06 Oct 2003  Research & Ideas

The Growth of the Social Enterprise

To branch or affiliate? Different organizational structures have different strategic implications for nonprofit expansion, say HBS’s Jane Wei-Skillern and Duke-based colleague Beth Battle Anderson.


Social enterprise groups are traditionally organized along one of two lines: The affiliation model favors decentralized control, while the branch model concentrates control at a central headquarters. Most social enterprise groups choose the branch model, even thought it often leads to slower growth. Why? HBS's Jane Wei-Skillern and Duke-based colleague Beth Battle Anderson discuss their analysis of some 300 social enterprises.

Tishler: How do you distinguish between branches and affiliates?

Wei-Skillern and Anderson: Branches are organizational units that are legally incorporated under the same 501(c) (3) as the central organization. Branches are managed and controlled by the central organization, with local staff operating the branch but reporting directly to the central office. Local branches may have their own advisory boards, but the governance responsibility lies primarily with the central board of directors.

In contrast, affiliates are independent, 501(c)(3) organizations that have an agreement with a coordinating, central organization to be part of an identifiable network. Each affiliate has its own governing board, as does the central organization. The affiliate relationship can range from loose to tight with respect to the financial and operational interactions between affiliates and the central organization. A loose affiliation generally refers to a network of organizations committed to exchanging knowledge, pursuing a similar mission, and implementing the same social programs or services, but with limited formal mechanisms for central control and few prescribed interactions between local sites and the center. Tight affiliates are quite similar to business franchises, where the central organization retains significant control over the network's brand and operations. Thus, the key distinctions between the two organizational forms, affiliates and branches, are ownership, governance, and control.

Nonprofits choosing the branch approach should consider the potential trade-offs between maintaining central control and having widespread impact.

Q: What makes a social enterprise more likely to choose branching?

A: While our survey did not explore the reasons for choosing branching over affiliation, some nonprofits may prefer branching because it allows the nonprofit leader to exert the most control since the new organizational units are established and managed centrally. This level of control is conducive to protecting the organization's brand and reputation and coordinating overall strategy and operations. A high degree of control might be particularly desirable when the program being implemented is complex or relies heavily on intangibles, an organization is in the early stages of growth and is interested in testing and refining its model in new locations, or a nonprofit has less ambitious, less geographically dispersed expansion plans.

In fact, our survey data show that pure branch organizations were smaller, grew at slower rates, and had less ambitious growth plans than affiliate organizations. While this finding is perhaps not surprising given that branching would likely entail a greater investment of resources by the central organization than expansion via affiliates would, it does suggest that nonprofits choosing this approach should consider the potential trade-offs between maintaining central control and having widespread impact.

Q: Are there one or two social enterprises that exemplify the characteristics that make them trend toward branching?

A: While not always the case, many branch organizations are tightly managed with very specific practices and identifiable cultures. For example, City Year, the national youth service organization, has programs in fourteen cities across the country. Its founders made a strategic decision from the start to spread a comprehensive organizational model that includes specific practices designed to promulgate a particular culture, commitment to diversity, strategy for corporate involvement, and other core elements of City Year's theory of social change. It is expanding geographically through branches, in part, because its founders believe its success has been dependent on intangibles (such as its distinctive culture) and tacit knowledge that could best be conveyed within a single organizational structure. Similarly, but on a much larger scale, the Salvation Army operates primarily as a branch organization, which is conducive to sustaining its very hierarchical, religious, and army-influenced culture.

Q: For social enterprises that choose the affiliate route, what are the likely factors that point them in that direction?

A: Nonprofits may adopt an affiliate model because they have a bias towards local ownership and autonomy or feel that expanding via affiliates will allow them to grow more quickly while still providing some brand benefits and quality control. While branch organizations, such as The Nature Conservancy, may choose to allow a significant amount of local autonomy by decentralizing authority and decision-making, affiliates are by nature locally governed and operated, which promotes local responsiveness, empowers entrepreneurial managers, and appeals to many nonprofit leaders.

Additionally, as our survey data supported, affiliate organizations do often grow more quickly than branches, which could be a result of several factors. As independent, local organizations, affiliates may have better success fundraising and mobilizing resources locally, and the fundraising responsibility may be more efficiently allocated between the central office and local affiliates. In fact, in our survey, relative to branch organizations, affiliates did report greater access to philanthropic funds as a result of their expansion efforts. Moreover, nonprofits expanding via affiliation may decide to partner with existing local organizations or networks of organizations, which can facilitate faster growth and better leverage outside resources. For example, in order to accelerate their expansion efforts, Jump Start, an early childhood education program that pairs college students with pre-schoolers struggling in Head Start, began entering into affiliate partnership agreements with colleges and universities to host the Jump Start program on their campuses.

Q: What is the major take-away you would like social enterprises to get from your research? What are the major implications in terms of building and running a nonprofit?

A: While perhaps not surprising, our research suggests that there is no single, optimal structure for nonprofit geographic expansion. Rather, different structures appear to have different strategic implications, and nonprofits must consider how these implications align with their particular organization and expansion goals. In addition to branch organizations being smaller, less ambitious, and growing more slowly, there were some notable differences across the branch, affiliate, and plural (organizations with both branches and affiliates) structures. For example, while plural organizations reported significantly greater benefits than pure branch or affiliate models, they also claimed to have faced greater challenges and seem to require stronger, more committed leadership. And branch organizations reported more challenges with human resource issues, while affiliates asserted greater challenges in governance.

While these patterns are not necessarily generalizable to all nonprofits, they do suggest that nonprofit leaders should anticipate some of the distinct challenges that seem to be associated with particular organizational structures.

Q: Did you come across any major surprises during your research?

A: The most surprising finding from our survey was that regardless of organizational structure, some of the anticipated benefits of geographic expansion failed to materialize, while other benefits seemed to exceed expectations. Economies of scale were actually less than anticipated for branch, affiliate, and plural organizations, and tapping into new funding sources appeared to be a significant benefit primarily for affiliates only. The fact that many of the key anticipated benefits for expansion were never realized is a stark reminder to nonprofit leaders to be realistic in their expectations of what benefits expansion will bring.

Building a brand was consistently one of the strongest motivators for expansion.

In contrast, across all organizational structures, the benefits from both brand and organizational learning were considerably greater than our survey respondents anticipated at the outset. This finding was particularly striking because building a brand was consistently one of the strongest motivators for expansion, yet the benefits nonetheless exceeded the expectations. By recognizing the potential benefits from brand and organizational learning from the beginning, nonprofit leaders may be able to increase their ability to capitalize on these factors through more deliberate planning and strategy.

Q: What is the next step in your research? Will you continue to focus on branching vs. affiliation?

A: In conjunction with this research, we have produced another paper with Professor J. Gregory Dees, Pathways to Social Impact: Strategies for Scaling Out Successful Social Innovations, that explores a broader range of strategies by which nonprofits can spread their social impact into new communities. Dees and Anderson plan to continue with conceptual research that will help nonprofit leaders spread their success in more timely, effective, and appropriate ways. Additionally, building on this research, Wei-Skillern is working on a long-term study on how multi-site nonprofits are structured and managed to achieve their goals. This research focuses on various structures ranging from branch to looser affiliate and network forms of organization. The study explores key management challenges of multi-site nonprofits and the facilitators and barriers to effectiveness of various multi-site, nonprofit structures. The goal of this research is to develop a framework to aid social sector managers in planning their growth strategies and managing multisite nonprofits more effectively.

Nonprofit Geographic Expansion: Branches, Affiliates or Both?

by Jane Wei-Skillern and Beth Battle Anderson

Throughout the nonprofit sector, countless innovative and effective programs remain limited to the immediate communities in which they were established and rarely, if ever, achieve any appreciable scale. Thus, there are at-risk children missing out on effective after-school programs, welfare recipients who lack access to new job training services, and innumerable others who do not benefit from high impact social services and programs that exist in other locales. Therefore, many nonprofit observers and practitioners express the need for more successful organizations to extend their services and impact more broadly. In some instances, nonprofits may be able to spread their social impact through loose forms of diffusion, such as informally through social or professional networks (Edwards and Hulme, 1992), or by means of active dissemination with or without technical assistance (Dees, Anderson, and Wei-Skillern, 2002). Alternatively, many nonprofit organizations looking to extend their reach move into new territories by establishing local branches or affiliates. Additionally, some organizations spread geographically with a combination of both branches and affiliates, resulting in a plural form. Our focus here is on these latter approaches to nonprofit expansion, scaling into new communities through branch, affiliate, or plural structures.

This paper reports the results of a large-scale survey of nonprofit leaders currently engaged in or seriously considering expanding their organizations by establishing branches and/or affiliates in new locations. By capturing data on their organization's geographic expansion and the related motivations, benefits, and challenges, we sought to better understand the process of expansion and to generate new hypotheses for further research. Additionally, our aim is to help nonprofit leaders considering expansion make more informed decisions regarding a branch, affiliate, or plural structure. We present our findings here and conclude by proposing new hypotheses for further research and contemplation.

  • There is no single optimal organizational structure for geographic expansion.
  • Many nonprofit leaders may prefer expanding via branches, yet pure branch organizations are generally smaller, grow more slowly, and have less ambitious growth plans than those with affiliates.
  • Economies of scale from geographic expansion may be less than anticipated at the outset.
  • Regardless of structure, when expanding into new territories, the benefits from building a brand and organizational learning may exceed initial expectations.

Key Findings
Structural preferences, size, and growth preference for branches

One of the most important decisions facing nonprofit leaders seeking to expand their organization is what organizational structure to adopt as they enter new communities.

