Basing budget baselines.
Measuring the cost of legislation or even projecting the course of the federal budget requires defining a budget baseline--a starting point capturing the current state of the budget. Budget baselines underlie most measures employed in federal budget debates and enforcement rules. Yet, despite their widespread use, budget baselines engender considerable confusion and abuse.
For instance, when legislators enact temporary tax breaks, the breaks are officially estimated to cost far less than they likely will because of a loophole in federal budget baseline rules. Then, later efforts to extend the tax cuts are counted as increasing deficits when, in fact, by more reasonable metrics, they do nothing of the sort and might even reduce deficits.
In response to such problems and the relative lack of scholarly attention, this Article seeks to ground budget baselines in a theoretical framework and then apply this framework to some of the leading debates involving baselines. For example, after presenting this new framework for understanding budget baselines, the Article proposes a way to fix the official baseline so that temporary tax cuts no longer appear less expensive than they really are and extensions no longer appear more expensive. This Article also uses this framework to describe why the long-term fiscal shortfall is smaller than often depicted and why a long-term budget metric now under consideration should be rejected.
By arriving at a better understanding of budget baselines, this Article helps to inform a number of key fiscal debates and makes recommendations for how to improve budget measures going forward.
TABLE OF CONTENTS INTRODUCTION I. HISTORY OF BUDGET BASELINES A. The Advent of the Federal Budget Baseline B. Baseline Disagreement and (Some) Consensus C. Long-Term Baselines: Confusion Always Reigns II. THE PURPOSES OF BUDGET BASELINES A. Setting the Agenda B. Policy Design: Expected Effects C. Policy Design: Using the Status Quo as a Reference Point D. Facilitating Political Accountability E. Enforcing Budget Targets F. Aiding Private Sector Planning G. Serving Irrational Biases III. A TAXONOMY OF BASELINES A. Current Law Baseline B. Low Attention Baseline C. Status Quo Baseline D. Expectations Baseline E. Policy Uncertainty Baseline F. Where Does the Official Baseline Fit? G. Summing Up: A Focus on Clarity and Expectations IV. LESSONS FOR BASELINE DEBATES A. Getting Rid of Baselines Is Not a Viable Option B. Baselines Should Not Be Frozen in Nominal Terms C. Ignoring Policy Cliffs (Much of the Time) 1. Why We Should Ignore Policy Cliffs 2. Applying This to the Official Baseline D. Clarifying Long-Term Baselines 1. How the Long-Term Fiscal Challenge Has Been Exaggerated 2. Confusion Compounds: The Curious Case of Generational Accounting CONCLUSION
It was the end of 2012, and Washington was embroiled in a fiscal standoff. Hundreds of billions per year in tax cuts (collectively known as the "2001 and 2003 tax cuts") were scheduled to expire. (1) Republicans wanted almost all of them extended, (2) but Democrats demanded that the high-income tax cuts be pared back. (3) The economy was on the line, as projections showed that failing to resolve the standoff and allowing all the tax cuts to expire at once could end the nascent recovery and drive the country back into recession. (4)
This "fiscal cliff" drew plenty of headlines, (5) but one relatively little-noted aspect of the cliff is that it had largely resulted from a quirk in what is known as a budget baseline. Furthermore, its resolution would be subject to the same quirk, making a final deal best seen as a tax increase appear as a very large tax cut, at least in the official cost estimates.
To quote from Merriam-Webster's Dictionary, a baseline is "a starting point." (6) Or, to offer a somewhat more technical variation on that--and applying it to a legal system--the baseline is the starting point of the legal regime and the reference point against which to measure legal change. The budget baseline is a subset of this: it is the starting point of the legal regime with budgetary effect. Anytime someone cites the cost of new legislation or describes how much a policymaker is increasing (or decreasing) spending, a baseline is at play--since any change must be measured relative to a starting point: the baseline.
The official budget baseline was a central character in the story of the fiscal cliff. Specifically, the official baseline treats expiration of temporary tax cuts as "real," even as it assumes extension of a number of major spending programs that also regularly expire. (7) The original authors of the 2001 and 2003 tax cuts took advantage of this loophole and wrote in an expiration to reduce the cost of the tax cut package and avoid budget enforcement rules that could have otherwise made enactment much more difficult. (8) Further, the same rule would govern in resolving the fiscal cliff. Despite the Democrats winning what some called a tax increase on the highest-income Americans, the Congressional Budget Office (CBO) estimated that the final deal added nearly $4 trillion to the deficit over the coming decade by making most of the tax cuts permanent. (9)
As this illustrates, when it comes to temporary tax changes, the baseline used to judge our current course and changes from it is broken. In this Article, I propose a way to amend the official baseline to close this loophole--which both parties continue to abuse--going forward. (10)
At the very same time that baseline manipulation was playing a large (albeit underappreciated) role in the fiscal cliff mess, it also featured in another fiscal debate dividing Washington: the long-term course of the federal budget. Back in 2012, CBO emphasized a long-term scenario--essentially, a baseline capturing our current course--that showed the budget on a track toward fiscal calamity. (11) CBO projected debt in excess of 250 percent of GDP by the 2040s. (12) These types of projections set the stage for the deficit-reduction recommendations from entities like the Bowles-Simpson fiscal commission, and helped justify the fiscal austerity that began in 2010. (13) However, this long-term baseline and others like it were fundamentally flawed and represented not so much a devastating long-term outlook as a fundamental confusion over that outlook--a point this author made at the time. (14)
And just one year later in its 2013 report, CBO highlighted a much rosier fiscal future. (15) In this new projection, debt stood at around 100 percent of GDP by 2040 (16)--not low, but much better than the projection from the year before and probably not deserving of the "fiscal crisis" label. In large part, this radical turnaround reflected different assumptions about the course of future law, but not anything that policymakers had done in the meantime. (17)
The radical change in the long-term trajectory has received less notice than it deserves; this reflects the more fundamental problem that many baselines--and perhaps long-term baselines especially--rely on unexamined assumptions. The inconsistency derives from confusion, even among the official budget agencies, as to what long-term baselines should measure. This Article offers a theory to ground long-term baselines by better describing what they can and should measure, and rejects a long-term metric advocated by some--called generational accounting--that represents an incoherent approach to long-term budgeting and reflects a further misunderstanding of baselines.
