00:00:00
Katy Milkman: As behavioral
scientists, we actually find it
00:00:02
quite unusual that people give
at all. So being generous,
00:00:07
making donations to
organizations, is something that
00:00:09
a strict rational actor model in
economics might say should never
00:00:13
happen, right? Unless you have a
financial interest in the
00:00:16
organization yourself, because
you'll somehow gain. And so this
00:00:20
is sort of a ripe area for
Behavioral Science Research.
00:00:23
Dan Loney: Welcome to
<i>The Ripple Effect</i>,
00:00:25
the podcast that takes you on a
00:00:27
journey through the minds of
Wharton faculty. I'm your host,
00:00:30
Dan Loney, and in each episode,
we'll be diving deep into the
00:00:33
inspiration behind the
groundbreaking research that
00:00:36
Wharton professors have
conducted, and exploring how
00:00:39
their findings resonate with the
world today.
00:00:42
- When you think about
donating to an organization,
00:00:45
when is the right
00:00:46
time for that organization to
make contact?
00:00:49
Research looked at
00:00:51
how hospitals contact potential
donors, especially ones who had
00:00:55
been at their facility for
treatment in the recent past.
00:00:59
One of the people involved in
that research is our guest here
00:01:01
in studio, Katy Milkman, who is a
Professor of Operations,
00:01:04
Information and Decisions here
at the Wharton School.
00:01:07
Great to see you again.
00:01:09
- Great to see you too.
00:01:10
How did this idea come about in
the first place, when you're
00:01:13
thinking about the charitable
component? Obviously, hospitals
00:01:17
do rely, a component, on their
revenue
00:01:20
being from a charitable avenue.
00:01:22
Yeah. Well, so first of all, of
course, we were really
00:01:24
interested in charitable giving,
because the team that did this
00:01:27
work, led by Amanda Chuan and
also with Judd Kessler, another
00:01:30
Professor here, we are
interested in charitable giving
00:01:34
because as behavioral
scientists, we actually find it
00:01:38
quite unusual that people give
at all. So being generous, making
00:01:43
donations to organizations, is
something that a strict rational
00:01:46
actor model in economics might
say should never happen, right?
00:01:50
Unless you have a financial
interest in the organization
00:01:52
yourself, because you'll somehow
gain. And so this is sort of a
00:01:56
ripe area for behavioral science
research. And then hospitals are
00:02:00
a place where a lot of
charitable giving takes place.
00:02:02
So many hospitals benefit
immensely from charitable giving
00:02:06
each year. And so that seemed
like an interesting place to
00:02:08
study this. But when you asked
specifically about this project,
00:02:12
where we're interested in the
speed at which an "ask" follows
00:02:16
the receipt of service—this was
really something that we got
00:02:19
interested in when we realized
that there's a lot of different
00:02:22
theories that might predict that
you'd want to capitalize quickly
00:02:26
on people's emotions and
gratitude when it's an
00:02:30
organization, say an animal
shelter or a hospital or a
00:02:33
religious organization, or even
an organization providing relief
00:02:38
after a natural disaster. People
may have a quick feeling of warm
00:02:43
glow, to use the economic
terminology of gratitude. And
00:02:48
that might crater as time
passes. And so you may want to
00:02:52
be quick to take advantage of
those warm feelings that are
00:02:55
going to arise right after this
interaction. On the other hand,
00:02:59
though, there's sort of the risk
of being perceived as too quick
00:03:06
to ask, too quick to capitalize
right after doing something that
00:03:10
was generous and charitable,
which is the nature of
00:03:12
charitable organizations. If you
quickly say, "Oh yeah. Now, could
00:03:16
you give us some money?" the
person who received that, you
00:03:20
know, positive encounter may
feel like this is kind of icky,
00:03:25
and I don't like being contacted
by my hospital. Or by the— you
00:03:29
know, shelter that— where I got
my puppy, so quickly after they
00:03:33
did something nice for me. And
it feels like they're trying to
00:03:35
capitalize. And so there could
be reactants, which is the word
00:03:38
we use for when it's negative.
So we thought that was
00:03:41
interesting. We always love
studying questions where you
00:03:44
could imagine the result going
in either direction, and that
00:03:48
got us excited about this
project.
00:03:49
But seemingly, then— that is the
issue, or the question that a
00:03:54
lot of organizations deal with—
is, how you formulate that plan
00:03:58
with the timing component being
right near the top in terms of
00:04:02
the importance of making that
contact with that person at the
00:04:06
right time and getting the best
positive outcome.
00:04:09
That's right.
00:04:10
We want to be top of mind and
sort of maximize the feelings of
00:04:14
gratitude, and minimize the sense
that we're being sort of too
00:04:18
grabby for money after a
positive encounter. So it seemed
00:04:22
like a really interesting
research question that would
00:04:24
also tell us something about
human nature. And to what extent
00:04:29
is gratitude a stable feeling,
and to what extent does it
00:04:33
really decay quickly over time?
00:04:34
So as we're taping this, we're
actually a couple weeks away
00:04:37
from two massive hurricanes that
hit the United States, and so
00:04:41
that dynamic in terms of the
donation really differentiates
00:04:45
from what would be, I guess,
traditional models or paths of
00:04:50
asking for donations. Because
you have such an immediate
00:04:54
impact and such a disastrous
impact on so many people in
00:04:58
communities across the
Eastern half of the US.
00:05:01
That's right. And of course,
charitable
00:05:02
giving goes up in response to
00:05:04
need, because people can
empathize with victims. And
00:05:07
there's some wonderful work
that's been done on how
00:05:10
important it is to have
identifiable victims. This is
00:05:12
work by former Wharton Professor
Deb Small and George Lowenstein
00:05:15
at Carnegie Mellon. When we
think about a person and their
00:05:18
suffering, we're much more
willing to open our wallets and
00:05:21
give if we think we can impact
them than if we think about
00:05:24
statistical victims. So just
describing how many people have
00:05:26
been displaced. And these
organizations know that. They
00:05:30
wisely fund raise at times of
need and do so in effective
00:05:34
ways. But what we're studying in
this project is actually for
00:05:39
those who were recipients of,
say, rescue workers' aid. What is
00:05:45
the optimal time to think about
asking them, "You've benefited,
00:05:50
and now let's think about sort
of turning around and asking
00:05:53
you, would you give because
someone else in the future is
00:05:55
going to also need help?" So we
didn't, again, study this in the
00:05:58
exact context of hurricanes or
disaster relief. We studied it with
00:06:01
hospitals, but the idea is that
there'd be a similar dynamic
00:06:04
where you'd want to think about,
what's the right timing after
00:06:06
someone has benefited from an
organization that's a charitable
00:06:09
organization's contributions, to
ask that recipient of the
00:06:15
benefit giveback.
00:06:16
And so with this work,
00:06:18
in many cases,
you were talking with
00:06:20
individuals, or you were finding
out about individuals who had
00:06:23
been at a hospital in recent—in
the recent past, they had had
00:06:26
some sort of procedure or
medical connection with the
00:06:30
hospital, and it's—so it is a
very unique dynamic of that
00:06:35
hospital having treated that
person, having that interaction
00:06:39
with that person, and then also
asking them for, potentially, a
00:06:44
donation to support the—
00:06:45
kind of the operation
of the hospital as well.
00:06:48
That's exactly right. So these are
first time patients at a large
00:06:50
hospital system, and within 20
to about 130 days after their
00:06:57
first patient encounter, they're
going to receive a letter from
00:07:00
the hospital system asking them
to make a charitable
00:07:04
contribution to the
organization. And what's sort of
00:07:07
interesting about this context
is that this is a large hospital
00:07:10
system that's sending lots and
lots of these asks for
00:07:13
donations. It's a big part of
how they can run the
00:07:15
organization, is the generosity
of these donors. Both patients
00:07:20
and non patients are contacted,
but we're focusing on the
00:07:22
patients. And what they do is
they send these mailings every
00:07:26
few months in a batch.
Basically, an administrator has
00:07:29
time on their calendar, and they
sort of sit down and all this
00:07:32
gets prepared, and the messages
all go out at once. And what
00:07:36
that means is that there's
natural variation in how long
00:07:40
after a person has their
encounter with a hospital
00:07:42
system— because maybe I came in
for surgery on April 1, and
00:07:46
you came in on May 1, and the
next batch goes out on June 1.
00:07:50
You, then, are going to have a
much shorter follow up period
00:07:53
after your encounter with a
hospital than I am. I'm going to have to
00:07:55
wait 60 days for my ask. You're
only waiting 30. And we're
00:07:58
taking advantage of that natural
variation to see—and over a
00:08:02
period of many years, when these
batches are sort of happening
00:08:04
erratically, is there a
consequence? Can we identify a
00:08:08
consequence of a faster versus a
longer lag time after that
00:08:12
patient encounter, for the
likelihood the patient says,
00:08:14
"Yeah, I would like to donate?"
00:08:15
So were you able to determine
what is kind of that sweet spot
00:08:20
for a hospital, for an
organization, to make that ask
00:08:24
of a person who'd been in a— you
know, in that facility?
00:08:28
Yes. And actually—
00:08:29
you talk about a
sweet spot, and that's because
00:08:31
we talked about these— this
tension between the two
00:08:34
competing theories that, you
know, you don't want to act too
00:08:37
fast and seem grabby, right—
like, "Oh, this wasn't done out
00:08:39
of the goodness of the
organization's mission. It was
00:08:43
just, you know, all a ploy to get
you to give us money as fast as
00:08:46
possible. It's almost like a
payment." And then the sort of, "We
00:08:49
want to capitalize on that
feeling of gratitude." So we
00:08:52
thought there might be sort of a
U-shaped relationship. I should
00:08:55
say, an upside down U. Like, it
might be better to wait— yeah,
00:08:59
hump— to wait a little while,
but not too long. And what we
00:09:03
actually found is, you can't act
fast enough, essentially. In our
00:09:06
data, we never saw anyone
getting a mailing asking for a
00:09:12
donation faster than 23 days
after their encounter. But
00:09:16
faster was simply better. We
just saw that there was a pretty
00:09:20
linear decline in giving. And
basically, every extra 30 days
00:09:24
after an encounter you wait to
contact a patient leads to a 30%
00:09:28
decline in the likelihood that
they will make a donation in
00:09:31
response to that ask. So a really
precipitous decline. And it
00:09:35
continues over the whole course
of the time that we were able to
00:09:37
follow patients. They— you know,
they don't ask years later.
00:09:41
- Right. Is that kind of—
00:09:43
you think— kind of
the format that hospitals
00:09:47
will take with all patients? And
I only ask that because I think
00:09:50
some people will react
differently to being contacted,
00:09:54
whatever the time is. You know,
some people, no matter the time,
00:09:57
will feel like, "Oh, my God, it's
another ask from the hospital."
00:10:00
Or they're like, "Oh, yeah. They
did such a great job of taking
00:10:03
care of me. Yes, let me write a
check to them."
00:10:05
That's right,
00:10:05
not everyone is going to be
affected, but it is remarkable
00:10:09
that you'd see this 30% decay
with every 30-day decline.
00:10:12
That's a pretty huge effect. So
it suggests that a lot of people
00:10:15
are impacted by the time frame
in which the ask is made. Now, I
00:10:19
do want to be clear that, again,
the fastest anyone was asked
00:10:22
here and that anyone got a
letter was at least three weeks
00:10:25
after their encounter. So it's
not like, you know, what you
00:10:28
should do is set up a Venmo and
ask people to give 30 seconds
00:10:32
after they leave your
organization having received
00:10:34
some sort of benefit, right? So
maybe that would work. It's
00:10:38
possible, but that would require
further research. But when
00:10:41
you're thinking about a range of,
you know, do I contact them
00:10:43
three weeks later or three
months later, you're much better
00:10:46
off with that three weeks later,
because we do see this decline.
00:10:49
And it suggests that the way we
think about the gratitude
00:10:52
that we feel towards
organizations that help— that
00:10:55
help us— much like many things
about human memory, just sort of
00:10:59
decays over time. Right? When you
have a positive encounter with
00:11:02
an instructor in the classroom
here at Wharton, when you have a
00:11:06
positive interaction on the
street with a friend, whatever
00:11:09
it might be— these things, we
tend to feel really great right
00:11:11
after them. And then pretty
quickly, you know, all the other
00:11:14
things going on in our lives
take over and dominate our mind
00:11:17
space. And so those feelings
decay. That— that strong positive
00:11:22
feeling decays. And so you just
want to make sure that you don't
00:11:25
wait too long. You want to
00:11:26
capitalize on that positive
feeling.
00:11:28
How important then, does it
become for the organization to
00:11:32
have a kind of understanding of
this component of behavior, you
00:11:36
know, in terms of making that
ask in the first place?
00:11:39
You know, again, the effect
sizes we saw
00:11:41
were remarkably large, with a
00:11:42
30% decline in the likelihood of
any donation for every extra 30
00:11:46
days they wait. That suggests it
really is important for the
00:11:49
organization to get its act
together and get these kinds of
00:11:51
asks out quickly. I think
organizations should be thinking
00:11:55
more about timeliness, given how
strong the effects are. And I
00:11:57
think more research is warranted
to try to figure out, what is
00:12:00
that sweet spot? And maybe there
is a too fast. Maybe there's
00:12:03
not. Maybe every organization
should be asking the day after.
00:12:07
My guess is that that would be a
little too fast. But again, more
00:12:09
research would be warranted,
because these are, you know—
00:12:12
there's a lot of organizations
that really do rely tremendously
00:12:16
on— on donations in order to do
their good work. And so being
00:12:20
more attentive to this issue of
speed seems important to me.
00:12:23
Well, and probably, you know, using
the example that you studied
00:12:26
with the hospital, in comparison
to, say, the SPCA or some other
00:12:29
organization like that, the type
of contact that you have with
00:12:33
that person probably plays a
factor in this as well.
00:12:35
I appreciate you asking about
that. Yeah, that's absolutely
00:12:37
right. One of the things we
thought it would be actually
00:12:39
interesting to look at in the
hospital was sort of the
00:12:41
acuteness of the illness that
brought the patient in. And so
00:12:44
it's a little bit tricky to say.
