delivered the opinion of the Court.

The federal priority statute, 31 U. S. C. § 3713(a), provides that a claim of the United States Government “shall be paid first” when a decedent’s estate cannot pay all of its debts.1 The question presented is whether that statute requires that a federal tax claim be given preference over a judgment creditor’s perfected lien on real property even though such a preference is not authorized by the Federal Tax Lien Act of 1966, 26 U. S. C. § 6321 et seq.

I

On January 25,1985, the Court of Common Pleas of Cam-bria County, Pennsylvania, entered a judgment for $400,000 in favor of Romani Industries, Inc., and against Francis *520J. Romani. The judgment was recorded in the clerk’s office and therefore, as a matter of Pennsylvania law, it became a lien on all of the defendant’s real property in Cambria County. Thereafter, the Internal Revenue Service filed a series of notices of tax liens on Mr. Romani’s property. The claims for unpaid taxes, interest, and penalties described in those notices amounted to approximately $490,000.

When Mr. Romani died on January 13,1992, his entire estate consisted of real estate worth only $53,001. Because the property was encumbered by both the judgment lien and the federal tax liens, the estate’s administrator sought permission from the Court of Common Pleas to transfer the property to the judgment creditor, Romani Industries, in lieu of execution. The Federal Government acknowledged that its tax liens were not valid as against the earlier judgment lien; but, giving new meaning to Franklin’s aphorism that "in this world nothing can be said to be certain, except death and taxes,”2 it opposed the transfer on the ground that the priority statute (§3713) gave it the right to “be paid first.”

The Court of Common Pleas overruled the Government’s objection and authorized the conveyance. The Superior Court of Pennsylvania affirmed, and the Supreme Court of the State also affirmed. 547 Pa. 41, 688 A. 2d 703 (1997). That court first determined that there was a “plain inconsistency” between § 3713, which appears to give the United States “absolute priority” over all competing claims, and the Tax Lien Act of 1966, which provides that the federal tax lien “shall not be valid” against judgment lien creditors until a prescribed notice has been given. Id., at 45, 688 A. 2d, *521at 705.3 Then, relying on the reasoning in United States v. Kimbell Foods, Inc., 440 U. S. 715 (1979), which had noted that the Tax Lien Act of 1966 modified the Federal Government’s preferred position in the tax area and recognized the priority of many state claims over federal tax liens, id., at 738, the court concluded that the 1966 Act had the effect of limiting the operation of § 3713 as to tax debts.

The decision of the Pennsylvania Supreme Court conflicts with two Federal Court of Appeals decisions, Kentucky ex rel. Luckett v. United States, 383 F. 2d 13 (CA6 1967), and Nesbitt v. United States, 622 F. 2d 433 (CA9 1980). Moreover, in its petition for certiorari, the Government submitted that the decision is inconsistent with our holding in Thelusson v. Smith, 2 Wheat. 396 (1817), and with the admonition that “ ‘[o]nly the plainest inconsistency would warrant our finding an implied exception to the operation of so clear a *522command as that of [31 U. S. C. § 3713],'” United States v. Key, 397 U. S. 322, 324-325 (1970) (quoting United States v. Emory, 314 U. S. 423, 433 (1941)). We granted certiorari, 521 U. S. 1117 (1997), to resolve the conflict and to consider whether Thelusson, Key, or any of our other cases construing the pí’iority statute requires a different result.

II

There is no dispute about the meaning of two of the three statutes that control the disposition of this case. It is therefore appropriate to comment on the Pennsylvania lien statute and the Federal Tax Lien Act before considering the applicability of the priority statute to property encumbered by an antecedent judgment creditor’s hen.

