Interval and Categorical Freezing

Drift occurs where policies are rigid but the economic, social, and political environment changes. Less ambiguous institutions and policies containing “plainer, less malleable language” elicit change through drift (Rocco and Thurston Reference Rocco and Thurston2014). Scholars examining the rigidity that produces drift have typically focused on rigid numerical formulas, such as the federal minimum wage ($7.25 since 2009, which is worth 27.4% less in 2022 after adjusting for inflation [Cooper, Martinez Hickey, and Zipperer Reference Cooper, Hickey and Zipperer2022]). However, there are several ways in which policies rigidify; in effect, freezing a policy at a particular level or rigidifying boundaries around specific activities or groups of people. I identify two main types of policy rigidity: interval freezing and categorical freezing (Table 1). Interval freezing is a rigid numerical formula, whereas categorical freezing is a fixed rule imposed on a specific category, with clear and stable rules about what counts as part of that category.

Table 1 Interval and Categorical Freezing

Interval freezing is the rigidification of a numerical threshold, either when a policy is first passed or at some later time. The amount of money expended or taxes collected is fixed and not changed to reflect inflation; for example, the value of Temporary Assistance to Needy Families (TANF) benefits, the federal minimum wage, or the $200 firearms tax imposed in 1934 under the National Firearms Act (NFA). Drift is the reduction in the real value of these benefits and taxes. If income thresholds for overtime eligibility remain static despite inflation, fewer workers are eligible to collect premium pay (Galvin Reference Galvin2016). Hence, overtime regulations with fixed numerical thresholds undermine employee protections over time (Koenig Reference Koenig2018). Interval freezing might also result in a fixed budget for staff or a predetermined number of full-time-equivalent employees charged with administering certain rules despite increasing coverage or population needs; for example, the regulatory agency created to enforce the Fair Labor Standards Act (FLSA), today has the same number of investigators as it did in 1948 but six times the caseload (Galvin Reference Galvin2016). Drift entails a reduction in the enforcement capacity of bureaucratic agencies.

Categorical freezing is the rigidification of category boundaries: fixed rules are imposed on a specific category of activity or persons, enumerating certain benefits or penalties without regard for changing group membership and other social, technological, or political changes that alter a particular standard and may impose unprecedented burdens; for instance, the requirement that states may collect sales taxes only from companies with a physical presence in that state (rendered obsolete by the rapid growth of e-commerce); the need to update highways and other infrastructure to accommodate driverless cars (Mettler Reference Mettler2016); residential zoning ordinances that do not account for population shifts (Quinn Reference Quinn1975); felon disenfranchisement laws imposed long before modern mass incarceration (Liles Reference Liles2006; Mauer Reference Mauer1999); and the granting of conscience exemptions to religious hospitals in reproductive healthcare, with the rapid recent expansion of the Catholic hospital system through mergers and acquisitions effectively denying reproductive healthcare options in certain jurisdictions (Fountain Reference Fountain2018; Solomon et al. Reference Solomon, Uttley, HasBrouck and Jung2020).

Conceptualizing the distinction between these two forms of policy rigidity—interval and categorical freezing—helps clarify the boundaries of the slippery concept of drift. Although many scholarly treatments of drift focus on what I term interval freezing—perhaps because it is easier to quantify the precise degree of drift that occurs through lack of change in numerical formulas or because many scholars working on policy drift are welfare state specialists concerned with un-indexed benefits—both interval and categorical freezing cause drift. There are more cases of drift than we have hitherto appreciated, and they invoke a much wider range of policy changes than the archetypal examples of welfare state retrenchment and declining bureaucratic capacity.

Welfare state retrenchment is a function not only of interval freezing—such as declining real minimum wages—but also of categorical freezing: rigid rules imposed on specific categories of persons. For example, independent contractors are not entitled to unemployment benefits, so the rapid rise of the gig economy has decreased the effective size of the welfare safety net. Recent surveys find that independent contractors represent as much as 15% of all workers, with Black and Hispanic workers more likely than white workers to be classified as independent contractors (Abraham et al. Reference Abraham, Hershbein, Houseman and Truesdale2023). Federal administrators have struggled to keep up with the speed of economic change, particularly the growth of ride-share companies.Footnote ¹ In October 2022 the Department of Labor proposed a new rule to reclassify more workers as employees, but efforts to change the rules face litigation and lobbying from organizations that would become newly responsible for worker benefits. Drift occurs when the frozen “employee” category confronts changing patterns of work.

