OFFICEMAX, INC., Plaintiff-Appellee, v. UNITED STATES of America, Defendant-Appellant, 428 F.3d 583

SUTTON, J., delivered the opinion of the court, in which ROSEN, D. J., joined. ROGERS, J. (pp. 600-05), delivered a separate dissenting opinion. OPINION SUTTON, Circuit Judge. When a party presents the question whether “and” means “or,” it is tempting to be dismissive of the claim or, worse, to make a crack about the demise of the rule of law.

But in this instance the disputed “and” appears in the context of several uses of the term that are alternately conjunctive and disjunctive and as much as nine billion dollars in potential tax refund claims (according to the government) rest on the resolution of the issue in this case and others, both of which prompt us to be anything but dismissive of the question. At issue is the meaning of “toll telephone service,” which Congress has subjected to a three-percent federal excise tax.

The relevant legislation defines the phrase as “a telephonic quality communication for which [ ] there is a toll charge which varies in amount with the distance and elapsed transmission time of each individual communication.” 26 U.S.C. § 4252(b)(1) (emphasis added). According to the IRS, the definition means that the tax applies when the telephone company assesses a toll charge that varies in amount by either the distance or elapsed transmission time of each individual communication, or both.

According to OfficeMax, the definition means that the tax applies when the telephone company assesses a toll charge that varies in amount by both the distance and elapsed transmission time of each individual communication.

Given the traditional presumption that Congress uses “and” conjunctively, given other contextual clues supporting a conjunctive reading, given the awkwardness of construing the provision as the government does and given the historical fact that the one provider of long-distance telephone service in 1965, when the definition was adopted, charged for phone calls both by distance and time, we conclude that a toll charge must vary by both distance and elapsed transmission time in order to be taxed. We thus hold for the taxpayer and affirm. I. A.

In 1898, 22 years after Alexander Graham Bell’s work led to the invention of the telephone, Congress imposed the first tax on telephone service. Designed to curb the federal deficit caused by the Spanish-American War, the temporary tax applied to “every person, firm or corporation owning or operating any telephone line or lines” and charged one cent for “messages or conversations transmitted over their respective lines ... for which a charge of fifteen cents or more was imposed.” 30 Stat. at 460, Pub.L; No. 55-133 (1898). Congress repealed the tax as scheduled in 1902.

See Louis Alan Talley, The Federal Excise Tax on Telephone Service: A History 1 (Congressional Research Service 2001), available at http://www. law.umary-land.edu /marshall/cr sreports/ crsdocu-ments/RL30553_01042001.pdf (hereinafter “Talley”).

Beginning in 1914 and continuing through 1916, in ■ response to falling revenues caused by the start of World War I, Congress reenacted the tax — this time applying it to owners or operators of “any telegraph or telephone line or lines” and charging the same one cent for “dispatches, messages, or conversations originated at each of their respective exchanges, toll stations, or offices, and transmitted thence over their lines ... for which a charge of 15 cents or more was imposed.” 30 Stat. at 761, Pub.L. No. 217; see Talley at 1-2.

Between 1917 and 1924, in response to the United States’ entry into the war, Congress imposed a higher tax — “5 cents upon each telegraph, telephone, or radio, dispatch, message, or conversation, which originates within the United States, and for the transmission of which a charge of 15 cents or more is imposed.” 40 Stat. 300, Pub.L. No. 50. Congress increased the tax to 10 cents for calls costing more than 50 cents in 1919. See Talley at 3. The Great Depression prompted the next iteration of the tax, and it has existed in one form or another ever since. Id.

In 1932, Congress enacted a tax on “each telegraph, telephone, cable, or radio dispatch, message, or conversation, which originates ... within the United States,” applying different rates to each service. 47 Stat. at 270, Pub.L. No. 154. In 1958, Congress changed the organization and labeling of the “communication services” tax, 72 Stat. at 1289, Pub.L. No. 85-859, dividing it into six categories, two of which concern us here: (1) “general telephone service,” i.e., local telephone services; and (2) “toll telephone service,” i.e., long-distance service involving a toll charge. 72 Stat. at 1290, Pub.L. No. 85-859.

The local (“general”) and long-distance (“toll”) telephone service taxes continued at varying rates through 1959, when Congress passed legislation repealing the tax on July 1, 1960. But by that date new legislation had enacted a one-year extension, and subsequent one-year extensions maintained the tax through 1965. Talley at 4-5.

In 1965, Congress enacted the Excise Tax Reduction Act, which called for the repeal of most excise taxes, including an immediate reduction of the telephone tax from ten percent to three percent followed by an annual one-percent reduction thereafter until the tax was completely repealed on January 1, 1969. 79 Stat. at 145, Pub.L. No. 89-44.

The 1965 legislation retained three categories of taxable services: “local telephone service,” “toll telephone service” and “teletypewriter exchange service.” As to toll telephone service, originally defined in the 1958 legislation as “a telephone or radio telephone message or conversation for which (1) there is a toll charge, and (2) the charge is paid within the United States,” Congress enacted the following definition, unchanged to this day: (1) a telephonic quality communication for which (

A) there is a toll charge which varies in amount with the distance and elapsed transmission time of each individual communication and (

B) the charge is paid within the United States, and (2) a service which entitles the subscriber, upon payment of a periodic charge (determined as a flat amount or upon the basis of total elapsed transmission time), to the privilege of an unlimited number of telephonic communications to or from all or a substantial portion of the persons having telephone or radio telephone stations in a specified area which is outside the local telephone system area in which the station provided with this service is located. 79 Stat. at 146, Pub.L. No. 89-44; 26 U.S.C. § 4252(b).

When Congress enacted this legislation (and for many years before and after its passage), the country had just one provider of long-distance service — AT & T. See, e.g., Am. Bankers Ins. Group v. United States, 408 F.3d 1328, 1333 (11th Cir.2005). And when Congress passed this version of the tax, AT & T already imposed a toll charge based on variations in both the time and distance of each call. Id. In 1966, in response to spending pressures brought on by the Vietnam War, Congress began another series of extensions of the tax.

The extensions continued through the 1970s and 1980s, with legislation maintaining the tax at varying rates through 1991. During these years, long-distance telecommunications transformed from a service offered by only one commercial provider to a service offered by several companies in a competitive market. The Federal Communications Commission issued a 1971 Specialized Common Carrier decision opening the long distance market to competition. See Michael R.

