Tax transitions, opportunistic retroactivity, and the benefits of government precommitment

Michigan Law Review

Volumn 94, Issue 5, 1996, Pages 1129-1196

  Logue, Kyle D ^a,b,c  

^a UNIVERSITY OF MICHIGAN LAW SCHOOL   (United States)

^b AUBURN UNIVERSITY   (United States)

^c YALE UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0345562962     PISSN: 00262234     EISSN: None     Source Type: Journal    
DOI: 10.2307/1289850     Document Type: Article

References (230)

1. JON ELSTER, ULYSSES AND THE SIRENS: STUDIES IN RATIONALITY AND IRRATIONALITY 10 (1984)
2. (paraphrasing JOSEPH SCHUMPETER, A HISTORY OF ECONOMIC ANALYSIS 87 (1954)).
3. Kenneth A. Shepsle, Discretion, Institutions, and the Problem of Government Commitment, in SOCIAL THEORY FOR A CHANGING SOCIETY 245-46 (Pierre Bourdieu & James S. Coleman eds., 1991).
4. See, e.g., ROBERT E. HALL & ALVIN RABUSHKA, THE FLAT TAX 1 (2d ed. 1995) (quoting Shirley Peterson, former Commissioner of the Internal Revenue Service) ("I would repeal the entire Internal Revenue Code and start over.").
5. Representative Richard Armey, for example, has proposed a flat tax of 17% to be applied only to wages and pension income with an exclusion for investment income. Because of the exclusion for investment income, the Armey flat tax is a form of consumption tax. For the classic description of a consumption-based flat tax upon which the Armey plan is based, see HALL & RUBUSHKA, supra note 3. Another recent proposal for replacing the current income tax with a broad-based consumption tax that has received a great deal of attention is the plan put forward by Senators Sam Nunn and Pete Domenici, called the Unlimited Savings Allowance (USA) Tax System. See H.R. 2060, 104th Cong., 1st Sess. § 101 (1995); S. 722, 104th Cong., 1st Sess. (1995) (dubbing the proposal the "USA Tax Act of 1995"). Under the USA tax system, taxpayers essentially would be allowed to deduct amounts set aside for savings or investment; thus, the USA tax amounts to a cash-flow consumption tax as well. For a general discussion of these and other recent flat-tax-reform proposals, see JOINT COMM. ON TAXATION, 104TH CONG., 1ST SESS., DISCUSSION OF ISSUES RELATING TO FLAT TAX RATE PROPOSALS (Comm. Print 1995).
6. See, e.g., Jerry Heaster, Reform, Via a Flat Tax, Seems Likely, KAN. CITY STAR, May 7, 1995, at G1 (quoting economists and political analysts who view radical tax reform as highly likely in coming years);
7. Sheldon D. Pollack, Consumption Taxes, Flat Taxes, Capital Gains, and Other Tax Fantasies, 66 TAX NOTES 577, 578 (1995) ("[T]here is surprisingly strong support on both sides of the aisle in Congress for both of these very radical tax proposals [cut in the capital gains tax and replacement of the federal income tax], and hence, they must be considered as viable political options.");
8. David E. Rosenbaum, Washington Memo: Debate of Flat Tax Revives Simplicity vs. Fairness Issue, N.Y. TIMES, April 18, 1995, at A1.
9. See, e.g., JOINT COMM. ON TAXATION, supra note 4, at 27-28.
10. For example, the Hall-Rabushka flat tax would eliminate all deductions other than the deduction for the costs of inputs used in the production of goods and services, including the costs of plant and equipment and employee wages and salaries; it also would eliminate all credits and exclusions, other than the personal exemption. See HALL & RABUSHKA, supra note 3, at 110-11. Note, however, that the flat tax, despite its name, would retain a degree of progressivity because of the personal exemption.
11. See, e.g., Alliance USA, Description and Explanation of the Unlimited Savings Allowance Income Tax System, 66 TAX NOTES 1481, 1487 (1995) ("The USA Tax System is designed to replace on a revenue-neutral basis the present corporate and individual income taxes . . . .").
12. Michael J. Graetz, Legal Transitions: The Case of Retroactivity in Income Tax Revision, 126 U. PA. L. REV. 47, 87 (1977)
13. [hereinafter Graetz, Retroactivity].
14. Louis Kaplow, An Economic Analysis of Legal Transitions, 99 HARV. L. REV. 509 (1986).
15. Graetz and Kaplow's positions with respect to tax transitions are discussed in more detail infra in Parts I and III.
16. There have been two noteworthy articles rejecting the Graetz-Kaplow conclusion, at least with respect to some types of tax transitions. See Daniel S. Goldberg, Tax Subsidies: One-Time v. Periodic An Economic Analysis of the Tax Policy Alternatives, 49 TAX L. REV. 305 (1994);
17. J. Mark Ramseyer & Minoru Nakazato, Tax Transitions and the Protection Racket: A Reply to Professors Graetz and Kaplow, 75 VA. L. REV. 1155 (1989). I discuss the Goldberg article infra in section V.B.4 and the Ramseyer-Nakazato article infra in section III.D.
18. See Saul Levmore, The Case for Retroactive Taxation, 22 J. LEGAL STUD. 265, 273-78 (1993). I discuss Levmore's arguments infra in section IV.B.2.
19. Although I sometimes use the language of "efficiency," the normative criterion that I apply, as do Graetz, Kaplow, and others, is a version of utilitarianism. See infra note 40.
20. See Kaplow, supra note 10, at 517; Michael J. McIntyre, Transition Rules: Learning to Live With Tax Reform, 4 TAX NOTES 7, 12 (1976).
21. See Graetz, Retroactivity, supra note 9, at 57-60.
22. For examples of nominally retroactive tax transitions, see infra section IV.A (discussing the Carlton case) and IV.B.2 (discussing retroactive income tax rate increases). Most nominally retroactive tax changes apply only to the year in which the change is enacted. Theoretically, a retroactive tax change also could apply to income earned in tax years prior to the year of enactment. If the tax change applies to earlier years, however, it might indeed run afoul of the Constitution. See infra note 130 (citing sources).
23. Professor Graetz was the first to draw the distinction between effective dates that are nominally prospective and those that are nominally retroactive. More important, he was the first to demonstrate that applying a transition nominally prospectively will not provide full transition relief. See Graetz, Retroactivity, supra note 9. Note, however, that, with respect to certain types of incentive-subsidy provisions, such as an "up-front" incentive subsidy, a nominally prospective effective date will provide almost full transition protection against the risk of a repeal. See infra section V.B.5 (discussing up-front incentive subsidies).
24. See Graetz, Retroactivity, supra note 9, at 57-60.
25. See DAVID F. BRADFORD & U.S. TREASURY TAX POLICY STAFF, BLUEPRINTS FOR BASIC TAX REFORM 166-68 (2d ed. 1984) [hereinafter BLUEPRINTS]; Graetz, supra note 9, at 60.
26. Code § 7805(b) empowers the Treasury Department to decide whether to apply regulations retroactively: "The Secretary may prescribe the extent, if any, to which any ruling or regulation, relating to the internal revenue laws, shall be applied without retroactive effect." I.R.C. § 7805(b) (1996).
27. The Treasury Department's discretion under § 7805(b) has limits. For example, if a regulation has existed unchanged for many years, it may acquire the force of law, in which case retroactive repeal can come only from Congress. See Helvering v. R.J. Reynolds Tobacco Co., 306 U.S. 110 (1939). See generally BORIS I. BITTKER & LAWRENCE LOKKEN, FEDERAL TAXATION OF INCOME, ESTATE, AND GIFTS ¶ 110.4.3 (1992).
28. See Kuhn v. Fairmont Coal Co., 215 U.S. 349, 372 (1910) (Holmes, J., dissenting) ("Judicial decisions have had retrospective operation for near a thousand years."). See generally Harper v. Virginia Dept. of Taxation, 113 S. Ct. 2510 (1993) (surveying the Supreme Court's jurisprudence on retroactive application of constitutional decisions). The primary exception to the rule of retroactive application of judicial decisions comes in the civil context, where a "new principle of law" can be applied prospectively, if prospective application would avoid "injustice and hardship" without unduly undermining the "purpose and effect" of the new rule. See Chevron Oil Co. v. Huson, 404 U.S. 97, 106-07 (1971).
