Acquiring the New Donor for Nonprofit Organizations: Evidence from a Field Experiment
Using Bonus Trigger Incentives
Sara E. Helms* Timothy M. Diette
Assistant Professor of Economics Associate Professor of Economics
Department of Economics,
Finance, and QA
Department of Economics
Brock School of Business Williams School of Commerce
Samford University Washington & Lee University
800 Lakeshore Drive
Birmingham, AL 35229 Lexington, VA
Voice: 205-726-4110 Voice: (540) 458-8220
[email protected] [email protected]
Betsy Bugg Holloway
Chief Marketing Officer and
Professor of Marketing
Department of Management,
Marketing, and Entrepreneurship
Brock School of Business
Samford University
800 Lakeshore Drive
Birmingham, AL 35229
Voice: 205-726-4109
Fax: 205-726-2464
*Corresponding author. Acknowledgements: Diette is appreciative of the financial support
provided by the Lenfest Summer Research Grant through Washington and Lee University.
Thanks to Russell James and participants at the Southern Economics Association Meetings for
comments and suggestions. All errors are the responsibility of the authors.
Author Biographies
Sara E. Helms is Assistant Professor of Economics and Brock Scholars Program Coordinator in
the Brock School of Business at Samford University in Birmingham, Alabama. Her research
focuses on giving and volunteering behavior, education policy, and health policy.
Timothy M. Diette is Associate Professor of Economics in the Williams School of Commerce
and is a faculty member of the Shepherd Program for the Interdisciplinary Study of Poverty and
Human Capability and the Africana Studies Program at Washington & Lee University in
Lexington, Virginia. He is also an affiliate member of the Research Network for Racial and
Ethnic Inequality at the Social Science Research Institute at Duke University. His research
focuses on education policy, health policy, and giving and volunteering behavior.
Betsy Bugg Holloway is Professor of Marketing in the Brock School of Business and Chief
Marketing Officer for Samford University in Birmingham, Alabama. She serves on the editorial
review board of Journal of Marketing Theory and Practice, and her research interests include
services marketing, retailing, and consumer behavior.
Acquiring the New Donor for Nonprofit Organizations: Evidence from a Field Experiment
Using Bonus Trigger Incentives
Abstract
There is a rich literature on fundraising and charitable giving by and for nonprofit organizations
which considers incentives applied to existing and potential donors. We add to the existing field
experiment research body in two important ways. First, we establish results for the impact of
incentives on acquisition of new donors for organizations which do not have clearly partisan
appeal, unlike much of the existing work. Second, we consider the influence of bonus trigger
incentive gifts, which fall somewhere between matching gifts and seed gifts. We find that
promising that a bonus gift of $10 be added to each donation significantly influences the
acquisition and behavior of new donors. While the $1 and $5 incentive conditions did not impact
the sample of potential donors, the promise of the $10 bonus increased the likelihood of
acquiring a new donor and likely changed the distribution of the new donors’ gifts.
JEL Code: C93 Field Experiments; D64 Altruism; Philanthropy; L31 Nonprofit Institutions,
NGOs; M31 Marketing
Keywords: donations; nonprofit; field experiment; new donors; incentives
Acquiring the New Donor – Page 2
Individuals in the United States donate nearly $300 billion annually to charitable organizations
and purposes, and more than one-quarter of the population volunteers each year (Giving USA
2011; BLS 2012). The business of raising money for nonprofits involves consulting firms, and
the variety of ways to make the “ask” (that is ask for money) is astounding. Economists have
been working to understand what affects charitable behavior, through the use of a variety of
research methods. Using large datasets to test hypotheses, experimental studies to isolate which
factors influence behavior, and field studies to observe real-world responses to varying
conditions, the full picture of donor behavior continues to develop. Despite the time, energy, and
resources committed to understanding donor behavior, there is much left to be done.
This study aims to understand the impact of bonus trigger incentive gifts (a low-dollar form of
seed money) on new donor acquisition for a health-related nonprofit firm providing highly-
visible support to households in need. We add to the literature in a few important ways. First,
while numerous studies consider the effect of matching grants and seed grants, we consider
another option available to nonprofit organizations. We use bonus trigger incentive gifts to
prompt donations by potential donors. Second, many if not most existing studies focus on
nonprofit organizations that appeal to a particular demographic; they consider donations to
public radio, environmental preservation organizations, and universities, for a few examples. We
instead offer evidence from an organization that arguably maintains appeal across demographic
and political interests, which is more applicable for similarly-identified nonprofit organizations.