Table 2. Organization Structure Frequencies

Organization Structure Organizations Already Expanded (N) % Organizations ConsideringExpansion (N) %
Branch Organization 62 48% 24 43%
Affiliate Organization 37 28% 8 14%
Plural Organization 31 24% 6 11%
No Preference -- -- 18 32%

As Table 2 indicates, nonprofit leaders responding to our survey exhibited a strong preference for branch expansion. Among organizations that had already expanded, nearly half had done so via branches as opposed to affiliate or plural structures. Among organizations considering expansion, 43 percent indicated a preference for doing so via branches, compared with only 14 percent via affiliates and 11 percent as plural organizations. The remaining 32 percent expressed no preference.

Branches: Smaller, slower, and less ambitious
While our survey data indicates a preference for branch expansion among organizations that have already expanded, pure branch organizations were smaller, grew at slower rates, and had less ambitious growth plans than those with affiliates. Excluding two outliers (999 for branches and 850 for affiliates), the average number of locations for branch organizations was 7, significantly smaller than the average of 36 for plural and 32 for affiliate organizations. This finding is entirely consistent with Oster's research (1992, 1996), which suggests that expansion via franchises (affiliates) is more effective for achieving large-scale growth. Additionally, the pattern echoes Oster's (1996) observation that among America's 100 largest nonprofits on the basis of revenues in 1990, the average number of units per organization was 461 for affiliates versus 9 for branches.

Table 3. Organization Size by Organization Structure

  Branches (N=62) >Affiliates (N=32) Plural (N=32)
<= 5 42 10 12
6-10 8 5 6
11-30 9 7 7
31-50 1 4 0
51-100 1 2 3
>100 1 4 4
Mean* 7.1 32.1 30.1
S.D.* 11.7 49.3 50.1
Median* 3 13 10
Maximum* 75 245 215

*While included in the total counts for number of units, the averages, standard deviations, medians, and maximums reported for branches and affiliates exclude two outliers, 999 for branches and 850 for affiliates.

Furthermore, recent growth rates for the organizations in our survey seem to support the argument that looser relationships can enable faster growth, with affiliate organizations growing twice as fast as branch organizations (16.9 percent CAGR vs. 8.3 percent ), and plural organizations falling in between at 14.8 percent . Finally, affiliate and plural organizations were much more ambitious in their future growth plans. Nearly half of the branch organizations did not anticipate establishing any new branches over the next five years, and none planned to open more than five. In contrast, over half of all plural organizations and 37 percent of affiliate organizations planned on establishing more than five new sites in the next five years, with the maximum being 108 for plurals and fifty for affiliates. Notably, overall, plural organizations planned on establishing far more affiliates than branches in the coming years. The data suggests that nonprofit leaders with ambitious growth goals may be better able to achieve those goals through affiliate or plural strategies, while those leaders who seek only to establish a handful of additional sites may opt to expand via branches. Consistent with Chandler's pioneering work on organizational structure, the structural form should follow from the organization's strategy (1962).

Taken together, the findings that our survey respondents prefer expansion via branches despite the fact that a branching strategy seems to entail slower, less ambitious growth presents a paradox. While the preference for branches can perhaps be understood by the fact that branches enable the central office to exert more control (Oster, 1996), this preference may be at odds with the objective to achieve greater social impact through widespread expansion. Since the data seem to indicate that affiliate or plural models are better than pure branching at enabling large-scale growth, this finding suggests that nonprofit leaders must reconcile their own objectives and preferences with the best interests of their organization and society. At the very least, the data suggest that a nonprofit leader must seriously consider which structure will most effectively enable the organization to reach its growth goals, and ultimately, achieve the greatest social impact.

Finally, these results are even more interesting when coupled with our comparison of motivations, benefits, and challenges according to size (small = five or fewer additional sites and large = more than five additional sites). For larger organizations, having both a leader and a board with a strong desire to grow into new communities played a significantly greater role in motivating expansion efforts. Moreover, all of the organizational benefits we measured were significantly or marginally significantly greater for organizations with more than five sites compared to those organizations with five or fewer sites. These findings suggest that regardless of structure, nonprofits may need to reach a certain size before they will really begin to reap the organizational benefits associated with geographic expansion. While the consistency across factors is striking, these results are not altogether surprising given that larger organizations have most likely committed more resources and effort to achieving these benefits. Furthermore, more sites simply increase the likelihood that organizational benefits are realized in at least some sites. From the survey data, it is not possible to determine whether the benefits for larger organizations were greater relative to the resources invested, but these results nonetheless are notable since 68 percent of branch organizations were relatively small.

Excerpted with permission from the HBS working paper Social Enterprise Series No. 27: Nonprofit Geographic Expansion: Branches, Affiliates, or Both? by Jane Wei-Skillern and Beth Battle Anderson, 2003.

About the authors

Carla Tishler is the director of HBS Baker Library Information Products.

Jane Wei-Skillern is an assistant professor at Harvard Business School.

Beth Battle Anderson is a lecturer and managing director of the Center for the Advancement of Social Entrepreneurship at the Fuqua School of Business at Duke University.