Journal of Management Vol. 39 No. 2, February 2013 496-530 DOI: 10.1177/0149206311398135 © The Author(s) 2011 Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

Geographic and Product Diversification in Charitable Organizations Geoffrey M. Kistruck The Ohio State University

Israr Qureshi Hong Kong Polytechnic University

Paul W. Beamish The University of Western Ontario

The diversification of firms into new geographic and product markets has received a tremendous amount of attention in the field of management. Charities, on the other hand, have garnered attention among management scholars primarily as social influencers on multinational corporations. However, over the past half-century, charities have become significant diversified entities in and of themselves. At their core, both charitable and for-profit organizations are bundles of routines that struggle to deal with administrative issues related to institutional isomorphism, organizational slack, and resource dependency. However, from a contextual standpoint, one of the primary mechanisms by which the diversification of charities impacts efficiency is the transaction costs associated with seeking out and maintaining external resources from donors in addition to charities’ internal set of capabilities and routines. Using panel data involving 3,616 charities over a five-year period, the authors’ findings suggest that while the main relationship between geographic diversification and efficiency is U shaped in nature, the main relationship between product diversification and efficiency is inverted U shaped. From an interaction perspective, the authors’ results also suggest that while charities that maintain lower levels of product diversification follow a similar U-shaped curve as they increasingly diversify into new regions, this curve is inverted for charities that are engaged in unrelated

Acknowledgements: This article was accepted under the editorship of Talya N. Bauer. Corresponding author: Geoffrey M. Kistruck, Fisher College of Business, The Ohio State University, Columbus, OH 43221, USA. Email: [email protected]

496

Kistruck et al. / Diversification in Charitable Organizations   497

types of product diversification. Therefore, the study suggests that the extent to which one type of diversification produces positive or negative efficiencies depends significantly on the level of the other type of diversification. Such findings have theoretical implications for both the charitable and for-profit sectors. Keywords:  diversification; internationalization; charities; nonprofit

While multinational corporations may struggle to identify new geographic and product market needs in seeking out profitable growth opportunities, charitable organizations have no such shortage of unmet social needs in regions throughout the world (Drayton, 2005). Rather, the engagement of charities in diversified activities to pare the already prodigious pile of global social needs, such as those created by poverty, illiteracy, and environmental degradation, is constrained primarily by their ability to garner donor and volunteer resources (Froelich, 1999). However, heightened competition for funding has caused such resources to become increasingly scarce and costly to acquire (Scott, 2003). Historically, donors have focused almost exclusively on measures of social effectiveness when evaluating the performance of charitable organizations. However, increased donor skepticism surrounding the percentage of their funds being spent directly on charitable activities, as opposed to administrative overhead and fund-raising, has led to an increased focus on financial efficiency as an important indicator of performance (Bennett & Savani, 2003; Moore, 2000). Recent studies have indicated that charities perceived as inefficient tend to experience lower future financial funding from stakeholders who are no longer willing to support such organizations (Thornton, 2006). Thus, the strategic focus of charities is being forced to shift away from unchecked growth and more toward maintaining a level of geographic and product scope that reflects an efficient allocation of resources as judged by the stakeholders upon whom the organization relies (Fox & Brown, 1998). Unfortunately, there has been very little theoretical or empirical work undertaken to explore how the increasingly heterogeneous set of activities within charitable organizations impacts their levels of efficiency. The relationship between efficiency and the diversification of organizations into new geographic and product markets has previously received a tremendous amount of scholarly attention in the field of for-profit management (Hoskisson & Hitt, 1990; Werner, 2002). Such research has highlighted the potential benefits of geographic diversification such as the exploitation of firm-specific resources (Bartlett & Ghoshal, 1989), achievement of greater economies of scale and scope (Kobrin, 1991; Kogut, 1985), and exploitation of location advantages (Dunning, 1988). Similarly, scholars have identified potential costs associated with diversification related to the liability of foreignness (Hymer, 1976), the liability of newness (Stinchcombe, 1965), and the increased requirement for mechanisms of strategic control (Doz & Prahalad, 1991). Thus, for-profit management scholars have focused primarily on the internal capabilities of firms to exploit these opportunities and mitigate these costs as the primary mechanisms by which diversification impacts efficiency (Chatterjee & Wernerfelt, 1991). There are many underlying theoretical similarities between for-profit firms and charities as organizational entities. At their core, both can be considered bundles of routines that struggle

498    Journal of Management / February 2013

to deal with administrative issues related to institutional isomorphism, organizational slack, and resource dependency (Brody, 1996; Kistruck & Beamish, 2010; Nelson & Winter, 1982; Useem, 1987). However, from a contextual standpoint, one of the primary mechanisms by which the diversification of charities impacts efficiency is the transaction costs associated with seeking out and maintaining external resources from donors and volunteers in addition to their internal set of capabilities and routines. As compared to for-profit organizations that often use accumulated earnings and organizational slack in the pursuit of new diversification opportunities, charitable organizations, with their absence of excess profits and full-budget approach to acquired resources, are typically forced to seek out new financial resources for the purpose of diversification (Galaskiewicz, Bielefeld, & Dowell, 2006; James, 1983). While some for-profit organizations may similarly seek out additional forms of capital by way of equity or debt offerings to fund growth, the use of proceeds of such funds is often much more discretionary than donor funds, which are typically narrow in scope and short-term project-oriented in nature (Kelly, 1991). Thus, one of the key additional mechanisms to consider when exploring the link between diversification and efficiency in charitable organizations is the extent to which the expanded set of activities would be sufficiently different from the organization’s current activities that new donor funding is required. Using panel data involving 3,616 charities over a five-year period, our study examined the relationship between geographic and product diversification on efficiency both independently and in combination. Our main effect findings indicate that while the relationship between geographic diversification and efficiency is U shaped in nature, the rela­tionship between product diversification and efficiency is inverted U shaped. However, our statistical results also suggest that, rather than being independent, the effects of geographic and product diversification on efficiency are highly interdependent. Thus, the effect of one type of diversification on efficiency is significantly dependent on the other, and their relative impacts should therefore be examined in combination rather than in isolation from one another. Our study makes a number of theoretical contributions. First, we contribute directly to the literature on charities by highlighting the underlying mechanisms by which diversification and efficiency are related within the charitable context as opposed to the for-profit context. The identification of such mechanisms provides theoretical justification for why geographic and product diversification exhibit contrasting effects on efficiency, as well as provides practical insight to managers of charities and donors who are seeking information on how these variables are connected. Our study also contributes to broader organizational theory in two ways. First, with the exception of a handful of studies (i.e., Hitt, Hoskisson, & Kim, 1997; Tallman & Li, 1996; Wiersema & Bowen, 2008), very little research has been undertaken to explore the interaction of product and geographic diversification within organizations, despite the prevalence of such combinations in both the for-profit and charitable sectors (Boli & Thomas, 1999; Hitt, Tihanyi, Miller, & Connelly, 2006). Our study directly contrasts these two effects and provides a theoretical rationale for the manner in which they interact. Second, our study suggests that higher transaction costs resulting from high levels of resource dependency on external organizations can play an important role, in addition to the more traditionally studied mechanisms associated with internal capabilities and routines, in the link between diversification and performance.

Kistruck et al. / Diversification in Charitable Organizations   499

We begin by defining and contrasting charities with for-profit organizations, followed by a description of the historical drivers of diversification in the charitable sector. We then develop our hypotheses related to the main and interaction effects of geographic and product diversification on efficiency in charities. After detailing our methodology, we present our results in tabular and graphic format. We conclude with a discussion of our finding and suggestions for future research directions.

Definition of Charitable Organization A charity is defined as “a nonprofit corporate body whose goals are charitable, cultural, scientific, or some other public goal, who make an appeal to the public liberality” (Wunderink, 2002: 21). While charities are a subset of the larger nonprofit sector that also includes cooperatives, associations, and unions, charities are unique in that their purpose must be “public serving” rather than private or “member serving” (Salamon, 1992). Furthermore, a charitable organization is “precluded by external regulation or its own governance structure from distributing its financial surplus to those who control the use of organizational assets” (Powell & Steinberg, 2006: 1). In pursuit of their social missions, charities rely a great deal upon donors and volunteers to deliver a range of different services or “products” such as promoting the arts, combating diseases, or enforcing human rights (Brody, 1996). In addition to the obvious nondistribution-of-profits constraint (Hansmann, 1980), charities differ from for-profit organizations in a number of overarching ways. While for-profit organizations are paying increased attention to their social responsibility as an adjunct to their pursuit of profit maximization, their primary focus remains to capture economic rents (Margolis & Walsh, 2003). Charities, while also being increasingly driven to focus on financial efficiencies, retain a primary focus on effectively addressing pressing social needs (Moore, 2000). The traditional market and nonmarket governance mechanisms within much of the for-profit sector, such as stock options, shareholder regulations, and the market for corporate control (Shleifer & Vishny, 1997), are contrasted in the charitable sector with primarily intrinsic rewards, extensive stakeholder reporting, and an intense competition for donor resources for survival (Rose-Ackerman, 1982). Charitable organizations tend to use a much more flat than hierarchical structure to accommodate more participative rather than topdown decision making (O’Regan & Oster, 2000), are attracted to weak rather than strong formal institutional environments (Austin, Stevenson, & Wei-Skillern, 2006), and have managers who are typically less professionally trained than their for-profit counterparts (Crittenden & Crittenden, 2000). However, at their core, charitable and for-profit organizations are similar in terms of their administrative struggles related to institutional isomorphism, organizational slack, and resource scarcity (Brody, 1996). Both charitable and for-profit organizations are a bundle of routines (Nelson & Winter, 1982) whose actions are often more practically governed by routinized human behaviors rather than by exogenously determined purposes (Mahoney, 2005). Indeed, Nelson and Winter’s early study of the evolutionary theory of organizations was groun­ ded in the study of nonprofit hospitals, and both agency theory (Jensen & Meckling, 1976) and institutional theory (DiMaggio & Powell, 1983) explicitly speak to the application of

500    Journal of Management / February 2013

their key tenets to both for-profit and nonprofit organizations. With the increased involvement of charities in commercial activities and the increased involvement of for-profit organizations in social activities, there also has been a significant blurring of sector boundaries (Dees & Anderson, 2003). Thus, while the contextual differences between for-profit firms and charities are sufficiently different to warrant unique study, their underlying theoretical foundations are sufficiently similar for diversification research grounded in either sector to be mutually beneficial.