29. See Kaplow, supra note 10, at 515-19.
30. A similar question often has been addressed in connection with nontax legal transitions. For example, much of the literature on the Takings Clause of the U.S. Constitution includes analyses of the question whether the government should provide compensation to citizens whose property it takes for a public purpose. E.g., Lawrence Blume & Daniel L. Rubinfeld, Compensation for Takings: An Economic Analysis, 72 CAL. L. REV. 569 (1984);
31. Lawrence Blume et al., The Taking of Land: When Should Compensation be Paid? 99 Q.J. ECON. 71 (1984);
32. Louis De Alessi, Implications of Property Rights for Government Investment Choices, 59 AM. ECON. REV. 13 (1969);
33. William A. Fischel & Perry Shapiro, Takings, Insurance, and Michelman: Comments on Economic Interpretations of "Just Compensation" Law, 17 J. LEGAL STUD. 269 (1988); Kaplow, supra note 10;
34. Frank Michelman, Property, Utility and Fairness: Comments on the Ethical Foundations of "Just Compensation" Law, 80 HARV. L. REV. 1165 (1967).
35. See Michael J. Graetz, Implementing a Progressive Consumption Tax, 92 HARV. L. REV. 1575, 1650 (1979)
36. [hereinafter Graetz, Consumption Tax] ("The politically dominant approach to significant changes in the tax law has been to protect the expectations of taxpayers who have 'relied' on existing law; protection typically takes the form of 'grandfathered' effective dates.") (citing, among other sources, BLUEPRINTS, supra note 20, at 181-215;
37. and INSTITUTE FOR FISCAL STUDIES, THE STRUCTURE AND REFORM OF DIRECT TAXATION 187-92, 198-200 (1978)).
38. E.g., Committee on Tax Policy, New York State Bar Association, Retroactivity of Tax Legislation, 29 TAX LAW. 21 (1975);
39. Note, Setting Effective Dates for Tax Legislation: A Rule of Prospectivity, 84 HARV. L. REV. 436 (1970).
40. See Levmore, supra note 13, at 267-68.
41. Other particularly cogent and influential analyses of tax transitions include BLUEPRLNTS, supra note 20, at 159-87; and McIntyre, supra note 15.
42. Kaplow concludes that, under certain assumptions, all forms of tax-transition relief are generally undesirable. See Kaplow, supra note 10, at 615. Graetz, however, would permit the occasional use of phased-in or delayed effective dates, depending upon the magnitude of the wealth loss the transition causes. See Graetz, Retroactivity, supra note 9, at 87. Another distinctive part of Kaplow's treatment of legal transitions is its comprehensiveness. Not only does he address transition losses that result from any type of legal transition (not just tax transitions), but he also addresses the symmetrical issue of transition gains (or "windfall gains") that result when taxpayers see the value of the pretransition investment rise as a result of the legal transition. Kaplow correctly notes that all the same issues apply to windfall gains. See Kaplow, supra note 10, at 553-55.
43. See generally Graetz, Retroactivity, supra note 9, at 65-66, 87; Kaplow, supra note 10, at 513-14, 520, 527-36.
44. Kaplow, supra note 10, at 513-14.
45. Id. at 615-16.
46. Graetz, Retroactivity, supra note 9, at 87.
47. In fairness, Graetz and Kaplow both acknowledge a number of possible exceptions to their general arguments against transition relief. See Graetz, Retroactivity, supra note 9, at 87; Kaplow, supra note 10, at 616 & n.336.
48. It is also worth noting that other scholars have come to view the Graetz-Kaplow analysis as standing for the rejection of grandfathered effective dates in tax transitions. See, e.g., Ramseyer & Nakazato, supra note 12, at 1155 ("[Graetz and Kaplow] argue that Congress should abandon [the use of grandfather clauses] as economically inefficient and ethically superfluous.").
49. See Graetz, Retroactivity, supra note 9, at 87;
50. Graetz, Consumption Tax, supra note 26, at 1650 ("[N]either fairness nor efficiency demands grandfathered effective dates, but . . . when the magnitude of change is large, its impact should be reduced through delayed or phased-in effective dates rather than grandfathering."); id. at 1653 ("[G]randfathered effective dates should not be enacted to protect assets that have received favored treatment under the income tax.").
51. See Kaplow, supra note 10, at 551-52.
52. See Levmore, supra note 13, at 268.
53. See id.; Ramseyer & Nakazato, supra note 12, at 1155-56. It is worth noting that, following Graetz's 1977 article, the Tax Section of the ABA issued a report adopting some of Graetz's recommendations. Special Comm. on Simplification, ABA Section of Taxation, Evaluation of the Proposed Model Comprehensive Income Tax, 32 TAX LAW. 563, 680-86 (1979).
54. The normative criterion applied through most of the article is a loose version of the Kaldor-Hicks variety, in other words, maximizing the size of the pie. Thus, typically when I say "efficient" or "optimal," I mean Kaldor-Hicks efficiency or optimality. At points in the article, however, I also address distributional concerns and how they may conflict with efficiency concerns.
55. See infra notes 70-78.
56. See Kaplow, supra note 10, at 587 (coining the term "up front" subsidy).
57. Goldberg explores the distinction between up-front subsidies and installment subsidies in considerable detail. Although he uses different terminology ("one-time" subsidy and "periodic" subsidy, respectively) the concepts are the same as those discussed by Kaplow and by me. Goldberg, supra note 12.
58. See id. at 314; Kaplow, supra note 10, at 587.
59. Interestingly, Professor Kaplow concludes generally that the government should not commit to any substantive tax rule. He also concludes, however, that the government should commit itself to maintain an optimal transition policy, which he distinguishes from substantive tax rules. Therefore, Kaplow recognizes that there are efficiency benefits to the government's precommitment to transition policy; however, he concludes that the efficient transition policy is for the government not to provide transition relief. See Kaplow, supra note 10, at 557-60.
60. See infra section III.B (discussing the distinction between opportunistic and nonopportunistic legal transitions).
61. Graetz acknowledges the potential efficiency benefits of reducing the default premium (which he calls an "uncertainty premium") that would flow from the government's precommitment to the transition policy I have described. See Graetz, Retroactivity, supra note 9, at 69-70.
62. He concludes, however, that a commitment by the government to provide grandfather treatment would be inefficient; further, he concludes that transition relief should be supplied only on an ad hoc basis, depending upon the magnitude of the transition loss in question. See id. at 87. Kaplow, on the other hand, argues that it is always cheaper for the government to pay the default premium than to commit to grandfathering an incentive-subsidy repeal. See Kaplow, supra note 10, at 528-29. This argument seems clearly wrong, for the reasons discussed in the text. Goldberg acknowledges the default-premium effect, and, like me, he regards it as a reason for providing transition relief in this context. See Goldberg, supra note 12, at 327. He does not explain why it would be efficient to have the government rather than the individual taxpayers bear the default premium. Indeed, Golberg's argument seems to be based primarily on an unspecified conception of fairness or horizontal equity. See id. at 323 ("If the subsidy is removed, however, transition rules should be enacted to prevent inequities . . . ."). In addition, Goldberg does not address the issue of government precommitment, and thus he does not suggest proposals for facilitating government precommitment to an optimal transition policy. Finally, Goldberg ignores the boundary between incentive subsidies and other types of tax provisions and the desirability of transition relief with respect to transitions in the latter type of provision. Graetz also provides a Rawlsian analysis of what would constitute a "fair" transition policy from the perspective of the "original position;" he concludes that, "[i]n the context of a progressive income tax, the repeal of tax-favored treatment for particular investments would be the least likely category of change for which the parties would agree to require grandfather clauses." Graetz, Retroactivity, supra note 9, at 85
63. (quoting JOHN RAWLS, A THEORY OF JUSTICE 199 (1971)). His argument is that the parties most likely to benefit from a rule requiring grandfathering of such provisions are likely to be the "most-advantaged persons" in society rather than the least. See id. at 86. But this conclusion does not follow. Which class of individuals will benefit most from such a transition policy depends upon how the government spends the money it saves in reduced default premiums resulting from the policy. That money could be distributed to the least-advantaged individuals in society, for example.
64. See Graetz, Retroactivity, supra note 9, at 70.
65. See Michelman, supra note 25, at 1214-18 (discussing the concept of "demoralization costs"). For Graetz's response to Michelman's demoralization cost story, see Graetz, Retroactivity, supra note 9, at 72 n.77.
