Abstract
This article contributes to a growing conversation about the role of numbers in promoting gendered agendas in potentially contradictory ways. Drawing from interviews with gender advisors—the professionals tasked with mainstreaming gender in development projects—in an East African country, I begin from the paradox that gender advisors articulate a strong preference for qualitative data to best capture the lives of the women they aim to assist while voicing a need for quantitative metrics. I demonstrate that (women) gender advisors come to imagine metrics as expeditious bureaucratic tools able to inspire cooperation from otherwise reluctant (men) coworkers. I argue that development organizations are gendered in ways—acutely seen in how advisors struggle, are sidelined, and attempt to advance their goals with numbers—that lead to the utility of valuing quantitative metrics over qualitative ones. I establish two theoretical contributions: (1) Gendered organizations theory is essential to understanding the adoption and globalization of performance metrics, and (2) in an age of evidence-based decision making, the utility of quantified data to garner resources is heightened, rewarding those who adopt quantified knowledge production. I coin the term “the paradox of quantified utility” to describe how these material advantages encourage even skeptics to value quantitative metrics.
Keywords
Quantitative data—Ha! I’m saying we need more of it!—is also a bias against the global women’s movement and Global South intelligence, expertise, and knowledge.
Social scientists are interested in the increasing role numbers play in modern society through valuation, evaluation, knowledge production, and allocation of resources (Espeland and Sauder 2007, 2016; Lamont 2012). Numeric data and indicators have recently captured the attention of gender scholars across disciplines (Bose 2015; Liebowitz and Zwingel 2014; Merry 2016). Still, little attention has been paid to how professionals respond to indicators and why. Indicators, or quantifiable measurements, are typically used as performance metrics to measure progress, such as prenatal care rates at a hospital, women staff and promotion statistics in Fortune 500 companies, and global progress toward Sustainable Development Goals (UN General Assembly 2015). Some development professionals, however, state a preference for qualitative data, which they see as more able to capture the experience of the people with whom they work than numbers. So why might professionals, like Irene, want more quantitative measures? I examine this puzzle using the occupational case of gender advisors—the professionals tasked with mainstreaming gender and engaging women as participants in international development projects. I demonstrate that this question can be answered by understanding the highly gendered nature of their occupation and the ability of performance metrics to alter behavior and decisions within organizations that prioritize metrics. Indeed, quantitative indicators help make women’s work visible and build solidarity around issues that disproportionately affect women, among other uses (DeVault 1996). Amid the rise of organizational decision making based on quantitative performance metrics, scholars must interrogate how gendered organizations and quantification intersect and with what implications for knowledge production.
The growing scholarship on governance by indicators focuses on how performance metrics combine with organizational structures to alter organizational and professional behavior (Berman and Hirschman 2018; Muller 2018). For example, Alexander Davis’s (2018) case study demonstrates how the quest for high rankings drove universities to unexpectedly adopt transgendered bathroom and housing facilities. My research advances scholarship on the role of numbers in society and the paradoxical reactions that quantitative performance metrics—a rising tool of neoliberal management—may elicit in organizations and from professionals, particularly in gendered organizations.
Getting gender equality on the development agenda was a hard-won victory of the 1980s and 1990s. Today, many development organizations in the Global North mandate gender mainstreaming—bringing a gendered lens to the design, implementation, and evaluation of development projects (Cornwall and Rivas 2015; for an insightful overview of gender in development, see Ransom and Bain 2011)—and employ a class of professionals, “gender advisors,” to accomplish mainstreaming (Ferguson 2015) in the Global South. Although advisors and mainstreaming exist to alter gender dynamics, raising the gender saliency of workplace interactions (Ridgeway 1997), Acker (2006) cautions that what appears as a progressive step may be a reconfiguration: The experience of a job may vary, yet occupational roles and gender typing may reinforce hierarchies. In response, I analyze the occupational experiences of gender advisors across development organizations—a sector vigilant about achieving results.
This article is based on interviews with international development professionals employed in donor and implementing organization offices in an East African country. I use the case of donor-mandated performance metrics in an agricultural initiative to show how and why gender advisors, skeptical of quantitative metrics as measurement tools, come to embrace them as bureaucratic tools able to alter coworkers’ behavior. By using gendered organizations as an analytic framework (Acker 1990, 2002, 2006; Ridgeway 1997, 2009), I demonstrate how gender advisors are structurally subordinated and emotionally fatigued from workplace interactions, which generates their interest in strategies capable of engaging their coworkers in gender mainstreaming. Gender advisors see the power of indicators in the development office workplace, and although they may view them as problematic measurements of women’s lives, I argue they come to value such numbers for their bureaucratic, micropolitical power. This case contributes to our understanding of how and why resistance to quantitative metrics is silenced in gendered organizations and their impact on knowledge production.
In what follows, I synthesize literature from development organizations, gendered organizations theory, and governance by indicators. Empirically, I detail gender advisors’ quotidian work experiences and their dismissal of quantified indicators as valuable knowledge about women’s lives, and demonstrate that their paradoxical desire for such metrics is driven by their imagined future workplace once gender-related metrics are more strongly present. I conclude by highlighting two theoretical contributions: First, professionals’ reactions to performance metrics in the workplace are indeed gendered, suggesting that future research must incorporate gendered organizations to understand the nuanced pathways by which metrics are globalized, and particularized, as managerial tools. Second, quantitative indicators are not necessarily a superior form of knowledge, but are perceived as such, which draws even those who may contest quantitative metrics into valuing them. Within a culture focused on demonstrating results, this produces what I term “the paradox of quantified utility.” As a consequence, knowledge production increasingly tightens around quantitative ways of knowing.
Gendering Development Organizations
Nongovernmental organizations were once considered altruistic, values-driven organizations operating outside of market logic (DiMaggio and Anheier 1990). These beliefs derived from the groups’ socially oriented goals and nonmarket funding sources (mostly grants, donations, and government contracts). Indeed, development organizations endure persistent funding instability, which alters their everyday work activities. Large bilateral and multilateral funding typically occurs through contracting—a chain of donors, brokers, and implementing organizations (Lewis and Mosse 2006) in which donors outline their developmental aspirations and contract development organizations to implement them. That is, donors are consumers (Krause 2014). Because funding is often sourced from taxpayers, politicians (and associated donor nations) are constantly concerned that lower-level implementing organizations are not completing the expected work and seek evidence capable of casting their financial stewardship in a positive light (Best 2017; Watkins, Swidler, and Hannan 2012). This peculiar funding source creates strong incentives for donors to mandate evaluation systems to monitor progress. These performance metrics become a key mode of communication between donors and the organizations they hire to implement development.
More recent work, however, has demonstrated that development organizations—both for-profit and nonprofit entities—operate within a shared social space of rules and, along with their donors, share broad similarities with other organizations (Barman 2016; Krause 2014; Watkins, Swidler, and Hannan 2012). In particular, they are gendered. Gender is omnipresent, operating through a “gender substructure” (Acker 1990), yet often unacknowledged until it is directly relevant to the workplace circumstance or the people interacting differ in sex category (Ridgeway 1997; Ridgeway and Correll 2004). Gendered hierarchies are reinforced and entrenched through mundane, often textual, organizational processes, such as the division of labor, wages, performance evaluations, and job descriptions (Acker 2002, 2006; Martin 2003; Ridgeway 1997, 2009). Moreover, intersecting systemic oppressions of race, class, and sexualities, among others, create inequality regimes (Acker 2006). One example of how workplace structure may coproduce interactional experiences (Ridgeway 2009) is the differential experiences of “tokens.” Employees may be relative minorities in their workplace (Kanter 1977), women may be numerically scarce in gender-inappropriate occupations (Yoder 1991), or tokenization may be mitigated by accessing the cultural resources of dominant groups (Turco 2010). The powerful emotive residuals these and other experiences produce for individuals are difficult to capture (Martin 2003). This scholarship offers valuable tools to analyze the often-invisible gendered dimensions of workplaces.
A small body of work confirms that gender influences what occurs inside development organizations, yet this typically focuses on international discourse at conferences and donor agendas (Ferguson 2015; Merry 2016). Others document how race, sexuality, and citizenship affect the development workplace, which may be understood as a contemporary manifestation of colonial relationships (Baines 2010). As a policy, gender mainstreaming within the workplace has been stymied by gendered power differentials and a failure to challenge dominant structures within organizations (Benschop and Verloo 2006; van Eerdewijk 2014). There is much room for gendered organizations theory to enhance analyses of development workplace dynamics, and specifically the field offices where the labor of gender advisors is understood as the implementation of gender mainstreaming.
Reactivity to Institutionalized Measurement
Development organizations, like other organizations, are invested in measurement practices. Metrics are crucial to a variety of organizational processes: rationalization (Weber [1922] 1978), standardization (Timmermans and Epstein 2010), and political governance (Scott 1998), to name but a few. Understanding neoliberalism as a broad term, I draw attention here to the role of metrics in the devolution of responsibility onto individuals rather than organizations (Connell, Fawcett, and Meagher 2009). Metrics play a role in this process because of reactivity—the idea that people alter their actions once they know they are being measured, evaluated, or observed (Espeland and Sauder 2007). Reactivity, then, is valued by managers and other stakeholders looking to mold worker behavior (Miller 2001; Power 1997), yet may also result in worker gaming strategies and “teaching to the test” (Lipsky 2010). Scholarship that explores the relationship between gendered organizations and metrics is nascent: Levy (2016), for example, demonstrates in a study of truck drivers that hypermasculine gender identity prompts resistance to metrics, and Van Oort (2018) coins “the emotional labor of surveillance” to capture how fast-fashion store cashiers, marginalized along multiple dimensions, absorb the anxieties of customer-time-to-payment metrics. The stress of long lines shifts from managers who must hire more employees to employees who must work faster. Acker (2006, 181) notes that “struggles for power and control are often struggles over bureaucratic tools.” This repositioning focuses attention on how performance metrics, as bureaucratic tools, may be wielded in organizational contexts and with what gendered effects.
