Dummy variable and multiple regression

Hi all,

I've recently encountered a problem. I've got alot of data from an aircarrier company. They've included dummy variables to check for monopoly-or duopololy routes.

The Norwegian airline industry was deregulated in 1994 allowing all Norwegian companies to freely enter on any Norwegian route. Up until 1994 the Norwegian routes were all monopoly routes served by one of the two companies Braathens or SAS.1 I have access to data on 11 Norwegian airline routes over a period of 13 years for the years

1985 to 1997. The routes differ in size, the smallest having 4 or 5 daily flights others more than 20. The data can be found in the Excel file ‘airline_case.xls’ (Se attachment please).

The dataset includes individual data on yearly aggregated capacity utilization on each route (CU) and various indicators/dummy variables (excel sheet ’capacity’). The indicators include deregulation timing (REG94=1 for the period 1994 to 1997, 0 otherwise), a monopoly routeindicator (MON=1 if monopoly route, 0 otherwise), a duopoly route indicator (DUO=1 if
duopoly route, 0 otherwise) and an indicator for large routes (LARGE=1 if route is large, 0 otherwise). Also indicator for route (flight_indicator), corresponding dummy variables (f1-11) and a year indicator (year) are included.

In addition I have access to a representative airline-ticket price index (API) aggregated over all 11 routes as well as a consumer price index (CPI) and a gasoline price index (GPI) for these 13 years (excel sheet ‘prices’).

My boss has asked me to use the dataset and make a summary statistics (averages, graphs, trends etc) and to discuss how the
deregulation affected prices and capacity utilization in the airline industry. I do not really understand how to make use the datasheet and which varuiables and dummyvariables to use. Please help me find a solution to this. Would mean alot!

Yours sincerely

bump bump bump

Hi,
I went over your data and it seems like the aggregations are biasing the data.
There are two things I found that might be interesting:
1) The relation between API and CPI in the prices table (see the regression in the attached file). The consumer price keeps going up with the API but slower starting from the year of regulation 94.
2) The flight capacity for groups 1,3,5,6,10,11 has great variation.

As a general approach try to use relative values and not absolute i.e. try to check price changes in percent over the years and eliminate the effect of gas price. Also try to get some sort of economic or world flight indicator that will help you to eliminate world wide economic effect.

Try to get better data with less grouping inorder to have better resolution and more indicators.

The files included are two images with the regression result and the capacity variation, an Analysis Studio 6.3 stp and data files with the project.

Hope this helped.