Undercover Economist: Check this out
By Tim Harford, FT.com site
Published: Dec 15, 2006
Few places are less festive than a supermarket checkout in December, with over-worked staff scanning Christmas puddings and bags of mixed nuts as fast as they can. Not all of us find it easy to spare a smile for these hard-working people at Christmas. But there is now one more reason to give checkout staff our sympathy: not only do they labour for low rewards, but they are also being spied upon by economists.
The researchers in question, Alexandre Mas and Enrico Moretti, decided that checkout staff would be ideal guinea pigs in an experiment to answer a vexed question: what happens when an unusually hard-working (or lazy) worker joins a team?
The question is part of the broader study of "peer effects". When my neighbour, classmate or housemate is particularly smart, dishonest or lazy, what does that do to me? The question is tricky because most people can select their peers. For example, observing that all the kids in a school play truant, we might conclude that they are a bad influence on each other, but we might also conclude that the school is in a deprived area where richer parents choose not to live.
Some economists have looked at situations where peers have been assigned randomly - to a college dormitory, for instance, or even (through a government housing programme) to a particular neighbourhood.
Mas and Moretti rely instead on scarily detailed data: having somehow sweet-talked a supermarket, their data-set tracks every single "beep", every transaction, for 370 workers in six stores, timed by the second, for two years. They can measure each worker's productivity by the second, and note how it changes depending on who else is working at the same time.
It is not obvious what they should find. Since shoppers can and do move to fast-moving queues, a quick worker will tend to lighten the burden on their colleagues. That might encourage them to slack off, or it might encourage them to work harder. The positive effect dominates, according to Mas and Moretti: they find that a shop assistant sitting near someone who is 10 per cent quicker than average will raise their own game by 1.7 per cent.
This might be an illusory effect. Perhaps at busy times, all workers increase their speed and managers also throw on the fastest workers. What looks like a peer effect would be the co-ordination of two different responses to a rush of shoppers. But Mas and Moretti can tell which times are busy and which times are not; they also know that checkout staff, not managers, choose their hours (one of the few benefits of the job); and they are measuring productivity changes every 10 minutes, not over the course of an entire shift. They are convinced that the positive peer effect is real.
But why? There are, broadly, two explanations. One is that workers are spurred to greater efforts when contemplating the superior speed of their colleague. This is psychologically plausible but economically irrational. A more cynical explanation is that workers do not like it when faster colleagues are looking at them, because they fear being accused of slacking off.
It might seem impossible to distinguish between these two possibilities, but at the checkout each worker is looking towards one colleague with his back to a second colleague who is looking at him.
It turns out that facing towards a fast worker makes no difference, but having a fast worker face towards you encourages frenetic scanner-wielding. So the cynics have it. And next time you hear the beep of a supermarket scanner, remember that Big Brother is watching you - and he's an economist.
Motivation article