66. See MICHAEL J. GRAETZ, FEDERAL INCOME TAXATION: PRINCIPLES AND POLICIES 92 (2d ed. 1988). There is evidence suggesting that taxpayers are less willing to comply with the tax laws if they perceive the system to be unfair or inconsistent.
67. See Steven M. Sheffrin & Robert K. Triest, Can Brute Deterrence Backfire? Perceptions and Attitudes in Taxpayer Compliance, in WHY PEOPLE PAY TAXES: TAX COMPLIANCE AND ENFORCEMENT 193 (Joel Slemrod ed., 1992). Goldberg makes a similar point. See Goldberg, supra note 12, at 329.
68. See infra Part V.
69. There is some psychological evidence to support the notion that humans and other animals have a tendency to make decisions that maximize their short-run utility at the expense of their total utility. See George Ainslie, Beyond microeconomics: Conflict among interests in a multiple self as a determinant of value, in THE MULTIPLE SELF: STUDIES IN RATIONALITY AND SOCIAL CHANGE 133 (Jon Elster ed., 1986);
70. Frank A. Logan, Decisionmaking by Rats: Delay Versus Amount of Reward, 59 J. OF COMP. & PHYSIOLOGICAL PSYCHOL. 1 (1965). This tendency can be understood as a weakness of will or as the application of an inappropriate discount rate.
71. See generally ELSTER, supra note 1, at 87-103.
72. The benefit of government precommitment is a common theme in constitutional theory. See, e.g., LAURENCE TRIBE, AMERICAN CONSTITUTIONAL LAW § 9-7, at 470, 473 (2d ed. 1988) (arguing that the Contract Clause can be understood as a means of precommitting the government);
73. Stephen Holmes, Precommitment and the Paradox of Democracy, in CONSTITUTIONALISM AND DEMOCRACY (Jon Elster & Rune Slagstad eds., 1988) (reviewing the intellectual history of the government-precommitment problem in political theory).
74. In recent years, the benefits of government precommitment and the harms of failing to honor such commitments have been formalized by both political theorists and macroeconomic theorists. These models attempt to quantify the theoretical efficiency gains associated with the government's ability to make binding commitments. The classic article in this area, which gave rise to the literature on the "time-consistency" problem, is Finn E. Kydland & Edward C. Prescott, Rules Rather than Discretion: The Inconsistency of Optimal Plans, 85 J. POL. ECON. 473 (1977).
75. The default-premium problem discussed in the text is a version of this time-consistency problem. In developing the time-consistency model, macroeconomists have borrowed from two literatures (game theory and rational-choice theory) that for years have been analyzing implicitly or explicitly the benefits of precommitment by individual and institutional actors. See, e.g., ELSTER, supra note 1, at 65-76; THOMAS SCHELLING, THE STRATEGY OF CONFLICT (1960);
76. Robert H. Strotz, Myopia and Inconsistency in Dynamic Utility Maximization, 23 REV. ECON. STUD. 165 (1955).
77. This is sometimes called "regulation."
78. This is sometimes called "eminent domain."
79. Cost-benefit analyses of these policy options are common. See, e.g., Richard Layard & Stephen Glaister, Introduction to COST-BENEFIT ANALYSIS 1, 3 (Richard Layard et al. eds., 2d ed. 1994) ("Three main methods of [government] intervention are . . . regulation, taxes and subsidies, and public direction of what is to be produced, be it via public enterprise or purchase from private firms. Each of these types of government activity can be subject to cost-benefit analysis. . . ."). Some commentators argue, however, that the choice among such policy options cannot be captured fully in a traditional cost-benefit analysis but instead require sensitivity to the "cultural" consequences or the "expressive effect" associated with each different policy option.
80. E.g., Richard H. Pildes, The Unintended Cultural Consequences of Public Policy: A Comment on the Symposium, 89 MICH. L. REV. 936, 940-41 (1991).
81. Compensated government coercion might be efficient, for example, during times of military or medical emergency, when there isn't time for options three, four, or five to be implemented effectively.
82. See Ronald H. Coase, The Nature of the Firm, in 4 ECONOMICA 386 (1937).
83. See Stanley S. Surrey, Tax Incentives as a Device for Implementing Government Policy: A Comparison with Direct Government Expenditures, 83 HARV. L. REV. 705, 715-19 (1970).
84. Surrey argues that, when it becomes desirable for the government to provide incentive subsidies (which he concedes happens), such subsidies should be provided through direct government expenditures rather than through the federal income tax laws in the form of so-called tax expenditures. Surrey's arguments against the use of the income tax law to alter behavior include the following: (a) tax incentives (at least those that take the form of deductions or credits) disproportionately benefit taxpayers with relatively high incomes and (b) by having the expenditures funneled through the tax-writing process rather than the normal expenditure process, we run the risk of implementing conflicting policies and of unnecessarily complicating the tax code. On the other side of that debate, however, some have argued that tax subsidies may be superior to direct subsidies because the former requires less bureaucracy than the latter. See, e.g., Edward A. Zelinsky, Efficiency and Income Taxes: The Rehabilitation of Tax Incentives, 64 TEXAS L. REV. 973, 1010-12 (1986). That particular debate is beyond the scope of this article, and, for now, I remain agnostic on the question whether incentive subsidies are provided best through tax expenditures or direct expenditures. The framework of this article applies in either case to any type of incentive subsidy, whether it be a tax expenditure or direct subsidy.
85. According to standard contract theory, the prevention of opportunistic behavior is one of the principal functions of contracts and contract law. See RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 81 (3d ed. 1986) ("Thus the fundamental function of contract law (and recognized as such at least since Hobbes's day) is to deter people from behaving opportunistically toward their contracting parties, in order to encourage the optimal timing of economic activity and make costly self-protective measures unnecessary." (footnote omitted));
86. see also John H. Barton, The Economic Basis of Damages for Breach of Contract, 1 J. LEGAL STUD. 277 (1972) (demonstrating how a legally enforceable contract permits parties to overcome the prisoner's dilemma and to achieve joint-wealth-increasing cooperation). Professor Oliver Williamson has written extensively about how markets can respond to overcome the possibility of opportunistic behavior.
87. See OLIVER E. WILLIAMSON, MARKETS AND HIERARCHIES: ANALYSIS AND ANTITRUST IMPLICATIONS (1975);
88. Oliver E. Williamson, Credible Commitments: Using Hostages to Support Exchange, 73 AM. ECON. REV. 519 (1983).
89. See, e.g., Bowen v. Public Agencies Opposed to Social Sec. Entrapment, 477 U.S. 41, 52 (1986) ("[T]he Federal Government, as sovereign, has the power to enter contracts that confer vested rights, and the concomitant duty to honor those rights . . . ." (citations omitted)); Perry v. United States, 294 U.S. 330, 351 (1935) ("To say that the Congress may withdraw or ignore [its] pledge, is to assume that the Constitution contemplates a vain promise, a pledge having no other sanction than the pleasure and convenience of the pledgor. This Court has given no sanction to such a conception of the obligations of our Government."); Lynch v. United States, 292 U.S. 571, 580 (1934) ("Congress was free to reduce gratuities deemed excessive. But Congress was without power to reduce expenditures by abrogating contractual obligations of the United States. To abrogate contracts, in the attempt to lessen government expenditure, would be not the practice of economy, but an act of repudiation."). Note, however, that Congress imposes a clear-statement restriction on its ability to bind itself into the future. See Bowen, 477 U.S. at 52 ("[W]e have declined in the context of commercial contracts to find that a 'sovereign forever waives the right to exercise one of its sovereign powers unless it expressly reserves the right to exercise that power in' the contract." (quoting Merrion v. Jicarilla Apache Tribe, 455 U.S. 130, 148 (1982))). This is sometimes called the "unmistakability doctrine." The remedy against the federal government for breach of contract has developed in stages. Prior to 1855, a citizen with a monetary claim against the United States could seek redress only through a private bill introduced in Congress. See Jeremy Travis, Note, Rethinking Sovereign Immunity after Bivens, 57 N.Y.U. L. REV. 597, 642 (1982).
