DEPARTMENT OF REVENUE OF KENTUCKY, et al.,
PETITIONERS v. GEORGE W. DAVIS et ux.
on writ of certiorari to the court of appeals ofkentucky
Justice Souter delivered the opinion of the Court, except as to Part III–B.**
For the better part of two centuries States and their political subdivisions have issued bonds for public purposes, and for nearly half that time some States have exempted interest on their own bonds from their state income taxes, which are imposed on bond interest from other States. The question here is whether Kentucky’s version of this differential tax scheme offends the Commerce Clause. We hold that it does not.
Like most other States, the Commonwealth of Kentucky taxes its residents’ income. See Ky. Rev. Stat. Ann. §141.020(1) (West 2006). The tax is assessed on “net income,” see ibid., calculated by reference to “gross income” as defined by the Internal Revenue Code, see §§141.010(9)–(11) (West Supp. 2007),1 which excludes “interest on any State or local bond” (“municipal bond,” for short2), 26 U. S. C. §103(a). Kentucky piggybacks on this exclusion, but only up to a point: it adds “interest income derived from obligations of sister states and political subdivisions thereof” back into the taxable net. Ky. Rev. Stat. Ann. §141.010(10)(c). Interest on bonds issued by Kentucky and its political subdivisions is thus entirely exempt,3 whereas interest on municipal bonds of other States and their subdivisions is taxable. (Interest on bonds issued by private entities is taxed by Kentucky regardless of the private issuer’s home.)
The ostensible reason for this regime is the attractiveness of tax-exempt bonds at “lower rates of interest … than that paid on taxable … bonds of comparable risk.” M. Graetz & D. Schenk, Federal Income Taxation 215 (5th ed. 2005) (hereinafter Graetz & Schenk). Under the Internal Revenue Code, for example, see 26 U. S. C. §103, “if the market rate of interest is 10 percent on a comparable corporate bond, a municipality could pay only 6.5 percent on its debt and a purchaser in a 35 percent marginal tax bracket would be indifferent between the municipal and the corporate bond, since the after-tax interest rate on the corporate bond is 6.5 percent,” Graetz & Schenk 215.4 The differential tax scheme in Kentucky works the same way; the Commonwealth’s tax benefit to residents who buy its bonds makes lower interest rates acceptable,5 while limiting the exception to Kentucky bonds raises in-state demand for them without also subsidizing other issuers.
The significance of the scheme is immense. Between 1996 and 2002, Kentucky and its subdivisions issued $7.7 billion in long-term bonds to pay for spending on transportation, public safety, education, utilities, and environmental protection, among other things. IRS, Statistics of Income Bulletin, C. Belmonte, Tax-Exempt Bonds, 1996–2002, pp. 169–170, http://www.irs.gov/pub/irs-soi/02govbnd.pdf (as visited Jan. 23, 2008, and available in Clerk of Court’s case file). Across the Nation during the same period, States issued over $750 billion in long-term bonds, with nearly a third of the money going to education, followed by transportation (13%) and utilities (11%). See ibid. Municipal bonds currently finance roughly two-thirds of capital expenditures by state and local governments. L. Thomas, Money, Banking and Financial Markets 55 (2006).
Funding the work of government this way follows a tradition going back as far as the 17th century. See C. Johnson & M. Rubin, The Municipal Bond Market: Structure and Changes, in Handbook of Public Finance 483, 485 (F. Thompson & M. Green eds. 1998) (“[In] 1690 … Massachusetts issued bills of credit to pay soldiers who had participated in an unsuccessful raid on the City of Quebec”). Municipal bonds first appeared in the United States in the early 19th century: “New York City began to float [debt] securities in about 1812,” A. Hillhouse, Municipal Bonds: A Century of Experience 31 (1936) (hereinafter Hillhouse), and by 1822 Boston “had a bonded debt of $100,000,” id., at 32. The municipal bond market had swelled by the mid-1840s, when the aggregate debt of American cities exceeded $27 million, and the total debt of the States was nearly 10 times that amount. See ibid. Bonds funded some of the great public works of the day, including New York City’s first water system, see id., at 31, and the Erie Canal, see R. Amdursky & C. Gillette, Municipal Debt Finance Law §1.2.1, p. 15 (1992) (hereinafter Amdursky & Gillette). At the turn of the 20th century, the total state and municipal debt was closing in on $2 billion, see Hillhouse 35, and by the turn of the millennium, over “$1.5 trillion in municipal bonds were outstanding,” J. Temel, The Fundamentals of Municipal Bonds, p. ix (5th ed. 2001).
Differential tax schemes like Kentucky’s have a long pedigree, too. State income taxation became widespread in the early 20th century, see A. Comstock, State Taxation of Personal Incomes 11 (1921) (reprinted 2005) (hereinafter Comstock), and along with the new tax regimes came exemptions and deductions, see id., at 171–184, to induce all sorts of economic behavior, including lending to state and local governments at favorable rates of untaxed interest. New York enacted the first of these statutes in 1919, see 1919 N. Y. Laws pp. 1641–1642, the same year it imposed an income tax, see Comstock 104,6 and other States followed, see, e.g., 1921 N. C. Sess. Laws p. 208; 1923 N. H. Laws p. 78; 1926 Va. Acts ch. 576, pp. 960–961, with Kentucky joining the pack in 1936, see 1936 Ky. Acts p. 71. Today, 41 States have laws like the one before us.7
Petitioners (for brevity, Kentucky or the Commonwealth) collect the Kentucky income tax. Respondents George and Catherine Davis are Kentucky residents who paid state income tax on interest from out-of-state municipal bonds, and then sued the tax collectors in state court on a refund claim that Kentucky’s differential taxation of municipal bond income impermissibly discriminates against interstate commerce in violation of the Commerce Clause of the National Constitution. The trial court granted judgment to the Commonwealth, relying in part on our cases recognizing the “market-participant” exception to the dormant Commerce Clause limit on state regulation. See App. to Pet. for Cert. A18–A19 (citing Reeves, Inc. v. Stake, 447 U. S. 429 (1980) , and Hughes v. Alexandria Scrap Corp., 426 U. S. 794 (1976) ).
