I sometimes commute on the Jubilee and Bakerloo Lines of the London Underground.

If I have to stand on the former, especially at this time of year, "straphanging" is a nuisance. The poles and grab rails are COLD! They are of uncoated metal with a high thermal conductivity!

By contrast, the poles and rails on the Bakerloo Line trains are coated and are much warmer to the touch.

Does it matter?

Does customer comfort matter? Does it matter if people are tempted not to hold tight because it's actually uncomfortable to do so?

Are there other examples of where a little thought would have made a difference?

While addressing a session of the Transportation Research Board's annual meeting in Washington DC last month, Federal Transit Administration Administrator Peter Rogoff was asked about Federal agencies looking at using international contractors. He said, "We are looking for a complete and unquestionable compliance with Buy America for recovery dollars".

While I am personally unsure of the precise implications of that, it doesn't sound good on a number of levels.

It restricts the flow of ideas to those which have been already taken on board by American industry - which is sad, because there are other good ideas out there too. So Mr. Rogoff may well get less for his dollars than he could.

It may mean paying more.

The Administrator was keen to suggest looking at alternative revenue sources: he could equally look for lower costs, which produces the same effect on the bottom line.

I remember reading about a case where a State deliberately did not apply for Federal funding for a project because they would then have had to comply with Buy America rules: the savings they made by buying abroad were large enough to mean that they didn't need the Federal funds!

If I wanted to buy a large commercial aircraft, I'd certainly look at the US company Boeing. They specialise in large commercial aircraft: one can assume that they are pretty efficient at making them and the price:value mix is pretty good. If on the other hand I wanted to buy a signalling system for a high speed railway, I wouldn't look to US companies. They do not specialise in this area and I'd assume that they had a large learning curve to climb - at my expense. So why not "Buy best"?

At a higher level, Tom Friedman, the New York Times columnist and author, made two wise observations in this context (and I paraphrase fairly widely).

First, he described a call centre which had been off-shored to India. But because the staff there used IT and communications equipment designed and engineered in Silicon Valley, it was valuable - the guys in India were doing what they were good at, as were the guys in California. To put a software designer into a call centre would be dumb.

Second, he commented on the effects of the inability of Pakistan to sell cotton t-shirts to the US market - only the raw cotton, because of Buy America considerations. It meant that they got less for their crop than they would have got for finished goods - so they were less affluent. How much impact does the resultant poverty in Pakistan have on Afghanistan, drugs and the safety of America?

Buy America? Bye Bye America?