The increased use of performance metrics to alter worker and organizational trajectories has fostered scholarship about governance by indicators. This literature focuses on how the production and communication of numbers are leveraged for the expression of power (Espeland and Stevens 2008). Beginning in the 1980s, financial accounting practices extended into other sectors, with funding decisions increasingly determined by demonstrations of performance (Strathern 2000). In this new global “audit culture,” metrics may become technologies of governance when they name and define the social world, simplify what they measure, and are institutionalized to evaluate performance and inform decision making (K. Davis et al. 2012; Merry 2011). The organizational effects are most visible when a single evaluation regime is present (Lamont 2012), such as that of the U.S. News & World Report higher education rankings or the No Child Left Behind rating system (Colyvas 2012; A. Davis 2018). Considering the global shift toward institutionalizing performance metrics across sectors, scholarship must attend to whether, how, and why measurement practices influence organizational and professional decision making.
In the development sector, “audit cultures” have manifested in administrative reconfigurations to meet “results” demands (Dar and Cooke 2008; Eyben et al. 2015; Natsios 2010). Donor mandates have created what Merry (2016) calls “evidence-based governance” and Best (2017) terms “measurement-driven governance.” These ideas are enshrined in the highest levels of development discourse and planning, from the Organisation for Economic Co-operation and Development’s Paris Agreement on Aid Effectiveness—where quantified results are classified as one of five pillars for development (Best 2017)—to the adoption of goals, indicators, and targets in the Millennium and Sustainable Development Goals (Fukuda-Parr, Greenstein, and Stewart 2013). At the meso level, development organizations risk budget cuts unless they quantify results for donors and politicians (Rottenburg et al. 2015; Watkins, Swidler, and Hannan 2012). A culture of vigilance around accountability, transparency, and performance metrics now permeates the development sector (Eyben et al. 2015), laying the institutional and normative context for indicators to proliferate and offering a meaningful case of performance metrics in gendered organizations.
Methods
To illustrate how gendered organizations, social indicators, and professionals intertwine, I use the occupational vantage point of gender advisors serving organizations involved in a single-country instance of a 20-country bilateral agricultural development initiative, here called Farming Sustenance. Informed by my lived experiences as a development practitioner and seeking to understand how everyday development happenings are shaped by the rising “results agenda,” I completed 60 interviews with professionals serving offices across the evaluation system of Farming Sustenance. This study is based on the perspectives of 19 gender advisors, contextualized by those of 41 of their coworkers (project leaders, evaluators, and donor representatives), participant observation at gender-related meetings, and document analysis of project evaluation artifacts. Of the 19 gender advisors, eight work in the donor country and 11 in the recipient East African country. Of these 11, five serve in supervisory or consulting positions and six are frontline gender advisors who are responsible for gender mainstreaming across a 5-year portfolio worth approximately $260 million.
These interviews represent nearly full coverage of the initiative’s core projects, 1 including gender advisors (all local staff), their managers (all expatriate staff), and gender experts and donor staff (local and expatriate staff). Interviews were 1 to 2 hours, semistructured, and took place in English—the working language of these projects. Verbal consent was obtained, and recordings were made when individuals consented. Using ATLAS.ti, I began analysis by grouping data by office location and job title, and was struck by how gender advisors consistently voiced frustration, foregrounding gender as the most important category in their work life. Understanding the development office as gendered, I then coded transcripts for how professionals value performance metrics. To access the structural nature of how gender advisors are, in turn, valued by the institutionalized, donor-mandated evaluation system, I analyzed their narratives in relation to evaluation artifacts. Viewing discrepancies in narratives as demonstrable of power contestation and therefore analytically valuable (Hesse-Biber 2012), I privileged and contextualized the standpoint of gender advisors. I use pseudonyms for individuals to protect their anonymity 2 and to ensure analytical focus on the interactions between gendered organizations and evaluation practices. Although the data are limited to a single context—with its particularized and variegated gender dynamics—the managerial techniques, bureaucratic policies, and employment roles at the heart of this research are widely practiced around the globe.
Gender Advisors’ Everyday Worklife
Ridgeway (2009) encourages a multilevel analysis of gendered organizations that places individuals and organizational structures into relation. I first synthesize interview data and evaluation artifacts to build a structural account of the gender advisor occupation, and then use this organizational context to make sense of their interactional experiences. By employing gendered organizations as an analytical framework, the gendered nature of the development workplace becomes visible.
The Structural Subordination of the Gender Advisor
Although gender advisors are hardly a monolithic group (ranging in age and marital and parenthood status, they are educated through degrees and/or work experience and wear everything from brown flats to pink stilettos), they share a structural location within large-scale development projects. Official development discourse encourages strong investment in women at the site of the project, yet when these intentions are refracted through development organizations, gender-related work becomes tenuously positioned. A woman manager pondered “How can we deliver women’s empowerment when it’s not in our offices?” In partial response, I show that the structural subordination of gender advisors results from low numbers of women professional staff, the tokenization of the gender advisor role, the placement of gender metrics at the lowest level of the monitoring system, and team orientation to performance metrics.
