Concept
The Four Dole Limits
Spending power
The Four Dole Limits
Congress’s power under Article I, Section 8, Clause 1 to tax and spend lets it attach conditions to federal money, but Dole took care to say that the power is “not unlimited.”1 In sustaining the highway-funds condition, the Court collected from its earlier cases several general restrictions on what Congress may demand in return for a grant.2 Lawyers call the resulting list the Dole test: the spending must pursue the general welfare, and the condition must be unambiguous, related to a federal interest, and not barred by any other constitutional provision.3 The Court stated the four limits in this order:
- The exercise of the spending power must pursue the general welfare, a limit drawn from the language of the Spending Clause itself.
- A condition on a state’s receipt of federal funds must be stated unambiguously, so that the state can choose knowingly.
- A condition may be illegitimate if it is unrelated to the federal interest in the particular national program the funds support.
- No other constitutional provision may independently bar the conditional grant.
The four limits share one aim: a condition that a state could accept knowingly and freely, for a national purpose, without violating anyone’s constitutional rights. A spending condition works through acceptance, because the state takes the money on the terms Congress offers, so the limits police both the legitimacy of the offer and the quality of the acceptance. The first limit asks whether the purpose is a proper one. The second asks whether the state understood the terms. The third asks whether the condition stays tied to the program the money funds, so that a grant cannot become a lever for unrelated regulation. The fourth ensures that a grant cannot buy what another constitutional provision forbids.
The limits do not bite equally. Most conditions meet the first two readily, so the live arguments arise under relatedness and the independent bars.4 Dole itself fits that pattern. The Court found the means reasonably calculated to advance the general welfare and the conditions as clearly stated as Congress could make them, and South Dakota admitted that it never contended the condition was unrelated to a national concern in the absence of the Twenty-first Amendment.5 The Court described the third limit as something earlier cases had suggested “without significant elaboration,” so the opinion offers only a sketch of it.6 Each limit is taught below with the authorities that fill it in. Dole’s remarks about coercion, and the anti-coercion principle later built on them, come in the next part.
General Welfare
The first limit comes straight from the words of the Spending Clause, which lets Congress tax and spend “to pay the Debts and provide for the common Defence and general Welfare of the United States.”7 Dole turned that language into a requirement: the exercise of the spending power must be in pursuit of “the general welfare.”8 Because the clause names the ends that taxing and spending may serve, this is the limit most directly rooted in constitutional text.9
How much the requirement restricts anything depends on how seriously courts enforce it. Hamilton read the clause as conferring a power separate and distinct from the enumerated powers that follow it, a substantive power to tax and appropriate limited only by the requirement that it be exercised to provide for the general welfare, and a competing reading has also had weighty support.10 On the Hamiltonian reading, the general welfare requirement is the only limit the clause itself imposes. In practice it is a gentle one.
Dole instructed courts deciding whether an expenditure is meant to serve general public purposes to “defer substantially to the judgment of Congress.”11 A footnote goes further: the deference is so great that the Court has more recently questioned whether the general welfare language is a judicially enforceable restriction at all, citing Buckley v. Valeo.12 Commentary on the clause agrees. The ends it names have been construed so broadly that Congress holds an almost unreviewable discretion to define them, and the power extends to any tax or expenditure that Congress perceives as beneficial to the common defense or general welfare.13
The deference reflects the kind of question the limit raises. Deciding which national goals deserve public money means ranking competing uses of limited funds, a judgment the Constitution entrusts to Congress and one for which courts have no workable measure; no judicial standard says that disaster relief serves the nation better than highway safety. The check on spending the public dislikes is therefore political. A parallel context shows the same instinct. The Court would have upheld direct cash payments or tax subsidies to in-state milk producers financed by the state’s own general taxpayers, because the people who pay for a subsidy will weigh its costs and benefits, and it saw a broader role for judicial review only where the costs were shifted to out-of-state persons.14 A challenger also has trouble reaching the question at all, because a person suing simply as a general taxpayer, who is affected only indirectly by a spending program, stands in a very different position from one who is directly harmed.15
The limit took only a few sentences to satisfy in Dole. Congress had addressed a dangerous interstate problem, young people combining drinking with driving, and the Court accepted that it called for a national solution.16 Conditioning a share of highway money on a uniform minimum drinking age was a means “reasonably calculated to advance the general welfare,” so the Court readily concluded that § 158 satisfied the limit.17 The ingredients that carried the day were a public problem that Congress identified and a funding condition aimed at that problem.