Drift caused by interval freezing is mostly (though not exclusively) a function of inflation and population growth. Because these factors are more easily anticipated—at least in ordinary times, barring supply chain shocks, pandemic, and war—we might be more willing to attribute intentional strategic action to policy makers who write interval-frozen laws than we are to those promulgating categorical-frozen laws. But the relative ease with which intentional action can be discerned in relation to interval-frozen laws does not mean that drift produced by categorical-frozen laws is innocent of strategic policy maker action.

Drift caused by categorical freezing is a function of a much wider range of social, cultural, economic, technological, and environmental changes, including population movements, the growth of the knowledge economy, declining marriage rates, economic crises, the digital revolution, partisan polarization, and mass incarceration. Some changes might be predicted at the outset, and others present a clear agenda for policy updates that policy makers might deliberately fail to heed. When a policy is categorically frozen, any change in circumstances exogenous to that policy (whether foreseen or not) that alters the nature or size of a particular category of activities or persons can produce drift. By broadening our focus from interval to categorical freezing, we can more easily discern the range of ways in which drift occurs and with what consequences.

The interval–categorical distinction incorporates not only liberal concerns about the declining value of welfare benefits but also conservative concerns about the categorical freezing of civil rights era measures in a changing society. There is no neutral standard of policy modification. Rigid categories and thresholds could be seen as “stable” and “firm of purpose” or “inflexible” and “outdated” (Bardach Reference Bardach1976); responsive policy updating or “drift reversal” (Shpaizman Reference Shpaizman2017) could also be described as “unstable,” “volatile,” or “arbitrary” (Wiseman and Wright Reference Wiseman and Wright2022). While liberals criticize conservatives’ failure to update interval-frozen benefit levels or bureaucratic capacity, conservatives in turn condemn the rigidity of affirmative action quotas or elections preclearance rules for specific geographical locations (Gorsuch Reference Gorsuch2023; Roberts Reference Roberts2013; Thomas Reference Thomas2023).

Drift does tend to suit conservatives’ preferences because interval freezing involves erosion of benefit levels, transfer of risk to the individual, and declining bureaucratic enforcement capacity. Conservatives are more likely to celebrate the declining government tax-collection ability, wider disenfranchisement, and location-specific rules of categorical freezing. Yet although drift is generally characterized as a conservative device that reinforces social inequality, incremental change is not always conservative in nature (Béland, Campbell, and Weaver Reference Béland, Campbell and Weaver2022). Conservatives might be fearful of drift caused by the failure to “delist” benefits once covered by public programs or the difficulty of excluding previously covered populations (Hacker Reference Hacker2004b). Snead’s (Reference Snead2022) examination of labor law suggests that the Supreme Court became an “agent of drift” long before Republicans came to dominate judicial selection.

These disputes about drift take place in multiple arenas of contestation, including not only Congress and state legislatures, executive agencies, and the media but also the courts. Both interval and categorical forms of freezing invoke legal questions, and we cannot understand the legal implications of policy drift without considering categorical and interval freezing, because legal questions are typically couched in categorical terms.

Policy Drift and Legal Disputes

Drift has repercussions for many important legal disputes, including such basic questions as whether people can vote, get an abortion, own a gun, obtain legal representation, and exercise their right to speak and engage in political campaigns. For example, felon disenfranchisement laws passed in the nineteenth century took on new significance at the dawn of the age of mass incarceration in the latter part of the twentieth century. This categorical freezing—withdrawal of voting rights from those classed as having committed a felony—eventually resulted in the denial of voting rights to a vast swathe of the population: up from 1.17 million in 1976 to 6.1 million by 2016 (Chung Reference Chung2021), of which more than one-third are African American, a population excluded at more than four times the rate of any other racial group. In this case, drift has implications for voting rights and equal protection. Despite recent efforts by some states to reverse this drift by expanding voting access, today the disenfranchised population remains five times larger than it was at the middle of the twentieth century, with a severe racial skew (Uggen et al. Reference Uggen, Larson, Shannon and Pulido-Nava2020).

