I am trying to use linear regression to analyze the demand model of a time series data of poultry imports. The data available are the years, quantity in (kg), amount in dollars, and per-capital income of the importing country, and I need help as to how to go about it.
could you also include a variable which reflects the amount of poultry meat that reaches the demands of the consumer (e. g. amount of sold poultry products, consumption data per capita, an assumption of the part of imported poultry meat that really is sold)? For my understanding of a "demand model" these variable should be adressed in some way....
Dear Kwamena,
If the physical weight data is available that is best. Then expenditure on poultry. In the absence of this, you can use national income (NI) of importing country. This NI preferably should be the real NI rather than the nominal. Then an important variable is price; preferably real import price. You may consider price of compliments and substitutes to poultry. These will enable you compute the price and cross price elasticities. Then consider population. In the analysis remember your unit rot test and cointegration.
To estimate a demand function, you need to have data on consumption of the good, the local price level and the international price of the good. By définitioon, overall demand at the macroeconomic level takes into account the final consumption and investment.
It should also be introduced into the model the exchange rate effect. Indeed, if the good is imported or exported ther is a link with consumption.
Moreover, in the estimation of a demand function, we need a functional form of the theoretical model. In this case I suggest you use a translog function
Thanks to you my good friends: Monika, Justice and Latif, your information is very helpful to me.
Kwamina. on top of what the other participants offer, I would suggest to search for more data on the product itself ie what size birds are they, are whole or pieces etc. Also consumer tastes and preferences should be checked. Reason for that is, a whole broiler chicken has x price. Same bird in pieces overall, a higher one(value added). Same time an egg laying old hen that is past her prime time(economically viable) has a significantly lower one. So if the taste of the locals, or the disposable income or both will allow such a choice, that will distort the outcome of your calculations or lead to wrong assumptions. Also, livers, hearts and stomachs are included? In some countries are quite desirable, in others are a nuisance. Again different prices there.
It's not quite linear after all, is it?
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing