Corporate Earnings Data Could Improve State Budgeting

The US government has faced a lot of criticism over its profligate ways in recent years. Under both parties, it has used its ability to borrow and print money to spend far more than it takes in from taxes. Many states, by contrast, require policymakers to balance their budgets annually and restrict borrowing to fund deficits. Such mandates mean that if fiscal planners are far off in their forecasts of how much tax revenue a state will bring in, it can have dire consequences for the public programs they fund.

But there may be a more accurate way to devise such forecasts, suggests research by Chicago Booth’s Anthony Welsch and University of Texas’s Braden Williams and Lillian Mills. Their work indicates that this can be done by adding a measure of corporate earnings growth to states’ tax revenue models. Additionally, the researchers compared four methods of measuring aggregate corporate earnings growth and conclude that a state-specific, industry-weighted measure produces the most accurate predictions. The findings should prove useful to fiscal planners in forecasting state tax revenue and setting budgets, the researchers write.

Fiscal planners face a complex challenge in accurately forecasting state tax revenues. When they overestimate how much tax revenue a state will bring in, the unexpected shortfall can force officials to make midyear spending cuts. When they underestimate tax revenues, policymakers can lose the chance to fund priorities such as public safety and education. That’s exactly what happened during the early days of the COVID-19 pandemic, when some state governments braced for larger revenue shortfalls than ultimately occurred and aggressively cut public services.

State governments generated about 33 percent of their 2023 tax revenue from personal income taxes, 47 percent from sales taxes, and 10 percent from corporate income taxes, according to the Pew Charitable Trusts. Public officials have traditionally relied on historical tax collection data and proprietary economic models to generate their revenue predictions. Unfortunately, these tax revenue streams have become more volatile since 2001, resulting in an increase in the average forecasting error since then, the researchers write.

Notably, the forecasting methods now in use do not include data on corporate earnings growth, despite the fact that a variety of prior studies have shown them to be a leading indicator on a national level of GDP growth, aggregate employment outcomes, and aggregate investment. Earnings growth should also be correlated with state tax bases, assert Welsch, Williams, and Mills.

To determine whether such information does, in fact, improve forecasting accuracy, the researchers compared several ways of incorporating it. They built a national earnings measure similar to the ones used in prior studies, as well as three separate measures that each assign earnings growth to individual states. These latter measures use the location of companies’ headquarters and operations and the industries involved.

Then they incorporated the various measures into existing state forecasting methods to create their own models, whose predictions they then compared with actual predictions and tax revenue growth from 1999 and 2018.

A state’s forecasting model became more accurate when it was supplemented by three of the four measures of earnings growth, according to the researchers. Industry-weighted earnings growth provided the largest improvement, of 12.5 percentage points, versus the baseline model, and nearbly doubled the explanatory power, according to the study. By controlling for aggregate tax revenue growth across all states, they determined that slightly over two-fifths of the models’ incremental predictive power related to state-specific tax revenue growth and the remaining portion to national-level tax revenue growth.

Together, this evidence indicates that earnings contain relevant information about tax revenues in specific states beyond that which is currently produced using inputs such as GDP forecasts and estimated tax payments. What’s more, the researchers find that earnings growth data improved forecasts of all three major forms of state taxes, with personal income tax predictions improving the most. One possible explanation is that when corporate earnings grow, employers hire more employees, increase pay, or return capital to investors, the researchers write.

States that include earnings growth in state tax revenue forecasts could reduce errors and increase the efficiency with which they allocate resources, the study suggests. City and other local governments could also benefit from using the new earnings growth measure, Welsch, Williams, and Mills say.