A widely used economic forecast is a series of predictions of some combination of the values of economic variables (e.g., gross domestic product, consumption, investment, interest rates, government expenditures, and employment) over some future period. This type of economic forecast is often presented in a way that emphasizes the differences between the underlying time series data and the predicted values. For example, the term quarterly growth is often used rather than annual growth. This convention, which can have some interesting effects if data for the current year are revised, is based on the fact that most models used to predict quarterly GDP are not as sensitive to changes in the actual value of the previous quarter’s output than are models that use the time series of averaged output over the entire calendar year.
The outlook for global economic activity continues to be clouded by rising trade barriers and heightened policy uncertainty. In 2025, growth is projected to slow in EMDEs that are highly reliant on global trade and investment, and the dimmer outlook for activity is likely to weigh on trading partners as well.
The BEA’s advance estimates of the components of GDP are derived using publicly released source data, which can be found in the Key Source Data and Assumptions table that accompanies each estimate. These estimated inputs are then combined with the results of a series of econometric models that are designed to predict the underlying dynamics of each component, in a manner that tries to account for the fact that these relationships are jointly determined and are therefore not independently determined by the time series data alone.