Ward, Bureau of Economics Staff Report: Measurements of Market Power in Long Distance Telecommunications (Federal Trade Commission 1995), available at http:// www.ftc.gov/ reports/teleeomm.pdf. An antitrust suit filed by the Department of Justice and the eventual settlement of that litigation in 1982 spurred further competition in the long-distance market. “While AT & T’s share of the long distance market’s revenues in 1982 was 95%, by 1987 its market share had fallen to 80%, and [in 1993 was] about 60%.

By 1991, MCI’s and Sprint’s revenue market shares had climbed to 17% and 10% respectively.” Id. at 4 (citation omitted). In 1990, Congress made the excise tax permanent at a three-percent rate, and, aside from a 1997 amendment clarifying that calling cards were taxable, it did not revisit the issue again during that decade. The 1990s, however, did see a change in the way companies billed for long-distance telephone services, with long-distance companies beginning to offer for the first time nationwide long-distance plans for flat per-minute rates.

AT & T abandoned its distance-and-time formula in favor of a time-only formula in 1997. Congress most recently considered the tax in 2000, when it enacted a package of spending and tax bills that proposed repealing the tax. A presidential veto, however, prevented the bill from becoming a law. B. From the first quarter of 1999 through the last quarter of 2002, OfficeMax purchased long distance telephone service from MCI.

During this time, MCI charged OfficeMax one flat per-minute rate for every long distance call it made between States, another flat per-minute rate for long distance calls within each State (the rate'of which varied by State) and another flat per-minute rate for long distance calls to international destinations (the rate of which varied by the destination country). At the end of each call, MCI rounded up to the nearest minute, multiplied the number of minutes by the flat per-minute rate, and assessed a total charge for the call.

At the end of the billing cycle, MCI added up the total charges for all of the calls and billed OfficeMax for the amount. In addition, MCI collected from OfficeMax, and remitted to the federal government, the three-percent excise tax ostensibly imposed on these calls. On April 6, 2002, and February 13, 2003, OfficeMax filed claims with the federal government for a full refund of the $380,296.72 in excise taxes it had paid through MCI.

In seeking a refund, OfficeMax argued that MCI was not providing a taxable service — neither a “toll telephone service,” a “local telephone service” nor a “teletypewriter exchange service,” as Congress has defined those phrases. Most pertinently, the company argued, it had not provided a “toll telephone service” because it did not provide “a telephonic quality communication for which [] there is . a toll charge which varies in amount with the distance and elapsed transmission time of each individual communication,” 26 U.S.C. § 4252(b)(1) (emphasis added). The district court granted OfficeMax’s motion for

summary judgment, reasoning that the statutory language requires a telephone service plan’s charges to vary both with the distance and elapsed time of each call. It also rejected the IRS’s alternative arguments that OfficeMax’s long-distance plan met (1) an alternative definition for “toll telephone service” provided in § 4252(b)(2) or (2) the definition of “local telephone service,” § 4252(a), both of which must pay the same tax. II.

Section 4251 of Title 26 imposes a tax on “amounts paid for communications services.” It defines “communications services” as “(

A) local telephone service; (

B) toll telephone service; and (

C) teletypewriter exchange service.” Id.

Section 4252 defines “toll telephone service” as: (1) a telephonic quality communication for which (

A) there is a toll charge which varies in amount with the distance and elapsed transmission time of each individual communication and (

B) the charge is paid within the United States, and (2) a service which entitles the subscriber, upon payment of a periodic charge (determined as a flat amount or upon the basis of total elapsed transmission time), to the privilege of an unlimited number of telephonic communications to or from all or a substantial portion of the persons having telephone or radio telephone stations in a specified area which is outside the local telephone system area in which the station provided with this service is located. 26 U.S.C. § 4252(b) (emphasis added). The

section defines “local telephone service” as: (1) the access to a local telephone system, and the privilege of telephonic quality communication with substantially all persons having telephone or radio telephone stations constituting a part of such local telephone system, and (2) any facility or service provided in connection with a service described in paragraph (1). 26 U.S.C. § 4252(a). The

section defines “teletypewriter exchange service” as: the access from a teletypewriter or other data station to the teletypewriter exchange system of which such station is a part, and the privilege of intercommunication by such station with substantially all persons having teletypewriter or other data stations constituting a part of the same teletypewriter exchange system, to which the subscriber is entitled upon payment of a charge or charges (whether such charge or charges are determined as a flat periodic amount, on the basis of distance and elapsed transmission time, or in some other manner). 26 U.S.C. § 4252(c).

As these sections of the tax code reveal, Congress used “and” in more than one sense when it enacted these provisions in 1965 — giving it a conjunctive meaning (requiring all items) in some places and giving it a disjunctive or cumulative meaning (allowing any of the items) in other places.

Section 4252(b)(1), for example, requires that the toll charge vary in amount with the distance and elapsed time of each individual communication “and” that the charge be paid within the United States. Both sides of this “and,” the parties agree, must be satisfied for the telephone service to be taxable. The same is true of other uses of the term in these provisions.

See 26 U.S.C. § 4252(a)(1) (defining one alternative meaning of “local telephone service” as “access to a local telephone system, and the privilege of telephonic quality communication with ... such local telephone system”) (emphasis added); § 4252(c) (defining “teletypewriter exchange service” as “access ... to the teletypewriter exchange system ... and the privilege of intercommunication ... with substantially all persons ... [in] the same teletypewriter exchange system”) (emphasis added). At times, however, Congress used the term differently.

Take § 4251: it defines “communications services” as “local telephone service, toll telephone service, and teletypewriter exchange service.” 26 U.S.C. § 4251 (emphasis added). No one doubts the disjunctive (or cumulative) nature of this usage because the

definitions of the services in § 4252 are generally mutually exclusive and because no service exists that can satisfy all three

definitions at once.

Section 4252 likewise uses “and” to connect a list of alternative

definitions of toll telephone service found in §§ 4252(b)(1) and (b)(2). Here, too, no one contests that a “toll charge which varies in amount with the distance and elapsed transmission time of each individual communication,” § 4252(b)(1), cannot simultaneously be a “periodic charge (determined as a flat amount or upon the basis of total elapsed transmission time) [for] the privilege of an unlimited number of telephonic communications,” § 4252(b)(2). As these different usages reveal, there is more to “and” than meets the eye.

