Institutionalizing Women’s Enterprise Policy: A Legitimacy-Based Perspective

Policy Institutionalization

The term “policy” represents a diversity of meanings to scholars. For the purposes of this study, we distinguish between policy as an expression of prevailing political rationale and policy as a complex multi-actor system (Cairney, 2015). Thus, policy can denote a “plan of action,” typically formulated by powerful government actors to guide political responses to societal challenges (Richards & Smith, 2002, p. 1). Under this perspective, it is largely an instrumental tool that is applied in a top-down fashion. The latter interpretation, however, conceives policies not as technical, neutral devices but as “institutions” that structure the collective activities of participating actors (Arshed et al., 2014; Lascoumes & Le Gales, 2007).

Distinctions between these two understandings of policy surface some important issues. Scholars working within a variety of empirical contexts (e.g., Singh, Heimans, & Glasswell, 2014) have noted how macro-level policy objectives, when refracted through multiple layers of localized interpretation, tend to deviate (sometimes dramatically) from their original ostensive purpose (Arshed, Mason, & Carter, 2016; Burch, 2007). Thus, while policies are representative of certain normative ideas, they are also changeable during the social processes of enactment, and their institutional maintenance is ultimately beholden to the ever-shifting agendas of diverse participating actor groups (Lascoumes & Le Gales, 2007).

In studies of policies introduced to support marginalized or under-represented groups, scholars have often observed how variances in the experiences of target and non-target audiences can lead to a range of unintended (often negative) consequences (Heilman, 1994; Leslie, Mayer, & Kravitz, 2014). Studies of WEP, however, have displayed comparatively limited recognition of the pluralistic nature of socially enacted policy. This is unsurprising, given that much of the early work in the field focused on explaining the economic performance of women-owned businesses (Hughes, Jennings, Brush, Carter, & Welter, 2012). WEP research has accordingly tended toward instrumental assessments of intervention efficacy (Orser et al., 2017; Robson, Jack, & Freel, 2008), policy rationales and designs (Marlow & McAdam, 2013; Wilson et al., 2004), female-specific barriers (Brush et al., 2018), or broad “fixes” to such barriers (see Foss, Henry, Ahl, & Mikalsen, 2018). Such perspectives primarily frame WEP as a monolithic, static discourse (e.g. Ahl & Nelson, 2015) and not as a distributed social institution. To address this gap, there is a need to examine policy from the perspective of the diverse actors that constitute the policy enactment ecosystem. We initiate theorization of the relationship between micro and macro perceptions of policy institutionalization by turning to institutional theory, and specifically to the concept of legitimacy.

Legitimacy-As-Perception

Legitimacy is a core concept in organizational theory, defined by Suchman (1995, p. 574) as a “generalized perception or assumption that the actions of an entity are desirable, proper or appropriate within some socially constructed system of norms, values, beliefs and definitions.” Legitimacy has formed a central component of neo-institutional theory (Meyer & Rowan, 1977), explaining both the functioning of institutions and the survival of organizations within institutional fields.

To address our research question, which seeks to understand how and under what circumstances individual actors make ongoing evaluations of a policy institution, we adopt a legitimacy-as-perception lens (LAP). LAP is an emerging strand of theory which advances a cognitivist perspective to analyze how legitimacy judgments are made by individual evaluators (Tost, 2011). In comparison to the more prevalent contingency views, whereby legitimacy exists between two entities (e.g. an organization and a regulator; Zimmerman & Zeitz, 2002), LAP proposes that legitimacy resides in the eye of the beholder (Fisher, Kuratko, Bloodgood, & Hornsby, 2017). This principle has led to scholars taking a less passive view of individual legitimacy evaluators by considering the relationship between their micro-level legitimacy judgments and a higher-level, aggregated legitimacy. Bitektine and Haack (2015) and Tost (2011) distinguish between two cross-level components of multilevel legitimacy judgments: “propriety” and “validity.” At a micro-level, propriety represents the degree of individual approval for a legitimacy object. At a macro-level, validity is the “extent to which there appears to be a general consensus within a collectivity that the entity is appropriate for the social context” (Tost, 2011, p. 689). Significantly, an individual can attribute a negative propriety judgment to an item that has high collective validity and vice versa. However, the extent to which this propriety view is expressed publicly is contingent on the macro “validity belief” formed by the evaluator of said object, and the associated pressures for conformity within the organization or system. That is, individuals may hold relatively negative personal judgments of a specific legitimacy object (propriety), but—in a strongly conformist context—these judgments are unlikely to be expressed if they are perceived to be at odds with the general judgment (validity).

This scope for misalignment between propriety and collective validity creates the potential for institutional change. In a stable institutional environment, pressures to suppress any negative or deviant judgments are considerable, leading to reinforcement of collective validity and a high validity belief by evaluators. During periods of institutional instability, however, individuals often invoke a broader range of evaluative frameworks when forming judgments (Bitektine & Haack, 2015). Furthermore, there is also increased scope for individuals to make public unacceptable judgments within the established “stable” environment. Such actions, in turn, may weaken the collective validity of a legitimacy entity. Thus the LAP lens has enabled scholars to unpack the multilevel nature of these institutional processes to gain an understanding of how macro-level institutions iteratively shape and are shaped by the judgments and actions of individual actors (Barley, 2008; Dacin, Munir, & Tracey, 2010).