The Diversification of Charitable Organizations While charities have garnered increased attention in the field of strategic management, this attention has been primarily as small social influencers on multinational corporations rather than as diversified entities in and of themselves (Prahalad & Brugmann, 2007). However, charities now employ more than 7% of the U.S. workforce (Sherlock & Gravelle, 2009) and are considered agents of globalization rather than simply respondents to international forces (Doh & Teegen 2003), and the social services they provide have become increasingly hetero­ geneous (Boli & Thomas, 1999). Thus, charities have become influential economic actors, and their levels of both geographic and product scope have become extensive. Around the middle of the 20th century, charitable organizations began to rapidly internationalize their operations into new regions around the globe (Lindenberg, 1999). As opposed to for-profit corporations that expand geographically for reasons of seeking higher consumer market returns (Hymer, 1976), lower internal transaction costs (Hennart, 1982), or access to cheaper labor and other factor markets (Buckley & Casson, 1976), charities internationalized because of the willingness of an expanding donor base to supply financial support for the rapidly growing social demands at that time. Such demands included the declining role of the nation-state (Owens & Shaw, 1972), the collapse of the former Soviet Union and other communist countries (Dichter, 1999), worldwide economic recessions (Lindenberg, 1999), and the increased need for “keeping an eye on” multinational corporations that were similarly experiencing rapid international growth (Johnson, 1986). As a result, charities during this period experienced a large increase in both the number of regions in which they were operating and the number of activities with which they became involved (Boli & Thomas, 1999). Toward the end of the 20th century, however, the number of charities had grown to such a level that extensive competition worldwide developed for the supply of resources from traditional sources of government funding, as well as for private giving from individuals, corporations, and religious institutions (Froelich, 1999; Lindenberg, 1999). In conjunction with this new environment of intense resource scarcity, the focus of donors began to shift from measuring success according to the quality of anecdotal stories of social achievements to administrative efficiency (Scott, 2003). As such, donors and charities began to question the degree to which the current levels of geographic and product diversification, resulting from the rapid expansion that occurred over the past several decades, were an efficient use of their increasingly limited resources (Dichter, 1999; Moore, 2000). This focus on financial efficiency continues today. A recent survey of individual and institu­ tional charitable donors indicated that a significant number had discontinued their contributions

Kistruck et al. / Diversification in Charitable Organizations   501

because they were concerned that the money would not be used efficiently (Hall, Lasby, Gumulka, & Tryon, 2006). Such concerns appear well founded given that one in six charities was shown to spend more on fund-raising and administration than on the delivery of social programs and that one in five spend less than the recommended minimum of $0.60 per $1.00 collected in revenues (Donovan, 2002). Additionally, a recent book on the topic of the efficiency of charities indicated that only 11% of Americans believe that “charities do a very good job of spending their money wisely” (Light, 2004). Thus, there remains a number of unanswered questions surrounding the degree to which charities are able to strategically recombine and deploy their resources in order to efficiently grow and adapt to new geographic and product needs.

Hypotheses The Relationship Between Geographic Diversification and Efficiency in Charities The charitable sector represents a previously unstudied context in the field of geographic diversification. Geographic diversification is defined as “the expansion across the borders of global regions and countries into different geographic locations, or markets” (Hitt et al., 1997: 767). The term “geographic diversification” has also been used interchangeably with other terms such as “internationalization,” “international expansion,” and “globalization” in previous research (Hitt, Tihanyi, Miller, & Connelly, 2006). The mechanisms that drive charities to internationalize differ from those in the for-profit sector, resulting in different behaviors that may impact efficiency levels. First and foremost, charities are primarily socially driven in their decisions of where to establish operations. Thus, as compared to for-profit corporations, which typically limit their geographic expansion to the 40 or 60 most developed countries in the world where financial opportunities exceed entry risks (Contractor, 2007), charities, by their very nature, are drawn to a much broader diversity of regions, including the least developed regions with weak formal institutions (Austin et al., 2006; Webb, Kistruck, Ireland, & Ketchen, 2010). Thus, as opposed to following the eclectic paradigm as an internationalization pattern (Dunning, 1988), charities are often quickly drawn to distant regions in need of social reform rather than those where they are most likely to capitalize on economic ownership, location, or internalization advantages. In the weak institutional environments where charities often operate, higher administrative costs may be incurred to create or maintain internal substitutes for strong institutions as well as to circumvent redundant bureaucratic barriers to foreigners (Hoskisson, Eden, Lau, & Wright, 2000). When entering new regions, charities also tend to establish new facilities to provide social services to local inhabitants (Quotah, 2004). Because of their focus on the quality of service delivery, and the desire to build local capacity as part of their social mission, charities often forgo incremental internationalization through modes such as licensing and export and instead jump right to building new on-the-ground facilities in their internationalization patterns. Thus, charities undertake a different route from the traditional Upsalla model (Johanson &

502    Journal of Management / February 2013

Vahlne, 1977) in that they tend to not gradually escalate from more indirect to direct levels of involvement. Using a more on-the-ground entry mode in diversifying into new regions would similarly require costly new administrative structures for service delivery and coordination (Kogut & Singh, 1988). As alluded to previously, charities also differ from for-profit organizations in that the scope of their diversification traditionally has been limited not by their internal capabilities to exploit new opportunities (Penrose, 1959; Wernerfelt, 1984) but rather by their ability to garner new donor resources for an expanded set of activities to meet existing social needs (Galaskiewicz et al., 2006). Much like for-profit organizations, expanding into new regions often involves seeking out and interacting with new customers or “beneficiaries.” However, as compared to for-profit organizations where the costs associated with exploring new customer activities are recouped by way of payment for products and services, beneficiaries of charities are typically not a corresponding source of revenues for the organization (Froelich, 1999). Combined with this lack of customer revenues as a source of financial resources, charities do not possess accumulated earnings with which to engage in international expansion efforts (James, 1983), and the capital they do receive is much more narrowly directed in terms of its targeted usage (Kelly, 1991). Due to heightened levels of resource scarcity, donors have considerably greater power relative to individual charities as a result of this resource dependency (Pfeffer & Salancik, 1978) and thus are exercising greater control in dictating how their funds are used (Carroll & Stater, 2009; Froelich, 1999). The result is that charities seeking to expand into new geographic regions often incur significant transaction costs in seeking out new donor funding as well as learning and executing the reporting requirements that arise as a result of dealing with different donors (Grønbjerg, 1993; Peterson, 1986; Powell & Friedkin, 1986). Therefore, charities looking to expand to different regions would typically be required to search and negotiate with new potential donors representing high ex ante transaction costs (Williamson, 1979). Similarly, the ex post monitoring costs imposed on the charity as a result of transacting with new donors are significant. Due to their focus on regions with weak institutional environments, their tendency to pursue expensive modes of internationalization at early stages, and the high transaction costs associated with seeking out new donor resources in conjunction with diversification, we would expect a negative relationship between geographic diversification and efficiency as charities begin to expand into new regions. However, as they continue to expand geographically, organizational theory has suggested that organizations can eventually learn through experimentation how to operate on the ground within diverse institutional environments and develop new internationalization capabilities (Capar & Kotabe, 2003; Hitt et al., 2006). These internationalization processes and routines can be leveraged to create greater efficiencies as the organization repeats the process of geographic expansion into new regions (Hitt et al., 1997). Although each new region will be different, the process-based knowledge that exists within the organization offers an existing template that can be adapted. While diversifying into new geographic regions would likely continue to require soliciting new resources, at higher levels of diversification, the charity will have already formed a broad network of linkages with diverse donors that can be leveraged, thereby lowering search and negotiation costs (Tsai, 2000). Such funds would increasingly be sought from one of the many existing donors with whom the charity now has existing relationships. Similarly, through

Kistruck et al. / Diversification in Charitable Organizations   503

previous experience in creating reporting templates and procedures to respond to a diverse set of donor demands for monitoring and reporting, the ex post transaction costs of dealing with additional funding requirements similarly decreases as the organization garners ex post transactional capabilities across diverse environments (Mayer & Salomon, 2006). From an internal resources and structural perspective, while smaller, more focused charities often use flat hierarchical structures and managers with less professional training than their for-profit counterparts (Crittenden & Crittenden, 2000; O’Regan & Oster, 2000), as charities become larger and more diversified in the regions they operate, they begin to adapt more hierarchical structures, employ more efficient-minded managers, and engage in more “business-like” processes (Dart, 2004; Odom & Boxx, 1988). Thus, in addition to the decreased transaction costs associated with external resource dependency mechanisms in charities, the more traditionally studied internal mechanisms linking diversification to efficiency would also contribute more positively at high levels of internationalization. Therefore, while during the initial stages of expansion into new regions we expect a negative relationship between geographic diversification and efficiency, at higher levels of geographic diversification we would expect the relationship to become positive. Hypothesis 1: The relationship between geographic diversification and efficiency will be U-shaped curvilinear.

The Relationship Between Product Diversification and Efficiency in Charities In addition to extensive geographic diversification on the part of charities throughout the latter half of the 20th century, many nongovernmental organizations also dramatically expanded their level of product diversification (Boli & Thomas, 1999). Product diversification can be defined as “the entry of a firm or a business unit into new lines of activity, either by processes of internal business development or acquisition” (Ramanujam & Varadarajan, 1989: 525). The notion of diversification in the context of charities reflects the scope of social programs or activities rather than the traditional notion of “products.” Hence, charities can engage in a host of different social activities such as education and research, religion, health, and so forth. Although most charities are initially founded with a single-focused social mission, res­earch has shown that many subsequently diversify into additional social activities (Scott, 2003). However, in contrast to geographic diversification, in which the charity’s initial foray into new regions represents a very dramatic difference in terms of the donor base, institutional environment, and operational undertakings, a charity’s expansion into new services has been characterized much more as a “drift” than a “leap” (DiMaggio, 1986; Jones, 2007). Thus, the diversification patterns of charities are much more incremental in terms of “product” diversification as compared to geographic diversification, as charities gradually shift away from services highly related to their core missions toward activities more and more unrelated. It is expected that this distinction will alter the slope of the relationship between product diversification and efficiency as compared to geographic diversification. Because related

504    Journal of Management / February 2013

product diversification involves only minor adjustments to existing activities, charities can leverage positive efficiencies due to the synergies that can be exploited within resource bundles across related activities (Chandler, 1962; Penrose, 1959). Similarly, expanding into similar activities that fit the charity’s overall mission can produce economies of scale and scope for the organization (Teece, 1982). Perhaps of greatest importance, charities that engage in additional activities that are highly related to their existing scope of services are able to do so without incurring the additional transaction costs of seeking out and negotiating with new sources of funding. While existing donors provide fairly specific instructions on how their funds are to be used, there is at least some degree of “wiggle-room” for activities that are highly related to the organization’s stated mission (Oster, 1996). Indeed, the limited evidence that has been presented on charities that diversify their missions suggests that, at least at related levels of product diversification, charities are able to produce some positive efficiencies (Kaplan & Norton, 1996; Nielsen, 1986). Thus, we would expect that diversification into closely related activities would be positively related to efficiency in charitable organizations. However, as charities engage in increasingly unrelated diversification, such efficiency gains may be significantly challenged. Diversification into more unrelated activities has been described as a much more “on-the-job training” process, with planning rarely taking place prior to expansion in charities (Stone, Bigelow, & Crittenden, 1999). This is particularly problematic given that organization theory suggests that unrelated diversification without appropriate planning and structural design typically produces diseconomies of scale and scope (Hoskisson & Hitt, 1990; Rumelt, 1974). Additionally, pursuing increasingly unrelated activities to the organization’s core mission requires seeking new sources of funding with a corresponding increase in ex ante and ex post transaction costs. Therefore, while we would expect a positive relationship between product diversification and efficiency with related diversification, we would anticipate a negative relationship at more unrelated levels. Hypothesis 2: The relationship between product diversification and efficiency will be inverted U-shaped curvilinear.