In sum, confusion around and abuse of budget baselines have shaped some of the most important fiscal policies of the current century, and failures in budget baselines have undermined political accountability, budget enforcement rules, and a proper understanding of the likely effects of these budget policies. In this Article, I seek to remedy this confusion and abuse.
I start by laying out a historical and theoretical groundwork in Parts I and II. Part I describes the origins of budget baselines in the 1970s and their subsequent widespread use in federal budgeting. Part II then engages in theoretical work that so far has been largely neglected by the literature, identifying how budget baselines can be constructively used in decision making. This is in contrast to some recent scholarship that sees baselines as largely arbitrary (18) and servicing irrational biases. (19)
Part III then brings organization to the jumble of employable baselines by creating a taxonomy of different budget baselines--differentiated by the part of the legal system these baselines hold constant going forward--and recommending which baselines to use and when. I argue that priority should be given to the expectations baseline--reflecting what is actually expected to occur. To give two examples: we should generally expect temporary tax cuts to be continued, and the baselines we use should reflect that expectation. Similarly, we should expect that the unsustainable will not be sustained in fiscal policy and that, one way or another, the budget deficits will be controlled in the coming years. Consequently, some of the predictions of fiscal calamity for future generations are overstated and certainly improperly framed.
With this taxonomy established, Part IV of this Article then applies these conclusions regarding which baselines to use to four separate ongoing debates that prominently feature baselines. First, I engage the longstanding debate over whether we should use baselines at all; I answer this "yes" in light of the considerable information that would be lost in the absence of baselines. Second, I describe why spending in the baseline should grow in nominal terms, despite calls (especially from conservatives) for a nominal freeze. Specifically, I conclude that a nominal freeze is not consistent with any of the constructive purposes of budget baselines, and serves only to misinform decision makers (if anything). Third, I engage how baselines (and cost estimates) should deal with temporary legislation, arguing that expirations should frequently be ignored. Finally, I discuss our long-term fiscal future and detail how baseline confusion is leading to a significant exaggeration of the magnitude and nature of the long-term fiscal shortfall, and should lead us to reject new methods like generational accounting that some are pushing to be officially adopted.
I. HISTORY OF BUDGET BASELINES
A. The Advent of the Federal Budget Baseline
The modern system of federal baseline budgeting began to take root in 1974 with the passage of the Congressional Budget Act. (20) That Act, which established the framework for the current budget process, included two separate provisions key to the development of budget baselines. First, it required the Office of Management and Budget (OMB) annually to submit to Congress an estimate of the amount of funding needed in the upcoming year to maintain federal programs at "the same level as the fiscal year in progess and without policy changes" (21)--a baseline often referred to as a "current services" or "current policy" baseline. Second, the Act created CBO as a competing counterpart to the executive budget office (22) and, among other things, asked CBO to produce each year a projection of the budget over the five years to come. (23)
Before 1974, federal budget analysts and policymakers used a baseline--but it was a very simple one, at least when it came to spending programs. In particular, spending levels were often compared to the nominal level (unadjusted for inflation or any other factors) of the prior year. (24) This was referred to as the "base" rather than a "baseline," (25) but the point remains the same. Policymakers and analysts used these figures as the starting point for budgeting for the upcoming fiscal year.
For revenues, the baseline was more sophisticated even in these early years. At least a few presidents' budgets prior to the Congressional Budget Act compared the proposed revenue level to what would occur under "current law" while taking into account changes in the economy. (26) This is akin to the current approach for calculating baseline revenues.
The early use of the current-year nominal figures as a baseline naturally precluded long-term budget projections. As will be discussed further in Part IV. B, the previous year's nominal funding is of questionable relevance as a "starting point," even for the upcoming year. Conditions--such as inflation, economic growth, and changes in caseload--mean that $100 in the current year for a program simply is not the same thing as $100 in the next year. Over longer periods of time, this becomes even more true. Looking only at what occurred in the previous year makes it difficult--or impossible--to project what will occur years down the road.