Legitimacy and Institutional Change

While a significant focus of institutional theory has been on the enduring nature of institutions (Scott, 2001), less attention has been directed toward explaining why institutions weaken or disappear (Lawrence & Suddaby, 2006). Deinstitutionalization happens when the legitimacy of an institution is called into question by an audience or evaluator (Oliver, 1992) and institutional entrepreneurs then work to transform the institution (Gilmore & Sillince, 2014; Maguire & Hardy, 2009). Such a process may be initiated by an “insider” within an institutional field, or by an outsider, for example, as in the seminal case of the deinstitutionalization of the chemical DDT (Maguire & Hardy, 2009).

While deinstitutionalization is inherently tied to the concept of legitimacy (Maguire & Hardy, 2009; Oliver, 1992), few studies have theorized the relationship between individual propriety legitimacy judgments and actors’ engagement in deinstitutionalizing work (see, for example, Huy, Corley, & Kraatz, 2014). Moreover, in past research, there has been a tendency to focus on individualized accounts of institutional entrepreneurs who destabilize institutions (Gilmore & Sillince, 2014), despite Maguire and Hardy’s (2009) explicit rejection of the “heroic” individual in such institutional work. Extending upon previous deinstitutionalization research therefore, we perceive a need to analyze everyday distributed examples of (de)institutionalizing work arising from individual judgments, formed across enterprise policy ecosystems, at the “coalface” of a policy institution (Barley, 2008). So, by combining these theoretical strands, we can more fully explore the gap in WEP research concerning the role of legitimacy in policy institutionalization.

Research Context

Our study was conducted in the UK between late 2008 and late 2009 during the latter stages of the Labor administration but prior to the ascendency of the Conservative-Liberal Democrat coalition in May 2010. The government of this period (1997–2010) established the first explicit WEP agenda, building on the publication of the “Strategic Framework for Women’s Enterprise” (DTI, 2003) to consolidate previously piecemeal support programs into a coherent strategic approach (Forson, 2006). The aim of the UK government’s “Enterprise Strategy” at this time was to foster a “culture of enterprise” (BERR, 2008, p. 13) through supportive regulatory frameworks, advisory services, enterprise education, and improved access to finance. Under this umbrella, WEP was primarily focused on removing specific barriers to female participation in enterprise, notably those concerning access to public sector procurement contracts, international trade, and bank finance (Alexander, Stone, Ahmad, Carter, & Dwyer, 2009). Key to this development were “female-friendly” business support services, which, in conjunction with a range of existing nongovernmental services and networks, were aimed at improving a perceived lack of accessibility to mainstream “universal” support provision (Alexander et al., 2009).

Data Sources

Our study is built upon a single in-depth case study detailing the workings of the UK WEP ecosystem. Consistent with other studies of large complex organizational systems (e.g. Wiedner, Barrett, & Oborn, 2017), our case encompassed a wide range of data sources—observational notes, a researcher diary, semistructured interviews, and archival data—which were designed to empirically mobilize three broad stakeholder groups (formulators, implementers, and users).

Viewing legitimacy as “a subjectively bestowed judgment that is variably ascribed by different audiences” (Suddaby, Bitekine, & Haack, 2017, p. 470), we captured actor behaviors at multiple localized sites. Our aim was to explore a range of research contexts reflective of where and how different stakeholder groups typically experienced their particular version of “doing” policy. In examining policy formulators, for example, we accessed senior civil servants and policy-makers working within relevant branches of central government (notably, the Enterprise Division of the Department for Business, Innovation and Skills—BIS). To investigate the implementation of policy, we focused on the workings of a Regional Development Agency (RDA) based in the West Midlands of England that was responsible for the translation of higher-level policies into actionable localized strategies. We also conducted research within the context of nine local advisory services operating under the purview of this RDA, which were responsible for localized program delivery. To draw boundaries around the data collection, we deliberately focused on implementation activities carried out by local business advisors. Thus we did not explore WEP within the context of broader agendas surrounding, for example, non-adult enterprise education or regulatory reform. Finally, we explored the usage of policy by examining women entrepreneurs who had either directly engaged with, or had specifically elected not to engage with, one or more of the services offered by a support agency.

The research design was inductive and our purpose was to build conceptual explanations for the multifaceted experiences reported by respondents rather than to impose an a priori framework. Consequently, we treated participants as sensemaking subjects, who, through their efforts to explain their realities, constructed rich portrayals of key phenomena and associated organizing processes (Gioia, Corley, & Hamilton, 2013). Table 1 provides an overview of the various actors, roles, contextual settings, and corresponding data sources that we used to represent the WEP ecosystem. We detail our data collection methods below.

OPEN IN VIEWER

Table 1.

Research Sites and Data Sources.

