed financial statements are not generally required where the applicant's net worth is plainly well below the cap; conservative practice is to include enough detail to make eligibility obvious without exposing more financial information than necessary.
A common practical question, particularly for pro se litigants and small-firm practitioners, is whether some recovery may be available under § 2412(a) even where § 2412(d) attorney's fees are unavailable. The answer is generally yes, with significant caveats.
§ 2412(a) authorizes recovery of "costs" against the United States on the same terms as would be available between private parties. The categories of taxable costs are set by 28 U.S.C. § 1920:
For a typical mandamus action, recoverable § 2412(a) costs are modest — usually the $405 filing fee, certified mail charges for Rule 4(i) service, and possibly copying costs. Attorney's fees, paralegal fees, and most attorney expenses fall outside § 1920 and are not recoverable as "costs."
In a case where the plaintiff cannot establish prevailing-party status under § 2412(d) — most importantly, where the case ends in voluntary dismissal after USCIS adjudication — § 2412(a) costs may nonetheless be available if the plaintiff obtained at least some judicial action in their favor. The threshold for § 2412(a) "prevailing party" status is generally treated as lower than under § 2412(d), though the doctrine here is less developed and the practical recovery is small.
For pro se litigants — including pro se litigants who themselves are licensed attorneys — EAJA attorney's-fee recovery is largely unavailable. The controlling authority is Kay v. Ehrler.
The plaintiff in Kay was an attorney who successfully represented himself in a civil-rights action and sought attorney's fees under 42 U.S.C. § 1988. The Supreme Court held unanimously that a pro se attorney litigant cannot recover attorney's fees under § 1988. The Court reasoned that the fee-shifting statute's purpose — encouraging effective and independent representation — is not served by paying a party for representing himself, even when that party is a competent lawyer. The same reasoning has been uniformly applied to EAJA: pro se litigants, whether or not they are attorneys, cannot recover § 2412(d) attorney's fees for their own time.
Pro se EAJA litigants can recover:
For an individual with a meritorious mandamus claim, the EAJA pro se limitation is one of the principal economic arguments for hiring counsel. A pro se filer who proceeds to a prevailing-party outcome may recover the $405 filing fee under § 2412(a) but nothing more. A represented plaintiff with the same case can potentially recover thousands of dollars in attorney's fees under § 2412(d), substantially defraying — and in some cases offsetting in full — the cost of the representation.
This calculation cuts both ways. A plaintiff filing pro se because the case is too small to justify retained counsel may be making a rational choice. A plaintiff filing pro se because retained counsel is unaffordable may be giving up not only the procedural advantages of representation but also the very fee-recovery mechanism Congress designed to make representation affordable.
A fee petition is a discrete pleading filed after the underlying case has reached a "final judgment" — meaning a judgment that is no longer appealable, either because the appeal period has expired or because the parties have stipulated to a non-appealable disposition.
28 U.S.C. § 2412(d)(1)(B) sets a thirty-day clock from "final judgment in the action." "Final judgment" means a judgment that is final and not appealable — which, for most district-court judgments, means the date the appeal period expires (60 days after entry of judgment where the United States is a party, under Federal Rule of Appellate Procedure 4(a)(1)(B)). The thirty days runs from that later date, not from the date of the underlying judgment itself.
The deadline is unforgiving. Scarborough v. Principi, 541 U.S. 401 (2004), holds that the substantive allegation of non-substantial-justification can be amended after the thirty-day deadline, provided a fee application was timely filed. But the deadline to file some application is firm and is not extendable absent compelling equitable grounds.
The fee petition must contain, at minimum:
The fee award is calculated using the lodestar method: reasonable hours multiplied by a reasonable rate. The "reasonable rate" is the CPI-adjusted EAJA rate (unless special-factor enhancement applies). The "reasonable hours" is the time actually expended, subject to reduction for time that was duplicative, excessive, or spent on unrelated matters. Lodestar enhancements above the calculated product are theoretically possible but extremely rare in EAJA practice.
The government has fourteen to thirty days, depending on local rules, to respond. The response typically contests one or more elements: lack of prevailing-party status, substantial justification, excessive hours, excessive rate, or eligibility caps. The applicant then files a reply. Briefing on EAJA petitions can be substantial — often longer and more substantive than the briefing on the underlying mandamus action.
Every district has local rules governing fee petitions, and most assigned judges have standing orders. Common variations include required meet-and-confer before filing, specific formats for time records, page limits, and required summary tables. Read the local rules and standing orders before drafting any EAJA petition.
EAJA doctrine is general; mandamus practice is specific. A few strategic considerations are worth flagging, because they are the points where EAJA recovery is most often won or lost.