Thus, the advent of the new baseline helped policymakers and budget analysts begin to look beyond the ends of their noses, and beyond just the previous year's funding level. Alice Rivlin, the first director of CBO, made this point in some of her early testimony before Congress responding to OMB's current services projection. (27) In tasking CBO with making a five-year budget projection, Congress, at the time, had not actually given direction as to how that projection should be made. Rivlin made clear that CBO would use a "current services" concept to project spending in its new five-year projections (as it then proceeded to do) (28) and she ended her testimony with a strong endorsement of the new baseline. As Rivlin declared, "I believe that the use of such estimates will become widespread, both in the Congress and in the Executive Branch." (29) She was right.
B. Baseline Disagreement and (Some) Consensus
In the early days of the "current services" baseline, there was uncertainty as to how to implement the new baseline, and both budget offices used a variety of assumptions. (30) These inconsistencies were temporarily alleviated when two laws--the first in 1985 and the second in 1990 amending the first--established a baseline for purposes of certain budget enforcement rules. (31) The assumptions set out in those baselines were then largely adopted in CBO and OMB budget projections. (32) That baseline continues to be referenced today, even if not always followed. (33) I refer to the variation now employed by CBO as the "official" baseline. As amended in 1990, the baseline dictated uniform inflation measures to be used for projecting annually appropriated budget authority. (34) It established exactly how mandatory programs were to be projected: follow current law--that is, the law on the books going forward--except for any temporary mandatory programs greater than $50 million per year, which were assumed to be made permanent. (35) It also gave the projection rule for receipts: follow current law except for temporary excise taxes devoted to trust funds, which were assumed to also be made permanent. (36) Further, in the 1990s, CBO expanded its medium-term projection window from five years to ten years. (37) OMB followed suit, although the length of OMB's budget window has fluctuated since then. (38)
The temporary consensus around the medium-term federal budget baseline entirely broke down entering the 2000s. The key turning point was the enactment of the 2001 and 2003 tax cuts. (39) As a way to make them both appear less expensive than they likely would be and evade budget rules, these large tax cuts were written into law as temporary measures, even though supporters intended to make them permanent. (40) The magnitude of these and the other also-expiring tax cuts enacted during the Bush Administration was large--equal to nearly 2 percent of the economy. (41) Under the official baseline rules, these tax cuts were assumed to be temporary, producing a rosy budget outlook based on a scenario (expiration of the tax cuts) that supporters of the tax cuts never intended to occur. (42) The result was a baseline disconnected from any future state of the world that analysts or policymakers thought would actually exist.
With unrealistic assumptions underlying the official baseline, a cottage industry quickly arose among independent budget analysts to produce more "realistic" baselines. (43) The industry included outside analysts working from the Brookings Institution, (44) as well as a coalition of budget-oriented Washington think tanks that, for a time, offered a unified alternative baseline. (45) These alternative baselines quickly took on a prominent role in fiscal debates as news reports looked to these baselines, rather than the official baseline, for information on the budget's future. (46)
The official budget offices also got into this game. By the time of the Fiscal Year 2005 Budget, OMB was using a revised version of the current services baseline that assumed extension of some of the temporary tax cuts (arguably, to disguise the full cost of temporary policies the Bush Administration had itself initiated). (47) When President Obama took office, the new Administration adopted a series of different revised baselines against which to gauge the cost of its policies--all of which assumed continuation of most, and sometimes all, of the 2001 and 2003 tax cuts. (48)
As the official baseline projection lost policy relevance, CBO responded. First, CBO began to regularly produce a table showing the cost of various options (relative to the official baseline) that allowed analysts to make their own baseline adjustments. (49) Eventually, CBO began doing this itself, placing increasing emphasis on its "alternative fiscal scenario," which reflected extension of the tax cuts, among other adjustments. (50) To be clear, these adjustments were not without attendant controversy. A number of observers have suggested sticking to the official baseline rules rather than assuming extension of temporary policy, like the 2001 and 2003 tax cuts. (51)
C. Long-Term Baselines: Confusion Always Reigns
The long-term budget baseline--extending beyond the traditional ten-year window--perhaps generates even more confusion than the medium-term projections. Economists like Alan Auerbach have led the effort to increase the prominence of such projections in budget debates, developing long-term budget metrics like the fiscal gap and generational accounting, (52) the latter of which is critiqued in Part IV.D.2. In the mid-1990s, CBO also began calculating such long-term projections, extending its projections decades into the future. (53) Since then, CBO has annually produced a report giving its long-term estimates of up to seventy-five years. OMB now also provides such long-term projections under the Obama Administration's policies. (54)
However, when it comes to long-term projections exceeding the now-standard ten-year window, there was never an "official" set of assumptions that analysts or the budget offices adopted, or even used as a reference point. This translated into pervasive inconsistency in methodology, including internal inconsistency in projections. Take one example: revenues. Projections tend to divide into two camps. The until-recently dominant school assumes that most revenues are frozen as a share of the economy after ten years, unlike in the first ten years when revenues are allowed to vary as a share of the economy. (55) An alternative assumption--and one now being more widely adopted, or at least recognized--assumes that revenues rise over the long-term as factors such as "real bracket creep" (real growth which pushes people into higher tax brackets) drive up receipts. (56) The difference is considerable; over seventy-five years this reduces the fiscal gap by an amount approaching two percent of GDP. (57) But there is little consensus as to the proper treatment and why. Still, despite the general confusion with regard to long-term baselines, they are frequently employed in policy debates, and have been helping to drive concern about the track of the federal budget. (58) As explained in Part IV.D, the confusion has led to considerable exaggeration over the severity of the long-term fiscal challenges facing the country or, at least, a misunderstanding of them.
