A Critical Assessment of the Originalist Case Against Administrative Regulatory Power: New Evidence from the Federal Tax on Private Real Estate in the 1790s
abstract. The Supreme Court is poised to toughen the nondelegation doctrine to strike down acts of Congress that give broad discretion to administrators, signaling a potential revolution in the separation of powers. A majority of the Justices have suggested in recent opinions that they are open to the far-reaching theory that all agency rulemaking is unconstitutional insofar as it coerces private parties and is not about foreign affairs. If adopted, this theory would invalidate most of the federal regulatory state. Jurists and scholars critical of rulemaking’s constitutionality base their claims on the original meaning of the Constitution. But these critics face a serious obstacle: early Congresses enacted several broad delegations of administrative rulemaking authority. The critics’ main response has been that these early statutes do not count, because they fall into two areas in which (say the critics) the original nondelegation doctrine did not apply, or applied only weakly: noncoercive legislation (e.g., giving benefits) or foreign-affairs legislation.
This Article finds that the originalist critics of rulemaking are mistaken to say that no early congressional grant of rulemaking power was coercive and domestic. There is a major counterexample missed by the literature on nondelegation, indeed by all of legal scholarship, and not discussed more than briefly even by historians: the rulemaking power under the “direct tax” of 1798. In that legislation, Congress apportioned a federal tax quota to the people of each state, to be paid predominantly by owners of real estate in proportion to their properties’ respective values. Thousands of federal assessors assigned taxable values to literally every house and farm in every state of the Union, deciding what each was “worth in money”—a standard that the legislation did not define. Because assessors in different parts of a state could differ greatly in how they did valuation, Congress established within each state a federal board of tax commissioners with the power to divide the state into districts and to raise or lower the assessors’ valuations of all real estate in any district by any proportion “as shall appear to be just and equitable”—a phrase undefined in the statute and not a term of art. The federal boards’ power to revise valuations en masse in each intrastate tax district is identical to the fact pattern in the leading Supreme Court precedent defining rulemaking. Thus, each federal board in 1798 controlled, by rule, the distribution of the federal real-estate tax burden within the state it covered.
This Article is the first study of the federal boards’ mass-revision power. It establishes that the mass revisions (a) were often aggressive, as when the federal board in Maryland raised the taxable value of all houses in Baltimore, then the nation’s third-largest city, by 100 percent; (b) involved much discretion, given serious data limitations and the absence of any consensus method; (c) had a major political aspect, as the federal boards were inheriting the contentious land-tax politics that had previously raged within the state legislatures, pitting the typical state’s rich commercial coast against its poor inland farms; (d) were not subject to judicial review; and (e) were accepted as constitutional by the Federalist majority and Jeffersonian opposition in 1798 and also by the Jeffersonians when they later took over, indicating the boards’ power was consistent with original meaning or, alternatively, with the Constitution’s liquidated meaning. In short, vesting administrators with discretionary power to make politically charged rules domestically affecting private rights was not alien to the first generation of lawmakers who put the Constitution into practice.
More broadly, this Article is the first in-depth treatment of the 1798 direct tax’s administration. It shows that the tax, measured by personnel, was the largest federal administrative endeavor, outside the military, of the Constitution’s first two decades. It is remarkable that today’s passionate debate on whether the administrative regulatory state violates the Framers’ Constitution has so far made no reckoning with this endeavor.
This Article’s dataset is available at: https://doi.org/10.7910/DVN/IGMJ7E.
author. William K. Townsend Professor of Law, Yale Law School, and Professor of History (secondary appointment), Yale University. For comments on all or parts of the manuscript, I thank Gregory Ablavsky, Bruce Ackerman, Jonathan Adler, Kevin Arlyck, Nicholas Bagley, Jack Beermann, Emily Bremer, Daniel Carpenter, Ronald Cass, Charlotte Crane, Saul Cornell, Katherine Mims Crocker, Blake Emerson, Frank W. Garmon, Jr., Philip Hamburger, Kristin Hickman, Daniel Hulsebosch, Gary Lawson, John Manning, Jennifer Mascott, Jerry Mashaw, Aaron Nielson, James Pfander, Michael Rappaport, Daphna Renan, Jed Shugerman, Chris Walker, Ilan Wurman, and participants in the Darling Foundation Originalism Works-in-Progress Conference at the University of San Diego, the Fordham Constitutional History Workshop, the George Mason University Nondelegation Roundtable, the Harvard Public Law Workshop, the Power in the Administrative State Workshop Series at the University of Pennsylvania, and the University of San Diego Faculty Workshop. For discussions about the project, I am grateful to Ian Ayres, Michael Blaakman, Maureen Brady, Peter Conti-Brown, Justin Driver, Farley Grubb, Naomi Lamoreaux, Sophia Lee, Scott C. Miller, Claire Priest, David Schleicher, Adam J. White, Robert E. Wright, and Taisu Zhang. I owe special thanks to Frank W. Garmon, Jr. and Carole Shammas, for generously sharing with me the important datasets they respectively created on the direct tax; to Yair Listokin, for in-depth discussions of certain economic aspects of the subject; and to Judith Green Watson, for advising me on the project in light of her extensive research in direct-tax records. For the crucial service of locating sources, I am indebted to the staffs of the Connecticut Historical Society, the Daughters of the American Revolution Library, the Princeton University Library Department of Rare Books and Special Collections, the Rhode Island Historical Society, the Rutgers University Library Special Collections and Archives, the Suffolk County Historical Society, the U.S. National Archives, and the Yale Law Library (especially Julian Aiken and Maryellen Larkin); and to Elizabeth Barnett and Karen Needles. For financial support, I thank Yale Law School and the Oscar M. Ruebhausen Fund. All errors are my own.
Introduction
Article I of the Constitution says: “All legislative Powers herein granted shall be vested in a Congress of the United States.”1 Since the nineteenth century, the Supreme Court has construed this language to mean Congress cannot give away its legislative powers: there is a constitutional limit on how much power Congress can delegate by statute to the President or to administrators in the executive branch.2 While every statute inevitably gives some discretion to those who implement it, the Court requires the discretion delegated to be confined to that which is “executive” in nature,3 not so broad as to be “legislative.” This principle is called the nondelegation doctrine.4 Since the 1920s, the Justices have said that a statute is constitutional under this doctrine so long as it lays down an “intelligible principle” for those administering it to follow.5
The doctrine has proven very loose. Only three statutes have ever been held to violate it, all in 1935-36.6 Statutes have passed muster despite providing only vague guidance, such as telling an administrative agency to make regulations to serve the “public interest, convenience, or necessity.”7 With little judicial constraint on its freedom to delegate, Congress has vested numerous wide-ranging powers in the President and the agencies, including everything from the Environmental Protection Agency’s statutory mandate to set national air-pollution standards that are “requisite to protect the public health” with “an adequate margin of safety,”8 to the President’s statutory power to declare a “national emergency” (an undefined term), which President Trump invoked to obtain some of the funding to build a border wall between the United States and Mexico.9
But now, for the first time in nearly a century, the Supreme Court is poised to reformulate the nondelegation doctrine, opening the possibility of a revolution in separation of powers and administrative law. In Gundy v. United States, handed down in June 2019, Justice Gorsuch, joined by Chief Justice Roberts and Justice Thomas, declared that the “intelligible principle” test should be abandoned in favor of a new approach that demands greater specificity from Congress when it delegates authority to the administrative state.10 The Court in Gundy had only eight Justices (because Justice Kavanaugh had not been confirmed at the time of the oral arguments), and Justices Ginsburg, Breyer, Sotomayor, and Kagan were unwilling to revisit the established, delegation-friendly test.11 Justice Alito expressed sympathy with Justice Gorsuch’s view and announced that he “would support” reconsidering the established test, but not with the Court short-staffed and divided 4-4. Justice Alito therefore opted to provide the fifth vote necessary to maintain the old test (and uphold the statute at issue), but only for the moment.12 In November 2019, Justice Kavanaugh—since confirmed to the Court and carrying the potential swing vote—publicly suggested that Justice Gorsuch’s “thoughtful” opinion rejecting the old test merited “further consideration.”13
What can we expect of this apparently fast-approaching revolution? What will the new nondelegation doctrine be? One important possibility was articulated by Justice Thomas in an extensive solo concurrence in Department of Transportation v. Association of American Railroads back in 2015.14 In his opinion, Justice Thomas argued that, “[u]nder the original understanding of the Constitution,” it is unconstitutional for Congress to give the President or an agency “the discretion to formulate generally applicable rules of private conduct.”15 To this principle, Justice Thomas recognized only two exceptions. The first would be if the making of the rule turned simply on a “factual determination” by the agency, for example, if the statute states a restriction that shall go into effect if a certain event occurs, and the agency then finds that the event has in fact occurred.16 Such an agency determination would be constitutionally permissible, said Justice Thomas, because it entails no “exercise of policy discretion.”17 The second exception would be if the delegated function “involved the external relations of the United States,”18 given that the Court has long viewed the nondelegation doctrine as weaker in this area because of the President’s independent constitutional authority therein.19 Regulating international trade might be an example of this second exception for matters related to foreign affairs.20 Overall, Justice Thomas’s theory of nondelegation is that all agency rulemaking governing private conduct is unconstitutional unless it turns solely on a factual determination or involves foreign relations.
