Slides here - Rene Bekkers

advertisement
Altruism, Warm Glow
and Charitable Giving
3rd SPI Conference, September 12, 2015
René Bekkers, Dave Verkaik
Philanthropic Studies, VU University Amsterdam
Mark Ottoni-Wilhelm
Lilly Family School of Philanthropy, IUPUI
Thanks to
• Co-authors: Mark Ottoni-Wilhelm (IUPUI)
and Dave Verkaik (VU) + undergrads
• Funders: Templeton Foundation through
the Science of Philanthropy Initiative
September 12, 2015
3rd SPI Conference 2015
2
Smart Philanthropy
• Before spending the $29,999 from SPI for a
field experiment, we invested ~$3,500 of
our own money on four pilot experiments.
• Today I will share insights from these
pilot experiments.
• We have learned some lessons about the
reliability of published research, as well as
fundraising materials used in practice.
September 12, 2015
3rd SPI Conference 2015
3
Envisioned field experiment
• Past participants in the Giving in the
Netherlands Panel Survey 2015 receive a
letter explaining the experiment and a link
to an online survey in which the
experiment is implemented.
• Informed consent + ERB approval.
• Six decisions, one implemented.
• Match donations in experiment with
Oxfam donation history from database.
September 12, 2015
3rd SPI Conference 2015
4
Key results thus far
• It is very difficult to get people to give
more by priming morality.
• Crowding-out effects are highly contextdependent.
• Dispositional empathic concern and the
principle of care are strong predictors of
donations.
• We understand little of the heterogeneity
in treatment effects.
September 12, 2015
3rd SPI Conference 2015
5
Three experiments
1. Van Vliet, R. (2014). Geefgedrag in Nederland:
Altruïsme of Egoïsme? Onderzoek naar empathie, het
principe van zorg, en het crowding-out effect. Master
Thesis VU Amsterdam.
2. Bekkers, R. & D.J. Verkaik (2015). Six Primes That Do
Not Affect the Principle of Care. Mimeo, VU
Amsterdam.
3. Verkaik, D.J., Bekkers, R. & Ottoni-Wilhelm, M. (2015).
“Do What’s Right” and the Principle of Care. Mimeo,
VU Amsterdam.
September 12, 2015
3rd SPI Conference 2015
6
The Big Questions - nested
1. Why do
people give?
2. When do altruism
and warm glow
motivate giving?
3. How much giving do
altruism and warm glow
motivate?
September 12, 2015
3rd SPI Conference 2015
7
Giving = Altruism + Warm Glow
• Giving can be motivated by altruism
(utility from well-being of recipients) as
well as warm glow (utility from giving).
• Under altruism, giving should be crowded
out by giving by others.
• Previous tests have examined crowdingout at relatively low levels of giving by
others.
September 12, 2015
3rd SPI Conference 2015
8
September 12, 2015
3rd SPI Conference 2015
9
Hypotheses
Empathic
concern
+
Third party
contributions
+
-
Private
giving
-
+
Principle
of care
September 12, 2015
3rd SPI Conference 2015
10
Challenges
• Developing a test of the effects of third
party contributions in a natural way;
• Developing a measure of the state of
principle of care;
• Developing a manipulation of the
principle of care that temporarily
increases the state of care.
September 12, 2015
3rd SPI Conference 2015
11
Solutions
• Talking to scholars to hear about null
findings and failed replications in
unpublished research. To our detriment,
we found that interventions based on
published research did not work.
• Looking at materials used in the practice
of fundraising.
September 12, 2015
3rd SPI Conference 2015
12
Budgets
Ottoni-Wilhelm, Vesterlund & Xie, 2014
Budget
€ funded by
sponsor
€ for participant
1
2
3
4
5
4
10
28
34
4
40
40
40
40
46
6
28
46
September 12, 2015
3rd SPI Conference 2015
13
Budgets
€ for participant
1
2
3
4
5
4
10
28
34
4
40
40
40
40
46
6
September 12, 2015
28
3rd SPI Conference 2015
Income
€ funded by
sponsor G-i
Crowding-out
Budget
46
14
Procedure
• All 6 budgets are evaluated in random
order by participants.