Stakeholder group	Actors, roles, & group description	Research setting(s)	Data sources
Formulators	The Enterprise Division of BIS was the core governmental department devoted to high-level policy design. Agenda setting, resource & funding allocations were primarily dictated by government ministers and senior civil servants. Research, public consultation & dissemination, communication with the implementation network was carried out by a hierarchy of policy officers	Enterprise Division, BIS, Central London	Interviews (8 senior policy makers; 2 senior civil servants) 3 months full-time participant observation, researcher diary, follow-up enquiries* Observational notes from 32 meetings Secondary/archival data (The “Green Book,” “White papers,” internal memos, organizational hierarchy maps)
Implementers	RDAs were responsible for improving regional competitiveness, building institutional capacity & fostering partnerships with local level delivery agents. RDA board comprised of business owners & representatives from local government, trade unions & voluntary organizations. The board was supported by a managerial & admin staff of 100+	RDA HQ, West Midlands	Interviews (4 RDA staff; 9 local agency advisors) Secondary/archival data (promotional descriptions of services offered, advisor training guidelines) 7 of the 32 observed meetings conducted within central government also involved representatives from RDAs and other associated delivery bodies
	Local advisory agencies represent the public facing hubs of the enterprise support system, and were typically located in easily accessible high-street sites. Their business support provision primarily took the form of by-appointment advisory services & support workshops	9 Local advisory branches throughout the West Midlands
Users	During 2009, there were an estimated 1.1 million self-employed females (Labour Force Survey, 2009–10). Government figures for the 2009–10 period (BIS, 2011) indicate that 55% of women-led SMEs had sought external advice or information on matters affecting their business	Entrepreneur’s typical place of business	Interviews (22 policy-using and non-using women entrepreneurs, 11 of each) Secondary/archival data (advisor/client correspondence, advisor recommendations)

Semistructured interviews: Between December 2008 and 2009, we conducted in-depth semistructured interviews in a series of phases. We carried out a total of 45 interviews (eight with policy-makers, two with senior civil servants, four with RDA senior staff, nine with local agency advisors, and 22 with women entrepreneurs). To recruit participants, we employed a hybrid purposeful and snowball sampling strategy, initially utilizing the extended professional networks of our research team to access respondents in senior policy-making roles, before leveraging interviewee recommendations to initiate further introductions. The sampling approach was used to capture data from respondents with relevant knowledge and generalize to theory rather to a larger population. Lasting between 1.5 and 2 hours, our semistructured interviews followed a broad thematic protocol aimed at elucidating the perceptions, motives, and rationales underpinning respondents’ participation in policy enactment processes.

Participant observation diary: Between August and December 2009, the lead author conducted ethnographic research within the Enterprise Division of BIS, acting as a participant observer while performing a 3-month full-time role as an enterprise Policy Advisor. Actors within the host organization were made aware of the research project and of the researcher’s dual role. Participants agreed to take part in our study under conditions of anonymity. Data were recorded in diary format and entries were submitted daily. Entries included observations of informal conversations, formal discussions, and day-to-day behaviors, as well as numerous follow-up interviews with key actors, designed to clarify specific points or reflections. The researcher also attended and took field notes from a total of 32 policy meetings, which involved actors both internal and external to the host organization.

Archival data: Finally, we collected a range of documentary evidence, including governmental guidelines (e.g., the “Green Book”), relevant policy proposals (“White Papers”), internal memos, published reports, training and advisory guidelines. The collection of archival data served two main purposes. Firstly, it allowed examination of ostensive policy objectives and implementation guidance, thus acting as a means for us to build a sense of the “idealized” policy discourse. This, in turn, provided a reference point for examinations of supportive or divergent enactment behaviors. Secondly, it served as a means of data triangulation, particularly with respect to investigating links between individual-level observations and any responses actioned at wider organizational levels.

Analysis

Initially our analysis sought to understand how the WEP ecosystem worked in practice. Starting from an analysis of secondary and archival data, we used policy guidelines, training manuals, internal memos, and promotional literature to form a baseline understanding of core policy rationales, and of the ostensive roles played by different actors within the ecosystem. We then used this information to map relationships between stakeholders and organizational entities.

Next, we focused on data collected from primary sources (observations and interviews), developing enactment narratives for each of the individual actors in our sample. The objective was to elucidate first-person perspectives of how actors “do” WEP, thus allowing us to draw out comparative variations between idealized practices and localized interpretations. We found that points of discrepancy and/or tension between ostensive and localized practice typically resided at the heart of actor judgment processes. Consequently, we conducted a systematic analysis of our interview and observational transcripts to identify such episodes. Comparing interlinked incidents and behaviors within and across interviews, as well as over the timeline of our ethnographic observation period, we constructed a series of multi-actor longitudinal narratives, each of which detailed the antecedents, content, and repercussions of various judgment processes.

Our analysis revealed that policy actors typically felt compelled to justify why they enacted or engaged with policy in the ways that they did. These justifications were typically manifested as responses to interview questions or as observed communications with other policy actors (cf. Huy et al., 2014). Our initial round of “open coding” (Gioia et al., 2013) uncovered three dimensions of actor justifications. Firstly, we examined participants’ reflections on the contextual pressures that enabled, prevented, or otherwise impacted their policy engagement behaviors. These included shifting policy targets, signaling from other actors, and competing policy institutions. We categorized these pressures as “evaluative triggers,” which typically acted as antecedents to judgment formation.