In each of the field offices of Farming Sustenance, gender advisors are uniformly women and local nationals. Among the local project staff, men are technical staff with knowledge in agricultural development—for example, livestock fattening or market systems—whereas women serve as the singular gender advisor. In contrast to tokens in the United States (Kanter 1977; Yoder 1991), women here entered the development sector in a role not previously held by men while also being a numerical minority. Further, inequality regimes at the intersections of citizenship and gender are present: Although each office has at least one other woman employed in a non-gender–related professional position (technical agricultural staff, upper management, or titled gender expert), 3 these women are expatriates. The only man engaged in gender-related tasks holds the title consultant, and the primary focus of his work is not gender-related. Taken together, the gender advisor role is both visibly gender-typed, with the women serving these roles as numerical tokens and with a tokenized responsibility to address gendered power relations at the project site.
The tokenization of the gender advisor role is created through the reactions of the technical agricultural staff to the task of gender mainstreaming. Project leaders confirmed that most staff do not consider gender dynamics, with one project lead saying this results in “some funny things like groups of men being trained by male trainers about improved breastfeeding practices.” Conceptually, gender mainstreaming is a collegial team effort, with the gender advisor working as a facilitator among coworkers similarly committed to mainstreaming. Yet in practice, “gender work” is offloaded onto gender advisors by coworkers who do not incorporate a gender perspective in their own technical work. Hamdiya, a frontline gender advisor who recently joined the project, described attending a workshop where the nutrition technical team had written scripts for a theater drama about improved child-feeding practices. She lamented their reproduction of gender norms, saying, “The developed scripts were promoting a strong husband who shouts at the household and has a traditional way of household management. Even he has two or three wives. Why would the project promote polygamy?” As her job requires, Hamdiya requested that the all-men technical team rewrite the scripts to dismantle, rather than reproduce, gender norms. Through scenarios like this, gender advisors are tasked with disrupting the men-dominated space of the development field office. Several told me that simply hiring a gender advisor like themselves generates resistance to incorporating gender into the project activities, to the point that one project leader refused to hire a gender advisor. In this context, women gender advisors are perceived as vanguards for gender, as they singularly carry out their occupational tasks by correcting and advocating for the inclusion of gender to reluctant, predominantly men coworkers.
Gender advisors face resistance from their coworkers, partly because of how gender mainstreaming is incorporated into the measurement practices of development projects. These practices matter because the donors may stop disbursing funds if they become unsatisfied with an implementing organization, thus incentivizing strong, consistent performance on metrics. In other words, performance metrics have been institutionalized and budget allocation is made based on “results.” This institutionalization occurs in two forms: The evaluation of gender mainstreaming exists at the lowest level of the monitoring system, and staff teams are configured toward the accomplishment of nongendered project objectives.
First, gender mainstreaming is a complex process meant to bring a gendered lens to all aspects of a project, yet it is often only tracked through sex-disaggregated data and the achievement of a target for women’s involvement. For example, an indicator on an education project may be “the enrollment of students” with a target of 30 percent girls. Although indicator targets are used to track progress toward the overall project objective, they may also be established for any level of disaggregation—typically breaking down counts of project participants by age and sex. This latter disaggregation is typically used to ensure women’s involvement in project activities, reducing gender to a simplistic understanding of woman or man. In a context firmly committed to measurement practices and a discursive focus on reaching smallholder farmer women, one would expect to see gender present in the monitoring system, yet Table 1 illustrates how gender is actually obscured.
Example of Gender’s Disappearance within Performance Metrics
Development staff further report that achieving the agricultural-related indicator targets takes precedence over achieving sex-disaggregated targets. One project lead, voicing frustration with the donor’s emphasis on results and given women’s current levels of literacy and numeracy, regretted that he could not “slow down” to better reach women and hit sex-disaggregated targets, noting that it is “more important” to achieve the main project results. Despite the conceptual underpinning of gender mainstreaming as gendering all aspects of a project, it is tracked only at the lowest level of these measurement practices, allowing “more important” components to take precedence.
The second way gender mainstreaming, and thus gender advisors, is precariously positioned occurs through staffing configurations by performance metric. Although it is common in bureaucracies for staff to be divided by work task, spurred by contractual obligations to donors to deliver results, development organizations often arrange their staff by “intermediate result” (commonly referred to as “IR”). Teams are designated as “IR1 team,” “IR2 team,” and so on, and progress reports to the donor are compartmentalized by IR. Using this lexicon normalizes the accomplishment of results as the primary goal: The work task is to achieve results, not the substance it is meant to measure. In contrast, gender advisors are not aligned to a single IR but must “mainstream across,” measured solely by hitting sex-disaggregated targets. Elaine, who is local staff and advises multiple projects, explained how their role is positioned outside of the core project: We have to show our coworkers how gender is important to reach their objective. Usually projects are designed to meet some other goal, like increase agricultural productivity, not to increase women’s standing. [Increase] nutrition outcome—sure, but to improve nutrition you must improve the status of women. It’s not enough to address gender through other [objectives], but people aren’t committed enough. [emphasis added]
Thus, gender advisors face a structural conflict: They need to work alongside the IR teams to programmatically mainstream gender, but the IR team does not inherently need to incorporate gender—they must accomplish the IR’s associated indicators, not the sex-disaggregated targets.
In addition to gender advisors being numerical minorities in a newly created and gender-typed role, the work of gender mainstreaming is buried at the lowest level and positioned relative to the core project objectives in ways easily dismissed as peripheral. These practices intertwine to compound the marginalization of “gender work” in the development workplace. This structural location affects how gender advisors interact with and are treated by their coworkers, eliciting emotional fatigue in gender advisors.
