Patterns of B Corps Certiﬁcation: The Role of Institutional, Economic, and Political Resources

: This paper explores the certiﬁcation of companies as B Corps from 2007 through 2016, the ﬁrst 10 years of certiﬁcation. B Corps are for proﬁt companies that promise to “Be a Force for Good” in our society. Over 2600 companies in over 50 countries are certiﬁed as B Corps, responding to demands for higher accountability, ethical behavior, and contributions to their environment and community. We focus here only on B Corps in the United States and analyze a state-level database we have developed of 851 companies that became certiﬁed in the ﬁrst 10 years of certiﬁcation, between 2007 and 2016. In the paper we ask: What conditions in the macro environment facilitate the spread of B Corps certiﬁcation? This paper uses the framework of resource dependence theory and institutional theory to explore the di ﬀ usion of certiﬁcation. We hypothesize that institutional, economic, and political resources in the external environment provide conditions that support B Corps certiﬁcation.


Introduction
B Corps certification is a promise by firms to be a "better corporation" by committing to improve their practices in relation to their employees, external stakeholders, and the environment. B Corps are certified by third party nonprofit B Lab, and promise to "Be a Force for Good" in our society while still seeking profits. They offer profit-driven solutions to some of society's most difficult challenges, including climate change, rising levels of inequality, political upheaval, and social injustice. B Corps respond to these challenges by promoting values that emphasize a commitment to social and environmental stewardship. B Corps practice a maximum view of corporate social responsibility (CSR) known as "behaviors and practices that extend beyond immediate profit maximization goals and are intended to increase social benefits or mitigate social problems for constituencies external to the firm" [1] (p. 925). Patagonia, for example, considers itself to be an activist company and promotes environmental stewardship, and a commitment to its people, far more than the profits it seeks through its specialty outdoor clothing business. According to its founder Yvon Chouinard [2] (p. 230): When you get away from the idea that a company is a product to be sold to the highest bidder in the shortest amount of time, all future decisions in the company are affected. The owners and the officers see that since the company will outlive them, they have responsibilities beyond the bottom line. Perhaps they will even see themselves as stewards, protectors of the corporate culture, the assets, and of course the employees.
King Arthur Flour, another B Corps, is a baking goods company that has long invested in its people. As an employee-owned company, they continually create high quality impacts for their employees and community. Marikar notes that "King Arthur employee benefits include job sharing, flexible work schedules, and remote office options; 40 h a year of paid volunteer time; free professional-development courses; and reimbursement accounts for things like smoking cessation, weight management, and child care" [3]. Patagonia and King Arthur provide just 2 of the over 2300 global examples of B Corps that are responding to demands for higher accountability, ethical behavior, and contributions to their environment and community.
Research on B Corps certification has thus far looked at the role of legislation in creating and supporting the form [4,5]; the nascent history and impact of certification on the bottom line [6,7]; and the motives for B Corps certification [8,9]. What is less understood is how the external environment impacts the diffusion of B Corps certification. Hickman, Byrd, and Hickman (2014) begin to explore this phenomenon by looking at several variables that may predict higher rates of certification [10]. They conclude by suggesting that regions with higher educational attainment, progressive political environments, healthier populations, higher ranks of greenness, and that are diverse are more likely to have higher certification rates. Their paper does not, however, provide a theoretical grounding for why these variables are of importance.
Research suggests that organizations depend on their external environment for survival as critical resources are located outside the firm. Resources in the environment include organizational networks, capital flows, access to employees, suppliers, customers, and business friendly political and regulatory policies. Firms often adapt their organizational practices and policies to facilitate resource flows [11,12]. This view, called resource dependency, suggests that when the external environment has resources to support particular practices, firms may be more likely to adopt the practice. B Corps certification may be more likely to spread when the resources in an organization's external environment are plentiful.
In this paper we explore patterns of B Corps certification by asking: What conditions in the macro environment facilitate the spread of B Corps certification? The paper first provides a theoretical framework that explores the literature and develops hypotheses that suggest relationships between factors in the external environment: institutional, economic, and political, and the diffusion of B Corps certification. Next, we review the materials and methods of the study describing how we develop and analyze a database of 851 organizations that were certified between 2007 and 2016. Our results section evaluates our hypotheses and discusses the findings based on patterns of certification throughout time, across the organizational population, and within the context of broader economic and political factors. The paper then discusses the connections that were made to the literature. Finally, we conclude with our main contributions and a discussion of avenues for future research.

