Structuring the revenue forecasting process.

• Author: Kavanagh, Shayne C.
• Position: Report

[ILLUSTRATION OMITTED]

Forecasting is very difficult, especially if it is about the future.

--Niels Bohr, Physicist, Nobel Prize winner, 1922

Financial forecasting, which defines a government's financial parameters, is one of the finance officer's most important tasks. Accurate forecasting also forewarns a government about financial imbalances, allowing it to take action before a potential imbalance becomes a crisis, and forecasts can promote discussion about the future and how the organization might develop long-term plans and strategies.

This article describes a step-by-step approach to conducting revenue forecasts in local governments. As Bohr said, forecasting is not easy--but a structured approach can help forecasters ask the right questions, given the environment and audience; use the most appropriate techniques; and apply the lessons from one round of forecasting to future rounds. Structuring the forecasting process provides the following potential advantages:

* Forecasting is easier to replicate each year, leading to greater organizational learning.

* The process is transparent and relatively easy to explain to others, which should make it easier to get others to accept the forecasts.

* Because a structured process encourages the forecaster to adhere to forecasting best practices more diligently than an unstructured method would, the results should be more accurate.

The GFOA has adapted the following forecasting approach from the advice of leading forecast scientists' and the experiences of public finance practitioners and academic researchers. It involves the following steps: (2)

1. Define the Problem. What issues affect the forecast and presentation?

2. Gather Information. Obtain statistical data, along with accumulated judgment and expertise, to support forecasting.

3. Conduct a Preliminary/Exploratory Analysis. Examine data to identify major drivers and important trends. This establishes basic familiarity with the revenue being forecast.

4. Select Methods. Determine the most appropriate quantitative and qualitative methods.

5. Implement Methods. Use the selected methods to make the long-range forecast.

STEP I: DEFINE THE PROBLEM

The first step in the forecasting process is to define the fundamental issues affecting the forecast, providing insight into which forecasting methods are most appropriate and how the forecast is analyzed, as well as providing a common understanding as to the goals of the forecasting process. It will also help the forecasters think about what their audience might be interested in. While each local government will have distinct issues to consider, there are three key questions that all governments should consider as part of Step 1.

What is the Time Horizon of the Forecast? The time horizon affects the techniques and approach the forecaster will use. For example, short- and medium-term forecasts demand higher levels of accuracy because they will be used for detailed budgeting decisions. Longer-term forecasts are used for more general planning, not detailed appropriations. Also, some forecasting techniques lend themselves better to shorter- than longer-term forecasting, and vice versa. Longer-term forecasts will benefit from presentation techniques that emphasize the degree of uncertainty inherent in the forecast and the choices decision makers face.

Is Our Forecasting Policy Conservative or Objective? An organization can adopt one of two basic policies on how forecasting will be conducted. A "conservative" approach systematically underestimates revenues to reduce the danger of budgeting more spending than actual revenues will be able to support. An "objective" approach estimates revenues as accurately as possible.

Some public officials prefer conservative forecasts, which reduce the risk of a revenue shortfall. But this kind of forecast can also cause unnecessary fiscal stress during the budget process as the organization closes a fictitious revenue gap. At best, the government will incur opportunity costs, losing out on the benefits that could have been realized from programs and projects not undertaken. At worst, this approach could lead to unnecessary lay-offs or other disruptive cuts. Further, overly conservative estimates could lead to lost credibility as budgeting personnel become increasingly weary of the pseudo-financial stress.

The downside of an objective approach is a greater chance of experiencing an actual revenue shortfall during the year and, thus, incurring actual fiscal stress during the course of the year, as the budget has to be adjusted. Organizations that pursue an objective policy of revenue estimating should develop policies and practices to guard against these risks, such as budgetary reserve policies and contingency plans.

Are We in a High or Low Growth Environment? Land uses underpin the fiscal health of local governments. Forecasters should consider if their community is experiencing: (3)

* High growth and development--the rate of population growth is increasing each year.

* Moderate growth--there is growth, but the rate is declining.

* Low growth--population is declining or flat, or the rate of growth is negligible.

Each of these categories has different implications for the forecast. For example, high-growth communities need to carefully consider the costs and benefits of potential new development, while low-growth communities are usually more concerned with modeling the financial impact of maintaining an aging infrastructure.

STEP 2: GATHER INFORMATION

The next step is to gather information supporting the forecasting process, including statistical data and the accumulated expertise of individuals inside and, perhaps, outside the organization.

Take Stock of Economic Measures. Economic measures help provide context and might be directly useful in developing revenue assumptions. But first, you'll need to understand the role of economic measures in forecasting for your community. The foremost question is whether relevant indicators are available. For example, small communities might not match up perfectly with an indicator for a large geographic area--unemployment, for instance. In this case, the indicator could provide context, but it probably wouldn't be used in forecasting equations.

For other governments, the geographic area covered by the indicator will match the jurisdiction well enough for direct use in forecasting. In this case, consider the quality of the forecast for the indicator. Are credible, reliable future values available? For example, Fairfax County, Virginia, gets estimates of future employment in the area from a third-party economic analysis firm. State and federal...