The effects of informational framing on charitable pledges: experimental evidence from a fund raising campaign.
Shah, Anwar ; Khan, Karim ; Majeed, Muhammad Tariq 等
We designed a field experiment to test the direction of the impact
of informational frame on charitable pledges. We solicited charitable
pledges from 395 students during a campaign aimed at helping students
through students at the School of Economics, Quaid-i-Azam University
(QAU), Islamabad. The participants are randomly divided into 5 different
treatments. In the Pledge Disclosed (PD) treatment, we provided
information to students about the average size of pledge we received
from participants in the Baseline (BL) treatment. Similarly, in the Need
Disclosed (ND) treatment, we provided information about the total need
of those who asked for assistantship. In the Pledge & Need Disclosed
(P&ND) treatment, we informed the students about both the need as
well as the pledge made by the students to meet that need. In All
Disclosed (AD) treatment, we provided details about the need, pledges,
the previous history of the project, and the pledge by Charity Australia
International. The findings show that relative to BL treatment,
charitable pledges decreased when participants were informed about the
previous pledges and the total required need. However, charitable pledge
increased when full information was provided to the participants.
JEL Classification: D64
Keywords: Charitable Pledges, Philanthropy, Helping Students
through Students, Field Experiment
1. INTRODUCTION
Lack of the financial resources to fund higher education is one of
the critical issues of developing countries like Pakistan. Most of the
students are financed by their parents for their higher education.
However, in recent years, there has been an increasing trend in the
contributions from philanthropic organisations. In Pakistan,
organisations such as karwan-e-ilm, Alfalah scholarship scheme, and Agha
Khan Foundation are working in the field of education to assist
students. (1) The main source of the income of these organisations is
religious donations such as Zakat and other charitable donations from
the public. In order to raise funds, they adopt various methods such as
advertisement, fund raising dinners etc. Likewise, they adopt different
strategies to assist students such as direct scholarships, purchasing
books, or providing uniform etc.
Researchers have analysed the effects of various factors on fund
raising. For instance, List and Lucking-Reiley (2002) analysed the
impact of seed money and refund on fund raising. The impact of
government grants on private donations in the form of crowding in and
crowding out has also attracted the attention of scholars (Andreoni and
Payne 2011). In crowding out, individual donors consider their voluntary
private contributions as a substitute for their involuntary
contributions through taxation. Hence, they reduce full amount to a
charity. In contrast, in crowding-in individual donors regard their
contributions complementary to the contributions of government. Mostly,
the concepts of crowding-in and crowding-out have been analysed by the
researchers for the cases where both the government and public are
contributors to philanthropy. To our knowledge, there is no commendable
work on the impact of informational framing on private philanthropic
pledges. We fill this gap with an experimental study where we test
whether the type of information provided to participants affect the size
of pledges or not. These experiments have been carried out in a fund
raising campaign for helping students through students in School of
Economics QAU, Islamabad.
Like the international literature, philanthropy in Pakistan has
been the focus of attention of the researchers [Agha Khan Development
Network (AKDN) (2000); GhausPasha, et al. (2002)]. However, none of them
addresses the hypotheses raised in this paper. For instance, Agha Khan
Development Network (2000) provides a report on the indigenous sources
of philanthropy. Ghaus-Pasha, et al. (2002) documents the key dimensions
such as size, structure, revenue, and the composition of the non-profit
sector in Pakistan. Similarly, Abbasi (2011) analyses the success of the
non-profit sector of Pakistan in terms of funds generated indigenously.
In particular, the study focuses on sectors where the activities of
non-profit sector have contributed in terms of infrastructure
development and the ripple effect. Additionally, the article explores
the philanthropic depth of the society by examining the ability to
handle natural or manmade catastrophes over the decades. Unlike the
previous studies, here we want to see the impact of information
disclosure on philanthropic pledges.
Our study is based on a fund raising campaign which comprises two
rounds. In the first round, we went to class rooms in the School of
Economics, Quaid-i-Azam University, Islamabad. We read out all the
written instructions loudly and invited applications from the needy
students for assisting them in the spring semester of 2014. At the end
of this round, we received 11 applications from students who were
seeking assistantship. In the second round we divided all the students,
approached in the first round, into five treatments. Our aim was to take
pledges in the form of donations from students for helping those needy
students who submitted applications in the first round. Again in the
second round, we read out all the written instructions to students in
all of the 5 treatments.
In the baseline (BL) treatment, the students were asked to give
written pledges on a given form. They could choose an option from the
list or could choose any other amount. In the Pledge Disclosed (PD)
treatment, the students were provided information about the amount of
pledge per student in the BL treatment and they were asked to give
pledges on the same written form. The same exercise was repeated in the
third treatment i.e. Need-Disclosed (ND) treatment, where the students
were provided with the information about the total need asked by the
deserving students in the first round. In the fourth treatment which is
called Pledge and Need-Disclosed (P&ND) treatment, the students were
asked to make pledges after being informed about the need as well as the
amount of pledge per student in all of the previous treatments. In the
final treatment, All-Disclosed (AD) treatment, the students were
provided all the information about the previous developments. For
instance, they were briefed about the history of the campaign, pledges
by the students in the previous treatments, the total need, and the
financial support from an Australian based charitable organisation. (2)
The findings show both the effects of crowding-out and crowding-in.
For instance, crowding-out is observed in the PD treatment while
crowding-in is observed in the P&ND and AD treatments. In the ND
treatment, the information regarding the need for donations decreases
the average rate of pledge relative to the BL treatment. The rest of the
paper is organised in five sections. Section 2 reviews some of
literature on the hypotheses of crowding-in and crowding-out.
Experimental procedure and the description of treatments are provided in
Section 3. Section 4 describes the theoretical framework and discusses
the main hypotheses of the study. Results are presented in Section 5
while the study is concluded in Section 6.
2. REVIEW OF LITERATURE
To our knowledge, there is no commendable work in the available
literature that could explain the effects of information regarding
previous pledges on private pledges to charitable organisations.
However, there is considerable literature that examines the displacement
of donations to the private charitable organisations due to the grant
from government. For example, Steinberg (1991), after reviewing 13
studies on the issue of crowding out, finds that a dollar of government
spending crowds out private donations by $0,005 to $0.35. In the same
way, Payne (1998) finds that the effect is $0.50 for shelter, human
services and similar organisations. Onwards, Ribar and Wilhelm (2002)
find that the government funding to the international relief
organisations crowds out private donations by 23 precent. In contrast,
Straub (2003) finds no crowding out for public radio matching grants.
In order to find the causes of crowding-out Andreoni and Payne
(2011) proclaim that govememnt grants reduce fund raising activities,
and therby reduce the amount raised from private donations.
Alternatively, the reduction of fund raising activities dominates the
reduction in the rate of donations in the crowding-out effect caused by
the government grants. (3) Besides crowding-out, there is also empirical
evidence that supports the possibility of crowding-in. For instance,
Khanna, et al. (1995) finds crowding-in of 9.4 percent in a study of the
non-profit organisations in the United Kingdom (UK). Likewise, Khanna
and Sandler (2000) find some evidence of crowding-in for government
grants in the health and social welfare sectors. In contrast, Payne
(2001) reports the evidence that the federal research grants in the
United States cause both crowding-in and crowding-out of private
donations to universities. For instance, the study shows that the
federal research funding increases private donations by 65 percent to
100 percent for research universities while similar research funding to
liberal arts colleges and non-research universities decreases private
donations by 45 percent and 9 percent respectively.
The above literature suggests that people's response to
government grant depends on their motive for giving, their information
about the nature of the organisations, and their knowledge about the
sources of funding. Regarding these issues, Andreoni (1993) provides an
important source of information to distinguish between the alternative
hypotheses regarding the pattern of contributions. In the study, the
crowding-out hypothesis is tested within the framework of a public good
game. The game consists of three players and has an interior Nash
equilibrium, which is below the level of contributions characterised as
Pareto efficient. In one treatment (no-tax), there is no minimum level
of contribution. The other treatment (tax) sets a minimum level of
investment in the public good. This minimum level is set lower than the
equilibrium level of contribution. The treatment is not framed as a tax,
but rather the payoff matrix is set up so that there is a two-token
minimum. The complete crowding out hypothesis predicts that the total
contributions to the public good in both the no-tax and tax treatments
(including the two-token tax) would be the same. However, the results
show a high level of crowding-out in the tax treatment. For instance,
contributions in the no tax treatment are 71.5 percent higher than the
contribution in the tax treatment across all the rounds.
In a similar study, Eckel, et al. (2005) examines through a
controlled experiment whether the third-party contributions crowd-out
private giving to a charity. A single dictator game is played by the
participants where they choose their preferred rate of charity from a
given list. The experiment has four treatments: two initial allocations
and two frames. Initial allocations are either $18 for the subject and
$2 for the charity or $15 for the subject and $5 for the charity. The
participants could then allocate additional funds if they wished so. In
the first frame, subjects are simply informed of the initial allocations
among themselves and their chosen charity. In the second frame, subjects
are told that their allocation of $20 has been taxed, and the tax
collected has been given to the charity of their choice. The structure
of payoffs is the same in both of the frames. The results show that no
participant contributes additional amount in the tax-frame, showing
nearly 100 percent crowding-out. In contrast, in the first frame, almost
all of the participants contribute some additional amount, showing a
close to zero crowding-out.
In most of the above studies, we can observe that the crowding-out
is mostly associated with the grants from the government. However, this
is quite possible that the information about individuals' grants to
charity displace the donations of the private donors. This question is
partly analysed by Nikiforakis (2010) in a different context in the
creation of a public good game. Nikiforakis (2010) provides feedback to
participants in a public good game regarding the earnings and
contribution of their peers. The paper shows that the level of
cooperation is significantly low in the treatment where subjects receive
information about the earnings of their peers as compared to the
treatment where subjects receive information about the contributions of
their peers. This is in-spite of the fact the feedback format does not
affect incentives. Besides Nikiforakis (2010), there is no study that
systemically answers this question in the context of charitable pledges.
We fill this gap by showing the effects of information revelation on
philanthropic pledges in a field experiment. Moreover, in the
informational frame of the earlier researchers, the donation is not
passed on to the needy among the same subject pool. While, we study the
impact of informational frame on charitable pledges where participants
clearly know that their donations will be actually passed on to the
needy among the same subject pool.
3. EXPERIMENTAL PROCEDURE AND TREATMENTS
The experiment for our analysis was conducted in the School of
Economics, Quaid-i-Azam University (QAU), Islamabad during the months of
November and December 2013. In the school, more than 600 students are
registered in four programs, i.e. BS, MSc, MPhil, and PhD. The
experiment comprised two rounds. In the first round, applications were
sought from the needy students for financial assistantship. During the
teaching hours, after getting approval from the concerned teachers, we
visited class rooms and provided the details of the project by reading
out the written instructions. (4) The students were asked to submit
their applications for financial assistantship within a week at the main
office of the School of Economics. In the application they were required
to provide a brief introduction and background of their financial need.
