BY-NC-ND 3.0 license Open Access Published by De Gruyter November 14, 2017

Do Government and Foreign Funding Influence Individual Donations to Religious Nonprofits? A Survey Experiment in Pakistan

Rafeel Wasif and Aseem Prakash
From the journal Nonprofit Policy Forum


Donors support faith-based organizations whose missions and work cohere with their religious values. Might such donors located in the local community be less likely to donate to a hypothetical Madrasa, an Islamic seminary providing K-12 education, which receives funding from non-local sources, specifically the federal government or foreign donors? We examine this question using a survey experiment in Lahore, Pakistan. Because some Madrasa students have indulged in acts of terrorism, the government of Pakistan has offered Madrasas financial assistance to secularize their curriculums. Most Madrasas, however, have refused government funding because they fear it will harm their reputations in the local community for piety and religious training. Alongside, several foreign donors also provide funds to Madrasas, some to aid governmental reform efforts, and others for religious reasons. Based on the survey of 530 respondents, we find that acceptance of government funding does not diminish the respondents’ willingness to donate to the hypothetical Madrasa. Additionally, we find modest evidence that willingness to donate diminishes when the Madrasa accepts money from donors in Saudi Arabia and the United States (but not Germany).

1 Introduction

Nonprofits are important policy actors delivering goods and services, often to underserved populations. Like any other organization, they seek resources to fulfill their organizational missions. In the classic civil society model, nonprofits generate resources locally from their communities (de Tocqueville. 2001; Putnam 1995; Berry 2005). Yet, nonprofits, especially in developing countries, often seek funding from non-local sources, namely the government and overseas donors. This resource generation model begs an important question: are citizens less likely to donate to a nonprofit located in their community if it receives funding from the government or overseas donors?

In addressing the above question, our paper contributes to the literature examining whether non-local funding “crowds out” (or “crowds in”) local donations. Because governments fund nonprofits using tax money, individual donors might believe that they have already fulfilled their civic duty and therefore do not have to provide additional support to nonprofits (Ferris 1984; Brooks 2000; Eckel et al., 2005; Andreoni and Payne 2011; Heutel, 2014). This classic crowding out argument (governmental support funded by taxes replaces direct charity) may not hold in the case of overseas charities, simply because overseas charities do not use citizens’ tax dollars to support nonprofits.

However, citizens might believe that by accepting government or foreign funding, a nonprofit will prioritize non-local preferences over its organizational mission and local preferences (mission drift, in short). This is particularly important if local and nonlocal perspectives clash about what the nonprofit should do and how it should function. Alongside mission drift issues, nonprofits might simply put in less effort to mobilize funds from local donors because they are able to secure resources more easily from the government or overseas donors. And even if nonprofits do not diminish their marketing efforts, local citizens might believe that the nonprofits with external grants are resource rich. Consequently, the marginal impact of their individual donations on nonprofits’ activities will be minimal.

The issue of local support for nonprofits is particularly important in the context of faith-based nonprofit organizations (FBNPOs). Although secularization scholars relegate religion to the realm of private life (Bruce 2011), FBNPOs play a major role in the public sphere because they represent a unique hybrid of religious beliefs and service delivery (Berger 2003; Bielefeld and Cleveland 2013). Policy scholars recognize the potential role of religious organizations in providing local public goods (Clerkin and Grønbjerg 2007; King, 2007; Houston et al. 2008; Lowery 2005). In some highly religious societies of South Asia, such as Pakistan, Bangladesh and Afghanistan, as well Middle Eastern and North African countries such as Egypt, Lebanon, and Jordan, FBNPOs have emerged as key local nongovernmental actors to provide public goods that resource-starved governments typically have failed to adequately provide. In part, FBNPOs are able to overcome the resource constraints by attracting funding from local donors (or volunteers) who want to combine philanthropy (or social action) with religious service.

We focus on a specific category of FBNPOs: Islamic seminaries or religious schools known as Madrasas. These FBNPOs impart K-12 education and train students in Islamic values. Because Madrasas often charge no tuition fees (sometimes even providing free room and board), they tend to attract students predominantly from the poorer sections of society (Nelson 2006). To finance their operations, Madrasas typically rely on donations from the local community, specifically individuals who want to promote Islamic values. Although Madrasas probably do not educate their children, supporting a Madrasa is an act of piety that fulfills these individuals’ religious obligations.

Of course, local donors might be able to support a number of Madrasas. How do they direct their donations? If promoting Islam is the key consideration, local donors must seek assurance that the Madrasa is faithfully delivering Islamic education. This issue resonates with the broader challenge faced by FBNPOs because they often provide a tangible product with an intangible religious component. What motivates individuals to donate to a religious cause is a complex topic, but motivating factors probably include the promise of salvation or a superior afterlife. Yet the donor cannot verify such promises of future benefits; they have features of a credence good (Iannaccone 1998; Clerkin and Grønbjerg 2007).

Of course, some donors might also want the Madrasas to impart technical skills that will allow students to become gainfully employed. This again poses an informational problem for donors. For one, countries vary in how they regulate NGOs (Bloodgood et al. 2014). In Pakistan, Madrasas face no legal requirements to regularly disclose financial information via something akin to the 990 Form. Donors cannot look up a charity watchdog such as Charity Navigator or BBB to assess how well a specific Madrasa is using the donated funds. Furthermore, because of the informal nature of the Madrasa system, most Madrasas do not publish annual reports outlining their activities (Bano 2012). Finally, Madrasa students do not have to appear in any formal standardized exams (say as a legal requirement for the Madrasa to retain its license to operate) that might shed light on the quality of technical educational that the Madrasa imparts. In effect, donors have access to very little hard evidence to judge whether the Madrasa is functioning as per its organizational mission, imparting true Islamic education, or equipping students with employable skills.

This is where organizational reputations come in. Reputations are judgements about a specific issue (Islamic teaching, technical skills) held by specific actors (local community) about other actors (Madrasas). Organizations can have multiple reputations on different dimensions even among the same set of actors. Reputations can substitute for hard evidence. How Madrasas acquire and maintain their reputation for imparting true Islamic education is a complex issue. It is plausible that the financial support a Madrasa receives from non-local actors such as the government or overseas donors could influence its reputation for imparting Islamic education. Local donors might believe outside funding causes Madrasa to ignore its mission. This is because non-local donors probably do not care about, or may be even hostile to, the Islamic component of the educational product that the Madras offers. Indeed, Madrasa officials we interviewed attribute their reluctance to accept government funding to this concern: the local community believes that by accepting government funding, a Madrasa will downplay Islamic teachings and education. Arguably, this concern might hold when the Madrasa receives funding from non-Muslim and Western overseas donors. Yet, receiving money from overseas Islamic donors might strengthen the Madrasas’ reputation for piety. Thus, it is not clear how outside funding per se shapes the perceptions of local donors about the Madrasa’s commitment to Islamic education. The identity of the non-local donor is important because it might explain their motivations for philanthropy, and how their funding strengthens or weakens the Islamic component of Madrasa’s educational product.

External funding for Madrasas has becoming an important issue in the so-called “war on terror.” Madrasas are controversial because they have produced students that have indulged in acts of terrorism (Abuza 2003). To reduce Madrasas’ potential as breeding grounds for radical youth, the Government of Pakistan (allegedly under US pressure) has initiated a reform program that offers financial assistance to Madrasas that secularize their curriculum and imbue them with marketable skills (Fair 2008). Overseas donors from Islamic and western countries have also stepped up their funding to Madrasas, although with different motivations.

How might such non-local funding influence the willingness of individual donors to support Madrasas? They might perceive the government’s or western donors’ attempts to secularize the curriculum negatively, because it potentially undermines the religious training and Islamic values the Madrasas seeks to impart. While Pakistan is allied with the United States in the “war on terror”, drone strikes and other issues have generated considerable suspicion among the ordinary Pakistanis, if not hostility, towards the United States. [1] While they may disapprove of the US, they may approve of funding from a European country such as Germany that has generously accepted Muslim refugees and challenged the US on some aspects of the war on terror. In contrast to governmental or Western funding, financial support from overseas Saudi Arabian charities should reaffirm the piety of the Madrasas, and therefore crowd-in local support.

Yet the perceptions of individual donors towards external funding are probably also shaped by ongoing political events. Because Pakistan has suffered large scale terrorist violence, [2] some individuals might oppose Saudi Arabian funding because they might link it with the radicalization of the Madrasas. Some in the local community might want the Madrasas to impart marketable skills, and therefore view government funding (and western funding) as a useful tool to encourage Madrasas in this direction.

To explore how non-local funding might influence an individual’s willingness to donate to a hypothetical local Madrasa, the authors contracted with Gallup-Pakistan for a survey experiment in Lahore, Pakistan in October 2016. Given the exploratory nature of this study, we worked with a sample of 530 respondents in randomly selected neighborhoods of Lahore. We randomly assigned respondents to one of the five frames: one control frame and four experimental frames.