The case, then, does not simply turn on the intuition that “and” means “and,” “or” means “or,” and never the twain shall meet. For several reasons, however, the “and” at issue in this dispute requires a telephone company’s toll charge to vary both by distance and time to be taxable. First, dictionary

definitions, legal usage guides and case law compel us to start from the premise that “and” usually does not mean “or.” Dictionaries consistently feature a conjunctive definition of “and” as the primary meaning of the word, see, e.g., Webster’s Third New International Dictionary 80 (2002) (“along with or together with ... added to or linked to ... as well as”), or the first usage of the word historically, see Oxford English Dictionary (2d ed, 1989) (“[introducing a word, clause, or sentence, which is to be taken side by side with, along with, or in addition to, that which precedes it”); Caleb Nelson, Origi- nalism and Interpretive Conventions, 70 U.

Chi. L.Rev. 519, 519 (2003) (“In all living languages, the conventional usages of individual words change over time. For illustrations, one need only consult the Oxford English Dictionary, which arranges its

definitions of each word so that they proceed from the earliest usages to those that were introduced more recently.”). Cf Webster’s Third New International Dictionary 80 (2002) (alternative six of the second definition of “and”: “reference to either or both of two alternatives ... especially] in legal language when also plainly intended to mean or ”). Legal usage guides are to the same effect. See 1A Norman J. Singer, Statutes and Statutory Construction § 21.14 at 179-80 (6th ed. 2002) (“Statutory phrases separated by the word ‘and’ are usually to be interpreted in the conjunctive.”); id. at 183-84 (“While there may be circumstances which call for an

interpretation of the words ‘and’ and ‘or,’ ordinarily these words are not interchangeable. The terms ‘and’ and ‘or’ are often misused in drafting statutes.... The literal meaning of these terms should be followed unless it renders the statute inoperable or the meaning becomes questionable.”); 1 Bouvier’s Law Dictionary and Concise Encyclopedia 194-95 (3d Revision 1914) (“A conjunction connecting words or phrases expressing the idea that the latter is to be added to or taken along with the first. It is said to be equivalent to ‘as well as.’ ”); id. at 195 (“It is sometimes construed as meaning ‘or.’ ”); Bryan A.

Garner, A Dictionary of Modem Legal Usage 55 (2d ed. 1995) (“Oddly, and is frequently misused for or where a single noun, or one of two nouns, is called for.... Sloppy drafting sometimes leads courts to recognize that and in a given context means or, much to the chagrin of some judges.”) (quoting MacDonald v. Pan Am. World Airways, Inc., 859 F.2d 742, 746 (9th Cir.1988)) (Kozinski, J., dissenting) (‘We give our language, and our language-dependent legal system, a body blow when we hold that it is reasonable to read ‘or’ for ‘and.’ ”).

Reflecting these traditional assumptions about the meaning of the term, the Supreme Court has said that “and” presumptively should be read in its “ordinary” conjunctive sense unless the “context” in which the term is used or “other provisions of the statute” dictate a contrary

interpretation. See Crooks v. Harrelson, 282 U.S. 55, 58, 51 S.Ct. 49, 75 L.Ed. 156 (1930) (We find nothing in the context or in other provisions of the statute which warrants the conclusion that the word ‘and’ was used otherwise than in its ordinary sense.”); United States v. Field, 255 U.S. 257, 262, 41 S.Ct. 256, 65 L.Ed. 617 (1921) (“These conditions are expressed conjunc-tively; and it would be inadmissible, in construing a taxing act, to read them as if prescribed disjunctively.”); City of Rome v.

United States, 446 U.S. 156, 172, 100 S.Ct. 1548, 64 L.Ed.2d 119 (1980) (“By describing the elements of discriminatory purpose and effect in the conjunctive, Congress plainly intended that a voting practice not be precleared unless both discriminatory purpose and effect are absent.”). The federal courts of appeals have done likewise. See, e.g., Bruce v. First Fed.

Sav. & Loan Ass’n of Conroe Inc., 837 F.2d 712, 715 (5th Cir.1988) (“The word ‘and’ is therefore to be accepted for its conjunctive connotation rather than as a word interchangeable with ‘or’ except where strict grammatical construction will frustrate clear legislative intent.”). When courts have interpreted “and” disjunctively, they have done so only to avoid an incoherent reading of a statute. See Slodov v.

United States, 436 U.S. 238, 246-47, 98 S.Ct. 1778, 56 L.Ed.2d 251 (1978) (interpreting a statute disjunctively that limited personal liability to “any person required to collect, truthfully account for and pay over any tax imposed by this title” because a conjunctive reading would lead to an absurd result: “Because the duty to pay over the tax arises only at the quarter’s end, a ‘responsible person’ who willfully failed to collect taxes would escape personal liability for that failure simply by resigning his position.”)(quotations and brackets omitted); Sosa v.

Chase Manhattan Mortgage Corp., 348 F.3d 979, 981, 983 (11th Cir.2003) (interpreting as disjunctive the phrase “[n]o person shall give and no person shall accept any portion” because “[extending liability only if there were both a culpable giver and acceptor of an unearned fee would lead to irrational results”); United States v.

Bonilla-Montenegro, 331 F.3d 1047, 1050-51 (9th Cir.2003) (interpreting as disjunctive a definition of “crime of violence” — “an offense ... that has as an element the use ... of physical force ... and [ ] includes murder, manslaughter, kidnapping, [etc.]” — because it would “def[y] common sense” to require an enumerated crime of violence also to have a use-of-force elementXquotation omitted); Peacock v.

Lubbock Compress Co., 252 F.2d 892, 893 (5th Cir.1958) (interpreting as disjunctive “an employer engaged in ... the ginning and compressing of cotton” because “compressing is an operation entirely removed from ginning and [] the two are never carried on together”). Second, consistent with the inquiry proposed by these

definitions and cases, there is nothing unusual (or for that matter absurd) about requiring a telephone toll charge to vary both with distance and time before subjecting it to tax. That is exactly the way long-distance calls were charged for a considerable period of time. In 1965, when Congress enacted this definition, the country had just one long-distance telephone provider, and AT & T billed for long-distance calls based both on the distance and time of the call or on a WATS-line basis (i.e., a flat rate for unlimited calling).