Secondly, we interpreted the expressed content of legitimacy judgments (see Tost, 2011). Our evaluations of judgment content gathered evidence of both the basis and the targets of actor evaluations. Judgment targets proved to an important element of the evaluative process, as actors would often base their assessments of the wider policy institution on interactions with a particular stakeholder or stakeholder group. In this way, individual interactions within the policy ecosystem (e.g., a single meeting with a local advisor) could often act as proxies for wider evaluations of WEP. By recording and coding these numerous individual judgments, we were able to construct a tapestry of the different views that underpinned overall evaluations.

Thirdly, we analyzed the talk and actions employed by policy participants, that is, the “externalized” discourses and tangible practices that caused internal judgments to become “consequential to the organization” (Bitektine & Haack, 2015, p. 53). Importantly, we observed frequent discrepancies between the internal evaluations and subsequent actions of policy actors. Using evidence of “consequential” actions built from triangulated observations and documented practices, we were able to compare the evaluative content of actor reflections with their externalized behaviors. This allowed us to conceptualize various institutionalizing and deinstitutionalizing “response tactics.”

Building on this set of inductive observations, we developed second-order themes and aggregate theoretical dimensions (Gioia et al., 2013; see Figure 1). These served as a basis for our theorizing. Throughout this process, we iteratively returned to key extant literature (e.g., Bitektine, 2011; Bitektine & Haack, 2015; Tost, 2011), comparing our observations and categorizations with existing theoretical constructs. Figure 1 provides a representation of the links between our raw data and the theoretical categories underpinning our contributions. Lastly, we refined our second-order constructs by scrutinizing them against our multi-actor vignettes. This allowed us to explore relationships between constructs, thus forming the basis for an emergent theoretical model. The following sections detail our themes and outline our legitimacy-informed interpretation of WEP institutionalizing. OPEN IN VIEWER

Evaluative Triggers

In a stable institutional environment, the legitimacy of an institutional entity is taken for granted by actors who remain in a passive cognitive mode (Bitektine & Haack, 2015; Tost, 2011). What then are the destabilizing triggers that induce those in the policy ecosystem to enter a mode in which the legitimacy of WEP is actively re-appraised? We identify three trigger categories through our field work, and in doing so, reveal the underlying catalysts for individual propriety judgments in this setting.

Shifting Strategic Agendas

A key source of instability within the policy ecosystem relates to the short-term approach politicians are perceived to adopt by other ecosystem actors. Often ministerial policy announcements are made with little warning and the delivery infrastructure (i.e., the “implementer” group within our study) is expected to enact the new policy within a very short timescale. This creates a tension between ministers and the various implementer organizations such as regional and local delivery agencies, who can often take years to transform a policy directive into effective support. In such circumstances, where ministers launch policy before previous initiatives have been fully implemented, stakeholders begin to perceive WEP as reactive and faddish as opposed to strategic and considered:

I think inevitably when we talk about enterprise policy it’s going to be subject to the comings and goings of particular ministers and policies and economic priorities…schemes that were flavour of the month a year or so ago, are no longer (DH, RDA).

Signaling From Other Stakeholders

Very often, ecosystem stakeholders are spurred to re-evaluate the legitimacy of WEP based on interpretations of the externalized legitimacy judgments of other actors. Through ongoing sensemaking processes (Weick, 1995), micro-interactional signals that may take the form of an “off-the-cuff” comment or even a gesture (an eye roll during one of the meetings in our ethnography, for example), reveal potentially deviant judgments that diverge from prescribed normative positions. In one case, a women entrepreneur entered an active evaluative mode following a chance conversation with a conference delegate that challenged her normative assumptions relating to WEP and equality:

I was at some lunchtime networking thing and it was all about diversity and I got to the ticket machines after this lunch and stood there putting my money in and these two men, British, mid-40s, white, stood at the ticket machine and said “well, I hope you feel better than we do.” I said “what do you mean?” They said, “you know what happened to us?”…he said “everything is for everyone else and we’re not allowed to do anything to help the British white, UK middle-manager, whatever, because everything is about diversity and different agendas and females and minorities.” And it was actually, like, “we really feel like we’re the bottom of the pile now” (GM, Entrepreneur).

Further trigger signals from ecosystem stakeholders emerged through the lack of reciprocal engagement between potential users and implementers of support, exemplified by the very low interest in some women’s enterprise events. The apparent rejection of the policy support measures by those intended to benefit from the provision, problematized the overarching legitimacy of WEP for some individuals involved in policy implementation:

… when you mail shot these businesses, only about 3 of them said “yes we would like to come.” So anyways, in the end we had “rent-a-mob” turn up in our region at short notice because that’s who you’re going to get (SB, Local Agency).