The Pennsylvania statute expressly provides that a judgment shall create a lien against real property when it is recorded in the county where the property is located. 42 Pa. Cons. Stat. § 4303(a) (1995). After the judgment has been recorded, the judgment creditor has the same right to notice of a tax sale as a mortgagee.4 The recording in one county does not, of course, create a lien on property located elsewhere. In this case, however, it is undisputed that the judgment creditor acquired a valid hen on the real property in *523Cambria County before the judgment debtor’s death and before the Government served notice of its tax liens. Romani Industries’ lien was “perfected in the sense that there is nothing more to be done to have a ehoate lien — when the identity of the lienor, the property subject to the lien, and the amount of the lien are established.” United States v. City of New Britain, 347 U. S. 81, 84 (1954); see also Illinois ex rel. Gordon v. Campbell, 329 U. S. 362, 375 (1946).

The Federal Government’s right to a lien on a delinquent taxpayer’s property has been a part of our law at least since 1865.5 Originally the lien applied, without exception, to all property of the taxpayer immediately upon the neglect or failure to pay the tax upon demand.6 An unrecorded tax lien against a delinquent taxpayer’s property was valid even against a bona fide purchaser who had no notice of the lien. United States v. Snyder, 149 U. S. 210, 213-215 (1893). In 1913, Congress amended the statute to provide that the fed*524eral tax lien “shall not be valid as against any mortgagee, purchaser, or judgment creditor5’ until notice has been filed with the clerk of the federal district court or with the appropriate local authorities in the district or county in which the property subject to the lien is located. Act of Mar. 4,1913, 37 Stat. 1016. In 1939, Congress broadened the protection against unfiled tax liens to include pledgees and the holders of certain securities. Act of June 29, 1939, §401, 53 Stat. 882-883. The Federal Tax Lien Act of 1966 again broadened that protection to encompass a variety of additional secured transactions, and also included detailed provisions protecting certain secured interests even when a notice of the federal lien previously has been filed. 80 Stat. 1125-1132, as amended, 26 U. S. C. § 6323.

In sum, each time Congress revisited the federal tax lien, it ameliorated its original harsh impact on other secured creditors of the delinquent taxpayer.7 In this ease, it is agreed that by the terms of § 6323(a), the Federal Government’s liens are not valid as against the lien created by the earlier recording of Romani Industries’ judgment.

III

The text of the priority statute on which the Government places its entire reliance is virtually unchanged since its enactment in 1797.8 As we pointed out in United States v. *525Moore, 423 U. S. 77 (1975), not only were there earlier versions of the statute,9 but “its roots reach back even further into the English common law,” id., at 80. The sovereign prerogative that was exercised by the English Crown and by many of the States as “an inherent incident of sovereignty,” ibid., applied only to unsecured claims. As Justice Brandéis noted in Marshall v. New York, 254 U. S. 380, 384 (1920), the common-law priority “[did] not obtain over a specific lien created by the debtor before the sovereign undertakes to enforce its right.” Moreover, the statute itself does not create a lien in favor of the United States.10 Given this background, respondent argues that the statute should be read as *526giving the United States a preference over other unsecured creditors but not over secured creditors.11

There are dicta in our earlier cases that support this contention as well as dicta that tend to refute it. Perhaps the strongest support is found in Justice Story’s statement:

“What then is the nature of the priority, thus limited and established in favour of the United States? Is it a right, which supersedes and overrules the assignment of the debtor, as to any property which the United States may afterwards elect to take in execution, so as to prevent such property from passing by virtue of such assignment to the assignees? Or, is it a mere right of prior payment, out of the general funds of the debtor, in the hands of the assignees? We are of opinion that it clearly falls, within the latter description. The language employed is that which naturally would be employed to express such an intent; and it must be strained from its ordinary import, to speak any other.” Concord v. Atlantic Ins. Co. of N. Y., 1 Pet. 386, 439 (1828).

Justice Story’s opinion that the language employed in the statute “must be strained” to give it any other meaning is entitled to special respect because he was more familiar with 18th-century usage than judges who view the statute from a 20th-century perspective.