The Interaction Effect of Product and Geographic Diversification in Charities Despite their obvious relation to one another, the application of organizational theories to product diversification and geographic diversification has undergone separate but parallel paths of development over the past several decades. However, many charities have become diversified in both their products and geographic locations throughout the latter half of the 20th century (Boli & Thomas, 1999). Thus, understanding how these two types of diversification interact with one another is important to understanding their net effect on efficiency levels. By comparing the two main effects hypothesized, the impact of either type of diversification on efficiency is similarly dependent on the degree to which such diversification represents a significant departure from the existing set of organizational activities. In instances where such diversification represented incremental changes from the charity’s existing donor relationships,

Kistruck et al. / Diversification in Charitable Organizations   505

structural configurations, and routines, the effects on efficiency were positive. However, when diversification was significantly different, new linkages, new structures, and new routines were required, resulting in a negative effect on efficiency. Therefore, while the respective curves of product and geographic diversification on efficiency had different predicted outcomes, the theoretical explanations underlying both types were congruent. We hypothesize that charities that are relatedly diversified in their set of activities will experience downward pressures on efficiency as they begin to internationalize into new regions. While there would be some efficiency gains associated with minor diversification into new services, prior organization theory has suggested that “the information-processing demands are more complex and greater when firms move into new international markets than when they move into different product markets within the same domestic setting” (Hitt et al., 1997: 773). Thus, we argue that the stronger negative effects of initial geographic diversification will produce net inefficiencies for charitable organizations even when relatedly diversified. However, as relatedly diversified organizations continue to expand into a greater number of diversified geographic regions and leverage their broader set of donor networks and operational routines, the positive main effects of both related product diversification and higher geographic diversification will combine for a net positive upswing. Thus, we hypothesize that relatedly diversified charities will follow a U-shaped curve as they increasingly diversify into new regions. However, we expect this relationship to be inverted for unrelatedly diversified charities. Charitable organizations that are unrelatedly diversified were at one time forced to seek out and negotiate with new donors, erect new administrative structures, and develop new processes. While such undertakings likely produced negative efficiencies to the organization from a standalone perspective, charities that had already built diverse linkages, structures, and routines by engaging in unrelated diversification may be able to leverage these in the early stages of diversifying into new geographic regions (Kim, Hwang, & Burgers, 1989). Thus, unrelated levels of product diversification in combination with low levels of geographic diversification can result in increased efficiency for firms as they are able to capitalize on such prior investments representing complementarities (Hitt et al., 2006). However, as unrelatedly diversified charities continue to expand into an increasing number of geographic regions, the complexity and lack of absorptive capacity resulting from combining both unrelated product and high geographic diversification will be too much of an administrative burden for charities and result in a downturn in efficiency (Ramanujam & Varadarajan, 1989; Zahra & George, 2002). While external transaction costs with acquiring funding may be much less of an issue to charities with incredibly diverse donor bases that were developed in becoming both unrelatedly diversified and highly geographically diversified, the internal costs of transacting and coordinating are sufficiently high to warrant overall inefficiencies (Doukas & Lang, 2003; Kumar, 2009). Thus, we hypothesize that unrelatedly diversified charities will follow an inverted U-shaped curve as they increasingly diversify into new geographic regions. Hypothesis 3: The effect of geographic diversification on efficiency will be moderated by the level of product diversification. Specifically, the U-shaped curvilinear relationship between geographic diversification and efficiency at related levels of product diversification will become inverted at unrelated levels of product diversification.

506    Journal of Management / February 2013

Data and Method Sample The data for this study were drawn from the Charities Assessing and Registration (CARE) database system that was provided by the Canada Revenue Agency for the five-year period from 1997 to 2001. The specific time period was selected for making consistent comparisons across time as there was a change in government reporting requirements in 1997 and then again in 2002. The longitudinal data set contains charities that were diversified geographically as well as involved in a diverse range of services including education and research, social services, and the environment. A large number contained missing data for one or more variables in a year, and the pairwise deletion method was adopted to exclude 220 records from the sample. Previous research has shown that including cases with missing observations using panel data is subject to biased results, even when using a host of value estimation techniques (Twisk, 2004). Our final sample consisted of 3,616 charities and 17,860 observations. Recent studies have stressed the importance of using longitudinal data to explore the rela­ tionship between geographic diversification and efficiency (Lu & Beamish, 2004; Thomas & Eden, 2004). Geringer, Tallman, and Olsen (2000) also indicated that the effects of both geographic and product diversification on efficiency can vary tremendously over time, and thus a cross-sectional analysis is potentially misleading. However, the overwhelming number of studies examining this relationship has relied on either cross-sectional or pooled data, thereby ignoring the time effects (Hitt et al., 2006). Thus, our sample is composed of panel data consisting of multiple observations of the same variables over time to redress these previous shortcomings.

Measures Efficiency. Efficiency was calculated by dividing the amount spent directly on charitable programs by the total expenditures of the organization. Highly efficient charitable organizations are expected to minimize administrative and fund-raising expenses, thereby maximizing the percentage of dollars spent on the actual delivery of charitable programs (Bennett & Savani, 2003). Thus, our measure attempts to capture the administrative ability of charities to efficiently manage the positive and negative effects of geographic and product diversification. It is important to note, however, that our measure of efficiency contains no information on the quality of services being produced, a component of charitable performance that continues to be extremely problematic to capture (Moore, 2000). However, due to increased pressures on charities to be more efficient, our measure of efficiency remains the most commonly used metric to evaluate the activities of charities by governments, potential donors, and other stakeholders (Callen, Klein, & Tinkelman, 2003; de Andres-Alonso, Cruz, & Romero-Merino, 2006; Glaser, 1994; Keating & Frumkin, 2003). This measure of efficiency has also become the common metric upon which “watchdog agencies” evaluate the performance of charities (Bennett & Savani, 2003). Charities deemed inefficient tend to experience

Kistruck et al. / Diversification in Charitable Organizations   507

lower future financial funding from stakeholders who are no longer willing to support such organizations (Thornton, 2006). Therefore, while our measure of efficiency may be incomplete as an overall assessment of performance, we believe it provides novel insight into the capacity of charities to manage diversification activities using a metric that their stakeholders, upon whom they rely for their ongoing survival, use to judge the organization’s worthiness. Geographic diversification. Many operationalizations of geographic diversification have been used in prior organizational theory literature including the percentage of foreign sales as a percentage of total sales (Capar & Kotabe, 2003; Geringer, Beamish, & daCosta, 1989; Grant, 1987), foreign assets as a percentage of total assets (Daniels & Bracker, 1989; Ramaswamy, 1993), and the number of countries within which a firm operates (Kogut, 1985; Lu & Beamish, 2001, 2004; Tallman & Li, 1996). Broadly speaking, these measures have fallen into two categories—geographic scale and geographic scope (Thomas & Eden, 2004). While geographic scale refers primarily to the depth of strategic focus of an organization’s international activities vis-à-vis its domestic activities, geographic scope refers to the breadth of an organization’s activities. We measured geographic diversification as the total number of regions within which the charities operated. We used a measure that captures geographic scope rather than scale because our theoretical interest was to examine how greater geographic diversity impacts efficiency levels. While increases in geographic scale may simply be capturing greater strategic focus within a single foreign market, the use of a scope measure allows us to capture the increase in the number of different foreign markets (Lu & Beamish, 2004). Furthermore, our measure of geographic diversification was calculated at the regional rather than country level because scholars have begun to emphasize interregional rather than intraregional differences as more salient to issues of internationalization (Ghemawat, 2003; Qian, Khoury, Peng, & Qian, in press; Rugman & Verbeke, 2004). Thus, geographic diversification was measured using 12 different global regions: (1) United States and Mexico; (2) Central America, Caribbean, and Antilles; (3) South America; (4) Western Europe; (5) Central and Eastern Europe; (6) Middle East; (7) South Asia; (8) China; (9) other Asian countries; (10) Eastern and Southern Africa; (11) Northern, Central, and Western Africa; (12) and Australia and Pacific. Product diversification. The notion of product diversification in charitable organizations reflects the scope of social activities or causes rather than typical “product” categories. Our sample contains a wide array of charitable activities under the primary headings of social services (A), international aid (B), education and research (C), culture and arts (D), religion (E), health (F), environment (G), and community benefits (H). Additionally, our sample contains a host of subheadings under each primary category. For instance, under the primary heading of culture and arts (D) were museums, galleries, and concert halls (D1); festivals and performing groups (D2); art schools (D3); cultural centers (D4); and historical sites and heritage societies (D5). Measures of diversification have taken several forms in previous research, including onedimensional and multidimensional measures. However, the entropy measure (Jacquemin & Berry, 1979; Palepu, 1985) provides a more fine-grained analysis that has been construct validated (Hoskisson, Hitt, Johnson, & Moesel, 1993). For each charity, the organization reported

508    Journal of Management / February 2013

not only its range of activities but also its approximate percentage of emphasis spent on each activity. Thus, the entropy measure was calculated as follows: Product Diversification = Si [Pi × ln(1/Pi)],

where Pi is the percentage of emphasis on activity i, and ln(1/Pi) represents the weight of each activity. By incorporating relative emphasis, the entropy measure is capable of capturing diversification beyond basic industry classifications. For instance, a charity that was engaged in only two different charitable activities but split 50% of its efforts on each would score higher on the entropy scale than one that was engaged in three different activities with 96% of its efforts on one activity and only 2% on each of the other two. While prior studies have suggested that there is a high degree of correlation between the Jacquemin and Berry (1979) objective entropy measure and the Rumelt (1974) subjective measure of relatedness (Baysinger & Hoskisson, 1989; Hoskisson et al., 1993), for statistical conclusion validity we conducted a correlation analysis on these two measures within our particular data set. The extent to which a charity’s diversification of activities was judged to be related versus unrelated was a factor of the degree to which the activities listed were included under one primary heading (i.e., A1, A3, and A6) versus under multiple primary headings (i.e., A1, B3, D5). Subsequently, we split the total number of charities in our entropy measure into two groups based on whether they were higher or lower than the midpoint of our measure. We then analyzed the degree to which charities in the lower half correlated with charities categorized as related and those in the upper half correlated with those categorized as unrelated. Our results indicated that those in the lower half correlated much more highly with related (.848, p < .01) than unrelated (.341, p < .01) and those in the upper half correlated much more highly with unrelated (.401, p < .01) than related (.215, p < .01). Control variables. Previous work involving both for-profit corporations and charities has indicated that organizational size can significantly impact efficiency levels (Geringer et al., 2000). Interestingly, charities have been shown in some cases to experience diseconomies of scale at much smaller sizes than for-profit corporations (O’Regan & Oster, 2000). Thus, we included the natural log of total assets as a control for organizational size. We also included a control for the industry sector of charities as defined by its primary purpose. Previous studies have shown that different types of charities may experience differing levels of economies of scale and scope that would impact their relative importance (Callen et al., 2003). Thus, we created dummy variables based upon the industry that the charity focused the majority of its efforts in, with the choices including social services, international aid and development, education and research, culture and arts, religion, health, the environment, and community benefits. A control for the extent of subcontractor monitoring was also used to account for the differences in charities’ attention to agency costs. While most charitable organizations operate directly in different geographical regions, there is a subset in our data that operates indirectly through local charities. Because our theoretical arguments refer to the transaction costs between the charity and potential donors in relation to diversification rather than downstream monitoring costs between the charity and its potential subcontractors, we wanted to control for

Kistruck et al. / Diversification in Charitable Organizations   509

this effect. In the forms charities must fill out, they are asked to report on the usage of seven possible subcontractor monitoring mechanisms for their international operations such as “receive a detailed breakdown of expenditures at least annually” and “give prior approval for the specific allocation of funds.” Thus, the total number of mechanisms was calculated for each charity to control for the extent of subcontractor monitoring efforts. A control for cultural distance was also included in our analyses. Much of the extant international business research suggests national culture as a key variable influencing the relationship between geographic diversification and performance (Hennart & Larimo, 1998; Li, Lam, & Qian, 2001). Because our geographic diversification arguments presented in this article are related more to strength of institutional environment rather than cultural heterogeneity, we included a control for cultural distance. The index developed by Kogut and Singh (1988) is the most frequently used measure of the cultural distance construct (Shenkar, 2001). For the purpose of this study, we calculated cultural distance based on the four original dimensions—PDI, IDV, MAS, and UAI (Hofstede, 1980). The fifth dimension of cultural distance, Long-Term Orientation, was not included in the analysis because too few scores of this dimension were available from the Hofstede index. The composite index was calculated as follows: 4

CDj = ∑ {(Iij – Iiu)2/Vi}/4, i=1

where i = cultural dimension (1-4 or PDI, IDV, MAS, UAI), j = country other than host country (Host country CANADA), and V = variance of the index of the ith dimension. We calculated regional scores using all the constituent countries for which information about their cultural dimension scores was available. If a charity worked in more than one region, average cultural distance was used. A final control variable we included in our models was revenue shifting. Although many charities based in the northern hemisphere (i.e., the United States, Europe, and Canada) continue to receive the majority of their funding from government, a growing number are being forced to shift their donor bases as government funding decreases (Boris, 1998). As a result, charities are increasingly looking to corporations, foundations, private individuals, and even commercial activities to fill this void. Important to our analyses, diversifying into different broad types of revenue sources can be accompanied by additional constraints and administrative costs (Grønbjerg, 1993). Because the theoretical arguments we make in this article relate to the increased transaction costs that arise from seeking out new donors in conjunction with diversification, we wanted to control for the natural shift that is taking place between donor types irrespective of diversification decisions. Therefore, we included a variable of revenue shifting as a control in our model by separating funding into three primary classes: private, government, and internal. An entropy score was then calculated to reflect both the existence of alternative types and the relative reliance of the charity on each of the potential types. Instrumental variables. As prior research has shown, strategic choices related to both geographic and product diversification may not be completely exogenous (Campa & Kedia, 2002;

510    Journal of Management / February 2013

Hamilton & Nickerson, 2003; Shaver, 1998). In such instances, a two-stage least squares approach using instrumental variables can be an effective method to address problems related to endogenous repressors, that is, when explanatory variables are correlated with the regression error term (Bascle, 2008; Baum, Schaffer, & Stillman, 2007a). Under these circumstances, instrumental variables provide a way to obtain consistent parameter estimates (Campa & Kedia, 2002; Doukas, Kim, & Pantzalis, 2005). In order to account for the potential endogeneity in our geographic and product diversification variables, we identified a number of instrumental variables that were correlated to our endogenous predictors but not correlated to our dependent variable. The instrumental variables were (1) board size, (2) number of non-arm’s-length board members, (3) amounts receivable from non-arm’s-length parties, (4) amounts payable to non-arm’s-length parties, (5) number of compensated positions, (6) international expenditures using a contracting approach, (7) international expenditures using a gift approach, (8) international expenditure using other approaches, and (9) quick ratio (please see Appendix A for a more detailed discussion of the theoretical rationale for selecting such variables as instruments).

Analysis and Results Analysis Our panel data were analyzed using an instrumental variable/two-stage least squares approach in Stata 11. We used the ivreg2 command, which is an extension of the Stata ivreg command created by Baum, Schaffer, and Stillman (2007b) for use with continuous dependent variables. The ivreg2 command automatically generates Sargan and Basman overidentification statistics, and its estimates are robust to the heteroscedasticity and autocorrelation common with panel data (Bascle, 2008; Baum et al., 2007b; Sargan, 1958). The ivreg2 command also allows for linear and nonlinear models in a single framework, which is a key component of our hypotheses. Furthermore, this technique is especially useful in longitudinal studies where the dependent variable under study in any given year is highly correlated with the dependent variable in the prior year, thereby violating a fundamental assumption of traditional methodologies such as ordinary least squares or chi-square tests. This analytical approach thus allows researchers to examine “within” variable differences over time without biasing the results for highly correlated observations. Appendix A outlines our step-by-step process for dealing with endogeneity concerns. For illustrative purposes, we present only the regression model with interaction terms (Table A.1), but the same process was adopted for analyzing main effects. We also show the first-stage regressions using our instrumental variables for geographic and product diversification sep­ arately. The first-stage regressions of our nine instrumental variables indicate that all F values were above 10 on either geographic or product diversification (Tables A.1.1 and A.1.2), and the p values demonstrate that all instrumental variables were strongly associated with the endogeneous variables used in our analysis (Hahn & Hausman, 2005; Staiger & Stock, 1997). Table A.2 presents our regression analysis corrected for endogeneity. As the Sargan statistic indicates (0.342, p value = .5589) the instrumental variables we included in our model

Kistruck et al. / Diversification in Charitable Organizations   511

Table 1 Mean, Standard Deviation, Minimum, and Maximum (N = 17,850) Variable Efficiency Geographic diversification Product diversification Cultural distance Subcontractor monitoring Organization size Revenue shifting Quick ratio Board size Non-arm’s-length board members Non-arm’s-length amounts receivable Non-arm’s-length amounts payable Compensated positions International contracting International gifts International other

Mean

Standard Deviation

Minimum

Maximum

0.77 3.95 0.60 2.49 1.30 2,710,273 0.27 0.50 5.20 0.71 16,150 18,487 16 70,496 8,651 44,956

0.13 3.21 0.43 0.64 2.43 29,000,000 0.26 0.44 6.16 1.81 345,938 268,050 192 1,797,047 235,112 728,012

0.01 1.00 0.00 0.39 0.00 30,000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

  1.00  12.00   1.39   4.81   7.00 1,560,000,000   1.10   2.68 112.00  32.00 36,400,000 13,800,000 9,999 143,000,000 16,400,000 32,500,000

are not themselves endogenous (Baum et al., 2007a; Sargan, 1958). Further­more, the PaganHall statistic (2.925, p value = .9982) indicates that there is no problem of heteroscedasticity (Pagan & Hall, 1983), and the Arellano-Bond statistic (1.88, p value = .06) suggests there is no statistical indication of autocorrelation (Arellano & Bond, 1991). However, as the Arellano-Bond statistic is only marginally greater than .05, we used the robust estimation option of the ivreg2 command to obtain final estimates robust to hetero­scedasticity and autocorrelation.

Results Both descriptive statistics and a correlation matrix for the variables were calculated and are reported (Tables 1 and 2). Correlations are reported using a Bonferroni correction to account for yearwise effects typically present in panel data. Model 1 was run using only the control variables included in the study. A test for overall model fit indicates that the model is significant (F value = 141.66) at the p < .001 level. As Model 1 indicates, both the effects of organization size (b = –0.030, p < .001) and cultural distance (b = –0.035, p < .001) on efficiency are statistically significant and negative. The control of subcontractor monitoring is also statistically significant but positive (b = 0.039, p < .001), and revenue shifting has no significant effect on efficiency in the control model. (See Table 3.) Model 2 was subsequently constructed to test the main effects of geographic and product diversification on efficiency. Model 2 was analyzed using the ivreg2 command with robust estimates. A test for overall model fit indicates that the model is significant (F value = 450.71)

512

†p < .10. *p < .05. **p < .01. ***p < .001.

Efficiency Geographic diversification Product diversification Cultural distance Subcontractor monitoring Organization size Revenue shifting Quick ratio Board size Non-arm’s-length board members Non-arm’s-length amounts receivable Non-arm’s-length amounts payable Compensated positions International contracts International other International gifts -.37*** -.15** -.03 .12* -.05 -.01 .03 -.05 .02 -.01 -.02 -.02 -.03 -.04 -.01

1

.11* .02 -.03 .04 -.01 -.12* .08* -.06† .01 .07† .03 .08* .07* .01

2

.03 -.02 -.03 .03 -.10* .04 .04 -.06† .01 -.09† .03 .04 .09*

3

.04 .02 -.02 .00 .00 -.02 .00 .01 .03 .00 .01 -.01

4

6

7

.10 .08* .10* .24 -.01 .05 .06† .09* .11* -.05 -.02 -.03 .02 .05 .02 .07† .11* .05 .10 .36*** .10* .08* .07† .03 .11* .01 .06† .05 .04 .03

5

Table 2 Correlations (N = 17,850)

-.03 -.03 .03 .00 -.03 .02 .05 .02

8

.15** .02 .01 .09* .03 .03 .03

9

-.01 -.02 -.03 -.01 -.02  .01

10

.18** .04 .04 .01 .00

11

.09* .01 .02 .00

12

.04 .01 .00

13

.00 .08*

14

0.02

15

Kistruck et al. / Diversification in Charitable Organizations   513

Table 3 Effects of Geographic and Product Diversification on Efficiency (N = 17,860) Independent Variable

Model 1

Geographic diversification Geographic diversification (squared) Product diversification Product diversification (squared) Geographic Diversification × Product Diversification Geographic Diversification (squared) × Product Diversification Geographic Diversification × Product Diversification (squared) Geographic Diversification (squared) × Product Diversification (squared) Organization size Cultural distance Subcontractor monitoring Revenue shifting F value Probability > F

–0.030*** (0.002) –0.035*** (0.009) 0.039*** (0.002) 0.003 (0.022) 141.66 .0000

Model 2

Model 3

–0.293*** (.011) 0.017*** (0.005) 0.171 (0.022) –0.639** (0.231)

–0.298*** (.077) 0.043*** (0.008) 0.310† (0.166) 0.645† (0.371) 0.162† (0.100) –0.039† (0.022) 0.141 (0.300) –0.112*** (0.027) 0.001 (0.002) 0.007 (0.006) 0.003 (0.002) –0.050* (0.024) 731.66 .0000

0.001 (0.003) –0.004 (0.004) 0.003 (0.003) –0.039* (0.022) 450.71 .0000

Note: Dummy variables of charitable industry sector included but not shown. Standard errors are in parentheses. †p < .10. *p < .05. **p < .01. ***p < .001.

at the p < .001 level. As Table 3 illustrates, the relationship between geographic diversification (squared) and efficiency is positive and statistically significant (b = 0.017, p < .001), suggesting an overall U-shaped relationship. Thus, Hypothesis 1 is statistically supported. Model 2 also tests the main effect of product diversification on efficiency. The relationship between product diversification (squared) and efficiency is negative and statistically significant (b = –0.639, p < .01), indicating an overall inverted U-shaped relationship. Thus, Hypothesis 2 is also supported. Prior to testing the interaction hypothesis, the variables within the model were standardized. Yearwise comparisons also were examined for high variance inflation factors, but no scores of concern were found. It is important to stress the need to include all combinations of linear and quadratic terms within the model to ensure accurate results. Thus, the model contains not only multiplicative terms related to the individual linear terms and individual quadratic terms but also multiplicative terms with the range of linear terms multiplied by the quadratic terms. Model 3 indicates that the interaction effects of product diver­sification

514    Journal of Management / February 2013

(squared) on the relationship between geographic diversification (squared) and efficiency is significant and negative (b = –0.112, p < .001), which supports our assertion that the U-shaped main effect between geographic diversification and efficiency becomes downward sloping at unrelated levels of product diversification. Thus, Hypothesis 3 is statistically supported.