Students were also informed that a committee comprising of the teachers
of the same school will conduct interviews of the applicants and only
those will be helped, who are recommended by the committee. After
visiting most of the classes, the same written instructions were
displayed on all the notice boards of the school. The purpose was to
ensure that the information reaches all the students who might have been
absent during our visit to their classes. (5) Within the due time, we
received 11 applications. One may wonder about the small number of
applicants. However, the students knew that cheating their own teachers
might not be possible; hence only those applied who were really
suffering financially. It is worth mentioning that almost all of the
applicants were recommended by the committee after interviews. In the
applications, the total demanded amount was 342,000 Pakistani rupees
(approximately $3420) for one semester.
In the second round we visited the same classes after a week. The
main purpose of the second round was to collect pledges from students in
order to support those students who had asked for help in the first
round. A question might arise here that donors and recipients belong to
the same subjects. This was done to make the students realise that the
needy were from them; however, the identity of the needy students was
not provided. After reading out the written instructions, we provided
the pledge forms to the students. The pledge form contained various
options. Each student was asked to select an option of his choice and
drop it in a large collection box that we had placed in the room. The
reason of placing a large collection box was to minimise the
experimenter demand effect if any. It is worth mentioning that the size
of classes was heterogeneous, hence students could pledge zero without
peer pressure. It is also worth mentioning that the students only knew
that the pledges are taken to help the needy among them; however, they
did not know that they are participating in an experiment as well. In
this round, all the students were divided into five treatments. The
details of the all 5 treatments are summarised in Table 1. Each of the
treatments differed from the baseline treatment only in terms of the
provision of additional information.
In the baseline (BL) treatment we distributed the pledge forms
after reading out the written instructions. The students deposited the
pledge form in the collection box. In the second treatment named as
Pledge Disclosed (PD) treatment, the students were provided additional
information about the average amount of pledge per students in the BL
treatment before submitting their pledges. The additional information in
the third treatment, i.e. Need Disclosed (ND) treatment, was about the
total need of the students who had asked for financial help in the first
round. In the fourth treatment, i.e. Pledge and Need Disclosed
(P&ND) treatment, students were asked to make pledges after
providing them with the information about the need as well as the
average amount of pledges per students in all of the previous three
treatments. The fifth and final treatment named as All Disclosed (AD)
treatment was similar to the fourth treatment except that the students
were provided with an additional set of information. The additional
information was about a pledge of 200,000 Pakistani rupees ($2000) by
the Charity Australia International which is an Australian based
charitable organisation.
4. THEORETICAL FRAMEWORK AND HYPOTHESES
Based on the assumption that economic agents are selfish, there are
at least five distinct theories that explain why one contributes to
philanthropic causes. Most of them predict that individual benefits, at
least partially, from the philanthropic contribution. The dynastic model
of family given by Barro (1974) asserts that individuals appear to act
altruistically by passing wealth to their children; however, this
philanthropic act is in reality a lateral shift within the family. Thus,
it assumes that family rather than the individual is the main unit of
analysis. Andreoni (1989) argues that an individual receives a private
good of "warm glow" from an act of philanthropy along with
more of a public good towards which he/she makes donation. Clotfelter
and Steuerje (1981) illustrate that income taxes have a negative effect
on the amount individuals contribute to philanthropy. This implies that
tax deductibility partially counteracts the discouragement caused by the
imposition of that tax. In other words, individuals are willing to make
donations only if the price of giving is low enough. Asheim (1991)
asserts that individuals consume private goods in conjunction with
altruistic giving such that they will only choose a level of private
consumption that is "just". Alternatively, the chosen level
must take into consideration the necessity of altruism. Rose-Ackerman
(1996) has made an argument similar to Andreoni (1989) in the separation
of public and private goods, theorising that individuals give
altruistically in order to receive, as a private good, greater social
capital in return.
In the design of our experiment, students create public good
privately; hence the theories of Andreoni (1989) and Rose-Ackerman
(1996) fit more to our design. (6) As mentioned earlier, these studies
assert that individuals' donations are like a public good; however,
individual donors receive private goods of "warm glow" from
their actions. Those who pledge donation in the design of our experiment
are not direct beneficiaries; but they are likely to receive "warm
glow". Hence, the theories of Andreoni (1989) and Rose-Ackerman
(1996) apply to the design of our experiment and we expected positive
amount of pledges in all of the five treatments. In addition, the
involvement of teachers is also an important factor for the positive
amount of pledges as Rose-Ackerman (1986) noted that when a third party
(especially the department teacher) acts as a monitor; it improves the
information available to donors, making the donors to contribute more.
Next, we discuss the question that how the informational frame
affects the level of pledges in different treatments. For instance, the
informational frame in the ND treatment is such that we informed the
students about the level of the need; however, we did not inform them
about the pledges made until that time to meet the required need. The
total need was about 342,000 Pakistani rupees (About $3420) based on the
applications of 11 needy students that we received in the first round.
Hence, considering the total demand, each student in the ND treatment
might underscore his pledge and instead of pledging high may pledge low.