Not surprisingly, when the drafters of § 4252 defined taxable long-distance service, they adopted a definition that turns on the distance-and-time measure of charging for the call, see § 4252(1), and a definition that turns on the WATS-line method of billing, see § 4252(2). There is something odd, by contrast, about giving the distance-and-time requirement a disjunctive

interpretation. Why would Congress write a statute that says toll charges may vary by time or distance? In the last decade, it is true, time-varied toll charges have proved to be a viable stand-alone option. How strange, however, to impose a toll charge that varies solely by distance, with all calls from, say, Columbus, Ohio to Bozeman, Montana costing the same amount regardless of whether they last 60 seconds, 60 minutes or 60 days. At no point in this litigation has the government given us any indication that there is now, or ever has been, such a curious billing practice. The government, we suppose, might respond that the drafters had a different disjunctive

interpretation of “and” in mind. Instead of taxing calls charged on one basis (distance) or the other (time), what Congress really meant to do was to tax calls charged on the basis of time and distance together or on the basis of time alone. Given the presumption that “and” has a conjunctive meaning and given the understandable reason why Congress would have embraced that meaning here, it is hard to see why we should adopt not just a disjunctive reading of the term but such a nuanced and changeable one. Third, context bolsters this conclusion.

In writing the 1965 legislation, Congress defined three types of taxable communication services: “toll telephone service,” “local telephone service” and “teletypewriter exchange service.” It chose to define local telephone service simply by reference to “access” to “a” local telephone system; it chose to define the other two services by reference to the method of charging for those services. As to toll telephone service, we have seen, Congress taxed long-distance phone calls charged on the basis of time and distance.

In defining teletypewriter exchange, it chose not to use a time- and-distance methodology alone and chose not to deploy the all-purpose, protean “and” that the government claims was used in defining toll telephone service. Rather than limit the taxable communication services to one billing method, it imposed the tax on charged teletypewriter services “whether such charge or charges are determined as a flat periodic amount, on the basis of distance and elapsed transmission time, or in some other manner.” 26 U.S.C. § 4252(c).

This phrase, enacted at the same time as § 4252(1)(A), shows that Congress used a distance-and-time formulation elsewhere in the same bill and plainly meant it to have a traditionally conjunctive connotation there. See Gustafson v. Alloyd Co., Inc., 513 U.S. 561, 570, 115 S.Ct. 1061, 131 L.Ed.2d 1 (1995) (recognizing “that identical words used in different parts of the same act are intended to have the same meaning”)(quotation omitted).

And it shows the ease with which Congress could have demonstrated the taxability of different methods of toll charges or all charges (“or in some other manner”) had it wished to do so when defining toll telephone service. Fourth, in reaching this conclusion, we are not alone. Every court to reach this issue — save one district court subsequently reversed — has concluded that the statute unambiguously requires variance by both distance and elapsed transmission time. See Am. Bankers Ins. Group v. United States, 408 F.3d 1328 (11th Cir.2005) (reversing Am. Bankers Ins. Group, Inc. v.

United States, 308 F.Supp.2d 1360 (S.D.Fla.2004)); Am. Online, Inc. v. United States, 64 Fed.Cl. 571 (Fed.Cl.2005); Honeywell Int’l, Inc. v. United States, 64 Fed.Cl. 188 (Fed.Cl.2005); National Railroad Passenger Corp., AMTRAK v. United States, 338 F.Supp.2d 22 (D.D.C.2004); Hewlett-Packard Co. v. United States, No. 04-03832, 2005 WL 1865419, 2005 U.S. Dist. LEXIS 199972 (N.D.Cal. Aug. 5, 2005); Reese Bros. v. United States, No. 03-745, 2004 U.S. Dist. LEXIS 27507 (W.D.Pa. Oct. 20, 2004); Fortis, Inc. v. United States, No. 03 Civ. 5137, 2004 WL 2085528, 2004 U.S. Dist. LEXIS 18686 (S.D.N.Y. Sept. 16, 2004).

We are comfortable in charting a similar course here. Attempting to counter this conclusion, the government first argues that we must account for the other uses of “and” in this legislation that have a disjunctive or cumulative meaning. But the case law referenced above adequately accounts for this argument. As those cases provide, ‘.‘and” presumptively has a conjunctive meaning save when that

interpretation makes little or no sense in context. Sections 4251 and 4252, it is true, contain three instances where “and” is not used in its customary conjunctive sense. But in each case, context requires the term to be construed disjunctively because it conjoins a list of mutually exclusive alternatives. See § 4251(b)(1) (taxing “(

A) local telephone service; (

B) toll telephone service; and (

C) teletypewriter exchange service”) (emphasis added); § 4252(b) (defining “toll telephone service” as the definition at issue in this case “and” an alternative definition that could not be satisfied simultaneously) (emphasis added); § 4252(a) (defining “local telephone service” as “(1) [a substantive definition] and (2) any facility or service provided in connection with a service described in paragraph (1)”) (emphasis added). In marked contrast, a toll charge simultaneously may be based on variations in distance and time.

The government next argues that these other uses of “and” at a minimum establish ambiguity about the meaning of the term in § 4252(b)(1)(A). “[Ajmbiguity,” however, “is a creature not of definitional possibilities but of statutory context.” Brown v. Gardner, 513 U.S. 115, 118, 115 S.Ct. 552, 130 L.Ed.2d 462 (1994). And the question whether a statute is ambiguous arises after, not before, a court applies traditional canons of

interpretation' — the most important here being the context in which the word appears and the presumption that “and” carries a conjunctive connotation. See Chevron, U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 843 n. 9, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) (“If a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue, that intention is the law and must be given effect.”); Robinson v.

Shell Oil Co., 519 U.S. 337, 341, 117 S.Ct. 843, 136 L.Ed.2d 808 (1997) (“The plainness or ambiguity of statutory language is determined by reference to the language itself, the specific context in which that language is used, and the broader context of the statute as a whole.”). Nor does the government’s invocation of legislative history advance its cause. For one, the government has not presented the necessary warrant — a case of statutory ambiguity — for commencing a search of the statute’s legislative history. See Ex parte Collett, 337 U.S. 55, 61, 69 S.Ct. 944, 93 L.Ed. 1207 (1949); Chapman v.