Discussion As our results indicate, the effects of diversification on efficiency in charitable organizations can be very complex. In terms of our main effect hypothesis of geographic diversification on efficiency in Model 2, as Figure 1 illustrates, while charities that expand into new geographic regions generally experience inefficiencies, as the organizations expand into a very high number of regions they achieve some positive efficiencies as compared to when they expand into a moderate number of regions. This finding supports our first main effect hypothesis that, when holding product diversification levels constant, charities initially experience inefficiencies when diversifying into new geographic regions, as they are forced to seek out new sources of donor capital in addition to building costly new, on-the-ground administrative structures and capabilities. However, at a very high number of geographic regions, charities can produce some improved efficiencies, as their now diverse network of donor sets and global brand, internationalization routines, and hierarchical structures built at earlier stages of internationalization can be more efficiently exploited. However, as Figure 1 suggests, the positive efficiencies obtained at a very high number of regions are still far below those of charities focused on only a low number of regions. Thus, when holding the level of product diversification constant, while diversifying into a high number of regions can produce some efficiencies as compared to diversifying into a moderate number of regions, expanding a charity’s international scope at any level appears less efficient than remaining focused on a single region. Similarly, as Figure 2 illustrates, charities that diversify into new related services experience some initial positive efficiencies at very low levels but then very quickly become more inefficient at increasingly unrelated levels. This supports our second hypothesis that, when holding geographic diversification levels constant, charities are able to garner some increase in efficiency by leveraging existing sources of donor funding, internal routines, and structures when initially engaging in highly related levels of product diversification. However, engaging in moderately to highly unrelated types of product diversification leads to a decrease in efficiency as new donors, structures, and processes are similarly required to manage a more diverse scope of activities. However, as Figure 2 suggests, much like geographic diversification the upward slope in efficiency is minimal, and it quickly becomes downward sloping at most levels of product diversification. Thus, when holding the level of geographic diversification constant in Model 2, while diversifying into a highly related service may produce some minimal efficiency gains for charities, in large part the charities that remain much more focused produce significantly higher efficiencies than those that diversify into even moderately unrelated services. However, as Model 3 suggests, geographic and product diversification do not influence efficiency independently but, rather, act in combination with one another. Therefore, the main

Kistruck et al. / Diversification in Charitable Organizations   515

Figure 1 Main Effect of Geographic Diversification on Efficiency

Figure 2 Main Effect of Product Diversification on Efficiency

516    Journal of Management / February 2013

Figure 3 Moderating Effect of Product Diversification on Geographic Diversification and Efficiency

effects must be interpreted with caution given that the degree to which one type of diversification will impact efficiency is significantly dependent on the level of the other type of diversification. With regard to the moderating effect of product diversification on the relationship between geographic diversification and efficiency as expressed in Hypothesis 3, as Figure 3 suggests,1 the shape of the relationship differs depending on the level of product diversification. At more related levels of product diversification, the effect is U shaped while at unrelated levels, the effect is inverted U shaped. This inversion does suggest that the new donor linkages, routines, and structures that were developed to deal with the complexity that occurred through engaging in more unrelated levels of product diversification can be exp­loited by charities in their very early stages of diversifying into new geographic regions to minimize inefficiencies. Figure 3 also illustrates that while charities that maintain related levels of product diversification follow a similar U-shaped curve as the main effect in Figure 1, at a very high number of geographic regions such charities are capable of producing similar efficiency levels as those that remained focused on a single region. Rather, it is only at more unrelated levels of product diversification that this curve is inverted and leads to rapidly increasing inefficiencies when combined with higher levels of geographic diversification. Thus, when allowing product diversification to vary rather than be held constant, our results suggest that charities may be able to expand into a large number of regions without significantly

Kistruck et al. / Diversification in Charitable Organizations   517

Figure 4 Three-Dimensional Interaction Effect of Product Diversification and Geographic Diversification on Efficiency

damaging efficiency levels. This distinction is particularly important for charities that are not simply trying to maximize efficiency but rather trying to maximize their “social bang” for their “financial buck.” Thus, being able to engage in a set of one or more related social activities in a high number of geographic regions without resulting in significant losses in efficiency levels allows charities to broaden their social reach without damaging their ability to elicit donor funds. To garner a better understanding of the overall range of interaction effects between geographic and product diversification at a very practical level, we plotted our interaction results in a three-dimensional graph. As Figure 4 reveals, there is a large range of geographic and product diversification combinations that allows charities to deliver more than one social activity in more than one region without causing dramatic losses in efficiency. Generally speaking, charities can engage in low to moderate levels of one type of diversification without significantly harming efficiency so long as the other type of diversification is also at low

518    Journal of Management / February 2013

to moderate levels. Additionally, charities can provide a highly unrelated set of activities without damaging efficiency as long as they remain focused on a single region and, similarly, can provide services to a very high number of regions as long as they remain focused on a single product. However, combinations in which the charity is focused on more unrelated products and a higher number of regions simultaneously do produce significant inefficiencies within charitable organizations, given the administrative strain to manage such complexity. Such findings not only provide prescriptive guidance for managers of charitable organizations but also make several contributions to organization theory in general. While prior research conducted by organizational theorists has suggested that the main effects of geographic diversification and product diversification on efficiency are very similar (Tallman & Li, 1996; Wiersema & Bowen, 2008), we propose that their main effects may follow very different patterns. While similar in our assertions that more disparate types of diversification often require new external resources, structures, and capabilities that cause inefficiencies within the organization, the levels at which these inefficiencies occur in product and geographic diversification can be very different. While even low levels of geographic diversification may represent the need for new resources, structures, and capabilities, related levels of product diversification may not. Thus, this contrasts the slightly upward and significantly downward sloping initial trends in the two types of diversification as main effects. However, while the expanded resource networks and experiences in geographic diversification eventually produce some small efficiencies at higher levels of internationalization, all but highly related levels of product diversification result in greater inefficiencies. It seems then, that while initially entering new product markets may represent lower costs and complexity than initially entering new geographic markets (Hitt et al., 1997), engaging in higher levels of product diversification represents more managerial complexity and internal transaction costs than operating in a higher number of geographic markets. While some prior research has suggested that very high levels of geographic diversification may produce similar inefficiencies within organizations, our post hoc test for an S-shaped curve as a type of “three-stage” model of internationalization (Lu & Beamish, 2004; Thomas & Eden, 2004) proved nonsignificant (p = .266). While charities and for-profit organizations possess similar mechanisms that would be expected to influence the curvature of these relationships—structural design, internal capabilities, and the nature of institutional environment—the charitable sector’s extreme dependency on donors as opposed to customers as the primary source of ongoing resources presents an additional external mechanism related to the transaction costs incurred with increased diversification activities (Froelich, 1999). Thus, in addition to similar costs associated with establishing new organizational structures, routines, and internal governance capabilities when engaging in diversification that is significantly different from the organization’s existing activities (Hitt et al., 2006), charities have the additional transaction costs of searching for and reporting to new donors, given the restricted discretion over donor funds and absence of accumulated earnings upon which to draw. Thus, while the mechanisms that have received the greatest amount of focus in the for-profit research domain on diversification revolve around internal capabilities to exploit potentially profitable opportunities, one of the most important mechanisms in the charitable sector is the external transaction costs.

Kistruck et al. / Diversification in Charitable Organizations   519

While previous for-profit studies have incorporated transaction cost arguments to predict diversification outcomes such as diversification type (Chatterjee & Wenerfelt, 1991), vertical boundaries (Hennart, 1982), or horizontal boundaries based on internal capabilities (Hitt et al., 1997), the charitable context allows us to integrate resource dependency with transaction cost perspectives to understand how relationships with external actors underpin the relationship between diversification and efficiency. Although this additional external transaction cost mechanism is certainly much more prominent in charitable as opposed to for-profit organizations, there are certain types of for-profit organizations in which these mechanisms may play a larger role than discussed in prior diversification literature. For instance, empirical studies of for-profit corporations that are small in size have similarly found support for a U-shaped relationship between geographic diversification and efficiency (Coviello & McAuley, 1999). The existing rationale provided for this linkage is that smaller organizations are more likely to incur higher initial levels of liabilities of foreignness and newness, compared to larger organizations (Lu & Beamish, 2001), and like charitable organizations, much of the adaptation of smaller firms occurs during the geographic diversification process itself rather than through careful planning (Ruigrok & Wagner, 2003). However, smaller organizations, much like charitable organizations, are also less likely to possess sufficiently large accumulated earnings with which to fund geographic diversification activities. As a result, the transaction costs associated with seeking out new external sources of financial capital for small, geographically diversifying firms from angel investors, venture capitalists, and so forth may also play a significant role as a theoretical mechanism underlying the empirically supported U-shaped relationship. Furthermore, empirical studies of service-based for-profit organizations have also exhibited a U-shaped relationship as compared to manufacturing firms that diversify geographically, because, much like charities, service-based firms are more likely to establish expensive on-the-ground modes of internationalization as opposed to indirect modes such as licensing or export (Capar & Kotabe, 2003). Again, not only may engaging in more expensive modes of internationalization impact the efficiency of service-based firms by way of increased operating costs, as suggested in previous studies, but such firms may also be more likely to incur external transaction costs to raise the larger amount of financial capital required to engage in more direct modes of internationalization. Our findings of an inverted U-shaped relationship between product diversification and efficiency are also somewhat similar to many organizational studies involving for-profit corporations (Palich, Cardinal, & Miller, 2000). However, much of the theoretical rationale provided for the initial increase in efficiency levels in the strategic management field has centered on the risk-mitigating properties of diversification (Ramanujam & Varadarajan, 1989). Charities, which are primarily driven to diversify into new activities for social reasons rather than for the mitigation of financial risk, are still able to achieve positive administrative efficiencies by engaging in related diversification. This suggests that the resource-based theoretical arguments related to product diversification remain valid even in the absence of the risk benefits put forth by portfolio theorists (Lubatkin & Chatterjee, 1994). However, it appears that in charities, as well as in for-profit corporations, the complexities of engaging in unrelated activities place overly high transaction costs and administrative burdens upon the organization.

520    Journal of Management / February 2013

Our results also contribute to the ongoing debate among organizational theorists regarding geographic and product diversification as complementary or contradictory activities when used in combination. With the exception of a handful of studies, very little research has been undertaken to contrast the theoretical rationale underlying the link between performance and these two types of diversification simultaneously (Hitt et al., 2006). There have, however, been several empirical studies that have attempted to bridge the geographic and product divide to examine their joint effects on efficiency. While a subset of these studies has failed to find any statistical support for an interaction effect (Geringer et al., 1989; Geringer et al., 2000; Tallman & Li, 1996), most can be grouped into two broad categories— those that find support for a positive interaction effect (Hitt et al., 1997; Kim et al., 1989) and those that support a negative interaction effect (Doukas & Lang, 2003; Franko, 1989; Kumar, 2009). The theoretical arguments underpinning the “positive group” are that the structures erected for one type of diversification are leverageable by the other, and engaging in both types of diversification allows for even greater mitigation of risk. Arguments supporting the “negative group” are that the strain on the organization’s resources, such as absorptive capacity, caused by engaging in both types of diversification are too high, and the internal transaction costs and complexity of combined diversification is overwhelming to the firm’s managers. Again, as Figure 3 illustrates, increased levels of product diversification not only result in a changing of the U-shaped geographic diversification curve but in fact invert the curve at high levels of unrelated product diversification. Thus, while prior studies examining the interaction effects of geographic and product diversification have tended to fall into either the “positive complements” or the “negative substitutes” camp (Kumar, 2009; Wiersema & Bowen, 2008), our study suggests that the effect can be both positive and negative depending on the specifics of their interaction. Furthermore, because for-profit organizations, compared to charitable organizations, tend to be more hierarchical and top-down in their decision-making structures and typically have more professionally trained managers (Crittenden & Crittenden, 2000; O’Regan & Oster, 2000), we suggest that the complementarity of geographic and product diversification may be even greater in corporate firms. For-profit organizations also typically seek out more culturally similar and geographically proximate locations with strong institutional environments for international expansion (Austin et al., 2006) that would lessen the overall administrative burden when coincidentally engaging in product diversification. And of course the ability of for-profit organizations to rely upon internal financial resources or more discretionary sources of external capital would reduce the dual transaction costs that can exist when engaging in both geographic and product diversification. Thus, we would expect the complementarity of product and geographic diversification may be even greater in for-profit forms of organization.