The studies, based on survey, reveal that when people perceive that
their contribution will not make any difference, they are less likely to
contribute [Radley and Kennedy (1992); Mathur (1996); Diamond and
Gooding Williams (2002); Duncan (2004); Arumi, et al. (2005); Smith and
McSweeney (2007)]. The individuals, who perceive so, believe in the
reasoning of free rider problem [Olson (1965)]. In other words, they
think that an additional dollar does not solve the problem; hence, not
giving does not make things worse.
In the PD and P&ND treatments we expect crowding-in. The
possible justification is that when students see that others give to a
charity; they can take this as a signal that others have confidence in
the organisers or organisation. (7) In particular, in P&ND
treatment, the students can also observe the need; hence, they are more
likely to pledge more. This leadership effect is described earlier by
social psychologists as a 'modelling effect' [Bryan and Test
(1967); Lincoln (1977); Reingen (1982)]. One can argue that the high
level of pledges in the ND treatment might force participants to think
that their pledges are no longer needed. This is a valid argument, but
is less likely in our case, as participants in the PD treatment could
not observe whether the need has been satisfied. And participants of
P&ND treatments could clearly observe that the pledged amount is
less than the need.
In the AD treatment, the students are also likely to show
crowding-in. The major reason is that a matching offer by a third party
"Charity Australia International (CAI)" can have a
legitimising effect. Students are likely to think that the third party
had enough confidence in the organisation. This may increase the
confidence level of the students in the organisation and as a result,
they might pledge more. In a field experiment of a health charity, Van
der Scheer, et al. (1998) found that a signature by a professor in
health care research raised donations by 2.4 percent. Similarly, a lab
experiment found that observing high status donators leads others to
increase their donations. In contrast, the leadership effect was not
found when low status individuals were observed as contributors [Kumru
and Vesterlund (2002)]. The giving by CAI may also increase the
perceived value of giving for students. This is because the students
might see themselves in line with the cause endorsed by a party having
superior information [Vesterlund (2003)].
Based on the above discussions, we test the following hypotheses in
this study:
Hypothesis 1
The pledges per students in the Pledge Disclosed (PD) treatment
will be higher while the pledges per students in the Need Disclosed (ND)
treatment will be lower than the pledges per students in the BL
treatment.
Hypothesis 2
The pledges per students in the Pledge and Need Disclosed
(P&ND) as well as in the All Disclosed treatments (AD) will be
higher than the pledges per students in the BL treatment.
5. RESULTS
In this section, we present the results of our analysis. First, we
will provide an overview of the average pledges in all the treatments.
Next, we will discuss the impact of informational frame on the total
pledges in each treatment relative to the baseline treatment.
5.1. Descriptive Statistics
The descriptive statistics of the charitable pledges across all the
treatments are shown in Table 1. Besides pledges, it also provides the
information regarding the number of observations per treatment and the
average expenditure of students per semester. (8) The second row in the
Table shows the percentage of students in each treatment who have made a
positive pledge. As is evident from the Table, this percentage is higher
in All Disclosed (AD) treatment. Moreover, the results show that the
level of pledges does not increase monotonically as we move from BL to
AD treatment. The average pledge in the Baseline (BL) treatment is
434.46 ($4.34) Pakistan rupees. As we provide information on pledge
requests, there are different effects with different degrees of
information provided. For instance, as the Table shows, the average
pledge size decreases in PD and ND treatments while it increases in
P&ND and AD treatments.
The results are further elaborated in Figures 1 and 2 where the
average pledges across BL, PD, and ND treatments are shown in figure 1
while the average pledges across BL, P&ND, and AD treatments are
shown in Figure 2. It can be observed that the average pledges show a
downward trend in cases of PD and ND treatments relative to the BL
treatment. In contrast, the average pledges in P&ND and AD
treatments show an increasing trend relative to the BL treatment. In
order to find the justification for these results, we provide the
detailed description of each treatment relative to the BL treatment.
5.2. Baseline and Pledge Disclosed Treatments
As stated earlier, in the PD treatment, we disclosed the
information about the average pledge by the participants in the BL
treatment. With the introduction of such a piece of information, the
participation ratio increased. For instance, we find that the percentage
of participants with positive pledge rises from 51.78 percent in the BL
treatment to 61.33 percent in the PD treatment. However, the average
pledge decreases in the PD treatment to 325.07 Pakistan rupees which is
434.46 rupees in the BL treatment. Likewise, the maximum pledge in the
BL treatment is 2688 rupees which decreases to 1344 rupee in the PD
treatment. Similar to the maximum pledge, the minimum level of pledge
also decreases from 100 rupees in the BL treatment to 50 rupees in the
PD treatment. Taking positive pledge of each participants as an
independent observation, Wilcoxon Rank-sum test shows that the
distribution of pledges is not similar across both treatments
(p<0.01).
The fall in the average size of pledge in the PD treatment is
against the hypothesis 1. It is astonishing for us as one might have
expected an increase in the size of pledge. However, it is possible that
the participants in PD treatment might have thought that as others have
already made a reasonable amount of pledge, so why not to free ride on
their pledges. Second, the other justification for the fall in the
average level of pledge in the PD treatment might be the lack of
information with regard to the need for donations. In other words, we
did not provide information regarding the total need for donations with
the pledge request in the PD treatment. (9) Hence, they might have
thought that the pledge made in'the BL treatment might satisfy the
total need. Third, it is also pertinent to mention that we provided
information about the amount of average pledge in the BL treatment but
did not provide the total number of participants that pledged that
amount. In order to justify free riding on the pledges of others, the
students in the PD treatment might have overestimated the total number
of participants in the BL treatment. Finally, the difference in the
sample of students in terms of their incomes across both the treatments
might explain the difference in their average level of pledges. This
conjecture, however, loses ground when we compare the average
expenditure of students across both the treatments. (10) As is evident
from Table 2, the average expenditure of students per semester in the BL
treatment is 32358.6 Pakistani rupees while in the PD treatment, it is
35167 rupees.
