Thanks to Michael Noetel, Luke Freeman, Duncan Mortimer, David Reinstein and Peter Slattery for input on this piece. 


- We should test opt-out donations on employment contracts

- We should test what default percentages lead to the most contributions
- Churn and lifespan contributions should also be considered

Changing default preferences is a low-cost and high-impact intervention in behaviour design that is commonly overlooked. Below are several interventions we would like to see trialled and tested to capture as much charitable giving as possible. This article will focus on initiatives to capture giving from regular income, such as the ‘Giving what we can’ initiative.

1. Testing of opt-out donations at the organisation level.

Several companies and organisations have taken various pledges for revenue for charitable donations. These companies have aligned values to effective giving and would probably like to continue to use their status to further the culture of giving within their organisation. 

Providing the option for organisation employees to donate a percentage of their income to charitable causes appears to be a pathway to increase giving through the sacrifice of salary by those employed by the organisation.

Better yet, if organisations could be persuaded to include this at the employment contract level and be the default terms of employment, with opt-out optionality. (Legal input to this idea is required)

I would like to contribute 1% of my income to effective causes.

◻ I would not want to contribute my income to effective causes.

This could set social norms within organisations and leverage defaults to reduce friction to giving from regular income. It could demonstrate commitment to charitable giving at the organisational and cultural level without loss of profits and would likely be seen as a marketing win-win.

There may also be individuals that would like to give in a more automated way through their income and an initiative that could streamline pre-tax donations for their organisations and may champion such an initiative in their workplaces. This could, however, also be perceived as heavy-handed.

There is grey literature that suggests this model could be effective, but Preference-heterogeneity is a big deal for a charitable donation, and there is evidence and theory to suggest that 'warm glow' donors will respond quite differently to information re effectiveness than pure or impure altruists. Of note, 'warm glow' givers may react negatively to information around effectiveness. 

Furthermore, endorsement effects may not be in the expected direction if the employer endorses behaviour that is inconsistent with employees' preferences. This suggests that testing in a single employer is unlikely to tell the whole story. Selection effects make the employees of one industry/employer quite different (e.g. charities, universities, government) from the employees of another (e.g. defence, legal, real estate). Testing in an industry/employer with a preponderance of 'warm glow' donors can be expected to give different results than testing in an industry/employer with a preponderance of 'pure or impure altruists'.

2. Testing of default suggested percentages of income.


For the above idea and organisations like the ‘Giving what we can’ pledge, a suggested percentage of income seems the most commonly recommended regular giving calculation. However, there is limited evidence and testing around what suggested amount or default would yield the most effective contributions.

Some common suggestions are ‘10%’ or ‘1%’, but these amounts may seem too low or too high to commit to for some individuals.

However, if 11x more people commit to a 1% pledge, it would be more effective than asking for or setting the initial contribution norm to 10%. 

We should aim to find the ‘sweet spot’ (possibly for specific subsets of personas) that leads to the most effective contribution commitments. 

There is no clear evidence with a definitive result on the tradeoff between ‘ask for less → more participation’ and ‘ask for more → some people who would participate in either condition donate more. There is consistent evidence of such a tradeoff, but ‘which one wins’ (‘sweet spot’) is likely contextual. 

Some evidence shows default options have a substantial effect on decisions to donate but may not vary the final amount due to this trade-off. However, matched amounts or co-contributions may increase default amounts.

We would hypothesise that asking for a low initial percentage contribution at the early career stages and intermittently asking for incremental increases would yield the highest lifespan contributions with the lowest churn rates, compared to ‘optimistic altruism’ of setting social norms to high levels. 

Previous A-B trials with the EA market testing group presented the GWWC pledge options showed slightly higher uptake with 10% pledges when presented centre and next to a 1% ambiguous but a higher option. However, this is likely attributed to the possibilities offered and some inherent social signalling of the options layout, rather than testing pure default option uptake. 


Perhaps giving people more options makes them indecisive. They may be particularly reluctant to choose a “relatively ambitious giving pledge” if a less ambitious option is highlighted. This could also involve issues of self and social- signalling; if the 'main thing' to do is 10% (as in "separate bullets", then this may seem a straightforward way of conveying 'I am generous. On the other hand, if the 'Further pledge' is fairly prominent, perhaps the signal feels less positive. And if the '1% pledge' is made central, 10% might seem more than a necessary signal.”


Several other key papers highlight ‘paltry’ contributions or asking for less, which increases the incidence of giving. This finding is repeatable but may decrease the amount given and is sensitive to influence by anchors and expectations.


3. We should also test churn and increased contributions on these defaults.

We should also capture data on churn rates at different contributions and those who increase their pledges to calculate potential lifespan contribution sustainability. It is possible that the initial 10% signups decrease after one year or stop, whereas those who commit 1% for 11 years are more impactful.


Where to from here?

There is potential to approach organisations that have charitable giving in place and offer to implement such a strategy to employees, test defaults and measure contribution amounts and churn. This could improve the lifespan donations of individuals, onboard more people to learn about effectively given principles and improve the data around scaling up regular charitable giving.

There would be interest in structuring a grant proposal around this, and if anyone with appropriate skills would like to contribute, please let me know.



New comment