Economic forecast is a discipline that seeks to understand the behavior of the economy and predict future trends. It involves a wide range of techniques, including statistical models, consensus forecasts, econometric methods, time series analysis and input-output modeling. Forecasting is a complex endeavor, and experts regularly get predictions wrong. The challenges and subjective human behavioral aspects of the practice have led some rational people to view government economic forecasts with healthy doses of skepticism. In addition, private-sector economists have a reputation for getting their predictions wrong as well.
To produce a forecast, an expert gathers data from many sources that provide a summary-level perspective on economic trends, issues and risks. This information is used to form assumptions about the behavior of various economic variables, which are then used as inputs into a model. The resulting forecasts are typically compiled into reports that include useful charts and commentary to help the reader understand them. The specific methodology used to create a forecast varies between the structural macroeconometric methods that are rooted in economic theory and the essentially statistical methodologies that are based on forming and estimating regression equations. Regardless of the exact method used to form a forecast, it is important for an analyst to consider non-linearities in the relationship between one series and another. These properties often reveal additional information about the underlying dynamics of the economy and may improve forecasts over frameworks that assume linear relationships. Non-linearities may also be more pronounced at times of change, such as at the start or end of a recession.