The Emotional Fatigue from Workplace Interactions
Gender advisors are required to campaign for the inclusion of gender in the project through overtly gendered interactions in workplaces in which they are structurally subordinated. In other words, structural subordination and interactional resistance coproduce emotional fatigue in gender advisors, which creates a differential work experience for them.
Despite official policies, gender advisors and project leaders describe efforts to incorporate gender as piecemeal and driven by individuals working “against the grain.” In contrast, most technical IR staff report positive engagement in gender efforts—albeit bureaucratically (e.g., “We have sex-disaggregated targets and target accordingly”)—while acknowledging gender’s relative absence from projects. None recognized themselves as a potential barrier. Elaine explained: I have to plan to meet people and assist them to better integrate gender. How can I help them? How can I convince them? You have to be a professional. I hate my job. Really. I have to analyze and interpret, I have to use diplomacy, advocacy, knowledge, and expertise to convince people.
If Elaine’s coworkers were truly participating in gender mainstreaming, “convincing” them would be unnecessary. This interactional resistance breeds frustration, resulting in Elaine hating her job as a de facto vanguard for gender. Thus, the gender advisor must constantly muster her skill set to garner coworker support.
Such feelings were persistently present as the gender advisors described their job tasks. Although gender mainstreaming is a goal for the project, gender advisors describe being “exhausted” by engaging their coworkers in the office. An expatriate gender expert, Kathee, spoke of such gendered fatigue, saying: I have to come to terms with how my coworkers think and take them on a journey to a new way of learning. And that is an exhausting process [she laughs] and there’s not a lot of stamina or support for gender advisors on how we do that, on how we communicate, on how we have to stand up and be ridiculed as step number one. And how we deal with that and how we move past that. And how we constantly feel like a nag, and then we’re told we’re a nag. And then we’re buying into gender stereotypes again because generally we’re female. [emphasis added]
Gender advisors “feel like a nag,” partly because they are positioned outside the core project structure and required to correct their coworkers—typically after they have completed planning, design, or even partial implementation. Not only are interactions substantively about gender, they are also about gendered power dynamics: It is hard to imagine women calling men “nags” in the workplace.
At times, these dynamics manifest as intimate, overtly gendered interactions between coworkers. Two gender advisors, working in separate offices, independently mentioned that men coworkers commented they were annoyed by the noise from women’s heeled shoes. In the first instance, both people were expatriates and, in the second, both were locals, foregrounding gender. The expatriate gender advisor said she was walking down the hallway, when another expatriate man ran out of his office, yelling “It’s you! You’re the one! Your heels—clack-clacking down the hallway! I cannot even concentrate!” Another example of highly personal interactions includes a local gender advisor who recounted a conversation with a local donor representative. He segued from the women’s empowerment agenda of his employer by saying to her “If you ladies keep on talking like this, you’ll be divorced.” The donor representative generalized from her to all gender advisors through the use of “you ladies” and then portrayed the donor goal of women’s empowerment as undermining the valued social institution of marriage. She rebuffed his statement, but contextualized the story as demonstrative of men’s resistance in the workplace.
In sum, the common practice in development projects of having one gender advisor who must “get gender in” to nongendered programmatic activities means that gender advisors must interact with their reluctant IR team coworkers. These interactional realities individualize the gender advisors’ structural subordination, producing emotional fatigue and a work environment markedly different from that of their men coworkers.
Gendered Valuations of Measurement Practices
Gender advisors face marginalization partly because of measurement practices that place gender mainstreaming at the lowest level, as sex-disaggregated indicator targets, whereas other staff focus on accomplishing the higher-level intermediate result indicators. Given this circumstance, how do gender advisors interpret the value of sex-disaggregated data? Most gender advisors noted that numbers do not provide insight into what “really” happens in women’s lives as a result of development programming: “We can say ‘20 percent women’ but it doesn’t show how we can transform their lives unless we put a qualitative indicator.” All the gender advisors recognized that quantitative data provide insight into the prevalence of a given issue, but, like Uwimana, a former frontline gender advisor turned consultant, believed that qualitative data provide a better picture of project efficacy. Uwimana said: Qualitative [data] really tells you women’s position: access to resources, how women really own it. . . . If you just have numbers, you’ll see the 40 percent [target is reached], but you have to look beyond the numbers to see the real situation. [emphasis added]
Uwimana understands qualitative data as more “real” and worries that quantitative indicators may be misleading. Chungwa, a former frontline gender advisor and current expert overseeing national agricultural policies, was frustrated that her office could not hire a consultant to qualitatively investigate the reasons for women’s lower project participation rates. She said: We couldn’t find a good gender researcher apart from “sex-disaggregated data.” Gender has different faces. If you go to the agricultural technology people, they say, “Women are not there, women don’t use extension.” Why is that? Because of culture. Okay, what kind of culture? It’s religion. When you go to religion, what kind of religion is that? Why is the religion biased? So you want to talk to people. [emphasis added]
In Chungwa’s interpretation, qualitative methods can trace backward from socially embedded effect to cause—an insight that sex-disaggregated data do not provide. Meti, a young frontline gender advisor recently trained at the master’s level in research methods, was skeptical of quantitative methods. She said: I believe that quantitative methodologies are predetermined. Maybe it’s the data collector, maybe the questions are leading, and the women just answer how they think they should. Qualitative is best, with in-depth interviews we can capture the real change of the women. It’d be better, the reliability, the validity. By using quantitative we may not get the real experience, it’s not enough on its own, it needs to be triangulated with qualitative. [With quantitative data] we can’t get reliable info.