A comparison of the distributions of pledges across BL and PD
treatments is shown in Figure 3. As the figure indicates, the number of
larger pledges decreases. For instance, the frequency of pledges below
500 rupees is high in the PD treatment as compared with those in the BL
treatment. In contrast, the number of larger pledges, in particular over
500 rupees, is more in the BL treatment.
[FIGURE 3 OMITTED]
5.3. Baseline and Need Disclosed Treatment
The rate of Participation in the ND treatment increases to 56
percent from 51.78 percent in the BL treatment. However, the inclusion
of additional information in the PD treatment does not seem to have a
beneficial effect on the pledge size as is shown in Figure 4.11 The
Figure indicates that the distribution of the pledge size in the ND
treatment is mostly lower than the distribution of the pledge size for
the corresponding BL treatment. The number of pledges lower than 500
rupees is more in the ND treatment while the number of pledges greater
than 500 rupees is less in the ND treatment. As a consequence, the
average pledge in the ND treatment drops to 341.41 rupees from 434.46
rupees in the BL treatment. This translates into a negative effect of
the additional information in the form of disclosing the need of
deserving students on the average pledge size. For instance, it
decreases the average pledge size by 93.05 rupees. If we take positive
value of individual pledges as independent observations, the Wilcoxon
RankSum test shows that the distribution of pledges in ND treatment is
higher than the distribution of pledges in the BL treatment (p<0.01).
The finding supports hypothesis 2 and substantiates the evidence of
earlier studies. For instance, the earlier research establishes that
when people perceive that their contribution will not make any
difference, they are less likely to contribute [Radley and Kennedy
(1992); Mathur (1996); Diamond and Gooding Williams (2002); Duncan
(2004); Arumi, et al. (2005); Smith and McSweeney (2007)].
[FIGURE 4 OMITTED]
5.4. Baseline and Pledge and Need Disclosed Treatment
In this treatment, we disclose both the earlier pledge and the
total need to the participants. With the introduction of this
information, the number of individuals who make positive pledges
increases. In addition to the increase in the number of pledge makers,
the sizes of the pledges also increase. The average pledge amount is
522.67 Pakistani rupees in the P&ND treatment which is larger than
that of the BL treatment by 88.20 rupees. The comparison of the
distributions of pledges in the P&ND and BL treatments is shown in
Figure 5. It is evident from the Figure that the level of pledge in the
P&ND treatment is mostly skewed to the upper tail of the
distribution. Also, the level of the highest donation, i.e. 2688 rupees,
almost doubles in the P&ND treatment. However, taking positive level
of the individual pledges as independent observations, Wilcoxon Rank-Sum
test does not show that the distribution of pledges across the two
treatments is much different (p=0.58).
[FIGURE 5 OMITTED]
5.5. Baseline and All Disclosed Treatment
In this treatment, all of the previous information was disclosed
with the pledge requests. The results are remarkable both in terms of
the participation rate and in terms of the size of pledges. Namely, both
the participation rate and the size of the pledges increase in the AD
treatment. The participation rate increases from 51.78 percent in the BL
treatment to 74.54 percent in the AD treatment. Likewise, the average
size of pledge increases from 434.46 rupees in the BL treatment to
644.62 rupees in the AD treatment. In the same way, the average donation
as a percentage of the average expenditure increases from 1.34 percent
in the BL treatment to 1.98 percent in the AD treatment. We expected
that the larger number of pledges would be composed mainly of small
pledges. However, Figure 6 shows that the level of pledges in the AD
treatment is larger than the level of pledges in the BL treatment. A
comparison of distributions in Figure 6 also indicates that the absolute
number of small donations actually decreases in the AD treatment. Though
there were some small pledges ranging from 10 rupees to 50 rupees, but
there is also a large pledge of 6000 rupees. This shows that by
disclosing all the information, a fund raiser can better achieve the
target fund. However, the Wilcoxon Rank-Sum test shows that the
distribution of AD treatment is weakly different from the distribution
of BL treatment (p=0.10).
[FIGURE 6 OMITTED]
6. CONCLUSION
This study is motivated by the previous literature that emphasises
the role of information framing on economic behaviour. Here, we analyse
the impact of informational frame on charitable pledges. The study is
based on an experiment in a fund raising project named as Helping
Students through Students. We solicit pledges from 395 students in the
School of Economics, Quaid-i-Azam University, Islamabad. We randomly
divide the students into 5 different treatments. In the PD treatment, we
provide information to students about the average size of pledge made by
students in the BL treatment. Similarly, in the ND treatment, we provide
information about the total need of those who had asked for donations in
the first round of the experiment. In the P&ND treatment, we inform
the students about both the need as well as the pledge made by students
in the BL treatment. Finally, in the AD treatment, we provide the
details about the need, the pledges made in the BL treatment, the
previous history of the project, and the pledge made by the Charity
Australia International.
We find an increase in the average level of pledge with the
informational frame of disclosing all types of information, i.e. AD
treatment. On the other hand, we find the lower size of average pledge
in the treatments where we only provide the pledge in the BL treatment
or the need of the others, i.e. PD and ND treatments. Moreover, we find
substitutability of charitable pledges in the PD treatment while
complementarity of charitable pledges in the P&ND and AD treatments.