90. In 1855, Congress created the Court of Claims and empowered it to hear claims against the United States founded upon any act of Congress or regulation of an executive department or upon any express or implied contract with the United States. See Act of Mar. 3, 1887, ch. 359, §§ 1, 2, 5, 24 Stat. 505, 506. Through the Tucker Act, 28 U.S.C. §§ 1346, 1491 (1994), Congress responded to the inadequacies of the original Court of Claims legislation. The Act added to that court's jurisdiction claims founded upon the Constitution and claims for liquidated or unliquidated damages not sounding in tort. In addition, district courts were given concurrent jurisdiction of any claims under the statute not exceeding $10,000. See 14 CHARLES A. WRIGHT ET AL., FEDERAL PRACTICE AND PROCEDURE § 3657 (1976). Federal law determines liability under the statute. See Roxfort Holding Co. v. United States, 176 F. Supp. 587, 589 (D.N.J. 1959) ("[T]he matter of liability of the federal government is to be determined by federal law.").
91. The Supreme Court addressed the issue of states' immunity from suit early in the development of American jurisprudence, deciding that the Contracts Clause, U.S. Const. art. I, § 10, cl. 1, could defeat a state's claim of sovereign immunity. See Dartmouth College v. Woodward, 17 U.S. (4 Wheat.) 518 (1819); Fletcher v. Peck, 10 U.S. (6 Cranch) 87 (1810). For a variety of reasons, the Contracts Clause fell into disuse until 1977. See United States Trust Co. v. New Jersey, 431 U.S. 1 (1977). See generally, TRIBE, supra note 54, at 613-28 (summarizing the history of Contracts-Clause jurisprudence).
92. Put in these terms, the notion of eliminating the government's discretion to break its contracts seems to be in tension with the general conclusion suggested by the Graetz-Kaplow analysis, the conclusion that losses caused by legal transitions, in most cases, should not be compensated by the government. As discussed more fully in the following section, however, in the case of the government's opportunistic breach of its own contracts, we have identified a type of government risk with respect to which the government is unambiguously the least-cost insurer and a type of transition loss that on efficiency grounds should be compensated. But see Kaplow, supra note 10, at n.54 (acknowledging that the government should pay for goods and services it uses but failing to generalize the point to all government contracts or to other legal transitions).
93. Note that an incentive subsidy has some of the characteristics of an old common law unilateral contract. See Carlill v. Carbolic Smoke Ball Co., 2 Q.B. 484 (1892).
94. Several previous analyses of tax transitions have recommended a similar transition policy with respect to incentive-subsidy transitions; however, none of them justifies this position by using the sort of efficiency argument that I offer. See BLUEPRINTS, supra note 20, at 176-77; McIntyre, supra note 15, at 13. In addition, as explained below, there are some important differences between the incentive-subsidy transition policy that those authors advance and the one that I advance.
95. Dani Rodrik & Richard Zeckhauser, The Dilemma of Government Responsiveness, 7 J. POLY. ANALYSIS & MGMT. 601, 602 (1988) (footnote omitted).
96. For a game-theoretic discussion of precisely this type of government opportunism in the context of utility regulation, see Glenn Blackmon & Richard Zeckhauser, Fragile Commitment and the Regulatory Process, 9 YALE J. ON REG. 73 (1992).
97. See Blackmon & Zeckhauser, supra note 67. In future research, I plan to analyze a number of specific cases of government opportunism in contexts other than federal income tax law and to emphasize the importance of taking into account default-premium effects. For example, one prominent instance of government opportunism involves the circumstances surrounding California's Proposition 103, the referendum passed in the fall of 1988 that ordered the rollback in liability and certain other insurance rates in California. This referendum generated an enormous amount of litigation and scholarly attention. See, e.g., Calfarm Ins. Co. v. Deukmejian, 771 P.2d 1247 (Cal. 1989); Steven D. Sugarman, California's Insurance Regulation Revolution: The First Two Years of Proposition 103, 27 SAN DIEGO L. REV. 683 (1990). One could plausibly argue that the insurance companies who were induced to enter the California insurance markets in reliance on the assumption that the prevailing regulatory structure would remain in force (a structure that allowed the insurers to charge premiums that cover their costs) were victims of opportunism at the hands of the State of California as a result of the mandatory rate rollbacks. My concern is that the California voters, in making such a radical change in their insurance-regulatory regime, may have failed to consider adequately the default-premium effects. To be specific, query the extent to which California voters understood and took into account the possibility that, following the rate rollbacks, it would become much more difficult to get insurers to write business in California.
98. Keep in mind, however, that, with respect to all of these examples (including incentive subsidies and government contracts) if it can be shown that the default-premium effect is relatively small, the efficient transition policy may be to provide no transition relief: no grandfather treatment, no contract damages, nothing.
99. I.R.C. § 42 (1996).
100. I.R.C. § 51 (1996).
101. I.R.C. § 43 (1996).
102. I.R.C. § 44 (1996).
103. I.R.C. § 45 (1996).
104. I.R.C. § 45A (1996).
105. I.R.C. § 1396 (1996).
106. I.R.C. § 41 (1996).
107. I.R.C. § 40 (1996). These credits are combined into a general-business credit for the purpose of computing how much of each credit will be allowed in a given year and in carryback and carryover years. See I.R.C §§ 38(b)-39 (1996).
108. The [investment] credit was first enacted in 1962 in connection with that year's liberalization of depreciation rules to stimulate purchases of new machinery and equipment to bolster a lagging economy. Since 1962, the credit has been utilized as an instrument of fiscal policy. The investment credit was increased in 1964, removed in 1966, reinstated in 1967, repealed in 1969, reinstituted in 1971, temporarily increased in 1975, "permanently increased" in 1978, reduced in 1982 and repealed once again in 1986. Although the repeal of the investment tax credit was used in 1986 to finance a large share of the revenue cost of reducing the top corporate tax rate from 46 to 34 percent, its history suggests that it would be foolish to regard the investment tax credit as a dead letter for the future. GRAETZ, supra note 50, at 407.
109. Taxpayers have long been able to take depreciation deductions to account for the "exhaustion, wear and tear (including a reasonable allowance for obsolescence)" of assets used in a trade or business or held for the production of income. I.R.C. § 167 (1996). In 1981, Congress enacted the Accelerated Cost Recovery System (ACRS), which replaced the pre-1981 depreciation rules with respect to most tangible property. The accelerated deductions permitted under ACRS had the effect of increasing the after-tax rate of return on investment in qualifying depreciable property, thereby giving taxpayers an incentive to increase their level of investment in such assets. In fact, when the Tax Reform of 1986 replaced the ACRS with the Modified Accelerated Cost Recovery System (MACRS), one of its principal motivations was to reduce the effect on investment incentives. See 2 U.S. TREASURY DEPT., REPORT TO THE PRESIDENT, TAX REFORM FOR FAIRNESS, SIMPLICITY AND ECONOMIC GROWTH 154-57 (1984) ("The low or negative effective tax rates on ACRS property and the tax deferral resulting from accelerated depreciation allowances distort investment decisions in a variety of ways."). See generally GRAETZ, supra note 50, at 392 ("[D]epreciation allowances have long been used to adjust the overall level of investment in plant and equipment for fiscal policy reasons.").
110. See I.R.C. § 179 (1996). A limit is placed on the total amount a taxpayer can expense under § 179 in a given tax year. The current limit is $17,500. See I.R.C. § 179 (b)(1) (1996).
111. See I.R.C. §§ 611-14 (1996).
112. See I.R.C. § 169 (1996).
113. See I.R.C. § 174(b) (1996).
114. See I.R.C. § 194 (1996).
115. See I.R.C. §§ 163(a), (h)(3) (1996).
116. See I.R.C. § 170 (1996).
117. See I.R.C. § 103 (1996).
118. For an extensive list of incentive-subsidy provisions, see generally JOINT COMM. ON TAXATION, 103D CONG., 2D SESS., ESTIMATES OF TAX EXPENDITURES FOR FISCAL YEARS 1995-99 (Comm. Print 1994).
119. It is still true, however, that some of the provisions I have listed as incentive tax subsidies might be better characterized differently. For example, the home-mortgage-interest deduction may be understood best, not as being designed to induce higher levels of home ownership, but rather as being designed to reduce the effective tax rate on middle- and high-income taxpayers. If the latter characterization is more accurate, and the home-mortgage interest deduction is not primarily an incentive subsidy, we may be less concerned about default-premium effects in the event of a repeal. In my view, this deduction (more so than most provisions) has substantial incentive and distributional effects. Interestingly, these two effects are importantly related; they pull in opposite directions. If the deduction has a large incentive effect (causing more taxpayers to buy homes) the distributional effect may be diminished because of the putative tax that results when the taxinduced incentive increase in demand for housing produces higher housing prices.