Given that today’s regulatory statutes ubiquitously authorize agencies to engage in rulemaking that governs domestic, private rights, Justice Thomas’s theory in American Railroads would seem—at least if its exception for nonpolicy “factual determinations” were interpreted narrowly—to invalidate most of today’s domestic regulatory state.
But will the Court follow this far-reaching theory? Two recent opinions suggest that at least five Justices would consider it. The first is Justice Gorsuch’s concurrence in Kisor v. Wilkie, decided in June 2019.21 Joined by Justices Thomas, Alito, and Kavanaugh, this opinion expressly left the door open for Justice Thomas’s theory. In discussing the question at issue (whether courts should defer to an agency’s interpretation of its own rules), Justice Gorsuch stated that such rules are legally binding under current doctrine, but he dropped a footnote citing Justice Thomas’s American Railroads concurrence and asserted that “our precedent allowing executive agencies to issue legally binding regulations to govern private conduct may raise constitutional questions.”22
The second opinion is Justice Gorsuch’s Gundy dissent, noted earlier. Recall that it was joined by Chief Justice Roberts and Justice Thomas and then praised implicitly by Justice Alito and expressly by Justice Kavanaugh. As in Justice Thomas’s American Railroads concurrence, Justice Gorsuch in Gundy contended that the current delegation-friendly approach “has no basis in the original meaning of the Constitution”23 and declared that “the framers understood [nondelegable legislative power] to mean the power to adopt generally applicable rules of conduct governing future actions by private persons.”24 He further argued that, in order for agency rulemaking about “private conduct” to be constitutional, it must either (a) confine itself to “details” rather than “policy decisions,”25 (b) turn simply on agency “fact-finding,”26 or (c) overlap with “authority the Constitution separately vests in another branch,” such as executive power over “foreign affairs.”27 Justice Gorsuch’s exceptions for fact finding and foreign affairs follow the exceptions in Justice Thomas’s American Railroads concurrence. Justice Gorsuch adds the exception for rulemakings that fill up “details,” but this exception may be narrow, as it seems not to allow any rulemakings that make “policy decisions.”28
Even if the Court does not categorically invalidate all agency rulemaking about domestic private conduct other than fact finding, rulemaking is so ubiquitous that mere doubt about its constitutionality could work major changes in the nondelegation doctrine and administrative law more generally. Two recent proposals illustrate the potential scope of such changes. First, Ronald Cass, in an article cited by Justice Gorsuch in Gundy,29 argues that the Court should be relatively more willing to demand congressional specificity—and to pull the trigger by striking down statutes for loose delegation—when it comes to authorizations of rulemaking on domestic private conduct, as compared to other types of authorizations.30 Similarly, Ilan Wurman, in a recent Yale Law Journal Feature, refrains from a categorical claim that rulemaking on domestic private conduct is unconstitutional, instead suggesting a looser test that bans rulemaking if it goes beyond “details” to decide “important subjects.”31 But in explaining how to apply this test, Wurman says that rulemaking power regarding private rights will be more likely to fail the test, because private rights tend to be more “important” than other matters.32
Even apart from the nondelegation doctrine proper, constitutional doubt about rulemaking could have sweeping effects. The “major questions” doctrine, which states that the judiciary should not defer to an agency statutory interpretation that concerns a question of “deep ‘economic and political significance,’”33 is essentially an application of the canon of constitutional avoidance to the nondelegation doctrine.34 If the nondelegation doctrine even possibly prohibits all rulemaking as to domestic private conduct, the “major questions” doctrine automatically becomes more potent.35 More broadly, constitutional doubt about rulemaking can instill judges with a sense that our entire modern regulatory state is suspect, and that sense may motivate them to constrain that state through every available doctrinal avenue. For example, such doubt makes the Chevron doctrine of judicial deference to agency statutory interpretation seem less legitimate as applied to rulemakings, as then-Judge Gorsuch suggested during his tenure on the Tenth Circuit (citing Justice Thomas’s American Railroads concurrence).36
As the quotations above indicate, the Justices’ doubts about rulemaking’s constitutionality rest on their sense that rulemaking violates the original meaning of the Constitution. But given that the Court has expressly upheld agency power to impose binding rules on private persons outside foreign affairs since 1911 at the latest37—and given that the body of regulatory legislation enacted in reliance upon this precedent is “massive”38—a judicial reversal on the constitutionality of rulemaking would demand strong originalist evidence.39
But is domestic coercive rulemaking actually doubtful as a matter of original meaning? Today’s debate on nondelegation and original meaning has focused, to a large degree, on early congressional statutes.40 This focus fits nicely with the Supreme Court’s original-meaning case law, which has often assigned much weight to statutes enacted in the period from Congress’s first session in 1789 through the very early 1800s.41 The focus on early federal statutes is also quite understandable given that sources bearing on original meaning besides early congressional acts—constitutional text, preratification discourse, and structure—say very little about constitutional limits on delegation.42 At most, these other sources might possibly indicate that there is some abstract, unspecified limit on delegation (I assume arguendo there is),43 but they give no useful specifics for what the content or stringency of that limit might be, to say nothing of specifying whether there is a categorical ban on domestic coercive rulemaking.44 Text, preratification discourse, and structure may tell us that there is some limit on delegation, but these sources tell us basically nothing about whether that limit should be the loose one that the Court has embraced for the last eighty-plus years,45 or the tremendously more stringent one (banning domestic coercive rulemaking) that some of today’s Justices are interested in adopting—or some other type of limit. Early legislation of Congress may tell us more about the content and stringency of any limit that originally existed, for that legislation is evidence of how practically capacious any such limit was. While practical applications are not definitive of original meaning, they may be the most usable evidence of it.46
What do acts of Congress circa 1789-1800 tell us? A long line of scholarship has pointed out that the early Congresses made many delegations of power to administrators with vague guidance for the power’s exercise, or no guidance whatsoever, several of which were delegations of rulemaking power.47 Originalist advocates for the nondelegation doctrine have often recognized that these early statutes confer broad rulemaking power but have insisted that the acts nonetheless do not count because they fall into one of two exceptional areas where the nondelegation doctrine supposedly does not apply or is supposedly weaker in a way that allows more latitude for rulemaking: (a) foreign affairs, including military affairs, or (b) the realm of voluntary transactions, government services, privileges, and benefits, as opposed to coercive regulation of private rights and private conduct. Invoking these two exceptions, originalist skeptics of rulemaking have found ways to explain away all of the early statutes delegating rulemaking power that have been invoked in prior literature.48 More than that, they have gone on the offensive: the fact that every early rulemaking authorization falls into one of the exceptions, the skeptics claim, indicates that rulemaking outside the exceptions was not constitutional.49
This Article finds, through new primary research, that the originalist skeptics of rulemaking are mistaken to say that no early congressional grant of rulemaking power was coercive and domestic. There is a major counterexample missed by the literature on nondelegation, indeed by all of legal scholarship, and not discussed more than briefly even by historians50: the rulemaking power under the “direct tax” of 1798. This tax was an enormous administrative undertaking, and it fell upon literally every farmer, homeowner, and slaveholder in every state of the Union, subjecting the farmers and homeowners to federal rulemakings that could determine their tax liabilities.51
Part I of this Article explains the fundamentals of the federal taxation of real estate that was enacted in 1798 and its early demonstration of congressional commitment to rulemaking particularly and to administrative power more generally. Although Congress had relied overwhelmingly on foreign-import duties to finance the federal government from 1789 up to the late 1790s, a fiscal shortfall struck in 1798 that pushed Congress to exercise, for the first time, its constitutional power to levy a “direct” tax (that is, roughly speaking, a tax on property52). Congress decided to raise $2 million nationwide and, per the Constitution’s requirement for direct taxes, apportioned that sum among the states according to each state’s free population plus three-fifths of its slave population. This meant a quota of $345,488 for Virginia, and $260,435 for Massachusetts, for example. In each state, slaveholders were to pay fifty cents for each enslaved person they owned.53 Once the sum levied on a state’s slaveholders was calculated, the remainder of the state’s quota—which proved to be the large majority of the quota in every one of the twelve states for which records survive, including major slave states54—was to be paid by the owners of the state’s real estate in proportion to the value of their respective properties55 (that is, what each property was “worth in money,” as the statute said, without definition56).
Well over 1,500 frontline federal assessors fanned out across the nation to assign a value to literally every house and farm in every state, and a corps of more than 600 higher-level federal assessors decided appeals from owners who thought they had been assessed out of proportion to properties in their local area. As I demonstrate for the first time, this nationwide assessment involved more federal officials than any nonmilitary undertaking of the Constitution’s first two decades.57 Yet for all the work these officials did, there was one problem they were not positioned to solve: the danger that officials in some parts of a state might generally value real estate in their respective areas in a way that was out of proportion to what officials did in other parts of the state. To address this problem, Congress established in each state a board of federal tax commissioners, appointed by the President and confirmed by the Senate, with power to divide the state into federal assessment districts and to raise or lower all assessments within any district by any percentage amount “as shall appear to be just and equitable”—a phrase the statute did not define.58 Each federal board’s district-wide mass revisions were final. There was no review by any other official, nor by any court. Intrastate mass tax valuation revisions practically identical to those authorized under the direct tax of 1798 were held not to be adjudications for due-process purposes in Bi-Metallic Investment Co. v. State Board of Equalization (1915),59 and that case has long been administrative law’s touchstone for defining rulemaking.60 Thus, the 1798 direct tax provides a clear Founding-era example of congressional delegation of rulemaking authority in a context that was both coercive and domestic: the taxation of real estate.
The direct tax of 1798 and its provision for the federal boards’ mass revisions are virtually unknown to the literatures on the nondelegation doctrine and on administrative law.61 The tax is known to tax-law scholars,62 but only one tax-law article mentions the federal boards,63 and none mentions their mass-revision power. The tax is better known outside the legal academy, to historians of slavery,64 of wealth distribution,65 of public finance,66 and of housing.67 Still, only four studies in these nonlaw subfields of history say anything about the mass revisions, and none more than briefly.68
This Article is the first study of the federal boards’ mass-revision power in any field. It is also, more broadly, the first in-depth study of the administrationof the 1798 direct tax—a study that is overdue, as the tax was the largest administrative endeavor of the federal government near in time to the adoption of the Constitution and outside the military. It is remarkable that today’s passionate debate on whether the administrative regulatory state violates the Framers’ Constitution has so far made no reckoning with this endeavor.
One cause of the neglect of the 1798 direct tax is that, although the federal boards’ mass-revision power appears on the face of the legislation, the boards’ dramatic and sweeping use of their power has been buried in sources that are quite obscure. The reports of the federal boards to the Treasury Department generally gave only the boards’ final postrevision numbers, not their prior decisions about mass revisions.69 Once the quotas in the states were assessed, the offices of the federal commissioners and assessors expired by their own terms—mostly by about 1800—and Congress chose to discontinue the smaller corps of officials who were to keep the assessment records.70 Thereafter, the records of the federal boards’ deliberations and decisions were scattered to the winds, and many were lost completely. Yet others have been preserved serendipitously—in miscellaneous federal field offices (when not used for kindling), or in the private papers of individual officials and their descendants.71 The extant records show that the federal boards were often aggressive.72 For instance, the federal board in Maryland raised the value of all houses in Baltimore (the nation’s third-largest city as of 1800) by 100% while making no change or much smaller changes in the other districts with extant records.73 The federal board in Pennsylvania raised the value of all lands in Allegheny County (a center of western Jeffersonian antagonism to the eastern part of the state) by 50% while keeping most other land the same.74 The federal board in Massachusetts raised or lowered more than half the districts in the state, including increases of 50% for Salem (the nation’s eighth-largest city as of 1800) and of 75% elsewhere.75 Each federal board had the power to determine the intrastate distribution of the real-estate tax burden, and many were not shy about using it.
Part II shows how the federal boards had wide discretion in their mass-revision rulemakings because their task was far from determinate. People could agree, say, that land was generally worth more in a thickly settled area proximate to water transport than in a remote frontier area—but how much more? It was a high-stakes question and a contestable one. Even today, real estate is the prime example of an asset whose value is hard to determine, for unlike commodities, which are standardized and traded on public exchanges that provide a knowable market price, real estate is heterogeneous and illiquid. Valuation was even harder in the 1790s, when markets were thinner, data harder to gather, and methods for reasoning from it less developed. Alexander Hamilton recognized this problem in 1797 when he urged those in power to drop the idea of value-based taxation of land and to adopt a tax on houses pegged not to value but instead to each house’s objective features (i.e., so many cents’ tax for each room, each chimney, each staircase, etc.)—an approach that he said would avoid “the very bad business of valuations.”76
Numerous sources indicate that Hamilton was right about the uncertainty of valuation. The Continental Congress in the 1780s had attempted a nationwide land-valuation project but concluded it would never be possible to reach consensus on land values.77 The states that drew upon assessors’ valuations of real estate for their own taxes circa 1798 often gave no specifics on the principles, methods, or evidence to be used, and the few states that gave specifics did not agree on a common principle.78
Moreover, valuation in 1798 entailed many data limitations and open methodological choices that were obstacles to building a consensus around any determinate approach. Methods based on income estimation required data that were hard to get. Other methods based on historical sale prices ran into trouble because deeds might not reflect recent or true prices, recent sales were often few, and sales in any event were not a random or representative sample of a district’s stock of land. Both methods required contestable guesses about whether past economic data fit with present and future conditions, especially considering that American land prices had been volatile for much of the late 1700s. Additionally, America in 1798 was suffering or just exiting a sudden recession and a money-supply slump at a time when such shocks had very disparate effects on land prices in different parts of a state—not to mention that Philadelphia and New York had been evacuated and cut off from commerce as of the legal date of the valuation due to yellow fever.79
Congress, in writing the direct-tax legislation, recognized the uncertainty of the federal boards’ mission and did not even try to tell them how to do it.80 The statutory standard for the boards’ mass revisions—“just and equitable”—connoted discretion; the phrase was not a term of art with a more specific meaning.81 As Treasury Secretary Wolcott told the House when he first proposed delegating valuation to a federal board in each state, “[T]here appears to be no necessity that the principles of valuation should be uniform in all the States.”82 After the tax was enacted, the Treasury Department’s guidance to the federal boards was reticent to recommend any particular data source beyond urban areas and was nonbinding even for those areas,83 and the boards in their own regulations took different approaches from each other on key matters like how much to rely upon historical sale prices and whether to take account of alleged abnormalities in the money supply.84 While federal officials’ postrevision average valuations per acre by district generally show a correlation with districts’ population density (which in turn is correlated with a host of things that officials might think are related to value), many districts nonetheless see wide departures between the officials’ valuations and the predictions of a density-based model (e.g., in Virginia, 38% of the state’s real estate by taxable value was located in districts whose official valuation was either more than 40% above or less than 40% below what a density-based model predicts).85 Thus, while people tended to find that land in more thickly settled areas was worth more, this was only a tendency, leaving officials to decide much on the basis of who-knows-what other considerations that were distinct from density and its many correlates.