The gap between that boundary case and Dole shows how much work the limit does. It screens out spending with no public justification at all, and almost nothing else, since a congressional judgment that a program serves the nation will almost always be accepted.18 Deference about ends does not turn the clause into a general license, however: an exercise of the power must still take the form of a tax or an expenditure.19 Because most conditions pass this limit and the clear-statement limit without difficulty, the most important criteria are the latter two.20
Unambiguous Conditions
The second limit is a clear-statement rule, and it comes from earlier decisions and not from the words of the Spending Clause. Dole, quoting Pennhurst State School and Hospital v. Halderman, said that if Congress desires to condition a state’s receipt of federal funds, it “must do so unambiguously . . . , enabl[ing] the States to exercise their choice knowingly, cognizant of the consequences of their participation.”22 The rule operates on the text of the grant statute. Congress must write the condition into the statute in terms a state official could read before accepting the money, so that the price of the funds is visible when the state decides whether to take them. A condition that Congress intended but did not clearly state does not bind the state, however sensible it may be as policy.
The rule follows from how conditional spending is justified. Spending legislation is much in the nature of a contract, so a condition’s legitimacy rests on the state’s accepting it voluntarily and knowingly, and states cannot knowingly accept conditions of which they are “unaware” or which they are “unable to ascertain.”23 A state free to refuse the money keeps its choice, which is why a funding offer differs from a federal command, but that choice has value only if the state can read the terms. A right to decline an offer whose contents are hidden is a right to decline blind, and an acceptance given blind is not the voluntary and knowing assent that the contract frame requires. The rule also forces Congress to announce a condition openly, so that state officials, and the voters who hold them accountable, can decide whether the money is worth the strings before the state commits.
The rule bites when a statute announces an obligation without saying that the money depends on it. A state reading such a statute cannot tell that taking the grant carries the obligation, so it cannot be held to it. Section 158 sits at the other pole. It told states in plain terms that a share of their federal highway money would be withheld if they allowed people under 21 to buy alcohol, so the condition and its financial consequence appeared together in the text. South Dakota could read the price of its highway money off the face of the statute, and the Court said that the conditions “could not be more clearly stated by Congress.”24 The importance of the underlying policy never decides the question. What matters is whether a state reading the statute would see that its funds depended on the obligation.
A statute need not be perfectly self-explanatory to pass. Notice is judged by the statutory language, but when that language is ambiguous a court may look to prior judicial decisions or administrative regulations that clarified it.25 In Jackson v. Birmingham Board of Education, the Court on this basis held that Title IX, which on its face bars sex discrimination by recipients of federal funds, also prohibits retaliation against those who complain of violations, even though Title IX is silent on retaliation.26 The reasoning follows from the contract frame. The question is whether the state could ascertain the condition when it took the money, so a clarifying decision or regulation already on the books at that time forms part of what the state could learn, while one that arrives later cannot supply notice that was missing at acceptance. That timing point matters again when this chapter reaches the Medicaid expansion, in the discussion of the anti-coercion principle.
Relatedness to the Program
The third limit asks whether a condition has anything to do with the money it accompanies. The Court drew the idea from earlier cases that had suggested, “without significant elaboration,” that conditions on federal grants might be illegitimate if they are unrelated “to the federal interest in particular national projects or programs.”28 Courts and commentators call the requirement relatedness or germaneness.29 Because the Court said so little, this is the limit students find hardest, and a careful answer to a hard application often has to say that the law has not settled the point.
