Woman covering her eyes
(Image credit: Diana Grytsku via Freepik)

Ignorance and moral wiggle room

Sometimes we prefer not to know something, for example to avoid worry or to escape accountability. But ignorance may also serve our moral identity and reputation

ore than 2/3 of all adults in the UK made charitable donations in the UK in 2017 — from regular contributions to specific charities to the occasional small change into a collection box or into the hat of a homeless person on a street corner. If ever there was a violation of standard economic assumptions, then here we have it: generosity would appear to go squarely against our self-interest. We do indeed manifestly end up materially poorer as a result of such donations.

A simple explanation would be that we are not only concerned about our own wellbeing, but also about that of others — including of random strangers. But donating is not always anonymous. We can choose whether or not our name will be visible when making online donations. And while those made in public view may not reveal who we are, we still signal generosity to anyone observing us. That suggests a potential ulterior motive: being seen to be generous improves our social image, and that may yield other rewards over time.

Busker playing a harp
Busker playing a harp
Streets of London on a harp — that must be worth £2, right? (photo: Diana Jeater CC BY)

Still, that doesn’t explain generosity when nobody is looking (or when those looking don’t know us). So, another reason for this apparently ‘irrational’ behaviour is that being generous reinforces our self-image. But we cannot fool ourselves so easily. An outsider who spots us popping something in the hat of a busker singing ‘Streets of London’ will not know how much we are donating — anything (even just an old button) will make us look generous. But for us, the boost to our self-image would be proportional to our generosity. We don’t want to be too generous towards the would-be Ralph McTell and give him £100 to give us that boost, but we know very well that a button or tuppence is not generous.

Ignorant about our generosity

What if we genuinely didn’t know how generous we really are? That is what a study by Danae Arroyos-Calvera, Rebecca McDonald, Daniel Read and Bruce Rigal at Birmingham and Warwick universities set out to explore.

Studies of generosity often use variants on the Dictator Game, an experimental tool in which the player (the ‘dictator’) decides how an amount of money is divided between themselves and another person (the ‘recipient’). For example, the dictator may have a choice between $6 for them and $1 for the recipient (the ‘selfish’ option), and $5 each for both of them (the ‘generous’ option) (Fig. A).

The researchers’ point of departure was an earlier experiment using a variant of this instrument, in which a dictator could choose to take either $6 or $5 for themselves, but without knowing whether the recipient would get $1 when they chose $6 for themselves and $5 if they chose $5, or vice versa — both scenarios were equally likely (Fig. B). However, before they made their choice, the dictators could opt to reveal which scenario applied to them. 38% of them decided to make their choice without knowing the scenario (and hence to risk only giving the recipient $1 with a probability of 50%). The vast majority of them (86%) chose the maximum $6 payoff for themselves. By remaining ignorant, they gave themselves moral wiggle room: even though they chose the maximum payoff for themselves, they could justify doing so, as there was a 50% chance this also maximized the benefit to the recipient.

As the authors suggest, this is a bit like grabbing a coin from your purse containing both 50p and £2 coins, and dropping it in the hat without looking. This, too, would give you moral wiggle room: it could have been the more generous donation, and that may be all you need for your self-image (even though it could also have been a miserly one).

But in this experiment the dictators themselves decided whether or not to find out which scenario applied (and hence how much the recipient would get). It can therefore not tell us whether they behaved generously (or selfishly) because of the information they received, or because their desire to be generous (or selfish) drove their decision to obtain (or not) the information.

The researchers thus set up an experiment* in which they could separate two potential influences: the dictators’ information preference, and the actual information provided to them (and/or to the recipients).

They used a lottery variant of the dictator game, in which the payoff to the dictator included a random element: it could be £2.50 or £4. The dictator had to decide whether to give the recipient 50p or £2, and their decision influenced their own payoff: if they handed the recipient 50p, their chance of collecting £4 themselves was higher than if they donated £2. This made being generous costly to them.

The possibility of moral wiggle room was introduced by establishing two scenarios, which differed in how generous or selfish the dictator would be. In the low endowment scenario, their chance of collecting the maximum payoff was relatively low (1/8 if they were generous, and 3/8 if they were selfish); in the high endowment scenario it was higher at respectively 3/8 and 5/8. The difference to the dictator between giving £2 and giving 50p was relatively much larger in the low endowment scenario than in the high endowment scenario. In the former, the chance of collecting £4 was three times higher (3/8) if they only gave 50p than if they gave £2 (1/8). In the high endowment scenario, the relative increase was just 5/3 higher, less than a factor two.

Giving 50p or 2 pounds, how generous is that really?

If the dictator was unaware of the scenario they were in, there was private wiggle room (the dictator could justify a selfish choice to themselves); if the recipient was unaware, there was public wiggle room (the dictator could justify a selfish action to the recipient). This allowed for four information states (both aware, both unaware, only dictator aware, only recipient aware).

To separate the preference of the dictators for remaining ignorant (or keeping the recipient ignorant) from their preference to be generous or selfish, the researchers created three experimental treatments:

  • Choice: dictators could decide whether to reveal the scenario to themselves, and/or to the recipient. This was then implemented
  • Forced: dictators had no choice, and were allocated to a given information state at random
  • Chance: dictators could state their information preferences as in the Choice treatment, but there was only a 50% chance this would be implemented (referred to as the Chance-Choice treatment); the alternative was the Chance-Forced treatment, in which an information state would be imposed.

The Forced treatment allowed the researchers to look at just the role of the information provided on the dictators’ choices, irrespective of their preferences. Conversely, in the Chance-Forced treatment they could look at the role of the dictators’ preferences, as the information was provided randomly. Finally, the Choice and Chance-Choice treatments shed light on the effect of private and public moral wiggle room.

Strategists rather than opportunists

Self-interested behaviour (i.e. giving 50p rather than £2) was, as expected, more prevalent when dictators had chosen to keep the information concealed. But this research also showed that the information itself had no effect on generosity.

It was the dictators’ information preference — regardless whether that preference was implemented or overridden — that predicted their generosity or selfishness. The authors suggest that this is because they already knew their decision, and didn’t change it according to the information state. But if they were able to indicate what information they would like disclosed or kept concealed, they chose the option that maximized the benefit (or minimized the harm) to their self-image and their social image.

This also explains the seemingly counterintuitive finding that dictators who chose to avoid knowing the endowment scenario were more likely to give £2 than those who chose to know it. Avoiding information creates wiggle room, and wiggle room is what helps justify selfish behaviour, so we’d expect a dictator to make good use of the uncertainty. However, if a dictator has already decided they want to be generous, the wiggle room allows them to imagine they were in the low endowment scenario, where £2 is a more generous gift than in the high endowment scenario.

So, it seems we are not so much opportunists who make our choice according to whether or not information about how generous or selfish we are is disclosed. Instead, we know very well how we want to act we and tend to be honest to our intentions. But if we can, we strategically seek to disclose information (or keep it concealed) to others and to ourselves, in order to enhance or maintain our image.

No matter our moral identity, we are all strategic users of ignorance and knowledge to give it a good polish.

*: The detailed description is a little too complicated to summarize in this post. If you are interested, do read the actual paper!

Originally published at http://koenfucius.wordpress.com on August 21, 2020.

Thank you for reading this article — I hope you enjoyed it. Please do share it far and wide — there are handy Twitter and Facebook buttons nearby, and you can click here to share it via LinkedIn, or simply copy and paste this link. See all my other articles featuring observations of human behaviour (I publish one every Friday) here. Thanks!

Accidental behavioural economist in search of wisdom. Uses insights from (behavioural) economics in organization development. On Twitter as @koenfucius