We find that the individual respondents’ willingness to donate is not influenced even when Madrasa accepts money from the federal government. This is an important finding because most Madrasas have declined governmental funding. Interviews with officials from ten Madrasas and with heads of Ittehad-e-Tanzimat-e-Madaris (the union of Madrasa boards), which comprises representatives from five Madrasa boards representing different Islamic sects, suggest that Madrasa officials believe that government funding will undermine their reputation for religious piety because the government has started this program primarily under U.S. pressure. [3]

Despite the high level of public support for Saudi Arabia (95 % approval), [4] we find modest evidence (p value of 0.09 in the full logistic regression model) that individuals are less willing to donate to Madrasas that accept money from overseas Saudi donors. In part, this might reflect the public’s desire not to support militancy and terrorism that Saudi funding is often accused of fomenting. We also find modest support that individuals are also less willing to donate to Madrasas that receive funds from the United States (p value of 0.06 in the full logistic regression model), a country that is viewed as hostile to Pakistan and to the Islamic world in general. [5] But respondents do not transfer this hostility to all western donors: funding from Germany does not have a statistically significant effect on individual willingness to donate to Madrasas in any of the models.

The paper has five sections. Section two reviews the literature on FBNPOs and how accepting governmental or foreign funding can crowd-out its local support. Section three places Madrasas and efforts to reform them in a historical context and develops hypotheses. Section four provides the research design and methods and section five offers findings and results. Section six concludes with implications for policy and future research.

2 Reputations, Information Problems, and Faith-Based Nonprofits

Nonprofits are policy actors that supposed to provide goods and services in the face of both market and government failures (Weisbrod 1977). Scholars suggest that nonprofits’ unique organizational form, the non-distributional constraint, makes them trustworthy (Hansmann 1980). Others attribute nonprofits’ trustworthiness to organizational culture and ideological disposition; for them, nonprofits are motivated by principles and not instrumental concerns (Keck and Sikkink 1998; Mitchell and Schmidt, 2014 but see Cooley 2002; Prakash and Gugerty, 2010). But going beyond principles, when nonprofits rely on the local community for financial support, they have incentives to be attentive to citizens’ expectations.

Theoretically, the above discussion can be cast in terms of information problems, agency theory, and signaling. Nonprofits can be conceptualized as agents working on behalf of their donors (but see Van Slyke 2007). Nonprofit agree to undertake specific tasks on behalf of these principals: providing K-12 education in our case. As in other agency relationships (Berle 1933), information asymmetries between principals and agents can lead to agency slippages (Tremblay-Boire and Prakash 2016). These asymmetries can manifest in many ways: from financial mismanagement to mission drift where the nonprofit engages in activities that that are at odds with its organizational mission.

Information asymmetries are particularly acute for nonprofits for three reasons. First, individual donors may not have the resources and expertise to monitor how nonprofits deploy their funds (though watchdogs such as Charity Navigator are important innovations in this regard but limited to developed countries). Second, these monitoring challenges are accentuated because nonprofits operate in institutional settings that do not necessarily compel them (or the regulations are not enforced) to disclose financial information on their finances (via say, form 990). Information problems are also magnified by the nature of the “products” they tend to supply, an issue that much of the nonprofit literature has focused on (Young 2000). Nonprofits’ customers tend to belong to the vulnerable sections of society; indeed Madrasa often attract students from the economically disadvantaged sections of society. These customers typically cannot vote with their feet (or dollars), or even voice their disapproval (Johnson and Prakash 2007).

The issue becomes complicated when nonprofits have multiple donors, and the preferences of these donors on what the nonprofit should do and function do not converge. This is particularly important in the context of Madrasas that often seek resources from the local community, and yet the same time, often accept resources from the federal government and different types of international donors. Given the information problems about Madrasa operations and effectiveness, what sorts of clues might local donors look for when deciding about their charitable giving? Because individual donors may find it difficult to assess the quality of product, organizational performance or commitment to their mission, they look for proxies such as the nonprofit’s reputation for fulfilling its organizational mission. Nonprofits, therefore, seek to protect, or sometimes enhance, their reputations through advertising and other sorts of publicity efforts (Stride and Lee 2007; Boenigk and Becker 2016; Willems et al. 2016). They may seek to signal their virtues by the personnel they recruit, the activities they undertake, and testimonials from their “customers.” FBNPOs, in particular, are sensitive to the issues of reputation and credibility, given the additional credence dimension of their products (Stolz 2014).

From a public policy perspective, FBNPOs are an important part of the nonprofit population. By some accounts, an increasing number of nonprofits have started to define themselves in religious terms (Berger 2006). The role of FBNPOs in development discourse is increasingly recognized because they tend to have stronger roots in the communities they serve and can draw on social capital to support their activities. Similarly, FBNPOs are uniquely situated at defining development goals in religious terms as many religions emphasize caring for the poor. Some suggest that religious beliefs incentivize individuals to become more philanthropic (Berger 2006); in deeply religious societies of South Asia and the Middle East, it is difficult to separate religion and philanthropy (Kandil 2004). Yet, FBNPOs are criticized because they can sometimes be selective in choosing their beneficiaries and use their activities to proselytize to their faith (Flanigan 2007; Taniguchi and Thomas 2011). There is also a darker side to some FBNPOs. Radical groups such as Hamas started as FBNPOs focusing on social charity.

Nonprofits might seek to raise resources from multiple sources (Keating and Frumkin 2003; Reddick and Ponomariov 2013). If the nonprofit is able to secure resources from non-local sources such as the government or overseas donors, how might this influence the willingness of local donors to support it? The “crowding-out” literature identifies several mechanisms connecting government funding with individual philanthropy. Individuals probably want to support charities but believe that they are already doing so through the taxes they pay to the government (Ferris 1984; Brooks 2000; Eckel et al., 2005; Andreoni and Payne 2011; Heutel, 2014).

Citizens might also believe that government or foreign monies will incentivize nonprofits to prioritize non-local preferences over local preferences. In the international context, scholars suggest that foreign funding for local NPOs undermines their local support. This is because the local community perceives such funding as undermining NPOs’ downward accountability to the local community by prioritizing foreign preferences over local concerns (Chahim and Prakash 2014; Dupuy et al. 2015; Jakimow 2010).

Of course, resource availability might make nonprofits less motivated to expend effort mobilizing funds from local donors funds (Dokko 2008; Andreoni and Payne 2011; Kim and Van Ryzin 2014). Here the idea is that resource scarcity leads nonprofits to work harder to raise funds from private individuals while resource abundance via government support does the opposite.

Finally, local citizens might believe that the nonprofits simply do not need their money and it makes no sense for them to support rich charities (Roberts 1984; Warr 1982; Andreoni, 1990). This sort of an argument is noted by social dilemma scholars whereby individuals are less likely to incur private cost to provide a public good if they believe that their contributions do not affect the overall quality or level of the public good provision. Thus, donors seek to support organization that is doing good work but faces resources problems. They want to see their monies “making a difference,” the marginal impact of their donations on the effectiveness of the nonprofit decreases as the nonprofit relies increasingly on government (or external) support.

3 The Madrasa Sector of Pakistan

Madrasas or Islamic religious seminaries are ubiquitous in Pakistan (Tavernise 2009).The popularity of Madrasas could be understood from both a demand and supply side. Madrasas address a fundamental problem of government and market failure: an unmet demand for affordable K-12 education. Further, in a highly religious society, they promise to impart Islamic teachings and train students in Islamic scriptures. Madrasas are funded mostly by the community in which they are located. For these donors, supporting a Madrasa is almost a religious duty and an act of piety. From the supply side, the Madrasa sector virtually has no entry barriers; any preacher can start a Madrasa. There is no system, public or private, of certification, audit, or governance. Not surprisingly, the level of competition for donations in the Madrasa sector is fierce and Madrasas seek ways to differentiate themselves from their competitors.