The Court of Appeals of Kentucky reversed. See 197 S. W. 3d 557 (2006). In a brief discussion, it rejected the reasoning of an Ohio case upholding a similar tax scheme challenged under the Commerce Clause, see id., at 563 (discussing Shaper v. Tracy, 97 Ohio App. 3d 750, 647 N. E. 2d 550 (1994)), and distinguished our market participant cases, see 197 S. W. 3d, at 564, as well as a decision from the 19th century the Commonwealth relied on, see id., at 563–564 (discussing Bonaparte v. Tax Court, 104 U. S. 592 (1882) ). The Court of Appeals thought it had “no choice but to find that Kentucky’s system of taxing only extraterritorial bonds runs afoul of the Commerce Clause,” 197 S. W. 3d, at 564, and the Supreme Court of Kentucky denied the Commonwealth’s motion for discretionary review, see App. to Pet. for Cert. A14.
We granted certiorari owing to the conflict this raised on an important question of constitutional law, and because the result reached casts constitutional doubt on a tax regime adopted by a majority of the States. 550 U. S. ___ (2007). We now reverse.
The Commerce Clause empowers Congress “[t]o regulate Commerce … among the several States,” Art. I, §8, cl. 3, and although its terms do not expressly restrain “the several States” in any way, we have sensed a negative implication in the provision since the early days, see, e.g., Cooley v. Board of Wardens of Port of Philadelphia ex rel. Soc. for Relief of Distressed Pilots, 12 How. 299, 318–319 (1852); cf. Gibbons v. Ogden, 9 Wheat. 1, 209 (1824) (Marshall, C. J.) (dictum). The modern law of what has come to be called the dormant Commerce Clause is driven by concern about “economic protectionism—that is, regulatory measures designed to benefit in-state economic interests by burdening out-of-state competitors.” New Energy Co. of Ind. v. Limbach, 486 U. S. 269, 273–274 (1988) . The point is to “effectuat[e] the Framers’ purpose to ‘prevent a State from retreating into [the] economic isolation,’ ” Fulton Corp. v. Faulkner, 516 U. S. 325, 330 (1996) (quoting Oklahoma Tax Comm’n v. Jefferson Lines, Inc., 514 U. S. 175, 180 (1995) ; brackets omitted), “that had plagued relations among the Colonies and later among the States under the Articles of Confederation,” Hughes v. Oklahoma, 441 U. S. 322, 325–326 (1979) .
The law has had to respect a cross purpose as well, for the Framers’ distrust of economic Balkanization was limited by their federalism favoring a degree of local autonomy. Compare The Federalist Nos. 7 (A. Hamilton), 11 (A. Hamilton), and 42 (J. Madison), with The Federalist No. 51 (J. Madison); see also Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528, 546 (1985) (“The essence of our federal system is that within the realm of authority left open to them under the Constitution, the States must be equally free to engage in any activity that their citizens choose for the common weal”).
Under the resulting protocol for dormant Commerce Clause analysis, we ask whether a challenged law discriminates against interstate commerce. See Oregon Waste Systems, Inc. v. Department of Environmental Quality of Ore., 511 U. S. 93, 99 (1994) . A discriminatory law is “virtually per se invalid,” ibid.; see also Philadelphia v. New Jersey, 437 U. S. 617, 624 (1978) , and will survive only if it “advances a legitimate local purpose that cannot be adequately served by reasonable nondiscriminatory alternatives,” Oregon Waste Systems, supra, at 101 (internal quotation marks omitted); see also Maine v. Taylor, 477 U. S. 131, 138 (1986) . Absent discrimination for the forbidden purpose, however, the law “will be upheld unless the burden imposed on [interstate] commerce is clearly excessive in relation to the putative local benefits.” Pike v. Bruce Church, Inc., 397 U. S. 137, 142 (1970) . State laws frequently survive this Pike scrutiny, see, e.g., United Haulers Assn., Inc. v. Oneida-Herkimer Solid Waste Management Authority, 550 U. S. ___, ___–___ (2007) (slip op., at 14–15) (plurality opinion); Northwest Central Pipeline Corp. v. State Corporation Comm’n of Kan., 489 U. S. 493, 525–526 (1989) ; Minnesota v. Clover Leaf Creamery Co., 449 U. S. 456, 472–474 (1981) , though not always, as in Pike itself, 397 U. S., at 146.
Some cases run a different course, however, and an exception covers States that go beyond regulation and themselves “participat[e] in the market” so as to “exercis[e] the right to favor [their] own citizens over others.” Alexandria Scrap, 426 U. S., at 810. This “market-participant” exception reflects a “basic distinction . . . between States as market participants and States as market regulators,” Reeves, 447 U. S., at 436, “[t]here [being] no indication of a constitutional plan to limit the ability of the States themselves to operate freely in the free market,” id., at 437. See also White v. Massachusetts Council of Constr. Employers, Inc., 460 U. S. 204, 208 (1983) (“[W]hen a state or local government enters the market as a participant it is not subject to the restraints of the Commerce Clause”). Thus, in Alexandria Scrap, we found that a state law authorizing state payments to processors of automobile hulks validly burdened out-of-state processors with more onerous documentation requirements than their in-state counterparts. Likewise, Reeves accepted South Dakota’s policy of giving in-state customers first dibs on cement produced by a state-owned plant, and White held that a Boston executive order requiring half the workers on city-financed construction projects to be city residents passed muster.