To summarize, baselines now have a long history in federal budgeting. It is a history not just of widespread use but also of confusion and abuse.
II. THE PURPOSES OF BUDGET BASELINES
Alleviating the confusion and abuse surrounding budget baselines requires both asking why we use baselines to begin with as well as understanding what purpose they can serve. This Part sets out a number of different ways in which budget baselines can be used to enhance decision making. This is meant to further our understanding of baselines in two different ways. First, scholarship on budget baselines (or any form of legal baseline) has rarely asked why we use baselines in decision making. Second, to the extent the purposes of budget baselines are discussed in modern scholarship, they tend to be either dismissed as arbitrary or described as purely preying on (as opposed to helping overcome) human irrationality: namely, a bias toward whatever gets framed as the status quo and an aversion to losing anything relative to that status quo. (59) I do not deny that baselines may sometimes be used in this fashion, but, as this Part makes clear, baselines can be and are used in ways that are far more constructive.
A. Setting the Agenda
A baseline can aid policymakers in determining what should be on the legislative agenda by helping to define and prioritize policy problems that need to be addressed. (60) This is a task that political scientists have increasingly recognized as central to the policy making process. (61) In particular, recent scholarship emphasizes Congress's limited capacity--both as an institution and in terms of individual members--to focus on particular policy issues and the resulting importance of agenda setting. (62)
The first step in agenda setting--defining problems that should be addressed--cannot be easily separated from the idea of establishing a baseline. To say that there are problems for policymakers to address assumes that there is a starting point for the legal system--a course that the legal system is on. Conversely, the assumption that there are some areas of policy without problems similarly presumes that policy is on a certain course (in this case, an unproblematic course).
Return to the case of the "fiscal cliff' invoked at the beginning of this Article. To make the concepts more concrete, I will return to this example repeatedly in the next two Parts. At the end of 2012, the country faced the scheduled expiration of temporary tax cuts equal to roughly 2 percent of GDP (plus other austerity measures). (63) The official baseline showed the country metaphorically jumping off the cliff and could have been seen as helping policymakers define the country's fiscal course as a large problem that needed to be addressed. CBO produced a series of analyses showing the dramatic negative effects of the massive scheduled austerity. (64)
The importance of a baseline to defining policy problems can also be illustrated by imagining the absence of a baseline entirely. Let us say policymakers looked at the world as if it were a tabula rasa. In that case, there would not be particular policy problems per se that would require attention, nor other policies that could be left alone. There would be no fiscal cliff because policy would not be on any course at all. Everything would have to be addressed because there would be no point from which to start, whether good or bad. Of course, some policy choices would be better or worse than others. Policymakers could create problems with their choices. But there would be no problems at the start that could be identified for future policy, other than the glaring problem that policymakers would have no place from which to start and would have to take on the world as a whole. (65)
This function of budget baselines--helping to set the policy agenda--results from the limited ability of both individuals and institutions to process policy decisions. This is a form of bounded rationality. That is, both individuals and institutions have a limited ability to address issues at any given point in time. Baselines can help address that limitation by showing what areas most deserve attention at a given point.
B. Policy Design: Expected Effects
It is one thing to decide that a policy item is a problem that should be on the legislative agenda. It is another thing to actually work toward a resolution of that problem and design a policy response. Here, again, baselines can come into play.
Baselines can show the expected course of policy, and thereby aid in policy design. In other words, baselines can do more than illustrate what would happen if policymakers sat back and did little or nothing; they can illustrate what is actually expected to occur. Whether that course is a good or bad one is an important piece of information for policy making. Policymakers should know whether their actions and the expected results are the right or wrong ones. Unlike the previous function, this does not derive from bounded rationality and the limited ability to process information. Rather, any decision maker, irrespective of such bounds, would probably want to know the expected effects of their actions.
In referring to expected effects, I am referring to a central estimate of the course of policy, around which there is likely to be a considerable range of uncertainty. Of course, different evaluators will have different guesses as to the central estimate. But some of these guesses should be better than others. Some will be more informed than others, and some will have a better record than others in making these predictions. The better guesses are the ones that should be more helpful in illuminating decision making.
At first, this may seem like a difficult function to differentiate from agenda setting. But again, the fiscal cliff can illustrate. In the lead up, it was certain that the cliff would be on the agenda, and there was high confidence that policymakers were going to extend a large share of the tax cuts, even if there was uncertainty as to exactly how that would happen and whether the country might temporarily jump off the cliff as policymakers negotiated. (66) In fact, President Obama had proposed extending roughly 70 percent of the temporary Bush tax cuts, (67) while the Republicans had proposed extending nearly all of them. (68) That was the difference in their positions. So judging whether the country was on a good or bad trajectory when it came to these tax cuts--and whether that should be changed--required looking not at what would happen if no action were taken but, instead, if the expected actions were taken.
C. Policy Design: Using the Status Quo as a Reference Point
Another purpose of baselines is to serve as a reference point in designing policies. By reference point, I mean a marker of where we are now. That marker can then help inform where we want to go.