Given the widespread presence of Madrasas, and the importance of religious parties and organization in Pakistan’s politics, various governments have unsuccessfully sought to exercise some sort of control over Madrasas and their curriculums. Indeed, the so-called Madrasa reforms have been on the Pakistani government’s agenda since independence in 1947. Among the earlier attempts was the establishment of the Dars-e-Nizami Committee. Its recommendations asked for “unnecessary religious subjects to be removed from the syllabus … to broaden the worldview of the Madrasa students and to increase their mental horizons.” Similarly, the education policy in 1970 was the government’s second attempt for Madrasa reform and asked for introduction of modern subjects in the Madrasa syllabus to produce Madrassa graduates capable of meeting the demands of science and industry. In 1979 another committee called the Jaiza Deeni Madrasas (Review of Religious Schools) was formed. It also asked for the introduction of modern subjects and establishment of a Central Government Board or Waqf to administer Madrasa exams (Bano 2008). Similar concerns about the Madrasa reforms were also shown by General Musharraf’s government before 9/11. Pakistan’s interior minister at the time Moinuddin Haider announced a reform plan that would require all Madrasas to register with the government, expand their curricula, disclose their financial resources, seek permission for admitting foreign students, and stop sending students to militant training camps. However, it also did not prove a success as most of the Madrasas refused to accept it (Stern, 2000). They were not willing to expand their curricula under the government’s aegis or be forced to disclose their financial figures. All the attempts at Madrasa reforms failed in these sense that there was little follow up concrete action in the face of pushback from religious parties and organizations.

9/11 was some sort of a game changer in Madrasa politics because Madrasas began getting linked with Taliban and terrorism more generally. The need to secularize Madrasas became an international security issue. U.S. Secretary of State Colin Powell characterized Madrasas as breeding grounds for “fundamentalists and terrorists.” [6] Defense Secretary Rumsfeld asked in a memorandum: “Are we capturing, killing or deterring and dissuading more terrorists every day than the madrassas and the radical clerics are recruiting, training and deploying against us? [7]“ In a bipartisan appeal, Senators Collins and Schumer noted the role of Saudi Arabia in promoting Madrasas: “We know that Saudi funded madrassas promote religious intolerance and violence in schools … … Despite supposed reform efforts, reports indicate that Saudi funded schools and mosques continue to teach hatred and preach violence around the world.” [8]

Consequent to this tremendous international (especially American) pressure, the Musharraf government introduced a new set of Madrasa reforms in 2002–03. During the 2002–03 Madrasa reforms, the government offered financial incentives to Madrasas to accept the reforms which included provision of teachers to Madrasas at the primary, secondary and higher secondary levels of education. The government also incentivized the Madrasas with provision of stationaries, computer and other accessories for the Madrasas. In return, however, the Madrasas were asked to introduce secular subjects in their curriculum, as well as remove all radical religious material. They also needed to provide the financial reports which were to be audited by government representatives, as well as provide the government list of all their funders. Additionally, they could not take any funding or students from abroad without explicit government permission. By most accounts, these reform efforts also failed because only 432 madrasas accepted the reform. [9]

Why are Madrasas rejecting the reform package? They understand that the government aims to secularize their curriculum to ensure that they are not breeding grounds for radicals. In fact, they are also offered funds by international donors (US and Germany) who want to modernize Madrasas with the same objectives. However, some Madrasas also receive funding from Saudi Arabia that promotes conservative Wahhabi Islam and this funding might make the curriculum more religious. These external funding options pose a problem for Madrasas. On the one hand, they want resources to enroll larger number of students and improve their facilities and infrastructure. This is particularly alluring because they are mostly underfunded and would welcome any external assistance. Yet they are also aware that if this information gets leaked out through the rumor mill, the local community may not view this funding in favorable light. The local donors might feel that external money will lead to mission drift, and the that Madrasas are working to satisfy the external donors and deemphasizing Islamic teachings. Given that there is no guarantee of perpetual external funding, Madrasas probably will try to safeguard their local donor support by protecting their reputations for piety and devotion to Islamic education.

This perspective was supported in authors’ interviews with the Madrasa officials. [10] Most of them believe that government’s attempts to secularize the curriculum through offer of funding will harm the reputation of their nonprofit. The logic is that the local community is suspicious about government’s intention, and perceives that this policy was adopted under western pressure in the “war against terror”. Consequently, the local community will see government funding as eroding Madrasa’s autonomy. They fear that with a secularized curriculum, Madrasas will deemphasize religious training and Islamic values. The chairman of the Ittehad-e-Tanzimat-e-Madaris (Madrasas’ Union) argued that government funding “will make us look bad if we received it.” Similarly, a Madrasa official noted that “the community will note that Madrasas are taking money from the government. This will impact the Madrasa’s credibility.” An official from another Madrasas was concerned that taking “government funding will impact their autonomy as they will have to listen to the government if they take its money.” Yet, we note that a small number of Madrasas have accepted government funding. Arguably, they might believe that the local community wants their children to learn marketable skills in order to succeed in the world, and government funding is a useful way to encourage Madrasas to impart such skills. Based on the above discussion, this paper proposes the following:

Hypothesis 1: Willingness to donate will be diminished for Madrasas that have accepted funds from the government.

Alongside governmental support, Madrasas also receive funds from overseas donors located in Islamic countries. Some citizens might view this as a reassuring testimonial that the Madrasa is serving the cause of Islamic education. But this funding could arguably diminish the image as well. For one, local community may not appreciate that their local Madrasa is engaged in spreading radicalism and not training students to acquire real-life skills. Indeed a 2015 Pew poll suggested that Pakistani’s have a very unfavorable view of Taliban with 76 % disapproving of it. [11] In particular, the role of Saudi Arabian donors in fostering radicalism is often noted in the media. Thus, this article proposes:

Hypothesis 2: Willingness to donate will be diminished if the Madrasa receives funding from an overseas Saudi Arabian charity.

Madrasas might also receive funds from overseas donors from Western countries. For some, this might be a positive signal that the Madrasa is serious about imparting technical and real-life skills. But on the other hand, this funding might undermine its reputation for providing a truly Islamic education. After all, “the war on terror” and the ongoing refugee crisis in Europe seems to have pit western powers, especially the United States, against Islamic countries. Some might suspect that western donors are working on behalf of the CIA and other western intelligence organizations that are perceived to be hostile to Islamic interests (Yusuf 2011). The CIA employed a local health charity involved in polio vaccination program to gather information about Osama bin Laden. Further, Raymond Davis, a CIA contractor, killed two Pakistanis in Lahore. Therefore, this paper proposes the following:

Hypothesis 3: Willingness to donate will be diminished if the Madrasa receives funding from an overseas American charity.

While the approval rating of all western countries remains below 50 % in Pakistan, the EU is viewed more favorably than the US (35 % versus 12 %) (Braghiroli and Salini 2014). On one hand German donors might be viewed differently from American donors because Germany has been most generous western country in terms of admitting predominantly Muslim refugees, yet, respondents may not differentiate between different western powers and view them as working together against the Islamic world. This leads to the next hypothesis:

Hypothesis 4: Willingness to donate is enhanced if the Madrasa receives funding from an overseas German charity.

4 Research Design and Methods

In recent years, social science scholars have shown considerable interest in using experimental techniques (Barabas and Jerit 2010; Hyde 2015; Grimmelikhuijsen et al. 2017). Public administration scholars have employed survey experiments to understand the perceptions and beliefs of policy actors (Karens et al. 2016; Grimmelikhuijsen et al. 2017). Dietrich (2016) tests how donor agency officials from different countries respond to corruption and inefficient state institutions in recipient countries. Andersen and Jakobsen (2017) test how bureaucrats react to communication cues and find that communication frames and cues that align policies with norms are more likely to move bureaucrats’ policy positions in favor of the policies.

In the nonprofit field, scholars have employed survey experiments to generate insights about donor motivations. Shang (2008) explores the role of social influences on individual giving and finds that informing individuals of how much another unnamed person had given increases donations. Karlan and List (2007) report that information on matching grant (when a donor commits to match the contributions of other donors up to a certain amount) leads to increased donations from individuals. Schlesinger et al. (2004) suggest that donors may be incentivized to fund nonprofits that are transparent about their ownership structure. Similarly, Bekkers (2010) has studied the effects of material, social, and psychological incentives for giving and volunteering. Hoogervorst et al. (2016) report that donors are more likely to trust volunteers as opposed to professional employees when making decisions to donate to a nonprofit. Panic et al. (2016) examine how different types of celebrity endorsements influence donor motivation in an online setting. Kim and Van Ryzin (2014) examine the behavioral aspect of the crowding-out hypothesis, through an online survey experiment to test if government funding of a hypothetical nonprofit would influence donations.

This article contributes to the public administration literature on charitable giving and altruism. It employs a survey experiment to assess motivations for altruism in the context of a FBNPO that receives funding from multiple sources, governmental and non-governmental, and also seek to raise resources from the local community. It uses a hypothetical situation in which the donors actually do not give money but express their intentions to do so. Arguably, it is debatable the extent to which intention to donate to a hypothetical organization reflects actual donations that these individual might make. While this is true, there is some work that suggests that survey experiments can generate accurate insights about real behavior (Hainmueller et al. 2015; McDermott 2002). This study is unique, perhaps the first one, because it focuses on the willingness to donate to a category of highly controversial FBNPO, the Madrasas. Further, it seeks to study altruism towards Madrasas in arguably one of the most volatile regions of the world, Pakistan. Given the substantial investments to “reform” Madrasas by the Pakistan government, western governments, and overseas private foundations, this project has important public policy implications.