Our finding in the ND treatment is in line with the earlier studies
which are based on survey instead of experiments [Arumi, et al. (2005);
Diamond and Kashyap (1997)].
Our results are of interest for fund-raising practitioners. For
instance, our data on pledges show that by providing all information,
the fund-raisers can increase the pledge amount. However, the limitation
of the study is a possible presence of experimenter demand effect. The
reason is that two of the authors are permanent faculty members in the
School of Economics (SOE) where this experiment was conducted. We
suspect the presence of the experimental demand effect due to the
difference between the pledged and received amount in the account of the
trust. It is worth mentioning that the total deposited amount was less
than the pledged amount till the due time given to students and the
final submission of this paper. Future research can explore whether
findings of our experiment will change if campaign is run by aliens in
the SOE or the same experimenters run campaign in other departments
without disclosing their designations. Likewise, it will be interesting
to know whether the amount of pledge changes if participants had to
disclose their identity or if they are informed about the tracking of
their pledges.
APPENDIX A
HELPING STUDENTS THROUGH STUDENTS
INSTRUCTIONS FOR THE RECEIPENTS
We know that there are many students in QAU who need financial help
for pursuing their studies smoothly. Howerver, their ego and
self-respect discourage them to ask for help and support. On the other
hand, there are many students who are willing to help such needy
stdudents, but are unable to find them due to time and information
constraints. In the jargons of economics there exists demand and supply
for help however, market forces are unable to match them efficently.
We (myself and some alumani of QAU) have started a project named
"Helping Students Through Students" to bridge this gap and
link the donors students with the needy students. (12) In this regard we
have registered a Trust named "ROSHNI TRUST" with the
government of Pakistan under 1882 trust act. After formal registration
of the trust and opening an account on the name of ROSHNI TRUST in
Askari Bank QAU branch, we are for the first time launching a compaign
to help students through students. The compaign has two objectives.
(1) Raising fund from students in this semester and transferring
the same fund to needy students in the next semester.
(2) Doing a systemtic analysis of the compaign for research
purposes
(3) In the first round of the compaign we need information about
deserving students. If you think that you need financial support next
semester from this project of "Helping Students Through
Students", then kindly provide us the following details on a plain
page.
* Name
* Father's Name
* CNICNo.
* Email:
* Contact No.
* Per month need in the next semester (Feburary to June 2014)
* Brief details of the background due to which you need financial
support next semester
We assure that the provided information will remain confidential
and never be disclosed. We will try our best that the self-respect of
the students is not compromised. Please send us the required information
on the following address:
Dr Anwar Shah
Assistant Professor, School of Economics, Quaid-i-Azam University,
Islamabad.
You can also submit your sealed enevelop after writing the above
full adress at the front desk of School of Economics. The deadline for
sending your details is Monday 25 November 2013. You will receive a
confirmation email or text once we receive your details.
Note: The final selection will be made by a committee
HELPING STUDENTS THROUGH STUDENTS
INSTRUCTIONS FOR THE DONORS
We know that there are many students in QAU who need financial help
for pursuing their studies smoothly. Howerver, their ego and
self-respect discourage them to ask for help and support. On the other
hand, there are many students who are willing to help such needy
stdudents, but are unable to find them due to time and information
constraints. In the jargons of economics there exists demand and supply
for help however, market forces are unable to match them efficently.
We (myself and some alumani of QAU) have started a project named
"Helping Students Through Students" to bridge this gap and
link the donors students with the needy students. (13) In this regard we
have registered a Trust named "ROSHNI TRUST" with the
government of Pakistan under 1882 trust act. After formal registration
of the trust and opening an account on the name of ROSHNI TRUST in
Askari Bank QAU branch, we are for the first time launching a compaign
to help students through students. The compaign has two objectives.
(1) Raising funds from students in this semester and transferring
the same funds to needy students in the next semester.
(2) Doing a systemtic analysis of the compaign for research
purposes
To day we are running the second round of this compaign. In this
round we ask you for donations in the form of pledges. Once we receive
your pledges, we will add them up and after due scrutiny pass them on to
the deserving students. Priority will be given to the deserving students
from the school of economics. We will appreciate, if you could indicate
a deserving student whom you would like your money to be passed on. All
information regarding your donations and the students whom you would
like to sponsor will not be disclosed in any platform without your prior
approval.
We will urge you to ensure that your pledge amount reaches to the
account of Roshni Trust before 30 December 2013. Please note that your
pledge is completely voluntary hence avoid making a pledge which you
cannot pay by the due date. If the promised amount will not be received
by the due date, we will be unable to start processing the applications
of needy students. Hence you are urged again to make realistic pledge
and transfer it to the account of Roshni Trust before the deadline with
in due time.
Following is the details of the account of ROSHNI TRUST.
Title of Account: Roshni Trust
Account Number: 1500 39000 4256
Bank: Askari Bank Limited
Branch: Quaid-i-Azam University Islamabad
Swift: ASCMPKKA
HELPING STUDENTS THROUGH STUDENTS
A WELFARE PROJECT OF ROSHNI TRUST PLEDGE FORM
Please encircle one option from the list (A to E) given below
(A) I want to sacrifice two cups of tea per day and donate
12x2x(16x7)= 2688
(B) I want to sacrifice one cup of tea per day and donate
12x1x(16x7)= 1344
(C) I want to sacrifice half a cup of tea and donate 6x1x(16x7)=
672
(D) I want to contribute Rs-------------------(write down the
amount)
(E) I wish to contribute; however, my budget constraint is low at
the moment so cannot make a pledge
Note: In option A to C, the amount of pledge has been calculated
assuming 16 weeks of teaching per semester.
