Why good people sometimes do bad things: 52 reflections on ethics at work
42. Perverse effects of transparency:
moral licensing and the magnetic middle
Transparency has become a magic word. As long as things are transparent, all is well, because
transparency puriies and corrects. For this reason organizations often publicize their inancial
and social achievements, the composition and origin of their products, and their top salaries.
Managers and employees should be open about their additional jobs, the gifts they receive,
and the shares they own. Transparency, however, is not unequivocally beneicial: it has a dark
side. This chapter discusses two ways in which transparency can perversely affect moral
behavior, which we should guard against.
Research by Daylian Cain and colleagues sheds light on one negative effect of transparency.
Their research was directed at experts and the information they provide about possible conlicts
of interest in their advice to customers. For example, a mortgage adviser might receive a
higher commission from one mortgage provider than another, or a doctor might expect better
remuneration from one pharmaceutical company than another for the medicine he prescribes
to his patient. People publish such conlicts of interest with the aim of preventing bias and
prejudice.The expert will be more cautious about providing biased advice if he realizes that the
customer is aware of his interest. To ind out if this was really the case, Cain and colleagues
came up with the following experiment.
The study centered on a glass bowl containing a number of small coins. The participants were
assigned the role of an adviser or valuer. The adviser was required to research how many
pennies were in the bowl and write a report. He was permitted to do whatever he wanted,
apart from taking the coins out of the bowl and counting them one by one. The valuer, seated
in another room, then received the report, and was also required to guess how many pennies
were in the bowl, with the difference that the bowl was placed at a greater distance and the
valuer had only ten seconds to make the estimation. The report by the adviser was therefore
a welcome support.
The experiment played out over six rounds. The researchers varied the adviser’s reward and
the extent to which this was made known to the valuer. The average value of the bowl over
42. Perverse effects of transparency: moral licensing and the magnetic middle