Thus, individuals in the policy ecosystem enter evaluative mode not only because of macro-level shifts to policy strategy, but, as we discover at a micro-interactional level, a cutting joke or provocative comment can signal a challenge to the validity of WEP that warrants re-appraisal by the individual evaluator.

Competing Policy Institution

A pervasive challenge to the legitimacy of WEP is the pre-eminence of mainstream enterprise policy support. It was clear from our interviews there was a preference amongst policy formulators, implementers, and even many women entrepreneurs for providing universal support, albeit support that adequately catered for women. There was a perception from some of the enterprise agencies that the provision of gender-based services results in a zero-sum-game, where resources are transferred away from mainstream provision: “There is a danger if there is too much gender provision the mainstream provision which is of high quality may suffer” (AR, Local Agency).

Conversely, other enterprise support providers found there are inadequate resources for delivering on gendered policy promises, and this undermined the legitimacy of the overall WEP institution: “You know women only support… we can’t afford to run it… a unit cost of activity that sustains interactions with an individual client over a period, is beyond the cost of the contract we’ve got” (PH, Local Agency).

Partly as a consequence of underfunding, and partly driven by individually-held values, a number of ecosystem stakeholders actively championed a focus on mainstream policy in lieu of specialized WEP. As one female enterprise agency lead responded when asked about the legitimacy of WEP: “Mainstream…mainstream….it’s about circumstances and not necessarily sex” (JL, Local Agency).

These opinions were expressed frequently during informal conversations between stakeholder groups as part of our ethnographic study and provide insight as to how tensions between mainstream and specialized policy institutions can trigger ongoing, active, legitimacy evaluations from individuals during the course of a normal workday.

Propriety Judgment Framing

Uncertainty and instability can result in multiple concurrent interpretations of a single situation or event (Louis, 1980). To answer the second component of our research question, which seeks to understand the basis on which stakeholders perceptually evaluate the legitimacy of the WEP institution, we identify three normative frames utilized by ecosystem stakeholders to make either favorable or unfavorable propriety judgments when faced with an evaluative trigger (Table 2).

OPEN IN VIEWER

Table 2.

Judgment Frames Employed by WEP Stakeholders.

Framing category	Judgment orientation	Description	Representative data
Fairness	Positive Judgments	It is legitimate to cater to underrepresented groups	I think there’s an equity element. If you look at the headline figures in particular women are under-represented in terms of business ownership. There’s a huge resource there that’s not being capitalized on (Policy-maker 6). To me, [WE Policy] is obviously very important. People should be given the opportunity to show what their ideas are and to make them happen, irrespective of their background or gender (Policy-maker 1).
		Women face specific barriers and/or circumstances; therefore, it is right to address these	I’m not saying that there aren’t men that lack confidence or fear of failure, but if it’s mainly a women’s issue there must be enough flexibility to give as many women as you need to help support (Policy-maker 6). It’s becoming clear that women have a different way of doing things. We’re now finding about the impact of our sort of brain wiring or the impact of hormones. Feminists have been very nervous about this argument because you know it’s the whole biological determinism argument which is usually used against women (Policy-maker 8).
	Negative Judgments	Gender shouldn’t matter; women are no different	You need a policy that’s relevant to your business, not necessarily relevant to you just because you’re a woman (JS, Entrepreneur). The line of business I’m in is female-oriented but generally in business it shouldn’t matter whether you’re male or female really (BP, Entrepreneur).
		Gender is an arbitrary (and therefore unfair) way of segmenting support provision	One of the things that drives me mad when women are talked about as a homogenous group ‘oh well let’s just print it like that and all women will come to that’ (Policy-maker 4). I don’t think it’s important to separate; what about a man who is shy or who lacks confidence…or what about people from Africa? Should they have agencies for Africans? I think there is enough agencies out there. I think it’s the person; who you are (IA, Entrepreneur).
		Policy should cater to all entrepreneurs; specialization unfairly marginalizes other groups	Sometimes there’s an over-emphasis on one issue [women’s entrepreneurship] to the exclusion of others, supporting everyone equally is important. We get a lot of complaints from men because we actually do so much for women-only entrepreneurs (JW, Local Agency). I think it should be for everybody because if you were to have one just for men then there would be an uproar so I think you should have the same for women (SB, Entrepreneur).
Impact	Positive Judgments	WE policy is legitimate if it has positive impact on the wider economy	[WE Policy] is a really important area. This isn’t just a nice to do. This is an actual economic sense to do it. This is about closing the productivity gap (Policy-maker 4).
		WE policy is legitimate if it has positive impact on individual businesses	Policy has to be about looking at what is the evidence base tells us, what is working well, what polices appear to be helping women develop businesses or grow businesses? And where are the failures? Where are the gaps? Good enterprise policy will be looking at those market failures and those barriers and will be addressing them (Policy-maker 5).
	Negative Judgments	WE Policy focuses too much on “lip service” rather than tangible impacts	Some of them [policy making agencies] continue to do lots of research on women’s enterprise and take very little action and it tends to be focused on bringing out a new leaflet or a website. So, it’s lip service (Policy-maker 8). I’m not sure how much is being done [for women entrepreneurs] but I think some of it is lip service (PH, Local Agency).
		Having specialist policies undermines the impact of mainstream support provision	I think anything that’s slightly more specific, often it becomes under-funded, it can become weakened by not being part of the mainstream (Policy-maker 2). I think there is an attitude from the mainstream that the gender-based provision is not of as high a quality as mainstream provision, it’s seen as an add on. There is a danger if there is too much gender provision the mainstream provision which is of high quality may suffer (AR, Local Agency).
		Tangible outcomes of WE policy are not worth the effort and/or resources	Women only support; we can’t afford it…we can’t afford to run it…we’ve just been looking at our numbers and our achievement is as good as you could expect given that we don’t do special things [for women] (KG, Local Agency). The unit cost of activity that sustains interactions with an individual client over a period, is beyond the cost of the contract we’ve got (PH, Local Agency).
Esteem	Positive Judgments	Specialist women-only policy builds the confidence of policy users	There’s no doubt in my mind that you need something gender specific because there are issues about confidence and walking into a room trying to network when it’s a room full of men (AR, Local Agency). If [women entrepreneurs] are surrounded by positive women who are there to support them, yes, that’s a fantastic idea. Some women feel intimidated by men…because you do actually have that where you go into somewhere and the guy automatically knows more because of your age and a woman is slightly more understanding about kids, etc. (CBD, Entrepreneur).
		WE policy is a source of personal pride	This year, we updated our strategy and I think maybe it’s four of five years to say where we are now. And I’m really genuinely proud that when we first set up our strategy in 2005, which is slightly less time, we set a target to open 10,000 new women-owned businesses in the South East and we hit that target this year (Policy-maker 4). We think that providing services for women is important. Although it’s not a formal part of our contract it’s something we aspire to do (AL, Local Agency).
	Negative Judgments	Specialist women-only policy undermines actors and makes women look weak	There’s a minority, we know it’s a minority because we’ve asked women, say about 10–15%, who are quite offended and actively not interested in targeted provision (Policy-maker 8). I don’t see why women should be treated as a special case. I think the idea is patronizing. I can’t imagine how the contents of such programs would differ from those given to men (CM, Entrepreneur)