Our most recent look at the reach of the dormant Commerce Clause came just last Term, in a case decided independently of the market participation precedents. United Haulers, supra, upheld a “flow control” ordinance requiring trash haulers to deliver solid waste to a processing plant owned and operated by a public authority in New York State. We found “[c]ompelling reasons” for “treating [the ordinance] differently from laws favoring particular private businesses over their competitors.” Id., at ___ (slip op., at 10). State and local governments that provide public goods and services on their own, unlike private businesses, are “vested with the responsibility of protecting the health, safety, and welfare of [their] citizens,” ibid., and laws favoring such States and their subdivisions may “be directed toward any number of legitimate goals unrelated to protectionism,” id., at ___ (slip op., at 11). That was true in United Haulers, where the ordinance addressed waste disposal, “both typically and traditionally a local government function.” Id., at ___ (slip op., at 12) (quoting United Haulers Assn., Inc. v. Oneida-Herkimer Solid Waste Management Authority, 261 F. 3d 245, 264 (CA2 2001) (Calabresi, J., concurring); internal quotation marks omitted). And if more had been needed to show that New York’s object was consequently different from forbidden protectionism, we pointed out that “the most palpable harm imposed by the ordinances—more expensive trash removal—[was] likely to fall upon the very people who voted for the laws,” rather than out-of-state interests. United Haulers, 550 U. S., at ___ (slip op., at 13). Being concerned that a “contrary approach . . . would lead to unprecedented and unbounded interference by the courts with state and local government,” id., at ___ (slip op., at 11), we held that the ordinance did “not discriminate against interstate commerce for purposes of the dormant Commerce Clause,” id., at ___ (slip op., at 10).8
It follows a fortiori from United Haulers that Kentucky must prevail. In United Haulers, we explained that a government function is not susceptible to standard dormant Commerce Clause scrutiny owing to its likely motivation by legitimate objectives distinct from the simple economic protectionism the Clause abhors. See id., at ___ (slip op., at 11) (“Laws favoring local government … may be directed toward any number of legitimate goals unrelated to protectionism”); see also id., at ___ (slip op., at 12) (noting that “[w]e should be particularly hesitant to interfere … under the guise of the Commerce Clause” where a local government engages in a traditional government function).9 This logic applies with even greater force to laws favoring a State’s municipal bonds, given that the issuance of debt securities to pay for public projects is a quintessentially public function, with the venerable history we have already sketched, see supra, at 4–5. By issuing bonds, state and local governments “sprea[d] the costs of public projects over time,” Amdursky & Gillette §1.1.3, at 11, much as one might buy a house with a loan subject to monthly payments. Bonds place the cost of a project on the citizens who benefit from it over the years, see ibid., and they allow for public work beyond what current revenues could support, see id., §1.2, at 12–13. Bond proceeds are thus the way to shoulder the cardinal civic responsibilities listed in United Haulers: protecting the health,10 safety,11 and welfare12 of citizens. It should go without saying that the apprehension in United Haulers about “unprecedented . . . interference” with a traditional government function is just as warranted here, where the Davises would have us invalidate a century-old taxing practice, see supra, at 5, presently employed by 41 States, see n. 7, supra, and affirmatively supported by all of them, see Brief for 49 States as Amici Curiae.
In fact, this emphasis on the public character of the enterprise supported by the tax preference is just a step in addressing a fundamental element of dormant Commerce Clause jurisprudence, the principle that “any notion of discrimination assumes a comparison of substantially similar entities.” United Haulers, 550 U. S.,at ___ (slip op., at 10) (internal quotation marks omitted) (quoting General Motors Corp. v. Tracy, 519 U. S. 278, 298 (1997) ). In Bonaparte v. Tax Court, 104 U. S. 592 (1882) , a case involving the Full Faith and Credit Clause, we held that a foreign State is properly treated as a private entity with respect to state-issued bonds that have traveled outside its borders. See id., at 595 (beyond its borders, a debtor State “is compelled to go into the market as a borrower, subject to the same disabilities in this particular as individuals,” and has none “of the attributes of sovereignty as to the debt it owes”). Viewed through this lens, the Kentucky tax scheme parallels the ordinance upheld in United Haulers: it “benefit[s] a clearly public [issuer, that is, Kentucky], while treating all private [issuers] exactly the same.” 550 U. S., at ___ (slip op., at 10). There is no forbidden discrimination because Kentucky, as a public entity, does not have to treat itself as being “substantially similar” to the other bond issuers in the market.13
Thus, United Haulers provides a firm basis for reversal. Just like the ordinances upheld there, Kentucky’s tax exemption favors a traditional government function without any differential treatment favoring local entities over substantially similar out-of-state interests. This type of law does “not ‘discriminate against interstate commerce’ for purposes of the dormant Commerce Clause.” Id., at ___ (slip op., at 13).
This case, like United Haulers, may also be seen under the broader rubric of the market participation doctrine, although the Davises say that market participant cases are inapposite here. In their view, we may not characterize state action under the Kentucky statutes as market activity for public purposes, because this would ignore a fact absent in United Haulers but central here: this is a case about differential taxation, and a difference that amounts to a heavier tax burden on interstate activity is forbidden, see, e.g., Camps Newfound/Owatonna, Inc. v. Town of Harrison, 520 U. S. 564 (1997) (invalidating statute exempting charities from real estate and personal property taxes unless conducted or operated principally for the benefit of out-of-state residents); Fulton Corp., 516 U. S. 325 (striking down tax on corporate stock held by state residents, where rate of tax was inversely proportional to the corporation’s exposure to the State’s income tax); Bacchus Imports, Ltd. v. Dias, 468 U. S. 263 (1984) (holding excise tax on sale of liquor at wholesale unconstitutional because it exempted some locally produced alcoholic beverages).
The Davises make a fair point to the extent that they argue that Kentucky acts in two roles at once, issuing bonds and setting taxes, and if looked at as a taxing authority it seems to invite dormant Commerce Clause scrutiny of its regulatory activity, see Walling v. Michigan, 116 U. S. 446, 455 (1886) (“A discriminating tax imposed by a State operating to the disadvantage of the products of other States when introduced into the first mentioned State, is, in effect, a regulation in restraint of commerce among the States, and as such is a usurpation of the power conferred by the Constitution upon the Congress”); see also Camps Newfound, supra, at 578 (“[I]t is clear that discriminatory burdens on interstate commerce imposed by regulation or taxation may . . . violate the Commerce Clause”); Tracy, supra, at 287 (“The negative or dormant implication of the Commerce Clause prohibits state taxation . . . that discriminates against or unduly burdens interstate commerce”).