Conclusion Diversification has been a topic of intense scholarly debate for decades (Wiersema & Bowen, 2008). Although the majority of earlier work involving both product and geographic diversification focused exclusively on large, U.S.-based manufacturing corporations, the

Kistruck et al. / Diversification in Charitable Organizations   521

field, over time, has expanded to include different organizational sizes, industry classifications, and corporations headquartered in foreign countries (Geringer et al., 2000; Hitt et al., 2006). We have attempted to broaden the field even further by exploring diversification in the context of charitable organizations. The drivers of diversification, the location and product choices, and the level of resource dependency all differ from those of for-profit corporations, thereby providing a unique setting in which to explore the theoretical tenets regarding the link between diversification and efficiency. While our study has attempted to make a significant contribution by examining diversification in the context of charitable organizations, we hope it will also spur a number of future research directions. Although efficiency is an increasingly important constraint on charitable scope, it is important to reemphasize that it is an incomplete measure of overall charitable performance. While efficiency as measured in our study provides an indication of the ability of charities to maximize program spending on charitable activities as opposed to fund-raising and administrative overhead, it provides very little information about the quality of social output being produced (Moore, 2000). Unfortunately, attempting to measure this aspect of charitable performance continues to be extremely problematic within the field (Herman & Renz, 1999; Sawhill & Williamson, 2001) and is beyond the scope of our study as well. While researchers and practitioners alike continue to rely upon efficiency measures in their evaluations of charitable performance (Bennett & Savani, 2003), we strongly encourage future studies to try to capture this elusive component and particularly its relationship to diversification activities. As mentioned in our study, while there are a number of different types of nonprofit organizations such as trusts, private foundations, and cooperatives (Hall, 2006), the scope of our study was limited to charitable nonprofits. Future studies examining the variance in both diversification patterns and performance outcomes of different types of nonprofit organizations may further contribute to our knowledge of geographic and product diversification. In particular, studies involving nonprofit hospitals and educational institutions that rely on feefor-service rather than donative revenue models (Goodrick & Salancik, 1996), as well as social enterprises as emerging organizational hybrids spanning the for-profit and nonprofit sectors (Mair & Marti, 2006), may be particularly interesting for bridging the intersectoral and intrasectoral domains. Additionally, cross-sector partnerships between the charitable and for-profit sectors have been growing rapidly over the past two decades (Selsky & Parker, 2005). While many such partnerships represent only a loose association (Austin, 2000), a number of them are much more strategic and involve instances of joint geographic and product diversification. For instance, many multinational corporations are increasingly interested in exploring base-ofthe-pyramid market opportunities as new regions for economic growth (Prahalad, 2004). However, research has suggested that partnering with charities to enter least-developed markets may prove advantageous given their experience in operating within such institutional environments (Webb et al., 2010). Furthermore, many for-profit corporations, in their efforts at engaging in corporate social responsibility, are codeveloping new socially relevant products and services with the charitable sector as a means of achieving both financial and social gains. Thus, further integration of diversification research combining both the for-profit and nonprofit sectors could provide a great deal of theoretical and practical insight.

522    Journal of Management / February 2013

Appendix A Endogeneity Analysis Table A.1 Summary Results for First-Stage Regressions Variable GD PD GD2 PD2 GD × PD GD2 × PD GD × PD2 GD2 × PD2

Shea Partial R2

Partial R2

F(9, 17,836)

p Value

.0004 .0017 .0013 .0021 .0005 .0004 .0004 .0019

.0250 .0086 .0056 .0295 .0059 .0059 .0196 .0183

50.78 17.19 11.08 60.28 11.76 13.68 39.68 36.89

.0000 .0000 .0000 .0000 .0000 .0001 .0000 .0000

Note: GD = geographic diversification; PD = product diversification. F value > 10 indicates that the instrumental variables we used are strong.

Logic for the Selection of Instrumental Variables The following variables were selected as instrumental variables based on the logic provided below: Board size. Charity board members are generally unpaid and thus do not represent direct administrative costs to the charity. They are also typically selected not for functional skills (i.e., accounting, marketing, etc.) but rather for their passion for social causes and willingness to serve. However, larger boards are more likely to represent a greater diversity of social interests, and thus a larger number of board members in positions of authority increase the direct likelihood of diversification into different social interests (either in different services or different geographic regions or both). Number of board members not at arm’s length with each other. Again, boards that have fewer members who are related to one another are most likely to represent a greater diversity in social interests. Such diversity would impact efficiency not necessarily directly but, rather, indirectly through the pursuit of more diversified activities. Amounts receivable from individuals and organizations not at arm’s length from the charity. These are typically related parties who have committed to donating funds to the charity but have not yet deposited the funds. As opposed to this new donation changing efficiency levels of existing activities directly, it more likely indirectly impacts efficiency through diversification in that individuals or organizations are more likely to bring their own set of social interests, which if somewhat different from those of the initial group would result in increased levels of diversification. Amounts payable to individuals and organizations not at arm’s length from the charity. Conversely to the above argument, these are related parties who are owed funds. While this may not directly influence efficiency levels, it may cause such related parties to become more financially interested in the organization and less risk averse in the pursuit of the charity diversifying into new activities before their funds are repaid. Thus, this may have a negative direct effect on diversification. Number of compensated positions. A greater number of compensated positions does not necessarily mean that the organization is more or less efficient in terms of the direct relationship. As opposed to for-profit organizations, charities do not necessarily exhibit any economies-of-scale relationships with (continued)

Kistruck et al. / Diversification in Charitable Organizations   523

Appendix A (continued) a greater number of employees. However, in echoing the above logic, a great number of compensated positions would indicate a greater financial investment on the part of organizational decision makers who may be more inclined to avoid increased diversification at the risk of having their compensation negatively affected. Funding through international contracts. While most charities diversify internationally through onthe-ground efforts, there remains a portion that contract indirectly through local charities. As such, this funding serves more as a flow-through mechanism and would not be expected to significantly impact efficiency directly. However, operating such contracts may increase the likelihood, through exposure, of the charity diversifying directly into the region in the future. Funding through international gifts. Similar to the arguments above, such transactions are a more flow-through mechanism that would be expected to have minimal effects on efficiency but significant effects on the likelihood of direct diversification by the charity. Funding through other international means. Again, this variable covers the remaining category of charities in the data that are indirectly involved in international activities that would be more likely to significantly impact diversification than have a significant direct effect on efficiency. Quick ratio. The quick ratio (current assets minus current liabilities) represents a charity’s ability to immediately retire its debts. While we would not expect this balance sheet effect to have a strong direct effect on how efficiently the organization performs from a gain and loss perspective, it is likely to affect the likelihood of diversification if the organization is facing bankruptcy pressures. However, we feel that this is a different effect than the broader measure of organizational slack, given that quick ratio is a measure of short-term liquidity while organizational slack is a much broader measure of asset underutilization, which would be related to the long-term impacts of diversifying into new products or geographic regions.

Table A.1.1 First-Stage Regression for Geographic Diversification (GD) Geographic Diversification Board size Number non-arm’slength Amount receivable Amount payable Number compensated positions International exposure (contract) International exposure (gifts) International exposure (other) Quick ratio Intercept

b

Standard Error

z

p > |z|

0.120183 –0.180156

0.0256063 0.0430625

4.69 -4.18

0.000 0.000

0.0699922 -0.2645627

0.1703738 -0.0957493

0.0149935 -0.0428787 0.3850614

0.0093185 0.0081038 0.0251578

1.61 -5.29 15.31

0.108 0.000 0.000

-0.0032716 -0.0587629 0.3357497

0.0332586 -0.0269944 0.4343731

0.0024371

0.0083202

0.29

0.770

-0.0138714

0.0187455

-0.0696451

0.0117458

-5.93

0.000

-0.0926679

-0.0466222

0.0530864

0.0090435

5.87

0.000

0.0353603

0.0708126

-0.2115447 -1.042275

0.0598407 0.1410583

-3.54 -7.39

0.000 0.000

-0.3288383 -1.318763

-0.0942511 -0.7657876

Note: To conserve space, only the relevant portion of the output is presented.

95% Confidence Interval

524    Journal of Management / February 2013

Table A.1.2 First Stage Regression for Product Diversification (PD) Product Diversification

b

Standard Error

Board size Number non-arm’slength Amount receivable Amount payable Number compensated positions International exposure (contract) International exposure (other) International exposure (gifts) Quick ratio Intercept

0.007334 0.0305257

z

p > |z|

95% Confidence Interval

0.0034866 0.0058634

2.10 5.21

0.035 0.000

0.0004999 0.0190327

0.014168 0.0420186

-0.0051213 0.0000539 -0.0058537

0.0012688 0.0011034 0.0034255

-4.04 0.05 -1.71

0.000 0.961 0.087

-0.0076083 -0.0021089 -0.0125681

-0.0026343 0.0022168 0.0008606

0.0023986

0.0011329

2.12

0.034

0.000178

0.0046192

0.0070318

0.0012314

5.71

0.000

0.0046182

0.0094454

0.0025374

0.0015993

1.59

0.113

-0.0005974

0.0056723

-0.0603878 -0.2457502

0.008148 0.0192067

-7.41 -12.80

0.000 0.000

-0.0763586 -0.2833972

-0.0444169 -0.2081032

Note: To conserve space, only the relevant portion of the output is presented.

Table A.2 IV (2SLS) Estimation Estimates efficient for homoscedasticity only Statistics consistent for homoscedasticity only Number of observations = 17,850 F(11, 17,838) = 731.66 Probability > F = 0.0000 Total (centered) SS = 10,213.91916 Total (uncentered) SS = 13,041.32561 Residual SS = 4,331.9728

Centered R2 = .5758 Uncentered R2 = .6678 Root mean square error = .4926

Efficiency

b

Standard Error

z

p > |z|

GD PD GD2 PD2 GD × PD GD2 × PD GD × PD2 GD2 × PD2 CD SM RS Size _cons

-0.2976509 0.3097309 0.0427818 0.6452847 0.1621355 -0.0387379 0.1408541 -0.1115839 0.006578 0.0032182 -0.0498552 0.0013874 0.0137073

0.076816 0.1664795 0.0080919 0.3714258 0.0996837 0.0221863 0.2997119 0.0266779 0.0064827 0.0022236 0.023651 0.00235 0.0702229

-3.87 1.86 5.29 1.74 1.63 -1.75 0.47 -4.18 1.01 1.45 -2.11 0.59 0.20

0.000 0.063 0.000 0.082 0.104 0.081 0.638 0.000 0.310 0.148 0.035 0.555 0.845

95% Confidence Interval -0.4482076 -0.016563 0.0269219 -0.0826965 -0.033241 -0.0822222 -0.4465705 -0.1638716 -0.0061278 -0.0011399 -0.0962102 -0.0032185 -0.123927

-0.1470942 0.6360249 0.0586417 1.373266 0.357512 0.0047464 0.7282787 -0.0592961 0.0192839 0.0075763 -0.0035001 0.0059934 0.1513416

Note: GD = geographic diversification; PD = product diversification; CD = cultural distance; SM = subcontractor monitoring; RS = revenue shifting. Sargan statistic (overidentification test of all instruments): 0.342; c2(1); p value = .5589.