Part III demonstrates that the federal boards’ discretion-laden rulemakings had an important political aspect. The key thing to understand is that each federal board was inheriting the intrastate political struggle that had been ongoing in its state with respect to the state government’s property taxes in the years leading up to 1798. With regard to those state taxes, all the state legislatures who taxed real estate by value felt the need to address the danger of inconsistency in officials’ approaches to valuation across the state, but the state legislatures, in addressing this danger, jealously kept the intrastate geographic distribution of state taxation in their own hands—without delegation. The state legislatures did this by hammering out various specifics in their property-tax statutes, such as the average per-acre value of land in each county, or each county’s (or town’s) percentage of the statewide burden of a purportedly value-based tax. For example, an act of the Maryland legislature mandated that the average per-acre taxable value of land in Prince George’s County had to come out to 31½ shillings; in Caroline County, 15¾ shillings; and so forth, with a number for each of the state’s counties.86 No state legislature allowed administrators to determine the distribution of taxable value or of tax liability across a geographic area larger than a county, to say nothing of allowing administrators to decide such things across a whole state. Such geographically sweeping distributive decisions were made solely by elected politicians.87 Indeed, when the state legislatures hammered out the intrastate geographic apportionments of value or of liability for their respective property taxes, their decisionmaking tended to follow a pattern of political struggle that repeated itself in state after state: on one side, the seaports and the rich, coastal, market-oriented farms; and on the other side, the “back country” inland farms that were poorer and more subsistence-oriented.88 Tax politics within a state often got nasty, and deciding the relative value of real estate in rival parts of a state partook of that nastiness. In 1783, an address adopted by the Continental Congress and drafted by James Madison recognized “the local injustice and discontents which have proceeded from valuations of the soil in every State where the experiment has been made.”89
When Congress in 1798 enacted the first federal property tax, it could have decided, in the text of the federal legislation, the average taxable value of real estate (or, equivalently, the total tax liability) of each locality of each state. That would have tracked what the state legislatures had always been doing (and it is actually what Congress itself did in one of the multiple direct taxes it enacted later, in the 1800s). Congress in 1798 also could have made specific distributive decisions in a variety of other ways. But it did not. Instead, Congress in 1798 established the federal boards and delegated to them the power to decide all these discretionary and politicized questions of mass tax valuation through rulemaking, “as shall appear to be just and equitable.” Each federal board was in position to manage the potential struggles within its own state, meaning that Congress would not have to.90 Notably, each federal board in 1798 was structured like a mini state legislature: it had one member from each of several congressionally drawn geographic “divisions” of the state; it operated by majority vote; and commonly the board members had previously served, or were simultaneously serving, as state legislators.91 Political controversy over valuation could indeed flare up on the federal boards, as in South Carolina, where one member (from a division in the backcountry) resigned in protest when his fellow members opted not to imitate the approach to land valuation that the state’s legislature had previously adopted in response to the backcountry’s complaints of unfairness.92
Part IV shows that the federal boards’ mass revisions were final and absolutely binding on taxpayers: none of the potential avenues for judicial review of tax administration in the period were available to review the revisions, either in general or as applied. The 1798 legislation provided each assessed property owner with a statutory appeal to a higher-level assessment official to ensure consistent assessment of properties within the same district; these appeals occurred prior to the federal boards’ mass revisions, and the legislation provided for no review of those mass revisions. Nor was there any nonstatutory means to obtain judicial review of the quantum of an assessment (as distinct from more categorical questions like whether property was taxable at all), whether in equity or at common law, including writs of error, certiorari, or mandamus. Nor was there any opportunity for judicial review of the quantum of the assessment during the enforcement process (in which the delinquent taxpayers’ goods could be seized by distress, or their land sold against their will). Nor was there such an opportunity in tax-title litigation after enforcement, nor in tort suits against officers for unlawful distress of goods after enforcement.
Part V shows that the federal boards’ power achieved wide, enduring, and bipartisan acceptance at a constitutional level, indicating the power’s consistency with the Constitution’s original meaning or, alternatively, with its liquidated meaning.93 There were no constitutional objections (indeed no objections at all) to the delegation of the mass-revision power in the extensively recorded debates on the 1798 direct-tax legislation. Though mainly a Federalist measure, the legislation passed by 3-to-1 majorities in the House (winning most of the Jeffersonian votes) and unanimously in the Senate, and the federal boards’ mass-revision power was explained and endorsed at length by Representative Albert Gallatin, the House Jeffersonian leader. The tax drew no nondelegation objections from lawmakers even as the Jeffersonian opposition made vociferous constitutional objections to the near-simultaneous Alien and Sedition Acts, and even as Gallatin himself made express nondelegation arguments against the near-simultaneous bill to authorize the President to raise a provisional army.94 To be sure, Jefferson himself opposed the direct tax as a political matter, but he did not question its constitutionality in public. Even in private letters he questioned its constitutionality only in the sense that it was part of a larger military buildup that he considered unrepublican; he never suggested—even in private—that the tax’s administration was unconstitutional.95 On the contrary, when Jefferson became President in 1801, he joined with his Treasury Secretary (Gallatin) and his congressional allies to take affirmative measures to complete the assessment process (including mass revisions) where it had been delayed in South Carolina.96 Years later, when the Jeffersonian Congress prosecuted the War of 1812, it enacted its own direct taxes. The first of these, in 1813, omitted the mass-revision power (instead apportioning the tax burden in the statute itself, county-by-county within each state), but lawmakers’ reasons for the shift were prudential (to avoid the delay that mass revisions would cause), not constitutional.97 As the War of 1812 worsened and created more need for revenue, Congress in the winter of 1814-15 imposed a permanent annual direct tax, and this time it reinstated the federal boards and their mass-revision power, using the same open-ended “just and equitable” language as in 1798.98 This confirms widespread acceptance of the power’s constitutionality under different parties. Indeed, when Congress resorted to direct taxation of real estate for the last time—to help finance the Civil War in 1861—it essentially copied the mass-revision power from the winter of 1814-15, right down to the language “just and equitable.”99
Overall, the sweeping rulemaking powers of the federal boards of tax commissioners who administered the direct tax of 1798—combined with their decisions’ binding power (insulated against judicial review) and the wide bipartisan acceptance of their power at the time and in subsequent enactments into the 1800s—are important evidence that the American political nation in the Founding era viewed administrative rulemaking as constitutional, even in the realm of domestic private rights. Congress’s willingness to delegate rulemaking power in that realm was similar to the willingness it showed in other realms during the Constitution’s early years of implementation.100 Vesting power in administrators to make sweeping discretionary decisions with high political stakes was not alien to the federal lawmakers who first put the Constitution into practice.
* * *
Before closing this Introduction, let me discuss two responses that skeptics of rulemaking’s constitutionality may give to these findings. First, no matter how sweeping, indeterminate, and politicized the questions of mass valuation were, the skeptics could potentially choose to label them as “factual” in nature, given that the definition of what questions are factual is quite malleable, especially if we do not require a factual question to be objective or specific.101 Labeling the federal boards’ mass revisions as factual would cause them to fall within Justice Thomas and Justice Gorsuch’s exceptions for determinations of fact (as distinct from policy), making the rulemakings technically consistent with those Justices’ theories. But if the skeptics do this, then the 1798 legislation provides an originalist basis for construing those Justices’ factual exceptions to a constitutional ban on rulemaking so broadly as to bless most and perhaps all statutory authorizations for rulemaking.102 That would actually fit nicely with a long line of Supreme Court cases that have upheld agency rulemaking powers as turning on supposedly factual questions when it was obvious they were actually highly subjective questions laden with policy choice.103
Second, the skeptics might recognize a third exception to the nondelegation doctrine, for taxation, on top of their existing exceptions for noncoercive measures and for foreign and military affairs. I think such a move would be untenable, for three reasons. First, the nondelegation doctrine is premised on Article I’s vesting of “legislative power” in Congress, and the Framers explicitly and repeatedly defined taxation as a legislative power. In The Federalist, Hamilton asked, “What is the power of laying and collecting taxes, but a legislative power, or a power of making laws, to lay and collect taxes?”104 And Madison said that “there is, perhaps, no legislative act in which greater opportunity and temptation are given to a predominant party to trample on the rules of justice,” than the “apportionment of taxes, on the various descriptions of property.”105 Second, the Supreme Court held unanimously in 1989 that the nondelegation doctrine applies equally to taxation as to other types of legislation.106 Third, Justice Gorsuch’s own formulation seems to leave no space for a tax exception. His Gundy opinion defines legislative power as “the power to adopt generally applicable rules of conduct governing future actions by private persons.”107 Ordering landowners to pay a sum of their money to the government, under penalty of having their goods or lands seized, is a rule of conduct for private persons. To be sure, it governs conduct as between a private person and the government, not conduct as between one private person and another.108 But I fail to see why the imperative for laws to be made by Congress would be weaker for governing the former type of conduct than the latter. (If anything, my intuition runs the opposite way.) And Justice Gorsuch seems to agree that conduct as between private persons and the government would fall within this definition. In Gundy, he defined as “rules of conduct” the requirements of the Sex Offender Registration and Notification Act, which generally required previously convicted persons to report and disclose information about themselves to the government, which would then disseminate that information to others.109
In using this term, I follow convention. But cf. Philip Hamburger, Delegating or Divesting?, 115 Nw. U. L. Rev. Online 88, 88 (2020) (arguing that speaking in terms of a constitutional ban on “vesting and divesting” legislative power, rather than “delegating” it, would better capture the Constitution’s meaning).
50 U.S.C. § 1621(a) (2018); 10 U.S.C. § 2808(a) (2018); see also A Guide to Emergency Powers and Their Use, Brennan Ctr. for Just. i, 3-43 (Dec. 5, 2018), https://www.brennancenter.org/sites/default/files/2019-10/2019_10_15_EmergencyPowersFULL.pdf [https://perma.cc/RU3H-425Z] (detailing the 136 statutory powers that can become available to the President if a national emergency is declared).
