A Tyranny of the Tardy?

There was a nice piece in the Guardian this weekend about TFL’s trial of a new scheme for moving people through Tube stations more quickly: getting them to stand on the left on the escalators, as opposed to standing only on the right and leaving the left clear for people in a hurry, of whom there are always a few.

TFL built a model that showed they ought to be able to get 31 more passengers through an escalator a minute if people could be persuaded or coerced to stand on the left as well as the right: that’s a modelled increase in throughput of 38%.

Of course, London commuters hate changes to their routine, and the trial was badly received. Throughput, however, increased almost as much as the model suggested it might.

After the trial, customers went back to doing what they did before, so TFL are experimenting with new ways to make the new behaviour stick.

But should they? Has the case for changing behaviour really been made out? Does the model reflect in sufficient detail the underlying intuitions that inform commuters’ resistance?

What neither the model nor its rationale take into account are the different needs and motivations of different groups of commuters. There’s a trade-off between getting everyone, on average, through the station a little more quickly and giving those who want to walk (or run) up the left the option of doing so.

A better question for TFL to address, and one that reflects better commuters’ concerns about the change in their beloved convention, is: to what extent is the disutility that accrues to commuters in a hurry in the new regime offset by the utility that accrues to everyone else?

To start with, it would be good to know whether hurriers are better or worse off in terms of the time it takes them to get through the station. We know from the TFL model and trial results that the average commuter will get up the escalator more quickly, hence presumably relieving upstream congestion and improving carriage to street times. But we don’t know whether hurriers do better or worse in the new regime. They will certainly transit the escalator more slowly, but they may also benefit from reduced congestion upstream: they, too, may go from carriage to street more quickly with the new arrangement. Or they may not: we just don’t know until we model out these effects (or study them empirically).

Another other factor that needs to be taken into account is the apparently higher subjective value that hurriers place on their time than (able-bodied) non-hurriers. There’s an argument for weighting time gained or lost by hurriers more heavily in the overall utility calculation.

Finally, there is also utility for non-hurriers in having the option to move faster if they want to: that is, to switch from non-hurrier to hurrier. I’m not in a hurry now, but I appreciate the option of being able to step left and walk briskly up the escalator should I get the urge. There but for the grace of an earlier passenger action at Waterloo go I.

What’s at stake here is something akin to a Millian principle of the rights of a minority (hurriers) against the tyranny of the majority (non-hurriers). It’s not clear that the minority would be in any way better off if the majority got their way, nor even that the overall change in the balance of utility would favour it.

If transport wonks are going to use models of commuter flow as the basis for an argument to change long-standing conventions, they need to be sure that their models reflect the nuances in the situation that the convention, and the intuitions of commuters, are sensitive to. Sometimes conventions are arbitrary, and are perpetuated by blind habit, but sometimes they embody a deeper knowledge about what works and what doesn’t, and what’s right and what isn’t. If we sweep them away without sensitive analysis, we do so at our peril.