This is a surprising reversal from dominant understandings of qualitative as more interpretive: By invoking “reliability” and “validity,” Meti describes qualitative data in words typically reserved for quantitative data. Meti, like her counterparts, views qualitative data as better able to accurately capture women’s lives in all their intersectional complexities.
Despite this skepticism, gender advisors voice a simultaneous desire for quantitative metrics. One advisor, discussing large-scale surveys, remarked that “It’s statistical, it’s scientific. I appreciate that we get some number.” Under the logic established by these gender advisors regarding qualitative data, we would not expect them to also request quantitative measures. So why does this happen? I demonstrate that their desire is tied to their understanding of quantified metrics not as meaningful measurements of the women they seek to assist in the project, but as tools for gaining power in negotiations with coworkers in the office.
Bureaucratic Tools: The Appeal of Quantified Metrics
The appeal of quantified metrics to gender advisors becomes clearer once viewed through the lens of gendered organizations. These workplaces are gendered in ways that affect the value of performance metrics, existing and imagined, to gender advisors. Although quantitative indicators are often interpreted as useful externally, gender advisors understand their utility internally for potential workplace changes. Gender advisors are interested in quantitative indicators to, first, convince coworkers who are dismissive of gender mainstreaming as a project imperative, and, second, to force their coworkers’ compliance with gender mainstreaming.
Gender advisors believe that numbers appeal to their coworkers as “evidence” and, as such, value numbers as a tool for garnering support in an otherwise hostile workplace. Elaine explained: I have to convince people why it’s important to include gender. They don’t want to [include gender], but it must be there. We must have data or evidence, otherwise how can we get people to work on this? I must be equipped enough to convince them.
Her use of “enough” implies a search for the minimum standard of evidence needed to engage coworkers in gender mainstreaming. Chungwa adds nuance by citing the numerical dominance of men in agriculture and their understanding of what constitutes “evidence.” She said: Most of the staff in agriculture are men and say, “Women are not involved—they want to be in [micro]finance, they want to be in the house.” That attitude is there, so the women are not trained. The staff want to have an evidence-base—“Why do we need gender and technology? Technologies are neutral, why do you say ‘female friendly technologies’?” . . . You need to have an evidence-base that says, “because of this and that,” “because it decreases women’s time from six hours to you know.” Numbers, research, evidence—they [the staff] look for that.
Chungwa continues that gender-related outcomes are unknowable in advance, but that her coworkers are not comfortable operating without an “evidence base.” Rosalie, a frontline gender advisor, highlights the importance of evidence to prove the worth of the gender advisor position, noting that qualitative descriptions of progress are undermined by accusations of limited representativeness or generalizability. She said: The reason I need quantitative thing in general is to say, “Ok this is our evidence.” I don’t want anyone to later say, “Yeah qualitatively some have progressed and some have not, so where is your result?” I want to be able to say, “Okay because of this intervention, it has moved from this to this much percentage” or whatever.
Rosalie needs numbers to proactively combat critiques about the effectiveness of her job, yet she went on to describe her preferred method of capturing project efficacy as diaries written by women participants. This suggests her interest in percentages is solely a reaction to coworker expectations of data: defined by them as quantitative. Her perceptions were confirmed by project leaders, one of whom bluntly stated “I don’t want qualitative, it’s too anecdotal.”
These gender advisors perceive their coworkers as associating numbers with “evidence.” Although the advisors desire qualitative data, they are drawn into a quantitative logic in an effort to “convince” their coworkers of the value and impact of engaging women. The imperative to convince is magnified by the structural location of gender mainstreaming; gender advisors have no choice but to meet the knowledge demands of their coworkers because gender mainstreaming is outside the core project and only successfully incorporated if coworkers are first convinced and then cooperate in mainstreaming gender through each project component.
Gender advisors, structurally marginalized by measurement practices and emotionally fatigued, search for expeditious strategies to garner support and cooperation. Bernadette, a former donor-based gender expert, explains, “Gender is an integral part of the project, but lots of times it’s language and not reality. It’s on paper, but it doesn’t happen. When we put down numbers, it’s to force that into reality.” When gender advisors discuss their desire for indicators, they present them not as meaningful measurement tools about the women development projects are meant to engage, but as compliance mechanisms for their coworkers. Gender advisors switch their sentence structure to position metrics as subjects capable of performing action: “numbers force” attention to gender, “targets hold people” accountable regardless of their attitude toward gender mainstreaming, and “indicators force” the inclusion of women. In other words, metrics are recast as neoliberal managerial tools meant to induce behavior toward the desired aim of gender mainstreaming.
