The key to a credible budget forecast is having transparent and realistic assumptions. This will communicate a true picture of your agency’s financial condition, and whether that is positive or negative, it will provide an empirical basis for future budgetary actions.

City of Davis’ 20-Year Forecast Model

Nitish Sharma, former Davis finance director and current CSMFO Sacramento Valley Chapter Chair, generously allowed Davis’s city’s forecast model to illustrate the points in this article.

The City of Davis is a full-service, 68,000 population, university town, with a $70 million General Fund budget. Prior to 2016, the City prepared a 5-year forecast for its annual budget. But it was looking for a more robust forecasting model that allowed staff to easily model various outcomes over 20 years, including funding of infrastructure maintenance. The forecast model we prepared has guided the last three City budgets and given the City a valuable planning tool. A comprehensive forecast discussion and a condensed version of the forecast model itself are available on the City’s website. Major forecast assumptions are as follows:

• Recessions – On average, a recession has occurred every 7 years since the 1920’s, ranging from moderate revenue reductions of 5-10% from normal growth, up to the substantial multi-year losses of the Great Recession. A typical local agency forecast ignores economic cycles and may rely on conservative linear trends to try to compensate. The value of a forecast model is that you can easily show the impacts for recessions of different timing and magnitude. Davis assumes moderate recessions every 7 years starting in FY20-21, with recovery of 90% of losses over the ensuing 3 years. This provides a realistic “stress test” for City finances.
• Property Tax – The growth of this key tax is determined by specific elements. Davis assumes the 2% Prop 13 inflator for 96% of existing parcels; an average 40% increase for the other 4% of parcels due to changes in ownership; new construction projected by City planners; and no additional Prop 8 recoveries. This results in pre-recession growth of 4.1%.
• Sales Tax – Growth is determined by the projections from the City’s tax auditing firm for both its statewide 1% tax rate and its local voter-approved 1% rate. The pre-recession growth rate for both taxes is 2.8%.
  Major Developments – The forecast includes the projected revenues and expenses associated with two major proposed housing developments, based on the City’s fiscal impact analyses of those developments.
• FTE Growth – Agencies cut staff significantly in the wake of the Great Recession, and some of that lost staffing has been restored during the economic recovery. In addition to the expense linked to the two major developments, Davis assumes the future addition of 1.0 FTE annually to respond to other population and workload growth.
• COLAs – Cost of living adjustments are determined by the meet and confer process, but a realistic forecast needs to reflect the inflationary pressure on wages, especially if the forecast grows all other revenues and expenses. The Davis forecast assumes annual 2% COLAs.
• Growth Within Pay Ranges – Eligible employees typically receive 5% merit increases before they hit top step of their job class. Given typical employee turnover rates of 5-8% annually for local agencies, one out of every 12 to 19 positions will be vacated each year, and savings will be realized when those positions are re-filled at a lower step. Based on an analysis of its positions, the Davis forecast includes a 0.25% net increase from the combined impact of merit increases and turnover savings.
• Pension Discount Rate – CalPERS has slowly lowered its discount rate (projected investment return) from 8.75% to 7% over the past 30 years. A trend of slowly declining average returns and a policy of reducing volatility in rate setting indicate a risk of continued slow reduction in the discount rate. Davis projects a reduction from the current 7% to 6% over the next 20 years. Figure 1 shows the long-term total pension costs as a percent of payroll. This is a pattern typical of most local agencies: increased fiscal capacity will become available as costs begin to fall in the 2030’s due to the amortization of unfunded accrued liability (UAL) costs.
• Vacancy Rate – A budgeted vacancy savings rate of 3% is applied to salaries and benefits (excluding overtime and pension UAL) to reflect the inevitable periods that various positions will remain vacant during the year. (This is in addition to the turnover savings noted above.)
• Other Expense – CPI is assumed to be 2.0%, and this is used for most other expenses.

Infrastructure Funding a Major Priority

The Achilles’ heel of most jurisdictions is infrastructure maintenance. These expenses are often the first to go when budget cuts are required, and most agencies are not well-informed as to what their long-term maintenance needs really are. Davis is an exception. An active Finance and Budget Commission has long advocated infrastructure investments. As a result, the Council has commissioned engineering studies that have identified long-term funding plans for street and bike path resurfacing, public facilities, traffic maintenance needs, and parks and swimming pools. From these plans, the City knows how much should be spent in each year, and the forecast shows what is projected to be spent. Supplementing other dedicated sources, the General Fund contributes a minimum of $3 million annually, plus any excess of its reserve goal (see Figure 2 for total funding versus need). This is projected to fund 50% of the City’s identified maintenance needs over 20 years. Less than 100%, to be sure, but it is half of a known obligation, which beats not knowing your long-term maintenance and repair needs.

Positive Outcome, If Sales Tax is Renewed

The 20-year outlook for Davis remains positive – despite incorporating moderate recessions, modest staffing growth, and a decline in the pension discount rate – largely because of its 1% local sales tax rate. Figure 3 shows the forecasted unassigned fund balance assuming voters renew that tax in 2020. If they don’t, a $9 million revenue loss would quickly drive the budget into deficit. However, the model can also generate budget balancing scenarios to show the consequences of such a revenue loss, which in itself may provide a compelling case for renewal of the tax.

In good times and bad, a flexible forecast model and realistic assumptions will give you the information you need for informed, prudent and timely budget policy decisions. Forecasting: it’s in your future!