Background and Literature Review
The existing body of economic research on the factors which influence both the acquisition and
behavior of new donors, and the maintenance and behavior of existing donors, shows the
Acquiring the New Donor – Page 3
complexity of the question. While both laboratory and field experiments are part of this research
body, we focus here on field experiments. List (2008) promotes the use of field experiments over
laboratory experiments when considering donor behavior, arguing that field experiments get
much closer to the true decision process. For a detailed history of the study of the economics of
charity, see List (2011).
In the existing literature, most studies focus on matching schemes, rebate schemes, and seed (or
lead donor) schemes. Matching grants match the donor’s gift at some level, usually on the order
of $1 match per $2 gift, or smaller. Rebate campaigns promise to refund any gifts if the
fundraising does not achieve a targeted dollar amount. Seed grants do not match to donors’ gifts,
but add to the overall total fundraising goal. One last (albeit less studied) strategy is the common
“gift with donation” set up, wherein the nonprofit organizations give donors “token” gifts in
appreciation of the donation. Examples include access to early ticket sales and events for a
symphony, or a coffee mug, and the range included between the two.
When considering the impact of the variety of methods used by nonprofit organizations to raise
funds, there are two important dimensions to consider. First, organizations are interested in
which factors influence the probability of giving and the size of the gift for a particular
campaign. Second, organizations must also consider the impact of the current campaign on
future donor behavior of both the newly-acquired and existing donor pool, as nonprofits typically
have repeated interactions with these pools (see Landry, et al (2006) for additional discourse).
Using student donations to an educational fund, Meier (2007) finds that matching incentives
increase giving in the current period, but may decrease future giving for match recipients in later
periods. Meier’s study highlights an important hazard in manipulating incentives for donors—
Acquiring the New Donor – Page 4
one must consider both the current and subsequent effects of the incentive, as long-run outcomes
may differ from current outcomes.
In addition to the interplay between the impact on new and existing donors, charities must also
consider the advantages of offering matching or seed grant incentives to donors. Andreoni (2006)
develops a theory which addresses the motivation for advertising seed gifts. He argues that large
gifts announced in conjunction with fundraising drives, which he terms “leadership giving,”
provide a signal to donors about the credibility of the charity. List and Lucking-Reiley (2002)
test Andreoni’s theory using a university fundraising campaign, and find that increasing seed
money increases contributions. As explained in Karlan and List (2007), matching grants are
effective because they lower the price of giving, they serve as credibility signals, and they signal
that now is the right time to give. Similarly, Bekkers and Wiepking (2011) describe matching
offers as signals for the efficacy of the gifts, and as having a “legitimizing effect” indicating the
trustworthiness of the organization (p. 943).
Karlan and List’s (2007) seminal paper on matching grants and giving behavior focuses on a
nonprofit organization they describe as sufficiently politically motivated that “giving might be a
form of political activism” (p. 1775). There are two major implications of their study for the
current work. They find that matching offers increase the propensity for giving and the revenue
per solicitation. Second, they find that relative to a $1:$1 match ratio, more generous matches of
$3:$1 and $2:$1 do not further alter behavior. Effectively, we extend their study by moving in
the other direction. If larger matches do not improve fundraising, perhaps even smaller matches
can yield similar results.
Acquiring the New Donor – Page 5
Several studies exploit fundraising by local public radio stations to expose potential donors to
different conditions. Croson and Shang (2008) find that priming donors with relative social
position changes the size of the gift. Eckel and Grossman (2008) compare the effects of match
grants and rebates, and find that matching grants are more effective. While rebates led to a net
loss for the organization, the matching gift led to increased total contributions. They also find
“that the match is mainly effective because it attracts more donors, but not larger donations” (p.
8). Using the Giving in the Netherlands Panel Study, Bekkers (2005) finds that matching offers
more effectively increase giving compared to rebate offers. Similarly, he finds that the increased
giving comes about through additional donors but not larger gifts. Perhaps closest to our study
design, Rondeau and List (2008) use a pool of existing donors to the Sierra Club to consider the
impact of matching and challenge gifts. They find that challenge gifts, designed appropriately,
increase giving, but that matching did not.
Mail-based fundraising campaigns are a common research topic. They are rather low-cost to
implement, and they provide many opportunities for varied treatments. Hager, et al, (2003)
consider the responsiveness of the organizations themselves to a mail-based survey, and explore
what factors might encourage nonprofit executives to return a survey. They find that a $5 gift
(given regardless of response), the $5 gift plus the promise of a $50 to the organization, and no
such incentive do not lead to differential response rates.
Rasul and Huck (2010) explore the importance of transactions costs with giving in response to
mail solicitations. They explore multiple types of transactions costs, and find that lowering ex
post transactions costs through repeated appeals to non-donors from the initial plea leads to
significantly more giving. Along the same lines, Chen, et al, (2006) examine an online
Acquiring the New Donor – Page 6
fundraising campaign, and find that any incentive led to greater click-through for the campaign
relative to no incentive.
Other studies aim to get at the many complex mechanisms affecting donor behavior. One study
on university fundraising finds that even the previously-ignored impact of the attractiveness of
the solicitor matters in door-to-door fundraising (Landry et al, 2006). Martin and Randal (2008)
study the social feedback inherent in observable and visible fundraising. Using a glass box to
collect donations for an art gallery, they find that when the box is staged with smaller currency,
donors give less than when it is staged with larger-value currency. Together, these studies lend
insight into the social aspect of charity.
Despite the abundance of existing studies, we aim to fill gaps in the understanding of factors
influencing donor behavior.
Experiment Design
In coordination with a local charity office, we implemented a randomized control experiment in
a mail campaign targeted at acquiring new donors.1 The charity is a large-scale, nationally-
recognized, health-related organization whose work is viewed as nonpartisan. The charity
normally uses purchased mailing lists to target potential new donors. In the fall of 2010, the
charity obtained three mailing lists of individuals from commercial companies with a total of
14,023 addresses. We examine whether a potential new donor is more likely to contribute to a
charity if their donation will also bring the charity an additional gift, which we term the bonus
trigger incentive. The additional (bonus) dollars are triggered by the individual donor gifts, thus
providing greater incentive to give. In addition, we explore whether the amount of a donation
Acquiring the New Donor – Page 7
varies conditional on the presence of an additional gift and the amount of the additional gift to
the charity. An external non-profit foundation provided funding for the bonus trigger gifts.
For each of the three mailing lists, we randomly assigned one of four conditions to each mailing
address. The control group received a two page letter from the Executive Director of the charity
with a story of an individual who had benefited from the work of this charity in the past. The
three treatment groups each received an identical letter to the control group with one additional
sentence bolded at the bottom of the first page stating, “In addition, the name of nonprofit
foundation will add $x to your donation, so your gift can have an even greater impact!” For the
three incentive groups, x was 1, 5, and 10 respectively. The charity then provided the results of
the mailing, including (for each household) the amount of the donation and whether the mailing
address was in the control or one of the three treatment groups.
Results
We first explore the response to the bonus trigger incentive by sequentially testing for
differences in the means and distribution of donations across incentive groups for three
questions. First, does the unconditional average revenue per mailing unit vary by incentive
group? Second, does the response rate to the mailing vary by incentive group? And finally,
conditional on the new donor making a gift, does the average donation amount vary by incentive
group? We conclude our analysis with an empirical estimation for each of these three questions
that adds binary control variables for the specific source list of the mailing address.
Unconditional Average Gift Amount
We examine the unconditional average gift donation, which is the average including households
which gave $0, and report the results in Table 1. By design the 14,023 mailing addresses are
Acquiring the New Donor – Page 8
approximately evenly distributed into the four conditions with the number of mailing addresses
by incentive group ranging from 3,493 addresses to 3,515 addresses. In Panel A, the mean
donation per mailing unit varies between $0.103 for the $1 incentive and $0.218 for the $10
incentive with the overall average of $0.142. Due to the low response rate typical for this type of
solicitation, these mean gifts are quite low. A formal cost benefit analysis of the mailing would
require understanding both the explicit cost of the mailing as well as expected future gifts
generated by acquiring first-time donors. The total amount paid by the foundation funding the
bonus trigger gifts was $481.