Theoretical Framework
As resource dependence theory asserts, organizations depend on resources in their external environment for survival, and, thus adapt their behavior in order to facilitate resource flows from the external environment [11,12]. These resources might come from the institutional, economic, or political environment. Institutional resources include resources that provide access to information and networks that are found by being a part of the community or group of B Corp certified organizations. Economic resources are found when access to capital provides support for the future growth of firms and resources to support certification. Political resources can include political ideologies and policies that support CSR behavior, including certification. Below, we outline the external conditions that could impact the spread of certification.

B Corps Certification and Institutional Factors
In order to understand whether the diffusion of B Corps certification is impacted by the external environment we first look at the institutional environment, here defined as the informal environment that makes up a B Corps organizational field. An organization field consists of "a community of organizations that partakes of a common meaning system and whose participants interact more frequently and fatefully with one another than with actors outside the field" [13] (p. 56). For the purposes of this paper the field includes actors such as B Corps, other social enterprises considering certification, professional associations that may support certification, local groups of certifiers (like the B Local communities), and third parties including B Lab and other interest groups that may follow the community. Research on the spread of CSR practice suggests that institutional pressures among organizations that wish to promote social and economic welfare may lead to the spread of CSR practices [1,14], including B Corps certification. Tolbert and Zucker (1983) argue that organizations are seen as legitimate actors the more they adopt and enact institutional norms that are valued at the field level [15]. By incorporating institutionalized norms into their firm, organizations can gain legitimacy within their field, thereby increasing their resources, their potential for survival, and avoidance of unwanted outside scrutiny that may be detrimental to the firm [16,17]. Organizational decision making-for example, whether or not to adopt a specific practice, like B Corps certification-is impacted by the social relationships organizational members have in the institutional environment. Social relationships can consist of relational networks (communication led by B Lab, B Local Groups, conferences, informal networks) and broader market environments that define institutional environments [18].
According to Budros (2004), "the expectation is that, as similar organizations use each other as a social frame of reference, new practices that enter a cultural network will flow through it" [19] (p. 361). Under this view, certification begets certification likely due to the spread of information about the legitimacy and process of certification.
We anticipate, therefore, that B Corps certification will increase over time, but accelerate in later years. The more that certification is perceived to be legitimate, the more likely future adopters will be to mimic the actions of prior adopters [19].
Hypothesis 1A (H1A). B Corps Certification will increase over time with later adoption of certification accelerating.
Research also suggests that early adopters of organizational practices can impact later adopters. Montiel and Husted (2009) argue that the role of early adopters of environmental programs in Mexico is critical to the future institutionalization of the practice [20]. Early adopters can facilitate the diffusion of new organizational practices. Early adopters of B Corps certification can act as resources for future adopters by providing access to information necessary to consider certification. According to Russo (2003), "one of the most important forms of social capital is the acquisition and dissemination of scarce information" [21] (p. 321). B Corps certification is one example of a new organizational practice with scarce and ambiguous information available to potential adopters. In in the early period, after certification became a viable option to organizations who wished to pursue a social or environmental goals, information about what a B Corps entailed and the process of becoming a B Corps needed to be disseminated to potential new adopters. This exchange of information often flows through informal and formal peer networks that are more likely to be found in spatial proximity [22]. If we anticipate that early adopters will be more likely to spread information about certification and provide early legitimacy to B Corps certification for those in social proximity, we can anticipate that first adopters will impact future B Corps certification.
Hypothesis 1B (H1B). States with early adopters will have higher numbers of future B Corps certification than states without early adopters.

