Attorneys Fees and Costs: The American Rule and Its Many Exceptions

Summary. In the United States each side normally pays its own lawyer. That default is why most small claims never get litigated and why fee-shifting statutes are among the most consequential provisions in American law. This article covers the origin of the American Rule, the five categories of exception, prevailing-party doctrine, the lodestar and its adjustments, the asymmetry between prevailing plaintiffs and defendants, Rule 68's trap, taxable costs, and sanctions as a separate fee-shifting track.

Here is a fact that explains more about American litigation than any other single rule: if you sue someone and win, they usually do not have to pay your lawyer.

That default — the American Rule — is invisible to most people until it lands on them. A tenant is owed $3,400 in a wrongfully withheld deposit and discovers that recovering it will cost $6,000 in legal fees. A small business is stiffed on a $22,000 invoice and learns that collecting will consume the entire amount. The rule does not just allocate a cost. It determines which wrongs are worth remedying at all.

And that is exactly why Congress and state legislatures have carved so many holes in it. Fee-shifting provisions are how a legislature converts an unenforceable right into an enforceable one. Understanding when fees shift, to whom, and how much, is therefore not a housekeeping topic. It is often the difference between a case that exists and a case that does not.

Part I: The American Rule and where it comes from

The rule is old and it is judicial, not statutory. The Supreme Court's definitive modern statement came in Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240 (1975). Environmental groups had successfully challenged permits for the trans-Alaska pipeline, and the D.C. Circuit awarded them fees on a "private attorney general" theory — they had vindicated an important public policy at their own expense.

The Supreme Court reversed. Federal courts, it held, have no general equitable power to shift fees based on their own assessment of which statutes are important enough to warrant private enforcement. That judgment belongs to Congress. The Court catalogued the narrow exceptions that do exist — common fund, contempt, bad faith — and stopped there.

Alyeska is why the fee-shifting landscape looks the way it does. Congress responded within a year by enacting the Civil Rights Attorney's Fees Awards Act, now 42 U.S.C. § 1988(b), which permits a court to award a reasonable attorney's fee to the prevailing party in actions under § 1983 and a list of other civil rights statutes. Hundreds of federal statutes and thousands of state provisions now contain similar language. The default remains no shifting; the exceptions are numerous, specific, and each has its own rules.

The policy debate in one paragraph. Supporters of the American Rule say it keeps courthouse doors open: a plaintiff with a plausible but losing claim should not be bankrupted for bringing it, and uncertainty about fee exposure would chill legitimate suits. Critics say it under-enforces rights, because a rational person will not spend $6,000 to recover $3,400, which means small violations are effectively lawful. Both are correct, which is why the compromise looks like it does — a general rule of no shifting, with one-way shifting in the areas where the legislature has decided under-enforcement is the greater danger.

Part II: The five ways fees shift

1. By statute. The largest category. Federal examples include § 1988 for civil rights, Title VII, the FLSA, the ADA, ERISA, the FDCPA, the Truth in Lending Act, the Copyright Act, the Lanham Act, the Clayton Act, the Equal Access to Justice Act (fees against the government), and the Freedom of Information Act. State examples run into the thousands — consumer protection acts, wage statutes, landlord-tenant acts, mechanic's lien statutes, and open records laws.

Statutes differ in three critical ways. Directionality: most civil rights statutes shift one way in practice (see Part VII); commercial statutes often shift both ways. Discretion: "the court may allow" versus "the court shall award" is a real difference. Trigger: prevailing party, substantially prevailing party, successful party, or — in some statutes — merely obtaining "some degree of success on the merits."

2. By contract. Parties may agree that the loser pays. The clause is enforceable in most jurisdictions, but note two twists: several states, California most prominently, convert a one-way contractual fee clause into a mutual one by statute, so a landlord's form lease that awards fees only to the landlord will award them to a prevailing tenant too. And the clause's scope matters enormously — "in any action to enforce this Agreement" may not cover a tort claim arising from the same relationship.

3. Common fund. A litigant who creates or preserves a fund for the benefit of others may recover fees from that fund. This is the equitable exception Alyeska preserved, and it is the basis for most class action fee awards. It is not fee shifting in the ordinary sense — the class pays, not the defendant.

4. Bad faith and sanctions. Courts have inherent authority to shift fees against a party who has acted in bad faith, vexatiously, wantonly, or for oppressive reasons. See Part XI.

5. The substantial benefit and private attorney general doctrines, surviving in some state courts. California's Code of Civil Procedure § 1021.5, for instance, authorizes fees where litigation enforced an important right affecting the public interest — the very theory Alyeska rejected as a matter of federal law.

Part III: Who is a "prevailing party"?

You cannot recover fees under a prevailing-party statute unless you are one, and the definition is narrower than intuition suggests.

The judicial imprimatur requirement. In Buckhannon Board & Care Home, Inc. v. West Virginia Department of Health and Human Resources, 532 U.S. 598 (2001), the Court rejected the "catalyst theory" — the idea that a plaintiff prevails when its lawsuit causes the defendant to voluntarily change its conduct. To be a prevailing party, the plaintiff must obtain a judicially sanctioned material alteration of the legal relationship between the parties: a judgment on the merits, or a court-ordered consent decree.

The practical consequence is large. A defendant who repeals the challenged ordinance the week before the hearing may moot the case and defeat the fee claim. Plaintiffs' counsel respond by seeking a consent decree rather than a private settlement, or by negotiating fees expressly into any settlement agreement — because after Buckhannon, a private settlement without court approval generally does not make you a prevailing party.

Degree of success matters, and can reduce the award to nothing. In Farrar v. Hobby, 506 U.S. 103 (1992), a plaintiff who sought $17 million and recovered one dollar in nominal damages was technically a prevailing party — the judgment altered the legal relationship — but the Court held the only reasonable fee was no fee at all. Justice O'Connor's concurrence supplies the working test: compare the amount sought to the amount recovered, ask whether the suit accomplished any public goal, and ask whether it resolved a significant legal question.

Some statutes use a looser standard. In Hardt v. Reliance Standard Life Insurance Co., 560 U.S. 242 (2010), the Court held that ERISA's fee provision does not require prevailing-party status at all; a claimant need only show "some degree of success on the merits" — more than a trivial success or a purely procedural victory. A remand for reconsideration, after a court had found the administrator's process deficient, qualified. Read the statute; do not assume it says "prevailing party."

Part IV: The lodestar

Once entitlement is established, the amount is calculated by the lodestar: reasonable hours multiplied by a reasonable hourly rate. The framework comes from Hensley v. Eckerhart, 461 U.S. 424 (1983), which supplies three propositions worth memorizing:

The fee applicant bears the burden of documenting hours and rates and must exercise "billing judgment" — excluding hours that are excessive, redundant, or otherwise unnecessary, just as a lawyer would in billing a paying client.
Where claims are unrelated, hours spent on unsuccessful claims must be excluded entirely.
Where claims share a common core of facts or related legal theories, the court should focus on the overall relief obtained rather than parsing claim by claim. "The result is what matters."

The reasonable rate. Blum v. Stenson, 465 U.S. 886 (1984) established that the rate is set by the prevailing market rate in the relevant community for lawyers of comparable skill, experience, and reputation — and, importantly, that non-profit legal services organizations get market rates, not their actual cost. The applicant proves the market rate with affidavits from other practitioners, fee awards in comparable cases, and published rate surveys.

Reasonable hours. Contemporaneous time records are effectively required in practice, even where no rule mandates them. The common reductions:

Block billing — a single 6.8-hour entry reading "research; draft brief; conference; review documents" invites a percentage cut, often 10–20%.
Vague entries — "attention to file," "work on case."
Overstaffing — three lawyers at a deposition, four at a hearing.
Clerical work billed at attorney rates — filing, calendaring, organizing exhibits.
Excessive time — thirty hours on a routine motion to compel.
Travel time, often compensated at half rate.

Part V: Enhancements, reductions, and what the Court has closed off

The lodestar is strongly presumed to be a reasonable fee, and the Supreme Court has spent forty years narrowing what can move it.

No contingency enhancement. City of Burlington v. Dague, 505 U.S. 557 (1992) held that a fee award under a federal fee-shifting statute may not be enhanced to reflect the contingent risk of nonpayment. The Court reasoned that risk is largely a function of the merits, and paying more for weaker cases inverts the incentive.

Enhancement for superior performance is available but rare. Perdue v. Kenny A. ex rel. Winn, 559 U.S. 542 (2010) held that the lodestar may be enhanced in "rare and exceptional" circumstances where it does not adequately measure an attorney's true market value — for example, where the hourly rate does not capture extraordinary performance, where the litigation required an exceptional and unanticipated outlay of expenses over a long period, or where there was extraordinary delay in payment. But the applicant must produce specific evidence supporting the enhancement, and the calculation must use an objective method rather than a court's impression.

Reduction for limited success is common. This is the Hensley inquiry, and it does the most work in practice. A plaintiff who wins on one of five theories and recovers a fraction of what was sought should expect a substantial cut — not necessarily proportional, but substantial.

Fees for work on the fee petition ("fees on fees") are generally recoverable, though some courts cap them as a percentage of the underlying award.

Part VI: Apportioning between successful and unsuccessful claims

The recurring practical problem: a plaintiff brings five claims, wins two, and the defendant argues that three-fifths of the fees should be disallowed.

Fox v. Vice, 563 U.S. 826 (2011) supplies the governing test in the mirror-image situation — a defendant seeking fees for defending frivolous claims joined with non-frivolous ones. The Court adopted a but-for standard: the defendant may recover only the fees it would not have incurred but for the frivolous claims. Work that would have been done anyway to defend the legitimate claims is not recoverable.

The same logic runs in the plaintiff's direction under Hensley. If the unsuccessful claims shared a common core of facts with the successful ones — same witnesses, same documents, same discovery — the hours are generally compensable, and the court should adjust based on the overall result rather than by counting claims. If the unsuccessful claims were genuinely distinct, their hours come out.

The Court also cautioned in Fox that fee litigation should not become a second major lawsuit: "trial courts need not, and indeed should not, become green-eyeshade accountants." Courts may take a rough percentage approach. That cuts both ways — it protects applicants from line-by-line evisceration and protects opponents from having to prove each entry unreasonable.

Part VII: Why prevailing defendants almost never recover

Read § 1988 and Title VII literally and they appear symmetrical: fees to the prevailing party. In practice they are not, and the reason is Christiansburg Garment Co. v. EEOC, 434 U.S. 412 (1978).

The Court held that a prevailing plaintiff should ordinarily recover fees unless special circumstances would render an award unjust, but a prevailing defendant may recover only where the plaintiff's action was frivolous, unreasonable, or without foundation, or where the plaintiff continued to litigate after it clearly became so. Crucially, the Court added that a district court must "resist the understandable temptation to engage in post hoc reasoning by concluding that, because a plaintiff did not ultimately prevail, his action must have been unreasonable or without foundation." Losing is not the same as frivolous.

Where the asymmetry does not apply. Commercial fee-shifting statutes and contractual clauses are typically symmetrical, and in intellectual property the standards are their own subject — the Patent Act's "exceptional case" standard and the Copyright Act's multi-factor discretionary approach both permit defendant recoveries far more readily than Christiansburg does.

Part VIII: Rule 68 — the trap that takes fees away

Federal Rule of Civil Procedure 68 lets a defendant serve an offer of judgment. If the plaintiff rejects it and the judgment finally obtained is not more favorable than the offer, the plaintiff must pay the costs incurred after the offer was made.

That sounds modest until you read Marek v. Chesny, 473 U.S. 1 (1985). The Court held that where the underlying statute defines attorney's fees as part of "costs" — as § 1988 does — Rule 68's cost-shifting includes attorney's fees. A civil rights plaintiff who rejects a $100,000 offer, wins $90,000 at trial, and has incurred $200,000 in fees after the offer recovers the $90,000 and loses the post-offer fees entirely.

Four features of the rule matter:

It is one-directional. Only defendants may make Rule 68 offers, and the rule does not apply where the defendant wins outright.
The comparison is to the judgment, not the verdict — including pre-offer costs and fees where they are recoverable, which makes the arithmetic subtle.
The offer should be unambiguous about whether it includes fees and costs. Ambiguity is construed against the offeror, and a lump-sum offer "inclusive of costs and attorney's fees" is treated differently from one silent on the subject.
Many states have analogues — some two-directional, some applying to any party, some shifting expert fees. Check the local rule; the differences are large.

For plaintiffs, the response is discipline: evaluate every Rule 68 offer against a realistic verdict range plus the fee exposure, and document the evaluation.

Part IX: Contractual fee shifting

A clause that says the prevailing party recovers fees changes the economics of every dispute under the contract. Points that decide cases:

Scope. "Any action to enforce this Agreement" may exclude tort, statutory, and declaratory claims. Broader language — "arising out of or relating to this Agreement" — captures more.
Reciprocity by statute. Several states convert unilateral clauses into mutual ones. California Civil Code § 1717 is the best-known example and applies to any action on the contract.
Who is the prevailing party when both sides win something? Many clauses are silent, leaving the court to decide. Better clauses define it — for example, by net monetary recovery.
Interaction with settlement. A defendant facing a fee clause has an incentive to settle early, before the fee exposure exceeds the claim. A plaintiff has the mirror incentive to resolve quickly for the same reason.
Enforceability limits. Consumer contracts in many states restrict or void one-way clauses, and some statutes prohibit fee-shifting in specified transactions altogether.

Part X: Costs are not fees

Costs are a separate, much smaller category, and the distinction trips people up constantly.

Federal Rule of Civil Procedure 54(d)(1) creates a presumption that costs other than attorney's fees are allowed to the prevailing party. What may be taxed is limited by 28 U.S.C. § 1920:

Clerk and marshal fees
Transcript fees necessarily obtained for use in the case
Printing and witness fees
Exemplification and the costs of making copies necessarily obtained for use in the case
Docket fees under § 1923
Compensation of court-appointed experts and interpreters

What is not on that list matters more than what is. Ordinary expert witness fees are not taxable beyond the statutory witness attendance fee, absent a fee-shifting statute that provides otherwise. Neither are most travel expenses, legal research charges, mediation fees, or the cost of demonstrative exhibits that were never admitted. Electronic discovery costs are taxable only to the narrow extent they constitute "making copies," and courts have read that narrowly — processing and hosting charges usually are not recoverable.

Deadlines are short. In federal court the bill of costs is typically due 14 days after entry of judgment. State deadlines vary and are just as unforgiving.

Part XI: Sanctions as a separate fee-shifting track

Three overlapping authorities allow a court to shift fees as a sanction, and each has different requirements.

Rule 11 applies to pleadings, written motions, and other papers, requiring that legal contentions be warranted by existing law or a nonfrivolous argument for changing it, and that factual contentions have evidentiary support. Sanctions are limited to what suffices to deter, may be nonmonetary, and are subject to the 21-day safe harbor: serve the motion, wait, and file only if the offending paper is not withdrawn. Missing the safe harbor is the single most common reason Rule 11 motions fail.

28 U.S.C. § 1927 reaches an attorney (not a party) who "multiplies the proceedings in any case unreasonably and vexatiously," and makes that attorney personally liable for the excess costs, expenses, and attorney's fees reasonably incurred because of the conduct.

Inherent authority is the broadest and the most carefully policed. Chambers v. NASCO, Inc., 501 U.S. 32 (1991) confirmed that federal courts possess inherent power to assess fees against a party who has acted in bad faith, vexatiously, wantonly, or for oppressive reasons, and that the power survives alongside the rules and statutes. But Goodyear Tire & Rubber Co. v. Haeger, 581 U.S. 101 (2017) imposed a strict limit: an inherent-authority fee sanction must be compensatory, not punitive, and is confined to the fees the innocent party incurred solely because of the misconduct. The court must establish a causal link between the sanctionable conduct and each dollar awarded.

Part XII: Common fund and class action fees

In a class action, fees usually come from the recovery rather than from the defendant's pocket. Two methods dominate: a percentage of the fund (commonly 25% as a benchmark in many circuits, adjusted up or down) and the lodestar with a multiplier. Many courts use one method and cross-check with the other.

The structural problem is that no one at the settlement table has an interest in objecting: class counsel wants the fee, the defendant cares only about the total, and absent class members are unrepresented. Courts therefore scrutinize these awards more closely than ordinary fee petitions, and Rule 23(h) requires notice of the fee motion to the class and an opportunity to object. Watch for the recurring red flags — a "clear sailing" clause in which the defendant agrees not to contest fees, a reverter of unclaimed funds to the defendant, and coupon relief valued at face rather than redemption value.

Part XIII: Contexts with their own rules

Equal Access to Justice Act (28 U.S.C. § 2412). Fees against the United States where the government's position was not "substantially justified," subject to a statutory hourly cap adjusted for cost of living, and to a net worth eligibility limit.
ERISA (29 U.S.C. § 1132(g)). Discretionary, and after Hardt, available on "some degree of success on the merits" without prevailing-party status.
Prison Litigation Reform Act. Caps the hourly rate at a multiple of the Criminal Justice Act rate and requires that a portion of the judgment — up to 25% — be applied to satisfy the fee award. Murphy v. Smith, 583 U.S. 220 (2018) held the statute requires the district court to apply as much of that 25% as necessary before any fee falls on the defendant, leaving discretion only within the 0–25% range.
Social Security. 42 U.S.C. § 406(b) permits a contingent fee of up to 25% of past-due benefits, subject to court review for reasonableness, and coordinates with EAJA fees.
Bankruptcy. Professional fees require court approval under § 330 and are reviewed for reasonableness and benefit to the estate.
State consumer statutes. Many shift one way to consumers and are the practical engine of private enforcement in landlord-tenant, auto sales, and debt collection cases. See Debt Collection and the FDCPA.

Part XIV: A worked fee petition, with the arithmetic

Doctrine becomes concrete when you watch a lodestar get built and then taken apart.

The case. A tenant sues a landlord under a state consumer protection act with a one-way fee provision, alleging an illegal late-fee scheme. The tenant also pleads negligence and intentional infliction of emotional distress. After eighteen months, the tenant wins $4,200 in statutory damages on the consumer claim; the tort claims are dismissed on summary judgment.

Step 1 — Entitlement. The consumer act shifts fees to a prevailing consumer. The tenant obtained a judgment on the merits, so Buckhannon is satisfied. This is not a Farrar nominal-damages case — $4,200 on a claim that sought roughly $5,000 is real success.

Step 2 — The raw lodestar.

Timekeeper	Rate	Hours	Amount
Partner (18 yrs)	$525	61.4	$32,235
Associate (3 yrs)	$295	118.9	$35,076
Paralegal	$145	44.2	$6,409
Raw total		224.5	$73,720

Step 3 — Billing judgment, applied by the applicant. Hensley requires the applicant to cut before the court does, and a petition that shows voluntary reductions is far more credible than one that does not. Counsel removes:

9.1 paralegal hours of pure clerical work (filing, calendaring, Bates-stamping) — −$1,320
6.4 duplicative associate hours where two lawyers attended the same deposition — −$1,888
Half-rate treatment for 8.0 hours of travel — −$1,180
4.2 hours of internal conferences with vague entries — −$1,239

Adjusted lodestar: $68,093 on 199.9 hours.

Step 4 — Apportioning the unsuccessful claims. The tort claims failed. Under Hensley, the question is whether they shared a common core of facts with the successful claim. Here the negligence claim relied on the same lease, the same ledger, and the same three depositions — related. The emotional distress claim required a separate medical records subpoena, a treating provider's deposition, and its own summary judgment briefing — distinct. Counsel identifies and removes the 27.3 hours attributable solely to that claim: −$8,940.

Running total: $59,153.

Step 5 — The rate fight. The landlord submits a declaration asserting that the prevailing local rate for consumer litigation is $350 for partners. The tenant answers with three practitioner affidavits, two published fee awards in the same district within eighteen months, and a state bar rate survey. Under Blum, the question is the prevailing market rate for comparable skill and experience in the relevant community, and the applicant carries the burden. The court sets the partner rate at $475 — −$3,070.

Step 6 — Degree of success. The landlord argues for a 60% cut because two of three claims failed. The court declines to count claims, focuses on the overall relief obtained as Hensley directs, notes that the tenant achieved essentially the full statutory recovery available, and applies a modest 10% reduction for limited success across the related claims. −$5,608.

Step 7 — Fees on fees. Twelve hours preparing the petition and reply, at blended rates: +$4,140.

Award: approximately $54,615 on a $4,200 judgment.

That ratio is not a mistake and it is not an abuse. It is the statute working as designed: a $4,200 violation is not worth litigating on its own, and without the fee provision the practice would continue indefinitely. Courts have repeatedly held that proportionality between fees and damages is not required under fee-shifting statutes, precisely because the alternative is no enforcement at all.

What would have changed the outcome. If the landlord had served a Rule 68 offer (or its state analogue) for $6,000 at month three, and the tenant had rejected it, every hour after that offer would be at risk. The offer costs the defendant nothing if accepted at a number it would have paid anyway, and it converts an open-ended fee exposure into a capped one. It is the most underused defensive tool in fee-shifting litigation.

Part XV: How fee oppositions actually succeed

Most oppositions fail because they attack everything. The ones that work pick three targets and document them.

Target one: the rate. This is the cleanest fight, because it is comparative and evidentiary. Submit declarations from local practitioners, prior fee awards from the same court, and a rate survey. Do not merely assert that the rate is high — show what the court itself has approved.

Target two: identifiable blocks of unnecessary time. Not "the hours are excessive" but "counsel billed 34.6 hours to a motion to compel that was three pages long and was granted without opposition," with the entries attached. Specificity is what moves a court. Percentage arguments with no anchor do not.

Target three: apportionment. Identify the claims that failed and the work traceable solely to them. Under Fox v. Vice, the causal standard is but-for, and the same logic constrains Hensley reductions. Build a table: entry, date, hours, claim, why it is exclusively attributable.

Structural attacks worth making when the facts support them:

Block billing — quantify the affected entries as a percentage of the total, then ask for a proportionate cut.
Clerical work at professional rates — list the entries.
Overstaffing — show the number of timekeepers at each event.
Vague entries — quote them; "attention to file" quoted forty times is persuasive on its own.
Post-offer fees where a Rule 68 or state offer was rejected.
Absence of billing judgment — a petition claiming 100% of recorded time invites skepticism, and saying so is fair.

What almost never works: arguing that the fee exceeds the damages; arguing that the plaintiff's lawyer took the case on contingency and therefore has no fee to shift; and post hoc reasoning that a claim must have been frivolous because it lost — the argument Christiansburg expressly warns courts against.

And a note on tone. Fee litigation is where professional relationships go to die, and courts notice. Fox v. Vice cautions that trial courts "need not, and indeed should not, become green-eyeshade accountants." An opposition that asks for a defensible percentage cut on three documented grounds will do better than one that disputes 900 line items.

Part XV-A: State practice, and why the federal cases are only half the picture

Everything above is federal. Most litigation is not, and state fee practice diverges from the federal model in ways that change outcomes.

