With the next meeting of the Revenue Estimating Conference approaching, there are some questions about how Iowa’s process works and how it compares to the ways other states forecast their future revenue. And as there are fifty states, there are fifty different processes for determining how much tax revenue a state will generate.
Iowa’s process to set revenue projections, the Revenue Estimating Conference, is laid out in Iowa Code section 8.22A. The law requires a three-member panel to convene at least three times a year where they review and set estimates for revenue, tax refunds, interest earned by the state, and gaming revenue. The three members include a representative of the Executive branch (usually the director of the Department of Management), a representative of the Legislative branch (usually the fiscal services director of LSA), and a member of the public mutually agreed to by the other members.
The group meets generally meets in October, December, and March to do their work. The key meeting is in December, where the panel sets the official revenue estimate for the upcoming fiscal year. The March meeting can play a key role in the budgeting process, if the panel’s revenue projection for the upcoming year is reduced from their December forecast. When this occurs, the General Assembly is required to use the lower number as the revenue estimate for the next fiscal year.
In 2015, the National Association of State Budget Officers published Budget Processes in the States which includes a review of process each state uses to project future revenue. The majority of states utilize some process where the executive and legislative branches come to agreement on a revenue forecast. How they get to those consensus projections can vary greatly.
South Dakota starts the budgeting process with the governor submitting his budget and revenue estimate to the Legislature in early December each year. The process then shifts to the Legislature’s Joint Appropriations Committee, which receives revenue projections from the Legislative Research Council (their version of LSA) and the governor’s Bureau of Finance and Management. Within the first 20 days, the Joint Appropriations Committee must decide which projection they will use for the state’s budget.
Missouri’s forecasting process is unique in the fact that it is not established in state law. While the governor is required by law to include a revenue estimate as part of their budget recommendation, the legislature is not statutorily bound by this. They instead use a practice where the state’s budget office, both chambers of the legislature, and a professor from the University of Missouri develop a staff-level consensus estimate. From here, the Governor and legislative leaders decide whether or not to accept the staff recommendations and agree on a consensus estimate. Since this is not required by state law, there have been years when actual consensus was not reached.
Of the neighboring states, the one with the process most similar to Iowa’s is Nebraska. Their revenue forecast is established by Nebraska Economic Forecasting Advisory Board. This is a nine-member board, with four members appointed by the Governor and five appointed by the Legislature. Those appointed to these positions are required to have expertise in economics, economic forecasting, or tax policy and they serve four year terms. The Board receives revenue estimates from the Nebraska Department of Revenue and the Legislature’s Fiscal Office and then develops a consensus revenue forecast.
In Minnesota, the process for estimating state revenue is significantly different than what happens in Iowa. Revenue forecasting is an exclusive function of the executive branch, with the governor’s budget agency – Minnesota Management and Budget - being responsible for publishing five-year revenue estimates twice a year. That agency’s Economic Analysis Division is the formal revenue estimating group, which makes formal estimates in November and February each year. They also issue economic updates in January, April, July, and October and makes revisions to the estimate. Wisconsin has a similar approach. The revenue estimates are made by the Department of Revenue.
So what is the best approach to estimating state revenue? Some groups have tried to lay out best practices for revenue forecasting. The left-leaning Center for Budget and Policy Priorities identifies five practices that they feel are necessary to improve states’ revenue forecasting process. The five practices are:
The governor & legislature should jointly produce the revenue estimate
The forecasting body should include outside experts
The forecast and its assumptions should be published and made easily accessible on the internet
Meetings of the forecasting group should be open to the public
Estimates should be revised during the year
According to their survey, Iowa is one of fifteen states that utilize all five practices. But, CBPP’s best practices all focus on process. They do not focus on the accuracy of the forecasts. This is really the key data point.
One group that has not only reviewed state forecasting practices, but also assists states to improve their process is the Pew Center for the States. The Pew Center worked with the Nelson Rockefeller Institute of Government in 2011 to study state revenue forecasts and why states were experiencing greater volatility in tax revenue. In their 23 year review of state revenue forecasts, the study came to an interesting conclusion. States were more likely to under-estimate state revenue, producing greater ending balances. The study found that in 16 of the 23 years examined, state revenue forecasts were lower than the actual revenue brought into the states.
During the timespan of the study, there were three recessions. The Pew-Rockefeller study determined that forecasting errors got larger in each succeeding recession. They also determined that the use of consensus forecasting processes does not mean a more accurate forecast. It does make the forecast less susceptible to political forces.
So how do states improve their forecasting process? The Pew study suggests that states should take time to analyze those situations where the estimate was off. States like Indiana and Michigan undertook reviews of their forecasting process in the aftermath of missing projections during the Great Recession. Another recommendation, which Iowa already does, is the frequency of projections and the need to update these forecasts during the legislative session. Iowa’s process is set up to protect against political influence, which also is a recommendation from the Pew study. Finally, they suggest looking for ways to expand the expertise used in making forecasts.
Today, Iowa is amongst many states looking for ways to help improve the accuracy of their revenue forecasts. While creating a process that is 100 percent accurate every year would be an impossible quest, exploring new ideas that help the process is a task that will be a priority in the remainder of the 2017 legislative session.