There are a few reasons we may want such a marker. First, we might have limited information about the world around us, and thus our experience with the starting point may be particularly useful. In other words, amidst an array of different policy choices, we often do not know how different alternatives would turn out, or, at least, we do not know with confidence. We probably have much better information about policies as they exist now simply because we have been able to observe their effects in the world around us. As such, a baseline built on continuation of what we are currently doing can serve as a useful reference point. To the extent one is satisfied with policies as they now exist, the reference point is a model for future policy. To the extent one is dissatisfied, it serves as a point of departure. To be clear, there does not have to be any preference for the status quo for a focus on the status quo to provide useful information. We need only to have more information about the status quo than about alternatives for this to be the case.
Second, some may have an actual preference for the status quo. Those in this camp use the status quo not only as a means by which to better understand alternative policies, but as a preferred outcome going forward. The desirability of giving priority to the status quo is an old and long-contested idea. It goes to the core of conservative theory. This is the Burkean suspicion of radical change as compared to gradual evolution; namely, the idea that we know so little about our complex world that priority should be given to what we know most--the status quo--and that change should proceed gradually. (69) In modern times, this same idea has motivated the theory of incrementalism in policy making, a theory that describes (and sometimes advocates) decision making as "continually building out from the current situation, step-by-step and by small degrees." (70) These are theories that seek to rationally adapt to human limitations, or what might be thought of as maximizing rational decision making within constraints. For these theories to be operational, though, the status quo must be defined. That is where the baseline comes in, providing information on what represents the status quo.
I do not mean to say that such conservatism is always justified or is a universal description of the way policy making works. Despite Burkean warnings to the contrary, there are situations that may very well justify radical change. And there are times that policy making does shift radically, rather than in small increments. (71) With that said, conservatism and incrementalism cannot be wholly dismissed as decision-making rules (or even as descriptions); they have some value. (72) But, for them to work at all, there must be a baseline.
Thus, when it comes to the fiscal cliff, policymakers may want to use the tax policies in place just before the cliff as a marker. That is a state of the world with which they are familiar--unlike the world on the other side of the fiscal cliff, which is considerably different from what Americans have experienced. Policymakers might think that marker is a good or bad state of the world, but, either way, the marker allows them to better gauge alternative tax policies by reference to this known set. And, for Burkean conservatives or incrementalists, it might even be a state of the world that is preferred over alternatives. Or, at the least, gradual change from the marker might be preferred, rather than sudden changes (as in going off the fiscal cliff).
D. Facilitating Political Accountability
Baselines can also help establish political accountability in two separate ways. First, baselines can help gauge the degree to which policymakers have changed the course that policy is on. Second, a baseline can relatedly be used to project the course that policymakers have put us on. Both of these metrics seem relevant to holding policymakers accountable--though I suspect that the first may be more important as a normative matter.
Take one conception of accountability offered by the political scientist Andreas Schedler: "[T]he notion of political accountability carries two basic connotations: answerability, the obligation of public officials to inform about and to explain what they are doing; and enforcement, the capacity of accounting agencies to impose sanctions on powerholders who have violated their public duties." (73) I assume here that political accountability is in fact normatively desirable.
Both prongs of accountability--policymakers reporting on what they have done and the public, or some other body, sanctioning policymakers for violating their duties--require delineating the policies for which the policymakers are responsible. (74) In reporting their actions to the public, it makes sense for officials to identify the policies that are theirs and for the public to then adjudge policymakers based on the policies of those officials. (75) Absent this ownership of policies by the public officials, accountability makes little sense; there is little point in holding officials accountable for policies for which they are not in some sense responsible. (76)
This is why I suspect the first way baselines can help establish accountability is probably more meaningful as a normative matter. By gauging the degree to which policymakers have changed policy, the focus is on outcomes that policymakers can (and have) affected. This is in contrast to simply looking at the course policymakers have set us on, which the policymakers may or may not have been able to change. (77)
Gauging the degree to which policymakers have changed the course of policy requires comparing the course before and after the relevant policymakers took power, in other words, comparing two baselines. (78) First, there is a pre-policy baseline: the course that policy was on before the relevant policymakers took the reins. Second, there is a post-policy baseline: the course that policy is on now, with the current policymakers in control.
The difference between these two baselines measures the change that occurred while a certain set of policymakers was in charge, and this change is one way of assessing policymakers' impact. From this perspective, policymakers are less responsible for anything in the pre-policy baseline, as compared to the change they have affected from that point.
Again, return to the temporary tax cuts and the fiscal cliff. The idea here is that previous policymakers had put the country on an expected course whereby most, if not all, of the tax cuts would be made permanent, even if the policymakers had not formally enacted their expectations into law. (79) This would constitute the pre-policy baseline. Current policymakers should be held responsible (or at least more responsible) for the difference between this pre-policy baseline and the post-policy baseline, rather than responsible for everything that is set to occur, namely, the full cost of the tax cuts and their effects. In this case, current policymakers would not be charged with the cost of extending most of the tax cuts, even as they enacted the law that did so. Importantly, this notion of political accountability requires a starting point from which to measure.
E. Enforcing Budget Targets
Given the chaos that sometimes reigns in federal budget negotiations, (80) one could be forgiven for thinking that budgeting is a lawless territory of partisan strife. Although at times it might seem that way, policymakers in fact sometimes set targets to help govern the game. These targets--sometimes self-proclaimed and at other times formally enacted into congressional rules or statutes--facilitate policy making and political accountability. And a baseline is often needed to get these targets to work.
