I'm estimating the demand of rail passengers in long haul in Italy with longitudinal data of 25 years. I'm using an Error Correction Model since the relationship between the passengers*km and real GDP, real average fare and train*km is cointegrated. I''m using the two step methodology and I found that income elasticity is inferior in long run than in short run. Moreover in short run is not significative. The same if I use overnight stays, but in short run they are significative. It's the first time I use this type of model and I'm wonderng on the plausibility of the findings. Moreover the aim is to forecast passenger demand, since there is a cointegrating relationship, may I use the step one regression (in levels) or do I necessarly have to use the ecm regression?
Thanks.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis