United States Supreme Court.
Stevens, J., delivered the opinion for a unanimous Court.
*107 Allen E. Shoenberger, by appointment of the Court, 507 U.S. 906, argued the cause and filed briefs for petitioner.
William K. Kelley argued the cause for the United States. With him on the brief were Acting Solicitor General Bryson, Acting Assistant Attorney General Schiffer, Deputy Solicitor General Mahoney, Mark B. Stern, and Henry D. Gabriel.[*]
Justice Stevens, delivered the opinion of the Court.
The Federal Tort Claims Act (FTCA) provides that an "action shall not be instituted upon a claim against the United States for money damages" unless the claimant has first exhausted his administrative remedies.[1] The question presented is whether such an action may be maintained when the claimant failed to exhaust his administrative remedies prior to filing suit, but did so before substantial progress was made in the litigation.
On March 6, 1989, petitioner, proceeding without counsel, lodged a complaint in the United States District Court for the Northern District of Illinois, alleging that the United States Public Health Service had caused him serious injuries while "conducting human research and experimentation on prisoners" in the custody of the Illinois Department of Corrections. *108 He invoked the federal court's jurisdiction under the FTCA and prayed for a judgment of $20 million. App. 3-7.
Four months later, on July 7, 1989, petitioner submitted a claim for damages to the Department of Health and Human Services.[2] The Department denied the claim on July 21, 1989. On August 7, 1989, petitioner sent a letter to the District Court enclosing a copy of the Department's denial of his administrative claim and an affidavit in support of an earlier motion for appointment of counsel. Petitioner asked that the court accept the letter "as a proper request, whereas plaintiff can properly commence his legal action accordingly." Id., at 10.
For reasons that are not entirely clear, the United States was not served with a copy of petitioner's complaint until July 30, 1990.[3]Id., at 2. On September 19, 1990, the United States moved to dismiss the complaint on the ground that petitioner's action was barred by the 6-month statute of limitation.[4] The motion was based on the assumption that *109 the complaint had been filed on April 15, 1990, when petitioner paid the court filing fees, and that that date was more than six months after the denial of petitioner's administrative claim. In response to the motion, petitioner submitted that the complaint was timely because his action had been commenced on March 6, 1989, the date when he actually lodged his complaint and the Clerk assigned it a docket number.
The District Court accepted March 6, 1989, as the operative date of filing, but nonetheless granted the Government's motion to dismiss. Petitioner's suit was not out of time, the District Court reasoned, but, rather, premature. The court concluded that it lacked jurisdiction to entertain an action "commenced before satisfaction of the administrative exhaustion requirement under § 2675(a)." Id., at 21.
The Court of Appeals for the Seventh Circuit affirmed. The court explained:
"According to 28 U.S. C. § 2401(b), a tort claim against the United States must be `begun within six months after the date of mailing . . . of notice of final denial of the claim by the agency to which it was presented.' The administrative denial was mailed on July 21, 1989, so McNeil had between then and January 21, 1990, to begin his action. The complaint filed in March 1989 was too early. This left two options. Perhaps the document filed in March 1989 loitered on the docket, springing into force when the agency acted. Or perhaps the request for counsel in August 1989, during the six-month period, marks the real `beginning' of the action. The district court rejected both options, and McNeil, with the assistance of counsel appointed by this court, renews the arguments here.
. . . . .
*110 "March 1989 was too early. The suit did not linger, awaiting administrative action. Unless McNeil began a fresh suit within six months after July 21, 1989, he loses." 964 F.2d 647, 648-649 (1992).
The court reviewed the materials filed in August 1989 and concluded that the District Court had not committed plain error in refusing to construe them as having commenced a new action.[5]
Because decisions in other Circuits permit a prematurely filed FTCA action to proceed if no substantial progress has taken place in the litigation before the administrative remedies are exhausted, see Kubrick v. United States, 581 F.2d 1092, 1098 (CA3 1978), rev'd on other grounds, 444 U.S. 111 (1979), and Celestine v. Veterans Administration Hospital, 746 F.2d 1360, 1363 (CA8 1984),[6] we granted certiorari to resolve the conflict. 506 U.S. 1074 (1993).