We cannot, however, ignore the Court’s earlier judgment in Thelusson v. Smith, 2 Wheat., at 426, or the more recent dicta in United States v. Key, 397 U. S., at 324-325. In Thelusson, the Court held that the priority statute gave the United States a preference over the claim of a judgment creditor who had a general lien on the debtor’s real property. *527The Court’s brief opinion12 is subject to the interpretation that the statutory priority always accords the Government a preference over judgment creditors. For two reasons, we do not accept that reading of the opinion.

First, as a factual matter, in 1817 when the ease was decided, there was no procedure for recording a judgment and thereby creating a ehoate lien on a specific parcel of real estate. See generally 2 L. Dembitz, A Treatise on Land Titles in the United States § 127, pp. 948-952 (1895). Notwithstanding the judgment, a bona fide purchaser could have acquired the debtor’s property free from any claims of the judgment creditor. See Semple v. Burd, 7 Serg. & Rawle 286, 291 (Pa. 1821) (“The prevailing object of the Legislature, has uniformly been, to support the security of a judgment creditor, by confirming his lien, except when it interferes with the circulation of property by embarrassing a fair purchaser”). That is not the ease with respect to *528Romani Industries’ choate lien on the property in Cambria Comity.

Second, and of greater importance, in his opinion for the Court in the Conard case, which was joined by Justice Washington, the author of Thelusson,13 Justice Story explained why that holding was fully consistent with his interpretation of the text of the priority statute:

“The real ground of the decision, was, that the judgment creditor had never perfected his title, by any execution and levy on the Sedgely estate; that he had acquired no title to the proceeds as his property, and that if the proceeds were to be deemed general funds of the debtor, the priority of the United States to payment had attached against all other creditors; and that a mere potential lien on land, did not carry a legal title to the proceeds of a sale, made under an adverse execution. This is the manner in which this case has been understood, by the Judges who concurred in the decision; and it is obvious, that it established no such proposition, as that a specific and perfected lien, can be displaced by the mere priority of the United States; since that priority is not of itself equivalent to a lien.” Conard, 1 Pet., at 444.14

The Government also relies upon dicta from our opinion in United States v. Key, 397 U. S., at 324-325, which quoted from our earlier opinion in United States v. Emory, 314 U. S., at 433: “Only the plainest inconsistency would warrant our *529finding an implied exception to the operation of so clear a command as that of [§ 3713].” Because both Key and Emory were eases in which the competing claims were unsecured, the statutory command was perfectly clear even under Justice Story’s construction of the statute. The statements made in that context, of course, shed no light on the clarity of the command when the United States relies on the statute as a basis for claiming a preference over a secured creditor. Indeed, the Key opinion itself made this specific point: “This ease does not raise the question, never decided by this Court, whether § 3466 grants the Government priority over the prior specific liens of secured creditors. See United States v. Gilbert Associates, Inc., 345 U. S. 361, 365-366 (1953).” 397 U. S., at 332, n. 11.

The Key opinion is only one of many in which the Court has noted that despite the age of the statute, and despite the fact that it has been the subject of a great deal of litigation, the question whether it has any application to antecedent perfected liens has never been answered definitively. See United States v. Vermont, 377 U. S. 351, 358, n. 8 (1964) (citing cases). In his dissent in United States v. Gilbert Associates, Inc., 345 U. S. 361 (1953), Justice Frankfurter referred to the Court’s reluctance to decide the issue “not only today but for almost a century and a half.” 345 U. S., at 367.

The Government’s priority as against specific, perfected security interests is, if possible, even less settled with regard to real property. The Court has sometimes concluded that a competing creditor who has not “divested” the debtor of “either title or possession” has only a “general, unperfeeted lien” that is defeated by the Government’s priority. E. g., id., at 366. Assuming the validity of this “title or possession” test for deciding whether a lien on personal property is sufficiently ehoate for purposes of the priority statute (a question of federal law, see Illinois ex rel. Gordon v. Campbell, 329 U. S., at 371), we are not aware of any decisions since Thelusson applying that theory to claims for real prop*530erty, or of any reason to require a lienor or mortgagee to acquire possession in order to perfect an interest in real estate.

