CBO's scoring window: why it matters

Michael J. O’Grady, PhD, a senior fellow at the National Opinion Research Center at the University of Chicago, writes about the Congressional Budget Office and how it scores proposals.  The CBO recently issued a price tag for the Senate Finance bill on health reform that is being voted on right now.

The current debate our country is having about overhauling our health care has highlighted the sometimes challenging interaction between the worlds of budget policy and health policy.  One such interaction is the issue of scoring windows.  The Congressional Budget Office (CBO), the official scorekeeper of reform proposals, typically projects spending for a 10-year period.  However, on September 14, Senator Kent Conrad (D-ND), Chairman of the Senate Budget Committee, requested that CBO provide a twenty—year estimate of the Senate Finance Committee bill.  He was right to do so, and here’s why.

First, let’s talk about why CBO moved from the original 5-year window to the current 10-year window.  The shift occurred for a number of reasons.  The Budget Committees began looking at a longer time horizon.   In the FY 1994 Budget Resolution Congress established a 10-year budget “point of order” in the Senate.  Then in 1995, Speaker Gingrich introduced a seven year balanced budget plan, adding further momentum to estimate budget impacts beyond the five-year window.  By 1996, CBO released its first 10-year budget baseline.

Another reason for the shift was that the five-year window was also easier to game for members of Congress.  Committee Chairmen and their staffs had become increasing facile at designing new federal programs, which showed atypically low spending until the sixth or seventh year after enactment.  In other words, they had effectively “gamed” the CBO scoring process and made their proposals look much less expensive than they really were.  Pushing costs to the eleventh or twelfth year was possible, but it was much more difficult.

There are specific goals in doing cost estimates whether it’s for five, ten, twenty or some other time period.  The primary and overarching goal is to provide decision-makers with the most accurate estimates of the possible budgetary impacts of a legislative proposal. There are secondary goals, such as providing policymakers with a consistent metric to compare proposals across different policy areas.  To serve this second goal CBO consistently applies the 10-year window across all policy areas from aircraft carriers to school lunches.

In the current health reform debate, the secondary goal of providing a consistent set of measures across policy areas is sometimes in conflict with the primary goal of providing policymakers with the best information possible. 

One example is in the fight against chronic disease. In developing new effective programs to fight chronic disease, much of the clinical and spending impact can occur outside the 10-year window. The National Institutes of Health and the Centers for Disease Control have found that early and aggressive programs to manage diabetes can prevent or delay costly and devastating complications such as kidney failure, heart attack, stroke, blindness, and amputation.  However, the NIH funded research also indicates that an intensive diabetes management program only begins to yield significantly lower spending on diabetes treatment and complications after about eight years, with the bulk of the reductions occurring between years 10 and 25.  The result is that CBO’s current scoring system underestimates the longer-term economic benefits associated with aggressive diabetes management programs.  This, in turn, makes it less likely for Congress to support government funding for such efforts.

A perfect example of how the scoring window might—or might not—work is the Medicare Modernization Act.  Although the bill passed at the end of 2003, Part D drug benefits were not offered until 2006.  Having served as the Assistant Secretary for Planning and Evaluation (ASPE) during the implementation of the Medicare Modernization Act, I can attest to the need for two years to build the risk adjusters, design the bidding systems and all the other infrastructure necessary to ensure the new system would work.  Still, a 10-year budget estimate that has very low, atypical spending for the first two years, gave policymakers a very unrepresentative picture of the ongoing costs of that program. 

The same concern exists in the current debate over health care reform.  In the President’s recent address to Congress, he indicated that his health care reform plan would not provide coverage for the first four years after enactment in order to build an effective infrastructure necessary for a new system.  Based on our Part D experience, four years sounds about right.  Still, a 10-year budget estimate that has atypically low spending for the first four years, will give policymakers a very unrepresentative picture of the cost of the new program.  Senator Conrad clearly has the right idea in asking for twenty-year estimates.

While a 10-year-estimate works well for the vast majority of congressional proposals.  Data from clinical trials and epidemiologic studies demonstrate that the course of treating chronic disease does not fit within this narrow framework.  Imposing a 10-year window runs the risk of significantly underestimating offsetting savings.  In major legislative changes that will take years to implement fully, imposing a 10-year window risks leaving policy makers with an inaccurate impression of the typical spending and savings associated with the new program. 

CBO understands and embraces their primary mission to better inform congressional decision-making.  In those cases where the debate is better served by the longer view and where the methodological rigor exists for reliable projections beyond ten tears, there should be the ability and the necessary tools to show longer time windows in the “official” estimates.