May 27, 2021
Our office recently released its multiyear budget outlook based on our evaluation of the Governor’s May Revision budget proposal. An important element of this annual exercise is our independent estimate of the total cost of the Governor’s May Revision budget proposals for several years into the future. A key finding from our analysis this year is that the Governor’s ongoing spending proposals are probably too high relative to his administration’s revenue projections. This manifests in operating deficits that persist through the outlook horizon (2024‑25). Adopting a budget with this combination of assumptions is inadvisable in our view. One reason is that—given the state’s balanced-budget requirement—doing so requires assuming that the multiyear budget projections are wrong. To appreciate why this is problematic, it is useful to revisit some of the reasons for doing multiyear fiscal forecasting in the first place.
Most substantively, multiyear estimates offer the Legislature a vital measure of the state’s ongoing budget capacity. Without a reasonable approximation of the state’s current fiscal trajectory and bottom line over the next several years, policymakers are essentially flying blind. While events of the past year show that attempting to predict the future is a fool’s errand, the value of multiyear fiscal estimates is not in their predictive precision. Rather, policymakers need insight on the condition of the state’s underlying fiscal operations—annual revenues relative to expenditures—outside of any near-term surplus or deficit. Many times, such surpluses or deficits are one time and not indicative of the state’s structural budget condition. Moreover, many programmatic initiatives take effect in January—midway through the fiscal year—or take several years to be fully implemented. Consequently, absent fiscal estimates that extend several years into the future, the Legislature lacks a basis on which to judge whether the full annualized cost of existing state commitments is sustainable.
Building on these policy-related benefits, another reason for regularly preparing multiyear fiscal estimates is that doing so sends a reassuring signal to financial markets and the state’s creditors. As a formal part of their credit review, for example, rating agencies explicitly consider whether state and municipal government bond issuers have a policy to develop multiyear budget estimates. Reflecting a belief that the practice strengthens a government’s credit profile, rating agencies have cited California’s multiyear projections as a positive feature of its financial management. While the state’s overall bond rating is a distillation of numerous individual factors—many of them exogenous—budget management is a category over which state officials have direct control. Proposition 2 (2014) enshrined in the State Constitution the Department of Finance’s (DOF’s) requirement to semiannually prepare a multiple-year General Fund forecast. Additionally, our office has a longstanding practice of preparing a fiscal outlook twice per year. Insofar as investors and analysts deem these projections credible, California’s institutional commitment to doing multiyear budget estimates contributes to a higher bond rating and, consequently, more favorable financing terms in the credit markets.
The value of doing multiyear budget estimates is evident when considering the role they can play in budget development. When developing our main forecasts, both our office and DOF make their best estimate, aiming for the median scenario. Whether the estimates indicate the state’s fiscal trajectory is on a course to remain balanced, generate a surplus, or fall into a deficit can—and should—be pivotal in the budget process. Prudence dictates that only when baseline projections show operating surpluses (annual revenues exceeding expenditures) can new ongoing commitments or tax reductions be considered. Conversely, if projections indicate a looming structural deficit, the estimates are helpful in functioning as an early warning system. In that case, the Legislature gains the critical advantage of additional time with which to make necessary budget adjustments or—just as crucially—a chance to avoid worsening the problem.
What do the new multiyear estimates tell us? For one, they suggest the Governor’s emphasis on one-time spending is appropriate. Of the large $38 billion discretionary surplus at the May Revision, we estimate that the Governor initially would allocate just $2 billion for new ongoing purposes. Yet even this amount may be too much. Whereas DOF projects future revenue growth to slow, we expect the out-year cost of these spending proposals (and underlying costs) would increase significantly. This combination is untenable and is what gives rise to the operating deficits shown in our multiyear outlook. As we see it, the Legislature has two main options when planning its budget, both of which also should influence its spending choices.
First, the Legislature could adopt a budget premised on our office’s higher revenue estimates. While this would eliminate the projected operating deficits, the assumption of higher revenue growth brings with it increased risk that actual revenues will underperform the forecast. To manage this risk, we recommend using more of the current surplus to build discretionary reserves and pay down more debts. Doing so would necessitate trimming some of the Governor’s proposed spending.
Second, if the Legislature preferred to use the more cautious DOF revenue estimates, we recommend significantly reducing the amount of ongoing spending proposed in order to eliminate the operating deficit. That is, pass a budget resulting in multiyear estimates that at least break even. Under this approach, we also still recommend building larger discretionary reserves by restoring the budget tools that were deployed in last year’s budget.
Note that under either revenue assumption, we recommend providing for larger discretionary reserves. While total reserves, proposed at $19.8 billion in the May Revision, are slightly higher than the $18.6 billion proposed in January 2020—just prior to the pandemic—they have not kept pace with spending growth. Consider that as a share of proposed General Fund expenditures, reserves in the May Revision would edge down to 10 percent of expenditures from 12.8 percent in the January 2020 proposal. A year during which the state anticipates its fastest year-over-year revenue increase in four decades is not the time for passivity when it comes to reserves.
While the choices described above are not easy, it is thanks to the multiyear estimates that the Legislature can clearly weigh the policy trade-offs involved. Yet in order for the state to continue to realize the various benefits associated with doing multiyear fiscal projections, the estimates must play a meaningful part in the budget process. In the present circumstance, this means adjusting the mix of revenue, spending, and reserves proposed in the May Revision.