Retroactive Taxation, Unfunded Pensions, and Shadow Bankruptcies

I.     Introduction 

Academics and journalists have been quick to criticize politicians for the dismal financial situations of many state and local jurisdictions. In addition to the usual allegations of mismanagement (“fraud, waste, and abuse”),1 critics have pointed out the perverse incentives created by the electoral process: politicians can further their own careers by hiding the costs of popular governmental programs and ballooning public debt from constituents, while leaving the eventual fiscal collapses to be dealt with by their successors in office.2 I am among those people who have advocated accounting and institutional correctives to make governmental fiscal situations more transparent and, therefore, less amenable to such schemes.3 Yet it would be wrong to let the voting public entirely off the hook. Even when the precise amount of governmental revenue shortfalls have been hidden, reasonable voters in most of today’s currently distressed jurisdictions knew or should have known that it had been years since governmental revenues matched governmental expenditures.4 Newspapers, news programs, and magazines—not to mention candidates for public office—detailed the financial sleights-of-hand politicians employed to “balance the budget” or generate operating funds to avoid running into the “third rail” of American politics—raising taxes.5 Time and again though, voters chose (and continue to choose) candidates who promise to provide public services without raising the taxes necessary to fund them.6 Although it is tempting to ascribe such misbehavior to ignorance or a lack of financial sophistication, it bears noting that politicians are not the only participants with perverse short-term incentives. It is economically rational for voters to elect politicians who will provide debt-financed public benefits when they have the option of moving to another jurisdiction before that debt falls due—or if one lives in Illinois, of reaching the tax-free nirvana of retirement.7 Such voters can have their proverbial cake—valuable public services—and “eat it too” by foisting the costs of debt repayments onto some combination of continuing and new residents, investors, and public employees. Moreover, the perverse incentives provided by this exit (or retirement)8 option can make it harder for jurisdictions to return to solvency, as any financial demands placed upon remaining residents encourages more to leave the jurisdiction, and discourages newcomers from entering. In short, the “exit option” sets the stage for the sort of “death spiral” experienced by cities such as Detroit, Gary, and Stockton—and perhaps soon to be seen in Illinois and New Jersey.9

This Article argues that the perverse effects of the exit (or delay until retirement) option can—and should—be reduced by retroactive taxation. Distressed jurisdictions should be encouraged to raise tax rates not just prospectively, but also retroactively, on those who lived and worked and earned income in the jurisdiction during times of overspending (or under-taxing), thereby creating new tax liabilities for past as well as (some) current residents.10 Such a taxation regime would combat the deleterious effects of the exit option in several ways. First, some of the taxes so imposed would fall on nonresidents who would not be subject to prospective taxation, thereby broadening distressed jurisdictions’ revenue base. Second, by reducing the financial advantages of leaving,11 such taxes would lessen the incentive for well-off residents to leave distressed jurisdictions, while encouraging the entrance of new, economically productive residents. Finally, and most importantly, the inability to escape some of the financial consequences of financial distress may dissuade some residents from voting for or advocating policies that contribute to jurisdictions’ financial distress. It may force them (and the politicians they elect) to confront the hard issues (the tradeoff between lower levels of taxes and higher levels of services) at an earlier and more remediable stage.

States, unlike municipalities, have no bankruptcy mechanism to help them structure their way out of fiscal distress.12 It is not even clear that Congress could provide them with such a structure if it wanted to.13 Yet distressed states, like municipalities and private businesses, need some way of achieving a fiscal reorganization that allows them to continue operating as sustainable entities. This Article will show that retroactive taxation can be used to attain several of the goals instantiated in our current bankruptcy rules as necessary for such sustainability. Accordingly, it argues that retroactive taxation ought to become a regular feature of state attempts to cope with fiscal distress. It might be understood as a shadow bankruptcy law tool.

Part II explains why retroactive taxation would be beneficial. Part III explores some of the design issues that would arise when implementing such a taxing regime. Part IV details the legal constraints on such taxation. Part V concludes.

II.     The Justification for Retroactive Taxation of Former Residents of Distressed Jurisdictions

A distressed jurisdiction, like a distressed business enterprise, cannot realistically pay all of its financial obligations as they fall due. Two questions arise in such situations. The first is the amount of the losses that will be suffered by someone—the excess of the amount owed to creditors over the value of the assets available for distribution to them. The second is which of the many affected parties should suffer the loss so calculated. Usually it is not particularly difficult to determine the amount of the loss when the distressed entity is going to be liquidated or extinguished. Its assets are gathered, sold to the highest bidders, and the amount received measured against the amount owed. The harder question may be how to distribute those proceeds among the claimants, although the statutory priorities often solve that problem as well.

Neither question, though, is easy when the underlying entity needs to be preserved—whether the distressed entity is a business to be reorganized under Chapter 11, or a municipal entity under Chapter 9. Determining the amount of the loss in a Chapter 11 situation requires determining the current value of assets that will not be sold in the near future. The absence of contemporaneous third-party sales often engenders valuation difficulties.14 Municipalities need to retain sufficient assets to remain functionally and financially viable, which is a contestable and often contested standard.15 These questions are even more difficult when operating outside of the formal bankruptcy system, as one must do when the distressed entity is a government entity, such as a state, which is not subject to the dictates of the bankruptcy laws. Outside of bankruptcy, one has fewer guideposts to reaching the correct decision.

However, even where the bankruptcy laws do not apply as a formal matter, they are not totally irrelevant. Their provisions reflect hard-won knowledge about what it takes to successfully reorganize a debtor, as well as the socially (and perhaps legally) acceptable distributions of the inevitable losses. And one of the interesting features of bankruptcy law is that, under certain circumstances, it allows and even requires that some pre-bankruptcy distributions from distressed debtors be “clawed back” and added to the bankrupt’s estate.16 The premise of this Article is that a retroactive tax is functionally identical to such a bankruptcy “claw back”; when properly structured, it fulfills the same laudable purposes. It can be both fair and efficient.

A.     A Defense of Retroactive Taxation

1.     The Analogy Between Retroactive Taxation and Debt Financing

Many legislatures, academics and the general public regard retroactive taxation as unfair and even immoral.17 It is not hard to rationalize this reflexive abhorrence. Retroactive taxation seems to fall into the category of “taxation without representation,” as those levying the tax may not be the same people who will find themselves subject to it. In the most extreme case, imagine a state—call it “Distress”—in which everyone who lived in the state in year one has moved to another state in year five. All managed to sell their homes, however, and those buyers, the new residents of Distress, vote in year six to retroactively increase the property tax rates for years one through four. None of the new residents will have to pay any portion of this tax increase; it will fall entirely on the former (and nonvoting) residents. Indeed, retroactive taxation may appear to be just another way of passing off the cost of government to “the man behind the tree”18 who is not only unseen, but in no position to defend against the imposition of an unfair burden.

Government debt, however, does exactly the same thing. Debt—regardless of its form—is not “free” money; it comes with an obligation to repay. And usually, the funds to repay that debt must come from the future imposition of taxes or fees. The individuals faced with the task of repaying that debt may not be the same individuals as those who incurred the debt. Imagine another state, this one called “Decline.” In year one, the residents of Decline float (sell to investors) $1 million of bonds and use the proceeds of that sale to pay all the expenses associated with providing fire and police protection in that year. By year five, when principal payments on these bonds come due, all of the year-one residents of Decline have retired and moved away. The new residents of Decline, none of whom voted to issue the bonds nor received any benefits from the year one expenditures on police and fire services, are left with the responsibility of raising the funds necessary to pay this outstanding debt. In all likelihood, they will have to raise tax rates—or live with diminished government services—to do so. Although as a formal matter, these new residents have the right to “vote,” their vote is limited to the choice between these unpleasant alternatives. They had no input, no vote (except for moving in) on the question of whether they should have been placed in this situation to begin with; that was decided by voters who benefited from the services but paid none of their costs.19

The ability to finance government benefits with debt rather than current tax and fee revenue, then, also seems like a path towards political unaccountability and wasteful government expenditures. Yet few regard the issuance of government bonds or other forms of taking on governmental debt as unfair and even immoral.

What accounts for the difference in attitude? One possibility is that often the proceeds of government bonds are used to provide long-lived benefits. If bond proceeds are used to build roads or schools that will be used for 20 years, spreading the costs of paying for those roads and schools over 20 years seems eminently reasonable—even desirable. After all, in the absence of such spreading, it is entirely possible that roads and schools would not be built even in situations in which their 20-year benefits outweigh their 20-year costs inasmuch as the current benefits would not outweigh construction costs in any single year. The original voters/bond floaters may guess wrong as to the eventual value of roads or schools—their actual value to future residents may be higher or lower than projected—but it is hard to characterize attempts to spread the costs of providing government benefits over the groups of residents who are expected to benefit from them as exploitative or a perversion of rational budgeting processes. By contrast, attempts by later taxpayers to weasel out of their responsibility to pay for their share of such projects by (say) retroactively taxing past residents would be unjustifiable, as later taxpayers properly would be viewed as shifting the costs of the (roads and schools) benefits provided to them onto outsiders.

But in many fiscally distressed states, the largest component of outstanding government indebtedness is not formally issued government bonds, but rather unfunded promises to provide pension and retiree health benefits to state employees.20 For the most part, these promises were made to employees at the time they were hired or during their working careers to induce them to perform services benefiting the then-current residents of the governmental jurisdiction. These promises, like all forms of deferred compensation, were an integral part of salary arrangements. It was as if governments promised to pay their workers (say) $40,000 per year—$30,000 in cash up front and $10,000 worth of pension and retiree health benefits. When governments paid the $30,000 in cash salary out of current tax revenues, but made little or no attempt to cover the additional $10,000 expense by putting money aside to cover the eventual costs of providing the promised pension and health benefits, as an economic matter they shifted this $10,000 cost onto (nonvoting) future residents. Such decisions meant that governments (and their then taxpayers/residents) chose to pay for only three-quarters of the costs of the services provided to them in that year by government employees, leaving the rest for future, nonbenefiting (and nonvoting) taxpayers/residents to pick up.

There is no question that it is inconvenient, even unpleasant, to fund promises for deferred compensation in the year such compensation is earned through the performance of services. The money so set aside has to come from somewhere. Cash earmarked for payment of deferred compensation is cash that cannot be spent on other, current needs, such as teachers, police, firemen, or subsidized child care. But putting off the obligation does not cause it to go away, unless the jurisdiction is both comfortable with, and legally allowed to break,21 its pension promises—thus retroactively reducing employee salaries.22 At some point, a state that has promised deferred compensation has to devote resources to making good on those promises. And those resources can come from only two sources: tax increases or the diversion of existing tax revenues that would have been spent on other priorities. Thus the question becomes who should pay the additional taxes or suffer the reduction in government services: the residents23 who received the services performed by the employees or a later group of residents.

The case for making later residents pay the costs of services rendered to, and benefiting only, an earlier set of residents is weak. It depends on one of two questionable assumptions. The first is that subsequent residents will be wealthier than the current residents. The second is that while some current expenses may have been left unpaid, past residents more than made up for that failure by paying for capital expenditures and other items which redound to the benefit of new residents. Neither is likely to be true.

The belief that subsequent residents will be wealthier than current residents is a convenient fallacy.24 Even if overall wealth or income increases over time, the likelihood that increases will be distributed evenly across jurisdictions, even jurisdictions as large as states, is small. As a historical matter, demographic and economic changes routinely turn wealthy jurisdictions into poor ones and vice versa irrespective of trends in the larger economy.25 And unfortunately, the larger economic trends have not been particularly favorable for a very long time. On an inflation adjusted basis, average wages have barely moved in the last 30 years.26 Further, changes in the economy have meant that individuals’ standards of living may well be declining in some respects despite the slight improvement in average wages. For example, it takes more private resources (both in terms of time and monetary commitment) to sustain a middle class existence when college (or at least some post-secondary) education becomes necessary to obtain a decent job, than when a (publicly funded) high-school education sufficed.27

Nor can distressed jurisdictions attribute (or claim an offset for) their financial distress by pointing to the valuable legacy of public assets they have built up that will benefit future generations of residents. Every jurisdiction contains some long-term assets, ranging from roads, sewer and water treatment plants, parks and other recreational facilities, to schools and government buildings. Some jurisdictions may pay for such assets up front, leaving future residents with a valuable stock of such assets. Much more typical, however, is to fund long-term projects with debt repayable over the life of the asset. Although past residents may have had to live through the aggravation of the construction process, the “valuable inheritance” often comes subject to claims for debt repayments which reduce the jurisdictions’ effective equity in the assets to something close to zero. Such inheritances are incapable of offsetting the detriment of pension and other current expense indebtedness. Moreover, in recent years, fiscally distressed jurisdictions have become increasingly adept at entering into complicated financial transactions to “monetize” even those government assets in which they do have equity, further reducing their value to future generations of ratepayers.28

The weaker the justification for imposing the costs of past pension promises on current residents of a jurisdiction, the more defensible the argument for imposing retroactive taxation of past residents becomes. It would of course have been preferable had those earlier residents been forced to internalize the costs of their decisions regarding the level of the government services they enjoyed when they could have made the necessary tradeoffs for themselves, in accordance with their values and their perceptions of their economic constraints. But that time is long past—the issue facing courts and jurisdictions today is, given the choices made by those earlier residents, who should be responsible for “paying the piper?” And when asked that way, the case for requiring past residents to contribute financially seems reasonably clear.

2.     The Analogy Between Retroactive Taxation and Claw Backs in Bankruptcy

This argument in favor of retroactive taxation is little different from the one underlying the statutory recoupment of “profits” (amounts received in excess of the amount initially invested) received by early investors in Ponzi schemes under bankruptcy law.29 Bankruptcy law includes several provisions that allow bankruptcy trustees to claw back amounts earned and withdrawn by investors—even investors who had closed out their accounts long before the criminal nature of the scheme was revealed. Section 548 provides that trustees may void conveyances made within two years of the filing of the bankruptcy petition.30 Section 544 allows, in addition, the voidance of conveyances deemed fraudulent under applicable state law for the period of time specified by that state’s statute of limitations.31 Although both provisions purport to allow trustees to recoup only past transfers resulting from actual or constructive fraud, recoupment of profits (as opposed to principal)32 does not depend on any proof that the transferee knew or participated in the fraud. “The illegitimacy of the enterprise seems to be held against the investor.”33 Thus, in many cases,34 profits may be recouped from wholly innocent investors when the underlying income-generating transaction was later revealed to be a criminal venture. It is of course not criminal for a jurisdiction to borrow money to pay current expenses, but it is generally considered unwise, at least at the sub-national level,35 and federal tax law sometimes explicitly discourages such maneuvers.36 Moreover, the voters that acceded in decisions to rely on the explicit and implicit debt rather than pay for the full costs of the services they were enjoying likely were more cognizant of their attempt to foist costs off onto others than were many Ponzi scheme investors. In the Madoff case, for example, investors had no knowledge—and no easy way of gaining the knowledge—that Madoff’s activities were less than legitimate.37 They had a far better claim of being truly innocent victims than former residents of many distressed jurisdictions, who could only have lacked awareness of their underpayment for services by actively avoiding that information.