Please fill the following brief survey
* Your Gender
Male
Female
* How much is your approximate montly expenditure in the
university: Rs--
* Who supports your above mentioned monthly expenditure?
Parents
Rrelatives
Friends
Others--(Please mention)
* Would you like to disclose your name?
Yes
No
* If yes, then please write your name:--
* Would you like a reminder for submitting your pledge before one
week of the due date
Yes
No
* If Yes please give us your email and contact number:
Email:
Contact No.
* Would you like to name the students whom your donation is to be
transferred?
Yes
No
If Yes, then please mention the name, semester and class of such
student
Name:--
Semester:-- Class: BS/MSs/MPhil
REFERENCES
Abbasi, A. S. (2011) Philanthropic Sector of Pakistan: Performance
and Ripple Effect. World Applied Sciences Journal 12:4, 491-495.
AKDN (2000) Philanthropy in Pakistan: A Report of the Initiative on
Indigenous Philanthropy. The Aga Khan Foundation.
Andreoni, J. (1989) Giving with Impure Altruism: Applications to
Charity and Ricardian Equivalence. The Journal of Political Economy
1447-1458.
Andreoni, J. (1990) Impure Altruism and Donations to Public Goods:
A Theory of Warm-glow Giving. The Economic Journal 100:401, 464-477.
Andreoni, J. (1993) An Experimental Test of the Public-goods
Crowding-out Hypothesis. The American Economic Review 1317-1327.
Andreoni, J. and A. A. Payne (2011) Is Crowding-out Due Entirely to
Fundraising? Evidence from a Panel of Charities. Journal of Public
Economics 95:5, 334-343.
Arumi, A. M., R. Wooden, J. Johnson, S. Farkas, A. Duffett, and A.
Ott (2005) The Charitable Impulse. New York: Public Agenda.
Asheim, G. B. (1991) Unjust Intergenerational Allocations. Journal
of Economic Theory 54:2, 350-371.
Barro, R. J. (1974) Are Government Bonds Net Wealth? Journal of
Political Economy 80:6, 1095-1117.
Bryan, J. H. and M. A. Test (1967) Models and Helping: Naturalistic
Studies in Aiding Behaviour. Journal of Personality and Social
Psychology 6:4, 400-415.
Clotfelter, Charles and E. Stuerle (1981) Charitable Contributions.
In Henry Aaron and Joseph Pechman (eds.) How Taxes Affect Economic
Behaviour. Washington, DC: The Brookings Institution. 403-437.
Diamond, W. D. and S. Gooding-Williams (2002) Using Advertising
Constructs and Methods to Understand Direct Mail Fundraising Appeals.
Nonprofit Management and Leadership 12:3, 225-242.
Duncan, B. (2004) A Theory of Impact Philanthropy. Journal of
Public Economics 88:9, 2159-2180.
Eckel, C. C., P. J. Grossman, et al. (2005) An Experimental Test of
the Crowding Out Hypothesis. Journal of Public Economics 89:8,
1543-1560.
Ghaus-Pasha, A., H. Jamal and M. Asif Iqbal (2002) Dimensions of
the Nonprofit Sector in Pakistan. Social Policy and Development Centre
in collaboration with Aga Khan Foundation (Pakistan) and Centre for
Civil Society. Johns Hopkins University, USA.
Khanna, J., J. Posnett, et al. (1995) Charity Donations in the UK:
New Evidence Based on Panel Data. Journal of Public Economics 56:2,
257-272.
Khanna, J. and T. Sandler (2000) Partners in Giving: The
Crowding-in Effects of UK Government Grants. European Economic Review
44:8, 1543-1556.
Kumru, C. S. and L. Vesterlund (2002) The Effect, of Status on
Voluntary Contribution. New York Times. A13.
Lincoln, A. J. (1977) Effects of the Sex of the Model and Donor on
Donating to Amsterdam Organ Grinders. The Journal of Social Psychology
103:1, 33-37.
List, J. A. and D. Lucking-Reiley (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.
Mathur, A. (1996) Older Adults' Motivations for Gift Giving to
Charitable Organisations: An Exchange Theory Perspective. Psychology and
Marketing 13:1, 107-123.
Nikiforakis, N. (2010) Feedback, Punishment and Cooperation in
Public Good Experiments. Games and Economic Behaviour 68:2, 689-702.
Olson, M. (1965) Logic of Collective Action; Public Goods and the
Theory of Groups. Cambridge: Harvard University Press.
Payne, A. A. (1998) Does the Government Crowd-out Private
Donations? New Evidence from a Sample of Non-profit Firms. Journal of
Public Economics 69:3, 323-345.
Payne, A. A. (2001) Measuring the Effect of Federal Research
Funding on Private Donations at Research Universities: Is Federal
Research Funding More Than a Substitute for Private Donations?
International
Tax and Public Finance 8:5-6, 731-751.
Radley, A. and M. Kennedy (1992) Reflections upon Charitable
Giving: A Comparison of Individuals from Business. 'Manual'
and Professional Backgrounds. Journal of Community and Applied Social
Psychology 2:2, 113-129.
Reingen, P. H. (1982) Test of a List Procedure for Inducing
Compliance with a Request to Donate Money. Journal of Applied Psychology
67:1, 110.