Fairness Framing

By some distance, the most polarizing framing of WEP legitimacy encountered in our study was when the “fairness” of the policy was invoked as an evaluative lens. Debates over fairness as a specific form of morality judgment are at the heart of controversies surrounding many affirmative action policies (Shteynberg, Leslie, Knight, & Mayer, 2011), and we found two competing perspectives in our interviews. The first, is that women have additional needs both as primary caregivers and as individuals who face discrimination in the workplace, and hence, are deserving of extra support to ameliorate such barriers. This framing was common, particularly amongst policy formulators and implementers, who are each targeted with increasing economic participation amongst women:

Women are significantly more likely to start a business from a position of unemployment than male counterparts. It’s a particular issue for women, it’s not only the numeric side of it but it’s also the fact that women start their businesses at a slightly more gradual pace. And the new deal for self-employment really doesn’t support that, it is very focused on the male model of you know quicker paces etc. (Policy-maker 8).

Impact Framing

An alternative frame for evaluating the legitimacy of WEP concerned the anticipated effect of the policy and associated support measures. Where individual evaluators had observed either ongoing successes or failures of WEP, there was a tendency to frame legitimacy judgments in terms of “impact.” Most typically, this “impact” frame was invoked to emphasize a focus on “talk” versus “action” in relation to policy delivery. Here stakeholders perceived a misalignment between high-level policy announcements and their outcomes, forming a negative judgment based on low perceived impact. As one RDA observed: “…it’s (WEP) scratching the surface really and I think fundamentally there’s been a lot of talk about encouraging women to go into business but it’s…that hasn’t been followed up by any action which has been disappointing” (JW, RDA).

The impact framing encompassed judgments relating to tensions between delivering quality support as opposed to quantity (i.e., reaching more potential users). Various stakeholders took a cynical perspective, perceiving government to only care about being able to report impact through “numbers,” with little concern for the actual effect on women entrepreneurs: “Rather than the numbers, the quality of the projects, the impact it’s having on society…is more important” (AR, Local Agency).

The managerialist focus on “numbers” was used to form a judgment on the legitimacy of policy for a number of enterprise agency employees. It was considered that reporting of policy support interventions was focused on “how many inquiries have you dealt with, how many businesses have you assisted, what’s your percentage customer satisfaction level, those sorts of things which are operational management information things” (DB, RDA) as opposed to more tangible socioeconomic impacts.

One female lead within a local agency took exception to the tokenistic approach toward evaluating impact, querying why the focus was on achieving gender parity with male entrepreneurs in terms of the volume of women entrepreneurs in the economy. She commented:

We got the female minister saying that we need 50% of women starting businesses. Why? Sometimes women don’t always want to start…why half? Just because over half the population is women but why do they have to start a business? (JL, Local Agency).

