The Implications of Uncertain Economic Paths for Revenue Projections

Abstract

This paper measures the effects of macroeconomic uncertainty on the accuracy of baseline federal revenue forecasts. We build a simple stochastic model of gross domestic product, stock prices, and employment over time. Using a cloud-based microsimulation model, we simulate individual income and payroll tax revenues for 5,000 realizations of the macroeconomic variables. We find a large amount of uncertainty in revenue projections that grows over time. We find a small downward bias from using a single baseline to predict revenues, but it never exceeds 1 percent of revenues over the 10-year budget horizon.