Econometric Models

Econometric models statistically "explain" the value of one or more variables using values of a number of other "explanatory variables". For instance, a family's electric water heating energy use might be explained by the number of adults and children in the family, household income, presence of a dishwasher and washing machine and other factors. A simple representation or "model" of this relationship can be represented as:

El_Wh = b0 + b1*Adults + b2 * Children +b3 * Income + b4 * dishwasher + ...

where values for the b0, b1, b2, ..., parameters or "coefficients" are provided as part of the statistical estimation process. In this example, an estimate of electric water heating use is generated when the appropriate values of the explanatory variables on the right-hand side of the equation are applied for an individual family or for a market segment.

In actual applications, the empirical relationships are more complicated and a variety of additional modeling and statistical issues must be addressed. JA has developed, estimated and applied econometric models to analyze and forecast a variety of energy-related issues including:

Hourly, monthly, quarterly and annual electricity use (kWh)
Hourly, monthly, quarterly and annual electricity demand (kW)
Monthly and annual natural gas and oil energy use
Utility customer fuel choices for space heating, water heating, electric cloths drying
New energy-technology choices
Energy-supplier switching behavior
Customer profitability