After seeking permission from the university’s Institutional Review Board (Human Subjects), the authors commissioned Gallup-Pakistan to conduct the survey in Lahore. This survey was conducted face-to-face. The authors chose Lahore because it is the second most populous city of Pakistan with a population of about 9 million. Lahore is also the capital of the Punjab province that accounts for over half of Pakistan’s population. While religious radicalism has spread across Pakistan, Lahore is relatively peaceful with no major ethnic or sectarian conflicts. This means that respondents are likely to provide more truthful answers than say in some other city such as Karachi, Quetta or Peshawar where sectarian and ethnic conflicts can possibly create serious challenges in conducting surveys. We recognize that this limits the generalizability of our findings to some other cities in Pakistan. While recognizing that it is difficult to find a “representative” city in Pakistan to explore our research question, we decided to focus on Lahore primarily because of safety issues and the desire to ensure that we get truthful answer from the respondents. Recognizing the above concerns, we also decided that the survey should avoid any references to issues that might instigate sectarian loyalties. This allows the authors to analyze how individuals are willing to donate to a religious NGO irrespective of its sectarian (Shia versus Sunni, specifically) affiliations. The ground reality is that a respondent is likely to be disinclined to support a Madrasa from another sect. Therefore, when faced with a neutral Madrasa name with no obvious sectarian connotations, respondents are more likely to assume that the Madrasa belongs to their own sect.

Does Lahore reflects the broader Madrasas universe say in terms of size, acceptance of foreign funding, or support from overseas donors? This is a very difficult issue because there is no census of Madrasas in Pakistan. Further, the legal system governing Madrasas is very weak and nonprofits are not expected to file an annual report to the government, something analogous to Form 990. Yet, in spite of the above data shortcomings, Lahore can be considered as representative city as it is home to Madrasas from different denominations including Deobandi, Barelvi, Jamaat-i-Islami, Shia and Ahl-e-Hadith.

How representative is our sample of Pakistan? We suggest that our sample of 530 respondents is fairly representative in terms of standard demographic variables such as education level, marital status, number of children in a household, gender and age. According to the most recent population census in 1998, [12] the male-female ratio in the country is 1.08–1 which is well-reflected by 50 % of our respondents being women (Population Census 1998) Overall, sample is more educated than overall literacy rate of Pakistan (70 % versus 53 %). This is expected because Lahore is a highly literate city with about 82 % level (Pakistan Social And Living Standards Measurement 2014–15). For variables on political preferences, country wide data are not available. For political affiliation, the sample matches with the support PML-N (the ruling party) received in the province of Punjab (Election Commission of Pakistan 2016). The survey was conducted in October 2016. Surveyors contacted respondents in markets across a random sample of Lahore neighborhoods.

We also gave considerable attention to the language of the survey. There are two major languages spoken in Lahore: Punjabi and Urdu. The survey was conducted in Urdu because this is the national language of Pakistan and does not have a provincial and sectarian connotation. Consequently, the authors believe that all respondents residing in Lahore, irrespective of whether they are Punjabis or not, should have felt comfortable in responding to our questions. The questionnaire was translated into Urdu (and reverse translated to check for accuracy) and pre-tested. The pre-test revealed that the questionnaire was clearly worded and only minor changes were necessary. The final questionnaire (control group and 4 treatment groups) was administered to a random sample of 530 respondents. [13]

The dependent variable is the respondents’ willingness to donate Rs. 1000 as expressed on a five point Likert Scale. This is akin to a “referendum method,” where respondents are presented with a hypothetical referendum that specifies a single “price” for a good, and asked for their opinions (Greenlee et al. 2007). Why Rs. 1000? To the best of authors’ knowledge, there is no systematic data on the average levels of individual donations to religious nonprofits in Pakistan. The authors therefore discussed this subject with nonprofit functionaries in Pakistan, and concluded that Rs. 1000 is an appropriate “price point.” For reference, Rs. 1000 approximates a donation of $10, a reasonable amount in a country with per capita income at about $1300. At the same time, given the more peaceful and non-sectarian conditions in Lahore, the authors are more confident that the willingness to donate is not going to be impacted by social desirability bias. This means that respondents are more likely to respond to reveal their preferences in Lahore than in an area with sectarian tension where people will be less likely to reveal their real preferences. [14]

In terms of survey logistics, as summarized in Table 1 below, we randomly assigned respondents to one of the five frames (the control frame, plus four treatment frames). Thus, the respondents are not assessing making donation to one type of Madrasa as oppose to some other type; this is not forced choice approach. Rather, we describe a hypothetical Madrasa with specific attributes and ask about respondents’ willingness to pay. We then compare the willingness to pay across frames, and fit a regression to estimate the marginal effects of specific variables on the willingness to pay.

Table 1:

Experimental design.

FrameAccepts government FundingAccepts funding from Saudi ArabiaAccepts funding from USAAccepts funding from Germany
Control groupNoNoNoNo
Treatment group 1YesNoNoNo
Treatment group 2NoYesNoNo
Treatment group 3NoNoYesNo
Treatment group 4NoNoNoYes

The control frame provides a description of the Madrasa and does not mention any external source of funding. Government funding, and different sources of international funding are introduced in the treatment frames. All respondents were randomly assigned to one of the five frames and the Balance Table is provided in Table 2

Table 2:

Balance table: Means of demographic covariates.

Control frameGovernmentSaudi ArabiaUSAGermany
Urban/Rural ratio1.
Financial contribution to Madrasa1.811.791.841.781.83
Family members in Madrasa2.662.622.742.702.78
Extended family in Madrasa2.782.812.712.802.96
Ever enrolled in Madrasa?3.773.793.803.743.74
Political party support1.562.042.101.862.19
View regarding military operation against militants1.0561.
Total household members3.804.214.103.813.92
Total household income5.565.685.655.585.56
Employment sector3.393.543.273.373.21
Ever lived outside pakistan?1.971.992.001.962.00
Highest education degree2.012.022.932.882.81
Marital status1.871.851.871.821.89

After seeking their responses regarding their willingness to pay, the interviewers collected (non-identifiable) information on demographic variables such as their gender, occupation, income, number of children in the household, political leanings, past support for religious causes, international travel, international stay, trust in the government, and their opinion of political issues that have bearing on their perceptions of politics and religion. Regressions estimated the marginal effects of specific variables on the respondents’ willingness to donate.

Recognizing that the underlying mechanism behind the funding or lack of trust of government or foreign donors may be difficult to measure directly simply because respondents are unlikely to reveal their real distrust about the government or the foreign donors, the survey included three indirect questions. These are: support for current ruling party (PML-N), support of military operations against the militants, as well as the amount of time someone has spent abroad with the intuition that those who have traveled abroad are more likely to have more favorable opinions about countries abroad. Table 3 presents the survey instrument.

Table 3:

Survey questionnaire.