But there is no ignoring the fact that imposing the differential tax scheme makes sense only because Kentucky is also a bond issuer. The Commonwealth has entered the market for debt securities, just as Maryland entered the market for automobile hulks, see Alexandria Scrap, 426 U. S., at 806, and South Dakota entered the cement market, see Reeves, 447 U. S., at 440. It simply blinks this reality to disaggregate the Commonwealth’s two roles and pretend that in exempting the income from its securities, Kentucky is independently regulating or regulating in the garden variety way that has made a State vulnerable to the dormant Commerce Clause. States that regulated the price of milk, see, e.g., West Lynn Creamery, Inc. v. Healy, 512 U. S. 186 (1994) ; Baldwin v. G. A. F. Seelig, Inc., 294 U. S. 511 (1935) , did not keep herds of cows or compete against dairy producers for the dollars of milk drinkers. But when Kentucky exempts its bond interest, it is competing in the market for limited investment dollars, alongside private bond issuers and its sister States, and its tax structure is one of the tools of competition.14
The failure to appreciate that regulation by taxation here goes hand in hand with market participation by selling bonds allows the Davises to advocate the error of focusing exclusively on the Commonwealth as regulator and ignoring the Commonwealth as bond seller, see Brief for Respondents 36–39, just as the state court did in saying that “ ‘when a state chooses to tax its citizens, it is acting as a market regulator[,]’ not as a market participant.” 197 S. W. 3d, at 564 (quoting Shaper, 97 Ohio App. 3d, at 764, 647 N. E. 2d, at 552).15 To indulge in this single vision, however, would require overruling most, if not all, of the cases on point decided since Alexandria Scrap.
White, for example, also scrutinized a government acting in dual roles. The mayor of Boston promulgated an executive order that bore the hallmarks of regulation: it applied to every construction project funded wholly or partially by city funds (or funds administered by the city), and it imposed general restrictions on the hiring practices of private contractors, mandating that 50% of their work forces be bona fide Boston residents and setting thresholds for minorities (25%) and women (10%) as well. See 460 U. S., at 205, n. 1; see also id., at 218–219 (Blackmun, J., concurring in part and dissenting in part) (“The executive order in this case … is a direct attempt to govern private economic relationships… . [It] is the essence of regulation”). At the same time, the city took part in the market by “expend[ing] its own funds in entering into construction contracts for public projects.” Id., at 214–215 (opinion of the Court). After speaking of “ ‘[t]he basic distinction … between States as market participants and States as market regulators,’ ” id., at 207 (quoting Reeves, supra, at 436–437), White did not dissect Boston’s conduct and ignore the former. Instead, the Court treated the regulatory activity in favor of local and minority labor as terms or conditions of the government’s efforts in its market role, which was treated as dispositive.
Similarly, in Alexandria Scrap, Maryland employed the tools of regulation to invigorate its participation in the market for automobile hulks. The specific controversy there was over documentation requirements included in a “comprehensive statute designed to speed up the scrap cycle.” 426 U. S., at 796. Superficially, the scheme was regulatory in nature; but the Court’s decision was premised on its view that, in practical terms, Maryland had not only regulated but had also “entered into the market itself to bid up [the] price” of automobile hulks. See id., at 806.
United Haulers, though not placed under the market participant umbrella, may be seen as another example. Not only did the public authority acting in that case process trash, but its governmental superiors forbade trash haulers to deal with any other processors. This latter fact did not determine the outcome, however; the dispositive fact was the government’s own activity in processing trash. We upheld the government’s decision to shut down the old market for trash processing only because it created a new one all by itself, and thereby became a participant in a market with just one supplier of a necessary service. If instead the government had created a monopoly in favor of a private hauler, we would have struck down the law just as we did in C & A Carbone, Inc. v. Clarkstown, 511 U. S. 383 (1994) . United Haulers accordingly turned on our decision to give paramount consideration to the public function in actively dealing in the trash market; if the Davises had their way, United Haulers would be overruled and the market participation doctrine would describe a null set (or maybe a set of one, see Reeves, supra).
In each of these cases the commercial activities by the governments and their regulatory efforts complemented each other in some way, and in each of them the fact of tying the regulation to the public object of the foray into the market was understood to give the regulation a civic objective different from the discrimination traditionally held to be unlawful: in the paradigm of unconstitutional discrimination the law chills interstate activity by creating a commercial advantage for goods or services marketed by local private actors, not by governments and those they employ to fulfill their civic objectives, see, e.g., Fulton Corp., 516 U. S. 325 (higher tax on the stock of corporations with little or no presence in the State); New Energy Co. of Ind., 486 U. S. 269 (tax credit to sellers of ethanol available only for ethanol produced in the State); Bacchus Imports, Ltd., 468 U. S. 263 (tax exemption that applied only to sales of certain locally produced liquors); Lewis v. BT Investment Managers, Inc., 447 U. S. 27 (1980) (prohibition on out-of-state banks owning in-state businesses that providedinvestment advisory services); Boston Stock Exchange v. State Tax Comm’n, 429 U. S. 318 (1977) (higher tax on sale of securities by nonresidents if the securities were sold on an out-of-state, not an in-state, exchange). In sum, our cases on market regulation without market participation prescribe standard dormant Commerce Clause analysis; our cases on market participation joined with regulation (the usual situation) prescribe exceptional treatment for this direct governmental activity in commercial markets for the public’s benefit.16
The Kentucky tax scheme falls outside the forbidden paradigm because the Commonwealth’s direct participation favors, not local private entrepreneurs, but the Commonwealth and local governments. The Commonwealth enacted its tax code with an eye toward making some or all of its bonds more marketable. When it issues them for sale in the bond market, it relies on that tax code, and seller and purchaser treat the bonds and the tax rate as joined just as intimately, say, as the work force require-ments and city construction contracts were in Boston. Issuing bonds must therefore have the same significance under the dormant Commerce Clause as government trash processing, junk car disposal, or construction; and United Haulers, Alexandria Scrap, and White can be followed only by rejecting the Davises’ argument that Kentucky’s regulatory activity should be viewed in isolation as Commerce Clause discrimination.17
A look at the specific markets in which the exemption’s effects are felt both confirms the conclusion that no traditionally forbidden discrimination is underway and points to the distinctive character of the tax policy. The market as most broadly conceived is one of issuers and holders of all fixed-income securities, whatever their source or ultimate destination. In this interstate market, Kentucky treats income from municipal bonds of other States just like income from bonds privately issued in Kentucky or elsewhere; no preference is given to any local issuer, and none to any local holder, beyond what is entailed in the preference Kentucky grants itself when it engages in activities serving public objectives.