Federal Statutory Health Care Provider Conscience Protections passed in the 1970s were designed to protect the conscience rights of individuals and entities that object to performing, or assisting in the performance of, abortion or sterilization procedures. Fifty years later, mergers and acquisitions by Catholic hospitals have greatly increased the number of Catholic institutions—by 28.5% between 2001 and 2020—and Catholic hospitals are the sole community provider in many jurisdictions (Solomon et al. Reference Solomon, Uttley, HasBrouck and Jung2020). In 10 states, more than 30% of hospitals are Catholic. The consequences of granting religious protections to this category of institutions (categorical freezing) have increased substantially over time: in states with high proportions of Catholic hospitals such as Colorado, Washington, Illinois, and Oregon—all with Democratic unified government in 2023—family planning services are restricted or even denied for those served solely or mostly by Catholic institutions ("Catholic Hospitals’ Growth" 2022; Hochberg Reference Hochberg1996), a fact that may not be immediately obvious to patients because the names of merged entities do not always sound religious (Solomon et al. Reference Solomon, Uttley, HasBrouck and Jung2020; Wascher et al. Reference Wascher, Hebert, Freedman and Stulberg2018).

In these cases of drift caused by categorically frozen laws, shifting circumstances have transformed the nature, scope, and meaning of a rigid category, even though the textual specifics of categorically frozen laws do not change: a burden is imposed (disenfranchisement) or a benefit granted (conscience exceptions) to a clearly enumerated category. Subsequently, if the social, economic, or political environment shifts enough, categorical freezing can produce drift. The policy transformations wrought by drift not only involve increases in policy scope (more people brought within a particular category) but also fundamental changes to the nature and meaning of a policy. The fact that there has been rapid growth in the disenfranchised felon population or of the number of people served by Catholic hospitals is important. But even more interesting are the unequal (racialized, gendered) distribution of burdens and benefits; the way multiple categories interact to produce unprecedented effects and compounded disadvantage; lifelong spillovers; and, in many cases, basic constitutional rights claims. These categories have an entirely different meaning today than they did when first passed into law, even as (indeed, because) their boundaries remain formally unchanged.

Any policy that specifies the boundaries of a category and makes a blanket rule with respect to that category is susceptible to drift caused by categorical freezing. Similarly, any policy that specifies a threshold in absolute numerical terms is liable to drift due to interval freezing. Hardship results when inflation consumes the real value of welfare benefits specified in fixed dollar amounts, but interval freezing drift also invokes a variety of legal questions. For instance, in 1864 Congress established a $10 limit on attorney fees in veterans’ benefits claims, a law designed to protect Civil War veterans from overcharging by greedy lawyers. The law is still in effect, but the $10 limit is worth less than 5% of its original value—in effect, denying legal representation to those seeking to obtain disability benefits from the Veterans Administration. Here drift implicates due process and the right to counsel. Between 1976 and 2002, before the Bipartisan Campaign Reform Act indexed them to inflation, individual election campaign contributions limits were interval frozen at $1,000 per candidate. Inflation reduced the value of this hard limit by more than two-thirds during that 26-year period, raising questions about the limit’s impact on freedom of speech, particularly given the soaring cost of increasingly professionalized campaigns (Engle, DiLorenzo, and Spies Reference Engle, DiLorenzo and Spies1998).

Inflation reduces the value of numerical thresholds specified in absolute dollar amounts, sometimes transforming the nature of the policy entirely. A threshold that seemed generous at the time of passage becomes so small that it renders a policy completely unworkable or even reverses the original stated intentions of the creators, as in the case of the limits on attorney fees. Protection becomes a cage. Interval-frozen policies are transformed not only by inflation but also by shifting social, political, and technological circumstances that imbue a particular numerical threshold with meaning, as the case of campaign contribution limits demonstrates. It is not just that $1,000 bought less airtime in 2000 than it did in 1976, but also that changes in campaigning, voter habits, and technologies altered the meaning of the limit in ways that affect First Amendment rights.

Legal questions arise with respect to both the diminution and expansion of individual freedom through drift. Drifting policies can sometimes expand rather than contract individuals’ freedom of action and reduce rather than increase hurdles to the exercise of certain prerogatives. For example, the interval freezing of the National Firearms Act (NFA)’s 1934 tax on all registered NFA firearms—$200, a sum held constant for 89 years and counting—has effectively eliminated its original function. When President Roosevelt signed the NFA into law, $200 was worth more than $4,000 in today’s dollars, a prohibitively expensive amount that effectively prevented most people from owning a machine gun or short-barreled shotgun, some of the NFA-regulated items. Today the tax is worth 5% of its original value. Hence this hurdle to the exercise of Second Amendment rights is far lower with respect to the NFA’s taxation element (though other regulatory hurdles to gun ownership remain and have indeed expanded at the federal level and in many states).