3.     The Analogy to Balanced Budget Requirements

A balanced budget “is widely considered to be the foundation of state fiscal practices.”38 Accordingly, the overwhelming majority of the states have some statutorily or constitutionally imposed balanced budget requirement,39 although their stringency (and thus effectiveness) varies from state to state.40 Indeed, the magnitude of current fiscal straits facing many jurisdictions illustrates the toothless nature of their balanced-budget rules. All too often, politicians devised techniques for meeting the literal requirements of a “balanced budget” while running large deficits from an economic perspective.41 For example, several scholars have pointed out that in many jurisdictions, politicians managed to meet balanced-budget requirements by manipulating the discount rates used to calculate public-pension liabilities.42

Another way of viewing the claw backs advocated in this Article, then, is to view them as a method of strengthening weak balanced-budget requirements. Retroactive taxes allow a jurisdiction to rectify or “true up” errors (deliberate or not) in the financial projections used in calculating whether a jurisdiction achieved a balanced budget. Most importantly, such taxes would discourage deliberate miscalculations, such as purposefully understating pension liabilities, by forcing taxpayers to internalize the real costs of such liabilities.

4.     Potential Claw Back Defenses

i.     “Paying Backward”

The actual facts often are not quite as clear-cut as jurisdictions setting nothing aside for payment of pension and other post-employment promises, completely free-riding on future taxpayers. When these programs for post-retirement benefits were first established, older employees were often “grandfathered” in. Pre-existing employees were granted pension credit and benefits for work performed prior to the institution of the pension plans—the economic equivalent of a retroactive pay raise or a gift. Although insufficient money was set aside to fund the pension promises made to current workers for contemporaneous services, current tax revenues were used to pay the benefits promised to retirees who had performed services for earlier generations of residents. Thus, past residents may claim that as they “paid backward,” new residents should do the same.43 However, even leaving aside the fact that it was the decision of those earlier residents to grandfather in employees who often had no expectation of receiving (and certainly no legal entitlement to receive) such post-retirement benefits, the implicit trade of past benefits for current benefits was rarely, if ever, an even exchange. The cost of providing benefits payable to previous generations of employees generally fell far short of the present value of benefits being accrued by current workers. In part, this was because of the growth in the number of government employees over time, as the functions assumed by state governments expanded; earlier generations of state residents enjoyed lower levels of governmental benefits than those provided to later generations.44 Rising cash salaries (often linked to the increased educational requirements of those necessary to provide the services demanded by residents of the new administrative states) probably also played a part in this phenomenon. Whatever the cause, from an actuarial standpoint, in many states unfunded liabilities always went up, rather than staying the same or decreasing, as would be expected had the combination of actual payments and funds set aside for future payments kept pace with new benefit accruals.45

Of course, the question is not necessarily whether new residents or old residents should be forced to contribute more money to pay outstanding liabilities; discharging those liabilities without payment is always another option. Some (but not all) jurisdictions have the option of failing to pay some or all of the post-retirement benefits promised to employees (retroactively decreasing their salaries).46 Alternatively, they can make more money available for payment of current expenses and other creditors by defaulting on their formally issued bonds. Even if it is reasonable to place a payment burden on old residents rather than new residents, perhaps it is even more reasonable to place all of the burden of nonpayment on one of the creditor classes.

ii.     “Overpaid Employees”

The argument for imposing losses on public employees is two-fold. The first is that these employees likely incorporated the risk of nonpayment in their salary arrangements. Paying them in full thus would result in an overall overpayment. The second argument is that even if the salaries of individual employees did not incorporate a risk element, employees as a group benefited from increases in overall rates of public employment as a result of deferring pension funding. Though both these arguments are plausible, they merely show the reasonableness of imposing some losses on employees, not that employees should absorb the full amount of the fiscal deficit. Neither suggests that it is better to impose all losses on employees rather than past residents.

The first argument is that employees likely incorporated the risk of nonpayment in their salary arrangements. For illustrative purposes, let us say that absent any risk of nonpayment, an employee would have agreed to a total salary package of $38,000 per year, divided between cash salary ($30,000) and deferred retirement and health benefits ($8,000). However, knowing that there was some chance that the deferred compensation would never be paid, the employee insisted on a total salary package of $40,000 per year, again divided between cash salary ($32,000) and deferred retirement and health benefits ($8,000). Thus, the employee would have received an extra $2,000 per year up front to compensate her for the risk that the deferred compensation would not be paid in full (or at all). If the risk was priced correctly, she could have used the extra cash salary to buy insurance against the possibility of nonpayment of the promised benefits—and to insist that such an employee be paid in full while imposing losses on others would be unreasonable. She would end up being overcompensated.

What is the likelihood that employees actually incorporated the full risk of nonpayment in their salaries? It is certainly true that the unions representing state employees were often fully aware of pension and health benefit underfunding and regularly communicated their disquiet about the possible implications of the funding shortfalls. Indeed, in Illinois, the union brought several unsuccessful lawsuits to try and force the state government to adequately fund its pension programs.47 The courts in those cases told the plaintiffs not to worry about the underfunding—that their benefits would be protected under the state constitution no matter what.48 (And so the court held when the state tried to revisit those promises.)49 Given that history of court reassurance, it is unlikely that Illinois employees priced much risk into their salary negotiations. Other government employees from states with comparable guarantees may have been equally disinclined to build in a risk premium, though when push came to shove, some found their post-retirement benefits reduced.

The more general point is that the inclusion (or not) of a risk premium, as well as the size of that premium, would have been based at least in part on the history of payment (or not) of such benefits and expectations about the underlying legal rules; until recently, few believed that defaults could or would occur.50 One can always contend that such a misperception of risk is due to negligence, for which the employees should bear the resulting financial penalty—but in many cases, that misperception was deliberately created by the actions and statements of courts and elected officials, both agents of those former residents. Of the two parties to these transactions, employee unions and now-distressed jurisdictions, it is hard to view the latter as less culpable. Again, this does not mean that employees should be protected against all losses from their participation in what might be regarded as a “lottery” for pension benefits,51 but it does suggest that they should not be responsible for bearing the entire financial shortfall.

The second overpayment argument may actually be more persuasive. Had the voters acted more responsibly and elected officials who set aside the money necessary to pay their deferred compensation promises, the money would have had to come from somewhere. It is possible that taxes would have been increased by enough to provide the same amount of services—but it is also possible that budgets would have been balanced by hiring fewer state employees and by voters becoming accustomed to a lower level of state services. If the latter occurred, some of the employees would not have been hired, at least not in their actual jobs in that particular state. Presumably they would have worked somewhere else, but by definition (or they would have chosen that other job at the time) at a less-preferred job, and possibly for less money. Thus, state employees as a group likely benefited from the deferred funding—though identifying which particular employees benefited, or by how much, would be very difficult. To the extent such a benefit was received, it would be reasonable to claw back that benefit by reducing the amount of benefits actually provided, or by reducing workers’ wages going forward.52 However, it is hard to believe that, even if explicit claw backs were legally permissible,53 they would be large enough to absorb the total amount of the asset shortfall.

iii.     “Overpaid Bondholders”

The overwhelming majority of municipal bonds have been purchased through voluntary market transactions.54 As a result, one might expect that the prices paid by their holders would have incorporated not only the then-current expectations for a riskless rate of return over the remaining term of the bond but also the risk of nonpayment, the possibility that the bond issuer would default or fail to make the promised payments. After all, absent some sort of compulsion to buy such instruments, prospective purchasers should walk away rather than purchase an instrument that failed to offer a return commensurate with the investment’s underlying risk. In a world in which risk-free bonds paid two percent interest, a prospective purchaser would require, say, a four percent interest return after examining the financial status of the issuing state. The higher interest rate, like the higher salary received by state employees, would compensate them for the risk of nonpayment at some point in the future.55 Like those employees, the contention is that since bondholders were paid in advance for an expected “haircut” at the end, care must be taken to make sure that they are not “overpaid” by forcing more losses on the distressed debtors or other more “innocent” creditors.56 The argument is, essentially, that the economic losses caused by default are no more sympathetic—or deserving of relief—than those that might have been caused by changes in prevailing interest rates.

Again the question is whether it is likely that bondholders actually incorporated sufficient risk premiums in their prices—and if not, to whom to ascribe the failure. Some states deliberately withheld information which could have triggered demands for higher interest rates to reflect default risks. Many bonds were highly rated by rating agencies—and thus incorporated very low-risk premiums—precisely because those agencies took the governments at their word.57 It is, moreover, more than a little odd to claim bondholders should be held financially responsible (vis-á-vis state residents) because the bondholders failed to project the extent of those residents’ political and fiscal irresponsibility in advance. The moral hazard on the residents’ side is just too obvious.58 Whatever risk premium the creditors extracted, the residents could undermine by choosing to act even more irresponsibly. It is one thing to make a bet on the occurrence of events that are under the control of neither party to the agreement, and another to expect one party to make a bet on the occurrence of events under the control of the other party; if bondholders thought they were doing that, the interest rates they demanded would have been astronomical. And they were not.

At any rate, the argument I am making in this Article is not that bondholders (or government employees) should be held harmless, that the combination of past and future residents ought to be forced to pay them the full amount due them under the terms of their respective contracts. The argument is only that, compared to the other participants in these slow-moving financial disasters, past residents are at least as blameworthy and therefore, they should be part of the solution. Of all the involved parties, they are the only ones likely to have walked away from the situation enjoying a surplus—the difference between the taxes that they paid and the value of the government services they received—and it would be defensible to require that they contribute at least part of this surplus to the pot to be divided among the other affected participants.59 Moreover, doing so would encourage them to behave more responsibly in the future.

B.     Beyond Defensibility: Would Retroactive Taxation Be Sensible?

Even if retroactive taxation is defensible in theory, it might be hazardous in practice. Retroactive taxation might encourage residents of distressed or potentially distressed jurisdictions to engage in behavior that would have deleterious consequences for those jurisdictions, not to mention their various creditors. However, as the discussion below explains, the opposite effect is more likely. The threat of retroactive taxation should encourage more fiscally responsible, rather than less fiscally responsible, behavior. In addition, it might help fiscally distressed communities recover. As explained below, the precise effects of such a scheme are to some extent indeterminate, as they will depend in part on exactly who benefits from, and who pays, the retroactive taxes.

1.     Who Benefits?

The effects of retroactive taxation would depend in part on the uses to which the revenue raised through such taxation is put. Would it be used to increase the amount available for distribution to creditors (including employees and retired employees), or would it be used to reduce the contribution demanded of current residents? If used to increase the amount available for distribution to creditors, the creditors will be happy, and (at least if one believes in the pricing theory) would possibly hold down future risk premiums. Such decreases (or constraints on future increases) would redound to the benefit of future residents, should they be interested in incurring new debt of any type. However, such benefits would be indirect and perhaps small.

If the funds raised through retroactive taxation are used to decrease the contribution of current residents, however, the benefits for the distressed jurisdiction may be more substantial. The less the current residents spend on paying back debts incurred by past residents, the more they can spend on current services or on moderating tax rates. Instead of being used to pay off creditors, new and existing residents’ tax revenues could be used to provide the sorts of governmental services that make living in a jurisdiction tolerable or even pleasurable. Alternatively, the jurisdiction’s fiscal demands (taxes) on its residents might be eased. Either outcome would make the jurisdiction a more attractive one in which to live or operate a business. The difference between the tax and service packages of other jurisdictions and the distressed jurisdiction would decrease. On the margin, then, fewer residents and businesses would feel impelled to decamp to those other jurisdictions.60 The financial pressures on the distressed jurisdiction that lead to the “death spiral” would be minimized; it might even be able to attract new, productive citizens. The analogy to the favorable treatment of “new money” in a reorganized entity is quite compelling. No one would be willing to invest in a reorganized business entity—or move to a financially distressed jurisdiction—if all the money they invested or the tax revenues they paid went to pay off old creditors rather than to generate profits—or services, in the case of the jurisdiction—for the new investors or residents. The Bankruptcy Code avoids penalizing newcomers by providing partial insulation of their contributions from the claims of old creditors;61 in addition, prospective lenders generally require the existing creditors to subordinate their debt claims as a condition of providing the bankrupt entity with new money.62 Pushing the tax obligations onto old residents rather than new ones would have a similar insulating effect. This is important for an ongoing enterprise that will continue to operate past bankruptcy as is true of most distressed jurisdictions.

2.     Who Pays?

As noted above, this retroactive tax regime leaves new entrants in a distressed jurisdiction untouched, or even benefited, if the retroactive taxes moderate the taxes they will have to pay on income (or property) earned while resident in the distressed jurisdiction. Additional taxes will be levied on the jurisdiction’s former residents—a group that would include continuing residents as well as former residents who moved to other jurisdictions.63 In the absence of population mobility—if no one moves into the distressed jurisdiction and no one moves out of it—there would be little difference between currently structured tax regimes (i.e. a regime which simply raised future tax rates) and the retroactive regime. The current group of residents would find themselves paying both sets of taxes. Assuming the total amount of taxes raised would be equivalent under both taxation schemes (retroactive vs. increasing future tax rates), the taxes paid by most residents would be about the same under either scheme. There would be some rearrangement of the liabilities within the group of current residents. For example, newly retired individuals would have to pay retroactive taxes whereas they may be able to escape paying taxes on future income due to the exclusions from the tax base of certain types of retirement income. Thus, a retroactive tax regime would shift some of the tax burden from younger taxpayers to older taxpayers.64 In addition, residents could find themselves better (or worse) off if their incomes rose (or declined) between the retroactive period and later years for other reasons.

The shifts in tax liabilities become more substantial as mobility between jurisdictions increases. The benefits for new entrants have already been noted. Their gains, though, will come largely at the expense of former residents, who will find themselves in the position of having to pay additional tax to jurisdictions they have left. The question is who these former residents are—and how many there are likely to be? Although interstate migration has declined over time,65 the number of movers remains significant.66 The largest cohort consists of people in their twenties.67 Since most of these individuals are likely to be at the beginning of their working careers, a retroactive tax scheme likely will have little impact on them; many will have had little to no income or have owned real property in the retroactive period. The regime will have a greater impact on older movers, including the “small peak in the early sixties related to retirement moves.”68 This should result in a (probably minor) shift in tax obligations from the relatively poor to the better off since long-distance moving rates are highest for “the most educated segments of the population” who presumably are also the most highly compensated.69

Businesses move, as well as individuals. A retroactive tax regime could serve as a “transition tax” on those that flee struggling jurisdictions. Of course, such taxes can only be imposed on businesses that establish residence elsewhere; it will be hard to collect additional taxes from those that have closed due to financial distress.

However, shifting tax obligations from one group of taxpayers to another is only part of the benefit conferred by a retroactive tax regime. Another possible benefit is a change to political dynamics.

3.     Nonrevenue Effects on Shifting Tax Liabilities

Some advocates of retroactive tax regimes have been criticized for failing to recognize that such schemes only work once.70 The efficiency gains of retroactive taxation schemes often have been predicated on the effects of surprise; the unexpected nature of the tax prevents people from distorting their original behavior.71 Once people realize that retroactive taxation is a possibility, they may start planning for its occurrence, thereby destroying (or raising the costs of providing) the incentives that, in many cases, the original tax rules were supposed to create.72 For example, taxing capital gains more favorably may not induce additional investment in capital assets when investors are afraid that those favorable tax rates will be eliminated retroactively.73

However, in this particular situation, such advance planning and acting would be beneficial, rather than detrimental.74 The way to avoid retroactive taxation is to ensure that one’s jurisdiction’s revenues cover the full cost of providing the services it provides.75 Fear of retroactive taxation may impel more residents to demand that their politicians engage in honest budgeting techniques and make the hard choices required to match government revenues with government expenses, rather than lazily assume that they can walk away from any financial mess that they have created.76

There is another way to avoid retroactive taxation, which is to abandon a jurisdiction as soon as it starts to encounter budgetary trouble. Taxpayers who leave before deficit spending is enacted cannot be said to have benefited from it. Some residents may decide that the prudent thing to do is to move as soon as a jurisdiction’s finances seem to be on an irresponsible course. If too many residents do this, retroactive taxation schemes would accelerate the timing of “death spirals” rather than eliminate their occurrence. However, such acceleration may actually be a positive development as it would lead to the unraveling of unsustainable fiscal behavior earlier rather than later. Just as it would have been better for Madoff’s Ponzi scheme to have been revealed after one or two years rather than 20—far fewer people would have been harmed and much less money would have been lost77—it would probably have been better for the taxpayers of Illinois to have been forced to cope with the unsustainability of the gap between its tax and spending practices before its unfunded pension funding obligation grew to its present height.

For better or worse, though, the advent of accelerated death spirals is probably not realistic. Moving has high transactions costs for most people. Changing residences is both expensive, time consuming, and socially disruptive; personal moves often involve switching jobs. Nor is it that easy to speedily relocate a business.78 Thus, once willful ignorance starts to pose a serious financial risk in the form of retroactive taxation, residents’ incentive to keep track of what their representatives are up to, from a budgetary standpoint, will increase. Although it is unrealistic to believe that a majority of residents will become versed in the intricacies of government finance, it may not take that many informed and engaged citizens to counterbalance their neighbors’ and their representatives’ natural inclination to avoid bad news, and to have some salutary effects on financial behavior. At any rate, the situation is unlikely to be any worse than it is today.

For any benefits to arise from retroactive taxation, though, the scheme must be effective. The next Part outlines what an effective retroactive tax scheme might look like.

III.     What Retroactive Taxation Would Look Like

Not all taxes can be effectively collected on a retroactive basis. Sales taxes, for example, which are collected at the point of sale and without (in most cases) any record of their payors, would not be susceptible to later adjustment. Most fees would also fall beyond the scope of any retroactive scheme, because even though it may be possible to identify those on whom they are levied, as a legal matter, many may not be raised above the amount necessary to finance the operation of a specific regulatory regime or associated program.79 The most obvious candidates for retroactive taxation are state income and property taxes.80 In both cases, the identity of individuals to be taxed on a retroactive basis would be clear, as would the amount of the tax base; all that a government would need to do would be to revise the tax rate and, based on information already in their possession, send out bills in accordance with procedures that they already use for assessing tax deficiencies or otherwise adjusting tax liabilities following an audit. The difficult question is deciding how large the bills should be. Answering that question requires facing several subsidiary issues: how far back in time retroactive taxation should extend, when the retroactive period should start, how much additional revenue should be collected with respect to each retroactive year, and how much of this revenue should be collected from different classes of taxpayers. Parameters for resolving these issues are discussed below.

A.     How Retroactive?

The financial problems, including the underfunding of pension obligations, of many of the states currently facing financial distress are long-standing. Illinois’ pensions have been underfunded since their inception, in 1917.81 It is ludicrous to think that any retroactive tax regime could extend that far back in time. From a practical perspective, many of the putative taxpayers would be dead, with hard-to-trace heirs. Given that such practical concerns militate that some responsible parties will successfully avoid making any financial contributions, the question of which ones will be subject to the tax is a line-drawing problem. Resolutions to such problems are inherently arbitrary, but analogous situations can be used to provide temporal parameters.

One such analogous situation is bankruptcy. Trustees administering bankruptcy estates are allowed to claw back amounts paid by the bankrupt entity prior to the bankruptcy filing. All insiders are subject to claw backs of any amounts paid within one year of filing a bankruptcy petition, even in circumstances when the payments were in return for valuable services or property transferred to the firm.82 Payments made in return for less than adequate consideration may be recouped for much longer periods. The Bankruptcy Code provides a general two-year claw back period for constructively fraudulent payments,83 but allows for that period to be extended by state law.84 Many states have done just that. For example, New York law provides that all distributions made to partners (and former partners) of bankrupt partnerships made during the partnership’s period of insolvency may be clawed back, regardless of whether the partnership had filed for bankruptcy or the partner knew of the insolvency at the time the payment was made.85 Claw back periods for other types of payees vary, but can be substantially longer than two years.86 New York, where the Madoff bankruptcy was administered, allows amounts paid during the six years prior to the filing of the bankruptcy petition to be recouped by a bankruptcy trustee in appropriate cases.87

The statute of limitations for federal tax deficiencies provides another set of parameters. The general statute of limitations for the assessment of federal income taxes is three years after the due date of the return.88 However, the statute of limitations extends to six years in the case of “substantial omissions” from gross income89 and there is no statute of limitations for civil tax fraud90 or for the failure to file a return.91 State tax laws have similar rules.92 Moreover, those statute of limitations apply only to the period for assessing the tax. Once a tax has been assessed, the federal government has another ten years in which to collect the assessed tax.93 Moreover, these deficiencies can be collected not only from the taxpayer, but also from other persons or entities to whom the taxpayer has transferred property or other assets for less than adequate consideration in an effort to avoid payment of tax liability.94 Such suits are not subject to state statutes of limitations on fraudulent conveyance suits,95 and the applicable federal statutes of limitations are extended by a year when “transferee” liability is involved.96

These comparisons suggest that a six year period may be the upper limit of a politically acceptable retroactive tax, and that a lower limit, of three-to- five years, may be more realistic.97 Although from a financial perspective such a period is far too short to recoup many of the excess benefits derived by state residents, it is short enough that taxpayers and tax authorities should still have the tax documents necessary to recalculate tax liabilities.98 It may even be possible to extract additional money from heirs of decedents who would have been subject to tax had they still been alive.

However, there is nothing magic about this three-to-five year period. Indeed, a case can be made that a longer period would be appropriate, given that one of the arguments for a retroactive tax is to encourage more responsible political decisions. Many elective offices have four or even six year terms; affecting the outcome of elections may require seven or even ten years of look-backs. As detailed in Part IV.A.2, though, courts may not countenance a look-back period of that length affecting years prior to enactment of the retroactive rules, deeming it too disruptive of settled expectations.99 If a lengthy look-back period is desired, a more fruitful approach may be to start with a three-to-five year retroactive period while phasing in a lengthier look-back period that would apply only to taxable years subsequent to the enactment of the taxing regime. Once the law is on the books, no taxpayer should be able to complain about unfair surprise or hardship.

B.     Three-to-Five Years from When?

Both tax returns and bankruptcies involve the filing of documents which serve as markers for the start of the retroactive or claw back period. No such definitive marker exists for determining when a state becomes fiscally distressed enough to justify the imposition of retroactive taxes. States cannot file for bankruptcy, and the concept of “insolvency” does not map very well onto governmental entities. Thus, there is no easy answer to the question of three-to-five years preceding what event? What financial circumstances should be considered sufficiently dire to trigger retroactive taxation?

Leaving an unfettered choice of the triggering event in the hands of the state risks allowing politicians to play games at the behest of favored constituents or former constituents. They may try to exempt or minimize the exposure of some by delay. Alternatively, they may try to penalize a large employer leaving a jurisdiction by using that departure as an excuse for an unnecessary declaration of fiscal emergency. It may be preferable to make an action of an independent third party—say a downgrade by a bond rating entity—at least a precondition for such taxation.

That assumes, of course, that retroactive taxation ought to occur infrequently, like filing for bankruptcy. One might want it to serve a different purpose, as a routine “true up” to conform projected to actual government expenditures and receipts. In that case, such a tax would be levied every three-to-five (or ten) years, depending on the length of the retroactive period. The routine retroactive adjustment of tax rates would decrease the number of jurisdictions falling into financial distress and probably decrease the scope of that distress in the remaining cases. However, it would also leave state governments with far less flexibility when dealing with temporary fiscal problems.

C.     How Much Revenue?

Financial distress is rarely a short-term development. Most, if not all, fiscally distressed states accumulated their problems over decades, if not longer. It is unrealistic to assume that such fiscal problems can be solved by a few years of retroactive tax increases; tax rates in those years would have to be well in excess of 100% to reach that goal. Just as importantly, the rationale underlying retroactive taxation lies in the discrepancy between the amount of taxes previously collected during the retroactive period and the services conferred on residents during those same years. Recompense is due for that unjust enrichment of those residents, rather than for their failure to pay the debts associated with the performance of services provided to some earlier generation(s) of residents. Rates should be set high enough to claw back that unjust enrichment but no more (or at least, not disproportionately more than the burden placed on future residents for those past expenditures).

Calculating the amount by which those residents were unjustly enriched may be far from easy. Although it should not be difficult to determine how much revenue was raised in particular years, calculating the cost of government services (for purposes of this Article, I assume that government services have a value at least equal to their cost) may be difficult.

Accounting for defined benefit pension plan liabilities is notoriously tricky. Determining the value of benefits accrued in a given year requires utilizing a number of contestable actuarial assumptions, ranging from life expectancies, expected wage growth, and rates of return on invested assets.100 Changes in any one of these assumptions can significantly increase or decrease the actuarial value of the pension benefits earned in a particular year.101 Not surprisingly, the validity or not of various actuarial assumptions has been the subject of ongoing dispute, and such disputes will impact the calculation of discrepancies in any year between the amount set aside in pension trusts and the amount by which pension liabilities increased.102 Calculating the value of unfunded retiree medical benefits poses similar problems.

It is easy to envision other sources of controversy. For example, it may be hard definitively to assign other governmental expenditures to current periods (as opposed to future periods). Should road repaving expenses—or the replacement of utility infrastructure—be treated as current expenses or capital, allocable at least in part to future years and future taxpayers? Accounting conventions undoubtedly exist to make many of these allocations, but they too may be attacked as unrealistic once such allocations become high-stakes exercises.

Some of the difficult issues may be finessed with the application of a consistency rule. It would seem quite unfair (and likely fiscally impossible) to require past residents to pay the full costs of their government services without imposing a similar burden on current residents. The same standards should be used to identify and quantify the costs of providing past and current services. Though the resulting cost calculations may not be totally accurate, the incentive to pile costs on past residents through expansive definitions of “current expenses” or the use of questionable actuarial assumptions should be offset by the burdens those definitions and assumptions would impose on current and future residents—residents who will vote in future elections.

D.     Allocating Tax Liabilities

Once one determines the total amount of tax to be raised with respect to any given year, a further question arises: how to allocate that amount between that year’s taxpayers. Though the answer will depend on the type of tax involved (property tax or income tax) and on the original rate structure, one concern stands out. With the benefit of hindsight, a polity can identify favored and disfavored groups and target the disfavored with a tax burden that would never have been thought acceptable if levied on a prospective basis. The allocation rules should try to reduce such exploitative behavior.

Most but not all states have income taxes. Some states, like the federal government, employ graduated rate structures for those taxes. Those states would face a choice as to how to allocate the additional tax liability between the rate brackets. Several plausible methods of adjusting the rate structure exist. For example, suppose a state (originally) levied an income tax of five percent on income up to $50,000 and seven percent on income above that amount. As a result of the graduated rate structure, 50% of the total income tax revenues came from the 70% of the population whose income was at or below $50,000 per year, with the remainder coming from the 30% with incomes above $50,000. Assume for this example that the amount of tax revenue necessary to recapture excess benefits could be raised by increasing everyone’s tax liability by three percent—that is, the five percent rate could become eight percent while the seven percent rate would become ten percent. However, raising everyone’s tax rate by three percent would change the overall distribution of tax liabilities; now, 60% of the total tax revenue would come from $50,000 or lower households, while only 40% would come from wealthier taxpayers. Meanwhile, to raise the same amount of tax revenue while matching the original income distribution would require raising the five percent rate to six percent and the seven percent rate to 12%. Yet another alternative would be to maintain the ratio between the lower and higher tax rates, i.e., the lower rate would be five-sevenths of the higher rate. Then the lower rate would go up to 7.8% while the higher rate would be 11%. It is probably impossible to determine which of the three alternatives best described the original political bargain, assuming one’s aim is to replicate it in the context of the larger revenue requirements. There is probably no alternative to leaving the distribution question to the political process. If worried that the results of this process would be affected by hindsight knowledge of which taxpayers left and which remained, it might be worthwhile to impose a consistency requirement here as well, such that any principle for establishing the new rates be applied to establish the rates going forward, at least for a reasonable period of time.

But it is not only individual residents who pay income taxes. Most states with income taxes levy corporate income taxes as well as individual income taxes. And there is no reason to exempt corporations from retroactive tax rate increases; like individuals, they benefit from a variety of state services.103 But that raises the question of how much of the overall tax increase ought to be borne by such taxpayers. Corporate taxpayers would be particularly attractive targets of retroactive taxes since the retroactive nature of such a tax forecloses the threat that such taxpayers move their business activities outside of the jurisdiction to reduce their tax liability. Moving income-producing operations to another lower-tax jurisdiction would not reduce or avoid a retroactive tax, which would be levied on income derived from in-state sources in years prior to such a move. Indeed, as with individuals, the imposition of a retroactive tax may reduce the incentive for corporations to move their business activities outside a state by allowing that state to hold down tax rates applicable to future corporate income. The impersonal nature of corporations also makes them an attractive target; although one can envision television ads claiming that a retroactive tax liability will upset an individual taxpayer’s college or marriage plans, or lead to missed mortgage payments or the like, corporate pleas for fiscal relief will lack any similar emotional tug. Although one does not ordinarily think of corporations as needing political protection against exploitative behavior, in this context it may be worth thinking about the imposition of some sort of consistency rule. The rule could take one of two forms. Either the percentage of the additional revenue raised from corporations should equal (or at least not exceed) the percentage of the original income tax revenue derived from the corporate income tax, or the tax rate imposed on a retroactive basis could be capped at the rate imposed on such income on a prospective basis. Either rule would keep corporations from being singled out for especially unfavorable treatment.

Some states levy property taxes in addition to or in lieu of income taxes.104 These do not tend to include graduated rates; however, many have either different rates or different assessment rules for different categories of property. Residential property, for example, is often taxed on more favorable terms than commercial property. Agricultural and manufacturing property also may be subject to alternative regimes.

If the difference plays itself out in assessment practices, retroactive taxation does not raise any special distributional issues. Across-the-board rate increases should suffice to replicate the pre-existing consensus regarding the relative contributions of different property classes and taxpayers. If different classes of property are subjected to different tax rates, however, polities will be faced with the same sorts of distributional issues as would be faced by jurisdictions with graduated income tax systems. Should all classes of property be subjected to identical rate increases, should an attempt be made to replicate the earlier distribution of tax liabilities among property classes, or should the ratio of tax rates as between classes remain the same? Underlying legal documents may mandate a choice of one particular alternative, but otherwise the decision among them has to be left to the polity, perhaps subject to a consistency requirement similar to that described for use in connection with graduated income taxes.

IV.     The Legality (or Not) of Retroactive Taxation

Even if the retroactive tax regime outlined above is economically sensible and politically acceptable, it may face legal impediments. These potential impediments stem from two distinct sources: the due process clauses found in the federal and most state constitutions and various tax caps found in many state statutes and constitutions. The tax caps found in state constitutions are likely to be the most serious legal impediments to such schemes. Fortunately, relatively few states have caps that would interfere with such a taxing scheme.

A.     Due Process Concerns

The regime sketched out above may be seen as troubling for two different reasons. First, it involves the imposition of tax on nonresidents, who may be seen as outsiders in need of protection against exploitation. Second, it involves the imposition of what appears, in many respects, to be a retroactive tax. As detailed below, these objections can be elided for all but a short transitional period if the taxing scheme is adopted prior to the triggering fiscal crisis. However, in the world in which we presently live, that “short transitional period” may be quite important, since states are unlikely to adopt these regimes before they find themselves in a financial crisis. Fortunately, the very factors that justify the proposed taxing regime should also protect later-adopted schemes against constitutional invalidation.