Before going on, it might be helpful to briefly consider why policymakers set budget targets in the first place. A collective of policymakers (such as Congress or Congress working with the President) can set budget targets to coordinate action. There are multiple actors involved in budget decision making, and communication among those actors can be challenging, especially in divisive political environments. (81) Budget targets can serve as a means for these various actors to agree on a unified framework that guides the budget decisions they will make. (82)
The economist Alan Auerbach expands on this point and explains specifically how budget targets can coordinate action by serving as a pre-commitment mechanism. (83) He describes how a budget rule set up by a majority in Congress--and that could be overturned by a majority--might still change legislative outcomes:
For example, suppose that each legislator prefers a low overall deficit to a higher one, but also wishes to promote his or her own spending priorities. With no budget rule in place, there may be no commitment mechanism in place to facilitate co-operation on keeping spending low.... Thus, the outcome achieved under a budget rule might be consistent with the contemporaneous wishes of the majority, while at the same time representing a different outcome than would occur without the budget rule in place. (84)
In other circumstances, budget targets are used not as a coordinative device, but instead as a means of flagging to the public a policymaker's fiscal position. This links back to the concept of political accountability. (85) For instance, a policymaker might unilaterally declare that a piece of legislation, perhaps even legislation that the policymaker is writing, should cost no more than a certain amount of money or that the legislation should not add anything to the deficit. (86) The idea is to highlight for the public the effects of a policymaker's legislation.
In setting budget targets, baselines come into play in two ways. First, if the target is set in terms of change in spending, revenues, or the deficit (or any other metric of change), there must be a baseline against which to measure. (87) In other words, there must be a starting point by which to identify and define change. Second, if the target is set in terms of projected outcomes, rather than change, there must again be a baseline set of policies defining the legislated course that underlie the projection.
The temporary tax cuts of the last decade can again serve as an example, as baselines played an important role in the way they were structured. In particular, policymakers set the tax cuts to expire in order to take advantage of the official baseline rules that respected temporary tax measures as actually temporary. First, they wanted to avoid one of the budget enforcement rules. Given the fast-track procedure policymakers were using, they could not increase the deficit outside the ten-year budget window without triggering a super-majority vote. (88) This would have required them to garner sixty votes rather than just fifty-one votes in the Senate. (89) Second, separate and apart from this issue of budget enforcement, proponents had set a target for the total cost of the tax cuts to which a number of policymakers were committed. (90) However, these two constraints were easily manipulated given how the baseline was constructed. The proponents in 2001 could make the tax cuts temporary to avoid a super-majority vote and help hit the publicly-stated cost target.
F. Aiding Private Sector Planning
So far, the uses for baselines I have identified are focused on the decision making of policymakers or the decision making of the voters in electing their policymakers. But baselines also aid the private sector in its own planning.
The tax and spending policies of the government matter to the private sector. That is, the expected course of those policies--and uncertainty around that expected course--affect the decisions private actors make. To take a few examples: this might be in the form of an individual planning for her retirement. In that case, the expected Social Security and Medicare benefits should affect how much she chooses to put away for retirement. (91) Or it might be in the form of a person deciding whether to make an investment and assessing the tax rate that the government might impose on the investment's return. (92) Or it might be in the form of a business assessing how much product to take to the shelves based on a projection of economic activity over the coming months, which will in part be dependent on government policies. (93)
Not surprisingly, there is, in fact, a market demand for predictions of government fiscal policy. Private analysts such as Goldman Sachs generate nuanced analysis of the course of the federal budget and its effect on the economy. (94) These analysts create their own budget baselines, projecting the expected course of government policies, and then employ those baselines in their analysis of economic trends. (95) Again taking the fiscal cliff as an example: in the midst of that debate, private sector analysts released reams of predictions about how that standoff would be resolved and the resultant effects on economic activity. (96)
G. Serving Irrational Biases
Finally, baselines can influence decision making through what can be described as purely "irrational" channels. In recent literature, this is the mechanism that has received, by far, the most attention, often to the complete exclusion of any other purpose for a baseline. (97)
There is now considerable behavioral literature exploring what is sometimes called "loss aversion," the "endowment effect," or "status quo bias." (98) As scholar David Gamage summarizes, also in the context of baselines: "These three labels refer to related phenomena--that individuals dislike losses more than they like gains, or that individual preferences are biased toward whatever they view as the status quo." (99) Importantly, recent behavioral research has questioned the degree to which this phenomenon is a significant force in human decision making. (100) To the degree this phenomenon is not important and other forces are at play, this could give even more reason to focus on the other purposes of budget baselines, as this Article has done.
To the extent they are real, these behavioral phenomena are thought to be driven by the framing of the decision, rather than the underlying substance. (101) Importantly, in this context, baselines do not help facilitate a rational decision-making process or constructively overcome problems of limited information or processing capacity. (102) Rather, the baselines potentially play directly into the underlying irrationality. (103) This distinguishes these behavioral phenomena from the other ways in which people can use baselines to make decisions--either in completely rational decision-making processes or as ways to overcome the limits of our information and rationality.