Given the fact that this basic question of interpretation i’emains unresolved, it does not seem appropriate to view the issue in this case as whether the Tax Lien Act of 1966 has implicitly amended or repealed the priority statute. Instead, we think the proper inquiry is how best to harmonize the impact of the two statutes on the Government’s power to collect delinquent taxes.

IV

In his dissent from a particularly harsh application of the priority statute, Justice Jackson emphasized the importance of considering other relevant federal policies. Joined by three other Justices, he wrote:

“This decision announces an unnecessarily ruthless interpretation of a statute that at its best is an arbitrary one. The statute by which the Federal Government gives its own claims against an insolvent priority over claims in favor of a state government must be applied by courts, not because federal claims are more meritorious or equitable, but only because that Government has more power. But the priority statute is an assertion of federal supremacy as against any contrary state policy. It is not a limitation on the Federal Government itself, not an assertion that the priority policy shall prevail over all other federal policies. Its generalities should not lightly be construed to frustrate a specific policy embodied in a later federal statute.” Massachusetts v. United States, 333 U. S. 611, 635 (1948).

On several prior occasions the Court had followed this approach and concluded that a specific policy embodied in a later federal statute should control our construction of the priority statute, even though it had not been expressly *531amended. Thus, in Cook County Nat. Bank v. United States, 107 U. S. 445, 448-451 (1883), the Court concluded that the priority statute did not apply to federal claims against national banks because the National Bank Act comprehensively regulated banks’ obligations and the distribution of insolvent banks’ assets. And in United States v. Guaranty Trust Co. of N. Y., 280 U. S. 478, 485 (1930), we determined that the Transportation Act of 1920 had effectively superseded the priority statute with respect to federal claims against the railroads arising under that Act.

The bankruptcy law provides an additional context in which another federal statute was given effect despite the priority statute’s literal, unconditional text. The early federal bankruptcy statutes had accorded to “ ‘all debts due to the United States, and all taxes and assessments under the laws thereof’” a preference that was “coextensive” with that established by the priority statute. Guarantee Title & Trust Co. v. Title Guaranty & Surety Co., 224 U. S. 152, 158 (1912) (quoting the Bankruptcy Act of 1867, Rev. Stat. § 5101). As such, the priority Act and the bankruptcy laws “were to be regarded as in pari materia, and both were unqualified;... as neither contained any qualification, none could be interpolated.” 224 U. S., at 158. The Bankruptcy Act of 1898, however, subordinated the priority of the Federal Government’s claims (except for taxes due) to certain other kinds of debts. This Court resolved the tension between the new bankruptcy provisions and the priority statute by applying the former and thus treating the Government like any other general creditor. Id., at 158-160; Davis v. Pringle, 268 U. S. 315, 317-319 (1925).15

*532There are sound reasons for treating the Tax Lien Act of 1966 as the governing statute when the Government is claiming a preference in the insolvent estate of a delinquent taxpayer. As was the case with the National Bank Act, the Transportation Act of 1920, and the Bankruptcy Act of 1898, the Tax Lien Act is the later statute, the more specific statute, and its provisions are comprehensive, reflecting an obvious attempt to accommodate the strong policy objections to the enforcement of secret liens. It represents Congress5 detailed judgment as to when the Government’s claims for unpaid taxes should yield to many different sorts of interests (including, for instance, judgment liens, mechanic’s liens, and attorney’s liens) in many different types of property (including, for example, real property, securities, and motor vehicles). See 26 U. S. C. §6323. Indeed, given our unambiguous determination that the federal interest in the collection of taxes is paramount to its interest in enforcing other claims, see United States v. Kimbell Foods, Inc., 440 U. S., at 733-735, it would be anomalous to conclude that Congress intended the priority statute to impose greater burdens on the citizen than those specifically crafted for tax collection purposes.

