1. You want to estimate a model for the demand for electricity by households in the U.S. States of the following form: Quantity of Electricity Consumed per Household in Statei = f (Price of Electricity in Statei , Price of Natural Gas in Statei , and Income per Household in Statei ).
Add each series to an EViews workfile and change the names to QELEC, PELEC, PGAS, and INCOME:
a. Carefully identify each series.
b. Tell why you think each data series is, or is not, exactly what you should use in your estimation.
c. Paste the variable names for each series and only the first and the last observations here:
2. Type the equation you will estimate and explain it including your theory and the expected signs for each estimated coefficient.
3. Re-estimate the equation to see if there is a quadratic effect for income and paste the output in your assignment.
4. Interpret the coefficients for income and calculate where the parabola turns.
5. Describe the procedure you should use to determine how high a power of income should be included in the equation.
6. Re-estimate the equation of #3 in log-log form and paste the results in your assignment.
7. Evaluate your results including your estimate of the own price elasticity of demand
8. Which is better, the linear model or the log-log model? Calculate the predicted values of consumption for each of the states and use these values to construct an approximate R-square to compare with the one from the linear model.
9. Re-estimate the linear model using the ratio of electric to natural gas prices and paste the results in your assignment.
10. Evaluate this model in comparison with the original one.
11. Is the consumption behavior different (i.e., do prices, incomes and housing units have different effects) in sates with incomes higher than the average? Estimate an equation that will let you determine this and paste the results in your assignment.
12. Are there a differences? (Interpret all of the coefficients)