1.     Due Process Concerns Outside the Transitional Period

A tax is not considered to be “retroactive” merely because it changes the tax effect of earlier-concluded transactions when the possibility of such a change was known at the time of the earlier transaction.105 Indeed, the tax treatment of many transactions depends on later-developed facts. For example, the characterization of distributions from certain pension accounts and insurance policies depends on whether the amounts are returned within a certain time period. For many years, the tax treatment of transfers denominated as gifts depended on whether the donor survived for three years following the property transfer.106 Courts considering challenges to such gift tax provisions point out that those rules, unlike truly retroactive rules, do not interfere with settled expectations, nor can taxpayers claim lack of notice.107 In the event a jurisdiction encounters severe financial distress after enacting a retroactive tax regime, no taxpayer should have a settled expectation of avoiding the tax increases it mandates

Nor can former residents affected by such a regime claim that they were unrepresented in the political process, and thus deserving of judicial protection under the Due Process Clause, if they were residents of the state either in the year in which such a statute was enacted or in some later year. If present during the year of enactment, they would have had as much opportunity for input into the legislation as any other resident. Further, they likely would have had the right to vote for the state legislators that enacted the legislation. Later-arriving residents are generally bound to laws enacted before their arrival; their objection to a retroactive tax would be no more sustainable than those of out of state residents made subject to “welcome stranger” property assessment rules.108

The only situation in which residents or former residents could make even a colorable claim of surprise or unfairness—and thus have a colorable claim of a due process violation—would be if and to the extent such a regime applied to pre-enactment tax years. For example, some taxpayers may leave a jurisdiction shortly prior to the enactment of a retroactive tax regime, and then find themselves subject to taxes levied under it. However, as explained in the next section, although colorable, this claim is hardly likely to be successful.

2.     Due Process Concerns Within the Transitional Period

As has often been pointed out, many tax (and legal) rules have retroactive effects, so that it may be hard to meaningfully distinguish between retroactive and purely prospective taxation.109 Nonetheless, courts have repeatedly expressed discomfort with tax provisions—and other forms of economic legislation—which are explicitly retroactive in form.110 One concern is that the benefit of hindsight allows more effective targeting of disfavored groups.111 Even absent offensive targeting, such legislation may “deprive citizens of legitimate expectations and upset settled transactions.”112 Despite these concerns, however, courts (including the Supreme Court) have repeatedly upheld retroactive tax legislation in the face of due process challenges.113 There is little reason to expect a different outcome should some version of this tax regime be litigated.

To run afoul of the federal Due Process Clause, a retroactive tax must be deemed “harsh and oppressive” or “arbitrary and irrational.”114 This test is applied both to the tax rule itself and to the decision to impose it on a retroactive basis.115 Thus, according to the majority in United States v. Carlton, the most recent Supreme Court case involving a due process challenge to a retroactive tax amendment,116 not only must the statute as a whole be “supported by a legitimate legislative purpose furthered by rational means”117 but the retroactive aspect of the legislation must separately qualify under this test.118 Specifically, the retroactive nature of the legislation itself must be a rational means of achieving a “rational legislative purpose.”119

Carlton involved “curative”120 legislation—the retroactive amendment of a tax statute that proved to have a much broader application than was envisioned by Congress when the original statute was passed.121 However, retroactive tax legislation need not be “curative” to be acceptable. Courts have long accepted raising revenue as a rational purpose for a tax statute.122 And it is at least as rational to raise revenue from past residents who failed to pay the full costs of the services bestowed upon them as from future residents who did not receive the benefits of such services. Indeed, it is hard to describe legislation requiring residents to pay their own way (rather than stick someone else with the costs of providing their benefits) as “harsh and oppressive.” The facts here are quite unlike those in Eastern Enterprises v. Apfel, in which the Supreme Court struck down a legislative attempt to retroactively expand employers’ liability for providing health care to retired coal miners and their dependents to an employer who had exited the coal business 35 years before—well before the miners’ labor agreement provided any employer obligation to provide retiree health care.123 In that case, the Court found the provision constituted a “regulatory taking” because it “place[d] a severe, disproportionate, and extremely retroactive burden on [the plaintiff].”124 Because the suggested taxing regime only calls for past residents to pay for the difference between the taxes they paid and the benefits they received, the facts are much closer to those in Usery v. Turner Elkhorn Mining Co.125 In that case, the Court upheld provisions of the Black Lung Benefits Act of 1972, which required coal operators to compensate former miners for death or disability caused by black lung disease.126 The Court reasoned that the retroactive change in liability rules did not violate due process because it was “a rational measure to spread the costs of the employees’ disabilities to those who have profited from the fruits of their labor.”127 As explained in Part II, retroactive increases in tax liability are a rational mechanism for spreading the costs of government services to those who benefited from them.

The suggested three-to-five year look-back period is longer than the period of retroactivity axiomatically accepted as acceptable in tax cases.128Carlton itself involved only a “modest”129 period of retroactivity, of slightly over one year.130 However, there is no authority for the proposition that the period of retroactivity must be limited to one or two years, and courts have found longer periods of retroactivity to meet due process standards.131 Most recently, the Michigan Court of Appeals upheld in the face of due process and other constitutional challenges Michigan’s 2014 decision to rescind its membership in the Multistate Tax Compact (“MTC”) effective January 1, 2008—a rescission that had the effect of forcing taxpayers to allocate their income between Michigan and out-of-state sources under the formula prescribed by the State rather than that allowed by the MTC.132 The court specifically found that the “6 1/2-year retroactive period was sufficiently modest relative to the time frames of other retroactive legislation that have been upheld by Michigan courts, federal courts, and other state courts.”133 Although this case is being appealed to the Michigan Supreme Court, and may well be appealed up to the Supreme Court, it illustrates that at least some courts are sympathetic to extensive periods of retroactivity.

Moreover, moving beyond the tax context to other forms of economic legislation, much longer periods of retroactivity have been deemed to conform to due process standards.134 Given that the Supreme Court has stated that the due process standards for retroactive tax legislation should be the same as that for other forms of economic legislation,135 three to five years should not be problematic. Indeed, it is hard to argue that anyone can have “settled expectations” regarding their tax liability until that year’s statute of limitations for tax adjustments has closed—and in most cases, the statute of limitations will be either three or six years.

3.     The Extension of Retroactive Taxation to Nonresidents and Former Residents

Even if the retroactive nature of the tax is unproblematic, its application to nonresidents136 may give rise to an additional due process objection.137 Those nonresidents will be subject to tax without having any input into the political process through which the tax was enacted. Hence, the argument could be made, that this would constitute “taxation without representation”—a legal bugaboo at least since the Declaration of Independence.138

However, nonresidents are routinely subject to state (and local) taxation despite lacking formal access to political power. For example, states routinely tax locally generated income derived by nonresidents. Due process requires only “some definite link, some minimum connection, between a state and the person, property or transaction it seeks to tax”139 and that “the income attributed to the State for tax purposes must be rationally related to ‘values connected with the taxing State.’”140 Residency in the state during the period the income was earned (or the property was held) certainly rises to the level of a “definite link” or “minimum connection,” as does income derived from sources within the state,141 and it is completely rational for a state to require those residents to pay the full cost of benefits provided to them. As long as the retroactive tax increase does not discriminate against nonresidents or former residents (i.e. it applies equally to continued residents), it should not be viewed as any different from a nonretroactive source-based tax, which is likewise enacted and imposed without any direct political input from nonresident taxpayers. In short, it should be no more offensive than the original tax would have been had it been levied at the higher rate.

It is critical, however, that the terms of the retroactive tax be the same for nonresidents, current residents and former residents of the state. That is, if financial distress in 2017 triggers a 2% increase in 2012’s income tax rate, that 2% increase must be applied without regard to the taxpayer’s state of residence in 2017 (or for that matter, in 2012). To put it another way, in-state (in 2017) taxpayers must not be excused in whole or in part from the additional levy relating to their 2012 income by virtue of their in-state status. The imposition of heavier tax burdens on former residents (relative to those current residents who were residents in the earlier year) would be viewed as an “exit tax.” Exit taxes not only impermissibly interfere with the “right to travel,”142 but could be deemed to “discriminate” against nonresidents and thus to violate  the Commerce Clause143 or the Privileges and Immunities Clause (for individuals).144 A generally applicable retroactive increase in income tax rates, by contrast, does not impose a burden on a taxpayer’s decision to move to another jurisdiction as neither a  move, nor a decision not to move, affects the amount of tax due under the regime; nonresidents and residents would pay the same amount of tax on past in-state earnings.  Although a retroactive tax regime may discourage some residents from leaving a financially troubled jurisdiction,145 it would do so only by providing that jurisdiction with a mechanism for holding down the rate of tax imposed on future income. If attracting residents through the maintenance of uniform, but relatively low, tax rates constituted a Commerce Clause violation, all variations in state tax structures would have to be struck down as unconstitutional.

Although federal law should not be a barrier to the enactment of the tax regimes advocated in this Article, state tax limitations may present a serious problem. These limitations, and their potential impact, are discussed next.

B.     State Tax and Expenditure Limitations

The adoption by referendum of California’s Proposition 13 in 1978 signaled the beginning of populist “tax revolts,” attempts to limit and even roll back the size and cost of state and local governments by imposing some combination of spending limits, tax caps, or procedural impediments to tax increases.146 By 2010, 30 states were subject to some form of tax or spending limitation,147 and that number continues to climb.148 The language of the various limitations (generally referred to as “TELs”)149 differs from state to state, as do the interpretations of that language, so that it is impossible to generalize about their impact on retroactive tax schemes.150 Nonetheless, it is likely that some of these limitations would present obstacles, and perhaps insuperable obstacles, to the adoption of such schemes in some states.

1.     Expenditure Limitations

The most common form of state limitation is the expenditure limit.151 Twenty three states had constitutional or statutory limitations on state expenditures in 2010.152 Some state constitutions limit state governmental spending to a percentage of total state personal income,153 while others limit the growth in spending over a base year amount by a factor reflecting some combination of increases in inflation, population, and personal income.154 Others impose similar limits on government appropriations.155 And in others, these limitations are imposed by statute.156 In addition, several states have balanced budget requirements.157

In theory, these limitations should not interfere with the operation of a retroactive taxation scheme. After all, retroactive taxation is about raising revenue—and specifically the revenue necessary to pay expenses already authorized by the state legislature. However, many of the expenses that need to be paid with the proceeds of the retroactive taxes are expenses which were incurred “off the books,” outside of the normal appropriations and budget process. For example, the financial problems caused by underfunded pension and retiree health benefits stem from the fact that these benefits were never budgeted nor paid for out of appropriated funds. Actual payment of the sums owed will necessitate turning these unacknowledged liabilities into actual, acknowledged “expenses” and require “appropriations” of cash, which (unlike the unacknowledged liabilities) could count against the constitutional and statutory limits on state expenditures and appropriations. Unless states can come up with offsetting reductions in other, traditionally budgeted (and appropriated) expenses, a number of states will find themselves bumping up against these caps. Of course, this problem would be encountered no matter the source of the funds used to pay the State’s obligations; the underlying problem is the State’s hidden expenditures.

Of course, not every distressed state is subject to one of these caps. Illinois, for example, is not—but New Jersey is.158 Some of the caps have relief provisions built in, which may provide the flexibility necessary to effectively use the funds derived from retroactive taxes.159 And statutory caps can be modified by legislation enacting a retroactive taxation scheme.160 Ultimately, though, any state considering the adoption of a retroactive tax will have to analyze, and perhaps amend, their expenditure and appropriation caps.

Spending and appropriation caps are merely one tool that has been used to restrain state spending. The next tool, revenue limitations, may be the more significant danger to retroactive tax regimes.

2.     Revenue Limitations, Including Tax Caps

Many states have adopted limitations on their revenue raising powers instead of or in addition to spending caps. There are three basic types of revenue limitations. The first is a generalized revenue “cap” which limits the amount of revenue the state can raise.161 Some states even require that any excess above the stated cap be returned to taxpayers.162 More recently, states have begun amending their state constitutions to include caps on the rate of state income taxes.163 Finally, about a third of states have imposed procedural limitations on tax increases. Three states—Colorado, Missouri and Washington—require a supermajority of referendum voters to approve tax increases. Sixteen states require their legislatures approve tax increases by a supermajority vote, ranging from three-fifths to three-fourths.164

The generalized revenue caps come in two varieties. Some states limit general state revenues to a stated percentage of the previous year’s state personal income.165 More frequently, though, the revenues collected in a stated year serve as the base amount; the revenue raised in subsequent years cannot exceed this base amount, adjusted for some combination of population, inflation or personal income growth.166 Almost all of these revenue limitations have been embedded in state constitutions.167

Since a retroactive tax proposal would involve an actual (or potential) tax increase, enactment would require fulfillment of any applicable supermajority voting regime. This will undoubtedly make its adoption more difficult in some states. In most cases, bipartisan support would be needed to pass the necessary legislation.

The bigger obstacle, though, may be the revenue caps. In many states already bumping against their caps,168 nothing less than a state constitutional amendment would be required to make use of additional revenues derived from retroactive tax schemes. In their absence, the most such schemes would be allowed to do would be to shift tax burdens from current residents to past ones—and even that would be problematic if prior year revenues were bumping against the cap. Fortunately, the constitutions of the most fiscally distressed states do not at present contain such caps. But there is no guarantee this state of affairs will continue. Taxpayers in distressed jurisdictions may turn to such devices in an effort to stave off tax increases—or jurisdictions whose revenue generating capacities have been crippled by such rules may find themselves falling into financial distress.

In sum, federal law should not impede the enactment of retroactive tax schemes, but some state TELs might. However, many states, including some of the most fiscally distressed, do not have TELs, and in others, the TEL is comparatively weak. At present, then, there is no legal impediment to enacting a retroactive taxation scheme in many jurisdictions.

V.     Conclusion

The institutional and economic factors underlying the fiscal distress afflicting many states of the United States are many and varied. But at least part of the problem can be laid at the feet of voters who continuously kicked financial cans down the road in hopes that they would be long gone before those costs became inescapable. One way of discouraging such behavior is to make escape from inevitable liabilities less lucrative through the imposition of retroactive taxes—in effect retroactively imposing an effective balanced budget requirement. In addition to encouraging more responsible civic behavior, such taxes would, like claw backs in bankruptcy, effect partial restitution of unjustly derived gains.

This proposal is, of course, no panacea. In addition to the fact that revenue shortfalls accumulated over generations cannot (and should not) be recouped by increasing a few years’ worth of taxes, there is no guarantee that the funds raised by retroactive tax regimes will be used to pay down accumulated indebtedness. The additional funds might be used to pay future expenses, or simply squandered. But the alternatives—none of which interrupt the current destructive political dynamic—are hardly attractive. Rather than reward feckless politicians and voters with a bailout funded by their more responsible counterparts, Congress might do better to think about ways it might help Illinois and California and New Jersey collect taxes from former residents who have tried to escape the financial messes to which they have contributed by moving to states that did a better job of keeping their financial houses in order. It may even keep those expatriates from encouraging their new states to embark on similarly destructive financial paths.


  1. [1]. See Richard F. Dye, Balancing Illinois’ Budget Will Be Painful, Guaranteed, St. J.-Reg. (Mar. 25, 2014, 1:07 AM), (“[O]fficeholders and candidates argu[e] that cutting ‘waste, fraud and mismanagement’ in state programs will solve our budget woes.”); Michael Tanner, Math vs. Myth, Nat’l Rev. (Oct. 19, 2011, 4:00 AM), www.national (“We can balance the [federal] budget by cutting ‘fraud, waste, and abuse.’ This [idea] is the Republican flip side of the Democrats’ reliance on higher taxes . . . .”).