The limit keeps conditional spending from becoming a general regulatory lever. A term that advances the funded purpose is part of what the money buys; a term with no connection to the program is regulation that merely travels with the money. Justice O’Connor’s dissent put the danger bluntly: on her view, a lax standard would let Congress regulate almost any area of a state’s social, political, or economic life on the theory that use of the interstate transportation system is somehow enhanced.30 A counterweight explains why the limit is applied loosely. Conditions often emerge from legislative bargaining, which leads some to ask why courts should insist on germaneness at all, and the Dole majority took a broad view of relatedness.31
The majority’s method has two steps. A court first identifies the purpose of the spending program to which the condition is attached, and then decides whether the condition is sufficiently relevant to achieving that purpose.32 The standard is essentially a rational-relationship test.33 Dole shows both steps at work. Highway funds are spent for many ends, and one of the main ones is safe interstate travel.34 Differing state drinking ages stood in the way of that goal, because they created a danger of unsafe cross-border movements, and a uniform minimum age attacked that problem.35 The Court therefore found the condition reasonably calculated to advance the federal interest that gave rise to the highway program.36 South Dakota offered little resistance, having said that it never contended the congressional action was unrelated to a national concern in the absence of the Twenty-first Amendment.37
The first step does more work than students expect, because the answer to the second depends on how the program’s purpose is described. The drinking-age condition that passed on highway funds would likely fail if Congress attached it to a state’s federal housing-assistance money, since housing funds have no evident link to drunken interstate driving.38 The condition stayed the same and the relationship disappeared, because the funded purpose changed. An exam answer should name the program’s purpose before it judges the condition.
The majority describes the condition as “directly related” to one of the main purposes for which highway funds are expended, and the phrase tempts students to write that Dole requires a direct relationship.39 The Court decided no such thing. A footnote explains that friends of the court urged a rule that a condition is legitimate only if it relates directly to the purpose of the expenditure, that the petitioner had not sought such a restriction, and that the Court found any such limitation satisfied in any event. The Court accordingly declined to address whether conditions less directly related to the purpose might fall outside the spending power.40 Two statements are therefore true together: Dole sustained a condition with a close tie, and it never said that a looser tie would be invalid.
Justice O’Connor’s dissent shows how much room that open question leaves. She argued that the majority’s deference to any conceivably rational relationship abdicated the judicial responsibility to ensure that a condition is “reasonably related to the expenditure of federal funds.”41 In her view a minimum drinking age of 21 cannot fairly be said to be reasonably related to the expenditure of funds for highway construction, because the only possible connection, highway safety, has nothing to do with how the funds are expended.42 She added that § 158, if its purpose is to deter drunken driving, is over-inclusive, because it stops teenagers from drinking even when they are not about to drive on interstate highways, and under-inclusive, because teenagers are only a small part of the drunken driving problem.43 Her sharpest illustration was a federal demand that a state move its capital, which the United States might defend by arguing that interstate travel is easier when state government sits near the highways.44 At argument, moreover, counsel for the United States conceded that a condition requiring a unicameral legislature would violate the germaneness requirement, so the limit has some force even on the government’s own account.45
O’Connor would also have changed the question, and not only applied the standard more strictly. Her line asks whether a requirement specifies in some way how the money should be spent so that Congress’s purpose in making the grant is carried out, and she treated any requirement that does more as a regulation, valid only if one of Congress’s delegated regulatory powers supports it.46 On that view the drinking-age condition fell into neither of the two classes she recognized, conditions determining how federal money is spent and conditions justified by a regulatory power, because it was a regulation determining who may drink liquor.47 Fullilove v. Klutznick illustrates the second class. The Court found the minority business enterprise program justified as a valid regulation under the commerce power and § 5 of the Fourteenth Amendment, and it reasoned that the reach of the spending power within its sphere is at least as broad as Congress’s regulatory powers, so that if Congress could have achieved the program’s objectives through a regulatory power, it may do so through spending.48 O’Connor cited Lau v. Nichols, which upheld nondiscrimination provisions applied to local schools receiving federal funds, alongside it.49 The authorities carry different weight. Fullilove supplies a fallback that the Court itself reasoned from, while the how-the-money-is-spent test is a dissenting position and should be presented on an exam as the road the majority did not take.