As discussed in Section 3, the most common reason for EAJA failure in mandamus is the absence of judicial imprimatur on the relief obtained. Three drafting moves preserve prevailing-party status:
A well-drafted pre-suit demand letter, sent to USCIS and the U.S. Attorney's office, can serve two EAJA purposes. First, it documents the government's position before litigation, which is part of the substantial-justification analysis. Second, in some circuits, an unanswered demand letter strengthens the unreasonableness showing for both the merits and EAJA. The downside is that a successful demand letter that results in pre-suit adjudication renders the matter unsuitable for EAJA recovery entirely — no civil action was filed, no costs were incurred, no fees are recoverable. Use demand letters strategically based on the likelihood that they will produce adjudication: in cases where they will, they save the client money even though they preclude EAJA; in cases where they will not, they strengthen the eventual EAJA petition.
Government counsel will sometimes offer to adjudicate the underlying application in exchange for a covenant not to seek EAJA fees. Accepting such a covenant is generally a mistake unless the EAJA claim is weak on independent grounds, because the client is then trading away a statutory entitlement for relief the government would have provided anyway. The right response is to insist on a consent order with EAJA expressly reserved.
Attorneys litigating mandamus actions should maintain contemporaneous time records from intake forward, even before filing the complaint. EAJA hours are scrutinized: blocks of time without specific descriptions, "billing entries reconstructed from memory," and large round numbers all draw fire from the government and are routinely reduced by courts. Itemize by task. Avoid block-billing. Bill in tenths of an hour.
Under Commissioner v. Jean, fees incurred in litigating the EAJA petition itself are recoverable without a separate substantial-justification analysis. Bill the time spent on the fee petition just like any other time. This recovery is meaningful because EAJA petitions, particularly contested ones, can consume more attorney time than the underlying merits.
Most denied EAJA petitions fail for one of a small set of recurring reasons. These are the recurring traps.
The single most common mistake. Once USCIS adjudicates, the temptation is to file a quick stipulation of dismissal and move on. That filing renders the plaintiff a non-prevailing party under Buckhannon. Insist on a consent order or court-approved settlement instead.
The thirty days runs from final judgment, not from entry of judgment. Calendar the deadline immediately upon any disposition that might become final. When in doubt, file the petition early — it can be supplemented later if the judgment turns out to be appealable.
Eligibility is the applicant's burden. A petition that does not affirmatively show net-worth eligibility — typically via a short declaration — invites the government to argue, and the court to find, that eligibility has not been demonstrated.
Some practitioners request their actual hourly rate and let the government argue down. This invites a rate reduction and, sometimes, a finding that the petition was over-aggressive. The conservative approach is to request the CPI-adjusted rate up front and reserve special-factor argument for cases where it genuinely applies.
Entries like "Drafted complaint, conferred with client, researched venue, reviewed exhibits — 6.5 hours" are reduced or eliminated by most courts. Itemize each task, in tenths of an hour, with enough description for the court to evaluate reasonableness. This standard applies even where contemporaneous practice in the firm is to block-bill private clients.
Fees can be claimed only for the civil action. Time spent on related but distinct matters — pre-suit administrative inquiries, communications with USCIS that did not become part of the litigation record, work on the underlying immigration application — is generally not recoverable. Segregate carefully.
The petition must affirmatively allege that the government's position was not substantially justified. A petition that addresses only fees and hours without making this allegation can be dismissed as failing to state a claim for EAJA relief. The allegation may be conclusory at the petition stage and developed later, but it must be made.
Government counsel will sometimes offer a modest negotiated EAJA payment in exchange for waiver of further claims. The negotiated number can be substantially below what the lodestar calculation would yield. Evaluate whether to accept based on the strength of the prevailing-party showing and the contestability of substantial justification; do not accept reflexively.
Stepping back from doctrine, the practical question for both clients and counsel is when EAJA recovery is realistic. The answer depends on three variables: how the case ends, how aggressive the government's pre-suit and litigation positions were, and how the settlement was drafted.
For straightforward represented mandamus cases that go to a court-ordered disposition, EAJA recovery in the range of $4,000 to $12,000 is common, depending on the complexity of the litigation, the CPI-adjusted rate at the relevant time, and the hours actually expended. Cases that are heavily litigated through motion practice, briefing, and contested EAJA petitions can yield substantially more. Cases that end with voluntary adjudication and stipulated dismissal generally yield only the § 2412(a) costs — typically a few hundred dollars.
For clients, the practical implication is that EAJA recovery should be factored into the economic decision to pursue a mandamus action and into the choice of attorney representation. EAJA exists to make representation affordable; using it that way requires both choosing counsel who will pursue the fee claim and structuring any settlement to preserve it.
This article is published as a public legal resource for individuals, fellow practitioners, and anyone interested in the doctrinal mechanics of EAJA fee recovery in federal immigration mandamus litigation. It is not legal advice for any specific case, and reading it does not create an attorney-client relationship. If you are facing a delayed immigration matter and would like to discuss whether mandamus — and the EAJA framework that goes with it — fits your situation, you are welcome to reach out for a no-cost case evaluation.