• Only 1 will be implemented.
• Yields measures of altruism (α) and warm
glow (β) within participants.
• In Ottoni-Wilhelm, Vesterlund & Xie
(2014) giving is influenced primarily by
altruism and less so by warm glow.
September 12, 2015
3rd SPI Conference 2015
15
Inducing empathy / morality
• Can we use subtle manipulations to
increase empathy and the principle of
care?
• We looked at the priming literature and
talked to several psychologists warning us
against using scrambled sentences,
subliminal and physical cues, in order to
avoid disappointment.
September 12, 2015
3rd SPI Conference 2015
16
• Perhaps asking participants about their
self-image as an empathic / moral person
primes empathy and the principle of care.
• Reminding people of the norm, either
secular or religious, should also increase
the principle of care.
September 12, 2015
3rd SPI Conference 2015
17
Priming empathy and morality
• Participants read an
article about Oxfam
helping a family in
need by providing
them with chickens.
• Participants either
completed empathy
and principle of care
Design: 2 (self-image questions: no, yes) x 4 (priming:
scales or not.
none, image of hands, religion, moral appeal)
Participants: MTurk (n = 457)
September 12, 2015
3rd SPI Conference 2015
18
The principle of care - trait
a. People should be willing to help others who are less fortunate.
b. Everybody in this world has a responsibility to help others when they need
assistance.
c. These days people need to look after themselves and not overly worry
about others.
d. When people are less fortunate, it is important to help them even if they are
very different from us.
e. It is important to help one another so that the community in general is a
better place.
f. Personally assisting people in trouble is very important to me.
g. When thinking about helping people in trouble, it is important to consider
whether the people are like us or not.
h. We should not care too much about the needs of people in other parts of the
world.
September 12, 2015
3rd SPI Conference 2015
19
A better measure of state care
• We decided to use an adjective checklist,
describing the principle of care, avoiding
emotions.
–
–
–
–
–
Moral
Rational
Principled
Deliberate
Responsible
September 12, 2015
Following the model of the
empathic state measure
developed by Batson et al.,
1997
3rd SPI Conference 2015
20
State care measure
Please indicate the degree to which you
thought about these principles after reading
the story. Responsibility, Helpfulness,
Beauty(F) Benevolence, Loyalty(F), Humanity,
Commonality, Authority(F), Assisting,
Commitment, Purity(F), Justice, Selfcentered(*), Liberty, Disregard(*), Remoteness(*)
Three factors: care, isolation(*), fillers(F)
Good reliability, α = .86; .70; .87
September 12, 2015
3rd SPI Conference 2015
21
Familiar correlations
Trait care
State care for
recipients
State empathy
adjectives
State care adjectives
Filler adjectives
Isolation adjectives
September 12, 2015
Trait
empathy
.69
.60
Trait
care
.57
.58
.75
.33
.01
-.06
.55
.02
-.22
.65
.27
-.22
3rd SPI Conference 2015
State care for
recipients
.69
22
BUT NO EFFECTS OF…
• Having participants complete questions
on empathy and the principle of care
• Asking participants ‘How religious are
you?’ before they make donation choices
• Adding the sentence ‘You can do the right
thing by donating..’
• Showing participants a pair of hands
adding ‘with your donation you are able to
help end injustices that cause poverty’
September 12, 2015
3rd SPI Conference 2015
23
Self-image on states…
1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
state care for state empathy
recipients
adjectives
-0.6
state care
adjectives
filler
adjectives
48 < n < 117; 0.256 < F < 1.686 (df = 3); p > .167
isolation
adjectives
48 < n < 65; F = 2.685 (df = 3); p = .047
-0.8
-1
no self-image
September 12, 2015
care only
empathy only
3rd SPI Conference 2015
care and empathy
24
Priming on states…
1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
state care for state empathy
recipients
adjectives
-0.6
state care
adjectives
filler
adjectives
isolation
adjectives
55 < n < 118; 0.271 < F < 1.435 (df = 3); p > .233
-0.8
-1
no priming
September 12, 2015
hands
religion
3rd SPI Conference 2015
moral appeal
25
…or giving behavior
100
90
80
70
60
50
40
30
20
10
0
Average gift (€)
no priming
September 12, 2015
hands
Gives > 0 (%)
religion
3rd SPI Conference 2015
moral appeal
26
Very little crowding out
25
% of responses
20
4
15
10
28
10
34
5
n = 457
0
gives nothing
September 12, 2015
3rd SPI Conference 2015
gives half
27
Budgets
ΔG-i
Δgi
Crowding-out
Effect
6: 4  10
6: 28  34
18: 10 28
24: 4  28
30: 4  34
-.37
-.04
-.65
-1.02
-1.68
-.06
-.01
-.03
-.04
-.06
24 (W46): 4  28
-.69
-.03
September 12, 2015
3rd SPI Conference 2015
28
ΔG-i
0
6
12
18
24
30
0
-0.2
-0.4
-0.6
-0.8
ΔGi
-1
-1.2
-1.4
-1.6
-1.8
-2
September 12, 2015
y = -0.0506x + 0.169
3rd SPI Conference 2015
29
ΔG-i
0
6
12
18
24
30
0
-5
Observed crowd-out
y = -0.0506x + 0.169
-10
ΔGi
-15
-20
-25
-30
September 12, 2015
3rd SPI Conference 2015
30
Why so little crowding-out?
• Mturk participants?
• Hypothetical donations?
• A large pool of potential donors.
• Donations benefit families like the Abejide
family, not one specific individual with a
specific need.
• Not a lab study.
September 12, 2015
3rd SPI Conference 2015
31
Source: De Wit, A. & Bekkers, R.
(2015). Government Support and
Charitable Donations: A MetaAnalysis of the Crowding-Out
Hypothesis. Paper under review.
September 12, 2015
3rd SPI Conference 2015
32
Crowding-in Oxfam donations
Van Vliet, 2014
100
90
94
88
82
No induction
(n=96)
80
70
No social information
60
50
VU University donates
€10
40
Remembered that VU
donates €10
30
20
10
0
Donates €20
September 12, 2015
3rd SPI Conference 2015
33
Empathy lowers crowding-in
Van Vliet, 2014
100
90
80
79
83
81
70
Empathy
induction (n=96)
No social information
60
50
VU University donates
€10
40
Remembered that VU
donates €10
30
20
10
0
Donates €20
September 12, 2015
3rd SPI Conference 2015
34
Care induces crowding-out
Van Vliet, 2014
100
90
Care induction
(n=107)
88
80
70
70
69
No social information
60
50
VU University donates
€10
40
Remembered that VU
donates €10
30
20
10
0
Donates €20
September 12, 2015
3rd SPI Conference 2015
35
Does care affect crowding out?
100
Gives > €0
90
% of responses
80
70
60
4
50
10
40
28
30
34
20
n = 55 (no
self-image
questions)
10
0
No priming
September 12, 2015
Moral appeal: "do what's right"
3rd SPI Conference 2015
36
Does care affect crowding out?
100
Gives > €0
90
% of responses
80
70
60
4
50
10
40
28
30
34
20
10
0
No priming
September 12, 2015
n = 57
(started
with selfMoral appeal: "do what's right" image
questions)
3rd SPI Conference 2015
37
Empathy / care and giving
Amount donated in W = 40 budgets
Q1
Q2
Q3
Q4
35
30
25
20
15
10
5
0
trait care
trait state care state care state
filler
isolation
empathy
for
adjectives empathy adjectives adjectives
recipients
adjectives
.40 < r < .51
September 12, 2015
3rd SPI Conference 2015
r = .11
r = -.36
38
Empathy / care do not affect
crowding-out
Difference between donation in
G-i=4 and G-i=34
Q1
Q2
Q3
Q4
state care
state
trait
for
state care empathy
filler
isolation
trait care empathy recipients adjectives adjectives adjectives adjectives
1
0.5
0
-0.5
-1
-1.5
-2
-2.5
-3
September 12, 2015
3rd SPI Conference 2015
39
Summary of results thus far
• Very little crowd-out.