Even before the 1966 amendments to the Tax Lien Act, this Court assumed that the more recent and specific provisions of that Act would apply were they to conflict with the older priority statute. In the Gilbert Associates ease, which concerned the relative priority of the Federal Government and a New Hampshire town to funds of an insolvent taxpayer, the Court first considered whether the town could qualify as a “judgment creditor” entitled to preference under the Tax Lien Act. 345 U. S., at 363-364. Only after deciding that question in the negative did the Court conclude that *533the United States obtained preference by operation of the priority statute. Id., at 365-366. The Government would now portray Gilbert Associates as a deviation from two other relatively recent opinions in which the Court held that the priority statute was not trumped by provisions of other statutes: United States v. Emory, 314 U. S., at 429-433 (the National Housing Act), and United States v. Key, 397 U. S., at 324-333 (Chapter X of the Bankruptcy Act). In each of those cases, however, there was no “plain inconsistency” between the commands of the priority statute and the other federal Act, nor was there reason to believe that application of the priority statute would frustrate Congress’ intent. Id., at 329. The same cannot be saidin the present suit.

The Government emphasizes that when Congress amended the Tax Lien Act in 1966, it declined to enact the American Bar Association’s proposal to modify the federal priority statute, and Congress again failed to enact a similar proposal in 1970. Both proposals would have expressly provided that the Government’s priority in insolvency does not displace valid liens and security interests, and therefore would have harmonized the priority statute with the Tax Lien Act. See Hearings on H. R. 11256 and 11290 before the House Committee on Ways and Means, 89th Cong., 2d Sess., 197 (1966) (hereinafter Hearings); S. 2197, 92d Cong., 1st Se1ss. (1971). But both proposals also would have significantly changed the priority statute in many other respects to follow the priority scheme created by the bankruptcy laws. See Hearings, at 85, 198; Plumb 10, n. 53, 33-37. The earlier proposal may have failed because its wide-ranging subject matter was beyond the House Ways and Means Committee’s jurisdiction. Id., at 8. The failure of the 1970 proposal in the Senate Judiciary Committee— explained by no reports or hearings — might merely reflect disagreement with the broad changes to the priority statute, or an assumption that the proposal was not needed because, as Justice Story had believed, the priority statute does not *534apply to prior perfected security interests, or any number of other views. Thus, the Committees’ failures to report the proposals to the entire Congress do not necessarily indicate that any legislator thought that the priority statute should supersede the Tax Lien Act in the adjudication of federal tax claims. They provide no support for the hypothesis that both Houses of Congress silently endorsed that position.

The actual measures taken by Congress provide a superior insight regarding its intent. As we have noted, the 1966 amendments to the Tax Lien Act bespeak a strong condemnation of secret liens, which unfairly defeat the expectations of innocent creditors and frustrate “the needs of our citizens for certainty and convenience in the legal rules governing their commercial dealings.” 112 Cong. Rec. 22227 (1966) (remarks of Rep. Byrnes); cf. United States v. Speers, 382 U. S. 266, 275 (1965) (referring to the “general policy against secret liens”). These policy concerns shed light on how Congress would want the conflicting statutory provisions to be harmonized:

“Liens may be a dry-as-dust part of the law, but they are not without significance in an industrial and commercial community where construction and credit are thought to have importance. One does not readily impute to Congress the intention that many common commercial liens should be congenitally unstable.” E. Brown, The Supreme Court, 1957 Term — Foreword: Process of Law, 72 Harv. L. Rev. 77, 87 (1958) (footnote omitted).

In sum, nothing in the text or the long history of interpreting the federal priority statute justifies the conclusion that it authorizes the equivalent of a secret lien as a substitute for the expressly authorized tax lien that Congress has said “shall not be valid” in a case of this kind.