Id. at 2136. More recently, Justice Gorsuch joined a separate opinion by Justice Thomas that cites the American Railroads concurrence and says it is “constitutionally suspect” for an agency to make “general rules.” Cty. of Maui v. Haw. Wildlife Fund, No. 18-260, slip op. at 5 (Apr. 23, 2020) (Thomas, J., dissenting) (citing Dep’t of Transp. v. Ass’n of Am. R.Rs., 575 U.S. 43, 67-87 (2015) (Thomas, J., concurring in the judgment)).
Cf. Paul v. United States, 140 S. Ct. 342, 342 (2019) (mem.) (Kavanaugh, J., respecting the denial of certiorari) (signaling interest in the major-questions doctrine); Aditya Bamzai, Delegation and Interpretive Discretion: Gundy, Kisor, and the Formation and Future of Administrative Law, 133 Harv. L. Rev. 164, 176 (2019) (noting that the adoption of new principles of nondelegation will affect statutory construction under the avoidance canon).
See Philip Hamburger, Is Administrative Law Unlawful? 83-110 (2014) (discussing rulemaking statutes); id. at 191-226 (discussing adjudicatory statutes); Jerry L. Mashaw, Creating the Administrative Constitution: The Lost One Hundred Years of American Administrative Law 44-48 (2012); Michael W. McConnell, The President Who Would Not Be King: Executive Power Under the Constitution 329-35 (2020); Joseph Postell, Bureaucracy in America: The Administrative State’s Challenge to Constitutional Government 73-79 (2017); Bamzai, supra note 35, at 182; Cass, supra note 29, at 155-58; Kenneth Culp Davis, A New Approach to Delegation, 36 U. Chi. L. Rev. 713, 719-20 (1969); Aaron Gordon, Nondelegation, 12 N.Y.U. J.L. & Liberty 718, 744-50 (2019) [hereinafter Gordon, Nondelegation]; Jennifer Mascott, Early Customs Laws and Delegation, 87 Geo. Wash. L. Rev. 1388 (2019); Julian Davis Mortenson & Nicholas Bagley, Delegation at the Founding, 121 Colum. L. Rev. 277, 332-66 (2021); Eric A. Posner & Adrian Vermeule, Interring the Nondelegation Doctrine, 69 U. Chi. L. Rev. 1721, 1735-36 (2002); Joseph Postell & Paul D. Moreno, Not Dead Yet—or Never Born? The Reality of the Nondelegation Doctrine, 3 Const. Stud. 41, 44-48 (2018); Michael B. Rappaport, The Selective Nondelegation Doctrine and the Line Item Veto: A New Approach to the Nondelegation Doctrine and Its Implications for Clinton v. City of New York, 76 Tul. L. Rev. 265, 335-38 (2001); Wurman, supra note 31, at 1503-18, 1540-56; Christine Kexel Chabot, The Lost History of Delegation at the Founding, 56 Ga. L. Rev. (forthcoming 2021), https://ssrn.com/abstract=3654564 [https://perma.cc/U33F-VYQ5]; Aaron Gordon, A Rebuttal to “Delegation at the Founding” 32-51 (Apr. 23, 2020) (unpublished manuscript), https://ssrn.com/abstract=3561062 [https://perma.cc/R3RE-HKTF] [hereinafter Gordon, Rebuttal]; cf. Gary Lawson, Delegation and Original Meaning, 88 Va. L. Rev. 327, 396-403 (2002) (denying such statutes’ importance to originalist methodology but arguing their merits in the alternative).
Although the First Congress of 1789-91 is sometimes especially privileged, opinions of the Court rely on statutes throughout the 1790s and as late as 1806 in interpreting the Constitution’s original meaning. Michael Bhargava, The First Congress Canon and the Supreme Court’s Use of History, 94 Calif. L. Rev. 1745, 1768-69 (2006); see, e.g., Cent. Va. Cmty. Coll. v. Katz, 546 U.S. 356, 373-78 (2006) (relying upon early federal statutes to interpret constitutional meaning, mainly one from the year 1800, on the ground that they were passed in the “immediate wake of the Constitution’s ratification”); id. at 385-86 (Thomas, J., dissenting) (agreeing with the majority that “the practice of the early Congresses can provide valuable insight into the Framers’ understanding of the Constitution” while disagreeing on what implications to draw from early practice in the particular case); Alden v. Maine, 527 U.S. 706, 743-44 (1999) (relying upon early federal statutes, including two from 1798); Printz v. United States, 521 U.S. 898, 905 (1997) (“[E]arly congressional enactments provide contemporaneous and weighty evidence of the Constitution’s meaning.” (internal quotation marks omitted)); id. at 905-10 (examining one statute from 1789, two from 1790, one from 1791, one from 1793, three from 1798, and one from 1802, without suggesting that any of these acts were more or less probative than the others by reason of temporal proximity to ratification, and making conclusions about what the acts “establish[ed]” as to how “the Constitution was originally understood”); id. at 948-54 (Stevens, J., dissenting) (analyzing statutes from the period between 1789 and 1802 without questioning their relevance); see also Nev. Comm’n on Ethics v. Carrigan, 564 U.S. 117, 122-23 (2011) (drawing upon, and making no differentiation in weight between, a House of Representatives rule of 1789 and a Senate rule of 1801). Conceptually, early statutes may be considered evidence of original meaning itself, or they may constitute the kind of regular practice that can (as James Madison said) “liquidate and settle the meaning” of open-ended constitutional terms. See William Baude, Constitutional Liquidation, 71 Stan. L. Rev. 1, 11, 61-63 (2019).
As to text: Article I vests “[a]ll legislative Powers” in Congress but does not define “legislative power” or specify the difference between such power and the “executive power” to implement a statute that Congress can legitimately confer. U.S. Const. art. I, § 1. At least two originalist defenses of the nondelegation doctrine have recognized this textual vagueness. Larry Alexander & Saikrishna Prakash, Reports of the Nondelegation Doctrine’s Death Are Greatly Exaggerated, 70 U. Chi. L. Rev. 1297, 1310-17, 1310 n.43 (2003) (analyzing Founding-era understandings of the term “legislative,” denying that “all rulemaking occurring outside of Congress necessarily amounts to an exercise of legislative power,” and stating that “legislative power was a matter of the degree/scope of the discretion” exercised); Rappaport, supra note 40, at 305 (“[T]he term ‘executive power’ is ambiguous.”). For elaboration on the textual question, see Nicholas R. Parrillo, Supplemental Paper to: “A Critical Assessment of the Originalist Case Against Administrative Regulatory Power: New Evidence from the Federal Tax on Private Real Estate in the 1790s,” at 3-4 (C. Boyden Gray Ctr. for the Study of the Admin. State Research Paper No. 20-17 Supplement, 2020), https://ssrn.com/abstract=3696902 [https://perma.cc/V52X-SWK4] [hereinafter Parrillo, Supplemental Paper].
As to preratification discourse: I have combed through what I believe to be all originalist scholarship supporting the nondelegation doctrine, and I have found citations to only a handful of American sources from the start of the revolutionary crisis in 1765 through ratification in 1788 that assert a constitutional or other legal limit on a legislature’s conferral of power on others. For the citations, see Hamburger, supra note 40, at 384; McConnell, supra note 40, at 329-32; Postell, supra note 40, at 22-24, 32-33; Gordon, Nondelegation, supra note 40, at 741, 744; Gordon, Rebuttal, supra note 40, at 21-32; Wurman, supra note 31, at 1531-32, 1542. As I explain at length in a supplemental paper that directly examines all the cited sources and their context, every such assertion has at least one of the following characteristics, any one of which prevents the assertion from being useful for learning about the content or stringency of any nondelegation doctrine: (a) the assertion was rejected by a majority of the state political community to which it was addressed; (b) it concerned a delegation more radical than anything the U.S. Congress has ever attempted (for example, authorizing revolutionary councils of safety with unbounded power to create capital crimes); (c) it concerned a delegation whose content is unknown from extant sources; (d) it concerned, by its most natural reading, at most the existence of some limit in principle, without specifics; or (e) it was stated equivocally and thus was not really an assertion of a limit. See Parrillo, Supplemental Paper, supra, at 5-8, 36-49.
As to structure: leading works arguing that the nondelegation doctrine follows from the Constitution’s original structure recognize that their method does not give any specifics on the doctrine’s content or stringency. They say the limit for which they argue is “vague and difficult to apply,” Rappaport, supra note 40, at 312, and that any attempt to verbalize the limit is so “self-referential” that a “rule-of-law devotee . . . flees from it as a vampire flees garlic,” Lawson, supra note 40, at 361. For elaboration on structure, see Parrillo, Supplemental Paper, supra, at 8-9.