  2. [2]. See Clayton P. Gillette, Fiscal Home Rule, 86 Denv. U. L. Rev. 1241, 1259 (2009) (“Debt can be a dangerous tool in the hands of local officials who have incentives to spend money in the short term, especially money that has to be repaid only when they have left office.”); Maria O’Brien Hylton, Combating Moral Hazard: The Case for Rationalizing Public Employee Benefits, 45 Ind. L. Rev. 413, 415 (2012) (“[T]he core moral hazard problem . . . [is] the apparently irresistible tendency of state legislators and executive branch officials to spend taxpayer dollars to enhance benefits and decrease contributions during flush economic times in exchange for voter support at the polls.”); David A. Skeel Jr., States of Bankruptcy, 79 U. Chi. L. Rev. 677, 688 (2012) (“[P]oliticians have strong incentives to spend in the present and push their repayment to the future.”); Matt Bai, State of Distress, N.Y. Times Mag. (Oct. 20, 2009), (“[New Jersey’s governors] find themselves retreating to a kind of fiscal Honalee, a make-believe world in which the state can magically raise less revenue and spend ever more of it.”).

  3. [3]. See Julie A. Roin, Privatization and the Sale of Tax Revenues, 95 Minn. L. Rev. 1965, 2029–31 (2011) (advocating for rules aimed at increasing the transparency of privatization transactions).

  4. [4]. For example, Illinois’ fiscal woes can be traced to the 1990s, if not before. See State Budget Crisis Task Force, Illinois Report 16 (2012),
    Report.pdf (“[T]he origins of the structural gap between spending growth and sustainable revenues can be traced to the 1990s.”); Eric M. Madiar, Illinois Public Pension Reform: What’s Past is Prologue, 31 Ill. Pub. Emp. Rel. Rep. 3–8 (2014),
    pension%20history.pdf (tracing the “chronic underfunding” of Illinois state and municipal pension-plan funding from 1917 to the present).

  5. [5]. See, e.g., Skeel, supra note 2, at 679 & n.8 (“Projecting a $25 billion deficit last year, California Governor Arnold Schwarzenegger proposed to sell the San Francisco Civic Center and other state properties to raise funds. . . . Governor Jerry Brown later canceled the sales.”); see also Hal Dardick & Bill Ruthhart, Mayor’s Record on Debt Mixed, Chi. Trib. (Sept. 6, 2015),
    usa/chicago-tribune/20150906/283283161450400 (describing the “scoop-and-toss,” long-term borrowing for short-term expenses and other ploys used by Chicago mayors to delay property-tax increases); Cezary Podkul & Allan Sloan, Behind Christie’s Budget Claims, a More Controversial Legacy, Wash. Post (Apr. 17, 2015, 4:05 PM),
    _story.html (“[Governor Christie] has resorted to many of the financial maneuvers used by some of his predecessors: reducing state payments to pension plans, shifting money out of trust funds dedicated for specific purposes and borrowing to patch chronic budget gaps.”); Elizabeth Lesly Stevens, State Poised to Sell Trophy Buildings to Unidentified Investors, N.Y. Times (Dec. 25, 2010), http://www.nytimes.

  6. [6]. For example, Illinois voters recently elected Bruce Rauner as governor, who supported allowing a scheduled drop in the rate of the Illinois income tax to proceed, rather than the incumbent, Pat Quinn, who supported extending the higher tax rate. See Bob Secter & Rick Pearson, Illinois Income Tax Rate Falls by 25 Percent, Chi. Trib. (Jan. 1, 2015, 6:00 AM),
    local/breaking/ct-illinois-income-tax-rate-falls-met-20141230-story.html (“The financial wisdom of allowing rates to drop, significantly reducing state revenue, was the subject of heated debate during the recent battle for governor between incumbent Democrat Pat Quinn and Bruce Rauner, the Republican who defeated him. Quinn said the state couldn’t afford it, while Rauner positioned himself in favor of tax reductions . . . .”).

  7. [7]. See Julie Roin, Planning Past Pensions, 46 Loy. U. Chi. L.J. 747, 773 (2015) (Illinois “excludes from its income tax base payments made under public pension plans, private pension plans, Individual Retirement Accounts (‘IRAs’), 401(k)s, social security benefits, redemptions of U.S. retirement bonds, qualified annuities, and Keogh plans.”).

  8. [8]. Illinois is not the only state that grants special tax favors to retirees. Four other states also exclude all forms of retirement income from their income tax bases. Many states have less extensive, but still substantial, exclusions for retirement income. See Nat’l Conference of State Legislatures, State Personal Income Taxes on Pensions and Retirement Income: Tax Year 2014, at 2 (2015), (reporting that 36 of 41 states with personal income taxes “offer exclusions for some or all specifically identified state or federal pension income or both, a retirement income exclusion, or a tax credit targeted at the elderly”).

  9. [9]. Indeed, the exodus may have already started. Several of the states experiencing financial distress have suffered recent population losses. See Elizabeth Barber, Got the Urge to Leave Your State? You Must be from Illinois, Christian Sci. Monitor (May 1, 2014), (“About half of residents in Illinois, Connecticut, and Maryland would move if they could, a poll finds.”); Editorial, Goodbye, Illinois: Residents are Leaving for Other States, Chi. Trib. (Jan. 6. 2015, 7:11 PM),; IRS 2012-2013 State Migration Data—NY Down Big, TX & FL Shine, Clean Slate Tax (Aug. 20, 2015), (referencing IRS data for 2012–2013 showing New York, Illinois, California, New Jersey, and Pennsylvania as suffering the greatest out-migrations); Kim Janssen, Illinois Still Losing Residents to South, West: Surveys, Chi. Trib. (Jan. 4, 2016, 10:10 AM), (“Two recent surveys—one published by the U.S. Census Bureau and another by movers United Van Lines—both put Illinois among the top three states losing residents to the rest of the U.S. in 2015.”).

  10. [10]. Current residents would be liable for all prospectively levied taxes. However, new entrants in a jurisdiction would not be subject to retroactive taxes, unless they had previously earned income in that jurisdiction, in which case that income would be subject to the higher tax rate.

  11. [11]. Although leaving the jurisdiction for a lower tax jurisdiction would decrease taxes due on subsequently earned income, it would not lower the taxpayer’s obligation to pay the retroactive tax. In addition, a retroactive tax may provide enough revenue to allow the jurisdiction to levy lower rates prospectively. See infra Part II.B.1. Existing residents would still benefit financially from fleeing to lower tax jurisdictions, but not by as much. As a result, other considerations (job or family or friendship ties) militating in favor of remaining in the distressed jurisdiction may outweigh those financial benefits. On the margin, fewer residents should leave distressed jurisdictions.

  12. [12]. Chapter 9 of the Federal Bankruptcy Code, while allowing states to authorize their municipalities to take advantage of the Code’s bankruptcy mechanisms, does not extend the same bankruptcy privilege to the states themselves. Some academics and politicians have advocated the enactment of provisions for state bankruptcies, but Congress has yet to take any steps to do so. See Skeel, supra note 2, at 679 (“Starting in late 2010, a few politicians and commentators insisted that state bankruptcy was an idea whose time had now come.”).

  13. [13]. See id. at 707 (stating that “[s]ome critics question the constitutionality of a state-bankruptcy regime” because it would impermissibly interfere with state sovereignty and Contracts Clause obligations); see also Jack M. Beermann, The Public Pension Crisis, 70 Wash. & Lee L. Rev. 3, 76 (2013) (“It is not absolutely clear that the approval of municipal bankruptcy is precedent for finding no constitutional difficulty with state bankruptcy.”). However, others believe that a “well-crafted state-bankruptcy law” would survive constitutional challenges. Skeel, supra note 2, at 710.

  14. [14]. In theory, any plan of reorganization by a private debtor in possession is supposed to guarantee creditors will receive at least as much as they would have received had the enterprise been liquidated—but in the absence of firm market values for the corporate assets, those amounts may be subject to dispute.

  15. [15]. See Michelle Wilde Anderson, The New Minimal Cities, 123 Yale L.J. 1118, 1122 (2014) (“While laws provide an entitlement to a public education, and we have long struggled to interpret what constitutes a legally adequate education, there is little to nothing to indicate what other services the local public sector must provide. . . . This is a humanitarian question, but it is also a doctrinal challenge.”).

  16. [16]. See 11 U.S.C. § 544 (2012) (bankruptcy trustee may avoid fraudulent conveyances); 11 U.S.C. § 547 (bankruptcy trustee may avoid certain transfers made by the debtor within 90 or 365 days preceding the filing of the bankruptcy petition); see also infra Part II.A.2.

  17. [17]. See Matthew D. Slepkow, Note, Resurrecting the Challenge Against Retroactive Estate Tax Legislation: Acquiescing to the Holding of United States v. Carlton—Over My Dead Body, 3 Roger Williams U. L. Rev. 119, 127 (1997) (“Retroactive legislation appears to offend the natural law of decency . . . .”); see also Saul Levmore, The Case for Retroactive Taxation, 22 J. Legal Stud. 265, 265 (1993) (“[R]etroactive taxation . . . is generally regarded as abhorrent, unwise, and even illegal.”).

  18. [18]. Versions of this humorous expression have been traced as far back as the 1930s. Don’t Tax You. Don’t Tax Me. Tax that Fellow Behind the Tree, Quote Investigator (Apr. 4, 2014),

  19. [19]. This may be an overstatement. If knowledge of the pre-existing debt is widespread, its economic burden might be capitalized in housing prices. That is, existing residents could “pay” for part of the debt through a reduction in the value of their homes when they sell, while new residents could be “compensated” for assuming that debt through lower home-acquisition prices. The extent to which public debt is incorporated in housing prices is unclear; recent studies suggest that the residential-housing market is, in general terms, remarkably inefficient. See Robert J. Shiller, The Housing Market Still Isn’t Rational, N.Y. Times: Upshot (July 24, 2015),
    the-housing-market-still-isnt-rational.html (“Efficient markets require the possibility of selling short . . . . but such negative bets cannot easily occur in the housing market.”). Such incorporation requires these public debts to be salient to bargaining homeowners. As discussed infra note 57 and accompanying text, in the past, states deliberately hid pension-funding shortfalls, making such salience, and thus incorporation, unlikely. Recent changes in governmental accounting standards as well as increased press coverage of relative state and local government indebtedness may increase future salience and thus capitalization. See Roin, supra note 7, at 749–51 (describing the history of pension accounting standards). However, since the infinitely more transparent market for municipal bonds fails to completely capitalize applicable tax benefits, I feel safe in assuming that the housing market will remain at least somewhat imperfect, so that an additional legal remedy is warranted. More generally, if one is willing to assume perfect markets, most areas of law—and legal scholarship—would be unnecessary. See Marvin A. Chirelstein & Lawrence Zelenak, Federal Income Taxation 493 (2015) (“Unfortunately for all this heady analysis, in the real world it appears that tax preferences are not fully capitalized.”).

  20. [20]. In fiscal year 2012, for example, formal state debt amounted to $488 billion while unfunded pension liabilities (exclusive of unfunded retiree health benefits) totaled $894 billion. John A. Sugden et al., Standard & Poor’s Rating Servs., U.S. State Pension Funding: Strong Investment Returns Could Lift Funded Ratios, But Longer-Term Challenges Remain 15 (2014), http://; see also Jennifer Weiner, Assessing the Affordability of State Debt 6 fig.B-1 (2013),
    economic/neppc/researchreports/2013/neppcrr1302.pdf. In Illinois, general-obligation and Build Illinois Bonds represent slightly less than 24% of its $127 billion of total debt; the remainder of the debt consists of unfunded, accrued actuarial pension liability. Benjamin VanMetre, Illinois Drowning in Debt: $127 Billion and Counting, Ill. Pol’y (Jan. 7, 2014), Its unfunded pension debt alone amounts to $7,346 per Illinois resident. Roin, supra note 7, at 751–52. Once unfunded local pensions are taken into account, each Chicago resident owes $18,924. Id. Other states have similar ratios between bond and unfunded pension indebtedness. See Cory Eucalitto, State Budget Solutions’ Fourth Annual State Debt Report, Am. Legis. Exchange Council (Jan. 8, 2014), (showing a table with a state-by-state breakdown of general-revenue bonds). These sources may understate the extent of informal (unfunded pension and other post-retirement benefits, such as retiree health benefits) debt. Some believe that states (and localities) vastly understate the true amount of underfunding of these benefits by calculating those liabilities using unduly high discount rates. See Frank Russek, Cong. Budget Office, The Underfunding of State and Local Pension Plans 3–6 (2011),
    120xx/doc12084/05-04-pensions.pdf (detailing the differences between approaches for determining discount rates); Robert Novy–Marx & Joshua D. Rauh, The Liabilities and Risks of State-Sponsored Pension Plans, 23 J. Econ. Persp. 191, 192 (2009) (“We show that government accounting standards require states to use procedures that severely understate their liabilities.”).

  21. [21]. Not all jurisdictions have that option. Some state constitutions contain provisions which effectively prevent not only the “readjustment” of already-earned pension benefits, but even the ability to make downward adjustments in existing employees’ future pension accruals. Fields v. Elected Officials’ Ret. Plan, 320 P.3d 1160, 1165–68 (Ariz. 2014) (finding that legislative changes to statutory formula for pension benefits violated Arizona constitution); In re Pension Reform Litig., 32 N.E.3d 1, 18­–29 (Ill. 2015) (finding the legislature’s attempt to reduce public employee pensions violated the Illinois constitution); Amy B. Monahan, Statutes as Contracts? The “California Rule” and Its Impact on Public Pension Reform, 97 Iowa L. Rev. 1029, 1032 (2012) (“[C]ourts in California and the twelve other states that have adopted California’s precedent have held not only that state retirement statutes create contracts, but that they do so as of the first day of employment. . . . [C]ourts interpreting the California Rule have held that the contract protects not only accrued benefits (a relatively uncontroversial position) but also the rate of future accrual.” (footnote omitted)).

  22. [22]. Failing to make good on its promises of deferred compensation means that employees who thought they were receiving $40,000 per year will actually receive a lesser amount, perhaps just $30,000. Even if successful in the short term, this option works only once. Knowing that deferred compensation promises cannot be trusted, employees will no longer accept them as part of their compensation packages. They will insist on receiving all $40,000 in cash, and try to self-finance their retirement. They may not be particularly successful at doing so (ordinary Americans are remarkably bad at saving for their retirement). See Barbara A. Butrica & Karen E. Smith, 401(k) Participant Behavior in a Volatile Economy 3 (2012), (“But recent studies have revealed less than encouraging information about retirees’ ability to adequately plan for retirement.”); see also Susan J. Stabile, Is It Time to Admit the Failure of an Employer-Based Pension System?, 11 Lewis & Clark L. Rev. 305, 311–12 (2007) (describing extent of failures to participate and to make sufficient contributions to 401(k) plans); Stabile, supra, at 313 (decrying employees lack of “knowledge to make the necessary financial decisions”). But they would still be better off than if they accepted $30,000 and an unenforceable promise of deferred compensation. Assuming society values maintaining adequate savings for retirement, of course, the socially efficient arrangement would involve a lower cash salary and fully funded pension.

  23. [23]. Although this Article uses the term “residents” throughout, services are rendered to, and income taxes imposed on, both actual residents and on nonresidents deriving income from in-state sources. There is no reason to exclude nonresidents subject to the state’s income tax on a source basis from any retroactive tax increase. Of course, the increase (like the original tax obligation) should be imposed only upon that portion of the taxpayer’s income that was originally subject to the state’s taxing authority.