A more specialized market can be understood as commerce solely in federally tax-exemptmunicipal bonds, much of it conducted through interstate municipal bond funds.18 Here, of course, the distinction between the taxing State’s bonds and their holders and issuers and holders of out-of-state counterparts is at its most stark. But what is remarkable about the issuers in this and the broader interstate market is that nearly every taxing State believes its public interests are served by the same tax-and-exemption feature, which is supported in this Court by every one of the States (with or without an income tax) despite the ranges of relative wealth and tax rates among them. See Brief for 49 States as Amici Curiae. These facts suggest that no State perceives any local advantage or disadvantage beyond the permissible ones open to a government and to those who deal with it when that government itself enters the market. See supra, at 14–18.
An equally significant perception emerges from examining the third type of market for municipal bonds: the one for bonds within the State of issue, a large proportion of which market in each State is managed by one or more single-state funds. By definition, there is no discrimination against interstate activity within the market itself, but one of its features reveals an important benefit of intrastate bond markets as they operate through these funds. The intrastate funds absorb securities issued by smaller or lesser known municipalities that the interstate markets tend to ignore. See National Federation Brief 15 (compared with single-state funds, “[n]ational mutual funds … are less likely to dedicate the time necessary to evaluate a small, obscure or infrequent municipal bond issuer or to purchase bonds issued by such public entities”); id., at 19 (“[N]ational mutual funds place a higher premium on the liquidity of their holdings than do single state funds, which are willing to purchase less liquid municipal bonds of smaller and less familiar issuers because of the state tax advantage and the fund’s mandate to purchase bonds issued within a specific state”).
There is little doubt that many single-state funds would disappear if the current differential tax schemes were upset. See id., at 18 (“[O]ne predictable impact of the elimination of tax incentives for the purchase of municipal bonds issued in a specific state would be the disappearance, through consolidation into national mutual funds, of single state mutual funds”); ibid. (“Although a handful of single state funds might continue to exist for a small number of states (such as Florida) with high populations that have a high affinity for local bond issuers, the current state tax system is the raison d’ętre for virtually all single state funds, and they would cease to be financially viable in the absence of a tax advantage that outweighed their relative lack of diversification vis-À-vis national funds and their reduced asset base”); accord, Brief for Respondents 29 (the States’ tax exemptions “have fostered the growth of funds that hold only the municipal bonds of a single state,” which “[a]s compared [with] national tax-exempt bonds funds … tend to be higher risk and higher cost”); 11 Kiplinger’s Retirement Report, Win With Home-State Muni Bond Funds, p. 2 (Dec. 2004) (noting that in States without a differential taxation scheme, “there’s little incentive to create [single-state] muni bond funds”).
Nor is there any suggestion that the interstate markets would discover some new reason to welcome the weaker municipal issues that would lose their local market homes after a victory for the Davises here. See National Federation Brief 18, 19 (“The main adverse impact of the disappearance of single state funds … would be felt by small municipal issuers” because they “would stand to lose much of the intrastate market for the bonds that has developed under the currently prevailing state tax system without gaining much of an interstate market from its elimination”). Financing for long-term municipal improvements would thus change radically if the differential tax feature disappeared.19
This probable indispensability of the current scheme to maintaining single-state markets serving smaller municipal borrowers not only underscores how far the States’ objectives probably lie from the forbidden protectionism for local business; it also tends to explain why the States are so committed to a taxing practice that much scholarship says often produces a net burden of tax revenues lost over interest expense saved. See, e.g., Brief for Alan D. Viard et al. as Amici Curiae 19 (“[S]tates routinely fail to recoup the cost of the tax subsidy in the form of lower financing rates” (citing Chalmers, Default Risk Cannot Explain the Muni Puzzle: Evidence from Municipal Bonds that are Secured by U. S. Treasury Obligations, 11 Rev. Financial Studies 281, 282–283 (1998))).
In sum, the differential tax scheme is critical to the operation of an identifiable segment of the municipal financial market as it currently functions, and this fact alone demonstrates that the unanimous desire of the States to preserve the tax feature is a far cry from the private protectionism that has driven the development of the dormant Commerce Clause. It is also fatal to the Davises’ backup argument that this case should be remanded for analysis under the rule in Pike, 397 U. S. 137 .
Concluding that a state law does not amount to forbidden discrimination against interstate commerce is not the death knell of all dormant Commerce Clause challenges, for we generally leave the courtroom door open to plaintiffs invoking the rule in Pike, that even nondiscriminatory burdens on commerce may be struck down on a showing that those burdens clearly outweigh the benefits of a state or local practice. See id., at 142. The Kentucky courts made no Pike enquiry, and the Davises ask us to remand for one now, see Brief for Respondents 43.
The Davises’ request for Pike balancing assumes an answer to an open question: whether Pike even applies to a case of this sort. United Haulers included a Pike analysis, see 550 U. S., at ___–___ (slip op., at 14–15) (plurality opinion), but our cases applying the market participant exception have not, see, e.g., White, 460 U. S. 204 ; Alexandria Scrap, 426 U. S. 794 . We need not decide this question today, however, for Kentucky has not argued that Pike is irrelevant, see Reply Brief for Petitioners 2, n. 1, and even on the assumption that a Pike examination might generally be in order in this type of case, the current record and scholarly material convince us that the Judicial Branch is not institutionally suited to draw reliable conclusions of the kind that would be necessary for the Davises to satisfy a Pike burden in this particular case.
The institutional difficulty is manifest in the very train of disadvantages that the Davises’ counsel attributes to the current differential tax scheme:
“First, it harms out-of-state issuers (i.e., other States and their subdivisions) by blocking their access to investment dollars in Kentucky. Second, it similarly harms out-of-state private sellers (e.g., underwriters, individuals, and investment funds) who wish to sell their bonds in Kentucky. Third, it harms the national municipal bond market and its participants by distorting and impeding the free flow of capital. Fourth, it harms Kentucky investors by promoting risky, high-cost investment vehicles. Fifth, it harms the States by compelling them to enact competing discriminatory laws that decrease their net revenues.” Brief for Respondents 9.
