The ‘Gift’ of Consensus Revenue Forecasting
Indiana’s fiscal policymakers got an early holiday present last week – a state revenue forecast bolstered by predictions of a more resilient economy and legislative action to expand our sales tax base. The optimistic analysis is good news on its own, but should also be a reminder to give thanks for the integrity of the revenue forecasting process itself.
The December forecast projects 2020-21 general fund revenues will come about $530 million higher than April’s results. Revenues have been exceeding monthly estimates since the start of the new fiscal year in July, and there’s growing confidence in these trends.
Tax changes passed by the General Assembly this year (requiring online market facilitators to collect state sales taxes) accounted for most of the growth and were already factored into the 2020-21 budget, leaving roughly $260 million in ‘new’ revenue compared to the two-year budget plan.
There’s still plenty of uncertainty about the future of our economy. Both the April and December forecasts put the odds of a recession starting in 2020 at roughly one-in-three. But a more positive outlook for the housing market and light truck sales leads to upgraded growth (within a single percentage point) in income and sales tax collections above and beyond the revised online tax rules.
These modest changes are products of a data-driven forecasting process informed by expert analysis. This annual ritual has played an understated but essential role in a string of balanced budgets, a record state surplus and AAA ratings from the major credit agencies.
Every two years, the Governor and General Assembly debate spending priorities and pass a budget. Every legislative session is a balancing act between the costs of new proposals and how the burden is shared by Hoosiers. But it all has to start with basic assumptions about available resources.
Revenue forecasts set the ‘ground rules’ for budget negotiations, predicting the levels of state tax collections available to support K-12 schools, fund Medicaid and other public health programs, aid higher education and advance workforce and economic development – all of Indiana’s government agencies and operations.
Indiana uses a ‘consensus forecast’ model: The legislative and executive branches collaborate on the forecast (minimizing political bias), starting with an annual economic outlook analysis produced by an outside consultant (IHS Markit). The outlook goes to the Revenue Forecast Technical Committee (RFTC), composed of the fiscal analysts from each legislative caucus, a member of the State Budget Agency and an economist from a state university (appointed by the Governor).
With support from State Budget Agency and Legislative Services Agency staff, the RFTC applies statistical models based on the state revenue system to the economic outlook to produce the forecast, revealed publicly at the December meeting of the State Budget Committee each December. (The forecast is also updated in April during budget sessions of the General Assembly, like this year.)
This is second-nature to Statehouse veterans, but the consensus model is only used in about half the states. According to a recent review by the National Association of State Budget Officers (NASBO), the other half use a forecast produced only by the executive branch or ‘dueling forecasts’ – one from the executive, one from the legislature. The potential for political gamesmanship is obvious either way.
Indeed, a 2017 briefing produced by the Indiana Fiscal Policy Institute (“Revenue Forecasting: Indiana’s Method and the Results It Produces”) notes that Indiana is one of just fifteen states using all the best practices identified by the Center on Budget & Policy Priorities: Consensus forecasting using outside experts, with public transparency (a published methodology and open meetings) and regular revisions.
The IFPI briefing compared forecasts to actual revenues from 2004 to 2016 and found revenues underestimated six times and overestimated seven, with the Great Recession driving the most dramatic divergence (the forecast was off more than 7.5% in 2010). Removing 2009-2011 from the data, the average forecast was within 2% of the fiscal year close-out.
A consensus model improves accuracy, but the business cycle can’t be perfectly predicted. For Indiana, there’s a higher degree of complexity: Sales taxes are our dominant revenue source, followed by individual income taxes…a more volatile mix than states that collect more property taxes, comparing the ups-and-downs of personal income and consumer confidence with the relative stability of real estate.
To raise the stakes even higher, revenue forecasts play a more critical (legal) role in budget-writing starting this year. Under the state balanced budget amendment approved by voters in 2018, a passed budget must be balanced according to the revised revenue forecast, with unanticipated shortfalls being subtracted from the next forecast and putting more pressure on the subsequent budget cycle.
Last week’s forecast is positive news that will still sharpen policy debates: How cautious do we need to be with our ‘rainy day’ reserves? If the state surplus continues to grow, how much of it should we spend and on what – capital projects or operating priorities like teacher pay? Should extra budgetary breathing room cause us to revisit K-12 and other spending before 2021?
These issues will be hashed out in the Statehouse and on the campaign trail in 2020 and beyond. The budget process, and all questions of taxing and spending, are inherently political…but our revenue projections should be above politics and professionally conducted. In Indiana, we’ve created such a system to support fiscal stability and serve the taxpayers – a fact that deserves recognition.