Spousal and parental consent laws are another example of drift that has reduced rather than increased constraints on individuals. Parental consent laws (which exist in many states today) require minors to obtain parental consent before proceeding with an abortion. Spousal consent laws (which existed in many states up until the 1992 Planned Parenthood v. Casey Footnote ² decision rendered them unconstitutional) required wives to obtain the consent of their husband before proceeding with an abortion. The categorical freezing of these consent laws, many of which were passed in the immediate aftermath of the Supreme Court’s landmark 1973 Roe v. Wade Footnote ³ decision, means these pieces of legislation came to have more limited scope to restrict abortion than they did when first passed—even as other restrictions on abortion proliferated in the periods immediately preceding and following the Supreme Court’s seminal Dobbs v. Jackson Women’s Health Footnote ⁴ decision in 2022.

The reason the scope of these laws became more limited is that the social environment changed. Both pregnancy and abortion rates have declined over the past two decades among women under the age of 24. In 1973 the pregnancy rate per 1,000 girls under 15 was 13.5, and for girls 15–17, it was 65.1. By 2016, the rates were 2.5 and 15.0, respectively (Maddow-Zimet, Kost, and Finn Reference Maddow-Zimet, Kost and Finn2020). Marriage rates have also changed dramatically. In 1970, there were 76.5 marriages for every 1,000 unmarried women aged 15 and older in the United States; by 2008 the rate was 34.8 (Lee and Payne Reference Lee and Payne2010). Declining teenage pregnancies, declining marriage rates, increased cohabitation, and increased divorce rates over time drew fewer people within the ambit of parental and spousal consent laws.

In the case of the NFA tax, and spousal and parental consent laws, legal questions concern the capacity of a limit specified in absolute numerical or categorical terms to constrain behaviors under radically altered circumstances. In these cases—as with all instances of drift—economic, social, and political changes transform the scope and meaning of policies, although their language remains fixed. Table 2 details policies and their original scope and purpose, changes in environmental circumstances that produced drift, the nature of the drift, and its legal implications.

Table 2 Policy Drift and Legal Disputes

Table 2 identifies interval- and categorical-frozen policies that provoke legal questions through drift (outlined in the final column). This nonexhaustive list illustrates the variety and importance of the legal issues at stake when a frozen category or threshold drifts. Many of these policies have been taken to court; some have not. Of those that have been litigated, some complaints concern the intrinsic value of a policy (the allegation that the policy was unconstitutional from the moment it was promulgated); others the claim that the policy has drifted, although “drift” is not explicitly named of course (the policy was constitutional at the outset, but environmental changes rendered it unconstitutional). Cases concerning physical presence rules for state sales taxes in the internet era are quintessential “drift-based” cases: litigants argue that the internet revolution makes frozen taxation categories obsolete.Footnote ¹⁰ Cases concerning affirmative action quotas, campaign contribution limits, or religious conscience protections often involve intrinsic rather than drift claims (the policy has always been unconstitutional rather than having merely become so), but disputes might still invoke—and in several cases have turned on—the problems produced by drift specifically, as outlined in Table 2.

With a few notable exceptions (Barnes Reference Barnes2008; Snead Reference Snead2022), scholarship on policy drift has focused on the political causes and consequences of drift, rather than legal imperatives (Béland Reference Béland2007; Carpenter Reference Carpenter2010; Hacker Reference Hacker2004a; Hacker, Pierson, and Thelen Reference Hacker, Pierson, Thelen, Mahoney and Thelen2015; Koenig Reference Koenig2018; Rocco Reference Rocco2017). Separated, veto-prone systems with super-majoritarian thresholds for legislating are prone to drift because determined legislative minorities can easily block efforts to update policies or enact new ones (Hacker and Pierson Reference Hacker and Pierson2014). Blocking often benefits political actors who want to stay out of the limelight or pursue unpopular aims because it is not as “visible or traceable to particular groups as authoritative revision” (Galvin and Hacker Reference Galvin and Hacker2020, 219).

But when policy makers pursue unpopular, controversial, or even potentially illegal and unconstitutional aims, policy drift also has advantages for them beyond the political realm. Drift protects policies in court, in at least four ways: (1) judicial restraint, (2) low-salience injuries and slow-moving environmental change, (3) traditional expectations and the benefits of bright-line rules, and (4) plausible deniability for lawmakers and executive agencies.

Judicial restraint. Courts are institutionally predisposed to respect precedent and avoid (at least the appearance of) “legislating from the bench.” Opportunities for judicial action are most likely when lawmakers and executive agencies craft ambiguous laws with wide latitude for legal interpretation, rather than the clear, fixed, and rigid thresholds or categorizations characteristic of drifting policies. Whether because they are intellectually persuaded by the merits of judicial restraint or because they are strategically concerned with institutional maintenance at a time when public approval of the Supreme Court is at historic lows (Jones Reference Jones2022), judges and justices do not see themselves as responsible for updating specific pieces of legislation—even when frozen thresholds and categories confront rapidly changing environments, and policies drift.Footnote ¹¹

Low-salience injuries and slow-moving environmental change. Some of the drifting policies in Table 2 have never been litigated. Those injured by a particular form of drift might not be able to identify the specific policy that caused it because the policy passed long before the injury. The impact of interval or categorical freezing may also not be obvious. For instance, women denied access to sterilization, contraception, or abortion procedures in Catholic hospital systems today may not be familiar with the Church Amendments of the 1970s or be able to identify Catholic institutions when mergers and acquisitions typically take place quietly and without fanfare. In a recent survey, more than one-third of women whose primary hospital for reproductive care is Catholic were unaware of this fact (Wascher et al. Reference Wascher, Hebert, Freedman and Stulberg2018). The subtle nature of drift reduces the likelihood of legal challenge taking place because changes are slow to accrue and typically fly under the radar.

To bring a lawsuit against a drifting policy, plaintiffs must establish standing to sue. But the length of time over which drift unfolds makes it particularly difficult to demonstrate an “injury in fact,” to show that the injury in fact was caused by the defendant’s actions, and that it would be redressed by a favorable judgment from the court (Bertagna Reference Bertagna2006). Slower-moving and subtler exogenous changes, such as changing societal mores and economic circumstances, make it harder to build a case against a policy because the changes wrought by drift might not easily be discernible.

Traditional expectations and the benefits of bright-line rules. If drifting policies do reach the courtroom, defendants might fall back on “traditional expectations” of what a policy means, discounting ways in which environmental changes have altered the nature and scope of that policy. Textualist forms of originalism facilitate drift by freezing the meaning of a statute in the conditions present at the time of enactment, whether understood as founder intent or public meaning and use (Colby and Smith Reference Colby and Smith2009; Gillman Reference Gillman1997; Griffin Reference Griffin2008). At least four members of the current Supreme Court are self-identified “originalists,” although they do not agree about the content of the doctrine. Originalists bring a formal regard for textual specifics and an emphasis on history and tradition to the task of adjudicating cases. Insofar as the Court follows this jurisprudential philosophy, it is harder for litigants to argue in favor of updating drifting policies whose meaning is understood to be clear and fixed at the time of enactment.

Courts have also relied on the purported benefits of bright-line rules to uphold interval- and categorical-frozen laws. According to the Supreme Court in Quill v. North Dakota (Stevens Reference Stevens1992),Footnote ¹² a clear rule “firmly establishes the boundaries of legitimate state authority,” reduces litigation, “encourages settled expectations and, in so doing, fosters investment by businesses and individuals.” Categorically frozen “minimal-change maps”—drifting without new majority-minority districts as the minority population expands—are justified as the restrained exercise of neutral, traditional mapmaking principles such as district core retention (Bradley Reference Bradley2021; Edwards et al. Reference Edwards, Crespin, Williamson and Palmer2017). Proponents of interval- and categorical-frozen laws highlight their clarity and stability.