Gender advisors’ desire for more quantitative indicators, now understood as managerial, demonstrates that a struggle for power and control is under way. Recasting quantitative metrics as compliance mechanisms brings their power as potential tools for the gender advisors into focus, as Uwimana described: Always gender is considered an added-on responsibility. What is on paper holds people accountable. Some don’t believe in targeting women. You hear that sometimes and it shows people’s attitude. Even if they have excuses, targets hold people [responsible].
In Uwimana’s formulation, indicator targets override coworkers’ unsupportive attitudes about gender as an integral project element. This attitude ultimately leads to poor inclusion of women project participants in IR-specific programming, and sex-disaggregated targets then become a solution. Hamdiya illustrated the ability of targets to require IR teams to include women: The gender advisor before me, she faced a challenge: people in IR4 said, “We don’t have indicators on gender, so nobody is going to measure us.” It was a challenge to push, pushing and forcing them to focus. . . . In my observation it is only the indicator [targets] that forced them to include women.
This perspective confirms the ability of targets as tools to mandate staff attention to gender, while also highlighting that IR teams may use measurement practices to justify not working on gender mainstreaming. In doing so, they substantiate the structural marginalization of measuring gender as sex-disaggregated targets. Still, gender advisors find that even the few, existing low-level indicator targets (see Table 1) help them make the case for engaging women. Because the IR team tries to dodge working on gender under the logic of “no measurement” (for them), this leads gender advisors to believe that if there were gender mainstreaming indicators, IR teams would comply. Rosalie agrees and advocates for developing such indicators. She said: We often say, “Gender is crosscutting, it doesn’t need indicators.” No! Crosscutting doesn’t mean it doesn’t require indicators—you need to have them! You need to have an indicator in there! Measure it then! Then they won’t be forgotten!
By advocating for the measurement of gender at the indicator level (rather than as sex-disaggregated targets), Rosalie imagines she can leverage such numbers to better ensure that attention and resources are allocated to gender mainstreaming. Yet gender advisors’ request for more higher-level indicators means they end up playing by the rules of a quantitative knowledge production game—one they do not actually want to play.
This is not to argue that strengthening evaluation systems is the solution, but rather to show that gender advisors strategically assess the tools available to help get their jobs done and see indicators as the most expeditious option. And with good reason: Every day, gender advisors observe coworkers and a project, organization, and donor oriented to performance metrics. The common mantra “what gets measured, gets done” is their lived reality. Thus, they imagine that greater inclusion in performance metrics is the best solution, not for creating knowledge about how to enrich women’s lives—their true desire—but for their pragmatic utility to overcome the emotional fatigue and structural marginalization they endure in the development workplace. From the gender advisor’s perspective, quantified metrics are hard and fast bureaucratic tools that shift sole responsibility for “gender-inclusive development” away from the gender advisor to a shared responsibility with the technical IR teams. In other words, gender advisors become interested in trying to access and harness the reactive power of indicators as a manager would do. A gendered organization lens draws our attention toward the contested workplaces in which gender advisors operate and reveals their strategic use of metrics to advance their own aims.
Conclusion
Although this case study speaks to an East African development landscape, the presence of quantitative indicators mandated by international development projects is global in reach. As I have shown here, development organizations are gendered, which leaves women gender advisors sidelined, struggling, and seeking ways to advance their goal of gender mainstreaming. In a bid to advance this agenda, gender advisors repurpose quantitative measures to assert more power in a gendered work context. This empirical case results in two theoretical contributions.
The first theoretical finding of this research is that gendered organizations theory is foundational to analyzing how professionals respond to the growing influence of performance metrics in the workplace. Much of the quantification scholarship views rankings as “engines of anxiety” (Espeland and Sauder 2016), yet these gender advisors imagine performance metrics as potential mechanisms of relief through the devolution of responsibility. The gender advisors’ assumption is that, if better included, gender-related metrics would devolve responsibility away from themselves to a collective staff responsibility, inspiring coworker compliance rather than complaint and dispute. In contrast to Levy (2016), who finds that masculinist identities prompt truckers’ resistance to quantified performance metrics, I find that the emotional labor of women gender advisors, marginalized structurally and interactionally, prompts endorsement of metrics. Levy (2016) finds that gendered notions of self animate resistance to metrics, yet I argue that gendered work experiences can transform the gender advisors’ self-preference for qualitative measures into a pragmatic advancement of quantitative metrics. Both truckers and gender advisors seek ways to optimize their feeling of control over their work, and these individual preferences interact with gendered organizational components to produce potentially opposite outcomes.
Incorporating gendered organizations as an analytical component is therefore critical to scholarship, especially as workplaces become increasingly quantified, digitized, and surveilled. Acker (2002) encourages research on the devolution of responsibility under globalization. Performance metrics offer an important, yet mundane, pathway for such devolution. These gender advisors are not merely “acted upon” by globalization; they further the internationalization of quantified metrics by expressing agency in conditions not entirely of their making. Therefore, future research on professionals’ reception of and reaction to performance metrics must account for intersectionality, inequality regimes, and/or gendered organizations.