Stage 1: Consent
This research is being conducted by AUTHORS to understand what motivates individuals to give donations to Madrasas.
Your participation is voluntary. Participation involves responding to a 10–15 minute survey. You may choose not to answer some or all of the questions.
Information collected from this survey will be kept confidential. Names and other identifying information will not be used in any report, presentations, or paper.
Questions about this survey can be directed to: AUTHORS
Respondents must say “Yes” before they are given the survey. If they say “No,” the survey will end.
Stage 2: Intervention
Pakistan was established in 1947 to provide a homeland to Muslims. Over several decades to educate the children of Pakistan in Islamic values, many organizations have established Madrasas that provide religious education (K-12) and teach real world skills. Most Madrasas are free for students, while some charge a nominal fee. They accept children from all backgrounds, but particularly cater to poorer segments of the society. There is a popular belief that Madrasas encourage sectarianism and terrorism and their curriculum needs to be changed. Madrasas vary in how they finance their operations. Some completely rely on local support, some accept funding from the government, while others are funded by overseas charities, both Islamic and non-Islamic.
Control Frame
Madrasa Fahm-ul-Quran (Understanding Quran) has approached you for a donation of Rs. 1000. On a scale of 1–5 (1=not likely, 5=likely), what is the likelihood of you donating the said amount to Madrasa Fahm-ul-Quran. For reference, this Madrasa does not accept government funding. It also does not accept funding from overseas charities.
Treatment Group 1
Madrasa Fahm-ul-Quran (Understanding Quran) has approached you for a donation of Rs 1,000. On a scale of 1–5 (1=not likely, 5=likely), what is the likelihood of you donating the said amount to Madrasa Fahm-ul-Quran. For reference, in prior years, Fahm-ul-Quran has accepted funding from the Pakistani government for Madrasa reforms.
Treatment Group 2
Madrasa Fahm-ul-Quran (Understanding Quran) has approached you for a donation of Rs 1,000. On a scale of 1–5 (1=not likely, 5=likely), what is the likelihood of you donating the said amount to Madrasa Fahm-ul-Quran. For reference, in prior years, Fahm-ul-Quran has received funding from the Muslim Charity for Education in Saudi Arabia.
Treatment Group 3
Madrasa Fahm-ul-Quran (Understanding Quran) has approached you for a donation of Rs 1,000. On a scale of 1–5 (1=not likely, 5=likely), what is the likelihood of you donating the said amount to Madrasa Fahm-ul-Quran. For reference, in prior years, Fahm-ul-Quran has received funding from the Education for All charity in Washington DC, USA.
Treatment Group 4
Madrasa Fahm-ul-Quran (Understanding Quran) has approached you for a donation of Rs 1,000. On a scale of 1–5 (1=not likely, 5=likely), what is the likelihood of you donating the said amount to Madrasa Fahm-ul-Quran . For reference, in prior years, Fahm-ul-Quran has received funding from the Education for All charity in Berlin, Germany.
Follow-up Questions
Have you made a financial contribution to a madrasa in the past 12 months?
- Yes
- No
- Don’t Know
- Prefer to Not Respond
Does anyone in your immediate family (son, daughter, brother, sister, wife or husband) study or have studied in a madrasa?
- Yes, but not at present.
- Yes, they are currently enrolled.
- No
- Prefer to Not Respond
Does anyone in your extended family (uncle, cousin, relative) study or have studied in a madrasa?
- Yes, but not at present.
- Yes, they are currently enrolled.
- No
- Prefer to Not Respond
Have you ever enrolled in a madrasa in the past, or do you intend to enroll in a madrasa in the future?
- Yes, I was enrolled in the past, but I am not enrolled at present.
- Yes, I am currently enrolled.
- Yes, I want to get enrolled in the future.
- No, I have not enrolled, and do not plan to enroll in the future.
Which party best serves the interests of Pakistan?
- Pakistan Muslim League(N)
- Pakistan Tehreek-e-Insaf
- Pakistan Peoples Party
- Others: Please Mention Name
What are your views regarding the Pakistan Army’s current operation against militants in FATA and Punjab?
- It was necessary
- It was unnecessary
Demographic Questions:
How many people, including yourself, live in your household?
What is your total income per month in Pakistani Rupees from all sources before taxes?
- 4,000–20,000
- 20,000–100,000
- 100,000–200,000
- Above 200,000
What sector do you work in?
- Business
- Agriculture
- Service Industry
- Informal Sector
- Other (Please Explain)
Have you ever lived outside Pakistan?
- Yes
- No
If yes, for how long did you live outside Pakistan?
- Less than 1 year
- 1–5 years
- 5–10 years
- More than 10 years (Please mention the amount of years)
Which foreign region have you lived in?
- Europe
- US
- Canada
- Australia
- Middle East
- Other (Please mention the country)
What is your highest education level achieved?
- No education.
- High School.
- Technical/Vocational Training
- Undergraduate Education (Bachelors)
- Graduate Education (Masters)
- Doctoral Education (PhD)
Do you have any children?
- Yes
- No
What is your gender?
- Male
- Female
What is your marital status?
- Single
- Married
- Widowed/Divorced
Which of the below describes your ethnicity? (You can mark more than one.)
- Punjabi Baloch
- Sindhi
- Pashtun
- Others (Please mention the ethnicity that you self-identify with).

5 Findings and Results

The analysis was conducted in two stages. In the first stage, exploiting the experimental survey design, the authors looked at the mean willingness to donate across frames. Mann-Whitney U-tests, the nonparametric equivalent of difference-of-means tests (independent t-tests) for data that is not normally distributed, confirm these initial findings. We conducted a two-tail test to assess whether there is any difference in the mean willingness to donate to a Madrasa between the control group and any of the treatment groups. As reported in Table 4, accepting government funding does not undermine the willingness to donate to our hypothetical Madrasa (Hypothesis 1). This is an important finding given that the explicit purpose of the Madrasa reform program was to secularize the curriculum and the program has had a very poor response: by some accounts, less than1 % among the thousands of Madrasas have accepted funding. Madrasas officials believe that accepting government funding would cast their FBNPO in poor light in terms of their commitment to Islamic values. Yet, while citizens may not support government’s efforts with their own resources, they are certainly not punishing Madrasas for accepting funds from government (Hypothesis 1). Arguably, one could infer that these citizens do not oppose or support government’s efforts to secularize the curriculum.

Table 4:

Mann-Whitney U-test results.

Saudi Arabia0.0969

While funding from Saudi Arabia does not enhance the respondents’ willingness to pay, there is weak support that funding from Saudi donors (Hypothesis 2) is associated with a lower willingness to pay (p=0.097). This is significant because citizens of Pakistan hold Saudi Arabia in high esteem. Yet, Pakistan has faced several terrorist and sectarian attacks and 98 % of Pakistanis believe that terrorism is a big problem. [15] Indeed, tackling terrorism has become the most important priority for the current government (PML-N) as well as in the previous Zardari government (PPP). The military has launched a major campaign (Operation Zarb-e-Azb) in FATA, North Waziristan, Balochistan, South Punjab, as well as in Karachi. Because Saudi funding is associated with radical groups, respondents probably have a negative perception of Madrasa receiving Saudi funding. We do not find support that funding from the US (Hypothesis 3) or German donors (Hypothesis 4) have a statistically significant influence on the willingness to donate.

Cohen’s d (Cohen 1988) statistics provide additional support for the effect size (the mean difference between the control and treatment group in standard deviation units) significance of the Saudi Arabian frame and the lack of significant effect size for the government frame (0.063). While the effect size for the US and German frames are also below 0.2 (0.126, 0.164), the effect size for the Saudi Arabian frame (0.202), while modest, is above the threshold of 0.2.

To explore the marginal effects of different factors in influencing the willingness to donate to our hypothetical Madrasa, the paper first employs a binary logistic regression estimator, as reported in Table 5, and then as robustness check, an ordered probit estimator as reported in Table 6. For the former, the authors coded the willingness to pay variable into two categories, ‘‘likely’’ (scores of 3 and above on the Likert scale) and ‘‘unlikely’ to donate. Model 1 includes variables reflecting different resource providers (the Pakistani Government, Saudi Arabia, US, and Germany). Model 2 includes demographic indicators such as gender, age, location, salary, job type, location (urban or rural), marriage status, ethnicity, and education level. Model 3 includes questions that pertain to political views, religious experiences such as support for PML-N (the ruling party), support for military operations against militants, previous enrollment in madrasas, whether any of their family members attends a madrasa, or whether they have made any financial contributions to madrasas in the past 12 months. Model 4, the full model, includes all the variables in the regression.

Table 5:

Binary logistic regression full sample (Question: Imagine that Madrasa Fahm-ul-Quran contacted you with a request for a donation of Rs. 1000. How likely are you willing to donate money) .

Dependent variable:
Willingness to Donate
Saudi Arabia−0.495*−0.592**−0.493*−0.544*
Madrasa donation in last 12 months No0.708**0.531
Madrasa donation in last 12 months –Don’t Know−0.00020.002
Madrasa donation in last 12 months -Prefer to not respond−0.131−0.164
Family members in Madrasa- currently enrolled1.324*1.373*
Family members in Madrasa- not currently enrolled0.0150.002
Family members in Madrasa -prefer to not respond−0.063−0.146
Extended family in Madrasa-No0.6050.961
Enrolled in Madrasa -Yes I want to get enrolled in the future−0.130−0.061
Enrolled in Madrasa- No−0.207−0.044
Political party- Pakistan Tehreek-e-Insaf0.0930.059
Political party-Pakistan peoples party0.6430.573
Political party -none29.26431.470
Political party - refused to answer11.92711.940
Political party -dont know13.40313.887
Military operation against Taliban- unnecessary13.27813.735
Military operation against Taliban- refused to answer13.76814.403
Military operation against Taliban -dont know14.93815.486
Highest education l0.3820.284
Not lived outside Pakistan0.3410.111
Service industry−0.530*−0.505*
Informal sector−0.419−0.311
Rs 5001 – Rs 11 0000.5820.550
Rs 11 001 – Rs 14 000−0.130−0.261
Rs 14 001 – Rs 16 000−0.772−0.252
Rs 16 001 – Rs 20 000−0.012−0.194
Rs 20 001 – Rs 37 0000.6930.544
More than Rs 37 000−13.820−14.500
Log likelihood−362.765−329.815−333.764−301.971
Akaike Inf. Crit.735.530705.631711.529683.942

  1. Note: *p<0.1; **p<0.05; ***p<0.01

Table 6:

Ordered probit model full sample dependent variable measured on a 5 point Likert Scale.

