1 Introduction: Effective giving, responses to analytical ‘effectiveness information’

This project aims to robustly measure and synthesize evidence about how charitable giving responds to the presentation of analytical ‘impact per dollar’ information.

Why do we care about this? From our academic-style abstract:

While hundreds of billions of dollars are donated to charity each year, the effectiveness of these charities differs by orders of magnitude, even within similar categories. Furthermore, many individuals do not donate substantially even though they believe that the cost of saving a life is small. Our collaborative and dynamic synthesis considers this ‘puzzle’.

For people to systematically choose the most efficient charities they must be aware of the differences in effectiveness, suggesting a role for advertising and publicity. The EA movement and organizations like GiveWell and SoGive are increasingly presenting these to larger and more mainstream settings. However, we have limited evidence on how potential donors react to quantitative measures of charitable impact. We survey and meta-analyze this evidence below, also incorporating ‘our new evidence’.

Charity Navigator’s use of Impact Matters data might be considered an impact megasure, but so far, it has not been used prominently, nor in a way that allows meaningful comparisons. See discussion here

There is a puzzle: people make inefficient giving choices, and give little to highest-impact causes*

*

See “Presenting the puzzle” in Increasing effective charitable giving web-book/ That work offers a Conceptual breakdown of ‘barriers’

It has been claimed¹ that ‘innumerate empathy’ is a key barrier to promoting effective giving. Previous work² concludes that emotional reactions to suffering leads to helping behavior, but this ’doesn’t scale’ with need, and few people actually ‘research effectiveness before giving’. In a ‘dual process’ model of cognition, analytical impact information may disrupt empathic giving.

Here, we are analysing a variety of giving experiments, across various contexts.

Preliminary suggestive results: Analytical info doesn’t seem to substantially disrupt emotional-empathic giving (but there may still be a lack of power to detect small effects), and may enhance it. There seem to be heterogeneous effects by type of donor (Karlan), but this needs further examination (as it’s an ex-post hypothesis). The ‘right’ empathic images can drive effective giving. In less precise comparisons, overstated analytical effectiveness information seems to boost giving, at least relative to more realistic effectiveness information.

Unfold for a further discussion (from (Bergh and Reinstein 2020) abstract)

A small share of charitable giving goes towards “highly-effective” charities as measured by GiveWell (or to charities in this category: health interventions in poor countries).

Previous work finds that helping behaviors are strongly driven by emotional reactions to individual suffering, but these do not scale with the magnitude of need. From a dual-process perspective, presenting analytical impact information may even disrupt empathic giving, shifting potential donors towards a system-II deliberative mode. We report on seven charitable giving experiments. Treatments (between subjects, mainly 2x2) present (i) measures of charity efficiency/effectiveness and (ii) empathy-inducing images. Variations across experiments include: subject pools (MTurk, UK students, US regular donors via direct mail), separate vs. joint evaluation, the nature of the effectiveness information, and treatment ordering/timing.

1.1 Definition of ‘effective giving’ for our purposes, motivation*

Please see Definitions - “Efficiency” versus impact from ‘Barriers’ web book for a more in-depth presentation of this.

We define the impact, also referred to as the effectiveness of a charity, as follows:

$B(G_i)$: The beneficial outcome achieved by charity $i$ with total donations $G_i$

• This is an ultimate outcomeL Lives saved, QALY added, etc.

• We are not referring to intermediate ‘outputs’ such as ‘nets provided’ nor ‘paintings purchased’

(Total or marginal) impact per dollar = output per dollar $\times$ impact per output

Steinberg & Morris, 2010 wrote about marginal vs average effectiveness.

:::

Impact of a donation:

$B'(G_i)$ for the marginal donor

• GiveWell and others attempt to measure this

We have a strong empirical case that (in-depth discussion here):**

$B'(G_i)$ is much larger for most impactful vs. most popular charities.

1.2 Raises questions

1. “Why don’t we give more to the most effective charities and to those most in need?”,

and

2. “Why are we not more efficient with our giving choices?”

Is this a ‘Puzzle’?

1.3 Barriers and the use of analytical effectiveness information

We give a breakdown and assesment of potential barriers to effective giving in the aforementioned ‘barriers’ book.

Unfold for a peek at our proposed definitions:

1. Psychological distance/emotional arousal: Whether a cause/charity is something people are aware of, feel is important/salient, and feel close to

2. Identity: Things that run against one’s self-perception and how one believes others will will see you. (e.g., signaling emotional empathy; avoiding cognitive dissonance from previous ineffective giving)

3. Inertia and systemic/institutional “Factors limiting individual & social change in behavior

4. Quantitative biases : Problems dealing with numbers and making logical evaluations involving numbers, values and amounts. Common cognitive errors or intuitive and emotional reactions triggered in these contexts. (E.g., Overhead aversion; Proportional dominance)

1.4 Presenting analytical effectiveness information: The key issues

Obstacles/aversion to doing evaluations; analytical/empathetic clash

• Reluctant to evaluate (taboo tradeoffs, ‘social v. market norm’, …)

• The evaluation process switches off empathy $\rightarrow$ the present focus

Aversion to doing evaluations: “Things that make people reluctant to evaluate and consider the effectiveness of charities, and ways that doing these evaluations reduce generosity.” (E.g., taboo tradeoffs; market vs social norm; dual-system)

1.4.1 ‘Processing of effectiveness information’ and spontaneous/deliberate responses

• Heuristic (fast) $\rightarrow$ spontaneous generosity?