Justice Scalia accepted the nondelegation doctrine in principle but considered its stringency when enforced by courts to be a separate question; indeed, he found the Court’s loose enforcement of the doctrine to be justifiable on the ground that judges were not capable of drawing a principled, workable boundary between permissible and excessive delegations. Mistretta v. United States, 488 U.S. 361, 415-16 (1989) (Scalia, J., dissenting).
This work began with James Hart, The Ordinance Making Powers of the President of the United States 72-89 (1925), and grew from there. See David P. Currie, The Constitution in Congress: The Federalist Period, 1789-1801, at 146-49, 163 n.240, 186-88, 244-48, 254-59 (1997); Mashaw, supra note 40, at 44-48; Chabot, supra note 40; Davis, supra note 40, at 719-20; Mortenson & Bagley, supra note 40, at 332-66; Posner & Vermeule, supra note 40, at 1735-36.
Thus, power delegated in 1790 to raise or lower the scale of all disabled soldiers’ benefits has been explained away on the ground that benefits are a privilege, not a right, see Hamburger, supra note 40, at 86; Bamzai, supra note 35, at 182, or that it pertains to military affairs, Lawson, supra note 40, at 400-01. Power delegated in 1790 to make regulations for persons trading with Native Americans has been explained away on the ground that trade with Native Americans was foreign affairs, see Lawson, supra note 40, at 401-02; Wurman, supra note 31, at 1545, or that such trading was a privilege, see Bamzai, supra note 35, at 182. Power delegated in 1790 to write regulations for when and where the federal government would repurchase its debt has been dismissed as governmental engagement in voluntary contracts, as distinct from coercive exercises of sovereignty. Gordon, Rebuttal, supra note 40, at 36. Power delegated in 1792 to decide the locations of all post offices and the frequency of all mail nationwide, see Currie, supra note 47, at 149, has been dismissed because mail was “merely a government service,” not anything coercive, Hamburger, supra note 40, at 202, 564 n.23. See also Gordon, Rebuttal, supra note 40, at 47 (similar). Power delegated in 1794 to lay an embargo on all international trade within a certain time period has been explained away as foreign affairs. See Dep’t of Transp. v. Ass’n of Am. R.Rs., 575 U.S. 43, 80 (2015) (Thomas, J., concurring in the judgment); Cass, supra note 29, at 157-58; Gordon, Rebuttal, supra note 40, at 40. Other delegations have not been expressly addressed by proponents of toughening nondelegation, but it is easy to imagine either of the two exceptions being invoked. Power delegated in 1792 to decide whether to double the size of the army can be dismissed as military affairsSee Marvin A. Kreidberg & Merton G. Henry, History of Military Mobilization in the United States Army, 1775-1945, at 29 (1955). Power delegated in 1798 to set the eligibility criteria for giving government money and medical care to disabled merchant seamen, see Gautham Rao, Administering Entitlement: Governance, Public Health Care, and the Early American State, 37 Law & Soc. Inquiry 627, 634-46 (2012), would presumably be dismissed as a privilege, not a right. For more on these early delegations and others, see Parrillo, Supplemental Paper, supra note 44, at 13-16. For a synthetic discussion of many of these delegations, suggesting they are all excepted from nondelegation constraints for the common reason that they fall within the historical bounds of the royal prerogative, see McConnell, supra note 40, at 333-34.
See, e.g., Cass, supra note 29, at 157-58, 198; Hamburger, supra note 4, at 104-08; Mascott, supra note 40, at 1392. In general, a negative inference can be drawn about the constitutionality of a type of congressional action if early Congresses refrained from doing the action. Alden v. Maine, 527 U.S. 706, 743-44 (1999).
The skeptics’ dismissal of all early rulemaking authorizations as falling within exceptions for noncoercive or nondomestic legislation is subject to two additional objections besides the clear counterexample of the direct tax. First, the exceptions themselves may have no originalist basis; in particular, the few lawmakers in the 1790s who talked expressly of a nondelegation doctrine apparently thought it encompassed nondomestic and noncoercive matters. See Mortenson & Bagley, supra note 40, at 279 n.7, 363; Parrillo, Supplemental Paper, supra note 44, at 18-20; Note, Nondelegation’s Unprincipled Foreign Affairs Exceptionalism, 134 Harv. L. Rev. 1132, 1140-46 (2021). Second, the two purported exceptions actually cover the very large majority of legislation enacted by Congress in 1789-99, whether it delegated authority to administrators or not. Statutes not covered by the two exceptions (acts that were domestic and coercive) were so few that, even if they contained no rulemaking authorizations, the absence could be attributed simply to the paucity of the acts, not to a supposed congressional norm against delegation peculiar to such types of acts. It is hard to infer the impossibility of a positive result from an empirical lack of positive results if you have only a “small N,” as the statisticians say. See Parrillo, Supplemental Paper, supra note 44, at 21-36 (making this claim based on a coding of all public acts of Congress in 1789-99).
Slave ownership was taxed only if the enslaved person was “above the age of twelve, and under the age of fifty years,” and not if “from fixed infirmity, or bodily disability” the enslaved person was “incapable of labor.” Act of July 9, 1798 (“Valuation and Enumeration Act” or “V&E Act”), ch. 70, § 8, 1 Stat. 580, 585. This Article does not focus on the direct tax’s administration as applied to slave ownership, for two reasons. First, the tax’s relationship to slavery has been analyzed in excellent scholarship elsewhere. See Robin L. Einhorn, American Taxation, American Slavery (2006), especially at 184-96. Second, the original-meaning debate has put my focus on rulemaking with binding effect on private rights, and while the determination of enslaved persons’ ages and disabilities may well have entailed substantial frontline adjudicatory discretion, I have not found evidence that the tax’s application to slave ownership involved the sweeping rulemaking discretion that the tax’s application to real property did. The legislation did not tax slave ownership according to value, but instead at the fixed sum of fifty cents per enslaved person. Therefore, slave ownership, unlike real-estate ownership, was not subject to mass valuation revisions, which were the administrative actions that most clearly entailed generic mass determinations uniformly deciding many individual outcomes (i.e., rulemakings) with binding effect on private rights. Significantly, when Congress, in financing the War of 1812, shifted to a valuation-based approach for taxing slave ownership, the mass-revision power of federal boards under the 1815 legislation (which I describe briefly infra Section V.D.2) extended to the valuation of enslaved persons. See infra note 755 and accompanying text. Future research on the use of that power, while perhaps too long after 1788 to be strong evidence of original meaning, would be an important contribution to the history of slavery and federal administration.
To be exact, the legislation divided real estate into “dwelling-houses” and land-without-dwelling-houses. The reason for making this division was to tax expensive homes (seen as proxies for mercantile wealth) at a set of fixed progressive rates while taxing all other real estate at a flat rate that would “float” up or down to whatever level was necessary to fill the state’s quota. For a full explanation, see infra note 143 and accompanying text.
239 U.S. 441, 445-46 (1915). The Bi-Metallic Court said there was no adjudication for due-process purposes where a statewide board made “a general determination” as to “the principle upon which all the assessments in a county had been laid.” Id. at 446. The Court distinguished the prior case of Londoner v. Denver, 210 U.S. 373 (1908). The Bi-Metallic Court explained that Londoner had involved a determination of “‘whether, in what amount, and upon whom’ a tax for paving a street should be levied for special benefits. A relatively small number of persons was concerned, who were exceptionally affected, in each case upon individual grounds, and it was held they had a right to a hearing.” Bi-Metallic, 239 U.S. at 446 (quoting Londoner, 210 U.S. at 385); see also Londoner, 210 U.S. at 381 (setting forth an administrative complaint indicating that the case concerned “Eighth avenue paving district No. 1”). Even if county-wide scope were the minimum scope necessary for a proceeding to cease being adjudicatory (which the Bi-Metallic Court certainly did not say), the 1798 legislation would still meet that test. Each federal board was to divide its state “into a suitable and convenient number of assessment districts.” V&E Act § 7, 1 Stat. at 584. A board could create districts the size of counties, or smaller or larger: boards covering five of the sixteen states drew districts that were generally each coextensive with a county or larger, and boards covering another three states drew several—though not all—of their districts coextensive with counties. See infra note 182 & tbl.1. For a board covering yet another state (North Carolina), records survive as to just one district, which was drawn coextensive with a county. See infra text accompanying note 263. When Congress reintroduced the boards in 1815, it provided for their mass revisions everywhere to be by county (or by “state district,” a reference to the county-like districts into which South Carolina was divided for local government purposes). See infra note 752 and accompanying text.
