This matter is discussed in the frame of p. 11 of the enclosed paper AJD 150F available on ResearchGate. Essentially, combinations of ports and hinterland modes are defined and a choice model estimated. The frame points to the correct and the incorrect ways to estimate the resulting share model in order to obtain coefficients that are truly independent of the choice of a reference mode as one applies a particular, well-known estimator..

Dear Prof Gaudry,

Thank you very much for your comment and for the paper.

I will read it carefully.

I remain at your disposal.

Best regards