120. Kaplow, supra note 10, at 535.
121. See id. at 587 ("[Tax subsidies] typically should not be designed so as to incorporate grandfathering, because it generally is undesirable.").
122. Id. at 551.
123. Graetz, Consumption Tax, supra note 26, at 1651.
124. Below I discuss one type of transition (the mistake case) where the government-risk argument, on incentive grounds, may call for no transition relief.
125. See POSNER, supra note 61, at 79-81.
126. Kaplow, supra note 10, at 529 (footnotes omitted) (emphasis added); see also id. at 551, 573 (drawing a distinction between mistake cases, where nominal retroactivity is appropriate, and change-of-circumstance cases, where nominal prospectivity is appropriate).
127. The allocation of the risk of a government mistake presents issues similar to those that gave rise to the contract doctrines of mistake, impossibility, and frustration. See FRIEDRICH KESSLER ET AL., CONTRACTS ch. 8 (1986).
128. See Kaplow, supra note 10, at 536-50.
129. See id.
130. That the government generally must comply with its contracts does not mean that one legislature has unlimited power to bind future legislatures. For example, the Supreme Court has long held that a current state legislature cannot contract to constrain a future legislature's police power or power of eminent domain. See Stone v. Mississippi, 101 U.S. 814 (1880). States, however, frequently enter into contracts restricting their future taxing power. These are called tax abatements, and they are a common method by which states compete for corporate relocations. See Andy Zipser, Civil War, Round Two, BARRON'S, April 3, 1995, at 23 (discussing the growing use of tax abatements by states to lure corporations into the state).
131. See Kaplow, supra note 10, at 573.
132. Graetz, Retroactivity, supra note 9, at 82.
133. Still, Graetz's general point regarding redistributive tax transitions is well taken, and it is an important point for those incentive-subsidy transitions that affect large classes of taxpayers and that begin to have the characteristics of a nonsubsidy tax transition, such as rate increases. See infra Part IV and supra note 9.
134. Graetz, Consumption Tax, supra note 26, at 1651.
135. See generally KESSLER ET AL., supra note 98, at ch. 11 (reviewing the law of third-party beneficiaries). It is worth noting that in recent years courts have been increasingly willing to allow claims brought by parties alleging to be third-party beneficiaries of government contracts. See id. at 1384-418; see also Anthony Jon Waters, The Property in the Promise: A Study of the Third Party Beneficiary Rule, 98 HARV. L. REV. 1109 (1985) (connecting the development of third-party-beneficiary doctrine in the context of government contracts to the rise of "new property" concepts pioneered by Charles Reich).
136. Graetz, Retroactivity, supra note 9, at 78.
137. See KESSLER ET AL., supra note 98, at ch. 5.
138. Ramseyer and Nakazato, it turns out, favor guaranteed grandfather treatment for incentive-subsidy transitions but for reasons different from those argued here. They rely on the following theory: that a policy of promising grandfathering to those who rely on incentive-subsidy provisions would reduce the amount of social resources spent lobbying against the repeal of tax subsidies in situations in which such a repeal would be efficient. See Ramseyer & Nakazato, supra note 12, at 1171-72; id. at 1174 ("Under a tax-guaranteed regime, investors will not care whether Congress revokes their tax-favored status; under a taxcontingent strategy, they will care dearly. Accordingly, under the former they will lobby and bribe less than under the latter, and to the extent that happens, society gains."). In this article, I ignore the effect of alternative transition policies on lobbying expenditures. Ramseyer and Nakazato's arguments with respect to that question, though, seem facially plausible and therefore provide another justification for a credible government commitment to grandfather any changes to an incentive-subsidy provision.
139. Id. at 1167.
140. The distinction that I draw between obvious legislative error (discussed in this section) and government mistake (discussed in section III.B above) is akin to the distinction in contract law between the treatment of obvious typographical errors in written contracts and the treatment of unilateral mistakes.
141. McIntyre suggests something similar to this policy: "When Congress amends the Code to correct what is generally regarded as an unintended defect in the statute, no special transition rules are justified." McIntyre, supra note 15, at 13. The transition policy that I would apply to corrections of obvious errors, in contrast, would entail a special transition rule. It would apply the correction nominally retroactively. Also, McIntyre does not discuss the difficulty of distinguishing transitions that are corrections of obvious errors from those that are (to use my term) "opportunistic." Kaplow contends that arguments for an obvious-error exception of the sort I have described serve merely to illustrate the benefits of his proposed rule: nominal retroactivity for all transitions. See Kaplow, supra note 10, at 608-09. In my view, the obvious-error exception is understood best as an exception, not as a rule.
142. Note that the Code already contains specific penalty provisions designed to induce taxpayers to interpret the tax rules reasonably and in good faith. For example, the tax-return-accuracy penalty, found in § 6662, imposes a 20% penalty on any portion of an underpayment of tax that is attributable to, among other things, negligence or disregard of rules or regulations or a substantial understatement of income tax liability (in the absence of "substantial authority" for the taxpayer's position). See I.R.C. § 6662(b) (1996). Note, however, that no accuracy-related penalty is imposed under this section with respect to any portion of the underpayment for which the taxpayer can establish that she had a "reasonable cause" and that she acted in good faith. See I.R.C. § 6664(c) (1996). Thus, the accuracy penalties would not apply in the context of an obvious legislative error, as I have defined that term. That is, when Congress has unintentionally but unambiguously enacted a loophole, there is no question that the taxpayer has reasonable cause for her position, as that term is understood by the Service and by the courts. In this section, however, I am suggesting that, although taxpayers who take aggressive positions in reliance on obvious legislative errors do not risk accuracy-related penalties, if such provisions are then repealed by Congress, the repeal should be made nominally retroactive.
143. See STAFF OF JOINT COMM. ON TAXATION, 99TH CONG., 2D SESS., TAX REFORM PROPOSALS: TAX TREATMENT OF EMPLOYEE STOCK OWNERSHIP PLANS (ESOPs) 37 (Comm. Print 1985). See generally United States v. Carlton, 114 S. Ct. 2018 (1994).
144. Tax Reform Act of 1986, Pub. L. No. 99-514, 100 Stat. 2085, 2514, amended by Omnibus Budget Reconciliation Act of 1987, Pub. L. No. 100-203, § 10411(a), 101 Stat. 1330, 1330-433 (repealed 1989).
145. See 133 CONG. REC. 4145, 4293 (1987). Senator Bentsen stated that "Congress did not intend for estates to be able to claim the deduction by virtue of purchasing stock in the market and simply reselling the stock to an ESOP . . . and Congress certainly did not anticipate a $7 billion revenue loss." Id. at 4294.
146. See Omnibus Budget Reconciliation Act of 1987, Pub. L. No. 100-203, § 10411(a), 101 Stat. 1330, 1330-432 (repealed 1989).
147. See Omnibus Budget Reconciliation Act of 1987, Pub. L. No. 100-203, § 10411(b), 101 Stat. 1330 (repealed 1989).
148. I.R.S. Notice 87-13, 1987-1 C.B. 432, 442.
149. 114 S. Ct. 2018 (1994).
150. The purchase and sale of the stock produced a loss, before taxes, of $631,000. However, applying the ESOP estate-tax deduction, the sale generated tax savings of $2,501,161. See 114 S. Ct. at 2021.
151. See 114 S. Ct. at 2020.
152. See infra note 131 (discussing retroactive-tax cases).
153. Congress acted to correct what it reasonably viewed as a mistake in the original 1986 provision that would have created a significant and unanticipated revenue loss. There is no plausible contention that Congress acted with an improper motive, as by targeting estate representatives such as Carlton after deliberately inducing them to engage in ESOP transactions. Carlton, 114 S. Ct. at 2023. In her concurring opinion, Justice O'Connor criticizes the majority opinion for relying on the argument that the amendment clearly was intended to correct a legislative "mistake." She rightly observes that, at some level, "[e]very law touching on an area in which Congress has previously legislated can be said to serve the legislative purpose of fixing a perceived problem with the prior state of affairs." 114 S. Ct. at 2025 (O'Connor, J., concurring). This argument seems to deny the difference between the repeal of an intended incentive subsidy and the repeal of a provision that has entirely unintended and unexpected incentive effects on taxpayers. Obviously, I think there is a difference. Moreover, it may be that the Court's jurisprudence on retroactive taxation would recognize such a difference as well. The majority's opinion at least suggests that the nominally retroactive repeal of an explicit incentive provision would have a greater chance of violating the due-process standard in this context. See 114 S. Ct. at 2023.
154. See 114 S. Ct. at 2023.
155. Pub. L. No. 103-66, § 13001, 107 Stat. 416 (codified at I.R.C. § 1 (1996)).
156. Pub. L. No. 103-66, § 13201(a), 107 Stat. 312 (codified at I.R.C. § 1 (1996)). Before the RRA, the top marginal rate, after all "phase outs" and "bubble" effects, had been 31%. The RRA in effect added two new marginal rates for individuals: 36% (applied to income between $140,000 and $250,000 for married couples filing jointly) and 39.6% (applied to income over $250,000 for married couples filing jointly). See I.R.C. § 1(a) (1996). The new 36% and 39.6% rates have been adjusted for inflation for taxable years since 1994. The RRA also added a new top bracket for corporations: 35% for income over $10 million. See I.R.C. § 11(b)(1)(D) (1996).
157. See Revenue Reconciliation Act of 1993, Pub. L. No. 103-66, § 13201(c), 107 Stat. 416, 459 (codified at I.R.C. § 1 (1996)).
158. See, e.g., 139 CONG. REC. H6392 (daily ed. Sept. 8, 1993) (statement of Rep. Ramstad) ("Mr. Speaker, we have all heard the axiom, 'At least we are safe as long as the legislature isn't in session.' But it turns out that the American people are never safe. With passage of the last tax bill, Congress rolled back the clock to a time it was not even in session, to raise taxes retroactively. For the first time in American history taxes were raised retroactive to a previous administration."); Repeal Retroactivity, WALL ST. J., Aug. 10, 1993, at A14.
159. The Supreme Court repeatedly has upheld retroactive tax legislation against constitutional challenges. See, e.g., United States v. Hemme, 476 U.S. 558 (1986); United States v. Darusmont, 449 U.S. 292 (1981); Welch v. Henry, 305 U.S. 134 (1938); United States v. Hudson, 299 U.S. 498 (1937); Milliken v. United States, 283 U.S. 15 (1931). According to the Supreme Court's most recent pronouncement on retroactive federal income taxation, to survive a due process challenge Congress's decision to apply a tax change retroactively need only pass the "rational-basis" test. See United States v. Carlton, 114 S. Ct. 2018 (1994); see also infra section IV.B.1 (describing earlier retroactive rate increases).
160. See, e.g., Stephen C. Glazier, Tax Bill: Retroactive, Unconstitutional . . ., WALL ST. J., Aug. 5, 1993, at A12 (asserting that a retroactive rate change was unconstitutional).
161. See Levmore, supra note 13, at 265 ("[R]etroactive taxation . . . is generally regarded as abhorrent, unwise, and even illegal."). We are unlikely to see any additional retroactive rate increases from Congress in the near future. In January 1995, the House of Representatives adopted a change to its rules that prevents consideration of a "bill, joint resolution, amendment, or conference report carrying a retroactive Federal income tax rate increase." H. Res. 6, 104th Cong., 1st Sess. § 106(b) (1995). Before the adoption of that new House rule, a number of proposals were introduced that similarly would have limited Congress's ability to change tax provisions retroactively. See, e.g., H.R. 3024, 103d Cong., 1st Sess. (1993) (proposing the elimination of the retroactive tax increases in the RRA); H.R.J. Res. 258, 103d Cong., 1st Sess. (1993) (proposing an amendment to the Constitution outlawing retroactive tax increases); H.R.J. Res. 256, 103d Cong., 1st Sess. (1993) (proposing an amendment to the Constitution to prohibit federal laws from imposing liability for conduct arising before enactment date); H.R. Res. 247, 103d Cong., 1st Sess. (1993) (establishing a point of order under House rules against consideration of measures that contain retroactive tax increases). The relevant language of Resolution 258 reads as follows: Section 1. In the case of any provision of law which modifies the tax laws of the United States and which results in increased revenues to the United States - (1) no such modification of any income tax shall apply to any taxable year beginning before the date of the enactment of such modification, and (2) no such modification of any excise tax shall apply to any event occurring before the date of the enactment of such modification. H.R.J. Res. 258, 103d Cong., 1st Sess. 2 (1993).
162. See Revenue Act of 1938, Pub. L. No. 75-554, 52 Stat. 447 (superceded by I.R.C. 1939).
163. See Individual Income Tax Act of 1944, Pub. L. No. 78-315, 58 Stat. 231 (superceded by I.R.C. 1954). These rates do not include the small surtax on "normal income."
164. Both Graetz and Kaplow make similar observations regarding tax transitions that are motivated by distributional concerns. See Graetz, Retroactivity, supra note 9, at 82; Kaplow, supra note 10, at 519.
165. In his 1993 State of the Union Address, President Clinton justified the 1993 rate increase for high-earning individuals and corporations by emphasizing the need to distribute the burden of paying off the deficit fairly across all taxpayers. See Ruth Marcus & Ann Devroy, Asking Americans to "Face Facts," Clinton Presents Plan to Raise Taxes, Cut Deficit, WASH. POST, Feb. 18, 1993, at A1.
166. Note also that a tax-rate increase can, under certain assumptions, increase rather than decrease taxpayers' work incentives. This would be true if the "income effect" of the rate increase overwhelmed the "substitution effect." See JOSEPH A. PECHMAN, FEDERAL TAX POLICY 76 (5th ed. 1987). There is no a priori basis for deciding how the individual income tax affects work incentives. On the one hand, the tax reduces the financial rewards of greater effort and thus tends to discourage work (the substitution effect). On the other hand, it may provide a greater incentive to obtain more income because it reduces the income left for spending (the income effect). Id. Taking both of these effects into consideration, however, it is quite possible that changes in federal income tax rates have little net effect on taxpayers' work incentive because the decisions whether to work, how much to work, and what career to pursue are influenced primarily by factors other than taxes. Taxation is only one of many factors affecting work incentives. This makes it extremely difficult to interpret the available statistical evidence or the results of direct interviews with taxpayers. The evidence suggests that income taxation does not greatly reduce the amount of labor supplied by workers and managers who are the primary family earners. Work habits are not easily changed, and for most people . . . there is little opportunity to vary their hours of work or the intensity of their efforts in response to changes in tax rates. Id. Pechman also notes that "secondary earners" in a household have greater opportunity to vary their work effort in response to tax-rate changes than do "primary earners." He concludes, however, that, although the evidence is mixed, "[t]he historical trends in the U.S. labor supply do not . . . support the view that taxes have had a significant effect on aggregate labor supply." Id. at 77. Some investment decisions other than the work-leisure tradeoff probably are influenced significantly by income tax considerations: for example, whether to invest in taxable or tax-exempt securities.
167. See Levmore, supra note 13, at 273.
168. Id. Levmore's analysis seems to assume a large one-time retroactive tax that is not attached to a nominally prospective rate increase. See id. at 276 ("a onetime, large-scale (progressive) expropriation of private property").
169. Levmore acknowledges this qualification. See id. at 274, 276. He responds, however, that even if the retroactive tax is only a partial surprise (it surprises only some taxpayers) there still will be some efficiency gains. See id. True enough. However, having been hit once with a "bolt from the blue," id. at 277, and knowing that Congress will be tempted to try the same trick again, taxpayers will begin to watch the sky for the next one. This is the quintessential example of the time-consistency problem. See Stanley Fischer, Dynamic Inconsistency, Cooperation and the Benevolent Dissembling Government, 2 J. ECON. DYNAMICS & CONTROL 93 (1980) (demonstrating the benefits of credible government commitment to optimal tax-policy plan); Kydland & Prescott, supra note 54. Moreover, given the salience of the first retroactive tax, taxpayers may have a tendency to overestimate the chance of its reoccurrence. To avoid these problems, a credible precommitment device is needed. Perhaps the new House rule, adopted by the House of Representatives following the RRA that prevents consideration of retroactive income tax rate increases, is just such a commitment device. See infra notes 169-70 and accompanying text.
170. See Kaplow, supra note 10, at 576 ("To a large degree, expectations concerning future government policy will depend on past decisions, so consistent action over time can be quite important in fostering desired expectations.").
171. For example, the potential efficiency benefits of nominal retroactivity were substantially greater with the large rate increases of 1938 and 1944 than with the relatively small rate increase of 1993. See supra notes 133-34.
172. See Revenue Reconciliation Act of 1993, Pub. L. No. 103-66, § 13201(d), 107 Stat. 416, 459-61 (codified at I.R.C. § 1 (1996)).
173. This terminology was first used by the Treasury Department in its discussion of transition issues presented by shifting to a "broadly based tax system," either a broad-based income tax or a broad-based consumption tax. See BLUEPRINTS, supra note 20, at 159-87. The distinction has been used subsequently and to some extent criticized by tax-policy analysts. See Graetz, Consumption Tax, supra note 26, at 1649-59 (acknowledging the distinction but arguing that the two should be treated analytically as one); Kaplow, supra note 10, at 611-14 (arguing that the two phenomena merit different treatment).
174. Graetz, Consumption Tax, supra note 26, at 1653-54. For a general discussion of several different carryover problems, see BLUEPRINTS, supra note 20, at 160-61.
175. The Treasury Department explains these effects as follows: "[I]f owners were allowed to treat those assets as tax-prepaid, they would receive a gain to the extent they planned to use them for future consumption. Future income on past accumulated wealth would then be free from future taxes, and the government would have to make up the difference by raising the tax rate on the
176. The Treasury Department proposes a tax-prepaid approach, but it also recommends a 10-year phase-in period during which taxpayers would be required to calculate their income under both the old income tax and under the consumption tax and pay whichever tax liability is greater. Then, at the end of the 10-year period, all unrealized capital gains would be subject to taxation. See id. at 184-85. Graetz contends that the Treasury proposal would pose undue administrative burdens, and he seems to suggest that the use of a delayed effective date would be a superior alternative. See Graetz, Consumption Tax, supra note 26, at 1655-58. In addition, Graetz suggests that, if the pretransition wealth-accumulation problem must be remedied, a better approach would be "an immediate deduction of the basis of assets held on the date of enactment (perhaps limited to a maximum dollar amount with a carryover or required spread over a period of years)." Id. at 1655. An additional complication is how to deal with those pretransition assets that were purchased with before-tax dollars, for example, accumulations in pension funds whose original contributions and earnings have been excluded from the income tax base. If such investments are likewise treated as tax prepaid, they would be receiving a subsidy rather than a tax penalty upon the shift to a consumption tax. The obvious solution would be to distinguish pretransition assets according to whether they were funded by before-tax or after-tax dollars. But this would increase the complexity of the system considerably. The increase in administrative costs associated with such a system may or may not outweigh the benefits.
177. Kaplow recognizes an analytical difference between price changes and carryover problems; he argues that the former should not receive transition relief for incentive reasons, and the latter should receive some type of transition relief for distributional reasons. See Kaplow, supra note 10, at 612-14. The difference in my argument is that I would give the former full transition relief, at least in connection with incentive-subsidy transitions. My analysis of carryover problems and of income tax rate increases also would apply to other broad-based tax transitions: for example, the decision to enact an income tax or consumption tax in the first place or the decision to increase the amount of social-security benefits that is subject to income taxation.
178. See BLUEPRINTS, supra note 20, at 167; Graetz, Retroactivity, supra note 9, at 53, 60-63. Graetz refers to nontransferable grandfather clauses as "holder-only grandfathered effective dates." Id. at 53.
179. See BLUEPRINTS, supra note 20, at 167.
180. See Goldberg, supra note 12, at 322 & n.42.
181. Under-the-wire investments can also be seen in other transition contexts. For example, when word gets out that a state government is planning to exercise its power of eminent domain to condemn all of the property in a given area for the purpose of building a road, landowners in the targeted area may have an increased incentive to build structures on their property because of the prospect of increasing the price the government must pay for their property. This is a form of moral hazard, and the optimal transition rule would discourage such investments. Kaplow suggests that there is no analytical difference between what I call under-the-wire investing in a tax-favored asset and any other pretransition investment in such an asset. He argues that they differ only in degree and thus that neither type of investment should receive grandfather treatment. See Kaplow, supra note 10, at 608. I think there is an analytical difference, which is revealed in the contract analogy. If the government were to enter into a contract with a private developer under which the developer was to construct low-income housing units, the contract would include specific terms setting forth what is expected of both sides, including how many units the government was willing to pay for, a number that the government presumably determined to be the optimal amount. If the government later determined that it no longer wanted to buy government housing from this developer, it would have to comply with the current contract, but it could decline to enter into future contracts with the developer. This is comparable to a transition rule guaranteeing grandfathered effective dates for incentive-subsidy transitions. Providing grandfather treatment for under-thewire investments in the context of an incentive-subsidy transition, on the other hand, would be akin to allowing the developer in the contract example to force the government to reimburse it for housing units that were not part of the original contract.
182. In section III.C supra, I discuss why the enactment date is the appropriate cutoff point.
183. See Kaplow, supra note 10, at 607-08 & nn.305-06.
184. One would not be surprised if these taxpayers had expensive tax counsel.
185. If the overinclusiveness problem were thought to be serious, exceptions could be made in certain cases of egregious abuse. For example, if it could be demonstrated that some taxpayer suddenly increased her investment in some tax-favored asset by an enormous amount the week before a proposal to repeal the tax preference is introduced in Congress, grandfather treatment could be denied. Likewise, with the underinclusiveness problem, exceptions could be made if taxpayers could prove that their post-announcement investments would have been made anyway. Both of these exceptions, however, add complexity and cost to the system. It is also worth observing that, when an incentive-subsidy provision is enacted, there is often a compelling reason to make the provision nominally retroactive to the date on which the provision was first proposed. If Congress announces that it is considering enacting a particular incentive-subsidy provision, and it announces that the provision will be applied nominally prospectively if enacted, there will be a large and inefficient lull (perhaps even a total stop) in the type of investment that the subsidy is intended to encourage. This lull will begin on the date the proposal is announced and will end on the effective date of its enactment, as taxpayers who have been waiting to get the benefit of the subsidy suddenly rush to make the investment. Because the period between when a subsidy is first proposed and when it is enacted and made effective sometimes can be quite long, Congress will often make the subsidy retroactive to the date on which it was first proposed; that transition rule will be announced from the very beginning, so that taxpayers will not have an incentive to wait for enactment before making the investment.
186. See generally Shepsle, supra note 2, at 246-47, 250-57 (discussing, in general terms, the efficiency benefits of, and outlining several approaches to, "disabling" the government's discretionary policymaking authority). Kaplow notes that there are incentive benefits to the government's credibly precommitting to an optimal transition policy; however, he argues that the optimal transition policy generally is not to provide transition relief. See Kaplow, supra note 10, at 576.
187. See Graetz, Retroactivity, supra note 9, at 48.
188. See Kydland & Prescott, supra note 54. at 474.
189. But see Kirk J. Stark, The Elusive Transition to a Tax Transition Policy and the Role for a Grandfather Rules Budget, AM. J. TAX POLY. (forthcoming 1996) (arguing that congressional precommitment is unlikely to work, given the history of frequent tax-law changes and given structural changes in congressional committees which allow more input by more diverse parties than previously).
190. 64 F.3d 1531 (Fed. Cir. 1995).
191. Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA), Pub. L. No. 101-73, 103 Stat. 183 (codified in relevant part at 12 U.S.C. § 1464 (1994)).
192. One commentator has argued that the Winstar court's reasoning was deeply flawed because it failed to recognize that the original incentive subsidy (the special accounting treatment promised for healthy thrifts that merged with ailing ones) was a mistake from the start and that Congress, in passing the FIRREA, "was right to negate these deals." Jonathan R. Macey, Rule of Law: The Court Gets It Half Right on Firrea, WALL ST. J., Sept. 13, 1995, at A13. It is unclear whether the author would apply the same reasoning to government contracts. If so, the analysis would seem to suggest that the government should have the right to annul any government contract that it determines ex post to have been ill-advised, without having to pay damages. In addition, the author at times seems to suggest that government "bureaucrats" should not have been in charge of determining which thrifts received the favorable accounting treatment: "Congress [when enacting FIRREA] should have gone much further and relieved all federal bureaucrats of the power to cut special deals with particular favored constituents." Id. But that argument proves too much; under similar reasoning, Congress should never delegate authority to enter into contracts on behalf of the federal government, merely because of the possibility that "special deals" might result. Elsewhere in the article, however, the author appears to agree with the Federal Circuit's holding in Winstar that the government must pay contract damages. See id.