  24. [24]. If one takes this argument seriously, governments—and indeed individuals—should rely exclusively on debt rather than taxes. Indeed, even the funds used to make interest payments on the debt should be borrowed. Outstanding debt should be rolled over as it comes due, since future generations will be “better able” to pay the principal. There is no logical stopping point, either in terms of amount of debt or time of payment. Although some people may be sufficiently optimistic to accept this approach, most (and certainly I) would balk at the implicit assumption that economic growth will continue indefinitely at the pace that has been sustained since the beginning of the Industrial Revolution, given that over the vast proportion of human history, per capita growth rates averaged approximately zero. See Angus Maddison, Contours of the World Economy 1–2030 AD: Essays in Macro-Economic History 69–70 (2007). Moreover, ability to pay is not the only concern when designing a tax scheme; one also should be concerned with discouraging inefficient behavior. The failure to link overall levels of taxation to the associated costs of governmental benefits can lead to social inefficiencies, including inefficient movements among jurisdictions and questionable decisions about the size and contour of governmental activities. See infra Part II.B.1, 3.

  25. [25]. For example, the five states with the highest per capita personal incomes in 1950 were Alaska, Delaware, Nevada, Connecticut, and California, while the top five in 1999 were Connecticut, Massachusetts, New Jersey, New York, and Maryland. See G. Andrew Bernat, Jr., Convergence in State Per Capita Personal Income, 1950–99, Surv. Current Bus., June 2001, at 36, 37 tbl.1. In 2014, the five states with the highest per capita income were Maryland, Alaska, New Jersey, Hawaii, and Connecticut. Delaware had sunk to 15th and Nevada to 26th. California was 10th. Richie Bernardo, 2014’s Richest and Poorest States, WalletHub (Oct. 8, 2014), Interestingly, state wealth is no guarantee of state fiscal solvency; California and New Jersey are among the most fiscally troubled states. See Vincent S.J. Buccola, An Ex Ante Approach to Excessive State Debt, 64 Duke L.J. 235, 237 (2014) (listing California, Illinois, Michigan and New Jersey as “the most troubled states” from a fiscal standpoint).

  26. [26]. Drew DeSilver, For Most Workers, Real Wages Have Barely Budged for Decades, Pew Res. Ctr. Facttank (Oct. 9, 2014), (“But after adjusting for inflation, today’s average hourly wage has just about the same purchasing power as it did in 1979 . . . . In fact, in real terms the average wage peaked more than 40 years ago . . . .”); see also Adam Davidson, Is College Tuition Really Too High?, N.Y. Times Mag. (Sept. 8, 2015), (noting that there has only been a slight increase in median family income over 40 years).

  27. [27]. See Davidson, supra note 26 (“Tuition at a private university is now roughly three times as expensive as it was in 1974 . . . [and] public tuition . . . has risen by nearly four times.”).

  28. [28]. See Anderson, supra note 15, at 1121 (“Cities undertaking austerity measures also shed their property—public assets like parks, pools, and government office buildings.”); Michelle Conlin, States and Cities Selling Public Assets to Cover their Costs, CNSNews (May 13, 2011, 9:04 AM), http:// (describing proposed and actual transactions).

  29. [29]. The recouped funds are redistributed to other claimants in bankruptcy proceedings. These provisions of bankruptcy law correspond to current theories of restitution. See Restatement (Third) of Restitution and Unjust Enrichment § 67 cmt.f (Am. Law Inst. 2011) (“[R]ule . . . § 67(1)(c) allows a defrauded investor without notice to retain payments received from the commingled fund, but only to the extent that such payments reduce the amount of the investor’s inchoate restitution claim against the wrongdoer. The effect of this rule is that an innocent payee may retain withdrawals or distributions up to the amount of his investment, but is liable in restitution for anything more.”).

  30. [30]. 11 U.S.C. § 548(a)(1) (2012).

  31. [31]. Id. § 544. Different states have different statute of limitation periods. See Investor Beware! What Everyone Should Know About Clawback Litigation, Bernstein-Burkley, P.C.,
    publications-list/investor-beware-what-everyone-should-know-about-clawback-litigation (last visited Nov. 10, 2016) (four years in Pennsylvania, six years in New York) [hereinafter Investor Beware!]. This period may be extended yet further—indeed, much further—if one of the creditors is the Internal Revenue Service. Several courts have held that a trustee in bankruptcy can step in the shoes of a federal creditor including the IRS to take advantage of the six year look-back period provided by 28 U.S.C.
    § 3304, part of the Federal Debt Collection Procedure Act, for clawing back fraudulent transfers. See Gordon v. Harrison (In re Alpha Protective Servs. Inc.), 531 B.R. 889, 906 (Bankr. M.D. Ga. 2015) (holding that longer look-back period applies); Ebner v. Kaiser (In re Kaiser), 525 B.R. 697, 713 (Bankr. N.D. Ill. 2014); Tronox Inc. v. Kerr McGee Corp. (In re Tronox Inc.), 503 B.R. 239, 272–75 (Bankr. S.D.N.Y. 2013); Anderson v. Architectural Glass Constr., Inc. (In re Pfister), 2012 WL 1144540, at *5 (Bankr. D.S.C. Apr. 4, 2012); Sergeant v. OneWest Bank, FSB (In re Walter), 462 B.R. 698, 704–06, 712 (Bankr. N.D. Iowa 2011); Allred v. Porter (In re Porter), 2009 WL 902662, at
    *20–21 (Bankr. D. S.D. Mar. 13, 2009). But see MC Asset Recovery LLC v. Commerzbank A.G. (In re Mirant Corp), 675 F.3d 530, 536 (5th Cir. 2012) (refusing to apply FDCPA to extend the statute of limitations period for purposes of section 544 claw back). At least one court has applied an even longer claw back period when the federal creditor was the IRS by treating the fraudulent transfer claim as a “collection suit” which, under section 6502(a) of the Internal Revenue Code, can be brought at any time within ten years of assessment of a tax, which in turn, under section 6501(a) must occur within three years of the filing of the return, or, under section 6901, four years if the assessment is made against a transferee. See Ebner, 525 B.R. at 709 (allowing IRS to serve as a “golden creditor” by extending statute of limitations to ten years for all creditors). Whether federal agencies will be among the creditors in state workouts is uncertain and indeed irrelevant; the point is simply that present law accepts the use of extensive claw back periods.

  32. [32]. Investors are entitled to raise an affirmative defense of “good faith” to forestall recoupment of distributions up to the amount of their original investment in such schemes. See Saul Levmore, Rethinking Ponzi-Scheme Remedies In and Out of Bankruptcy, 92 B.U. L. Rev. 969, 972 (2012); Investor Beware!, supra note 31 (describing “good faith” affirmative defense to claims for the return of their initial investment; no such defense exists for “fictitious profits”).

  33. [33]. Levmore, supra note 32, at 971 n.9 (citation omitted); see also Investor Beware!, supra note 31 (“The very nature of a Ponzi scheme is fraudulent; therefore in almost all cases the trustee or receiver can show proof of the intent to defraud based upon the admission of a principal that they perpetrated a Ponzi scheme.”).

  34. [34]. Interestingly, many of the “winning” investors in the most recent Ponzi scheme, the Madoff debacle, escaped the full force of these recoupment provisions because section 546(e) of the Bankruptcy Code specifically exempts the proceeds of transfers made by stockbrokers either as “settlement payments” or “in connection with a securities contract.” See In re Bernard L. Madoff Inv. Secs. LLC, 773 F.3d 411, 417 (2d Cir. 2014). Madoff’s company was a stock brokerage, and although in most cases the funds deposited by investors were not used to purchase actual securities, the agreements pursuant to which Madoff’s customers entrusted their money to the firm constituted “securities contracts,” so that sums paid to customers pursuant to them fell within the definition of “settlement payments.” Id. at 419–23. The court dismissed all the recoupment claims filed by the bankruptcy trustee aside from those falling under section 548(a)(1)(A)—transfers involving actual fraud made within two years of the bankruptcy filing. Id. at 423. This decision created a “new breed of ‘net losers’”—investors who settled claw back actions with the trustee before the issuance of the opinion. Bankruptcy Code § 546(e) Shields Madoff Ponzi Scheme Payments, Buchanan Ingersoll & Rooney PC (Mar. 17, 2015),

  35. [35]. See Weiner, supra note 20, at 20–21(using long-term borrowing to finance noncapital costs is “generally considered a poor financial practice”).

  36. [36]. Interest on debt incurred to reimburse governments for current expenses generally does not qualify for tax-exempt status. See Treas. Reg. § 1.150­-2 (1993). Nonetheless, some governments issue such bonds. See Heather Gillers, $730M Deal Shows How Chicago May Use Long-Term Debt for Short-Term Costs, Chi. Trib. (July 9, 2015, 10:53 PM), (“Governments must resort to taxable bonds when they want to borrow for costs with no benefit over an extended period.”).

  37. [37]. Although rumors of fraud periodically swirled around Madoff, the Securities Exchange Commission (“SEC”) investigated Madoff’s operations on several occasions and found nothing wrong. See Donald C. Langevoort, The SEC and the Madoff Scandal: Three Narratives in Search of a Story, 2009 Mich. St. L. Rev. 899, 900 (“Markopolos was the Boston-based financial analyst who early on came to the accurate conclusion that Madoff was probably running a Ponzi scheme, and starting around 1999 tried to convince the SEC to expose it. Although he was neither the first nor the only whistleblower, his efforts were the most detailed and persistent.”). It is hard to fault investors for failing to be more prescient (or competent) than the SEC; indeed, some may have relied on the SEC declining to take action against Madoff as a guarantee of his legitimacy.

  38. [38]. Nat’l Conference of State Legislatures, NCSL Fiscal Brief: State Balanced Budget Provisions 1 (2010),

  39. [39]. See id. at 2 (“The National Conference of State Legislatures (NCSL) has traditionally reported that 49 states must balance their budgets, with Vermont being the exception. Other authorities add Wyoming and North Dakota as exceptions, and some authorities in Alaska contend that it does not have an explicit requirement for a balanced budget.”).

  40. [40]. See id. (describing differences between states); id. at 5 fig.1 (illustrating the relative stringency of state rules, as measured by the Advisory Commission on Intergovernmental Relations). Loopholes exist in even stringent balanced-budget requirements. See Evan O’Connor, Caught Off-Balance: How Implementing Structural Changes to State Balanced Budget Requirements Can Foster Fiscal Responsibility and Promote Long-Term Economic Health, 56 B.C. L. Rev. 351, 369-76 (2015) (describing mechanisms used to undercut balanced-budget requirements).

  41. [41]. For example, Illinois’ constitution provides that “[p]roposed expenditures shall not exceed funds estimated to be available for the fiscal year as shown in the budget.” Ill. Const. art. VIII, § 2(a). However, borrowed funds count toward the funds deemed available for this purpose. See Richard F. Dye, How Can Illinois Improve Its Budget?, Reboot Ill. (Jan. 23, 2014),
    2014/01/23/can-illinois-improve-budget/4973 (“The balanced budget requirement in the Illinois Constitution . . . counts as revenue . . . spending pre-existing balances, borrowing or using one-time revenue, and incurring obligations that do not have to be paid until later.”). Illinois is not alone in this regard. See Cory Eucalitto, Unbalanced: Why State Balanced Budget Requirements Are Not Enough, Am. Legis. Exchange Council (Apr. 4, 2013), (discussing the use of these techniques in other states).

  42. [42]. See David Splinter, State Pension Contributions and Fiscal Stress 3 (Am. Econ. Ass’n Working Paper 2011), (confirming the work of others by showing that states with stringent balanced budget requirements, when fiscally stressed, “both undercontribute to their pensions and choose discount rates that obscure actuarial underfunding”).

  43. [43]. It is worth noting that under the federal rules for funding privately sponsored defined benefit pension plans, plan sponsors must amortize the cost of grandfathered benefits, see Peter J. Wiedenbeck, ERISA: Principles of Employee Benefit Law 260–61 (2010) (amortization period gradually shortened from 40 years to seven years), as well as cover the costs of benefits earned during the year. See id. at 264 (describing funding rules).

  44. [44]. See Thomas A. Garrett & John C. Leatherman, An Introduction to State and Local Public Finance: Part 1—Government Growth, Taxes and Tax Theory, Web Book Reg’l Sci.,
    Book/Garrett/chapterone.htm (last visited Nov. 10, 2016) (showing in figures 1 and 2 the growth in real per-capital state and local expenditures and as a percentage of GDP from 1960 to 1996); US State Government Spending Since 1900,, http://www.usgovernmentspending
    .com/state_spending_chart (last visited Nov. 10, 2016) (illustrating that state-government spending rose from less than 1% of GDP in 1900 to over 8% of GDP by 2000).

  45. [45]. See Madiar, supra note 4, at 9 chart 2 (illustrating the growth of Illinois pension liabilities from 1970 to 2013).

  46. [46]. See Lance Christensen & Adrian Moore, Reason Found., Best Practices in Pension Reform: Lessons Learned From Successful Reformers 13–28 (Sept. 2014), http://www.reason.
    org/files/best_practices_in_pension_reform.pdf (describing pension reforms adopted in Alaska, Rhode Island, Utah, San Diego, and San Jose).

  47. [47]. See generally People ex rel. Sklodowski v. State, 695 N.E.2d 374 (Ill. 1998); McNamee v. State, 672 N.E.2d 1159 (Ill. 1996); People ex rel. Ill. Fed’n of Teachers v. Lindberg, 326 N.E.2d 749 (Ill. 1975).

  48. [48]. See Sklodowski, 695 N.E.2d at 379 (“The framers of the Illinois Constitution were careful to craft in the pension protection clause an amendment that would create a contractual right to benefits, while not freezing the politically sensitive area of pension financing.”); McNamee, 672 N.E.2d at 1162–63 (“[T]he purpose of the [Pension Protection Clause] was to clarify and strengthen the right of state and municipal employees to receive their pension benefits, but not to control funding.”); Ill. Fed’n of Teachers, 326 N.E.2d at 751 (Pension Protection Clause created contractual right “that they would receive the money due them at the time of their retirement”).

  49. [49]. In re Pension Reform Litig., 32 N.E.3d 1, 30 (Ill. 2015) (holding unconstitutional legislation that would have reduced state employee pensions).

  50. [50]. And that includes Congress. It excluded governmental plans from ERISA’s funding requirements because it believed that governmental taxing powers would be an adequate substitute. See Jack VanDerhei, Funding Public and Private Pensions, in Pension Funding & Taxation: Implications for Tomorrow 59, 75 (Dallas L. Salisbury & Nora Super Jones, eds., 1994) (“The legislature considered the ability of the governmental entities to fulfill the obligation to employees through their taxing powers an adequate substitute for minimum funding standards.”).

  51. [51]. Whether this bet on the receipt of promised pension benefits constitutes a true “lottery” is open to dispute. Few, if any, would willingly participate in a “lottery” where the organizer is entitled to unilaterally proclaim itself to be the winner. There would be nothing “random” about the results of the lottery; indeed, it would be obvious that no outsider would ever win. In addition, even if one could characterize the arrangement as a “lottery,” some compensation should be provided for state officials’ deliberate understatement of their jurisdictions’ financial difficulties, and thus the odds facing those thinking of playing the lottery. See infra note 57 (describing SEC actions).

