As Director General of the International Air Rail Organisation, I am sometimes asked how big an airport has to be before it justifies a rail connection. I've always believed that this is the wrong question - the right question is more along the lines of, given the geography, catchment area, business and situation of an airport, what kind of rail connection does it justify?

My classic example comes from Portland, in Oregon. Portland International Airport is served by the Tri-Met MAX light rail line. Way out on the other end of the line is Hillsboro airport - a business and general aviation airport. This too is served by Tri-Met, by the station called State Fair/Hillsboro Airport which primarily serves the Oregon State Fairground.

The extra costs of calling it that and not just State Fair were, I have always suspected, totally trivial. In the past I've always assumed that the revenue is only just less trivial - the odd airport employee might use it, perhaps someone visiting one of the offices at the airport might take the light rail.

But at the end of August was the Oregon International Air Fair, held at the airport. And spectators were actively encouraged to use the light rail to get there. So use of that station by airport users is even less trivial than I'd first thought!

This is an occasion where an airport with no scheduled air service and a very low passenger usage can justify a rail connection of some kind. The railway is close - a short walk - and calling it "State Fair/Hillsboro Airport" seems to be paying off. Excellent!