Extra work
1. Calculate aggregate real consumption and aggregate price of several (4-6) commodities, based on formulas for aggregate indices (for example consumption and price of fruit, the fruit being an aggregate of apples, oranges etc. ).
2. Create a regression model of aggregate consumption, as well as regression models of consumption of individual commodities included in the aggregate (e.g. a model of consumption of apples, oranges etc.).
3. Calculate predictions (for the sample period: 1980-2001) of aggregate consumption based on (a) the aggregate model, and (b) models for individual commodities.
4. Compare the results of alternative predictions. Which approach generates more accurate predictions – the one based on aggregate or detailed consumption models?
Remarks
• Each (volunteer) student should choose a different set of commodities. Please e-mail at jborat@uni.lodz.pl to verify the choice.
• Predictions are available in G7 right after running the “r” command – they are represented by the variable “predict” (try the following commands: “gr predict” and “ty predict” to graph or type prediction results).
• The results (of estimation and prediction) should be printed. The student should also supply his/her G7 command file and verbally describe the results during the office hours.
• If you decide to prepare this extra work you need to send me an e-mail with such a decision and a proposal of the choice of commodities before June 4 (at jborat@uni.lodz.pl).