  52. [52]. Of course, reducing future wages only operates as a “claw back” if the benefited workers continue in their jobs after the wage reduction; if the over-benefited workers have already retired or are on the verge of retirement, wage reductions foist the cost of past promises onto new employees and indirectly onto residents who can expect to attract less competent service providers.

  53. [53]. Although some jurisdictions have been allowed to cut back on their pension promises, particularly with respect to benefits for work yet to be performed, others have found such plans stymied by state law restrictions. See supra note 21 (describing impediments to reductions in pension benefits under some states’ laws).

  54. [54]. This would be true regardless of whether their owners purchased the bonds at the time of their original issuance, or at some later time on a secondary market. See Robert S. Amdursky et al., Municipal Debt Finance Law Theory and Practice 45, 48 (2d ed. 2013) (describing the mechanics of original issuance).

  55. [55]. There is some question as to how efficient the municipal bond market is at differentiating risk. See Maria O’Brien Hylton, Central Falls Retirees v. Bondholders: Assessing Fear of Contagion in Chapter 9 Proceedings, 59 Wayne L. Rev. 525, 550–52 (2013) (describing contrasting scholarly views). Of course, the market could become better at risk differentiation if investors start paying a price for its absence. See Clayton P. Gillette, Bondholders and Financially Stressed Municipalities, 39 Fordham Urb. L.J. 639, 657 (2012) (“Assigning priority to residents in the event of fiscal distress, therefore, would induce bondholders to ensure that those involved in the bond issuance process would exploit their monitoring capacity to avoid bondholder losses or provide ex ante compensation in the form of higher interest rates.”). For a more general discussion of structural inadequacies in the market for municipal bonds and limits on its regulatory supervision, see Amdursky et al., supra note 54, at 538–41 (discussing basis for “perception of holders of municipal securities as ‘second class citizens’” when it comes to disclosure requirements); id. at 641 (discussing problems created by the thin market for many municipal securities).

  56. [56]. See Gillette, supra note 55, at 677 (“Bondholders who have been compensated ex ante to take risks have little basis for complaint when those risks materialize.”); Kevin A. Kordana, Tax Increases in Municipal Bankruptcies, 83 Va. L. Rev. 1035, 1096–99 (1997) (detailing how bondholders have been paid ex ante to assume the risk).

  57. [57]. Indeed, the SEC has brought enforcement actions against both states and municipalities for misrepresenting the financial condition of their pension plans to investors in order to entice them to buy bonds at favorable rates. See U.S. Gov’t Accountability Office, GAO-12-322, State & Local Government Pension Plans: Economic Downturn Spurs Efforts to Address Costs and Sustainability 7 (2012) (“In 2008 and 2010, respectively, the Securities and Exchange Commission took enforcement actions against the city of San Diego and the state of New Jersey for misrepresenting the financial condition of their pension funds in information provided to investors.”); Michael Corkery & Jeannette Neumann, SEC Says Illinois Hid Pension Troubles, Wall Street J. (Mar. 14, 2013, 4:34 PM), (announcing settlement of security fraud charges under which Illinois avoided paying a penalty or admitting wrongdoing).

  58. [58]. See Kordana, supra note 56, at 1067 (explaining how opportunistic behavior by municipal debtors could become “a self-fulfilling prophecy” ending with the “collapse” of the municipal bond market “with no borrowing or lending taking place”). But see id. at 1085 (this is “a theoretical possibility, not a necessity” and empirical evidence suggests that “such moral hazard is constrained, as predicted by the dynamic game models” in international settings).

  59. [59]. Implicit in this statement is the argument that taxation should be based on the receipt of benefits, not just on the “ability to pay.” This argument is in some sense uncontroversial—after all, jurisdictions levy taxes, the proceeds of which are used to pay for government services for residents rather than to a random selection of insiders and outsiders. Eliminating the connection between paying taxes and receiving government benefits not only undermines public support for taxation, but also removes an important disciplining device on governments’ choice of the amount and type of benefits to provide. Finally, in a mobile world with exit possibilities, a single-minded focus on redistribution causes the wealthy—and thus the ability to raise revenue—to disappear to more welcoming jurisdictions. This is not to imply that ability-to-pay concerns are irrelevant. Indeed, within a given jurisdiction, such factors often influence how to allocate the overall cost of providing government services among residents. One legitimate concern with retroactive taxation is that individuals (or businesses) who would have had an ability to pay higher taxes had they been levied and collected at an earlier point in time may have dissipated that ability by the time of the retroactive tax-triggering event. Of course, the same can be said about employees who counted on receiving now-threatened pensions when deciding how much of their cash salary to save for retirement and how much to spend on current desires.

  60. [60]. See Kordana, supra note 56, at 1103 (arguing against the imposition of compulsory tax increases in Chapter 9 proceedings because they will cause some “infra-marginal residents, businesses, and investors” to leave and “importantly, residents and investors considering which location to move to will certainly be less likely to choose such a ‘loser’ municipality”). Moreover, the scheme would make moving to other jurisdictions less remunerative. Although former residents of the distressed jurisdiction might be able to obtain a better combination of tax rates and government services by moving to a less-distressed jurisdiction, moving would not enable them to escape the taxes they would owe to the distressed jurisdiction for services received in prior years.

  61. [61]. See 11 U.S.C. § 364(a) (2012) (allowing trustee to obtain credit to operate the business of the debtor and treat its repayment as an administrative expense). Indeed, if necessary, this section allows trustees to obtain credit or incur debt “with priority over any or all administrative expenses.” See id. § 364(c)(1). However, this priority does not extend to pre-existing secured debt; such debt retains its priority to the extent of its value prior to the capital infusion. But other provisions of the Bankruptcy Code may diminish such priority somewhat in the interests of providing liquidity for the bankrupt entity. See Kenneth Ayotte & David A. Skeel, Jr., Bankruptcy Law as a Liquidity Provider, 80 U. Chi. L. Rev. 1557, 1589–1602 (2013) (describing priming loans and other mechanisms in bankruptcy law which protect “new money” to provide liquidity to distressed entities).

  62. [62]. See Douglas G. Baird & Robert K. Rasmussen, Private Debt and the Missing Lever of Corporate Governance, 154 U. Pa. L. Rev. 1209, 1238 (2006) (“A new lender tends to enter the scene only with the blessings of the existing one. The debtor is going to need to use the cash collateral of the existing lender, and the new lender will generally insist on a lien that primes that of the existing lenders.” (footnotes omitted)).

  63. [63]. The tax should apply regardless of the location to which the former resident moves. Not all moves are tax-motivated; there is admittedly a chance that a resident of one distressed jurisdiction could move to another such jurisdiction. If the first distressed jurisdiction operates a retroactive tax regime while the second chooses to raise future tax rates, the unfortunate individual could find herself in the position of paying for the past revenue failures of both jurisdictions; alternatively, the individual may be able to escape liability for either if the first jurisdiction raises future taxes and the second operates a retroactive regime. However, there is no way to coordinate these liabilities without implicitly favoring one of two equally legitimate state tax regimes. See Moorman Mfg. Co. v. Bair, 437 U.S. 267, 279 (1978) (refusing to strike down single factor apportionment of corporate income taxes despite possibility of duplicative taxation caused by other states’ use of three factor apportionment formulas because “[t]he prevention of duplicative taxation . . . would require national uniform rules for the division of income”). It is therefore up to taxpayers to protect themselves by choosing where to live. Even in the worst case (double taxation), of course, the individual is no worse off than if, prior to his decision to move to the second distressed jurisdiction, he had paid tax to the first jurisdiction at a level appropriate to cover that state’s level of expenditures—i.e. he would be in the same position as an individual moving from a nondistressed jurisdiction to a distressed jurisdiction.

  64. [64]. This shift would make sense given that it is those older taxpayers who benefited from the government services provided in earlier years. They are also, as a group, generally better off financially than younger residents. See Roin, supra note 7, at 777 (“In absolute terms, the elderly are now better off than younger workers.”).

  65. [65]. See Greg Kaplan & Sam Schulhofer–Wohl, Understanding the Long-Run Decline in Interstate Migration 1–2 (Fed. Reserve Bank of Minneapolis, Working Paper No. 697, 2015), https://www. (showing that net migration flows have fallen by half since the early 1990s to about 1.5% per year).

  66. [66]. See Raven Molloy et al., Internal Migration in the United States 8 (Nat’l Bureau of Econ. Research, Working Paper No. 17307, 2011), (“[S]lightly less than one-third of the population lives in a different state than they were born . . . .”).

  67. [67]. See William H. Frey, The Great American Migration Slowdown: Regional and Metropolitan Dimensions 5 (2009),

  68. [68]. Id.

  69. [69]. Id.

  70. [70]. See Kyle D. Logue, Tax Transitions, Opportunistic Retroactivity, and the Benefits of Government Precommitment, 94 Mich. L. Rev. 1129, 1172 (1996) (“The benefits of applying tax-rate increases retroactively, however, should not be overstated. To produce the desired efficiency effects, Congress would have to promise credibly that it would never enact a retroactive rate increase again . . . . [or] taxpayers would alter their behavior in anticipation of future retroactive rate increases, thereby reducing and perhaps eliminating the efficiency benefits of the initial retroactive tax.”).

  71. [71]. Levmore, supra note 17, at 273–78 (indicating that occasional, unexpected use of retroactive taxation can raise revenue without creating behavioral distortions); see also Kyle D. Logue, If Taxpayers Can’t Be Fooled, Maybe Congress Can: A Public Choice Perspective on the Tax Transition Debate, 67 U. Chi. L. Rev. 1507, 1508 (2000) (“[A] number of tax policy scholars have for years encouraged the use of retroactive taxation precisely because such taxes have the capacity to surprise investors and thus to raise revenue without distorting incentives.”).

  72. [72]. See Logue, supra note 71, at 1513.

  73. [73]. See id. (“[I]f the norm essentially said that whatever incentive subsidy the government provides will be opportunistically confiscated ex post . . . then inducing reliance on the government’s incentive subsidies would become impossible.”).

  74. [74]. Scholars have also noted that advanced planning for retroactive legal changes may be beneficial:

    [A] transition policy that credibly promises not to alleviate the retroactivity effects of rule changes will encourage investors to take into account the possibility of such changes—even to anticipate them. And this will generally be a good thing, if we assume that government policy, over time, will tend toward the maximization of social welfare.

    Id. at 1512; see also Michael J. Graetz, Legal Transitions: The Case of Retroactivity in Income Tax Revision, 126 U. Pa. L. Rev. 47, 65–66 (1977) (indicating that behavioral changes by investors anticipating rule changes can advance social welfare).

  75. [75]. For example, if voters know that they will be responsible for paying the taxes necessary to fully fund state pension obligations, they may be less resistant to the adoption of mandatory funding standards matching plan contributions to the economic accrual of those obligations.

  76. [76]. This assumes that residents can impact policy decisions made by their government representatives. Some express skepticism at residents’ ability to monitor against financial misbehavior. See, e.g., Gillette, supra note 55, at 659 (arguing that citizen monitoring is difficult because of heterogeneous preferences and distributions of benefits and burdens). Such arguments, though, must be taken with a grain of salt—taken to their extreme, they would be an argument against the electoral process. Kordana, supra note 56, at 1092–93 (arguing that collective action problems make residents poor monitors of municipal finances); Richard C. Schragger, Citizens Versus Bondholders, 39 Fordham Urb. L.J. 787, 790–91 (2012) (expressing skepticism about citizen monitoring).

  77. [77]. See Levmore, supra note 32, at 980 (“Fellow creditors and investors can hope that a sophisticated and skeptical investor blows the whistle on the developing Ponzi scheme before much money is lost.”).

  78. [78]. See Kordana, supra note 56, at 1102–03 (describing costs of relocating).

  79. [79]. See Lynn A. Baker et al., Local Government Law Cases and Materials 599 (5th ed. 2015) (“Several courts have indicated that fees imposed in excess of the cost of service constitute a tax.”).

  80. [80]. Although local governments are the primary users of property taxes, some state governments also rely on them for a substantial portion of their revenues. Nat’l Conference of State Legislatures, Which States Rely on Which Tax 3,
    WhichStatesRelyonWhichTax.pdf (last visited Nov. 11, 2016) (“Property taxes are very significant for state and local collections when considered together. . . . However, when that is divided between the two levels of government the direct impact of property taxes on state tax systems is quite small. Still, for a few state governments, property taxes are a significant share of their total tax revenue. For fiscal years 2006 and 2007, Vermont’s property tax collections have averaged 34.5 percent of state tax revenue . . . . New Hampshire’s reliance . . . [was] 18 percent of total state tax revenue . . . . Wyoming’s reliance is nearly 12 percent. Washington, Michigan, and Montana each count on property taxes for roughly 10 percent of their total tax collections.”).

  81. [81]. See generally Madiar, supra note 4.

  82. [82]. See 11 U.S.C. § 547 (2012).

  83. [83]. See id. § 548(a)(1).

  84. [84]. See id. § 544.

  85. [85]. See generally In re Dewey & LeBoeuf LLP, 478 B.R. 627 (Bankr. S.D.N.Y. 2012).

  86. [86]. See Investor Beware!, supra note 31 (the statute of limitations for constructive fraud is four years in Pennsylvania and six years in New York).

  87. [87]. See id. The Madoff investors were exempted from this claw back provision because the Madoff scheme involved securities. See supra note 34.

  88. [88]. I.R.C. § 6501(a) (2012).

  89. [89]. Id. § 6501(e).

  90. [90]. Id. § 6501(c)(1).

  91. [91]. Id. § 6501(c)(3).

  92. [92]. See Walter Hellerstein et al., State and Local Taxation Cases and Materials 1005 (10th ed. 2014) (describing rules).

  93. [93]. I.R.C. § 6502(a).

  94. [94]. See United States v. Perrina, 877 F. Supp. 215, 217 (D.N.J. 1994) (“Certainly the United States as a creditor has the right, like any other creditor, to bring an action . . . against the transferee of a taxpayer for a fraudulent conveyance.”).

  95. [95]. See id.

  96. [96]. See I.R.C. § 6901(c) (2012).

  97. [97]. To the extent one wants to encourage better voting behavior, of course, the longer the recapture period the better. Indeed, four years might be the minimum necessary to affect one election cycle.

  98. [98]. The IRS’s website suggests taxpayers should retain all of their tax records for three to seven years past the due date of the return and in some cases retain copies of their income tax returns essentially forever. See How Long Should I Keep Records?, IRS, (last visited Nov. 10, 2016). Of course, only the tax return needs to be retained to calculate the effect of a tax rate increase, so from a documentary standpoint, a longer retroactive period would be possible.

  99. [99]. See Gen. Motors Corp. v. Romein, 503 U.S. 181, 191 (1992) (courts disfavor legislation that “deprive[s] citizens of legitimate expectations and upset[s] settled transactions”).

  100. [100]. See VanDerhei, supra note 50, at 61–64 (description of required actuarial assumptions).

  101. [101]. See id. at 61 (“The primary concerns are that overly conservative estimates will lead to an increase in funding . . . while overly optimistic assumptions will reduce funding and perhaps threaten benefit security . . . .”); see also Liston & LaBombarde, supra note 50, at 128 (noting “some concerns that . . . actuarial assumptions are being manipulated in order to reduce required plan contributions”).

  102. [102]. Some of the room for strategically adjusting actuarial assumptions may have been reduced by recent changes in governmental accounting standards. See News Release 06/25/12: GASB Improves Pension Accounting and Financial Reporting Standards, Governmental Acct. Standards Board (June 25, 2012),
    cid=1176160126951 (describing required actuarial assumptions).

  103. [103]. They also play a role in state politics through direct campaign contributions and, in some cases, uncoordinated issue advertising.

