The following factors (variables) are being considered by you in developing an econometric model in order to analyze annual peak summer electricity demand (ESUM , in MW) for New South Wales:
Y = real average income ($/person)
P = real electricity prices (¢/KWh)
R = real interest rates (% per year)
M = proportion of immigrants in NSW population (%) - a proxy for cultural diversity
A = proportion of apartments in total dwellings (%) - a proxy for housing pattern
D = percentage of income for the highest quantile (%) - a proxy for income distribution
Tmax= maximum temperature.
Data for these variables for the period 1990-2012 was provided earlier as part of Assignment # 1.
TASKS
a) What factors (as given above) and mathematical functional form (i.e., linear, non-linear, log-log, log-linear) will you consider in developing a first-cut econometric model? (present your model in an equation form as shown below). Provide reasoning for your selections (no more than two dot points with each dot point no more than three lines).
ESUMt = α + αt.Xit + ut
where Xi= factors that you consider in the model
u = error-terms
t = subscripts representing year
α0,αi = coefficients (parameters) to be estimated.
b) Estimate the parameters of the model in a), using OLS method (present your result in a tabular form, together with the values of t-stat, Durbin-Watson d-stat, and adjusted-R2).
Comment on the signs of the estimated partial slope coefficients, i.e., do they concur with your prior expectations (no more than three lines).
On the basis of t-stat, Durbin-Watson (or Durbin h) test and adjusted-R2, comment on the statistical acceptability of your estimates for the above model (no more than five lines).
c) Investigate whether multicollinearity, heteroscedasticity, and autocorrelation are serious issues in the model. Provide justification for your answer (no more than five lines - also provide a summary table of your test-statistics).
d) Does the model that you have considered in a) require further correction with respect to the issues of multicollinearity, heteroscedasticity, autocorrelation and/or specification errors (i.e., in terms of factors selected and functional form chosen)? (Yes or No). Provide reasoning for the corrections (no more than two dot points with each dot point no more than three lines)
If the correction is needed, take the appropriate corrective action and present the revised model (present your revised model in a tabular form, together with appropriate statistical values).
Comment on the statistical acceptability of the revised model, on the basis of t-stat, Durbin- Watson (or Durbin h) test, adjusted-R2
e) Provide a philosophical critique of your first-cut model, as developed in a)? (no more than two dot points with each dot point no more than three lines) How does your revised model, as developed in d), overcome that critique?
Its an assignment in field of econometric and analysis. I need to find the best econometric model based on the question requirement and answer the following questions.
Attachment:- 1.xlsx