The Quill Court entrenched a categorically frozen “physical presence rule” for state sales and use taxes that prevented states from imposing such taxes on companies that were not physically located within state boundaries. Over the ensuing 26 years, the internet revolution transformed the relative advantages of out-of-state sellers and ravaged state revenues. The physical presence rule drifted. In 2018, as this drift became apparent, the Court took the opposite line, arguing against “the sort of arbitrary, formalistic distinction that the Court’s modern Commerce Clause precedents disavow” (Kennedy Reference Kennedy2018). In South Dakota v. Wayfair (2018),Footnote ¹³ the Court allowed South Dakota to dispense with the physical presence rule at last and collect sales taxes from out-of-state sellers. Wayfair remedied drift caused by the categorically frozen presence rule.

Yet in Wayfair’s wake, a different form of drift continued. New laws passed in all states with a sales tax specified absolute thresholds for determining whether a seller was engaged in business in a state; they were typically $100,000 of in-state sales or more than 200 in-state transactions. Over the next five years these interval-frozen laws began to drift again as further growth in the size and complexity of e-commerce rendered the limits increasingly burdensome for small businesses.Footnote ¹⁴ South Dakota repealed its transaction threshold in 2023, but most of these interval-frozen rules endured. Quill shows how bright-line rules become entrenched through court action and produce drift. The aftermath of the belated Wayfair remedy demonstrates that drift can persist through interval, as well as categorical freezing, even when courts and legislatures attempt to respond to its damaging effects.

Plausible deniability. If a policy is challenged in court, judges and justices often consider not only the nature of the policy but also the intentions of its creators when deciding on its constitutionality (James Reference James2011). One advantage that drifting policies have over newly imposed laws is that policy makers can argue in court that they are not responsible for any injuries that claimants allege. For instance, in Harris v. McRae (1980)Footnote ¹⁵ the Supreme Court upheld the constitutionality of strict limitations on the use of federal funds to cover the cost of abortions (categorical freezing), reasoning that—even as abortion has become increasingly concentrated among low-income women (Boonstra Reference Boonstra2016)—the obstacles to the exercise of that constitutional right were not of the government’s making: “Although government may not place obstacles in the path of a woman’s exercise of her freedom of choice, it need not remove those not of its own creation, and indigency falls within the latter category” (Stewart Reference Stewart1980).

Drift offers legal cushioning because it obscures the government’s role in producing a controversial outcome. When a policy has drifted, policy makers plausibly deny any intention to produce unconstitutional outcomes: they are merely the distal rather than proximate cause of any alleged injury, even if they might reasonably have foreseen certain environmental changes.

Because we cannot observe policy maker intentions directly, there is no way to assess precisely how strategic and intentional drift is. But we can consider policy maker actions and stated intentions, and what changes they might reasonably have anticipated when crafting a policy—the same factors courts consider when called on to adjudicate the intentions underpinning a policy. Intentional action could occur at two key stages: (1) when a law is first written and (2) when rules are enforced, subsequent updates are blocked, or opportunities to alter a rule are foregone within legislatures or executive agencies.

Some changes are more easily predicted than others. Policy makers might be expected to forecast at least some level of future inflation and population growth, even if forecasters find it difficult to predict specifics during periods of volatility, economic crisis, pandemic, and war. But, challenging as it might be to anticipate the precise level of inflation or migration flows, it is even harder to predict future technological advances and changes in societal mores. Hence, we are more willing to attribute intentional action to policy decisions that result in drift due to (relatively) predictable changes in circumstances than to unpredictable ones.

If policy makers explicitly block policy updates, say by voting down cost-of-living adjustments or quashing new legislation, they also facilitate drift. We are more willing to attribute intentional action to policy makers who actively marshal their legislative forces to defeat policy updates explicitly than we are to those who merely fail to act. There are many demands on policy makers’ time, so their failure to update a policy is not necessarily an indicator of hostile intent. Because stage 2 is temporally spread out over the entire lifespan of a policy, and it is impossible to count the number of times that a policy update was defeated before even receiving a committee hearing, we might be most confident attributing strategic action at stage 1: the language of the policy as written (and the associated statements of policy maker intent). Judges often attempt to discern policy makers’ intentions, but it is extremely difficult to identify them when a policy drifts.

All these features of legal disputes over drifting policies—the importance of precedent and textual formalism for courts, the problem of standing, the role of traditional expectations, the purported benefits of clear bright-line rules, and the difficulty of discerning intentional action—make it challenging for plaintiffs to build a successful legal case. Drift helps insulate policy outputs and outcomes that would provoke immediate opposition if they had been imposed through new legislation.