Dependent variable:
Willingness to Donate
Saudi Arabia−0.367**−0.400**−0.319*−0.334*
Madrasa donation in last 12 months No−0.031−0.202
Madrasa donation in last 12 months -dont know−0.007−0.008
Madrasa donation in last 12 months -prefer to not respond0.1010.189
Family members in Madrasa- currently enrolled1.070**1.168**
Family members in Madrasa- not currently enrolled0.0250.020
Family members in Madrasa -prefer to Not Respond0.1450.027
Extended family in Madrasa-No0.4690.594
Enrolled in Madrasa -Yes I want to get enrolled in the future−0.149−0.198
Enrolled in Madrasa- No0.0260.083
Political party- Pakistan Tehreek-e-Insaf0.1340.114
Political party-Pakistan peoples party0.4060.421
Political party -none9.53510.126
Political party - refused to answer3.3462.889
Political party -dont know4.9634.973
Military operation against Taliban- unnecessary4.8814.910
Military operation against Taliban- refused to answer4.9285.036
Military operation against Taliban -dont know5.7155.755
Highest education level0.3440.465
Not lived outside Pakistan−0.191−0.393*
Service industry−0.054−0.010
Informal sector−0.717*−0.713
Rs 5001 – Rs 11 0000.9280.953
Rs 11 001 – Rs 14 000−0.178−0.150
Rs 14 001 – Rs 16 000−1.272*−1.240*
Rs 16 001 – Rs 20 0000.5690.708
Rs 20 001 – Rs 37 0000.718**0.794**
More than Rs 37 0004.6415.112
Log likelihood−330.864−308.653−306.243−284.251
Akaike Inf. Crit.671.729663.307656.487648.502

  1. Note: *p<0.1; **p<0.05; ***p<0.01

Across all models, government funding has no statistically significant impact on the willingness to support (Hypothesis 1). This is an important finding and this is contrary to the perceptions of Madrasas officials we interviewed who believe that receiving government funding will undermine their reputation for piety and excellence in religious teaching.

Willingness to donate in response to Saudi funding is negative (Hypothesis 2) but with weak level of statistical significance (p value=0.09 in the full model). Similarly, willingness to donate in response to US funding is also negative (Hypothesis 3) and again with weak statistical significance (p value=0.06 in the full model). Importantly, in any model, willingness to donate is not affected by information on German support (Hypothesis 4). These results are similar across different specifications.

While not of our central interest, we find interesting results in some control variables. Interestingly, party affiliation, and income levels are not statistically significant predictors of the willingness to donate. Age is a significant predictor of willingness to donate, as is enrollment of donors’ immediate family in a Madrasa. Gender (women) and marital status (widowed/single) shows a statistically significant association with willingness to donate. This is consistent with the broader literature that reveals that women tend to be more charitable than men (Mesch 2010) and married individuals (both men and women) tend to be more charitable than unmarried ones (Rooney et al. 2005).

The authors re-ran the regressions using an ordered probit model where the dependent variable is measured on five-point Likert scale. As presented in Table 5 below, our results hold for the government frame: willingness to donate to the government is not significant. Our results also hold for the willingness to donate to the Saudi Arabian frame: the coefficient is negative and with weak level of statistical significance (0.09). However, willingness to donate is not significant in the case of US or Germany.

6 Conclusion

Does government and foreign funding undermine local support for Madrasas? While there is a widely held perception among Madrasa officials that government funding does so, this paper suggests that Madrasa officials have probably misread the preferences of the local community in this regard. While this is an exploratory study limited to one city in Pakistan only and looking at only willingness to donate to a hypothetical Madrasa, this finding potentially signals important policy implications. If accepting government funding does not crowd out individual donors’ willingness to donate to Madrasas, the poor response to government’s Madrasa reform program was probably rooted in an incorrect understanding Madrasas officials have about the communities they serve. Madrasa officials might not appreciate that even highly religious citizens would like Madrasas to impart marketable skills and do not support the idea that religious seminaries have become the breeding grounds for terrorism and militancy. Yet citizens might be less willing to speak out openly on these subjects given the threat of violent retaliation by the militants. This article suggests that governments need to appreciate the sentiments of the silent majority and develop innovative strategies to combat militancy and deradicalize Madrasas.

And they should take care that they are not perceived as doing so under foreign pressure.

In some ways, the tremendous public support for and the lionization by the Pakistani press of the recently retired military chief, Raheel Sharif, who oversaw the military operations against the militants, supports our finding.

Yet, our paper also offers cautionary advice on external funding of Madrasas. While local population might not disapprove of government funding, they may not like the idea of foreign donors, specifically American and Saudi Arabian, funding. Because America has credibility crisis in Pakistan, American support, governmental or non-governmental, creates suspicion and undermines credibility. The policy implication is that any form of American intervention to support the K-12 education sector may not achieve the desired results. Our paper did not test whether respondents differentiate between receiving support from the US government (via say USAID) as opposed to support from US Foundation. While our sense is that all type of American intervention breeds suspicion (the use of polio workers in the hunt for bin-Laden did not enhance confidence about nonprofits), this is an issue that future research should examine.

Of course, one might wonder as to why any Madrasas accepted government reform package and to what extent these Madrasas also accepted foreign monies? What explains their cost-benefit calculus regarding external donations, and how does this differ from the ones that have declined the reform package? We sought to discern some trends through our interview work but could not find any systematic pattern based on observable attributes (such as size of the Madrasa or its longevity). Future work should explore this issue in a systematic way; specifically, find ways to create a centralized database on Madrasas. There are periodic statements about foreign donations but these provide information at the aggregate level and seem to be contradictory. For example, while a 2015 Punjab Police (Lahore is the capital of Punjab) reports that 1,000 out of 12,000 Madrasas they tracked received funded from foreign sources (Gishkori 2015), the Minister for State for Interior reports only 23 Madrasas were being funded from overseas (Haider 2015) Neither Punjab police nor the Minister identified which specific Madrasas received foreign funding. A systematic database could have enormous policy value in this regard.

Our paper also suggests that the level of radicalism in the local community is much lower than what media reports tend to suggest. After all, donor support never increases when the Madrasa received monies from Saudi Arabia. In fact, in some models, donor support diminished in this regard. This is very surprising given that Saudi Arabia has very high approval ratings in public opinion polls. This again suggests that the level of radicalism in the local population is perhaps less than what is reported in the media. The policy challenge is to find nuanced way to mobilize the silent (and terrified) majority without the appearance that these policies are an outcome of American coercion. This sort of nuanced diplomatic policy is the need of the hour.

One might suggest an alternative explanation for the Saudi Arabian frame: arguably, this signals that the Madrasas is rich and does not require its financial support from the local community. If this were true, funding from rich Western countries should have had a comparable effect of diminishing individuals’ willingness to donate; after all, western countries are affluent as well. Yet, the willingness to donate is not diminished when the Madrasa receives funding from Germany. This suggests that the local community does not perceive international funding influencing the Madrasas in the same way. It even differentiates among different western donors. Again, while appreciating the exploratory nature of this study, we believe that this paper points to the need for a more nuanced understanding about the educational needs of the local community, and their desire to see their children acquiring real world skills during their studies in Madrasas.

Finally, the issue whether the finding about German could be generalized to all European countries. We decided on Germany (as opposed to the United Kingdom or France) because Germany’s response to the refugee crisis differentiates it from the US, and arguably from other European countries as well. Consequently, it should probably be held in higher esteem in the Muslim world in relation to other European powers. Future research could rest how potential donors might differentiate among different European powers; for example, given that the UK is closely aligned with the US in Afghanistan, whether Germany is held in higher esteem in relation to the UK. This has important policy implications in terms of western intervention in K12 education in Pakistan. If indeed the objective is to help modernize Madrasas, then arguably a country such as Germany with some level of goodwill should take the lead, either in terms of governmental or private efforts.

Appendix Section

Appendix 1: Descriptive statistics

NMeanStandard DeviationMinimumMaximum
Overall willingness to donate [scale 1–5]5302.361.2015
Number of individuals in household5303.972.06220
Number of boys in household5301.321.0006
Number of girls in household5301.221.0805
Highest education level achieved5302.537.27198
Any children?4402.991.5208

Appendix 2: Overall Willingness to Donate

Scale of likelihood to donateNumber of respondents

Appendix 3: Cross-Table of Willingness to Donate by Various Frames

(Willingness to donate is on y-axis while treatment group is on x-axis. Willingness to donate is coded 1 for scale 3–5 and 0 for scale 1–2.)

Treatment group
Willingness to donateControlGovernmentSaudi ArabiaUSAGermanyTotal

Appendix 4: Cross-Table of Willingness to Donate by Various Groups

(Willingness to donate is on y-axis while donation to Madrasas in last 12 months is on x-axis. Willingness to donate is from scale 1–5.)