From a knowledge standpoint, gender advisors want thick description about the “real change” in women’s lives elicited by development projects but are caught up in a gendered organization that institutionalizes achieving quantitative targets for women’s participation. Building on the feminist debate about quantitative and qualitative data (DeVault 1996), I show that attention should be less about whether one form is better at capturing women’s lives but about the interactions, practices, policies, and organizational relationships that encourage and value particular forms of knowledge over others. I demonstrate how the gendered nature of organizations can change potential resistors of quantified metrics into promoters because of their pragmatic utility.
This leads to my second theoretical contribution: When gendered organizations institutionalize quantified performance metrics, knowledge production shifts to match the context. In industries caught up within calls for “evidence-based” decision making, one would expect organizations to be interested in whether quantitative or qualitative data best capture the “real” situation of women and gender relations to inform efficacious programming, policies, training, marketing, and more. Yet this dismisses the reactive power of quantitative metrics to alter organizational behaviors: Development organizations instead assess the utility of data to communicate results to their funders. This case has implications for organizations more broadly, in which quantified indicators are used to manage from a distance, devolve responsibility, and externalize risk. Numbers remove context, quieting probing questions, and enable aggregation, standardization, and comparison among diverse geographies and cultures across time. In doing so, numbers gain authority as masculine: objective, trustworthy, and suitable as evidence for decision making. Numbers are not necessarily superior measures, but are perceived as better, which increases their utility to organizations.
In a context where the allocation of funding requires demonstrable and uncontestable results, the value of quantified metrics is increased and reproduced. I coin the term “the paradox of quantified utility” to describe how, in contexts of institutionalized performance metrics, professionals re-orient to the metrics, not necessarily because they want to participate in this way, but because they are (or imagine they will be) rewarded for engaging in quantification and quantified knowledge production even at the cost of what they may recognize as the substance of their work. This paradox is likely felt most by marginalized people, such as gender advisors, and communities who engage in a betrayal of their own preferences to try to garner attention and resources. Yet given the very real material advantages that engaging in quantification may provide, their actions are strategic, albeit constrained.
In contexts that institutionalize “results,” resources increasingly flow to easily quantifiable outputs or processes, and knowledge production becomes recursive. The pragmatic utility of quantitative metrics for organizational survival outweighs rich knowledge production. Quantitative indicators ignore diversities in gender relations, religions, sexualities, reproduction, and productive labor, among other areas—resulting in knowledge useful for policy makers who must demonstrate return on investment. Ultimately, development organizations, by relying on sex-disaggregated indicator data to capture “gender,” produce knowledge that presents women as individuals to be acted on, rather than as relational and dynamic community members grappling with gender norms as they respond to urbanization, migration, climate change, and capitalist development, among others. I demonstrate how the pragmatic utility of performance metrics as bureaucratic tools recruits professionals and organizations into quantified knowledge production in ways that, paradoxically, may impede knowledge creation about the structural forces embedded in women’s everyday lives around the globe—knowledge that may actually assist organizations to engage with women on their own terms. The importance of what constitutes knowledge production under the “results agenda,” its relative salience, and source of value is a key theoretical finding of this research.
Practically, attention to organizational policies and pressures that steer knowledge production down particular paths is paramount. Professionals in development bureaucracies at the donor level have argued for engaging women because they produce a higher return on investment (i.e., the “business case”). My research suggests this action is likely explained by the utility of the business case to garner budget allocation in a gendered organization. Yet it also suggests a potential (partial) way out of the paradox: Feminists, practitioners, and scholars should labor to showcase the utility of qualitative data to provide rich insights for organizational decision making. As Irene noted in the opening epigraph, quantitative knowledge is biased against Global South knowledge. Grassroots organizations’ ability to deliver meaningful and sustainable change in people’s lives—substantive results—is often informed by knowledge that is situated, deeply contextual, and not standardizable. A movement to showcase the utility of qualitative ways of knowing would aid grassroots organizations to maintain their competitive advantage within the results landscape. Although this imagined solution may create more space for different ways of knowing more immediately, it still churns within a logic of utility, begging the question: At what cost in the long run? Conceptually, scholars must remain vigilant about modes of knowledge production that devalue the relational and intersectional dimensions of social issues. Thus, gender scholars must rapidly engage the dynamic intersections of quantification and gendered organizations theory to analyze the variegated role of numbers in society and their impact on knowledge production.
Footnotes
Author’s Note:
I thank Liz Boyle, Phyllis Moen, Rachel Schurman, Annie Jollymore, Erin Hoekstra, Hanife Cakici, Devika Nayaran, Madison Van Oort, Sarah Springer, the anonymous reviewers, and editors for their thoughtful feedback. This research was supported by Social Science Research Council’s Mellon International Dissertation Research Fellowship and the University of Minnesota’s Schwartzberg Workable World Trust Fellowship.
1.
The exception was one project that had concluded, and former staff were unreachable.
2.
For discussion regarding the importance of anonymity when researching evaluation practices, see Eyben et al. (2015), Mosse (2006), and 
.
3.
This article is focused solely on technical project staff. Women play crucial yet gender-typed roles in development offices as secretarial and cleaning staff.
Emily Springer is a PhD candidate in sociology at the University of Minnesota. Her research focuses on evaluation practices, complex organizations, knowledge production, technology, and gender in the development sector.