Because we— you know, we
00:12:47
didn't have detailed patient
medical records in addition to
00:12:51
the data we had on giving. But we
did know which portion of the
00:12:55
hospital the patient had been
seen in. So we asked— we had—
00:12:58
there were 11 different sort of
divisions. And we asked a set of
00:13:02
doctors who worked in the
hospital system to classify
00:13:05
which were the most severe types
of patient encounters, and there
00:13:07
was 100% agreement that it was
patients who came in for
00:13:10
surgery, for oncology— right? So
for cancer or for cardiac—
00:13:17
cardiology, right? Something
related to their heart. Those
00:13:19
are the patients who are having
the most significant health
00:13:23
issues, according to these
raters. And we just took a look
00:13:26
to see whether or not these
effects, this decay, this
00:13:29
decline, was any larger for the
patients who had had those kinds
00:13:32
of encounters than patients and
other groups. And we found that
00:13:35
it was substantially larger,
suggesting sort of these more
00:13:37
acute and emotional extreme
encounters might show the
00:13:41
biggest benefit to making sure
that you follow up quickly
00:13:45
before people sort of have gone
back to their normal lives and
00:13:47
aren't thinking so much about
their gratitude for the life-
00:13:50
saving service that potentially
was provided.
00:13:52
You talk a little bit in the paper about
the element of reciprocity.
00:13:57
Is there an expectation of an
organization, whether
00:14:02
intentional or not, that they
should receive some sort of a
00:14:06
donation because of the work
that they have done, or is it
00:14:10
just, you know, we will make our
connections with people and see,
00:14:15
you know, how they react to us
asking them for a donation?
00:14:18
Yeah, that's a great question. I
think reciprocity is a very
00:14:21
fundamental human reaction. We
see reciprocity in lots of
00:14:26
contexts. When— when I scratch
your back, you scratch mine,
00:14:29
it's very common. And so it is
very natural for the strongest
00:14:33
donor bases to organizations
like hospitals, for instance, to
00:14:36
be their former patients. So I
think there is an expectation of
00:14:40
that. Certainly it's going to be
less extreme in cases where
00:14:43
people don't have the means to
provide it, right? So asking
00:14:47
donors who are very low-income
and could barely afford the care
00:14:51
would, I think, be seen
differently than asking someone
00:14:54
who has the financial means to
be a contributor and received
00:14:58
care at a reasonable rate. And—
and you know, maybe had their
00:15:01
life saved. Then you'd expect to
see a larger degree of
00:15:03
reciprocity. So it's going to be
on a scale depending on how much
00:15:06
someone can give. But in
general, that's a major motive
00:15:09
for donations to many
organizations.
00:15:12
Having done this research,
what do you think is—
00:15:14
is the message then to
00:15:15
organizations about how they
should put together their plan
00:15:20
of contact, how they should look
at the process of asking for a
00:15:26
donation from individuals that
have been connected with them?
00:15:30
I think the main message is that
timeliness matters. And I think
00:15:33
we don't think enough about how
time changes the way that
00:15:37
people think about what
decisions they'll make. I've done
00:15:40
a lot of research in this area.
I have some work looking at
00:15:44
something called "the fresh start
effect", where at the beginning
00:15:46
of a new year or a new week or a
new month, people are more
00:15:49
motivated to pursue their goals.
We see these ebbs and flows and
00:15:52
motivation over time in a lot of
parts of life. And I think it's
00:15:57
something that doesn't get
enough attention when we're
00:16:00
thinking about everything from
fund raising to promoting
00:16:04
employee well-being— is
recognizing that every day we
00:16:08
don't wake up equally motivated
to take action on behalf of
00:16:11
another organization or on
behalf of ourselves. There are
00:16:14
moments when we feel more open
to making a change in our life
00:16:19
or expressing gratitude to
another individual, and that we
00:16:23
have to think about that, that
time course of people's feelings
00:16:26
and people's motivation more
carefully, in order to get the
00:16:30
best results if we're an
organization trying to motivate
00:16:32
our workforce or an organization
trying to motivate a donor base.
00:16:37
So, attention to timeliness, I
think, is just a critical thing
00:16:41
that all organizations need to
keep top of mind.
00:16:44
Katy, great to see you again.
Thanks very much.
00:16:46
Thanks so much for having me.
00:16:47
Thank you. Katy Milkman,
00:16:48
Professor of
Operations, Information and
00:16:50
Decisions here at the Wharton
School.
00:16:52
- Thank you for listening
00:16:53
to <i>The Ripple Effect</i>. We hope
you found this episode
00:16:56
informative and engaging. Don't
forget to subscribe and leave us
00:16:59
a review so that we can continue
to bring you the best insight
00:17:03
from the Wharton School.