Previous donations in last 12 months
Willingness to donateControlGovernmentSaudi ArabiaUSAGermanyTotal

Appendix 5: Cross-Table of Willingness to Donate (Binary 0–1) and self-reported previous donations to a Madrasa in Last 12 Months

Previous donations in last 12 months
Willingness to donateYesNoDon’t knowPrefer to not respondTotal

Appendix 6: Cross-Table of Willingness to Donate (Scale 1–5) and self-reported previous donation to a Madrasa in the Last 12 Months

Previous donations in last 12 months
Willingness to donateYesNoDon’t knowPrefer to not respondTotal


Abuza, Zachary. 2003. Militant Islam in Southeast Asia: Crucible of Terror. Boulder: Lynne Rienner Publishers. Search in Google Scholar

Andersen, Simon Calmar, and Morten Jakobsen. 2017. “Policy Positions of Bureaucrats at the Front Lines: Are They Susceptible to Strategic Communication?” Public Administration Review 77 (1):57–66. Search in Google Scholar

Andreoni, James. 1990. “Impure Altruism and Donations to Public Goods: A Theory of Warm – Glow Giving?” Economic Journal 100 (401):464–77. Search in Google Scholar

Andreoni, James, and A. Abigail Payne. 2011. “Is Crowding out Due Entirely to Fundraising? Evidence from a Panel of Charities.” Journal of Public Economics 95 (5):334–343.10.1016/j.jpubeco.2010.11.011 Search in Google Scholar

Bano, Masooda. 2008. “Dangerous Correlations: Aid’s Impact on NGOs’ Performance and Ability to Mobilize Members in Pakistan.” World Development 36 (11):2297–2313.10.1016/j.worlddev.2007.11.001 Search in Google Scholar

Bano, Masooda. 2012. The Rational Believer Choices and Decisions in the Madrasas of Pakistan. Ithaca: Cornell University Press. Search in Google Scholar

Barabas, Jason, and Jennifer Jerit. 2010. “Are Survey Experiments Externally Valid?.” American Political Science Review 104 (2):226–242.10.1017/S0003055410000092 Search in Google Scholar

Bekkers, René. 2010. “Who Gives What and When? A Scenario Study of Intentions to Give Time and Money.” Social Science Research 39 (3):369–381.10.1016/j.ssresearch.2009.08.008 Search in Google Scholar

Berger, Ida. 2006. “The Influence of Religion on Philanthropy in Canada.” Voluntas: International Journal of Voluntary and Nonprofit Organizations 17 (2):110–127.10.1007/s11266-006-9007-3 Search in Google Scholar

Berger, Julia. 2003. “Religious Nongovernmental Organizations: An Exploratory Analysis.” Voluntas: International Journal of Voluntary and Nonprofit Organizations 14 (1):15–39.10.1023/A:1022988804887 Search in Google Scholar

Berle, Adolf A., Jr. 1933. The Modern Corporation and Private Property. New York: Macmillan. Search in Google Scholar

Berry, Jeffrey M. 2005. “Nonprofits and Civic Engagement.” Public Administration Review 65 (5):568–578.10.1111/j.1540-6210.2005.00484.x Search in Google Scholar

Bielefeld, Wolfgang, and William Suhs Cleveland. 2013. “Defining Faith-Based Organizations and Understanding Them Through Research.” Nonprofit and Voluntary Sector Quarterly 42 (3):442–467.10.1177/0899764013484090 Search in Google Scholar

Bloodgood, E. A., J. Tremblay-Boire, and A. Prakash. 2014. “National styles of NGO Regulation.” Nonprofit and Voluntary Sector Quarterly 43 (4):716–736.10.1177/0899764013481111 Search in Google Scholar

Boenigk, S., and A. Becker. 2016. “Toward the Importance of Nonprofit Brand Equity.” Nonprofit Management and Leadership 27 (2):181–198.10.1002/nml.21233 Search in Google Scholar

Braghiroli, Stefano, and Luca Salini. 2014. “How Do the Others See Us? An Analysis of Public Opinion Perceptions of the EU and USA in Third Countries.” Working Paper. Transworld. Search in Google Scholar

Brooks, Arthur C. 2000. “Is There a Dark Side to Government Supportfor Nonprofits?.” Public Administration Review 60 (3):211–218.10.1111/0033-3352.00081 Search in Google Scholar

Bruce, Steve. 2011. Describing Secularization. Oxford, New York: Oxford University Press. Search in Google Scholar

Chahim, Dean, and Aseem Prakash. 2014. “NGOization, Foreign Funding, and the Nicaraguan Civil Society.” VOLUNTAS 25 (2):487–513.10.1007/s11266-012-9348-z Search in Google Scholar

Clerkin, Richard M., and Kirsten A. Grønbjerg. 2007. “The Capacities and Challenges of Faith-Based Human Service Organizations.” Public Administration Review 67 (1):115–126.10.1111/j.1540-6210.2006.00701.x Search in Google Scholar

Cohen, Jacob. 1988. Statistical Power Analysis for the Behavioral Sciences, 2nd ed. Hillsdale, NJ: Routledge. Search in Google Scholar

Cooley, Alexander. 2002. “The NGO Scramble: Organizational Insecurity and the Political Economy of Transnational Action.” International Security International Security 27 (1):5–39.10.1162/016228802320231217 Search in Google Scholar

de Tocqueville., Alexis. 2001. Democracy in America. New York, N.Y.: Signet Classic. Search in Google Scholar

Dietrich, Simone. 2016. “Donor Political Economies and the Pursuit of Aid Effectiveness.” International Organization 70 (1):65–102. Search in Google Scholar

Dokko, Jane. 2008. “Does the NEA Crowd Out Private Charitable Contributions to the Arts?” SSRN Scholarly Paper ID 1327111. Rochester, NY: Social Science Research Network. Search in Google Scholar

Dupuy, Kendra E., James Ron, and Aseem Prakash. 2015. “Who Survived? Ethiopia’s Regulatory Crackdown on Foreign-Funded NGOs.” Review of International Political Economy 22 (2):419–456.10.1080/09692290.2014.903854 Search in Google Scholar

Eckel, C. C., P. J. Grossman, and R. M. Johnston. 2005. “An Experimental Test of the Crowding out Hypothesis.” Journal of Public Economics, The Experimental Approaches to Public Economics 89 (8):1543–1560. Search in Google Scholar

“Election Commission of Pakistan.” 2016. Accessed November 11. Search in Google Scholar

Fair, C. Christine. 2008. The Madrassah Challenge: Militancy and Religious Education in Pakistan. Washington, DC: United States Institute of Peace Press. Search in Google Scholar

Ferris, James M. 1984. “Coprovision: Citizen Time and Money Donations in Public Service Provision.” Public Administration Review 44 (4):324–333.10.2307/976078 Search in Google Scholar

Flanigan, Shawn Teresa. 2007. “Paying for God’s Work: A Rights-Based Examination of Faith-Based NGOs in Romania.” Voluntas 18 (2):156–175.10.1007/s11266-007-9035-7 Search in Google Scholar

General Elections 2013. n.d. Islamabad: Election Comission of Pakistan. Accessed October 17, 2017. Search in Google Scholar

Greenlee, Janet, Mary Fischer, Teresa Gordon, and Elizabeth Keating. 2007. “An Investigation of Fraud in Nonprofit Organizations: Occurrences and Deterrents.” Nonprofit and Voluntary Sector Quarterly 36 (4):676–694.10.1177/0899764007300407 Search in Google Scholar

Gishkori, Zahid. 2015. “Punjab Police Disclosure: 1,000 Madrassas Foreign – Funded.” The Express Tribune, February 17, 2015. Search in Google Scholar

Grimmelikhuijsen, Stephan, Sebastian Jilke, Asmus Leth Olsen, and Lars Tummers. 2017. “Behavioral Public Administration: Combining Insights from Public Administration and Psychology.” Public Administration Review 77 (1):45–56.10.1111/puar.12609 Search in Google Scholar

Haider, Irfan. 2015. “Saudi Arabia Denies Funding ‘Extremist Mindset’ in Pakistan.” DAWN.COM, February 9, 2015. Search in Google Scholar

Hainmueller, Jens, Dominik Hangartner, and Teppei Yamamoto. 2015. “Validating Vignette and Conjoint Survey Experiments against Real-World Behavior.” Proceedings of the National Academy of Sciences 112 (8):2395–2400.10.1073/pnas.1416587112 Search in Google Scholar

Hansmann, Henry B. 1980. “The Role of Nonprofit Enterprise.” The Yale Law Journal 89 (5):835–901. Search in Google Scholar