B Corps Certification and the Role of Economic Resources
Resources found in the economic environment are likely to promote or discourage certification behavior. B Corps pursue social and environmental objectives in addition to traditional market based, for-profit goals. These organizations, therefore, are embedded in the broader economic system and should respond to traditional market mechanisms that encourage or discourage economic activity. According to Taylor (2005), certification schemes, "must operate in the mainstream market to effectively pursue their objectives" [23] (p. 130). The broader economic market can impact the extent to which organizations pursue socially responsible behavior. Campbell (2007) asserts, "if firms are operating in an economic climate where, for instance, inflation is high, productivity growth is low, consumer confidence is weak, and in short, it appears that it will be relatively difficult for firms to turn a healthy profit in the near term, they will be less likely to behave in socially responsible ways than would otherwise be the case" [14] (p. 952). Studies have shown that poor economic climates leave firms with fewer resources to invest in corporate social responsibility [23]. When organizations make the decision to become B Corps certified they are more likely to do so if the economic environment has the resources to support future growth. Becoming certified comes with a cost. B Corps certification fees include an annual fee that depends on the size of the firm. Organizations are likely to weigh these costs when making the decision whether or not to certify. When economic resources in the environment are plentiful, new enterprises are more likely to form. We use the growth of new firms as a proxy for strong economic climates. We argue that organizations will be more likely to certify when the broader economic climate is strong.

Hypothesis 2 (H2).
There will be a positive relationship between # of enterprises and # of new B Corp Certifications.

B Corps Certification and the Role of Political Resources
The political context, also, could have a strong influence on certification behavior. There is substantial support to suggest political ideology impacts corporate social responsibility behavior [9,20,23]. For example, Detomasi (2008) looks at political ideology from a political institutional perspective focusing on a country's political environment [24]. He argues that a firm's decision to adopt CSR behavior stems, in part, from the home country's political ideology and structure. Detomasi (2008) asserts, "Political pressure engendered via the ballot box drives public policy; and elected officials design and/or support particular policies because they believe those policies will maximize positive electoral gains and ensure continued tenure, or access to, elected office" [24] (p. 811). If we view certification as a corporate social responsibility activity, it follows that the political environment could impact certification rates. We hypothesize that liberal leaning states, specifically measured through percentage of vote for Democrats, will have higher numbers of new B Corps certifications. DiGuili and Kostovetsky (2014, p. 4) assert that "The Democratic Party platform places more emphasis on CSR-related issues such as environmental protection, anti-discrimination laws and affirmative action, employee protection, and helping the poor and disadvantaged" [25]. Furthermore, their study found that firms in Democrat leaning states had more socially responsible corporate behavior. Liberal ideology has been found to be more supportive of CSR behavior among organizations including certification behavior (see, e.g., Hickman et. al. [10]).

Hypothesis 3 (H3).
The higher the percentage of the presidential vote for Democrats, the higher the number of new B Corp certifications.

