An economic forecast is an estimate of the future performance of a country’s economy. In the past, before the work of Jan Tinbergen in about 1954 that produced the first macroeconometric model to be used for this purpose, predictions were often made only for individual prices or a set of other individual economic, generally microeconomic, variables. The choice of methodologies for the preparation of economic forecasts varies considerably. Some of the most widely employed methods are statistical, in the sense that they depend on the estimated behavioral patterns of time-series economic variables and their persistence into the future. This is a fundamentally different approach to forecasting from that of the structural macroeconometric model, which seeks to represent the relationships between economic and other relevant variables in terms of modern economic theory.

It is also possible to employ hybrid methodologies. It may be useful, for example, to incorporate aspects of the statistical models in order to adjust them to the particular properties of some economic variables, such as their volatility and stickiness.

Frequently, economic forecasts contain simultaneous estimates of values for a number of different economically relevant variables. Typical of these are estimates of gross domestic product (GDP), one or more interest rates and various measures of unemployment, employment and price inflation. For such combinations of independent variables, it is important to note that the measurement of relative accuracy rankings for competing forecasts can differ greatly, depending on which variables are taken into account.