Budget projections can provide a valuable look at what the world will look like if current laws remain in place, but they cannot tell us how or when to act in the face of projected imbalances.
All projections are based on assumptions around some set of variables and are vulnerable to error, or the variability in predictable factors, as well as uncertainty stemming from unforeseeable circumstances. Because error and uncertainty grow as the projection horizon is lengthened, in some cases, lengthening the window is not useful and can degrade decision making.
This paper reviews the utility of four sets of projections: long-term budget projections by the Congressional Budget Office, 75-year projections for Social Security, 75-year projections for Medicare, and infinite-horizon generational accounts. Within its 75-year projections, the CBO has recently shifted attention towards the first 25 years, a welcome change, as the later years are based on unrealistic assumptions. In contrast, the 75-year Social Security projections are useful, as the variables that the projections are based upon can be reliably predicted. Because Medicare costs depend heavily on unpredictable advancements in medical technology, the 75-year projection window should be pared down to a 25-year period. Infinite horizon generational account projections simply freeze the underlying variables at their value in the 75th year and extrapolate into perpetuity, providing no new information after the 75th year.
Whether and when action should be taken based on a projection depends not just on the uncertainty surrounding a projection, but also on the political options available. Taking action now can raise the cost of, or even preclude, future action. If political constraints only allow for less-than-optimal policy action, it may be beneficial to wait for political or economic conditions to change.