Between Dole’s close fit and the state-capital hypothetical lies a wide band of conditions the Court has never sorted. A condition requiring a state to employ dog catchers in every county crossed by a federal highway shows the difficulty. A court might accept it because loose animals plausibly cause crashes, might demand proof that dog-catcher employment actually improves safe interstate travel, or might treat Dole as inapposite because the condition solves no problem of interstate non-uniformity.50 The Court has not said which approach governs, and the footnote in Dole leaves the outer line undefined.51
The inquiry looks to the program’s purpose. O’Connor’s over- and under-inclusiveness argument did not carry the majority, so a condition does not fail merely because it reaches too much or too little of the problem it targets. A condition that passes relatedness can still fail the other limits, which the following sections take up. On a hard fact pattern, the best answer names the program’s purpose, applies the relevance question, identifies the unsettled question of how close the fit must be, and commits to a conclusion with a reason.
Independent Constitutional Bars
The Spending Clause grants Congress power to tax and spend for the general welfare, and like every other grant in Article I it operates inside a Constitution that contains other commands.53 The fourth limit takes that structure into account by asking whether some other constitutional provision independently bars the conditional grant. In Dole the only candidate was the Twenty-first Amendment.54 The label “independent constitutional bar” sounds capacious, and South Dakota read it that way. Relying on its view that the Amendment prohibits Congress from regulating drinking ages directly, the State argued that Congress may not use the spending power to regulate what it is forbidden to regulate directly.55 On that reading every limit on Congress’s regulatory power would double as a limit on its funding conditions, and a challenger would need only to show that Congress could not command the subject.
The Court rejected that reading on the strength of its earlier cases. United States v. Butler had established that the constitutional limits on Congress when it exercises the spending power are less exacting than those on its authority to regulate directly.56 Oklahoma v. Civil Service Commission had applied the same logic. The Court agreed that the United States has no power to regulate the local political activities of state officials as such, yet held that it may fix the terms on which its money allotments are disbursed, and it found no violation of the State’s sovereignty because the State could adopt the “simple expedient” of not yielding.57 A limit on what Congress may regulate therefore does not shrink the conditions it may attach to a grant.
From those decisions the Court drew a narrow meaning for the fourth limit. The bar does not forbid Congress to achieve indirectly what it could not achieve directly; it stands for the proposition that the spending power may not be used to induce states to engage in activities that would themselves be unconstitutional.58 A grant conditioned on invidiously discriminatory state action, or on the infliction of cruel and unusual punishment, would be an illegitimate use of the power.59 Dole fell on the other side of that line. If South Dakota yielded to Congress’s offer and raised its drinking age to 21, the State’s action would violate no one’s constitutional rights, so the Twenty-first Amendment supplied no independent bar.60 Under this reading, whether the Amendment would limit direct regulation of drinking ages does not matter to a spending condition.
The narrow reading rests on two reasons. The first is that individual rights bind government however it acts. A state that takes federal money and then discriminates invidiously violates a person’s rights just as it would without the money, and Congress cannot buy that violation by attaching a price to it. The second is structural. If every limit on direct regulation also limited funding conditions, spending would be no freer than regulation, which contradicts Butler’s teaching that the limits on spending are less exacting.61 Fullilove points the same way: the reach of the spending power within its sphere is at least as broad as Congress’s regulatory powers.62
The fourth limit also reaches programs that Congress writes itself. In Fullilove, the Court stressed the need for careful judicial evaluation to assure that any congressional program employing racial or ethnic criteria to remedy the present effects of past discrimination is narrowly tailored to that goal.63 A spending program thus gets no exemption from other constitutional commands, whether the violation would be committed by a state that accepts the money or by Congress in designing the grant.
The dissent in Dole shows where the fight moves once the bar is read narrowly. Justice O’Connor was willing to assume that the Twenty-first Amendment is no independent bar to a spending condition, and she placed her objection under relatedness, concluding that § 158 is a regulation determining who may drink liquor and not a condition on how federal highway money is spent.64 On an exam, an argument that a condition invades a field reserved to the states belongs under relatedness or the anti-coercion principle. The fourth limit asks only whether the conduct the condition induces is itself unconstitutional. Conditions that burden the rights of private recipients, such as First Amendment limits on how grantees may speak, raise the separate unconstitutional-conditions problem taken up later in this chapter.