Heutel, Garth. 2014. “Crowding Out and Crowding In of Private Donations and Government Grants.” Public Finance Review 42 (2):143–175. Search in Google Scholar

Hoogervorst, Niek, Judith Metz, Lonneke Roza, and Van Baren. Eva. 2016. “How Perceptions of Altruism and Sincerity Affect Client Trust in Volunteers Versus Paid Workers.” Nonprofit and Voluntary Sector Quarterly 45 (3):593–611.10.1177/0899764015597778 Search in Google Scholar

Houston, David J., Patricia K. Freeman, and David L. Feldman. 2008. “How Naked Is the Public Square? Religion, Public Service, and Implications for Public Administration.” Public Administration Review 68 (3):428–444.10.1111/j.1540-6210.2008.00880.x Search in Google Scholar

Hyde, Susan D. 2015. “Experiments in International Relations: Lab, Survey, and Field.” Annual Review of Political Science 18:403–424. doi:10.1146/annurev-polisci-020614-094854. Search in Google Scholar

Iannaccone, Lr. 1998. “Introduction to the Economics of Religion.” Journal Of Economic Literature 36 (3):1465–1495. Search in Google Scholar

Jakimow, Tanya. 2010. “Negotiating the Boundaries of Voluntarism: Values in the Indian NGO Sector.” VOLUNTAS 21 (4):546–568.10.1007/s11266-010-9154-4 Search in Google Scholar

Johnson, E., and A. Prakash. 2007. “NGO Research program: A Collective Action Perspective.” Policy Sciences 40 (3):221–240.10.1007/s11077-007-9043-x Search in Google Scholar

Kandil, Amani. 2004. “Civic Service in the Arab Region.” Nonprofit and Voluntary Sector Quarterly 33 (4):39S–50S. Search in Google Scholar

Karens, René, Jasper Eshuis, Erik-Hans Klijn, and Joris Voets. 2016. “The Impact of Public Branding: An Experimental Study on the Effects of Branding Policy on Citizen Trust.” Public Administration Review 76 (3):486–494. Search in Google Scholar

Karlan, Dean, and John A. List. 2007. “Does Price Matter in Charitable Giving? Evidence from a Large-Scale Natural Field Experiment.” American Economic Review 97 (5):1774–1793.10.1257/aer.97.5.1774 Search in Google Scholar

Keating, Elizabeth K., and Peter Frumkin. 2003. “Reengineering Nonprofit Financial Accountability: Toward a More Reliable Foundation for Regulation.” Public Administration Review 63 (1):3–15.10.1111/1540-6210.00260 Search in Google Scholar

Keck, Margaret E., and Kathryn Sikkink. 1998. Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell University Press. Search in Google Scholar

Kim, Mirae, and Gregg G. Van Ryzin. 2014. “Impact of Government Funding on Donations to Arts Organizations.” Nonprofit and Voluntary Sector Quarterly 43 (5):910–925.10.1177/0899764013487800 Search in Google Scholar

King, Stephen M. 2007. “Religion, Spirituality, and the Workplace: Challenges for Public Administration.” Public Administration Review 67 (1):103–114.10.1111/j.1540-6210.2006.00700.x Search in Google Scholar

Lowery, Daniel. 2005. “Self-Reflexivity: A Place for Religion and Spirituality in Public Administration.” Public Administration Review 65 (3):324–334.10.1111/j.1540-6210.2005.00457.x Search in Google Scholar

McDermott, Rose. 2002. “Experimental Methodology in Political Science.” Political Analysis 10 (4):325–342.10.1093/pan/10.4.325 Search in Google Scholar

Mitchell, George E., and Hans Peter Schmitz. 2014. “Principled Instrumentalism: A Theory of Transnational NGO Behaviour.” Review of International Studies 40 (3):487–504. Search in Google Scholar

Nelson, Matthew. 2006. “Muslims, Markets, and the Meaning of ‘A Good Education’ in Pakistan.” Asian Survey 46 (5):699–720.10.1525/as.2006.46.5.699 Search in Google Scholar

“Pakistan Social And Living Standards Measurement 2014–15.” 2015. Pakistan Bureau of Statistics. Search in Google Scholar

Panic, Katarina, Liselot Hudders, and Veroline Cauberghe. 2016. “Fundraising in an Interactive Online Environment.” Nonprofit and Voluntary Sector Quarterly 45 (2):333–350.10.1177/0899764015587715 Search in Google Scholar

“Population Census 1998.” Islamabad: Pakistan Bureau of Statistics. Accessed October 17, 2017. Search in Google Scholar

Prakash, Aseem, and Mary Kay Gugerty. 2010. “Trust but Verify? Voluntary Regulation Programs in the Nonprofit Sector.” Regulation & Governance 4 (1):22–47. Search in Google Scholar

Putnam, Robert. 1995. “Bowling Alone: America’s Declining Social Capital.” Journal of Democracy 6 (1):65–78.10.1353/jod.1995.0002 Search in Google Scholar

Reddick, Christopher, and Branco Ponomariov. 2013. “The Effect of Individuals’ Organization Affiliation on Their Internet Donations.” Nonprofit and Voluntary Sector Quarterly 42 (6):1197–1223.10.1177/0899764012452670 Search in Google Scholar

Roberts, Russell. 1984. “A Positive Model of Private Charity and Public Transfers.” Journal of Political Economy 92 (1):136–148.10.1086/261212 Search in Google Scholar

Rooney, P. M., D. J. Mesch, W. Chin, and K. S. Steinberg. 2005. “The Effects of Race, Gender, and Survey Methodologies on Giving in the US.” Economics Letters 86 (2):173–180.10.1016/j.econlet.2004.06.015 Search in Google Scholar

Schlesinger, Mark, Shannon Mitchell, and Bradford H. Gray. 2004. “Restoring Public Legitimacy to the Nonprofit Sector: A Survey Experiment Using Descriptions of Nonprofit Ownership.” Nonprofit and Voluntary Sector Quarterly 33 (4):673–710. Search in Google Scholar

Shang, Yue. 2008. “The Effects of Social Information, Social Norms and Social Identity on Giving.” ProQuest Dissertations Publishing. Search in Google Scholar

Stern, Jessica. 2000. “Pakistan’s Jihad Culture | Foreign Affairs.” Search in Google Scholar

Stolz, Dr Joerg. 2014. Religions as Brands: New Perspectives on the Marketization of Religion and Spirituality. Farnham, Surrey: Ashgate Publishing, Ltd Search in Google Scholar

Stride, Helen, and Stephen Lee. 2007. “No Logo? No Way. Branding in the Non-Profit Sector.” Journal of Marketing Management 23 (1–2):107–122.10.1362/026725707X178585 Search in Google Scholar

Taniguchi, Hiromi, and Leonard Thomas. 2011. “The Influences of Religious Attitudes on Volunteering.” VOLUNTAS 22 (2):335–355.10.1007/s11266-010-9158-0 Search in Google Scholar

Tavernise, Sabrina. 2009. “Pakistan’s Islamic Schools Fill Void, but Fuel Militancy.” The New York Times, May 3, sec. Asia Pacific. Search in Google Scholar

Tremblay-Boire, Joannie, and Aseem Prakash. 2016. “Will You Trust Me?: How Individual American Donors Respond to Informational Signals Regarding Local and Global Humanitarian Charities.” VOLUNTAS. 28 (2):621–647. Search in Google Scholar

Van Slyke, David M. 2007. “Agents or Stewards: Using Theory to Understand the Government-Nonprofit Social Service Contracting Relationship.” Journal of Public Administration Research and Theory: 17 (2):157–187. Search in Google Scholar

Warr, Peter. 1982. “Pareto Optimal Redistribution and Private Charity.” Journal of Public Economics 19 (1):131–138.10.1016/0047-2727(82)90056-1 Search in Google Scholar

Weisbrod, Burton Allen. 1977. The Voluntary Nonprofit Sector: An Economic Analysis. Lexington, Mass.: Lexington Books. Search in Google Scholar

Willems, J., M. Jegers, and L. Faulk. 2016. “Organizational effectiveness reputation in the nonprofit sector.” Public Performance & Management Review 39 (2):454–475.10.1080/15309576.2015.1108802 Search in Google Scholar

Young, Dennis. 2000. “Alternative Models of Government-Nonprofit Sector Relations: Theoretical and International Perspectives.” Nonprofit and Voluntary Sector Quarterly 29 (1):149–172.10.1177/0899764000291009 Search in Google Scholar

Yusuf, Huma. 2011. “Conspiracy Fever: The US, Pakistan and Its Media.” Survival 53 (4):95–118. Search in Google Scholar

Published Online: 2017-11-14
Published in Print: 2017-12-20

© 2017 Walter de Gruyter GmbH, Berlin/Boston

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.