The Higbee Co., 319 F.3d 825, 829 (6th Cir.2003); see also United States v. R.L.C., 503 U.S. 291, 307-11, 112 S.Ct. 1329, 117 L.Ed.2d 559 (1992) (Scalia, J., concurring). For another, even the most far-reaching search of the legislative history behind the 1965 Act uncovers little. The committee report describing this provision elucidates nothing; it merely parrots the language that Congress ultimately enacted into law. Cf. Pub.L. No. 89-44, 79 Stat. 136, 146, with H.R.Rep.

No. 433 (1965), reprinted in 1965 U.S.C.C.A.N. 1645, 1677 (“Toll telephone service is defined as being a telephonic quality communication for which a toll charge is made which varies in amount with the distance and the lapsed transmission time of individual communications, but only if the charge is paid within the United States.”). And the explanations for passing the 1965 Act do as much to undermine the government’s position as to support it.

When Congress enacted the present definition of toll telephone service in 1965— changing it from a broad definition of a service for which a toll was charged and paid within the United States, Pub.L. No. 85-859, 72 Stat. 1275, 1290 (1958) — it defined the phrase in a way that accounted for two distinct billing methods employed in 1965 by AT & T, the sole long-distance provider at that time. As we have noted, the first method calculated a charge for each call based on elapsed time and distance.

The distance between the originator and receiver of the call would fall within one of several concentric mileage bands measured from the call’s origin. The per-minute rate of the call varied depending on which mileage band applied. This system was codified in § 4252(b)(1). The second method, known as “Wide Area Telephone Service” or “WATS,” calculated one charge per billing period, based on either a flat rate or the total elapsed time of all calls made dining the period. This charge entitled the customer to an unlimited number of calls regardless of distance, and it was codified in § 4252(b)(2).

See 1965 USCCAN 1676-77. All of this supports our

interpretation. It shows why Congress would define toll telephone service as requiring a charge that varied both by time and distance. And it shows why Congress would define taxable telephone calls based on how a private company billed the calls: There was just one company; AT & T’s job was to collect the tax, not to pay it; and its monopoly status eliminated any commercial pressure to accommodate tax-sensitive clients. Persisting, the government points out that

(1) Congress crafted the new

definitions “to make it clear that it is the service as such which is being taxed and not merely the equipment being supplied,” 1965 USCCAN 1676, not to exempt certain long-distance phone calls from taxation, and

(2) Congress added an important exemption for “private communication services” (e.g., intercoms) but again did not aim to eliminate from taxation certain long-distance calls because of the way they were billed. Even if we were to infer from these shards of legislative evidence that Congress meant to tax all long-distance service as it existed in 1965 — through the creation of a definition based on extant billing methods — that does not mean we should interpret the same language to include all long distance service as it exists today.

A statute enacted in 1890 that imposed an excise tax on sales, say, of “any vehicle of transportation” would cover airplanes, even though they were not invented until several years after 1890. Yet a statute enacted in 1890 that taxed sales of “any vehicle of transportation designed to convey passengers by land or sea” would not cover sales of traditional airplanes, even though the legislator’s purpose in 1890 could fairly be characterized as taxing all modes of transportation.

Why should it be any different for this tax when time brings to an end an uncompetitive communications market and leads to wireless telephone communication (for which distance-based billing may make less sense)? As these examples suggest and as the terms of the 1965 legislation require, “[i]t is ultimately the provisions of our laws rather than the principal concerns of our legislators by which we are governed.” Oncale v. Sundowner Offshore Serv., Inc., 523 U.S. 75, 79-80, 118 S.Ct. 998, 140 L.Ed.2d 201 (1998).

A legislature that chooses to define eligibility for taxation based on how a private entity chooses to charge for the service, moreover, can hardly be treated as a body that means to impose a tax for all time. Just the opposite seems true, as the approach ultimately cedes the taxing authority to private entities, which is presumably why this method of taxation infrequently appears in Title 26. A party willing to invoke broad generalizations of legislative purpose, at any rate, must account for equally plausible — yet vastly different — characterizations of that intent.

When Congress wrote the present definition of toll telephone service in 1965, it planned to eliminate the tax in its entirety over the following three years. Entitled the “Excise Tax Reduction Act,” the 1965 legislation imposed an immediate reduction of the telephone tax (from 10 percent to 3 percent), established an annual one-percent reduction of the tax, then called for the elimination of the tax on January 1, 1969.

See 1965 USCCAN 1676; id. at 1645, 1649 (noting that the legislation “either in the current fiscal year, or over the next 3 subsequent years, removes entirely [the] miscellany of selective excise taxes [including the communications services excise tax] except those designed to serve certain specific purposes”); id. at 1655 (House committee report noted that “[i]n large part, these taxes were imposed as emergency wartime measures (either in World War II or the Korean War) or as an emergency method of raising revenues at the time of the depression of the 1930’s”).

The Congress that enacted this legislation thus did not contemplate the continued application of the language past January 1, 1969, let alone to 1999. No matter how clear its design to include the lion’s share of long-distance service in existence between 1965 and 1969, that does not lead to the inference that the 1965 Congress meant to tax the lion’s share of long distance service in existence between 1999 and 2002.

In the end, “application of ‘broad purposes’ of legislation at the expense of specific provisions ignores the complexity of the problems Congress is called upon to address and the dynamics of legislative action.” Board of Governors of Federal Reserve System v. Dimension Financial Corp., 474 U.S. 361, 373-74, 106 S.Ct. 681, 88 L.Ed.2d 691 (1986); Rodriguez v. United States, 480 U.S. 522, 526, 107 S.Ct. 1391, 94 L.Ed.2d 533 (1987) (per curiam) (“[I]t frustrates rather than effectuates legislative intent simplistically to assume that whatever furthers the statute’s primary objective must be the law.”).

The government also fails to gain ground by emphasizing that “the statute does not state that the charge must vary with ‘both distance and elapsed time.’ ” U.S. Br. at 27. Thé presumption that “and” has a conjunctive meaning, however, spares legislative drafters from having to use a belt-and-suspenders approach (or, shall we say, both belt-and-suspenders approach) every time they deploy the term. The argument also comes perilously close to suggesting an interpretive presumption in favor of the government and against the taxpayer.