Even if each of these drawbacks does to some degree eventuate from the system, it must be apparent to anyone that weighing or quantifying them for a cost-benefit analysis would be a very subtle exercise. It is striking, after all, that most of the harms allegedly flowing directly or indirectly to Kentucky’s sister States and their citizens have failed to dissuade even a single State from supporting the current system; every one of them, including States with no income tax, have lined up with Kentucky in this case.
The prospect for reliable Pike comparison dims even further when we turn to the benign function of the current system flagged a moment ago. Is any court in a position to evaluate the advantage of the current market for bonds issued by the smaller municipalities, the ones with no ready access to any other bond market than single-state funds? Consider that any attempt to place a definite value on this feature of the existing system would have to confront the what-if questions. If termination of the differential tax scheme jeopardized or eliminated most single-state funds (as the cited authorities predict) would some new source of capital take their place? Would the interstate markets accommodate the small issuers (as no cited authorities predict), or would the financing in question be replaced by current local taxation for long-term projects (unlikely, considering that financially weaker borrowers are involved), or would state governments assume responsibility through their own bonds or by state taxation? Or would capital to some degree simply dry up, eliminating a class of municipal improvements?20 And if some new source or sources of capital became available for these improvements in a given State, how likely is it that the new scheme would produce measurable net benefits to other States seeking capital, and how perceptibly would it produce a freer flow of funds? Money spent up front on increased local or state taxation is no more available for out-of-state investment than money invested in local bonds; sinking funds would be obviated, but what would the effect be on interstate capital flows?
What is most significant about these cost-benefit questions is not even the difficulty of answering them or the inevitable uncertainty of the predictions that might be made in trying to come up with answers, but the unsuitability of the judicial process and judicial forums for making whatever predictions and reaching whatever answers are possible at all. See Tracy, 519 U. S., at 308 (“[T]he Court is institutionally unsuited to gather the facts upon which economic predictions can be made, and professionally untrained to make them”); cf. Fulton Corp., 516 U. S., at 342 (“ ‘[C]ourts as institutions are poorly equipped to evaluate with precision the relative burdens of various methods of taxation. The complexities of factual economic proof always present a certain potential for error, and courts have little familiarity with the process of evaluating the relative economic burden of taxes’ ” (quoting Minneapolis Star & Tribune Co. v. Minnesota Comm’r of Revenue, 460 U. S. 575, 589–590 (1983) )).
While it is not our business to suggest that the current system be reconsidered, if it is to be placed in question a congressional forum has two advantages. Congress has some hope of acquiring more complete information than adversary trials may produce, and an elected legislature is the preferable institution for incurring the economic risks of any alteration in the way things have traditionally been done.And risk is the essence of what the Davises are urging here. It would miss the mark to think that the Kentucky courts, and ultimately this Court, are being invited merely to tinker with details of a tax scheme; we are being asked to apply a federal rule to throw out the system of financing municipal improvements throughout most of the United States, and the rule in Pike was never intended to authorize a court to expose the States to the uncertainties of the economic experimentation the Davises request.
* * *
The dissent rightly praises the virtues of the free market, and it warns that our decision to uphold Kentucky’s tax scheme will result in untoward consequences for that market. See, e.g., post, at 15. But the warning is alarmism; going back to 1919 the state regimes of differential bond taxation have been elements of the national commerce without wilting the Commerce Clause. The threat would come, instead, from the dissent’s approach, which to a certainty would upset the market in bonds and the settled expectations of their issuers based on the experience of nearly a century.
We have been here before. Our predecessors on this Court responded to an earlier invitation to the adventurism of overturning a traditional local taxing practice. Justice Holmes answered that “the mode of taxation is of long standing, and upon questions of constitutional law the long settled habits of the community play a part … . [T]he fact that the system has been in force for a very long time is of itself a strong reason … for leaving any improvement that may be desired to the legislature.” Paddell v. City of New York, 211 U. S. 446, 448 (1908) .21
The judgment of the Court of Appeals of Kentucky is reversed, and the case is remanded for further proceedings not inconsistent with this opinion.
It is so ordered.
** Justice Ginsburg joins all but Part III–B of this opinion.
1 Specifically, Kentucky defines “net income” for noncorporate taxpayers as “adjusted gross income,” minus certain deductions. See Ky. Rev. Stat. Ann. §141.010(11). “Adjusted gross income,” in turn, is defined as “gross income” minus other deductions spelled out in the Internal Revenue Code and elsewhere in the Kentucky statutes. See §141.010(10). Finally, “gross income” has the same meaning set out in §61 of the Internal Revenue Code. See §141.010(9); see also 26 U. S. C. §61.
2 “Municipal bond” is commonly defined as a “debt obligation of a state or local government entity.” J. Downes & J. Goodman, Dictionary of Finance and Investment Terms 439 (7th ed. 2006). We use that definition here; our references to “municipal bonds” thus include bonds issued by States and their political subdivisions. An argument raised by one of the Davises’ amici focuses on so-called “private-activity,” “industrial-revenue,” or “conduit” bonds, a subset of municipal bonds used to finance projects by private entities. These bonds are often (but not always) exempt under the Kentucky scheme. Amici contend that Kentucky’s exemption of these bonds, at the very least, plainly violates the Commerce Clause. See Brief for Alan D. Viard et al. as Amici Curiae 25–26. This argument, however, was not considered below, was never pressed by the Davises themselves, and is barely developed by amici. Moreover, we cannot tell with certainty what the consequences would be of holding that Kentucky violates the Commerce Clause by exempting such bonds; we must assume that it could disrupt important projects that the States have deemed to have public purposes. Accordingly, it is best to set this argument aside and leave for another day any claim that differential treatment of interest on private-activity bonds should be evaluated differently from the treatment of municipal bond interest generally.