As the case comes to us, we assume that the Court of Appeals correctly held that nothing done by petitioner after the denial of his administrative claim on July 21, 1989, constituted the commencement of a new action. The narrow question before us is whether his action was timely either because *111 it was commenced when he lodged his complaint with the District Court on March 6,1989, or because it should be viewed as having been "instituted" on the date when his administrative claim was denied.
The text of the statute requires rejection of the first possibility. The command that an "action shall not be instituted. .. unless the claimant shall have first presented the claim to the appropriate Federal agency and his claim shall have been finally denied by the agency in writing and sent by certified or registered mail" is unambiguous. We are not free to rewrite the statutory text. As of March 6, 1989, petitioner had neither presented his claim to the Public Health Service, nor had his claim been "finally denied" by that agency. As the Court of Appeals held, petitioner's complaint was filed too early.
The statutory text does not speak with equal clarity to the argument that petitioner's subsequent receipt of a formal denial from the agency might be treated as the event that "instituted" his action. Petitioner argues the word "instituted" that is used in § 2675(a), see n. 1, supra, is not synonymous with the word "begun" in § 2401(b), see n. 4, supra, or with the word "commence" as used in certain other statutes and rules. See, e. g., Hallstrom v. Tillamook County, 493 U.S. 20 (1989). He suggests that an action is not "instituted" until the occurrence of the events that are necessary predicates to the invocation of the court's jurisdiction namely, the filing of his complaint and the formal denial of the administrative claim. This construction, he argues, is consistent with the underlying purpose of § 2675(a): As long as no substantial progress has been made in the litigation by the time the claimant has exhausted his administrative remedies, the federal agency will have had a fair opportunity to investigate and possibly settle the claim before the parties *112 must assume the burden of costly and time-consuming litigation.[7]
We find this argument unpersuasive. In its statutory context, we think the normal interpretation of the word "institute" is synonymous with the words "begin" and "commence." The most natural reading of the statute indicates that Congress intended to require complete exhaustion of Executive remedies before invocation of the judicial process. Every premature filing of an action under the FTCA imposes some burden on the judicial system[8] and on the Department of Justice which must assume the defense of such actions. Although the burden may be slight in an individual case, the statute governs the processing of a vast multitude of claims. The interest in orderly administration of this body of litigation is best served by adherence to the straightforward statutory command.
*113 Moreover, given the clarity of the statutory text, it is certainly not a "trap for the unwary." It is no doubt true that there are cases in which a litigant proceeding without counsel may make a fatal procedural error, but the risk that a lawyer will be unable to understand the exhaustion requirement is virtually nonexistent. Our rules of procedure are based on the assumption that litigation is normally conducted by lawyers. While we have insisted that the pleadings prepared by prisoners who do not have access to counsel be liberally construed, see Haines v. Kerner, 404 U.S. 519 (1972); Estelle v. Gamble, 429 U.S. 97, 106 (1976),[9] and have held that some procedural rules must give way because of the unique circumstance of incarceration, see Houston v. Lack, 487 U.S. 266 (1988) (pro se prisoner's notice of appeal deemed filed at time of delivery to prison authorities), we have never suggested that procedural rules in ordinary civil litigation should be interpreted so as to excuse mistakes by those who proceed without counsel.[10] As we have noted before, "in the long run, experience teaches that strict adherence to the procedural requirements specified by the legislature is the best guarantee of evenhanded administration of the law." Mohasco Corp. v. Silver, 447 U.S. 807, 826 (1980).
The FTCA bars claimants from bringing suit in federal court until they have exhausted their administrative remedies. Because petitioner failed to heed that clear statutory command, the District Court properly dismissed his suit.
The judgment of the Court of Appeals is Affirmed.
[*] Joseph A. Power, Jr., and Arthur H. Bryant filed a brief for Trial Lawyers for Public Justice, P. C., as amicus curiae urging reversal.