The judgment of the Pennsylvania Supreme Court is affirmed.

It is so ordered.

concurring in part and concurring in the judgment.

I join the opinion of the Court except that portion which takes seriously, and thus encourages in the future, an argument that should be laughed out of court. The Government contended that 31 U. S. C. § 3713(a) must have priority over the Federal Tax Lien Act of 1966, because in 1966 and again in 1970 Congress “failed to enact” a proposal put forward by the American Bar Association that would have subordinated § 3713(a) to the Tax Lien Act, citing hearings before the House Committee on Ways and Means, and a bill proposed in, but not passed by, the Senate. See Brief for United States 25-27, and n. 10 (citing American Bar Association, Final Report of the Committee on Federal Liens 7,122-124 (1959), contained in Hearings on H. R. 11256 and 11290 before the House Committee on Ways and Means, 89th Cong., 2d Sess., 85, 199 (1966); S. 2197, 92d Cong., 1st Sess. (1971)). The Court responds that these rejected proposals “provide no support for the hypothesis that both Houses of Congress silently endorsed” the supremacy of §3713, ante, at 534, because those proposals contained other provisions as well, and might have been rejected because of those other provisions, or because Congress thought the existing law already made § 3713 supreme. This implies that, if the proposals had not contained those additional features, or if Members of Congress (or some part of them) had somehow made clear in the course of rejecting them that they wanted the existing supremacy of the Tax Lien Act to subsist, the rejection would “provide support” for the Government’s case.

That is not so, for several reasons. First and most obviously, Congress cannot express its will by a failure to legislate. The act of refusing to enact a law (if that can be called an act) has utterly no legal effect, and thus has utterly no place in a serious discussion of the law. The Constitution sets forth the only manner in which .the Members of Congress have the power to impose their will upon the country: *536by a bill that passes both Houses and is either signed by the President or repassed by a supermajority after his veto. Art. I, § 7. Everything else the Members of Congress do is either prelude or internal organization. Congress can no more express its will by not legislating than an individual Member can express his will by not voting.

Second, even if Congress could express its will by not legislating, the will of a later Congress that a law enacted by an earlier Congress should bear a particular meaning is of no effect whatever. The Constitution puts Congress in the business of writing new laws, not interpreting old ones. “[Ljater enacted laws ... do not declare the meaning of earlier law.” Almendarez-Torres v. United States, ante, at 237; ante, at 269-270 (Scalia, J., dissenting) (“This later amendment can of course not cause [the statute] to have meant, at the time of petitioner’s conviction, something different from what it then said”). If the enacted intent of a later Congress cannot change the meaning of an earlier statute, then it should go without saying that the later unenacted intent cannot possibly do so. It should go without saying, and it should go without arguing as well.

I have in the past been critical of the Court’s using the so-called legislative history of an enactment (hearings, committee reports, and floor debates) to determine its meaning. See, e. g., Conroy v. Aniskoff, 507 U. S. 511, 518-529 (1993) (Scalia, J., concurring in judgment); United States v. Thompson/Center Arms Co., 504 U. S. 505, 521 (1992) (Scalia, J., concurring in judgment); Blanchard v. Bergeron, 489 U. S. 87, 98-100 (1989) (Scalia, J., concurring in part and concurring in judgment). Today, however, the Court’s fascination with the files of Congress (we must consult them, because they are there) is carried to a new silly extreme. Today’s opinion ever-so-earefully analyzes, not legislative history, but the history of legislation-that-never-was. If we take this sort of material seriously, we require conscientious counsel to investigate (at clients’ expense) not only the hear*537ings, committee reports, and floor debates pertaining to the history of the law at issue (which is bad enough), but to find, and then investigate the hearings, committee reports, and floor debates pertaining to, later bills on the same subject that were never enacted. This is beyond all reason, and we should say so.