Finally, we encountered a number of women entrepreneurs who framed legitimacy judgments of WEP based on the perceived impact of the support on their businesses. The poor reputation of some implementer organizations who were contracted to provide WEP support, led to negative propriety judgments from users:

I’d heard of Business Gateway but I’ve heard so many bad things about it that I just thought I’m avoiding it because the last thing I want is somebody to put me down or put me off starting something … what are they going to tell me that I don’t already know, basically? (PH, Entrepreneur).

Esteem Framing

The final frame used to judge WEP echoes the “relational” norm identified by Tost (2011, p. 690) whereby legitimacy is afforded to entities that affirm “individuals’ social identities and bolsters their sense of self-worth.” We specifically identified “esteem” as a key judgment lens, which emerged as a strong theme amidst the highly politicized and contested nature of contemporary gender-identity debates (Marlow & Dy, 2017). Women entrepreneurs often rejected specialized support, and WEP more generally, on the basis that it undermined their status as competent business people. As SC, a successful women entrepreneur, reflects: “I get the equality, but I don’t want to go out there and start saying I’m something different. I’m a business leader, my gender doesn’t make a difference!” Such a view is reflected by other entrepreneurs who feel that WEP signifies a lower status relative to their peers.

Conversely, other entrepreneurs identified mainstream policy support as leaving them open to discrimination and sexism. For example, PG (Entrepreneur) noted that at traditional mixed-gender networking functions, some men treated events as a “dating opportunity,” in doing so challenging the self-image of women entrepreneur attendees:

So, yes, it’s almost like, do men take us seriously enough when we’re actually at networking? Yes, we might have a skirt on, we might have high heels on, we might have a top that might be a blouse that shows a bit of chest, whatever, but we’re not there to have a date. And that’s where I’m coming from having the Women in Networking events because I know that’s not going to happen at the Women in Networking event. I know it’s clear-cut; I know we’re there to do business; I know none of the women are going to come back to me and say somebody emailed me and they’re asking to go for a coffee, for a date.

Suppressed judgment tactic

As Bitektine and Haack (2015) note, it is possible for individual evaluators to form a low propriety evaluation of an entity that has high collective validity. In such cases, the perceived strength of the institution has some bearing on the likelihood of an individual expressing a deviant evaluation. Similarly, we found a common tactic for dealing with incongruent propriety judgments, was for individuals to suppress negative judgments, thus maintaining the collective validity of the policy entity. This was evident within the policy-making group for example, where some individuals confided to us that they thought WEP was unfair and rarely worked yet felt powerful normative pressure to align with the ministerial agenda. As one policy-maker noted, “we have a political driver, and ministers want to be re-elected.”

Championing Tactic

As the policy ecosystem reflects a diversity of opinions and perspectives, we encountered many individuals who passionately supported WEP. When these individuals sensed that the collective validity of WEP was weakened or under threat, they engaged in “championing” activities that shored up or repaired the legitimacy of policy. This often entailed providing support and assistance that was not contracted for, and generally going “above and beyond” normative expectations. This is exemplified by one enterprise agency head who said: “We think (support for women) is very important and one of the things we are working on—not because we are contracted to do it—we’re doing it because it’s the right thing to do” (AW, Local Agency).

Playing-the-Game Tactic

We encountered many instances of individual stakeholders paying “lip service” to the overall institution by abiding to minimum regulative or social expectations, while simultaneously signaling disapproving legitimacy judgments to other stakeholders, often triggering further propriety judgments in others. These individuals thought of themselves as “playing-the-game” to either avoid conflict, maintain personal legitimacy, or acquire resources, suggesting that there was at least a moderate, or conditional belief in the validity of the WEP institution. Examples include women entrepreneurs who openly mocked women’s enterprise support, but nevertheless signed up and attended meetings in order to access financial assistance. At the more extreme end, some enterprise agencies accepted funding to provide dedicated women-only support, but in practice, fudged the contractual requirement by directing women entrepreneurs towards a beefed-up mainstream offering. Such decoupling (Elsbach & Sutton, 1992) has a deinstitutionalizing effect as other stakeholders who observe the intransigencies, either through changes to everyday practices or through signaling from fellow stakeholders, are likely to form a lower validity belief of WEP which may affect subsequent legitimacy judgments.

Guerrilla Tactics

Although rarer, there were instances of individual stakeholders electing to defy or reject WEP as a legitimate institution. These included our “non-user” women entrepreneur cohort who were each entitled to business assistance and financial resources, but purposively withdrew support for the policy institution. In many instances, this involved publicly denouncing the policy and working to institutionalize mainstream enterprise policy instead. Other examples included an “implementer,” who, many years ago had pioneered women-only support in the UK but recently decided to end segregated practices owing to a belief that gender discrimination had been ameliorated:

Then a few years later and I had got half of my women advisers as women anyway and a lot of things had changed, the attitude of the banks had changed, banks no longer laughed at women who wanted funding, they wanted to see their proposal and there didn’t seem to be any real issues and we did a little bit of research, they said they were not bothered whether we see a man or women. We then disbanded the women’s agency because more often than not they would be seeing a woman, and we treated them the same, and there was no big issue (JL, Local Agency).