193. See Bowen v. Public Agencies Opposed to Social Sec. Entrapment, 477 U.S. 41, 51-54 (1986); see also Winstar, 64 F.3d at 1546-47.
194. Under the Winstar holding, if Congress were then to repeal the incentive tax credit without providing transition relief, arguably the U.S. government could be sued in contract. Moreover, consistent with Winstar, legislation repealing such a credit would not be exempted from contract principles by the "sovereign-acts" doctrine. See Winstar, 64 F.3d at 1548-50.
195. If this rule seems too strict, it could be changed as follows: If legislation repealing or cutting back an incentive-subsidy provision does not include a grandfathered effective date and does not satisfy the "error-correction" exception, it must be passed by a three-fifths vote rather than a normal majority of lawmakers. Also, Congress could create a list of Code provisions that qualify as "tax subsidies," much as it has done with tax expenditures. The former list would likely be shorter than the latter because it would include only those taxexpenditure provisions that are designed for their incentive effect (that is, for their effect on taxpayers' investment decisions) and not those that are primarily redistributive in nature.
196. H.R. Res. 6, 104th Cong., 1st Sess. § 106 (1995).
197. H.R. Res. 6, 104th Cong., 1st Sess. § 106 (1995).
198. See Comment, An Open Letter to Congressman Gingrich, 104 YALE L.J. 1539 (1995) (calling for reconsideration of the resolution regarding three-fifths-vote requirement) (signed by 17 law professors).
199. See also John O. McGinnis and Michael B. Rappaport, The Constitutionality of Legislative Supermajority Requirements: A Defense, 105 YALE L.J. 483 (1995) (defending the constitutionality of the three-fifths-vote requirement).
200. Another example of a procedural rule designed to limit Congress's legislative discretion is the Gramm-Rudman Act. See generally Kate Stith, Rewriting the Fiscal Constitution: The Case of Gramm-Rudman-Hollings, 76 CAL. L. REV. 593 (1988). Of course, the ultimate rule of this sort would be a constitutional amendment requiring Congress to provide transition relief in these contexts. See, e.g., supra note 132 (citing proposals to amend the Constitution to prohibit retroactive tax rate increases).
201. See Shepsle, supra note 2, at 256 (arguing that a "division of labor committee system . . . enables credible public commitments because it disables discretion of momentary majorities"); id. at 254-57.
202. Again, this may already be the practice of congressional committees most of the time. If so, the practice should be clearly articulated and publicized. One example of this practice occurred during recent hearings of the Ways and Means Committee when Committee Chairman Bill Archer stated publicly that no "loophole" would be reported out of committee unless accompanied by equivalent revenue offsets. See Archer Says No to "Rifleshots" at Hearings on Miscellaneous Proposals, 68 TAX NOTES 239 (1995). One commentator has noted that congressional precommitment via the committee structure has become more difficult in recent decades because of procedural reforms that serve to "open" the tax-writing process to more interested parties than before. Stark, supra note 160. If that is so, this article sheds light on one significant cost of such a trend: the increased default-premium effect resulting from Congress's reduced ability to make credible precommitments in the tax area.
203. Another, perhaps more realistic way of describing the second reason for delegation is this: to allow Congress to avoid political accountability for hard decisions that must be made. I do not focus on that justification in this article.
204. See generally ALFRED BROADDUS, A PRIMER ON THE FED (1988).
205. See A PRIMER ON THE FED id. at 14.
206. See A PRIMER ON THE FED id. at 14.
207. See A PRIMER ON THE FED id. at 14. ("The purpose of this [14-year] term of office is to insulate members from routine day-to-day political pressures.").
208. See, e.g, Kenneth Rogoff, Reputation, Coordination, and Monetary Policy, in MODERN BUSINESS CYCLE THEORY 236 (Robert J. Barro ed., 1989) (demonstrating the benefits of credible government commitment to optimal monetary policy).
209. All of this is not to say that the Board of Governors is entirely independent of Congress or the President. Together, Congress and the President could abolish the Fed, just as they created it, by statute. Moreover, there is a strong expectation that the Board will follow a monetary policy that is generally consistent with the current fiscal policy.
210. See I.R.C. § 7805 (1996).
211. That being said, one of the principal dangers of allocating authority to an agency is the risk of capture by the industry being regulated. See generally George J. Stigler, The Theory of Economic Regulation, 2 BELL J. ECON. & MGMT. SCI. 3 (1971). Arguably, however, the Treasury Department and the IRS are less susceptible to capture than a typical administrative agency, in part because of the relative size of their constituency.
212. ELSTER, supra note 1, at 87-103.
213. See I.R.C. § 41(h) (1996).
214. See Revenue Reconciliation Act of 1990, Pub. L. No. 101-508, 104 Stat. 1388-480 (codified in scattered sections of I.R.C.).
215. Before 1978, the investment tax credit had always been "temporary" in the sense that it always had included a termination date. In 1978, however, Congress made the ITC "permanent," which meant that the credit would remain in force until specific legislation was enacted to repeal or reduce it. See Revenue Act of 1978, Pub. L. No. 95-600, 92 Stat. 2763, 2824 (codified in scattered sections of I.R.C).
216. Although it is fairly common for Congress to extend termination dates or even to make them "permanent," I have been unable to find a single example of Congress prematurely repealing a provision that had an explicit termination date.
217. Graetz recommends increased use of termination dates, but for a different purpose. He sees them as a means of putting taxpayers on notice that they should expect the tax laws to change and therefore should take that fact into account. See Graetz, supra note 9, at 87.
218. A one-year period may be insufficient in some circumstances. For example, the legislative history of the 1978 Act states that the reason the ITC was made permanent was that the uncertainty created by the "temporary" status was reducing the effectiveness of the ITC as an incentive device. See S. REP. NO. 1263, 95th Cong., 2d Sess. 112 (1978).
219. For a definition of these terms, see supra Part I.
220. Golberg asserts that once an up-front subsidy has been enacted it cannot be repealed retroactively. See Goldberg, supra note 12, at 310. He also argues that "a one-time [up-front] subsidy is completely predictable because there is 100% certainty that it will be obtained." Id. at 327. Whereas, he contends that an installment subsidy "can never attain that level of predictability so long as there is a risk of uncompensated termination." Id. Although, as explained in the text that follows, I agree generally that up-front subsidies produce smaller default premiums than installment subsidies do, I think Goldberg overstates the case a bit.
221. See Kaplow, supra note 10, at 587.
222. See id. ("Thus, the analysis of grandfather provisions should be applied to initial program design decisions concerning the appropriate timing of taxes and subsidies. Programs typically should not be designed so as to incorporate grandfathering, because it is generally undesirable." (footnote omitted)).
223. See Goldberg, supra note 12.
224. See id. at 306 ("Periodic subsidies are inefficient and are likely to decrease the horizontal equity of the tax system.").
225. Goldberg suggests that the only purpose served by spreading the cost of a subsidy over several budgetary periods is obfuscation, that is, to hide the true cost of the subsidy. See id. at 312. Indeed, obfuscation may play a role in the use of installment subsidies. However, there are economic as well as political reasons to use the installment form. For example, the use of the installment subsidy rather than an up-front subsidy of equal expected value may permit the government to avoid having to fund the subsidy with deficit financing.
226. Goldberg seems to disagree with this argument, even as a theoretical matter. See id. at 327 ("[T]he need for risk premiums for periodic subsidies cannot be avoided.").
227. See supra text accompanying notes 67-69.
228. See Fred Vogelstein, Moody's Puts Treasury Debt Up for Review, WALL ST. J., Jan. 25, 1996, at C1.
229. Christopher Georges & David Rogers, A Compromise is Developing on the Budget, WALL ST. J., Jan. 25, 1996, at A24.
230. See Vogelstein, supra note 199.