Materials and Methods
One of the challenges that immediately emerged for the authors as this study moved from conceptualization to operationalization was to identify available data needed to test the above hypotheses on for-profit companies certified as B Corporations by the B Lab, a non-profit organization. This certification is described as certifying business in the same way that other certifications certify products (i.e., Fair-Trade certifying coffee or Forest Stewardship Council certifying paper products). The certification takes place based on a rigorous assessment of social and environmental performance, accountability and transparency [26]. This process first includes successful completion of the B Impact Assessment (requiring a minimum score of 80 out of 200), which includes categories measuring impact related to the environment, workers, customers, community, and governance, as appropriate for each company. The second step involves a legal requirement, which ensures the protection of the benefit mission by determining what is required based in the state or area of incorporation and on the present corporate structure. Companies that go through these steps finally sign the B-Certification Agreement and pay an annual fee to B Lab, which is pro-rated based on annual sales revenue. Certification lasts for two years, at which time corporations can apply for re-certification if they choose [26].
The B Lab website (http://www.bcorporation.net/) provides a searchable database of currently certified companies, totaling 2358 B Corporations in 44 states, 2 territories and the District of Columbia, 50+ countries, and roughly 130 industries as of January 2018 when this study was initiated [26]. However, this site does not provide a way to easily aggregate the collective data in a manageable form (i.e., in a spreadsheet or a format ready for use in a statistical software such as SPSS or Stata). Thus, the authors engaged in a multi-step process to construct a new dataset of corporations certified as B Corps to enable us to examine meaningful evidence regarding the hypotheses proposed in the previous section.
The data utilized in this study were first located as part of dataset on the Data.World website in March 2017 [27]. The file on this site had been posted by B Lab itself and was current as of February 2017. However, it was quickly determined to be incomplete, and was thus not usable in its form at that time. Once this determination was made, the dataset was developed manually, by identifying the information necessary to extract from each individual entry through the "Find a B Corporation" search tool on the B Lab website. The file extracted from Data.World was utilized as a base file once the accuracy of the data was validated, and from there added the other certified B Corps, along with the additional variables needed to help assess the hypotheses. At this point, the authors decided to focus only on B Corps certified in the U.S., as data for the hypotheses would be exceptionally difficult to locate for those non-U.S. B Corps.
Once this process was completed, there was a dataset consisting of information on 953 certified B Corps in 44 states, 2 territories and the District of Columbia. However, in subsequent evaluation, it was determined that there were additional discrepancies involving 93 of the corporations which were in the Data.World spreadsheet from March 2017 but were not on the B Corps website. Through analysis of company websites, it was determined that these were corporations that were no longer certified by B Corps. It is unclear whether they simply opted out or no longer met certification criteria, but given a lack of information, these corporations were excluded from the analyses. Furthermore, for purposes of consistency and comparability, seven additional corporations which were certified in early 2017 were excluded, since we were not able to include all corporations from that year. Finally, the two B Corps from Puerto Rico and the U.S. Virgin Islands (one each) were excluded from the analyses due to missing data issues for certain variables. This left a final tally of 851 currently certified B Corps that are included in the final analyses presented here.
To test the hypotheses described above, the following data were included in the final dataset:  Table 1). Year of certification and state of incorporation are straightforward to understand. B Corp concentration is a categorization of a percent of the total B Corps certified in each state in a given year, categorized as 1 = None, 2 = Low, 3 = High. A state that has at least one certified B Corps in 2007 or 2008 is defined as an early adopter state here. The number of private sector enterprises per state per year is an annual measure of economic activity in a state at a given point in time (March of each year). It is drawn from the Bureau of Labor Statistics annual National Business Employment Dynamics Data [28]. During the period of this study, 2007-2016, there were no certified B Corps that were public corporations. We use private sector enterprises as a measure of economic growth as it is the most appropriate measure given the fact that the majority of B Corps tend to be small, private enterprises. Finally, the percent of vote for the Democratic political candidate is a measure of the socio-political context of each state at a given point in time. This data was drawn from Federal Election Commission records and included the elections of 2004, 2008, 2012, and 2016 [29]. Given that these elections occur only every fourth year, it was decided to create an annualized measure for the percent Democratic vote in each state. The measure was created by utilizing a standard process-for the election years themselves, the percentage of vote for the Democratic candidate was converted to a proportion and utilized as is. For the other years, an average was calculated based on the gap in time between the previous and subsequent election-for example, the value for the year 2007 was calculated by taking 75% of the 2008 percentage and adding it to 25% of the 2004 percentage, while for the year 2010 it was calculated as 50% of 2008 and 50% of 2012.

Data Analysis
The dataset constructed in this study includes quantitative state-level annual panel data for a 10 year period (2007 through 2016). There is no missing data, meaning that the sample size is N = 510 observations, representing 10 years of 51 data points (i.e., 50 states plus District of Columbia). The hypotheses examine change over time (years) and/or variation across space (US states and DC), thus pooled OLS regression was determined as an appropriate technique for the statistical analyses undertaken. SPSS Version 25 was used for all analyses.

Results
Descriptive information for each of the variables is presented in Table 2  None 60.59% n/a n/a n/a 309 Low 27.65% n/a n/a n/a 141 High 11.76% n/a n/a n/a 60 Early Adopter State 0.27 n/a 0 1 510 Establishments (×100,000) Next, bivariate correlation coefficients for the association of each of the variables included in regression modelling is presented in Table 3. All relationships were statistically significant at the 0.001 level of significance; however, it is important to note that no coefficient was so strong as to provide concern over multicollinearity. The measures of number of B Corps certified by state per year and the state level B Corps concentration were the most strongly associated (r = 0.790), and the associations between these two variable and number of establishments in a state in a given year were also rather strong (r = 0.670 and 0.728, respectively). The political measure of percent Democratic vote in a state in a given year was modestly associated with each of the other three variables (ranging from r = 0.192 to 0.267).  As described above, Hypothesis 1A stated the following: B Corps Certification will increase over time with later adoption of the form accelerating. The regression analysis shows strong support for this hypothesis. The year of certification is highly significant by itself, but it is then reduced to insignificance when the annualized B Corps concentration variable is added to the model. In fact, adding the concentration variable increases R-square from 5% to 46% (see Table 4, Models 1 and 2). This suggests that B Corps certification does not follow predictable growth patterns, but is impacted by the concentration of B Corps in the population. The more B Corps in a state in a given year, the more likely future B Corps will exist. Here certification begets certification. Furthermore, if we look at the basic trend line (see Figure 1), there is a clear upward trend in the number of certified B Corps between 2007 and 2016. There may be some measured error here, since we are unable to account for those that are no longer certified; however, that number is relatively small and should be assumed to be spread out across the early years with some uniformity. Thus, the trend line observed in Figure 1  Furthermore, if we look at the basic trend line (see Figure 1), there is a clear upward trend in the number of certified B Corps between 2007 and 2016. There may be some measured error here, since we are unable to account for those that are no longer certified; however, that number is relatively small and should be assumed to be spread out across the early years with some uniformity. Thus, the trend line observed in Figure 1   Hypothesis 1B posits a positive relationship between early adopter status and number of new B Corp Certifications. This hypothesis was also strongly supported (see Table 4, Model 3). States that had B Corps certified in the first two years of this certification form had significantly more B Corps certified than those that did not have early adopters. Hypothesis 2 predicts there will be a positive relationship between number of enterprises and number of new B Corp Certifications. Economic resources in the environment that support the survival of enterprises will likely also support the certification of B Corps. In the panel regression analysis, this hypothesis was strongly supported (see Table 4, Model 3). As the number of Private Sector Establishments (PSEs) increases, the number of new B Corps in a state increases. The R-squared value strongly supports this hypothesis; the number of new enterprises accounts for 45% of the variance in the number of new B Corp Certifications.
Hypothesis 3 suggests that the political environment shapes certification patterns: The higher the percentage of the presidential vote for Democrats, the higher the number of new B Corp certifications. This hypothesis was supported, as seen in Table 4, Model 5. The number of B Corp certifications in a state in a given year is positively related to the percentage of the presidential vote for Democrats. Political resources are more likely to be found in liberal leaning states to support B Corps certification. Finally, a model incorporating each predictor was examined as well, with the expectation being that each hypothesized relationship would remain significant in this full model, while accounting for all other variables in the model. Results show that all variables remain highly significant, except for percent Democratic vote in the state in a given year. Further, this model substantially improves the R-square value, which rises to 0.62, or 62% of variance explained by the model (see Table 4, Model 6). Hypothesis 1B posits a positive relationship between early adopter status and number of new B Corp Certifications. This hypothesis was also strongly supported (see Table 4, Model 3). States that had B Corps certified in the first two years of this certification form had significantly more B Corps certified than those that did not have early adopters. Hypothesis 2 predicts there will be a positive relationship between number of enterprises and number of new B Corp Certifications. Economic resources in the environment that support the survival of enterprises will likely also support the certification of B Corps. In the panel regression analysis, this hypothesis was strongly supported (see Table 4, Model 3). As the number of Private Sector Establishments (PSEs) increases, the number of new B Corps in a state increases. The R-squared value strongly supports this hypothesis; the number of new enterprises accounts for 45% of the variance in the number of new B Corp Certifications.
Hypothesis 3 suggests that the political environment shapes certification patterns: The higher the percentage of the presidential vote for Democrats, the higher the number of new B Corp certifications. This hypothesis was supported, as seen in Table 4, Model 5. The number of B Corp certifications in a state in a given year is positively related to the percentage of the presidential vote for Democrats. Political resources are more likely to be found in liberal leaning states to support B Corps certification.
Finally, a model incorporating each predictor was examined as well, with the expectation being that each hypothesized relationship would remain significant in this full model, while accounting for all other variables in the model. Results show that all variables remain highly significant, except for percent Democratic vote in the state in a given year. Further, this model substantially improves the R-square value, which rises to 0.62, or 62% of variance explained by the model (see Table 4, Model 6).