Regardless of the current status of the “traditional canon that construes revenue-raising laws against their drafter,” United Dominion Indus. v. United States, 532 U.S. 822, 839, 121 S.Ct. 1934, 150 L.Ed.2d 45 (2001) (Thomas, J., concurring), the government has not identified any case establishing an opposite presumption. See Bowers v. New York & Albany Lighterage Co., 273 U.S. 346, 350, 47 S.Ct. 389, 71 L.Ed. 676 (1927) (“The provision is part of a taxing statute; and such laws are to be interpreted liberally in favor of the taxpayers.”); United States v.

Merriam, 263 U.S. 179, 188, 44 S.Ct. 69, 68 L.Ed. 240 (1923) (“If the words are doubtful, the doubt must be resolved against the government and in favor of the taxpayer.”); Benziger v. United States, 192 U.S. 38, 55, 24 S.Ct. 189, 48 L.Ed. 331 (1904); American Net & Twine Co. v. Worthington, 141 U.S. 468, 474, 12 S.Ct. 55, 35 L.Ed. 821 (1891); Leavell v. Blades, 237 Mo. 695, 141 S.W. 893, 894 (1911) (“When the tax gatherer puts his finger on the citizen, he must also put his finger on the law permitting it.”).

The government next urges us to defer to a 1979 IRS revenue ruling holding that a ship-to-shore satellite service that charged users without reference to distance satisfied the definition of “toll telephone service.” Rev. Rule 79^404, 1979-2 C.B. 382. But the one-page analysis in this revenue ruling does not contain the traditional hallmarks for receiving deference, whether under Chevron, or under Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944).

The revenue ruling acknowledges that “[laterally, the service provided in this case does not come within the definition of ... ‘toll telephone service’ ... because the charge for such service does not vary with distance and therefore does not meet the requirement of

section 4252(b)(1).” Rev. Rule 79-404, 1979-2 C.B. 382, *3. It then echoes the point, explaining that “[t]he toll charges described in

section 4252(b)(1) ... vary in amount with both distance and elapsed transmission time of the individual communication.” Id. at *6 (emphasis added). Despite this “literal[]” meaning of the statute, the IRS determined that the communications service was “essentially” a toll telephone service “in order to comport with the legislative intent” of the statute, which it discerned as seeking to encompass all long-distance service in existence in 1965 and therefore all long-distance service in the future (including what was being provided in 1979). In the IRS’s words: The service in this case is essentially “toll telephone service” as described in

section 4252(b)(1) of the Code, even though the charge for calls between remote maritime stations and stations in the United States vary with elapsed transmission time only. The toll charges described in

section 4252(b)(1), that vary in amount with both distance and elapsed transmission time of the individual communication, reflect Congress’ understanding of how the charges for long distance calls were computed at the time the

section was enacted. The intent of the statute would be frustrated if a new type of service otherwise within such intent were held to be nontaxable merely because charges for it are determined in a manner which is not within the literal language of the statute. Id. at *6-7. When the Court decided Chevron, five years after this revenue ruling, we do not think it had in mind deferring to tax administrators who wished to impose taxes on services that were “essentially,” but not “literally,” within the contours of a taxing statute.

Even if this circuit gave Chevron deference to revenue rulings, which it does not, see Aeroquip-Vickers, Inc. v. Comm’r, 347 F.3d 173, 181 (6th Cir.2003) (declining to give Chevron deference to revenue rulings because they lack the “force of law”); Ammex, Inc. v. United States, 367 F.3d 530, 535 n. 2 (6th Cir.2004) (same), this revenue ruling would not clear step one of the Chevron test. Likewise, even assuming revenue rulings were entitled to Skidmore respect, cf. United States v.

Cleveland Indians Baseball Co., 532 U.S. 200, 220, 121 S.Ct. 1433, 149 L.Ed.2d 401 (2001) (declining to decide whether revenue rulings are entitled to deference at all), that deference would apply to a revenue ruling “only to the extent that [it has] the ‘power to persuade.’ ” Christensen v. Harris County, 529 U.S. 576, 587, 120 S.Ct. 1655, 146 L.Ed.2d 621 (2000) (quoting Skidmore, 323 U.S. at 140, 65 S.Ct. 161).

Under Skidmore, courts should consider “ ‘the thoroughness evident in [the ruling’s] consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control.’ ” United States v. Mead Corp., 533 U.S. 218, 228, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) (quoting Skidmore, 323 U.S. at 140, 65 S.Ct. 161). Generated by an agency that frequently insists that its citizens “turn square corners,” Rock Island, Ark. & La. R.Co. v.

United States, 254 U.S. 141, 143, 41 S.Ct. 55, 65 L.Ed. 188 (1920), this revenue ruling no more supports an “essential” method of tax imposition than it does an “essential” method of tax payment. As other parts of our opinion explain, neither the legislative history nor one-sided, generalizations about the purpose of the law make the case for straying from the ordinary meaning of the language in the statute. Agencies in the end receive Skidmore respect because of the persuasiveness of their reasoning, not in spite of it.

No more availing is the government’s argument that, no matter how unpersuasive the revenue ruling is, we must defer to it because Congress repeatedly has reenacted the excise tax since 1979. The Supreme Court has not spoken with one mind on this issue. On the one hand, the Court has said that “[w]here an agency’s statutory construction has been fully brought to the attention of the public and the Congress, and the latter has not sought to alter that

interpretation although it has amended the statute in other respects, then presumably the legislative intent has been correctly discerned.” North Haven Bd. of Educ. v. Bell, 456 U.S. 512, 535, 102 S.Ct. 1912, 72 L.Ed.2d 299 (1982). On the other hand, the Court has said: “We walk on quicksand when we try to find in the absence of corrective legislation a controlling legal principle.” Helvering v. Hallock, 309 U.S. 106, 121, 60 S.Ct. 444, 84 L.Ed. 604 (1940); see also Central Bank, N.A. v. First Interstate Bank, N.A., 511 U.S. 164, 186, 114 S.Ct. 1439, 128 L.Ed.2d 119 (1994).