Philip Hamburger touches on the 1798 direct tax, noting the individual adjudicatory discretion of lower-level assessors; he does not discuss the tax’s rulemaking power (that is, the federal boards’ district-wide mass revisions). Hamburger, supra note 40, at 209-10. Leonard White briefly describes the tax, erroneously characterizing the federal boards’ power as an individualized appellate-adjudicatory authority and ignoring their rulemaking authority. Leonard D. White, The Federalists: A Study in Administrative History 452, 455-56 (1948).
E.g., James R. Campbell, Dispelling the Fog About Direct Taxation, 1 Brit. J. Am. Leg. Stud. 109, 142-44 (2012); Joseph M. Dodge, What Federal Taxes Are Subject to the Rule of Apportionment Under the Constitution?, 11 J. Const. L. 839, 872-75 (2009); Erik M. Jensen, The Apportionment of “Direct Taxes”: Are Consumption Taxes Constitutional?, 97 Colum. L. Rev. 2334, 2355 (1997).
Charlotte Crane, Reclaiming the Meaning of “Direct Tax,” 21-22 n.25 (Feb. 15, 2010) (unpublished manuscript), https://ssrn.com/abstract=1553230 [https://perma.cc/HB65 -QREJ].
Einhorn, supra note 53, at 184-96. Besides the just-cited passage discussing the federal direct tax of 1798 (and those of 1813-1815), Einhorn provides a larger two-hundred-year narrative that is essential background for understanding the politics of taxing property in America, including taxation of slave ownership. See id.
Robert E. Gallman & Paul W. Rhode, Capital in the Nineteenth Century 298-300 (2019); Peter H. Lindert & Jeffrey G. Williamson, Unequal Gains: American Growth and Inequality Since 1700, at 275-78 (2016); Lee Soltow, Distribution of Wealth and Income in the United States in 1798 (1989); Frank W. Garmon, Jr., Population Density and the Accuracy of the Land Valuations in the 1798 Federal Direct Tax, 53 Hist. Methods 1 (2020).
Dall W. Forsythe, Taxation and Political Change in the Young Nation 1781-1833, at 51-57 (1977); Donald R. Stabile, The Origins of American Public Finance: Debates over Money, Debt, and Taxes in the Constitutional Era, 1776-1836, at 127-31 (1998); Frederick Arthur Baldwin Dalzell, Taxation with Representation: Federal Revenue in the Early Republic 321-36 (Oct. 1993) (unpublished Ph.D. dissertation, Harvard University), ProQuest Dissertations and Theses (PQDT) No. 9412326. The tax also figures in histories of politics in the early republic, but almost exclusively in that it led to Fries’s Rebellion, a disturbance in one Pennsylvania county and some adjoining townships. But the Fries incident was far more minor than the Whiskey Rebellion of 1794 and, according to the standard history of the period, was notable mainly for the Federalist government’s overreaction to it. Stanley Elkins & Eric McKitrick, The Age of Federalism 696-700 (1993); see also Paul Douglas Newman, Fries’s Rebellion: The Enduring Struggle for the American Revolution (2004) (treating the incident in depth); Romain Huret, The Contested State: Revenue Agents, Resistance, and Popular Consent in the United States from the Early Republic to the End of the Nineteenth Century, 33 Tocqueville Rev. 87, 91-93 (2012) (focusing mainly on resistance). What is most striking, in my view, is that the early federal government levied a tax on literally every farm and house in every state and met no organized resistance much beyond a single county in a single state.
Carole Shammas, The Housing Stock of the Early United States: Refinement Meets Migration, 64 Wm. & Mary Q. 549 (2007) [hereinafter Shammas, Housing]; Carole Shammas, The Space Problem in Early United States Cities, 57 Wm. & Mary Q. 505 (2000) [hereinafter Shammas, Space Problem]; Judith Green Watson, Kent and Warren, Connecticut in 1798: Dwelling House Size and Quality, Multi-Family Living Arrangements, and Crowding, 56 Conn. Hist. Rev. 144 (2017).
Einhorn, supra note 53, at 306 n.69; Garmon, supra note 65, at 4, 9 n.6; Shammas, Housing, supra note 67, at 565 n.16; Judith Green Watson, The Implementation of the Federal 1798 Direct Tax in Connecticut, 45 Conn. Hist. Rev. 229, 232, 234-35 (2006). Einhorn briefly discusses the district-wide mass revisions under the analogous federal direct tax of 1815. Einhorn, supra note 53, at 307 n.75. She also provides important background on administrative bodies in other jurisdictions with powers analogous to those of the federal boards in 1798 and 1815. E.g., id. at 49-50, 66-71, 91-92, 141-42.
For a lucid account of the scattering and loss of the records, see Judith Green Watson, A Discovery: 1798 Federal Direct Tax Records for Connecticut, 39 Prologue Mag. 6 (2007). On the use of direct-tax records of 1798 for kindling at the Boston customhouse in the 1840s, see New Eng. Historic Genealogical Soc’y, An Index and Guide to the Microfilm Edition of the Massachusetts and Maine Direct Tax Census of 1798, at 1 (1979) [hereinafter NEHGS, Index].
Letter from Alexander Hamilton to Oliver Wolcott, Jr. (June 6, 1797), in Founders Online, Nat’l Archives, https://founders.archives.gov/documents/Hamilton/01-21-02-0059 [https://perma.cc/6346-KBC3]. (This database contains papers of leading Founders as edited and annotated in the principal modern academic collections of the various Founders’ papers.) For Hamilton’s plan to tax each house by a flat sum for each of certain listed objective features and not by value, see Alexander Hamilton, Enclosure: [Ideas on the Subject of Direct Taxes] ([Jan. 1797]), in Founders Online, Nat’l Archives, https://founders.archives.gov/documents/Hamilton/01-20-02-0319-0002 [https://perma.cc/ZT82-CVWK].
See infra Section II.E.2. One federal board, covering Connecticut, conducted substantial empirical research on historical sale prices but used this research only as one nondispositive factor in its revisions, see infra Section II.E.3, and even that level of reliance on data gathering was unusual compared to other federal boards, see infra text accompanying notes 421-422.
See infra Section III.B. The case of direct taxation thus involved a different relationship between congressional delegation and federalism values than that which Jennifer Mascott finds for the case of import duties. Mascott, supra note 40, at 1394-96 (finding that Congress’s refusal to delegate decisions about import duties furthered federalism values).
Consistency of early practice over time is evidence that the practice follows true original meaning. Vasan Kesavan & Michael Stokes Paulsen, The Interpretive Force of the Constitution’s Secret Drafting History, 91 Geo. L.J. 1113, 1166 (2003). Consistency of practice across multiple branches of government or multiple political parties strengthens the case that the Constitution’s meaning has been liquidated in a manner embodied in the practice. Baude, supra note 41, at 18-19; Curtis A. Bradley & Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411, 415 (2012).
A leading proponent of strengthening the nondelegation doctrine has noted in a commentary on Justice Gorsuch’s Gundy dissent that “there is no clear principle that can be used to draw” the distinction between factual questions and other kinds of questions; the distinction is “conventional, not metaphysical or epistemological” and “must often be drawn solely on the basis of policy.” Gary Lawson, “I’m Leavin’ It (All) up to You”: Gundy and the (Sort-of) Resurrection of the Subdelegation Doctrine, 2019 Cato Sup. Ct. Rev. 31, 66-67. Compare the view of another leading proponent, who argues that, for purposes of a strengthened nondelegation doctrine, legitimately delegable factual determinations do not include those whose uncertainty is too great, e.g., inferring the health effect of a substance at low-exposure levels from its observed effect at high levels—the kind of evidence-light determination that “can only be made on a policy basis.” Michael B. Rappaport, A Two Tiered and Categorical Approach to the Nondelegation Doctrine 23 (Oct. 17, 2020) (San Diego Legal Studies Paper No. 20-471), https://ssrn.com/abstract=3710048 [https://perma.cc/5YNZ -SK8P].