4 The amount of this benefit to municipal issuers can be approximated by comparing the interest rates on municipal bonds to those on Treasury bonds, which are also exempt from state taxation but are subject to federal taxation. “[A]t the end of 2006, the borrowing costs on AAA-rated, 10-year municipal bonds on average were 80.3 percent of comparable, but federally taxable, U. S. Treasury securities, [and] at the end of 2005 the borrowing costs on such municipal bonds were 88.4 percent of comparable U. S. Treasury bonds.” Brief for National Federation of Municipal Analysts as Amicus Curiae 8, n. 4.
5 The precise reduction in interest rates depends on the federal and state income tax rates, the credit rating of the issuer, the term of the bond, and market factors. See id., at 8. The reduction in interest rates is generally greater the higher are a State’s income tax rates. See id., at 9, and n. 6.
6 The Federal Government got in the game even earlier. Municipal bonds were exempted from “every federal income tax act enacted since passage of the Sixteenth Amendment ” in 1913. Amdursky & Gillette §7.2.1, at 440.
7 This figure includes Kentucky and 36 other States that have schemes that are nearly identical to Kentucky’s. See Ala. Code §§40–18–4, 40–18–14(3)(f) (2003); Ariz. Rev. Stat. Ann. §43–1021(3) (West Supp. 2007); Ark. Code Ann. §26–51–404(b)(5) (Supp. 2007); Cal. Rev. & Tax. Code Ann. §17133 (West 2004); Colo. Rev. Stat. Ann. §39–22–104(3)(b) (2007); Conn. Gen. Stat. §12–701(a)(20)(A)(i) (2007); Del. Code Ann., Tit. 30, §1106(a)(1) (1997); Ga. Code Ann. §48–7–27(b)(1)(A) (2005); Haw. Rev. Stat. §§39–11, 47–13 (1994), 235–7(a)(6), (b)(2) (2001); Idaho Code §§63–3022M(1), (3)(b) (Lexis 2007); Kan. Stat. Ann. §79–32,117(b)(i) (2006 Cum. Supp.); La. Stat. Ann. §§47:48, 47:293(9)(a), (b) (West 2001 and Supp. 2008); Me. Rev. Stat. Ann., Tit. 36, §5122(1)(A) (Supp. 2007); Md. Tax-Gen. Code Ann. §10–204(b) (Lexis Supp. 2007); Mass. Gen. Laws, ch. 62, §2(a)(1)(A) (West 2006); Mich. Comp. Laws Ann. §206.30(1)(a) (West Supp. 2007); Minn. Stat. §290.01, subd. 19a(1)(i) (2006); Miss. Code Ann. §27–7–15(4)(d) (Supp. 2007); Mo. Rev. Stat. §143.121(2)(b) (2007 Supp.); Mont. Code Ann. §15–30–111(2)(a)(i) (2007); Neb. Rev. Stat. §77–2716(1)(c) (2007 Supp.); N. H. Rev. Stat. Ann. §77:4(I) (Supp. 2007); N. J. Stat. Ann. §54A:6–14 (West 2002); N. M. Stat. Ann. §§7–2–2(B)(3), (V) (2005); N. Y. Tax Law Ann. §612(b)(1) (West 2006); N. C. Gen. Stat. Ann. §§105–134.6(b)(1)(b), (c)(1) (Lexis 2005); N. D. Cent. Code Ann. §57–38–01.2(1)(g) (Lexis Supp. 2007); Ohio Rev. Code Ann. §5747.01(A)(1) (Lexis Supp. 2007); Ore. Rev. Stat. §316.680(2)(a) (2003); Pa. Stat. Ann., Tit. 72, §9901 (Purdon 2000); R. I. Gen. Laws §44–30–12(b)(1) (Supp. 2007); S. C. Code Ann. §12–6–1120(1) (2000); Tenn. Code Ann. §67–2–104(e)(1) (2006); Vt. Stat. Ann., Tit. 32, §5811(18)(A)(i)(II) (2007); Va. Code Ann. §§58.1–322(B)(1), (C)(2) (Lexis Supp. 2007); W. Va. Code Ann. §§11–21–12(b)(1), (c)(2) (Lexis Supp. 2007). It also includes four States that tax out-of-state municipal bonds and exempt some, but not all, in-state municipal bonds. See Iowa Code §422.7(36) (2005); Okla. Stat., Tit. 68, §§2358.5, 2358.5A (West 2007 Supp.); Wis. Stat. §71.05(1)(c) (2003–2004); compare Ill. Comp. Stat., ch. 35, §5/203(a)(2)(A) (West 2006) with ch. 45, §35/80(e). Of the remaining States, Utah exempts its own bonds, and extends reciprocal treatment to the bonds of States that do not tax Utah bonds, see Utah Code Ann. §§59–10–114(1)(g), (6) (Lexis 2007 Supp.); Indiana exempts all municipal bonds, see Ind. Code §6–3–1–3.5 (West 2004); and the balance have no personal income tax.
8 In so holding, we distinguished our decision in C & A Carbone, Inc. v. Clarkstown, 511 U. S. 383 (1994) , which struck down a very similar ordinance on Commerce Clause grounds. The Carbone ordinance, however, benefited a private processing facility, and we found “this difference constitutionally significant” for the reasons adverted to in the main text. See United Haulers, 550 U. S., at ___ (slip op., at 1). Although the Carbone dissent argued that the private facility was “essentially a municipal facility,” 511 U. S., at 419 (opinion of Souter, J.), United Haulers relied on the apparent view of the Carbone majority that the facility was properly characterized as private, see 550 U. S., at ___–___ (slip op., at 7–8).
9 Justice Kennedy’s dissent (hereinafter dissent) says this is just circular rationalization, that the United Haulers acceptance of governmental preference in support of public health, safety, and welfare is the equivalent of justifying the law as an exercise of the “police power” and thus an exercise in “tautology,” since almost any state law could be so justified. See post, at 5. But this misunderstands what we said in United Haulers. The point of asking whether the challenged governmental preference operated to support a traditional public function was not to draw fine distinctions among governmental functions, but to find out whether the preference was for the benefit of a government fulfilling governmental obligations or for the benefit of private interests, favored because they were local. Under United Haulers, governmental public preference is constitutionally different from commercial private preference, and we make the governmental responsibility enquiry to identify the beneficiary as one or the other. See supra, at 9–10; United Haulers, supra, at ___ (slip op., at 11). Because this is the distinction at which the enquiry about traditional governmental activity is aimed, it entails neither tautology nor the hopeless effort to pick and choose among legitimate governmental activity that led to Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528 (1985) . One of the two fundamental points of difference between the Court and the dissenters is their rejection of the constitutional distinction between public and private preference, see post, at 6, 10, 12; the dissenters thus carry on the battle that was fought in United Haulers. (The second fundamental difference goes to the realism and legitimacy of treating bond issuance and tax provisions as aggregated features of a single scheme of public finance. Compare infra, at 15–16, with post, at 6, 14.)