  104. [104]. See supra note 80.

  105. [105]. See Milliken v. United States, 283 U.S. 15, 21 (1931) (upholding against a due process claim an increase in the rate of a gift tax imposed when the donor died within three years; the increase was imposed by legislation passed in the interim period between the date of the gift and the donor’s death).

  106. [106]. See Milliken, 283 U.S. at 23 (describing the treatment of gifts “made in contemplation of death”).

  107. [107]. See Welch v. Henry, 305 U.S. 134, 147 (1938) (“In the cases in which this Court has held invalid the taxation of gifts made and completely vested before the enactment of the taxing statute, decision was rested on the ground that the nature or amount of the tax could not reasonably have been anticipated by the taxpayer at the time of the particular voluntary act which the statute later made the taxable event.”); Milliken, 283 U.S. at 23 (“Not only was the decedent left in no uncertainty that the gift he was then making was subject to the provisions of the existing statute, but in view of its well understood purpose he should be regarded as taking his chances of any increase in the tax burden which might result from carrying out the established policy of taxation . . . .”).

  108. [108]. Such assessment systems were held constitutionally acceptable in Nordlinger v. Hahn, 505 U.S. 1 (1992).

  109. [109]. See, e.g., Levmore, supra note 17, at 266 (“Nominally prospective taxes can affect values and upset expectations every bit as much as explicitly retroactive assessments.”).

  110. [110]. See, e.g., E. Enters. v. Apfel, 524 U.S. 498, 532–33 (1998) (tracing disfavoring of retroactive legislation to English common law); Gen. Motors Corp. v. Romein, 503 U.S. 181, 191 (1992) (“Retroactive legislation presents problems of unfairness that are more serious than those posed by prospective legislation . . . .”); Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 208 (1988) (noting that “[r]etroactivity is not favored in the law”).

  111. [111]. See Landgraf v. USI Film Prods., 511 U.S. 244, 265–66 (1994) (discussing rationale for the Ex Post Facto Clause).

  112. [112]. Gen. Motors Corp., 503 U.S. at 191.

  113. [113]. See generally, e.g., United States v. Carlton, 512 U.S. 26 (1994); United States v. Hemme, 476 U.S. 558 (1986); United States v. Darusmont, 449 U.S. 292 (1981) (per curiam); Welch v. Henry, 305 U.S. 134 (1938).

  114. [114]. See Carlton, 512 U.S. at 30 (“The ‘harsh and oppressive’ formulation, however, ‘does not differ from the prohibition against arbitrary and irrational legislation’ that applies generally to enactments in the sphere of economic policy.” (quoting Pension Benefit Guar. Corp. v. R.A. Gray & Co., 467 U.S. 717, 733 (1984))).

  115. [115]. Id. at 31 (“‘The retroactive aspects of legislation, as well as the prospective aspects, must meet the test of due process, and the justifications for the latter may not suffice for the former’ . . . . But that burden is met simply by showing that the retroactive application of the legislation is itself justified by a rational legislative purpose.” (quoting Gray, 467 U.S. at 730)).

  116. [116]. The Supreme Court recently denied certiorari in a case challenging, on due process grounds, the retroactive application of an amendment to the Washington State estate tax. See Hambleton v. Dep’t of Revenue, 335 P.3d 398 (Wash. 2014), cert. denied, 136 S. Ct. 318 (2015). The Washington State legislature passed a statute including in the taxable estate of surviving spouses assets in qualified terminable interest property trusts received by them prior to the enactment of the explicitly prospective estate tax. Id. at 404–05. The Supreme Court of Washington upheld the amendment in the face of a due process claim. See id. at 411–12.

  117. [117]. Carlton, 512 U.S. at 30–31 (quoting Gray, 467 U.S. at 729).

  118. [118]. See id. at 31 (“The retroactive aspects of legislation, as well as the prospective aspects, must meet the test of due process, and the justifications for the latter may not suffice for the former.” (quoting Gray, 467 U.S. at 730)).

  119. [119]. Id.

  120. [120]. Id.

  121. [121]. Id. at 31–32 (“It seems clear that Congress did not contemplate such broad applicability of the deduction when it originally adopted § 2057. . . . It became evident shortly after passage of the 1986 Act, however, that the expected revenue loss under § 2057 could be as much as $7 billion—over 20 times greater than anticipated—because the deduction was not limited to situations in which the decedent owned the securities immediately before death.”).

  122. [122]. Quarty v. United States, 170 F.3d 961, 967 (9th Cir. 1999) (“There can be no dispute that the purpose of raising government revenue is a legitimate legislative purpose.”).

  123. [123]. See generally E. Enters. v. Apfel, 524 U.S. 498 (1998).

  124. [124]. Id. at 538. The plaintiff also raised a due process claim; however, the Court decided not to address it in light of its finding on the Takings Clause claim. See generally id.

  125. [125]. See generally Usery v. Turner Elkhorn Mining Co., 428 U.S. 1 (1976).

  126. [126]. Id. at 38.

  127. [127]. Id. at 18.

  128. [128]. See Welch v. Henry, 305 U.S. 134, 148 (1938) (“For more than seventy-five years it has been the familiar legislative practice of Congress in the enactment of revenue laws to tax retroactively income or profits received during the year of the session in which the taxing statute is enacted, and in some instances during the year of the preceding session.”).

  129. [129]. United States v. Carlton, 512 U.S. 26, 32 (1994).

  130. [130]. Id. at 33 (“[T]he actual retroactive effect of the 1987 amendment extended for a period only slightly greater than one year. Moreover, the amendment was proposed by the IRS in January 1987 . . . within a few months of § 2057’s original enactment.”).

  131. [131]. See, e.g., Welch, 305 U.S. at 150–51 (upholding Wisconsin tax enacted in 1935 but applicable to income derived in 1933 and 1934); Milliken v. United States, 283 U.S. 15, 23–24 (1931) (upholding application of tax statute regarding gifts causa mortis, enacted in 1918, to gift made in 1916 by 1920 decedent); Licari v. Comm’r, 946 F.2d 690, 694–95 (9th Cir. 1991) (upholding application of tax penalty passed in 1986 to returns filed from 1982 to 1984); Rocanova v. United States, 955 F. Supp. 27, 31 (S.D.N.Y. 1996) (upholding retroactive four-year extension of a tax collection statute of limitations); Mont. Rail Link, Inc. v. United States, 873 F. Supp. 1415, 1422 (D. Mont. 1994) (upholding 1989 legislative change which increased railroad retirement tax payments with respect to wages paid in 1987 and 1988); see also Quarty v. United States, 170 F.3d 961, 968 (9th Cir. 1998) (“The fact that the Carlton Court noted that the retroactive amendment before it was ‘curative’ of a legislative mistake does not make being ‘curative’ in that sense a requirement for the rationality of retroactive tax changes under the Due Process Clause.”).

  132. [132]. See Gillette Commercial Operations N. Am. & Subsidiaries v. Dept. of Treasury, 878 N.W.2d 891, 903 (Mich. Ct. App. 2015).

  133. [133]. Id. at 911.

  134. [134]. See, e.g., Gen. Motors Corp. v. Romein, 503 U.S. 181, 191–92 (1992) (approximately six years of retroactive effect); United States v. Heinszen & Co., 206 U.S. 370, 379–81, 390–91 (1907) (approximately seven years of retroactive effect); GPX Int’l Tire Corp. v. United States, 780 F.3d 1136, 1143 (Fed. Cir. 2015) (“[L]aw applies retroactively for a period of a little over five years . . . .”).

  135. [135]. In United States v. Carlton, 512 U.S. 26, 30 (1994), the Court stated that “[t]he due process standard to be applied to tax statutes with retroactive effect, therefore, is the same as that generally applicable to retroactive economic legislation.”

  136. [136]. This term encompasses both former residents who left before enactment of the taxing regime and nonresidents with income derived from sources within the state during the retroactive period.

  137. [137]. There should be no equal-protection argument as long as the same retroactive tax obligation is levied on in-state residents.

  138. [138]. The Declaration of Independence includes the “imposing Taxes on us without our Consent” on the list of “repeated injuries and usurpations” of the “present King of Great Britain.” The Declaration of Independence para. 2, 3 (U.S. 1776).

  139. [139]. Quill Corp. v. North Dakota, 504 U.S. 298, 306 (1992) (quoting Miller Bros. Co. v. Maryland, 347 U.S. 340, 344–45 (1954)).

  140. [140]. Id. (quoting Moorman Mfg. Co. v. Bair, 437 U.S. 267, 273 (1978)).

  141. [141]. See Nw. States Portland Cement Co. v. Minnesota, 358 U.S. 450, 452, 465 (1959) (“We conclude that net income from the interstate operations of a foreign corporation may be subjected to state taxation provided the levy is not discriminatory and is properly apportioned to local activities within the taxing State . . . .”).

  142. [142]. See Saenz v. Roe, 526 U.S. 489, 500 (1999) (interpreting the “right to travel” to include the right to “leave another State”); Crandall v. Nevada, 73 U.S. (6 Wall.) 35, 36 (1867) (invalidating a $1 tax imposed “upon every person leaving the State by any railroad, stage coach or other vehicle”).

  143. [143]. See Jerome R. Hellerstein & John A. Swain, State Taxation ¶ 4.13[1][a] (database updated 2016) (“When a state taxing statute discriminates on its face against interstate commerce (i.e., when the taxing statute explicitly subjects out-of-state products, out-of-state taxpayers, or interstate transactions to higher tax burdens than competing local products, taxpayers, or transactions), the Court has adopted ‘a virtually per se rule of invalidity’ for such taxes.”); Saul Levmore, Interstate Exploitation and Judicial Intervention, 69 Va. L. Rev. 563, 592 (1983) (“The Court has generally struck down taxes that superficially appear designed to fall more heavily on nonresidents—for example, commuters or multistate businesses—and has upheld other, apparently neutral taxes.”).

  144. [144]. See Lunding v. N.Y. Tax Appeals Tribunal, 522 U.S. 287, 290 (1998) (striking down New York law that denies “only nonresident taxpayers an income tax deduction for alimony paid” under the Privileges and Immunities Clause); Austin v. New Hampshire, 420 U.S. 656, 665–66 (1975) (striking down an income tax applicable only to nonresidents under Privileges and Immunities Clause).

  145. [145]. Indeed, that is one of the benefits of this taxing mechanism. See supra note 60 and accompanying text.

  146. [146]. See Tucker Staley, The Effect of TELs on State Revenue Volatility: Evidence from the American States, 35 Pub. Budgeting & Fin. 29, 29 (2015) (tax revolt sparked by Howard Jarvis’ Proposition 13 in California); see also Susan P. Fino, A Cure Worse Than the Disease? Taxation and Finance Provisions in State Constitutions, 34 Rutgers L.J. 959, 983 (2003) (“The tax revolt . . . began in California and Michigan in 1978.”).

  147. [147]. Bert Waisanen, State Tax and Expenditure Limits—2010, Nat’l Conf. St. Legislatures, (last visited Nov. 10, 2016) (“[N]o two TELs are exactly alike in their design and characteristics.”).

  148. [148]. See Colin Campbell, NC Senate Looks to Amend Constitution to Cap Taxes, Spending, News & Observer (Aug. 6, 2015, 3:30 PM), (the North Carolina Senate Finance Committee voted to put up a proposal to cap state’s personal income tax rate at five percent, tie state budget increases to the rate of population growth and inflation, and to create an emergency savings fund); Margaret Newkirk, Georgia Caps Tax Rate It Hasn’t Raised in 30 Years: Muni Credit, Bloomberg (Nov. 25, 2014, 7:00 PM), (“This month, Georgia became the first U.S. state to make raising its income-tax rate unconstitutional . . . .”).

  149. [149]. See Staley, supra note 146, at 29 (identifying tax and expenditure limitations as TELs).

  150. [150]. Id. at 37 (describing “six distinct characteristics” that need to be taken into account to measure the “stringency” of particular state tax or expenditure limitations).

  151. [151]. Waisanen, supra note 147.

  152. [152]. Id.

  153. [153]. Id. (showing that Arizona and South Carolina use a percentage of total state personal income to limit state spending).

  154. [154]. Id. (showing that Hawaii, Louisiana, Oklahoma, South Carolina and Texas use a combination of inflation, population, and personal income to limit state spending).

  155. [155]. Id. Arizona’s constitution imposes a percentage of income cap on appropriations. Id. States with limits tied to prior year appropriations include Alaska, California, Hawaii, Louisiana, Tennessee and Texas. Id.

  156. [156]. Id. States with statutes imposing percentage of personal income caps on spending or appropriations include Colorado, Idaho, and North Carolina. See id. States with statutes limiting government spending or appropriations by reference to population and/or income growth and inflation include Connecticut, Nevada, Ohio, Utah, and Washington. See id.

  157. [157]. Id. For example, Delaware’s constitution provides that appropriations may not exceed 98% of the year’s revenue estimate while the Iowa statute sets a 99% limit. Mississippi’s 98% limitation is contained in a statute. Rhode Island’s constitution has a 97% limit. See id.

  158. [158]. See id.

  159. [159]. See Staley, supra note 146, at 38–39 (“While most studies view all TELs as the same, this is an incredibly misguided perception. Depending on the type, nature, restrictions, approval, exemptions, and overrides of the TEL, some are more stringently binding than others. . . . Less restrictive TELs are more likely to have built in loop-holes and other mechanisms which allow policymakers to ignore the limitations established by the TEL.”).

  160. [160]. A legislative supermajority may be required to pass such legislation, however. See infra note 164 and accompanying text (discussing supermajority requirements).

  161. [161]. See Alison McCarthy & Elaine Maag, Limits on State Revenue, 112 Tax Notes 443, 443 (2006), (stating that these caps “typically restrict the growth of tax increases to the growth of population and inflation or personal income . . . As of [2006], six states had a tax [cap] in place”).

  162. [162]. See id.; see also Tax Policy Ctr., The State of State (and Local) Tax Policy: What Are Tax and Expenditure Limits? (2016), (“The most stringent revenue limits require that surplus revenues go back to taxpayers as rebates or be sequestered in rainy day funds.”).

  163. [163]. See Campbell, supra note 148; Liz Farmer, Georgia Becomes First State to Cap Income Taxes, Governing The States & Localities (Nov. 4, 2014),

  164. [164]. McCarthy & Maag, supra note 161, at 443.

  165. [165]. See Waisanen, supra note 147 (discussing Michigan and Missouri as examples).

  166. [166]. The precise adjustment varies from state-to-state. See id. (describing state revenue limitations).

  167. [167]. See id. (discussing Colorado, Florida, Michigan, and Missouri’s revenue limitations). Massachusetts’ revenue limitation is statutory. See id.

  168. [168]. Colorado, for example, found itself in the position of having to consider methods for returning tax revenues to its citizens when revenues from its marijuana tax caused total tax revenues to exceed its revenue cap. See Jack Healy, In Colorado, Marijuana Taxes May Have to Be Passed Back, N.Y. Times (Apr. 1, 2015), (“Technical tripwires in that voter-approved provision, known as the Taxpayer’s Bill of Rights, may require Colorado to refund nearly $60 million in marijuana taxes because the state’s overall revenue estimates ended up being too low when the marijuana tax question was put to voters. . . . In rare bipartisan agreement on taxes, legislators are piecing together a bill that would seek voters’ permission to hold on to the marijuana money.”).


Seymour Logan Professor of Law, University of Chicago Law School. 

I am indebted to Saul Levmore, Douglas Baird, and participants in the University of Chicago Law School’s Works in Progress workshop for their comments on earlier drafts. Mistakes remain my own.