Take "loss aversion": Here, the framing of the starting point matters. If a policy is described as a tax increase, the policy might be less popular than if it is described as a tax cut, even if the resulting level of taxation is exactly the same (since the tax increase is framing the policy as a "loss"). (104) Or in terms of programs, benefit cuts may be seen as bad policy compared to maintaining benefits or increasing them, even if the level of benefits is the same in all cases. And whether a policy represents an increase or decrease in taxes or spending is defined by the baseline--the starting point.
Some baseline battles may be about one side or another trying to take advantage of these irrationalities to achieve their desired policy preferences. Again looking back at the temporary tax cuts: When advocating for extension of the tax cuts, those who wanted to make them permanent tended to describe a failure to do so as a tax increase--measuring from a baseline in which the tax cuts were permanent. (105) By contrast, those who wanted some part of the tax cuts to expire would, more often, describe continuing them as a further tax cut. (106) While neither side would say it outright, it seems that both sides were, in part, trying to frame the debate to lead people to their preferred policy outcome. With that said, the use of baselines in the tax debate did not always cleanly align in this way. For instance, the Obama Administration often used a baseline in which expiration of the high-income tax cuts was measured as a tax increase--its preferred policy--in part to claim credit for the deficit reduction. (107)
However, it is wrong to see baselines as solely playing into irrational decision making as they have sometimes been described in the past. As explained here, baselines provide useful information in a number of ways for enhancing decision making.
III. A TAXONOMY OF BASELINES
Even if baselines are useful in the ways described above, this does not tell us how a baseline should be constructed. And, in particular, it does not explain what set of laws should underlie the baseline, or why. This question has also been left largely unexplored, as if different kinds of baselines all answer the same questions. (108)
To put it differently, it is not sufficient to say that the baseline is the "starting point" for the legal regime. That is because there are different ways to define the starting point. To borrow Adam Samaha's new terminology for defining legal change, it in part depends on the "object." (109) That is, it depends on what particular aspect of the legal regime is being held constant (the object) in defining the baseline. (110) Thus, simply casting off a few words describing the baseline without any deeper thought is insufficient and leads to confusion and incoherence.
In answer to this confusion, this Part offers a taxonomy of budget baselines, describing what different baselines can offer, relating back to the purposes I identified in the previous Part. (111) Importantly, the labels used in this taxonomy have sometimes been employed in other (and often confusing) ways. (112) I differentiate from these earlier definitions. Finally, I conclude by emphasizing which baselines are most meaningful. In my view, priority should be given to variants of what I call the "expectations" baseline, with less important roles for what I describe as "low attention," "status quo," and "uncertainty" baselines. I also find that a true "current law" baseline is of only very limited use and simply not viable for most purposes.
A. Current Law Baseline
I have defined the baseline as the starting point of the legal regime, and, if that is the case, it seems like there should be a natural and leading candidate for budget baselines--namely, current law. The "object" should be literally the words on the page of the codebooks. Simply crack open the books, and voila, the baseline is there. Yes, it would require interpretation and analysis, which could be challenging, but it should largely avoid deeper questions about what should underlie the baseline. Some intelligent commentators have in fact expressed support for a current law baseline, at least in the context of tax provisions, as a way to minimize budget gimmicks. (113) However, the current law baseline is of only limited interest. In particular, it helps in a certain form of agenda setting but, even in that regard, "low attention" baselines, discussed in the next section, may be more useful.
A true current law baseline--as the name suggests--shows what would happen if no further legislation were enacted. And, in a very literal sense, it aids policymakers setting the calendar to determine what legislation should be considered and enacted. As described earlier, a current law baseline projection of the temporary tax cuts in 2012 showed policymakers that there was trouble on the horizon absent action. (114) And, every year, a true current law baseline would show large parts of the government shutting down absent renewal of annual appropriations--again, indicating appropriations should be on the "to do" list. (115)
Beyond its limited role in setting the legislative calendar, current law can also help to some degree in enforcing fiscal targets. Specifically, at least when it comes to some discrete programmatic areas, current law can serve as a baseline against which to measure costs. But even with regard to discrete areas, the current law baseline can be gamed by enacting temporary, costly legislation that is intended to be permanent, as described earlier and elaborated on in Part IV.C. With that said, the current law baseline cannot be used in enforcing overall fiscal goals because such a projection would assume expiration of all annually appropriated programs among other outcomes that no one expects to occur.
The current law baseline supplies little information beyond this, and the above examples illustrate its limitations. Congress was going to address the fiscal cliff one way or other--the only uncertainty was how much of the tax cuts would be made permanent and when. The same is true of annual appropriations; they will get done, even if not always on time. Congress does not need much of a reminder to put these on the legislative agenda. Further, current law does not show the expected effects of policies, and therefore cannot help policy making or private sector planning in that way. Nor does the current law baseline hold current service levels constant and so serve as a reference point. Finally, it does little to establish political accountability because it does not indicate the actual expected effect of policy action. Take the 2001 and 2003 tax cuts; they were enacted as temporary, but the Bush Administration seems to deserve some of the blame or credit (depending on your point of view) for their now permanent effects, and current policymakers deserve less of the blame or credit for having made them permanent than a current law baseline would suggest.
The bottom line is that "current law"--although perhaps the obvious candidate for building a baseline--is useful only in limited circumstances as a way of setting the legislative calendar and perhaps helping to enforce budget targets, though with the caution that current law has proven eminently gameable in enforcing budget rules.
B. Low Attention Baseline
Another type of baseline is what I call a "low attention" baseline. In this case, the "object" is law, assuming policymakers pay little attention to it. It is a baseline explicitly showing policymakers what would happen if they do not make a certain area of policy a focus of their agenda.
This is to be distinguished from the current law baseline. Despite the barriers to legislation, sometimes the easiest course of action--the way to pay the least attention to an area--is, in fact, to legislate. Take annual appropriations. Failing to appropriate at all is not the way to pay relatively little attention to the area. That case--a government shutdown--would tend to focus maximal attention on at least some parts of the appropriations process. (116) Rather, "low attention" might look like continuing the same real funding from the previous year.
Like the current law baseline, this baseline can help with agenda setting--and perhaps even more so. It does not simply aid in setting the legislative calendar. It helps indicate what areas actually require substantial focus from lawmakers. In fact, when analysts use projections to show that a certain area needs attention from Congress, it is this low attention baseline, as opposed to a current law baseline, on which they should be focused. To be clear, the low attention baseline may coincide with current law in some cases, but it need not--and in important cases (like appropriations), it will not.
With that said, the low attention baseline does not help much in terms of the other purposes of baselines. It is like the current law baseline in this way. The baseline can serve to set the agenda, but probably not much beyond that. When it comes to some areas (like the 2001 and 2003 tax cuts), there may not even be a plausible "low attention" path. In that case, a political battle was brewing that would garner much attention, and the question was how it would be resolved--not what would happen if policymakers did not put it on the agenda. And, even in serving the purpose of helping to set the agenda, the baseline is highly speculative; in many circumstances, it is not obvious what the path would be if policymakers truly paid little attention to an area.
C. Status Quo Baseline
Yet another type of baseline looks at what we are experiencing at this very moment (the object) and then holds constant what we are experiencing now (the benefits provided by a given government program or the amount of taxes people are paying right now) going forward. I will refer to this as a "status quo" baseline. This is meant to capture a concept similar to the Congressional Budget Act's "current services" baseline, or at least some interpretations of that baseline. (117) Notably, I do not mean this baseline to incorporate future changes in law if they differ from what is being experienced now. (118)
This baseline can serve the purpose of being a reference point, providing policymakers with a sense of what it means to continue the status quo into the future. As I described in Part II, there is something helpful about knowing what law is in place at this very moment. (119) At the least, it is a state of the law we have experienced, and some may see rational reasons to prefer the status quo as a matter of course over alternatives. (120)
However, this status quo baseline comes with limitations on its face. In particular, it does not provide information that would fulfill the other purposes of budget baselines. For instance, a status quo baseline, as defined here, tells us little about the expected course of budget policy. The tax and spending programs in place may be expected to change. Looking back at the fiscal cliff: one of the temporary tax cuts at issue in 2012 was an over-$100-billion-per-year payroll tax cut which had been enacted two years before, (121) and that most expected to go away. A status quo baseline would have projected it to continue, even as no major political leader was advocating to make it permanent. (122) Continuing the payroll tax cut in the baseline would therefore tell us little about the expected effects of budget policy. A status quo baseline also would not be particularly helpful in establishing political responsibility. It seems odd to blame or credit a policymaker with responsibility for an over-$100-billion-per-year tax increase for allowing a tax cut to expire that no one ever aimed or expected to make permanent.
In application, the limitations of this type of baseline are even more significant. This is because the baseline is meant to inform policy decisions about the future; although what is happening at this very moment may be descriptively interesting, we cannot change what has already occurred. And what is relevant to the future is not simply what is happening now but what it would mean for the same policies we are experiencing now to continue in the future. However, defining what it means to continue current policies into the future can, at times, seem like a dive down a rabbit hole.
For instance, what does it mean to continue the 2001 and 2003 tax cuts twenty or thirty years down the line when incomes and the economy more broadly are likely to be very different? One could argue that holding policy constant might mean increasing the tax cuts with either inflation or, alternatively, incomes. Over time, these different strategies would produce two very different results. When it comes to government programs, the problems become perhaps even more difficult. For instance, what does it mean to hold defense services constant after September 11th? With the United States facing a new terrorism threat to its security, simply adjusting funding levels for inflation would seem to produce a different kind of service than before. Or what does it mean to hold Social Security constant as the Baby Boomers retire?
This does not mean that a status quo baseline is useless. Especially for relatively short projection periods, a status quo baseline can still provide useful information. It can be possible, or at least easier, to translate what is happening right now to those future circumstances, which are similar to today, even if not exactly the same. We might adjust for inflation and program participation, while recognizing that this estimate of the "status quo" is not quite the literal "status quo." It is an adjusted status quo that reflects some changes but holds a number of meaningful items constant.
However, that information can degrade significantly as the projection continues deeper into the future. How far into the future does it become meaningless? That probably depends on the tax policy or program, but my intuition is that even the current ten-year budget window is pushing the limits for how meaningful a status quo baseline can be. At some point, the world is sufficiently different that trying to hold policy constant becomes a fruitless endeavor.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Abstract through III. A Taxonomy of Baselines C. Status Quo Baseline, p. 143-181|
|Publication:||William and Mary Law Review|
|Date:||Oct 1, 2015|
|Previous Article:||Regulating drones under the First and Fourth Amendments.|
|Next Article:||Basing budget baselines.|