To begin with, the mandate for the federal boards “to revise, adjust and vary the valuations” in each district “as shall appear to be just and equitable,” V&E Act, ch. 70, § 22, 1 Stat. 580, 589 (1798), does not fit easily with the idea that the boards were simply to make factual determinations devoid of policy choice; the provision does not expressly call for any factual finding at all and is very open-ended. But say we ignore that problem and interpret the mandate to mean—more specifically—that each board was to bring valuations across districts in line with the board’s view of the relative average per-acre values of the districts. Even this would hardly mean the board’s task was factual to the exclusion of policy. In 1798 there was no generally accepted definition of real-estate value or method for deciding it. See infra notes 272-273 and accompanying text and Sections II.A, II.B, II.E.2. Such valuation was subject to many indeterminacies of method, not to mention problems in obtaining data if and when a method was chosen. See infra Section II.C. Contemporaries in the late 1700s and early 1800s often characterized real-estate valuation as uncertain or as variable depending upon different opinions or approaches. See infra notes 264-269, 280-286, 287-288, 374-375, 388, 397, 412, 540, 542 and accompanying text. The direct-tax legislation did not specify any definition or method; on the contrary, lawmakers voted down a mandate for lower-level officials to value land at what it “might be sold for immediate payment in money” and opted instead for the vaguer term “worth in money.” See infra notes 147-150 and accompanying text and Section II.D. And even if—contrary to what happened—Congress had adopted something like today’s concept of fair market value (that is, what the land would sell for under normal market conditions, however defined), this would have required each federal board to predict the average per-acre price at which land in each district of its state would sell. But of course, the vast majority of parcels in a district had not been sold recently and would not be sold in the near future, that is, the vast majority never actually sold under market conditions identical or close to those on the statutory date of valuation. See infra notes 353-361, 429-430 and accompanying text (estimating that usual annual rates of turnover were around 1%). Thus, the board’s determination for each district would involve taking the average of the predicted prices of a large number of future transactions (sales of all the parcels in each district of the state) that would, with few exceptions, never come to pass. It would require a mass prediction of hypothetical, never-to-occur events. Gathering data on historical sale prices in different districts would not yield determinate predictions, as the few lands sold in each district would not be a random or representative sample of all the district’s lands, and (worse) the relationship between an owner’s tendency to sell and the value of an owner’s land could vary across districts. See infra text and accompanying notes 362-365. In light of problems like these, two economic historians say “it was impossible for eighteenth-century contemporaries to produce a meaningful average price per acre,” as hedonic regression was not invented until the twentieth century. David B. Ryden & Russell R. Menard, South Carolina’s Colonial Land Market: An Analysis of Rural Property Sales, 1720-1775, 29 Soc. Sci. Hist. 599, 611-12 (2005). Consistent with the idea that resolving such indeterminacy is unavoidably a matter of policy and therefore political, the relative taxable value of real estate in different parts of each state, for purposes of state property taxes, was generally decided by the respective state legislatures as of 1798. See infra notes 455-461 (on the state legislatures’ control of mass valuation), 462-509 (on political controversy over legislative decisions regarding mass valuation) and accompanying text; see also infra Section III.D (on political controversy over federal boards’ mass revisions); Louis L. Jaffe, Judicial Review: Question of Law, 69 Harv. L. Rev. 239, 245 (1955) (noting that factual inferences from evidence depend on “theories of probability” having the status of “rules” that “are partially determined by policy, i.e., by law-making considerations”—an “aspect of the fact-finding process [that] is particularly pronounced in administrative fact-finding”).
All that said, and despite all this indeterminacy, the category of “fact,” as distinct from “policy” or “value judgment,” is susceptible to so many definitions (for a review, see Hilary Putnam, The Collapse of the Fact/Value Dichotomy and Other Essays 7-45, 135-39 (2002)) that one could presumably find some definition that would encompass the boards’ mass valuations. But that definition of “fact” would need to be very broad—encompassing determinations that are predictive, uncertain, and sweeping in scope, in which evidence-light presumptions can hardly be avoided. Such determinations play a central role in agency rulemaking. See, e.g., 1 Hickman & Pierce, supra note 60, at 597-99; Jacob Gersen & Adrian Vermeule, Thin Rationality Review, 114 Mich. L. Rev. 1355, 1384-96 (2016). As Kenneth Culp Davis said, such findings of “judgmental facts” are “mixed with judgment, policy ideas, opinion, discretion, or philosophical preference.” 3 Kenneth Culp Davis, Administrative Law Treatise § 15.10, at 180 (2d ed. 1980). Categorizing such determinations as factual and thus permissible to delegate would be to uphold the constitutionality of a large proportion of rulemaking delegations. And the “factual” category becomes even broader if we define it to encompass the power to specify the definition (or, equivalently, the methodology) for vague concepts like “worth in money,” to say nothing of deciding what “shall appear to be just and equitable.”
A side note: one might think we could shed light on the fact-or-policy status of the federal boards’ determinations by considering that courts of the early republic, in litigation over land, viewed valuation as a question for the jury. See, e.g., Faw v. Marsteller, 6 U.S. (2 Cranch) 10, 32 (1804). But jury questions were not confined to factual matters in this period. See Daniel D. Blinka, Jefferson and Juries: The Problem of Law, Reason, and Politics in the New Republic, 47 Am. J. Legal Hist. 35, 36-40 (2005). More importantly, even if valuation were considered a factual question in court, this would only shed light on valuation’s status for purposes of the binary fact/law distinction characteristic of the judiciary, not for purposes of the tripartite fact/law/policy distinction at issue in today’s debate on legislative delegation to the executive. See, e.g., Rappaport, supra note 101, at 11-24 (arguing that it is constitutional for Congress to delegate law questions and fact questions but not policy questions). And even if we were to categorize a court’s determination of land’s value as factual for purposes of a tripartite fact/law/policy distinction, we must keep in mind the vast difference between such a determination (confined to one or a few specific parcels in a single lawsuit) and the task of each federal board (determining the relative average per-acre value of land between all districts across an entire state). The latter was more general and complex by orders of magnitude. Prior to 1798, such statewide mass determinations of relative value had been undertaken only by state legislatures. See infra notes 455-461 and accompanying text; see also Ronald J. Allen & Michael S. Pardo, The Myth of the Law-Fact Distinction, 97 Nw. U. L. Rev. 1769, 1769-70 (2003) (finding that, “in practice,” courts drawing the law/fact distinction tend, apart from relying simply upon convention or pragmatic views of judges’ and juries’ relative competence, to find matters factual if they are more “specific” and “localized” and have less “general import”).
See Sotirios A. Barber, The Constitution and the Delegation of Congressional Power 57-58 (1975) (arguing that the delegated power to lift trade restrictions on a foreign nation if and when that nation “shall cease to violate the neutral commerce of the United States,” upheld as a fact determination in Cargo of the Brig Aurora v. United States (The Aurora), 11 U.S. (7 Cranch) 382 (1813), really called for “a policy choice, not a factual judgment”); Louis L. Jaffe, Judicial Control of Administrative Action 56 (1965) (characterizing the delegated power to retaliate against foreign tariffs that the President “may deem to be reciprocally unequal and unreasonable” and to continue retaliation “for such time as he shall deem just,” upheld as a fact determination in Field v. Clark, 143 U.S. 649 (1892), as really a “wide and uncertain area of judgment” and “not a formula at all but a bargaining power put into the President’s hands”).
Skinner v. Mid-Am. Pipeline Co., 490 U.S. 212, 214 (1989). In this case, the Court treated taxation as typical, not exceptional, for purposes of nondelegation under Article I. The Court has also treated taxation as typical for the purpose of other constitutional doctrines, such as procedural due process under the Fifth and Fourteenth Amendments, evidenced by Bi-Metallic’s status as a leading case. See supra notes 59-60 and accompanying text. That said, the Court has sometimes treated taxation as exceptional, perhaps most notably with regard to Article III’s “judicial power” clause. Under Article III, the Court has distinguished between “public rights” that may be adjudicated by administrators and “private rights” that must be adjudicated by courts, and it has long placed taxation in the “public rights” category, relying upon the long history of administrative tax adjudication, some of which is noted in Part IV infra. See Erwin Chemerinsky, Federal Jurisdiction 248-53 (7th ed. 2016); Caleb Nelson, Adjudication in the Political Branches, 107 Colum. L. Rev. 559, 586-90 (2007). Two scholars have recently suggested that Article III’s distinction between public and private rights could serve as a model for reformulating the Article I nondelegation doctrine—in particular, that privileges and benefits (which take up most of the category of public rights) should be free (or relatively free) from nondelegation constraints, whereas private rights should be fully subject to such constraints. Bamzai, supra note 35, at 178, 180-82; Rappaport, supra note 101, at 4-11. But neither scholar addresses whether taxation’s placement in the “public rights” category for Article III purposes means taxation should also fall in that category for Article I purposes. Apart from the question of whether an exception to the nondelegation doctrine for even noncoercive public rights has an originalist basis, see supra note 51, it seems very strange to think that the Framers, if they thought it more important for some matters to be under legislative control than others, would put taxation on the lesser end of that scale.
Gundy, 139 S. Ct. at 2131 (Gorsuch, J., dissenting) (describing the Act’s requirements); id. at 2144 (characterizing the Attorney General’s discretion to impose such requirements as the power to write “a code of conduct governing private conduct for a half-million people”); cf. Hamburger, supra note 40, at 60 (referring to the “power to tax or otherwise bind subjects”).