Loser-pays regimes. A handful of states have experimented with broader fee shifting. Alaska is the outlier that has done it longest: its civil rule awards partial fees to the prevailing party in most civil cases as a matter of course, on a schedule keyed to the judgment amount and whether the case was tried. Texas permits fee recovery by statute in a broad category of contract and services claims without requiring a contractual clause. Several states have adopted offer-of-judgment rules far more aggressive than Rule 68 — two-directional, applying to any party, and shifting expert fees as well as attorney's fees. In those jurisdictions the settlement calculus is different from the first day of the case.

Statutory reciprocity. California Civil Code § 1717 is the most-cited example of a statute that converts a one-way contractual fee clause into a mutual one in any action on the contract. Several other states have analogous provisions, some limited to consumer or residential leases. A landlord or lender using a national form with a unilateral clause may be handing the other side a fee remedy it never intended to create.

Prevailing-party definitions. Some states define the term by statute or rule — net monetary recovery, or the party in whose favor judgment is entered — rather than leaving it to case law. Where both sides win something, that definition decides the fee motion.

Consumer and wage statutes. These are the practical engine of private enforcement at the state level, and they are frequently one-way. A state wage act that shifts fees to a prevailing employee makes a $3,000 unpaid-overtime claim viable; without it, the claim is unenforceable in practice.

Procedure. State deadlines for fee motions and bills of costs differ from the federal 14-day norm, and several states require the fee request to be pleaded in the complaint or the answer to be preserved at all. Check that requirement before filing, not after judgment.

The practical instruction. Two research questions belong in every case memo: (1) does any statute or contract shift fees here, and in which direction? and (2) does this jurisdiction have an offer-of-judgment rule, and what does it shift? The answers determine what the case is worth, what an early settlement is worth, and — in a surprising number of matters — whether the case should be filed at all.

Part XVI: Frequently asked questions

If I win my case, will the other side pay my lawyer? Usually not, unless a statute or a contract says so, or the other side's conduct was sanctionable.

Does a contingency fee agreement affect a statutory fee award? The statutory award is calculated by the lodestar, not by the contingency percentage, and it belongs to the client unless the fee agreement provides otherwise. Most agreements address the interaction expressly; read the clause before the case ends.

Can I recover fees if I represent myself? Generally no. Pro se litigants — including pro se lawyers, in most circuits — cannot recover attorney's fees under fee-shifting statutes, because there was no attorney-client relationship and no fee incurred.

Are my expert's fees recoverable? Not as taxable costs beyond the nominal witness fee, unless a fee-shifting statute expressly authorizes expert fees. Several do; most do not.

What is the difference between "costs" and "expenses"? Taxable costs are the narrow § 1920 list. Litigation expenses — research, travel, couriers, mediation — are recoverable only where a statute or contract permits.

Can fees exceed the damages? Yes, routinely, in cases brought under fee-shifting statutes, and courts have repeatedly held that proportionality is not required. That is the design: a $900 statutory violation would otherwise go unremedied.

Part XVII: For non-lawyers — the practical questions to ask

Ask your lawyer, at the first meeting, whether any statute or contract shifts fees in your case. The answer changes what the case is worth and whether it should be brought at all.
Read the fee clause in any contract before you sign it. A one-way clause means you may pay their lawyer if you lose and your own if you win.
Understand your own fee agreement. Hourly, flat, contingent, or hybrid; what expenses you owe regardless of outcome; what happens to a statutory fee award if you win.
Take offers of judgment seriously. In fee-shifting cases, rejecting one and doing worse at trial can cost you more than the case is worth.
Keep costs in perspective. Even when you win costs, you are recovering filing fees and transcripts — not your lawyer's time.
In small disputes, ask about fee-shifting statutes specifically by name — your state's consumer protection act, wage act, or landlord-tenant act. These provisions exist precisely so that small claims can be brought, and they are frequently overlooked.