[1] Title 28 U.S. C. § 2675(a) provides, in pertinent part:
"An action shall not be instituted upon a claim against the United States for money damages for injury or loss of property or personal injury or death caused by the negligent or wrongful act or omission of any employee of the Government while acting within the scope of his office or employment, unless the claimant shall have first presented the claim to the appropriate Federal agency and his claim shall have been finally denied by the agency in writing and sent by certified or registered mail. The failure of an agency to make final disposition of a claim within six months after it is filed shall, at the option of the claimant any time thereafter, be deemed a final denial of the claim for purposes of this section."
[2] Petitioner sought damages of $500,000 in his administrative claim, not the $20 million for which he prayed in his earlier federal court action. Pursuant to 28 U.S. C. § 2675(b), a claimant is barred from seeking in federal court "any sum in excess of the amount of the claim presented to the federal agency." That is, had petitioner properly filed an action in district court after his administrative claim was denied, he would have been limited in his recovery to $500,000.
[3] Entries in the District Court docket indicate that plaintiff had previously filed a motion for leave to proceed in forma pauperis, that later in August he filed a motion for appointment of counsel, and that he ultimately paid a filing fee that caused the District Court to dismiss the motion for leave to file in forma pauperis as moot. In all events, in April 1990, the District Court ordered service to be effected by a United States Marshal "because plaintiff is incarcerated and proceeding pro se." App. 1.
[4] Title 28 U.S. C. § 2401(b) provides: "A tort claim against the United States shall be forever barred unless it is presented in writing to the appropriate Federal agency within two years after such claim accrues or unless action is begun within six months after the date of mailing, by certified or registered mail, of notice of final denial of the claim by the agency to which it was presented."
[5] In dissent, Judge Ripple expressed the opinion that petitioner had properly raised the issue in the District Court and on appeal, 964 F.2d, at 649, n. 1, and that in any event it was "clear that the plaintiff, a prisoner proceeding pro se, attempted to refile the action after the denial of the administrative claim." Id., at 649. Our grant of certiorari did not encompass the question whether a new action had been filed in August and we therefore express no opinion as to the correctness of the Court of Appeals' ruling on that issue.
[6] Decisions in the Fifth and Ninth Circuits agree with the position taken in the Seventh Circuit in this case. See Gregory v. Mitchell, 634 F.2d 199, 204 (CA5 1981); Reynolds v. United States, 748 F.2d 291, 292 (CA5 1984); Jerves v. United States, 966 F.2d 517, 521 (CA9 1992).
[7] Prior to 1966, FTCA claimants had the option of filing suit in federal court without first presenting their claims to the appropriate federal agency. Moreover, federal agencies had only limited authority to settle claims. See Federal Tort Claims Act of 1946, ch. 753, §§ 403(a), 420, 60 Stat. 843, 845. Because the vast majority of claims ultimately were settled before trial, the Department of Justice proposed that Congress amend the FTCA to "requir[e] all claims to be presented to the appropriate agency for consideration and possible settlement before a court action could be instituted. This procedure would make it possible for the claim first to be considered by the agency whose employee's activity allegedly caused the damage. That agency would have the best information concerning the activity which gave rise to the claim. Since it is the one directly concerned, it can be expected that claims which are found to be meritorious can be settled more quickly without the need for filing suit and possible expensive and time-consuming litigation." S. Rep. No. 1327, 89th Cong., 2d Sess., 3 (1966).
The Senate Judiciary Committee further noted that "the improvements contemplated by [the 1966 amendments] would not only benefit private litigants, but would also be beneficial to the courts, the agencies, and the Department of Justice itself." Id., at 2.
[8] Even petitioner concedes that at least one objective of the 1966 amendments to the FTCA was to "reduce unnecessary congestion in the courts." Id., at 4. See Brief for Petitioner 24.
[9] Again, the question whether the Court of Appeals should have liberally construed petitioner's letter of August 7, 1989, as instituting a new action is not before us. See n. 5, supra.
[10] Indeed, we have previously recognized a systemic interest in having a party represented by independent counsel even when the party is a lawyer. See Kay v. Ehrler, 499 U.S. 432 (1991).