Discussion
Our analysis dispels the assumption that certification trends reflect common patterns of organizational growth. We focus on two patterns discovered in the research that point to support for the impact of external factors on the diffusion of certification.
When we look at our hypothesized predictors we find strong support for institutional factors impacting the adoption of certification. Both the concentration of B Corps across the organizational population and the presence of early adopters of B Corps certification impacted the number of future adopters. This suggests that organizations look to other organizations in their field to provide legitimacy to certification. When others are certifying, potential adopters are more likely to mimic that behavior and make the decision to certify. The more organizations certify, the more likely the conditions will be in place to support future adopters.
In addition, when we look more closely at the hypotheses, we also find support to suggest that economic factors support certification. This finding contrasts with Hickman et al. (2014), who did not find an expected positive relationship between economic variables: median household income and charitable giving and B Corps certification [10]. Their results may have been hindered by the unit of analysis-the household-not correlating with organizational level phenomenon. Our results, instead, support an economic resources argument at an organizational level. As the number of private sector enterprises increases, so does the number of certified B Corps. This suggests that resources in the economic environment that facilitate capital flows to all organizations also encourage certification behavior. B Corps are likely to respond to traditional market mechanisms. In periods of economic strength, certification is more likely to be pursued. If we turn again to Figure 1, we see this pattern supported as certification flattened during the economic recession. During weaker economic periods, certification is less likely to be pursued.
What is also interesting to note is that a liberal political ideology was not supported as a condition that supported the diffusion of B Corps certification.

Conclusions
This paper highlights how institutional and economic resources in the external environment provide conditions that support the diffusion of B Corps certification. Our dataset was unique, as we were able to create and analyze a database of 851 B Corps between 2007 and 2016, the first 10 years of certification. Our findings and analysis showed that institutional and economic resources support the diffusion of B Corps certification. Our research suggests that companies will be more likely to certify when other organizations in the entire B Corps field and specific organizations in their state have done so. Furthermore, economic resources in the external environment support the adoption of certification. The decision to certify, then, should not only be attributed to internal decision making, but is also a result of favorable conditions in an organization's external environment.
Our study highlights the importance of institutional and economic resources in supporting certification-one type of corporate social responsibility behavior. The study helps to build the case that institutional and economic support are important factors that promote CSR behavior. In order for certification to become more legitimized, prevalent, and adopted by organizations, economic and institutional conditions must support certification. B Corps have been building communities to help support future adopters. B Local groups, in particular, have arisen throughout the world to support organizations that are considering certification. Their role is to foster connections among organizations who share values that promote social and environmental well-being and desire to use business as a force for good. Groups, like these, can help to foster peer, social, and network resources that can lead to tangible gains in terms of economic partnerships. Furthermore, these groups can provide the institutional support needed to promote certification. The more B Corps certified organizations, the more likely businesses will contribute to the social and environmental wellbeing of the planet.
B Local activity has grown since our dataset was constructed. The first B Local group, B Local Bay Area, was formed around 2013, and, today, there are 19 B Local groups in the United States. Future research could examine the impact of these groups on B Corps certification to see if institutional factors, like peer networks, continue to be a powerful force in promoting future adoption.
Future research should focus on other types of corporate social responsibility practices. This could include the legal form-the Benefit Corporation, employee stock option plans, the socially responsible investment fund, and other types of social enterprise. Comparative studies will help to show whether B Corps follow unique patterns in comparison to other CSR practice. This would provide additional insight as to the future of CSR practice and the broader movement for social and environmental responsibility.
Future research can also build on this research to identify the difference between organizations that certify and organizations that do not. We continue to build the database and future research will look at both B Corps certification, the Benefit Corporation (the legal form), and organizations that do not adopt B Forms. Future research should identify the different factors that might compel organizations to pursue one form over the other (B Corps or Benefit Corporation) and whether there are different patterns that predict certification over legal adoption.