[Table 1 about here.]
Panel B displays the results from the t-test for whether the unconditional means vary by
incentive group. The final column in Panel B reports results of the Mann-Whitney rank-order
test for whether the samples have the same distribution and we report the p-values for the null
hypothesis of identical distributions. The first three rows show the results for each incentive
group relative to the control (or no bonus trigger incentive) group. The mean difference column
reports the mean of the relevant treatment group less the mean of the control group. While
unconditional means for the $1 and $5 bonus trigger incentive groups are lower than the control
group, they are not statistically different from the control group. We report tests for both tails as
it is theoretically possible that some potential donors are offended by the bonus trigger incentive
gift from the foundation to the charity and the incentive actually harms the unconditional mean
donation. The $10 bonus trigger incentive group has a statistically significant larger average gift
per mailing unit than the control group. The subsequent rows test whether the $5 or $10
incentive unconditional means are statistically different than the $1 incentive. There is no
difference between the $1 incentive and the $5 incentive, but the unconditional mean donation in
Acquiring the New Donor – Page 9
the $10 incentive ($0.218) is over twice as large as that of the $1 incentive ($0.103), and the
difference is statistically significant. The final row in Panel B displays the result that the
unconditional mean donation of the $10 incentive is also statistically larger than the $5 incentive.
The results for the rank order tests of the distribution of donations mirror the results for the
unconditional means. The $10 bonus trigger incentive group is statistically different than all
other groups. The $1 and $5 incentive distributions do not statistically differ from each other or
the control group.
Response Rates by Incentive Group
In many cases, generating a response from a potential new donor may be more important to a
charity than the actual gift amount. The new donor presents an opportunity for additional future
giving. The overall response rate is 0.7 percent with response rates varying between 0.57 percent
for the no incentive group to a high of 1 percent for the group in the $10 incentive group. This
level of response is consistent with those found in other studies.2 Panels A and B in Table 2
mirror the results from Panels A and B of Table 1. Those households in the $10 bonus trigger
incentive group are statistically significantly more likely to respond with a donation than either
of the other incentive groups or the control group. In addition, the $1 and $5 incentive are not
statistically different from each other or from the control group. These findings suggest that a
bonus trigger incentive threshold is required with this type of offer to elicit a response from the
targeted new donor.
[Table 2 about here.]
Average Gift Amount among New Donors
Acquiring the New Donor – Page 10
We now turn to examining the average gift amount conditional on the potential donor responding
with a donation. Theoretically, the incentive may increase or decrease the average size of the
donation. The bonus trigger incentive is a fixed amount that the nonprofit foundation will make
if the individual responds with a donation to the charity. The incentive may increase the
likelihood that the marginal prospect will give a donation. This individual may give a very small
amount with the goal of entitling the charity to the additional gift from the foundation and
therefore lower the average donation. The bonus trigger incentive could increase the average gift
amount if individuals feel pressure to give more in light of generosity of the foundation, or
because they view it as a leadership gift. Panel A and Panel B of Table 3 appear to support to the
theory that the average gift will be smaller. The mean gift among the control group is $22.75,
the highest of the four groups. The mean gift of the $1 incentive group is $17.19 and statistically
smaller than the average gift among those in the control group. None of the other comparisons
yield differences in average donations between incentive groups. The rank order test reported in
the last column of Panel B suggests that the $10 incentive group has a different distribution of
gifts than the control group. As shown in Panel A, the $10 incentive group received one $100
donation while the maximum gift for all other incentive groups was $50.
[Table 3 about here.]
Controlling for the Source of the Mailing Address
The purchased mailing lists came from three different consumer-based businesses. The customer
base of these three businesses may be systemically different from each other. Therefore, our
final analysis predicts the unconditional amount of the gift, likelihood of giving, and amount of
the donation conditional on giving, controlling for the incentive group as well as an indicator
Acquiring the New Donor – Page 11
variable for two of the three lists. The reference group is comprised of those in the control group
and with addresses from the company providing List 1. The results are in Table 4. Column (1)
contains the linear probability estimates from an ordinary least squares regression of the
likelihood that a letter will generate a donation. As previously mentioned only a small
percentage of letters resulted in a donation to the nonprofit. Due to the typical concerns
regarding linear probability models we also estimate the probability of responding to the direct
mailing using maximum likelihood estimation, assuming the specifications take on a probit
distribution. The marginal effects reported in column (2) are estimates of the variable
contributions evaluated at the sample means for all the other controls. Finally, we use ordinary
least squares to estimate the effect of the incentive groups and the source of the mailing list on
the gift amount for the 98 mailing units that responded with a donation. These results are
reported in column (3).
[Table 4 about here.]
For each of the three estimations, the results are consistent with findings from the tests of
differences in means. From column (1), the $10 bonus trigger incentive increases the likelihood
of a donation relative to the excluded incentive group—the control group with no bonus trigger
incentive. The results from the χ2 tests confirm that the $10 incentive group is distinct from the
$1 and $5 incentive groups. In addition, we see that each of the three sources of the mailing
addresses is distinct from each other. As we designed the experiment to equally distribute the
incentive groups within each of the mailing lists, we did not expect that controlling for the
mailing lists would change our results for the incentive groups. However, due to the source of the
confidential lists, our results are consistent with the idea that those already sympathetic to the
Acquiring the New Donor – Page 12
general mission of the nonprofit we worked with are more likely to give, while those with no
connection to the issues addressed by the nonprofit will be less likely to give.
The results for the likelihood of giving a donation also match our findings from Table 2. Again
the $10 incentive group is statistically significantly more likely to respond than either the control
group or the $1 or $5 incentive groups. The other three groups do not have a differential effect
on the likelihood of giving. Each of the three lists is again significantly different from each
other. Finally, the OLS regression of the amount donated conditional on giving suggests that
none of our observable characteristics influence the donation amount.
Conclusions
Acquiring new donors is an important goal for many charities looking to increase contributions
in order to expand the assistance they provide in communities. One potential incentive
mechanism is to find a foundation or individual willing to increase the value of a new donor’s
gift. Matching gifts are commonly used, but impose risk upon the foundation if the match does
not include a specific total limit on either the match per gift or total amount to be matched across
all givers. An alternative is to provide a bonus trigger incentive, a fixed amount to be added to
the gift of a new donor. We examine the influence of $1, $5, and $10 bonus trigger gifts on the
mean unconditional gift amount, the likelihood that individuals respond with a gift, and the
average gift conditional on a donation being received. The results from a mailing to over 14,000
household units suggest that the additional gift can matter, but only for the largest promised
bonus trigger gift within our experiment, $10. Additional research is needed to better understand
what bonus trigger gift amounts may be related to discontinuous jumps in the likelihood of
giving. The incentives did not alter the average gift conditional on giving, but if the goal of the
Acquiring the New Donor – Page 13
nonprofit is new donor acquisition then the promise of a fixed dollar addition to the initial
donation may be beneficial to charities.
Acquiring the New Donor – Page 14
Endnotes
1 We acquired the appropriate Institutional Review Board clearance for the study through
Samford University’s IRB process.
2 For example, Eckel and Grossman (2008) had a response rate of 0.5% for prospective donors in
their study, and 0.9% for lapsed donors.
Acquiring the New Donor – Page 15
References
Andreoni, James. 2006. Leadership giving in charitable fund-raising. Journal of Public
Economic Theory, 8(1), 1-22.
Baker III, R.J., J.M. Walker, and A.W. Williams. 2009. Matching contributions and the
voluntary provision of a pure public good: Experimental evidence. Journal of Economic
Behavior and Organization, 70(1-2), 122-134.
Bekkers, R. 2005. When and Why Matches are More Effective Subsidies Than Rebates.
http://igitur-archive.library.uu.nl/fss/2006-0727-201343/Bekkers_05_When-and-Why-Matches-
are-More-Effective.pdf (downloaded 19 June 2013).
Bekkers, R. and P. Wiepking. 2011. A literature review of empirical studies of philanthropy:
Eight mechanisms that drive charitable giving. Nonprofit and Voluntary Sector Quarterly, 40(5),
924-973.
Bureau of Labor Statistics. 2012. Volunteering in the United States 2011. United States
Department of Labor. USDL-12-0329.
Chen, Y., X. Li and J.K MacKie-Mason. 2006. Online fund-raising mechanisms: A field
experiment. The Berkeley Electronic Press Contributions to Economic Analysis and Policy, 5(2),
Article 4, 1-37. Last accessed July 13, 2011.
Croson, Rachel, and Shang, Jen. The Impact of Downward Social Information on Contribution
Decisions. Experimental Economics, 11(3), 221-233.
Eckel, C. and P.J. Grossman. 2008. Subsidizing charitable contributions; a natural field
experiment comparing matching and rebate subsidies. Experimental Economics, 11(3), 234-252.
Giving USA Foundation. 2011. The Annual Report on Philanthropy for the Year 2010:
Executive Summary. Giving USA and The Center on Philanthropy at Indiana University. Last
accessed 11/10/2012.
Hager, Mark A., Sarah Wilson, Thomas H. Pollak and Patrick Michael Rooney. 2003. Response
Rates for Mail Surveys of Nonprofit Organizations: A Review and Empirical Test. Nonprofit and
Voluntary Sector Quarterly, 32: 252. doi: 10.1177/0899764003032002005
Karlan, D. and J. List. (2007). Does price matter in charitable giving? Evidence from a large-
scale natural field experiment. The American Economic Review, 97(5), 1774-1793.
Landry, C., Lange, A., List, J., Price, M., and Rupp, N. 2006. Toward an understanding of the
economics of charity: Evidence from a field experiment. The Quarterly Journal of Economics,
121(2), 747-782.
Acquiring the New Donor – Page 16
List, John. 2008. Introduction to field experiments in economics with applications to the
economics of charity. Experimental Economics, 11(3), 203-212.
List, John. 2011. The Market for Charitable Giving. Journal of Economic Perspectives, 25(2),
157-80.
List, John, Lucking-Reiley, David. 2002. The effects of seed money and refunds on charitable
giving: Experimental evidence from a university capital campaign. Journal of Political
Economy, 110(1), 215-233.
Martin, R. and J. Randal 2008. How is donation behaviour affected by the donations of others?
Journal of Economic Behavior and Organization, 67(1), 228-238.
Meier, Stephan. 2007. Do subsidies increase charitable giving in the long run? Matching
donations in a field experiment. Journal of European Economic Association, 5(6), 1203-1222.
Rasul, Imran, and Huck, Steffen. 2010. Transactions costs in charitable giving: Evidence from
two field experiments. The B.E. Journal of Economic Analysis & Policy, 10(1), Article 31.
Rondeau, Daniel, and List, John. 2008. Matching and challenge gifts to charity: evidence from
laboratory and natural field experiments. Experimental Economics, 11(3), 253-267.
Acquiring the New Donor – Page 17
Tables
Table 1
Comparison of Average Gift Amount
Full Sample
Panel A
Unconditional Mean Donation per Mailing Unit by Incentive Group, in $US
Mean Std. Dev. Obs.
No Incentive $0.130 1.986 3,504
$1 Incentive $0.103 1.593 3,493
$5 Incentive $0.115 1.731 3,515
$10 Incentive $0.218 3.234 3,511
Full Sample $0.142 2.233 14,023
Panel B
Test for Differences by Incentive Group
T-Test of Unconditional Mean Donations and Rank Order Tests of the Distribution
Mean
Difference
Pr(T<t) Pr(|T|>|t|) Pr(T>t) Mann-
Whitneya
Versus No Incentive
$1 Incentive -0.027 (0.269) (0.538) (0.731) (0.871)
$5 Incentive -0.015 (0.371) (0.742) (0.629) (0.768)
$10 Incentive 0.088 (0.915) (0.170) (0.085)* (0.044)
**
Versus $1 Incentive
$5 Incentive 0.012 (0.617) (0.765) (0.383) (0.894)
$10 Incentive 0.115 (0.970) (0.060)* (0.030)
** (0.063)
*
Versus $5 Incentive
$10 Incentive 0.103 (0.951) (0.097)* (0.049)
** (0.084)
*
aTest for the probability that the samples have the same distribution. Results are consistent with
Kruskal-Wallis Rank Order Test. p-value in parentheses indicates null hypothesis of identical
distributions.
Notes: significant results in bold; *** p<0.01, ** p<0.05, * p<0.1
Acquiring the New Donor – Page 18
Table 2
Comparison of Response Rates by Incentive Group
Panel A
Average Response Rate by Incentive Group
Mean Std. Dev. Observations
No Incentive 0.0057 0.0753 3,504
$1 Incentive 0.0060 0.0773 3,493
$5 Incentive 0.0063 0.0789 3,515
$10 Incentive 0.0100 0.0994 3,511
Full Sample 0.0070 0.0833 14,023
Panel B
T-Test for Differences of Response Rates by Incentive Group
Mean
Difference
Pr(T<t) Pr(|T|>|t|) Pr(T>t)
Versus No Incentive
$1 Incentive 0.0003 (0.5662) (0.8676) (0.4338)
$5 Incentive 0.0006 (0.6176) (0.7647) (0.3824)
$10 Incentive 0.0043 (0.9785) (0.0431)**
(0.0215)**
Versus $1 Incentive
$5 Incentive 0.0003 (0.5526) (0.8947) (0.4474)
$10 Incentive 0.0040 (0.9685) (0.0630)* (0.0315)
**
Versus $5 Incentive
$10 Incentive 0.0037 (0.9585) (0.0831)* (0.0415)
**
P-Value in parentheses, Significant results in bold, *** p<0.01, ** p<0.05, * p<0.1
Acquiring the New Donor – Page 19
Table 3
Comparison of Average Gift Amount
Conditional on Giving a Donation
Panel A
Mean Donation per Mailing Unit by Incentive Group, in $US
Mean Std. Dev. Observations Min. Max.
No Incentive $22.75 13.618 20 10 50
$1 Incentive $17.19 11.609 21 1 50
$5 Incentive $18.41 12.188 22 10 50
$10 Incentive $21.85 24.347 35 5 100
Full Sample $20.27 17.580 98 1 100
Panel B
Test for Differences by Incentive Group
T-Test of Conditional Mean Donation and Rank Order Tests
Mean
Difference
Pr(T<t) Pr(|T|>|t|) Pr(T>t) Mann-
Whitneya
Versus No Incentive
$1 Incentive -$5.56 (0.083)* (0.167) (0.917) (0.150)
$5 Incentive -$4.34 (0.141) (0.282) (0.859) (0.213)
$10 Incentive -$0.89 (0.440) (0.881) (0.560) (0.089)*
Versus $1 Incentive
$5 Incentive $1.21 (0.631) (0.739) (0.370) (0.701)
$10 Incentive $4.67 (0.793) (0.415) (0.207) (0.942)
Versus $5 Incentive
$10 Incentive $3.45 (0.730) (0.540) (0.270) (0.569)
aTest for the probability that the samples have the same distribution. Results are consistent with Kruskal-Wallis
Rank Order Test. p-value in parentheses indicates null hypothesis of identical distributions p-value in parentheses, ^ used Chi-squared probability with ties, Significant results in bold, *** p<0.01, ** p<0.05,
* p<0.1
Acquiring the New Donor – Page 20
Table 4: Multivariate Estimation Results
(1) (2) (3)
Variables
OLS:
Donate?
Probit:
Donate?
OLS:
Donation Amount
Conditional on Giving
$1 Incentive 0.0003 0.0006 -6.0426
(0.8673) (0.7802) (0.2849)
$5 Incentive 0.0005 0.0007 -3.6438
(0.7652) (0.7456) (0.5082)
$10 Incentive 0.0043** 0.0040** -0.2574
(0.0421) (0.0436) (0.9589)
List 2 0.0027** 0.0046** 9.1016
(0.0369) (0.0404) (0.1497)
List 3 0.0105*** 0.0109*** 4.2082
(0.0000) (0.0000) (0.4566)
Constant 0.0013 17.5289***
(0.3045) (0.0087)
χ2 Tests
$1=$5 0.8956 0.9642 0.6637
$1=$10 0.0646* 0.0787* 0.2482
$5=$10 0.0841* 0.0854* 0.4833
List 2=List 3 0.0000*** 0.0000*** 0.3125
Observations 14,023 14,023 98
R-squared 0.0033 0.0408
Robust p-values in parentheses, *** p<0.01, ** p<0.05, * p<0.1
(1) Linear Probability-Donation; (2) Probit-Donation; (3) OLS-Amount; Column
(2) displays the marginal effect of the variable calculated at the mean value of all
the other variables.