Hearings of the Committee on Rules
"Biennial Budgeting: A Tool for Improving Government Fiscal Management and Oversight"
James R. Horney
Senior Fellow, Center on Budget and Policy Priorities
1. Your experience at CBO highlighting the fact that estimates can change in "just a few months" could be used to argue against a biennial budget, as you do. However, couldn't that argument be turned against our current process just as easily - by saying that things change every few months so why do we still do an annual budget? Your argument might suggest we should be most concerned about uncertainty and therefore should probably be doing new budgets every few months. Could you respond? What makes an annual budget that is fraught with uncertainty any better than a biennial budget fraught with the same uncertainty?
Ideally, changes in the economic and budget outlook would be taken into account more often than once a year. If all Congressional budget decisions were made centrally, or if the actions of the various committees of Congress were not subject to any collective direction and control, then it would be possible to make nearly constant adjustments in response to changes in the economic and budget outlook without disrupting the budget process. But, as long as Congress wants to exercise strong collective control over budget legislation that is crafted by numerous subcommittees and committees in two Houses of Congress, frequent changes in the directions given to those subcommittees and committees will almost certainly disrupt the process.
Under the current budget process, the Congress adopts a budget resolution (and the economic and technical assumptions that were used in formulating the resolution and will be used to determine whether subsequent legislation complies with the resolution) early in a session of Congress. The appropriations and authorizing committees can then proceed to consider legislation knowing what they are allowed to do and how the legislation they are considering fits into the overall budget plan. Changing what legislation is allowed and how legislation will be "scored" in the middle of that process would promote confusion and uncertainty and make it harder for committees to do their work and for the Congress to exercise central control over budget decisions (it is hard to enforce rules when the rules are constantly changing).
Of course, that logic could be used to argue that no changes should be made to reflect updated economic and budget information between sessions of Congress. But I think that, in the absence of truly dramatic changes, updating the budget annually to reflect changing economic and budget conditions provides the most appropriate balance between the competing goals of being responsive to changing conditions and maintaining an orderly budget process.
2. Based upon your concerns about the uncertainty of predictions, and understanding your basic opposition to biennial budgeting, I would still like to ask you for your views on what technical mechanisms should be put in place to mitigate the problems you foresee if biennial budgeting were to be enacted? Are there specific recommendations you could make with respect to the work done by CBO and OMB, the timing of budget preparations and review and forecasting?
I think it is important to avoid erecting too high a barrier to revisions in the budget resolution in the non-budget year. I realize that making it easy to revise the budget resolution in every non-budget year could reduce the anticipated benefits of having more time to spend on non-budget issues. But if the resolution cannot be readily revised when there is general agreement that changed circumstances call for changes in the budget, the Congress will be less able to make appropriate collective decisions about budget needs and priorities in those years and to enforce those decisions during the consideration of subsequent legislation.
I think it is also important to devise mechanisms for taking changes in the economic and budget outlook into account in the Congressional Budget Office's cost estimates when budget legislation is considered in the non-budget year and the budget resolution has not been revised. If budget enforcement in the non-budget year is based on cost estimates of legislation that rely on clearly outdated economic and technical assumptions from the previous year, the credibility and effectiveness of that enforcement would be undermined. I do not know of a simple way to avoid this problem. I believe that the Rules Committee staff should work with the Congressional Budget Office, House and Senate Budget Committee staff, and perhaps other committee staff to develop ways to update estimates when necessary.
3. Some have suggested possibly including a sunset provision for biennial budgeting. How long do you think it would take before we could assess whether a biennial process is working? What would be the criteria you would use to judge its success or failure? Do you have any thoughts on another suggestion that has been made that we should consider implementing biennial budgeting on a pilot basis just for certain segments of the budget?
If there is going to be a move toward biennial budgeting, I believe it would be wise to first try biennial budget resolutions before moving to biennial appropriations. In any case, I believe that establishing a sunset date for the biennial budgeting legislation would be appropriate. It probably would take two cycles — four years — to allow observers to reach even a tentative conclusion about the success or failure of the experiment.