Kistruck et al. / Diversification in Charitable Organizations   525

Appendix A (continued) To test for presence of heteroscedasticity, we use following command: . ivhettest IV heteroscedasticity test(s) using levels of IVs only Ho: Disturbance is homoskedastic Pagan-Hall general test statistic: 2.925 c2(13), p value = .9982 The insignificant p value indicates there is no problem of heteroskedasticity (Baum, Schaffer, & Stillman, 2007a; Pagan & Hall, 1983). Once we have tested heteroskedasticity, we then move on to test autocorrelations (serial correlations; Arellano & Bond, 1991; Baum et al., 2007a) using the following command: . abar Arellano-Bond test for AR(1): z = 1.88 Pr > z = 0.06 The Arellano-Bond test indicates there is no threat to autocorrelations (serial correlations; Arellano & Bond, 1991; Bascle, 2008; Baum et al. (2007a), as p > .05. However, as it is only marginally greater than .05, it may be prudent to obtain robust estimates (though this is not mandatory). Following command estimates ivreg2 model robust to autocorrelations; the option bw(2) ensures this. . ivreg2 neweff_raw ( Gscen Pdcen GSQ PDQ GSXPD GSQXPD GSXPDQ GSQXPDQ=LnBS lnnal LnAR LnAP Lncompos Lnintcont lnintgift lnitoth QuickR) aveCD SM RS org_size, ffirst r bw(2)

Note 1. Our interactions were graphed using the range of -1 standard deviations to +2 standard deviations to accurately depict the actual range of our data as constrained by the means, standard deviations, and minimum and maximum values in Table 1.

References Arellano, M., & Bond, S. 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies, 58: 277-297. Austin, J. 2000. Strategic collaboration between nonprofits and businesses. Nonprofit and Voluntary Sector Quarterly, 29: 69-97. Austin, J. E., Stevenson, H. H., & Wei-Skillern, J. 2006. Social and commercial entrepreneurship: Same, different, or both? Entrepreneurship Theory and Practice, 30: 1-22. Bartlett, C., & Ghoshal, S. 1989. Beyond global management: Transnational solution. Boston: Harvard Business School Press.

526    Journal of Management / February 2013 Bascle, G. 2008. Controlling for endogeneity with instrumental variables in strategic management research. Strategic Organization, 6: 285-327. Baum, C. F., Schaffer, M. E., & Stillman, S. 2007a. Enhanced routines for instrumental variables/GMM estimation and testing. Working paper no. 667, Boston College Department of Economics, Chestnut Hill, MA. Baum, C. F., Schaffer, M. E., & Stillman, S. 2007b. ivreg2Stata module for extended instrumental variables/2SLS, GMM and AC/HAC, LIML and k-class regression. Retrieved from http://ideas.repec.org/c/boc/bocode/s425401 .html Baysinger, B., & Hoskisson, R. E. 1989. Diversification strategy and R&D intensity in multiproduct firms. Academy of Management Journal, 32: 310-332. Bennett, R., & Savani, S. 2003. Predicting the accuracy of public perceptions of charity performance. Journal of Targeting, Measurement and Analysis for Marketing, 11: 326-342. Boli, J., & Thomas, G. M. 1999. Constructing world culture: International nongovernmental organizations since 1875. Stanford, CA: Stanford University Press. Boris, E. T. 1998. Myths about the nonprofit sector. In E. T. Boris (Eds.), Charting civil society: 1-4. Washington, DC: Urban Institute. Brody, E. 1996. Agents without principals: The economic convergence of the nonprofit and for-profit organizational forms. New York Law School Law Review, 40: 457-536. Buckley, P. J., & Casson, M. 1976. The future of the multinational enterprise. London: Macmillan. Callen, J. L., Klein, A., & Tinkelman, D. 2003. Board composition, committees, and organizational efficiency: The case of nonprofits. Nonprofit and Voluntary Sector Quarterly, 32: 493-520. Campa, J. M., & Kedia, S. 2002. Explaining the diversification discount. Journal of Finance, 57: 1731-1762. Capar, N., & Kotabe, M. 2003. The relationship between international diversification and performance in service firms. Journal of International Business Studies, 34: 345-355. Carroll, D. A., & Stater, K. J. 2009. Revenue diversification in nonprofit organizations: Does it lead to financial stability? Journal of Public Administration Theory and Behavior, 19: 947-966. Chandler, A. D. 1962. Strategy and structure: Chapters in the history of the industrial enterprise. Cambridge, MA: MIT Press. Chatterjee, S., & Wernerfelt, B. 1991. The link between resources and type of diversification: Theory and evidence. Strategic Management Journal, 12: 33-48. Contractor, F. 2007. Is international business good for companies? The evolutionary or multi-stage theory of internationalization vs. the transaction cost perspective. Management International Review, 47: 453-476. Coviello, N., & McAuley, A. 1999. Internationalization and the smaller firm: A review of contemporary empirical research. Management International Review, 39: 223-240. Crittenden, W., & Crittenden, V. 2000. Relationships between organizational characteristics and strategic planning processes in nonprofit organizations. Journal of Managerial Issues, 12: 150-168. Daniels, J., & Bracker, J. 1989. Profit performance: Do foreign operations make a difference? Management International Review, 29: 46-56. Dart, R. 2004. The legitimacy of social enterprise. Nonprofit Management and Leadership, 1: 411-424. de Andres-Alonso, P., Cruz, N. M., & Romero-Merino, M. E. 2006. The governance of nonprofit organizations: Empirical evidence from nongovernmental development organizations in Spain. Nonprofit and Voluntary Sector Quarterly, 35: 588-604. Dees, J. G., & Anderson, B. B. 2003. Sector-bending: Blurring lines between nonprofit and for-profit. Society, 40: 16-27. Dichter, T. 1999. Globalization and its effects on NGOs: Effloresence or a blurring of roles and relevance? Nonprofit and Voluntary Sector Quarterly, 28: 38-86. DiMaggio, P. J. 1986. Nonprofit enterprise in the arts: Studies in mission and constraint. New York: Oxford University Press. DiMaggio, P. J., & Powell, W. W. 1983. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48: 147-160. Doh, J. P., & Teegen, H. 2003. Globalization and NGOs: Transforming business, government, and society. Westport, CT: Praeger. Donovan, K. 2002. Billions lost in charity scandal. Toronto Star, November 12.

Kistruck et al. / Diversification in Charitable Organizations   527 Doukas, J. A., Kim, C. F., & Pantzalis, C. 2005. The two faces of analyst coverage. Financial Management, 34: 99-125. Doukas, J. A., & Lang, L. 2003. Foreign direct investment, diversification and firm performance. Journal of International Business Studies, 34: 153-172. Doz, Y. L., & Prahalad, C. K. 1991. Managing DMNCs: A search for a new paradigm. Strategic Management Journal, 12: 145-164. Drayton, W. 2005. The citizen sector transformed. In A. Nicholls (Eds.), Social entrepreneurship: New models of sustainable social change: 45-55. Oxford, UK: Oxford University Press. Dunning, J. H. 1988. The electric paradigm of international production: A restatement and some possible extensions. Journal of International Business Studies, 19: 1-31. Fox, J., & Brown, D. L. 1998. Assessing the impact of NGO advocacy campaigns on World Bank projects and policies: The struggle for accountability: The World Bank, NGOs and grassroots movements. Boston: MIT Press. Franko, L. G. 1989. Global corporate competition: Who’s winning, who’s losing, and the R&D factor as one reason why. Strategic Management Journal, 10: 449-474. Froelich, K. A. 1999. Diversification of revenue strategies: Evolving resource dependence in nonprofit organizations. Nonprofit and Voluntary Sector Quarterly, 28: 246-268. Galaskiewicz, J., Bielefeld, W., & Dowell, M. 2006. Networks and organizational growth: A study of community based nonprofits. Administrative Science Quarterly, 51: 337-380. Geringer, J. M., Beamish, P. W., & daCosta, R. C. 1989. Diversification strategy and internationalization: Implications for MNE performance. Strategic Management Journal, 10: 109-119. Geringer, J. M., Tallman, S., & Olsen, D. M. 2000. Product and international diversification among Japanese multinational firms. Strategic Management Journal, 21: 51-80. Ghemawat, P. 2003. Semiglobalization and international business strategy. Journal of International Business Studies, 34: 138-152. Glaser, J. S. 1994. The United Way scandal: An insider’s account of what went wrong and why. New York: Wiley. Goodrick, E., & Salancik, G. 1996. Organizational discretion in responding to institutional practices: Hospitals and cesarean births. Administrative Science Quarterly, 41: 1-28. Grant, R. M. 1987. Multinationality and performance among British manufacturing companies. Journal of Inter­ national Business Studies, 18: 79-89. Grønbjerg, K. A. 1993. Understanding nonprofit funding: Managing revenues in social services and community development organizations. San Francisco: Jossey-Bass. Hahn, J., & Hausman, J. 2005. Instrumental variable estimation with valid and invalid instruments. Working paper, MIT Department of Economics, Cambridge, MA. Hall, M., Lasby, D., Gumulka, G., & Tryon, C. 2006. Caring Canadians, involved Canadians: Highlights from the 2004 Canada Survey of Giving, Volunteering, and Participating. Ottawa, Ontario: Statistics Canada. Hall, P. D. 2006. A historical overview of philanthropy, voluntary associations, and nonprofit organizations in the United States, 1600-2000. In W.W. Powell & R. Steinberg, The nonprofit sector: A research handbook (2nd ed.): 32-65. New Haven, CT: Yale University Press. Hamilton, B. H., & Nickerson, J. A. 2003. Correcting for endogeneity in strategic management research. Strategic Organization, 1: 51-78. Hansmann, H. 1980. The role of nonprofit enterprise. Yale Law Journal, 89: 835-901. Hennart, J. F. 1982. A theory of multinational enterprise. Ann Arbor: University of Michigan Press. Hennart, J. F., & Larimo, J. 1998. The impact of culture on the strategy of multinational enterprises: Does national origin affect ownership decisions? Journal of International Business Studies, 29: 515-538. Herman, R. D., & Renz, D. O. 1999. Theses on nonprofit organizational effectiveness. Nonprofit and Voluntary Sector Quarterly, 28: 107-126. Hitt, M., Hoskisson, R. E., & Kim, H. 1997. International diversification: Effects on innovation and firm performance in product-diversified firms. Academy of Management Journal, 40: 767-798. Hitt, M. A., Tihanyi, L., Miller, T., & Connelly, B. 2006. International diversification: Antecedents, outcomes, and moderators. Journal of Management, 32: 831-867. Hofstede, G. 1980. Culture’s consequences. New York: Sage. Hoskisson, R. E., Eden, L., Lau, C. M., & Wright, M. 2000. Strategy in emerging economies. Academy of Management Journal, 43: 249-267.