• No priming effects whatsoever on
empathic or moral states or giving.
• Crowding-out does not vary with primes
or self-image.
• Hypothetical giving increases strongly
with dispositional empathy / care.
September 12, 2015
3rd SPI Conference 2015
40
Learning from practice
• How would practitioners prime the
principle of care?
• Can we develop an effective manipulation
that can easily be used in practice?
• To what extent do real world fundraising
materials affect the principle of care?
September 12, 2015
3rd SPI Conference 2015
41
Testing
• Participants are first
exposed to Oxfam’s
mission statement in
different visual forms.
• Then participants
completed state and
dispositional measures.
Design: 3 (visual: text only, text with still image, video) x
• No giving.
2 (moral appeal: no, yes)
Participants: Crowdflower (n = 287); MTurk (n = 304)
September 12, 2015
3rd SPI Conference 2015
42
Oxfam
America
mission
statement
September 12, 2015
3rd SPI Conference 2015
43
Excluding
references to
injustice,
‘right the
wrong’
September 12, 2015
3rd SPI Conference 2015
44
Principle of care state
5.4
5.3
5.2
5.1
5
4.9
Excluding appeal
4.8
Including appeal
4.7
4.6
4.5
4.4
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
45
Principle of care disposition
4.2
4.1
*
4
3.9
Excluding appeal
3.8
Including appeal
3.7
3.6
3.5
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
46
Empathic concern state
5.5
5.4
5.3
5.2
5.1
5
Excluding appeal
4.9
Including appeal
4.8
4.7
4.6
4.5
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
47
Empathic concern disposition
4
*
3.9
3.8
Excluding appeal
3.7
Including appeal
3.6
3.5
3.4
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
48
Personal distress state
3.2
3
*
2.8
Excluding appeal
2.6
Including appeal
2.4
2.2
2
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
49
Personal distress disposition
2.7
2.6
2.5
2.4
Excluding appeal
2.3
Including appeal
2.2
2.1
2
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
50
Perspective taking disposition
4.5
4
*
3.5
Excluding appeal
Including appeal
3
2.5
2
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
51
Universalism values
5.7
5.6
5.5
Excluding appeal
5.4
Including appeal
5.3
5.2
5.1
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
52
Benevolence values
5.8
5.7
5.6
5.5
5.4
Excluding appeal
5.3
Including appeal
5.2
5.1
5
Text
September 12, 2015
Text + image
Video
3rd SPI Conference 2015
53
What we learned from practice
• Exposure to the Oxfam America mission
statement marginally increased (+5-10%)
the state of care, empathy, and distress.
• Also trait empathy and care are affected.
• These effects are most pronounced for the
video, followed by the text only version.
September 12, 2015
3rd SPI Conference 2015
54
Remember the model
Empathic
concern
+
Third party
contributions
+
-
Private
giving
-
+
Principle
of care
September 12, 2015
3rd SPI Conference 2015
55
Our findings thus far
Empathic
concern
+
Third party
contributions
-
Private
giving
(very small)
+
Principle
of care
September 12, 2015
3rd SPI Conference 2015
56
Remaining questions
MTurk:
• How can we more effectively manipulate
the principle of care?
Field experiment:
• Does it increase giving, its motivation, and
affect crowding-out?
• In the long run? Among whom?
September 12, 2015
3rd SPI Conference 2015
57
René Bekkers
Center for Philanthropic Studies
Vrije Universiteit (VU) Amsterdam
r.bekkers@vu.nl
www.giving.nl
Twitter: @renebekkers
Download