• Deliberative (slow) $\rightarrow$ thoughtful giving … or ‘calculated greed’

• Also: relational models theory (Fisk, ’91); Motivated reasoning (e.g., Exley ’19) $\rightarrow$ all suggests analytical information $\rightarrow$ less giving

Further discussion and channels (unfold; needs cleanup)

Spontaneous Emotions and Deliberation: Dual-Process Perspectives What happens when we combine emotional and reasoning appeals?

Dual process models of prosocial behaviors:

theoretical arguments for this perspective comes from other research on generosity and cooperativeness, suggesting that people are spontaneously prosocial and more self-serving when they think longer/more (e.g., Rand et al., 2012, Rand, 2016).

cf Fisk: Relational models theory (x1991 ) … all thinking about social relationships is based on four elementary mental models: communal sharing, authority ranking, equality matching, and market pricing.

Alt. channel: Exley, Motivated reasoning: ’ …if participants desire to keep money for themselves and to think of themselves as generous, they may overweight their dislike of less-than-perfect performance metrics as an excuse not to give.’

1.5 Evidence for related effects and channels that seem similar (but not analytical effectiveness information)

Before presenting the direct and indirect evidence on the impact of, and response to analytical effectiveness (particularly ‘impact per dollar’) information… … we consider evidence on some related, but distinct responses and motivations.

Note: in the next section we also have a section considering ‘indirectly relevant work’: work that informs the main question, but only obliquely. (We may want to merge these discussions.)

Identifiable victims effect/ deservingness vs deliberation

Claim: Better to portray an individual (child) than convey the total affected Small & Loewenstein (03); Small et al (07); Kogut & R (05)

• Driven by System-1 empathy, switched off by analytic thinking

---

Small, Lowenstein, Slovic (2007):

[Study 3] “individuals who faced an identifiable victim donated more…”

• “…than those who faced victim statistics, p < .01,”

• …“than those who faced [both] an identifiable victim [and] statistics, p < .05.””

---

Small et al, ’07, Study 4

Priming analytic thinking reduced donations to an identifiable victim relative to a feeling-based thinking prime.

Yet, the primes had no distinct effect on donations to statistical victims, which is symptomatic of the difficulty in generating feelings for these victims.

Tightly bounded null, but … nonlinearity?

Note, the latter non-effect appears tightly bounded; but this could simply be driven by nonlinearity. If people gave little to statistical victims, there is less room for this to decrease further. A classic problem when considering interactions.

Verkaik (2016)

While previous studies have convincingly shown that providing output information, informing donors of how their donation is used, increases generosity (Cryder & L, ’10; Cryder ea ’13; Aknin ea ’13)

Discussion:

Robin: I personally wouldn’t count the Cryder studies as examples of studies demonstrating that effectiveness has a positive effect on giving. There’s a fundamental difference between psychological perceptions of impact (and “tweaks” of info to make it seem more impactful to donate) versus evaluating actual evidence of effectiveness. For instance, the fact that donors are more willing to donate when a fundraiser is close to reaching a goal tells us something about how people feel about donating, but it tells us absolutely nothing about the actual effectiveness of a charity and how people process such information. Effectiveness is an attribute of the charity, not a perception of making a difference (if we reduce effectiveness to a subjective impact belief then guide dogs could also argued to be highly effective).

DR: True; But I wouldn’t say it is completely unrelated. For actual cost-effectiveness/impact information to be used to increase/improve giving we presumably need both

1. Individuals care about their own impact per $, as they perceive it (Cryder speaks to this),2. Individuals can reasonably draw conclusions about effectiveness per$ from the information properly presented, and 3. The presentation of this information does not itself substantially reduce the actualized motivation to give,

…the evidence on the effects of impact information are more mixed, with mainly null effects (Metzger & G ’15; Karlan & W, ’14; Baron & S, ’10; Caviola ea ’14, Berman ea ’15)

1.5.1 Ratings and information in general: mixed evidence

• Yörük (2016, JEMS): RD w/ Charity Navigator; significant for ‘small’ charities only

  • See also Brown ea (2017), Gordon ea (2009)

• Charity Navigator stars based on continuous score across categories (not EA criteria)
• Identification via RD: Impact of crossing a ‘star’ threshold on amounts raised Results: Significant impact for small charities only: finds a nearly 20% effect of a one star increase

References

Bergh, Robin, and David Reinstein. 2020. “Empathic and Numerate Giving: The Joint Effects of Victim Images and Charity Evaluations.” Social Psychological and Personality Science, 1948550619893968.