In judging that success or failure, it will be important to look both at the extent to which promised benefits — increased time spent on more effective oversight and other important Congressional activities — have been realized and the degree to which potential bad effects — increased use of supplementals and other ad hoc budget legislation or failure to respond in a timely manner to national needs — have occurred. It is important not to ignore the qualitative aspects of this assessment. For instance, spending more time on "oversight" will not mean much if the oversight is not carried out in a way that is likely to lead to a real improvement in the way federal programs operate. Similarly, limiting the number of supplementals and other non-budget year budget legislation will not represent a good outcome if the Congress is simply failing to deal with real needs in the non-budget year.
4. What role do you see the authorization process currently playing in the budget process and do you think that role should change? Do you think changing or strengthening its role would improve the role and condition of programmatic oversight?
Obviously, authorizing committees cannot ignore the budgetary effects of programs within their jurisdiction — especially the budgetary effects of mandatory programs within their jurisdiction. But I think the authorization process should focus more than it currently does on the goals of federal programs and whether those programs are being run in an effective and efficient manner.
5. You raise the concern that off-year budgeting, i.e. supplemental work, will tend to be done in a more ad hoc basis than that which occurs under the current system. Could you explain further what you mean by an ad hoc basis?
When I say that more legislation will be considered on an ad hoc basis I mean that more legislation will be enacted that is not considered in the context of an overall budget plan determined by Congress and that has not gone through the full process of consideration that most budgetary legislation goes through now. The current system allows the Congress to establish overall budget priorities each year and to influence the particular budget legislation that will subsequently be adopted during the current session of Congress by setting a revenue floor and allocating spending authority to various committees. That subsequent legislation must be consistent with the budget resolution or face political and procedural hurdles. In addition, most budgetary legislation goes through a process of extensive hearings, subcommittee mark-up, full committee mark-up, and lengthy consideration by the full House and Senate.
I believe that the Congress will find it necessary to pass a significant amount of budgetary legislation in the non-budget year. Unless a revised budget resolution is adopted each year, that legislation will not be consistent with the priorities set in the budget resolution and there will be no mechanism for the Congress to consider the policies embodied in that legislation in the larger context of overall national needs. In addition, it is likely that such non-budget year legislation — like current supplemental appropriations — will go through a less extended and less thorough review process during its consideration.
6. You discuss in great detail in your written testimony the difficulty of making projections. We already now make 5 and 10 year projections that everyone readily admits are not without significant error. However, we also make 1 and 2 year estimates now that are much less volatile and which many argue have actually improved. Would you care to comment on this?
Although 1 and 2-year baseline budget projections are less uncertain than longer-term projections, even the 1 and 2-year projections are subject to large misestimates. This is shown in an analysis CBO presented in Appendix B of its January 1999 Economic and Budget Outlook: Fiscal Years 2000-2009. In that analysis, CBO calculated the difference between the budget resolution's assumptions — which are generally based on CBO's revised baseline projections that are usually released in March or April each year — and actual outcomes for the fiscal year starting on October 1 of the same year. The analysis covers the budget resolutions for fiscal years 1980 through 1998. (CBO has not updated the analysis to include fiscal year 1999 since no budget resolution was adopted for that year.) The errors shown in this appendix include the effects of actual policies that differed from those assumed by the budget resolution. But if those effects are subtracted, the average absolute value of the remaining economic and technical errors (that is, ignoring whether the errors represent over- or underestimates) in the deficit projections for fiscal years 1980 through 1998 is 4.5 percent of actual annual outlays. Based on CBO's most recent projections, a 4.5 percent error in the surplus projection for fiscal year 2001 would represent an over- or underestimate of more than $80 billion.
CBO's analysis does not indicate that these projections have become less uncertain in recent years: the absolute economic and technical error as a percentage of actual outlays increased each year in 1996 through 1998 and exceeded the 1980-1998 average absolute error in each of those years.
7. When you discuss the difficulty in projections are you specifically concerned with projections of spending revenue?
I am concerned about the uncertainty of both revenue projections and outlay projections. CBO's analysis of projection errors in Appendix B of the its Economic and Budget Outlook: Fiscal Years 2000-2009 indicate that projections of revenues and outlays are both subject to significant error. Excluding the effect of differences between actual policies and those assumed in the budget resolution, the average absolute error over the 1980-1998 period for outlay projections is slightly higher than the average absolute error for revenues.
