Having an idea of where the economy is headed can help everyone from policymakers to businesses. But predicting the future isn’t easy, and even expert economists get it wrong sometimes. Fortunately, many techniques can be used to improve the accuracy of economic forecasts.
A key step is to gather historical data on a wide range of economic variables. This information is then used to create a model that uses a series of inputs to generate outputs such as a forecast for the economy, GDP or employment. The model is then tested using historical data and the results are reported. Often, these reports contain commentary and information graphics to explain the findings. Common models include econometric models, consensus forecasts and economic base analysis.
The forecaster then determines the relationships that exist between independent variables and the dependent variable under study. These relationships are known as causal effects. A good understanding of causal effects can be a powerful tool in creating economic forecasts, McCracken said.
Another tool is to make a nowcast, or an initial estimate of a quarterly economic statistic such as gross domestic product (GDP). Nowcasts can also be revised later, but the earlier the nowcast, the more useful it is because the forecaster can incorporate any surprises into the final estimate.
Judgment is also a critical factor in economic forecasting, McCracken said. A forecaster may decide that the current circumstances are unique enough to require a modification of the standard statistical methods that are normally used in producing the forecast. This is particularly true when the event in question reflects some kind of unusual, or even unforeseen, impact on the economy.