528    Journal of Management / February 2013 Hoskisson, R. E., & Hitt, M. A. 1990. Antecedents and performance outcomes of diversification: A review and critique of theoretical perspectives. Journal of Management, 16: 461-510. Hoskisson, R. E., Hitt, M. A., Johnson, R. A., & Moesel, D. D. 1993. Construct validity of an objective (entropy) categorical measure of diversification strategy. Strategic Management Journal, 14: 215-235. Hymer, S. 1976. The international operations of national firms: A study of direct foreign investment. Boston: Massachusetts Institute of Technology. Jacquemin, A. P., & Berry, C. H. 1979. Entropy measure of diversification and corporate growth. Journal of Industrial Economics, 27: 359-369. James, E. 1983. How nonprofits grow: A model. Journal of Policy Analysis and Management, 2: 350-365. Jensen, M., & Meckling, W. 1976. Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3: 305-360. Johanson, J., & Vahlne, J. 1977. The internationalization process of the firm: A model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8: 25-34. Johnson, D. A. 1986. Confronting corporate power: Strategies and phases of the Nestle boycott. In L. Preston & J. Post (Eds.), Research in corporate social performance and policy: 323-344. Greenwich, CT: JAI Press. Jones, M. B. 2007. The multiple sources of mission drift. Nonprofit and Voluntary Sector Quarterly, 36: 299-307. Kaplan, R. S., & Norton, D. P. 1996. The balanced scorecard: Translating strategy into action. Boston: Harvard Business School Press. Keating, E. K., & Frumkin, P. 2003. Reengineering nonprofit financial accountability: Toward a more reliable foundation for regulation. Public Administration Review, 63: 3-15. Kelly, K. S. 1991. Fund raising and public relations. Mahwah, NJ: Lawrence Erlbaum. Kim, W. C., Hwang, P., & Burgers, W. P. 1989. Global diversification strategy and corporate profit performance. Strategic Management Journal, 10: 45-57. Kistruck, G., & Beamish, P. (2010). The interplay of form, structure, and embeddedness in social intrapreneurship. Entrepreneurship Theory and Practice, 34, 735-761. Kobrin, S. J. 1991. An empirical analysis of the determinants of global integration. Strategic Management Journal, 12: 17-31. Kogut, B. 1985. Designing global strategies: Profiting from operational flexibility. Sloan Management Review, 27: 27-38. Kogut, B., & Singh, H. 1988. The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19: 411-432. Kumar, M. S. 2009. The relationship between product and international diversification: The effects of short-run constraints and endogeneity. Strategic Management Journal, 30: 99-116. Li, J., Lam, K., & Qian, G. 2001. Does culture affect behavior and performance of firms? The case of joint ventures in China. Journal of International Business Studies, 32: 115-131. Light, P. 2004. Sustaining nonprofit performance. Washington, DC: Brookings Institute Press. Lindenberg, M. 1999. Declining state capacity, voluntarism, and the globalization of the not-for-profit sector. Nonprofit and Voluntary Sector Quarterly, 28: 147-167. Lu, J. W., & Beamish, P. W. 2001. The internationalization and performance of SMEs. Strategic Management Journal, 22: 565-586. Lu, J. W., & Beamish, P. W. 2004. International diversification and firm performance: The S-curve hypothesis. Academy of Management Journal, 47: 598-609. Lubatkin, M., & Chatterjee, S. 1994. Extending modern portfolio theory into the domain of corporate diversification: Does it apply? Academy of Management Journal, 37: 109-136. Mahoney, J. T. 2005. Economic foundations of strategy. Thousand Oaks, CA: Sage. Mair, J., & Marti, I. 2006. Social entrepreneurship research: A source of explanation, prediction, and delight. Journal of World Business, 41: 26-44. Margolis, J. D., & Walsh, J. P. 2003. Misery loves companies: Rethinking social initiatives by business. Administrative Science Quarterly, 48: 655-689. Mayer, K. J., & Salomon, R. M. 2006. Capabilities, contractual hazards, and governance: Integrating resourcebased and transaction cost perspectives. Academy of Management Journal, 49: 942-959. Moore, M. H. 2000. Managing for value: Organizational strategy in for-profit, nonprofit, and governmental organizations. Nonprofit and Voluntary Sector Quarterly, 29: 183-204.

Kistruck et al. / Diversification in Charitable Organizations   529 Nelson, R., & Winter, S. 1982. An evolutionary theory of economic change. Cambridge, MA: Harvard University Press. Nielsen, R. P. 1986. Piggybacking strategies for nonprofits: A shared costs approach. Strategic Management Journal, 7: 201-215. Odom, R. Y., & Boxx, W. R. 1988. Environment, planning processes, and organizational performance of churches. Strategic Management Journal, 9: 197-205. O’Regan, K. M., & Oster, S. M. 2000. Nonprofit and for-profit partnerships: Rationale and challenges of crosssector contracting. Nonprofit and Voluntary Sector Quarterly, 29: 120-140. Oster, S. 1996. Nonprofit organizations and their local affiliates: A study in organizational forms. Journal of Economic Behavior and Organization, 30: 83-95. Owens, E., & Shaw, R. 1972. Development reconsidered: Bridging the gap between government and people. Toronto, Ontario, Canada: Lexington Press. Pagan, A. R., & Hall, A. D. 1983. Diagnostic tests as residual analysis. Econometric Reviews, 2: 159-218. Palepu, K. 1985. Diversification strategy, profit performance and the entropy measure. Strategic Management Journal, 6: 239-255. Palich, L. E., Cardinal, L. B., & Miller, C. C. 2000. Curvilinearity in the diversification–performance linkage: An examination of over three decades. Strategic Management Journal, 21: 155-175. Penrose, E. G. 1959. The theory of the growth of the firm. New York: Wiley Peterson, P. A. 1986. From impresario to arts administrator. In P. DiMaggio (Eds.), Nonprofit enterprise and the arts: 161-183. New York: Oxford University Press. Pfeffer, J. & Salancik, G. R. 1978. The external control of organizations: A resource dependence perspective. New York: Harper & Row. Powell, W. W., & Friedkin, R. 1986. Politics and programs: Organizational factors in public television decision making. In P. DiMaggio (Eds.), Nonprofit enterprise in the arts: 245-278. New York: Oxford University Press. Powell, W. W., & Steinberg, R. 2006. Introduction. In W.W. Powell & R. Steinberg (Eds.), The nonprofit sector: A research handbook (2nd ed.): 1-12. New Haven, CT: Yale University Press. Prahalad, C. K. (2004). The fortune at the bottom of the pyramid: Eradicating poverty through profits. Upper Saddle River, NJ: Wharton School Publishing. Prahalad, C. K., & Brugmann, J. 2007. Cocreating business’s new social compact. Harvard Business Review, February: 80-90. Qian, G., Khoury, T., Peng, M., & Qian, Z. (in press). The performance implications of intra- and inter-regional geographic diversification. Strategic Management Journal. Quotah, E. 2004. Good to grow: Local charities seek out best ways to expand nationally. Chronicle of Philanthropy, August 19. Ramanujam, V., & Varadarajan, P. 1989. Research on corporate diversification: A synthesis. Strategic Management Journal, 10: 523-551. Ramaswamy, R. 1993. Multinationality and performance: An empirical examination of the moderating effect of configuration. Academy of Management Best Papers Proceedings: 142-146. Briarcliff Manor, NY: Academy of Management. Rose-Ackerman, S. 1982. Charitable giving and “excessive” fundraising. Quarterly Journal of Economics, 97: 193-212. Rugman, A. M., & Verbeke, A. 2004. A perspective on regional and global strategies of multinational enterprises. Journal of International Business Studies, 35: 3-18. Ruigrok, W., & Wagner, H. 2003. Internationalization and performance: An organizational learning perspective. Management International Review, 43: 63-83. Rumelt, R. P. 1974. Strategy, structure, and economic performance. Boston: Harvard Business School Press. Salamon, L. M. 1992. America’s nonprofit sector: A primer. New York: The Foundation Center. Sargan, J. 1958. The estimation of economic relationships using instrumental variables. Econometrica, 26: 393-415. Sawhill, J. C., & Williamson, D. 2001. Mission impossible? Measuring success in nonprofit organizations. Nonprofit Management and Leadership, 11: 371-386. Scott, K. 2003. Funding matters: The impact of Canada’s new funding regime on nonprofit and voluntary organizations. Ottawa, Ontario: Canadian Council on Social Development. Selsky, J., & Parker, B. 2005. Cross-sector partnerships to address social issues: Challenges to theory and practice. Journal of Management, 31: 849-873.

530    Journal of Management / February 2013 Shaver, J. M. 1998. Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect. Management Science, 44: 571-585. Shenkar, O. 2001. Cultural distance revisited: Towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32: 519-535. Sherlock, M. F., & Gravelle, J. G. 2009. An overview of the nonprofit and charitable sector. Washington, DC: Congressional Research Service. Shleifer, A., & Vishny, R.W. 1997. A survey of corporate governance. Journal of Finance, 52: 737-783. Staiger, D., & Stock J. H. 1997. Instrumental variables regression with weak instruments. Econometrica, 65: 557-586. Stinchcombe, A. L. 1965. Organizations and social structure. In J. G. March (Ed.), Handbook of organizations: 142-193. Chicago: Rand McNally. Stone, M., Bigelow, B., & Crittenden, W. 1999. Research on strategic management in nonprofit organizations: Synthesis, analysis, and future directions. Administration and Society, 31: 378-423. Tallman, S., & Li, J. 1996. Effects of international diversity and product diversity on the performance of multinational firms. Academy of Management Journal, 39: 179-196. Teece, D. J. 1982. Toward an economic theory of the multiproduct firm. Journal of Economic Behavior and Organization, 3: 39-63. Thomas, D. E., & Eden, L. 2004. What is the shape of the multinationality–performance relationship? Multinational Business Review, 12: 89-101. Thornton, J. 2006. Nonprofit fund-raising in competitive donor markets. Nonprofit and Voluntary Sector Quarterly, 35: 204-224. Tsai, W. 2000. Social capital, strategic relatedness and the formation of intraorganizational linkages. Strategic Management Journal, 21: 925-939. Twisk, J. W. R. 2004. Longitudinal data analysis. A comparison between generalized estimating equations and random coefficient analysis. European Journal of Epidemiology, 19: 769-776. Useem, M. 1987. Corporate philanthropy. In W. Powell (Ed.), The nonprofit sector: A research handbook: 340-359. New Haven, CT: Yale University Press. Webb, J. W., Kistruck, G., Ireland, R. D., & Ketchen, D. J., Jr. (2010). The entrepreneurship process in base of the pyramid markets: The case of multinational/non-government organization alliances. Entrepreneurship: Theory and Practice, 34: 555-581. Werner, S. 2002. Recent developments in international management research: A review of 20 top management journals. Journal of Management, 28: 277-305. Wernerfelt, B. 1984. A resource-based view of the firm. Strategic Management Journal, 5: 171-180. Wiersema, M. F., & Bowen, H. P. 2008. Corporate diversification: The impact of foreign competition, industry globalization, and product diversification. Strategic Management Journal, 29: 115-132. Williamson, O. E. 1979. Transaction-cost economics: The governance of contractual relations, Journal of Law and Economics, 22: 233-261. Wunderink, S. R. 2002. Individual financial donations to charities in the Netherlands: Why, how and how much. Journal of Nonprofit and Public Sector Marketing, 10: 21-40. Zahra, A. S., & George, G. 2002. Absorptive capacity: A review, reconceptualization and extension. Academy of Management Review, 27: 185-203.