Never mind: even the cases applying the reenactment doctrine impose limitations on it that are applicable here — “where the law is plain, subsequent reenactment does not constitute an adoption of a previous administrative construction.” Brown, 513 U.S. at 121, 115 S.Ct. 552. And the doctrine does not apply when “the record of congressionral discussion preceding reenactment makes no reference to the [interpretation], and there is no other evidence to suggest that Congress was even aware of the [agency’s] interpretive position.” Id.

Here, the government fails to identify any evidence in the telephone-excise-tax legislation from 1979 to the present that discusses or acknowledges the existence of the 1979 revenue ruling. The government alternatively urges us to accept another

interpretation of the provision, one not yet embraced by any administrative ruling. It claims that “ ‘distance’ clearly is shorthand for toll rate,” so that even if the statute requires that the toll charge vary by both distance and elapsed time, a service would qualify if the toll charge varied by both toll rate (even a constant toll rate) and elapsed time. U.S. Br. at 35. Under AT & T’s billing method in 1965, the government explains, all calls within a certain mileage band were charged at the same per-minute rate regardless of any difference in actual distance spanned by a given call.

Because this system could produce equal billing for calls of modestly different distances (but the same time), all that the statute must require is some per-minute rate, however constant, by which to multiply the elapsed time. But, in making this argument, the government cannot escape the reality that the distance of each individual call under the AT & T system, measured by miles spanned and nothing more, did play a crucial role in the calculation of that call’s charge.

Without knowing the call’s distance in miles — hence the reason they were called “mileage bands” — no one could ascertain the appropriate charge for the call. That AT & T bundled concentric mileage ranges into a series of rate quanta does not change this fact. In juxtaposition to the 1965 billing method used by AT & T, MCI’s model for billing OfficeMax during the tax years at issue does not take into account the distance of the call.

Nor may the government prevail by comparing AT & T’s mileage bands with certain political zones — namely different States for intrastate calling and different countries for international calling — that yield different per-minute rates under the MCI plan for OfficeMax. The crucial variable under OfficeMax’s plan is the location of the call’s destination (for international calls) or the location of the call’s origin and destination (for intrastate calls) within the legal/political boundaries that serve as the dividing line between federal and state regulation of telephone service.

That is, the differences in rates do not serve as a proxy for call distance, but instead accommodate the FCC’s regulatory jurisdiction over interstate and international calls and the states’ regulatory jurisdiction over intrastate calls. See La. Pub. Serv. Comm’n v. FCC, 476 U.S. 355, 360, 106 S.Ct. 1890, 90 L.Ed.2d 369 (1986); AT & T Corp. v. Iowa Utils. Bd., 525 U.S. 366, 381 n. 8, 119 S.Ct. 721, 142 L.Ed.2d 835 (1999). Moreover, any possible correlation between political boundaries and a call’s total distance is imperfect at best and does not convert any of the political zones into a measure of call distance.

In many instances, for example, intrastate calls (which on average would seem to span a shorter distance than the longest of interstate calls), command higher per-minute rates than interstate calls. Judge Rogers’ dissent prompts a few responses. We agree with him that “and” may have a cumulative meaning or what comes to the same thing — a non-exclusive disjunctive meaning. And we agree that his examples — “I like bourbon and water” and “I like beer and wine” — begin to illustrate the difference.

While context and common sense suggest that the former means both together, they suggest that the latter means either or both independently. But from the vantage point of the regulator or the regulated in 1965, the better analogy for “time and distance” was “bourbon and water,” not “beer and wine.” In 1965, “time and distance” were used in combination to measure telephone fees by the only telephone company on the market.

Just as “bourbon and water” in context unambiguously requires both items, so does “time and distance.” But whether one prefers “bourbon and water” or “beer and wine,” each analogy is missing an ingredient featured here. Bourbon, water, beer and wine are all stand-alone items. Not so of “time and distance.” Distance, as we have shown, was not then, is not now and never has been used as a stand-alone measure for billing a phone call.

Because one half of the pair in this case cannot operate independently, the better analogy is to “bread and butter,” “pancakes and syrup,” “chips and dip,” or less abstemiously, to “Corona and lime,” or least appetizingly, to “tequila and worm.” As in these examples, context and common sense indicate that the inability of one item in a conjoined pair to function independently means that they must be combined, that they must come two by two, not one by one. The same is true of “time and distance”: While time can function alone, distance cannot.

The same problem undermines the dissent’s Yenn-diagram argument, which mistakenly suggests that distance can exist as a standalone item. Lastly, for the dissent, it suffices to say three things in resolving this case: (1) “and” may have an in-combination meaning or a cumulative meaning; (2) the statute thus is ambiguous; and (3) the government thus should prevail under Skidmore deference. Leaving to the side the question of how many statutory “ands” in the United States Code would be eligible for ambiguity status under this

interpretation, let us start by reiterating that context, common sense, history and the conjunctive time-and-distance formulation deployed in the tax for teletypewriter service in 1965 (“whether such charge or charges are determined as a flat periodic amount, on the basis of distance and elapsed transmission time, or in some other manner,” 26 U.S.C. § 4252(c)) all show that this “and” unambiguously has a conjunctive, in-combination meaning. But even were that not the case, it would not alter our conclusion. Skidmore breathes deference into an agency’s

interpretation based “upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements” — based upon, in short, its “power to persuade.” 323 U.S. at 140, 65 S.Ct. 161. But the revenue ruling at issue here takes an approach that the dissent castigates. The ruling acknowledges that “and” “literally” requires both time and distance: “The toll charges described in

section 4252(b)(1), that vary in amount with both distance and elapsed transmission time of the individual communication, reflect Congress’ understanding of how the charges for long distance calls were computed at the time the

section was enacted.” Rev. Rule 79-404, 1979-2 C.B. 382. In abandoning this natural

interpretation, the revenue ruling says that “a new type of service” cannot avoid taxation “merely because charges for it are determined in a manner which is not within the literal language of the statute.” Id. The dissent does not defend that reasoning, and it does not defend the government’s litigating position in this case. Cf. Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 213, 109 S.Ct. 468, 102 L.Ed.2d 493 (1988) (“Deference to what appears to be nothing more than an agency’s bonvenient litigating position would be entirely inappropriate.”).

