Basing budget baselines.

• Author: Kamin, David
• Position: III. A Taxonomy of Baselines D. Expectations Baseline through Conclusion, with footnotes and tables, p. 181-219

4. Expectations Baseline

   An expectations baseline reflects what is expected to occur in the future. The object here is the law as it is expected to be. The future is not some undifferentiated mass of possibilities. There are more and less likely outcomes. In short, there is an expected path, even if it may be an uncertain one.

   When expectations are invoked, this immediately raises the question of whose expectations. Here, I mean to refer to what is objectively the best central estimate of the future path of the budget. That is, some predictions should be better than others--with smaller errors and less bias relative to what actually occurs. Admittedly, identifying the "best" projection is no small feat. Perhaps it would be delivered by some combination of nonpartisan budget experts, whether at official government agencies like CBO or at independent think tanks, that closely watch budget developments and on which policymakers, the press, and the public rely. But the point is that I do not mean expectations to refer to anyone's subjective expectations of the course of the budget irrespective of the likely accuracy of the guess.

   This expectations baseline has the potential to offer considerable information. As suggested earlier, the expected effects of policies should matter to both policy making and private sector planning. (123) The 2001 and 2003 tax cuts were expected to have a very different effect than the one actually stated in law. (124) At the time, they were expected to be continued at least in significant part. (125) Thus, that expectation should have informed the decision to enact them.

   Perhaps less obviously, the expectations baseline can prove useful in setting the policy agenda, establishing political accountability, and enforcing budget rules. (126) In terms of agenda setting, the most relevant information is the question of what would happen if an issue is not brought onto the agenda as of now--information provided by what I call the "low priority" baseline. And the "low priority" baseline can actually be conceptualized as a particular type of expectations baseline. It is what is expected to occur, contingent on policymakers not paying much attention to an area. In terms of political accountability, it seems particularly relevant to look at the expected course on which policymakers have set us (what I described earlier as the post-policy baseline) and to compare that with the expected course before policymakers took over (the pre-policy baseline). Finally, in terms of budget enforcement, using a baseline that better reflects expectations can potentially avoid certain kinds of gamesmanship. Whereas a pure current law baseline can lead policymakers to enact temporary legislation and claim that it costs less than its actual expected effect, (127) an expectations baseline avoids this gaming. Furthermore, the baseline can be used to enforce targets in terms of overall projected budgetary outcomes (deficit and debt levels), (128) unlike a current law baseline, which is useless for such purposes.

   To be clear, setting an expectations baseline involves its own share of speculation. There is the actual technical question of how to estimate the expected policies in the face of uncertainty. The best method would likely be some probabilistically weighted average of various policies, interacted with economic factors. In practice, that may prove complicated across the full range of the budget. Another possibility is a projection of the most likely set of policy outcomes, which is not necessarily the same thing but may still provide similarly useful information.

   The long-term fiscal challenge generally, and the 2001 and 2003 tax cuts specifically, highlight the issue of uncertainty. The strong expectation of experts and markets is that the government will finance itself without defaulting on its debt. (129) By this I mean the expectation is that taxes will eventually be raised and spending cut in some combination that is sufficient to balance the government's books--or, more technically, sufficient so that the federal government fits within its long-term budget constraint. (130) In the case of the 2001 and 2003 tax cuts, this means that, as a matter of budgetary arithmetic, their enactment will eventually lead to offsetting tax increases and spending cuts of equivalent magnitude to finance these tax cuts. (131) To be clear, the long-term fiscal challenge may not be resolved with some combination of policies that people consider optimal--that part is by no means necessary--and the same is true specifically of the policies needed to finance the 2001 and 2003 tax cuts. Nonetheless, an expectations baseline should, over the long-term, reflect whatever combination of policies is expected to close the long-term gap and finance deficit-increasing policies.

   In previous work, I made an effort to identify what policies might be employed to close the long-term fiscal gap, noting that a combination of policies already proposed, though not enacted, by policymakers could do so. (132) My approach of projecting "least common denominator" deficit reduction--that is, the least each side had proposed in any given area--was one way to try to identify the expected course of policy. (133) There are certainly other ways, but efforts like these are needed to guess the future course of budget policy over the long-term and then assess whether that course is a good or bad one.

5. Policy Uncertainty Baseline

   Some might say, though, that an expectations baseline showing no fiscal shortfall--and the 2001 and 2003 tax cuts, and other deficit increasing policies, as financed--assumes away a rather large problem. Perhaps we expect that policymakers will eventually figure out a way to deal with the deficit, but that does not mean they have done so already. It is a rather curious chicken-and-egg problem. Should an expectations baseline show the long-term fiscal problem as solved when we remain uncertain as to how that will actually occur? Should we sit back and relax, secure in the relative certainty of future action?

   The answer to this is "no." We should probably differentiate the fiscal policies that are highly uncertain from those that are less so. At the least, we can try to measure the degree to which policymakers have generated more uncertainty than they would have if they had better specified a sustainable long-term fiscal path.

   This brings us to the "policy uncertainty" baseline. While the name might seem exotic, it is a baseline that should be familiar to anyone who follows federal budget debates, even from afar. When you see a projection of the federal budget on an unsustainable trend--as in deficits or debt rising perpetually into the future--you are probably looking at a policy uncertainty baseline, or at least a baseline that is akin to this. And, of course, this is familiar. The many warnings of the unsustainability of the federal budget often come with just such graphs, with lines ascending into the heavens. (134)

   Given how often such baselines get rolled out, it might come as a bit of a surprise that it is rarely (if ever) specified what those baselines are in fact projecting. (135) Yes--the technical assumptions underlying the projections are given. (136) But the theoretical framework--as in what all of those assumptions are meant to capture--is not actually specified. (137) It certainly is not a current law baseline, as it does not reflect the debt limit in place and much of the federal government disappearing as appropriations expire, among other items in current law. It probably is not a low attention baseline or, if it is, it is not a particularly meaningful one--since, even if deficits do not get placed at the top of the agenda today, they seem very likely to be on the agenda if debt rocketed up. Even debt at current levels has sparked negotiation after negotiation. (138) Further, it probably is not a status quo baseline, as these baselines often incorporate changes in services (like spending cuts or tax increases) that people expect to occur but have not yet happened. (139) Additionally, as I discussed, status quo baselines tend to lose coherence as they look further and further into the future. (140) And, finally, it is not an expectations baseline, at least as defined above, because few expect that the United States will actually have a debt load equivalent to several hundred percent of GDP (and rising) at some point in the future.

   This brings me to what I am calling the "policy uncertainty" baseline. Based on policymakers' legislative actions and proposals, we may project a certain minimum level of taxes--a level below which it is unlikely taxes will fall--and a certain maximum level of spending--a level above which spending is unlikely to rise. The gap between these two levels will be closed, but there is substantial uncertainty about the composition of the policies that will fill that gap and the timing. This uncertainty is the focus of the policy uncertainty baseline. To put this in Samaha's terminology, the "object" in this case is the amount of uncertainty that policymakers generate due to their failure to better specify a path to a fiscally sustainable future. (141)

   This can be conceptualized in terms of the fiscal gap--a metric of the fiscal shortfall popularized by Alan Auerbach. (142) This gauges the size of the tax increases or spending cuts needed to stabilize the debt-to-GDP ratio over given periods of time, which is one metric of fiscal sustainability. (143) The fiscal gap essentially quantifies the amount of policy uncertainty that exists. Compared to the baseline, spending will have to be cut by a certain amount and taxes raised by a certain amount, and there is confidence that such actions will occur; (144) what is uncertain is the timing and composition of that deficit reduction. (145)

   There are costs to such uncertainty, making the uncertainty worthy of quantifying. (146) For instance...