10 See, e.g., The Bond Buyer, Apr. 20, 2007, p. 31, col. 2 (describing bond issue by the Grayson County Public Hospital District Corporation).
11 See, e.g., id., June 20, 2007, p. 29, col. 3 (describing bond issue by Todd County for a “Detention Facility Project”).
12 See, e.g., id., Apr. 20, 2007, p. 31, cols. 2–3 (describing bond issue by the Johnson County School District Finance Corporation).
13 Contrary to the dissent, see post, at 11, we do not suggest that the only market at issue here is a discrete market for Kentucky bonds. In fact, we recognize that the relevant market can be conceived more broadly. See infra, at 20–21. Our point goes not to the contours of the market, but to the proper characterization of the various entities acting in the market.
14 The dissent overlooks this discussion when it claims that we contend Kentucky does not compete with other municipal bond issuers. See post, at 8.
15 The dissent does the same. See post, at 6, 14.
16 Significantly, our market participant cases are not limited to cases where the government supplies a uniquely public product. This much is manifest from Reeves, Inc. v. Stake, 447 U. S. 429 (1980) . There is nothing remarkable or inherently governmental about the cement South Dakota produced, and yet we recognized that the State may engage in clear discrimination against out-of-state buyers that regular dormant Commerce Clause analysis would undoubtedly have held unconstitutional.
17 The dissent criticizes this analysis on the basis of our statement in Camps Newfound/Owatonna, Inc. v. Town of Harrison, 520 U. S. 564, 593 (1997) , that “[a] tax exemption is not the sort of direct state involvement in the market that falls within the market-participation doctrine.” See post, at 14. This both misses the point and leaves the language from Camps Newfound shorn of context. In Camps Newfound, the tax exemption was unaccompanied by any market activity by the State; it favored only private charitable institutions. We correctly rejected the argument that a tax exemption without more constitutes market participation. But we had no occasion to consider the scheme here, where a State employs a tax exemption to facilitate its own participation in the market. As noted before, one of the dissent’s critical premises is the disaggregation of bond issuance and tax treatment, see post, at 6, 14; that strikes us as a denial of economic reality. The dissent also suggests that our reasoning conflicts with South-Central Timber Development, Inc. v. Wunnicke, 467 U. S. 82 (1984) , see post, at 14–15, but there is no conflict. In South-Central, Alaska conditioned the sale of state timber to private purchasers by requiring that the timber be processed within the State prior to export, and a plurality struck down the condition under the Commerce Clause. The case turned on the plurality’s conclusion that the processing requirement constituted a “restrictio[n] on dispositions subsequent to the goods coming to rest in private hands.” 467 U. S., at 98; see id., at 95 (“Under the Alaska requirement, … the choice is made for [the purchaser]: if he buys timber from the State he is not free to take the timber out of state prior to processing”). Kentucky imposes no such restrictions on the disposition of Kentucky bonds; bond holders are free to sell the bonds to whomever they please. Thus, the type of “downstream regulation” that South-Central found objectionable is simply not present here. Id., at 99. We note also that South-Central expressly applied “more rigorous” Commerce Clause scrutiny because the case involved “foreign commerce” and restrictions on the resale of “a natural resource.” Id., at 100, 96. Neither of those elements appears here.
18 See Brief for National Federation of Municipal Analysts as Amicus Curiae 11 (hereinafter National Federation Brief) (“In 2006, tax-exempt mutual funds held approximately $365 billion in long-term [municipal] bonds, of which approximately $155 billion were held in 481 single-state funds and approximately $210 billion in 230 national funds … [and as of March 2007,] approximately $254 billion [in short-term municipal bonds] were held in national tax-exempt money market funds and approximately $125 billion in single state tax-exempt money market funds” (citing Investment Company Institute, 2007 Investment Company Fact Book 96, 98; Lipper Analytical Services, Tax-Exempt Fixed Income Fund Performance Analysis, 1st Quarter 2007 Report)); National Federation Brief 12 (“[A]pproximately 58% of … long-term municipal bonds [owned by mutual funds] and approximately 67% of … short-term municipal securities were purchased without regard to a match between the state of the bond issuer and the state of the fund’s shareholders”).
19 The Davises themselves, in their opposition to the petition, explain that if the tax exemptions are removed, “states will open their investment sales to the entire national market for debt instruments.” Brief in Opposition 10–11. As a result, the Davises say, “[o]nce states compete in the financial markets without the protective benefit of coercive tax schemes, they will have to be more selective in what projects they choose to fund… . [T]he market will provide incentives for governments to be more careful in selecting and funding projects through bond sales.” Id., at 11, n. 5.
20 History bears out the concern that poorer places may have a harder time taking on at least some types of local investments. See Goldin & Katz, The Shaping of Higher Education: The Formative Years in the United States, 1890 to 1940, 13 J. Econ. Perspectives 37, 50–55 (Winter 1999) (per capita spending on public universities depended on local wealth); Goldin, America’s Graduation from High School: The Evolution and Spread of Secondary Schooling in the Twentieth Century, 58 J. Econ. Hist. 345, 369–372 (1998) (likewise for public high schools).
21 The dissent thinks the need to preserve existing financing practices is the true “controlling rationale” of our holding, post, at 15, but not acknowledged as such. As Justice Holmes’s opinion shows, practical consequences have always been relevant in deciding the constitutionality of local tax laws. The practical considerations discussed here support the traditional distinction between permissible public preferences and the forbidden discriminations for the benefit of local private interests.