Skidmore deference does not apply to a line of reasoning that an agency could have, but has not yet, adopted. That leaves only the possibility of deferring to the agency’s

interpretation, not because it makes sense on its own terms, but because of its longevity. While the 1979 revenue ruling concerned the same definition of “toll telephone service” at issue in this case, it arose long before accessible cellular service (and other developments) led telephone companies to begin charging on the basis of time alone and it arose in the context of ship-to-shore satellites, a service that (contrary to the dissent’s suggestion) would not necessarily grab the attention of members of Congress and their staff.

The real test for the agency came in the last decade when taxpayers began filing refund claims based on the decisions of long-distance carriers to begin charging on the basis of time alone. What the agency did at that point remains unclear. The taxpayer in this case argues that the IRS has settled many of these claims for well less than 100 cents on the dollar but claims it has been stymied in its attempts to obtain access to IRS records to establish whether that is true. At oral argument, we did not receive a clear answer one way or another.

Under these circumstances and with respect to this anomalously reasoned revenue ruling, we are not prepared to defer to a ruling that only became relevant to today’s debate in the last decade and that may or may not have been followed consistently by the agency. Accordingly, whether the “and” in § 4252(b)(1) is ambiguous or not, the government has not identified any tenable bases for deferring to its position — save for the facts that it is the government and we depend on its fiscal health, which are not enough. III.

The government argues that if it may not tax OfficeMax’s service under § 4252(b)(1), it may tax the company under the so-called WATS-line definition of “toll telephone service,” § 4252(b)(2), or under the definition of “local telephone service,” § 4252(a). We disagree.

The alternative definition of “toll telephone service,” recall, covers: a service which entitles the subscriber, upon payment of a periodic charge (determined as a flat amount or upon the basis of total elapsed transmission time), to the privilege of an unlimited number of telephonic communications to or from all or a substantial portion of the persons having telephone or radio telephone stations in a specified area which is outside the local telephone system area in which the station provided with this service is located. 26 U.S.C. § 4252(b)(2) (emphasis added).

The government does not contend that OfficeMax paid a flat amount for unlimited service. It instead contends that because the toll charge for each individual communication varied by elapsed transmission time alone, the service amounted to a periodic-charge service determined on the basis of total elapsed transmission time. Not so. OfficeMax’s monthly bill was a total of toll charges calculated for each individual call. And for each of these calls, MCI rounded up to the next billing increment and multiplied by a variable rate to determine the individual call’s charge.

The rate was subject to variation depending on other individual characteristics of the call as well, including “the time of day in which the call was made,” Office Max, Inc. v. United States, 309 F.Supp.2d 984, 987 (N.D.Ohio 2004). The end result is not a “periodic charge” based on total. elapsed time (or even total elapsed time itself rounded up to the next billing increment) but rather a monthly bill based on a summation of toll charges for individual communications. Due to the rounding, the total actual elapsed time of a bill could vary significantly without a change in the total amount due.

The government’s argument also eschews the dichotomy that the legislation sets in place between services billed by call and services billed by period regardless of the number of calls. OfficeMax’s phone services fall on the by-call side of this line. Finally, the government’s argument departs from the requirements that the service must be both “unlimited” and to “a specified area.” See 26 U.S.C. § 4252(b)(2). The statutory language mirrors the concept of AT & T’s traditional WATS service, which for a flat rate allowed customers to make as many calls as they wanted to one of AT & T’s six U.S. service areas.

OfficeMax’s phone service, as noted, levels charges for each call made, eschewing any reasonable definition of “unlimited” and allows calls to any non-local portion of the country, stretching the limits of “specified area.” All other courts reaching this issue have concluded similarly. Am. Bankers Ins. Group v. United States, 408 F.3d 1328 (11th Cir.2005) (reversing Am. Bankers Ins. Group, Inc. v. United States, 308 F.Supp.2d 1360 (S.D.Fla.2004)); Am. Online, Inc. v. United States, 64 Fed.Cl. 571 (Fed.Cl.2005); Honeywell Int’l, Inc. v. United States, 64 Fed.Cl. 188 (Fed.Cl.2005); Hewlett-Packard Co. v.

United States, No. 04-03832, 2005 U.S. Dist. LEXIS 19972 (N.D.Cal. Aug. 5, 2005); Reese Bros. v. United States, No. 03-745, 2004 U.S. Dist. LEXIS 27507 (W.D.Pa. Oct. 20, 2004); Fortis, Inc. v. United States, No. 03 Civ. 5137, 2004 WL 2085528, 2004 U.S. Dist. LEXIS 18686 (S.D.N.Y. Sept. 16, 2004). Indeed, even the government has elsewhere argued that § 4252(b)(2) applies exclusively to the traditional AT & T WATS service. See National Railroad Passenger Corp., AMTRAK v.

United States, 338 F.Supp.2d 22, 28 (D.D.C.2004) (“The IRS argues that ‘Congress drafted § 4252(b)(2) purely to cover WATS service.’ ” (emphasis added)).

Next, the provision defines “local telephone service” as “(1) the access to a local telephone system, and the privilege of telephonic quality communication with substantially all persons having telephone or radio telephone stations constituting a part of such local telephone system, and (2) any facility or service provided in connection with a service described in paragraph (1).” 26 U.S.C. § 4252(a). “The term ‘local telephone service,’ ” the statute adds, “does not include any service which is a ‘toll telephone service.’ ” Id.

The government argues that OfficeMax’s plan must be a service provided in connection with a local telephone service because all long-distance services eventually require access to local telephone systems. The only obstacle to a long distance service qualifying as a local telephone service, it adds, is the language excluding toll telephone service from the definition of local telephone service. Once it is determined that the OfficeMax plan does not qualify as a toll telephone service, the government concludes, it is eligible to be taxed as a local telephone service.

Yet the definition of local-telephone service requires “access to a” local telephone system, not “access to every” local telephone system included within the boundaries of a long-distance plan. Not surprisingly, given its application to “local” telephone service, the definition contemplates a service with limited geographic reach, not a plan that makes use of an untold number of local services. Applying the local-telephone definition to the long-distance telephone definition, moreover, not only blurs the line between two statutorily defined services but also is puzzling.

It makes one wonder why Congress would not eliminate the definition of toll telephone service altogether and apply the tax to all telephone services of any kind. IV. For these reasons, we affirm.