Ribar, D. C. and M. O. Wilhelm (2002) Altruistic and Joy-of-Giving
Motivations in Charitable Behaviour. Journal of Political Economy 110:2,
425-457.
Rose-Ackerman, S. (1996) Altruism, Nonprofits, and Economic Theory.
Journal of Economic Literature 701-728.
Rose-Ackerman, S. (1986) Do Government Grants to Charity Reduce
Private Donations? In S. Rose-Ackerman (ed.) The Economics of Nonprofit
Institutions. New York: Oxford University Press.
Siegel, S. (1956) Nonparametric Statistics for Behavioural
Sciences. New York: McGraw-Hill.
Smith, J. R. and A. McSweeney (2007) Charitable Giving: The
Effectiveness of a Revised Theory of Planned Behaviour Model in
Predicting Donating Intentions and Behaviour. Journal of Community and
Applied Social Psychology 17:5, 363-386.
Steinberg, R. (1991) Does Government Spending Crowd Out Donations?
Annals of Public and Cooperative Economics 62:4, 591-612.
Straub, J. D. (2003) Fund-raising and Crowd-out of Charitable
Contributions: New Evidence from Contributions to Public Radio.
Manuscript, Texas A&M University. Vesterlund, L. (2003) The
Informational Value of Sequential Fundraising. Journal of Public
Economics 87:3, 627-657.
Vriens, M., H. R. Van der Scheer, et al. (1998) Conjoint
Experiments for Direct Mail Response Optimisation. European Journal of
Marketing 32:3/4, 323-339.
(1) http://www.kif.com.pk/, http://www.alfalahss.org/,
http://www.akdn.org/akf_scholarships.asp
(2) Charity Australia International was the organisation which had
offered the financial support before the experiments.
(3) Using a panel of more than 8,000 charities, the study finds
significant crowding out but primarily due to reduced fund-raising.
(4) All instructions are available in the Appendix.
(5) It is pertinent to mention that there are about 4 Notice boards
in the school and the instructions were displayed on all of them.
(6) The creation of a pool from where needy students can benefit is
like a public good. The production of this public good comes through
voluntary contribution. As this public good is run by private
organisation (Roshni Trust), hence, we are of the view that students
create a public good privately.
(7) In this case, the organisation is Roshni Trust which, as stated
earlier, sponsors tuition fees of students in Quaid-i-Azam University,
Islamabad.
(8) Semester in Quaid-i-Azam University generally consists of five
months and there are two semesters per year, i.e. fall and spring.
(9) The total need for donations was collected in the first round
of the experiment through applications from the applicants.
(10) The average expenditure per semester can be regarded as a
proxy of income as the spending pattern or the spending level is a
strong indicator of the level of income.
(11) The additional information was the disclosure of the need of
deserving students from round 1 of the experiment.
(12) My Name is Dr Anwar Shah, Assistant Professor in the School of
Economics QAU.
(13) My Name is Dr Anwar Shah, Assistant Professor in the School of
Economics QAU.
Anwar Shah <
[email protected]> is Assistant Professor,
Department of Economics, Quaid-i-Azam University, Islamabad. Karim Khan
<
[email protected]> is Assistant Professor, Pakistan
Institute of Development Economics, Islamabad. Muhammad Tariq Majeed
<
[email protected]> is Assistant Professor, Department of
Economics, Quaid-i-Azam University, Islamabad.
Authors' Note: The authors are thankful to the editor of PDR
and the two anonymous referees for their useful comments.
Table 1
The List of Experimental Treatments
No. of
Treatment Informational Frame Observations
Baseline (BL) No prior information 112
about pledges or needs
Pledge Disclosed (PD) Prior information about 75
the average pledge
in BL
Need Disclosed (ND) Prior information about
the total demand for 75
financial assistance
Need and Pledge Prior information about 78
Disclosed (P&ND) total demand and the
average pledge
calculated from BL,
PD and ND treatments
All Disclosed (AD) Prior information about 55
the history, pledge
calculated from all
previous treatments,
total demand and
pledge by Charity
Australia International
Table 2
Descriptive Statistics of Charitable Pledges
across Treatments in PKR (Where Approximately 100 PKR=S1)
Pledge Need
Baseline Disclosed Disclosed
(BL) (PD) (ND)
Total number of 112 75 75
observations
% of observations 51.78 61.33 56.00
with positive
pledges
The maximum amount 2688 1344 2688
of positive
pledges
The minimum amount 100 50 50
of positive pledges
Average pledges 434.46 325.07 341.41
per semester
Average expenditure 32358.6 35167 28773
per semester
Average pledges as 1.34 0.92 1.19
a % of average
expenditure
Pledge
and Need All
Disclosed Disclosed
P&ND) (AD)
Total number of 78 55
observations
% of observations 55.13 74.54
with positive
pledges
The maximum amount 2688 6000
of positive
pledges
The minimum amount 100 10
of positive pledges
Average pledges 522.67 644.62
per semester
Average expenditure 27063.5 32471.7
per semester
Average pledges as 1.93 1.98
a % of average
expenditure
Fig. 1. Comparison of Average Pledges
across BL, PD and ND Treatments
Treatment
BL 434.46
PD 325.07
ND 341.41
Note: Table made from bar graph.
Fig. 2. Comparison of Average Pledges
across BL, P&ND and AD Treatments
Treatment
BL 434.46
P&ND 522.67
AD 644.62
Note: Table made from bar graph.