Distributed Agency in the Enterprise Policy Ecosystem

Our study addresses limitations in past WEP research where ecosystem actors have been marginalized in theory, thereby overlooking their potential agency in policy institutionalizing processes. Perhaps understandably, the primary focus of gender-based enterprise research has been on the individual (female) entrepreneur (Hughes et al., 2012). Yet, recently, scholars have highlighted the need to “study both the resource providers and the connectors within the ecosystem” (Foss et al., 2018, p. 2). We believe this is a particularly instructive call, as our empirical case reveals ways in which notionally peripheral actors (e.g., non-using women entrepreneurs, business advisors) can materially influence the generalized validity of WEP, and therefore policy institutionalization.

Furthermore, our ethnographic analysis of the enterprise policy ecosystem enabled us to observe the policy institution from the rarely utilized micro-interactional vantage point. Here we connected to practice-based approaches in institutional theory literature (Smets & Jarzabkowski, 2013), which posits that institutions are reproduced through the effortful accomplishment of actors (Jarzabkowski, 2005). We suggest this offers a much-needed corrective to the balance of the WEP literature, which has mostly examined policy as it is reproduced through high-level discourses, typically involving policy documentation (Ahl & Nelson, 2015). While these policy texts are a vital source of data for understanding enterprise policy structures, they offer little guidance as to how policy is reproduced through practice. Thus, through our ethnographic perspective, it becomes possible to observe the mundane reproduction of WEP through meetings, coffee breaks, and other episodes of practical work where individuals adapt WEP within the constraints of their own local contexts and networks of relations. While our model identifies a viral effect, where diverging individual legitimacy judgments and response tactics can be amplified by triggering legitimacy judgments in others, we suggest more work is required to understand how this distributed agency integrates to become “more than the sum of its parts” (Lawrence, 2017, p. 1792).

Studying individual-level WEP legitimacy judgments also led us to some surprising findings around judgment framing. We discovered for example, that the notional “stakeholder” categories we identified in our research (e.g., women entrepreneur, implementer, and formulator), which we thought may reflect the “thought worlds” in Lamin and Zaheer’s (2012) Wall Street versus Main Street distinction, did not hold in our study. Across the three judgment framing categories (fairness, impact, esteem), there was scant evidence of any consistency by stakeholder type in terms of which frame was invoked and whether a positive or negative propriety judgment was formed. This underlines the lack of any strong conformity of opinion across policy ecosystem stakeholder groups. ¹ It also affirms that that macro-level “consensus” (Suddaby et al., 2017) afforded to WEP is not replaced by meso-level group “consensus,” in turn obscuring individual examples of distributed agency that actually drive change (e.g., Lawrence, 2017; Van Wijk, Stam, Elfring, Zietsma, & Hond, 2013). These findings underline the need to further explore the heterogeneity of women entrepreneurs (Hughes et al., 2012) and other policy ecosystem stakeholders to fully understand policy institutionalization.

Gender and Enterprise Policy

Our study also sheds new light on how the social positioning of women in enterprise policy can influence their entrepreneurial activity. Scholars have observed, for example, that masculine norms shape both entrepreneurial meanings and practices (Ahl, 2006; Ahl & Marlow, 2012; Datta & Gailey, 2012; Malmström et al., 2017). In a recent study of WEP discourse from Sweden and the USA, Ahl and Nelson (2015) found that the subordination of sex-segregated policy contributed to the “othering” of women-owned businesses in the economy. Here, reification of women as somehow “lesser” or inferior to men was institutionalized across society through policy support measures that were developed and delivered by national governments and partner organizations. Our study builds upon these recent insights into policy discourse formulation (Ahl & Nelson, 2015; Arshed et al., 2014) by elaborating on the practical consequences of gendered social positioning to the ensuing policy institutionalization. This was achieved through exploring the “other side” of the policy process, which examined how policy is institutionalized (and deinstitutionalized) through the everyday practices of actors in the policy ecosystem. Notably, our study differed from other feminist and post-feminist analyses of enterprise policy, by exploring how male social actors—in addition to female—experience and reproduce gendered institutions, something Ahl and Nelson (2010) have called for to add depth to gender-based critiques of entrepreneurship.

We also found ample evidence of the reification of gender stereotyping (Gupta, Turban, Wasti, & Sikdar, 2009) of male and female businesses in policy and support services (Marlow & McAdam, 2012). Here, our work echoes Saridakis, Marlow, and Storey (2014), who identified a bias toward the “feminized” social (i.e., caring roles and domestic responsibilities) rather than the “masculine” economic sphere for women making self-employment decisions. This finding suggests that the top-down reproduction of stereotypical gender norms through policy is problematic for successful policy implementation (Tillmar, 2007), a finding that also surfaces in analyses of ethnic minority entrepreneurs in terms of racial stereotyping (Ram, Trehan, Rouse, Woldesenbet, & Jones, 2012). Finally, we were able to further contribute to these studies by outlining the consequences of perpetuated subordination in enterprise policy support, in the form of “playing-the-game” and “guerrilla” response tactics, that we show can reduce the generalized validity of WEP.