The Isolated Choice Effect and Its Implications for Gender Diversity in Organizations
People are more likely to choose candidates whose gender would increase group diversity (e.g., women in a male-dominated environment) when hiring multiple group members rather than when making a single, isolated hiring decision.
Many organizations and businesses in the U.S. acknowledge the importance of increasing diversity in their workplaces. For example, the majority of Fortune 500 companies have executive positions in charge of managing workplace diversity initiatives and publicizing diversity statements. However, progress in gender diversity in many organizations has been slow, and research has shown that well-intentioned diversity messaging may cause backlash.
Prior research suggests that individuals’ beliefs and implicit attitudes about gender are extremely difficult to change. As a result, interventions that aim to shift people’s attitudes to improve diversity and inclusion may have a low success rate. Instead, a growing area of research focuses on how manipulating “choice architecture” (that is, the design and structure of a decision-making environment) when people are making critical workplace decisions (e.g., selecting a candidate, writing a performance review) can impact gender inequality. For example, research suggests that implicit gender biases exert a stronger influence on evaluations of candidates when evaluators look at one candidate at a time for an open role (separate evaluation) than when they look at multiple candidates for the role at the same time (joint evaluation). Thus, a manipulation of the choice architecture of hiring decisions (whether one candidate is evaluated at a time or whether all candidates are evaluated at the same time) can improve gender equity.
The present research also identifies an intervention that manipulates aspects of the decision-making environment to improve gender diversity in hiring decisions. Specifically, the authors draw on the insight that diversity is a group-level property, not an individual property. People may therefore be more likely to (1) notice diversity and (2) weigh diversity in their decision-making when they make multiple hiring decisions at once (a collective choice) rather than one hiring decision at a time (an isolated choice).1 Thus, the authors hypothesize that groups formed by aggregating isolated hiring decisions will be less diverse than groups formed by an individual making multiple hiring decisions simultaneously.
1Notably, all hiring decisions in this work are “joint” decisions, as multiple candidates are considered at the same time for each open role. The manipulation does not affect the number of candidates per role but the number of open roles that each decision-maker has to fill.
People are more likely to choose candidates whose gender would increase group diversity when hiring multiple group members rather than when making a single, isolated hiring decision. These findings demonstrate what these researchers call the “isolated choice effect,” suggesting that a group whose members are selected by aggregating isolated decisions will be less diverse than a group whose members are selected simultaneously, as a set.
- When making hiring decisions for a technology company in which women made up 26.7% of the candidates, people making isolated hiring decisions selected women 7.4% of the time, whereas those making hiring decisions in a set selected women 18.0% of the time. A similar effect was found when people were asked to make real decisions, such as hiring for a business pitch competition or recommending speakers for an academic conference.
- When selecting authors for a high school curriculum in which women made up 32% of the candidates, people making isolated decisions selected women authors 24% of the time, whereas people selecting authors in a set selected women 29.5% of the time.
- People making hiring decisions in a set reported that diversity was more salient to them (that they thought more about diversity when making their hiring choice(s)), compared to people making a single, isolated hiring decision.
- Drawing attention to diversity (by emphasizing that diversity is one of the company’s hiring priorities) greatly increased the likelihood that people would select women when making isolated hiring decisions (causing them to choose women 36.5% of the time instead of 15.0% of the time).
Organizations typically make hires in isolation, rather than in sets. These findings show that changing the choice architecture of the hiring process so that hires are made in sets rather than in isolation can be a successful intervention for increasing organizations’ gender diversity.
In this paper, researchers conducted a series of six randomized experiments to investigate the impact of making hiring decisions in a set rather than in isolation on the gender diversity of hires. The randomized experiments were conducted on 3,509 participants recruited from Amazon Mechanical Turk (MTurk). The MTurk participants were excluded if they did not follow instructions, as stated in the researchers’ pre-registered analysis plans. In all studies, the authors compared the proportion of hires that were women across the two conditions.
- In Study 1, participants were given a prompt in which a technology company was looking to fill five different roles. Participants were randomly assigned to either make a collective choice (to make a hiring choice for each of the five roles) or to make an isolated choice (to hire one person for one of the roles). In either condition, the participants were shown the same candidate profiles (including their photos, gender, work experience, etc.) for each role.
- In Study 2, the same procedure was repeated with a non-hypothetical scenario. Participants were asked to recommend authors for a high school English curriculum and were assigned to either share five of their top-ranked authors, or just one of their top-ranked authors.
- Study 3A followed the same procedure as Study 1 with one addition: in order to assess the mechanism for the documented effect, participants were asked how salient diversity was to them, on a scale of 1 to 7, when making their hiring decision(s). In Study 3B, a similar procedure was used, but some participants were also randomly assigned to a “diversity valued” condition, in which they were explicitly told that the company values diversity, or a control condition, in which this statement was omitted.
- In Study 4, real, instead of hypothetical, decisions were tested. In Study 4A, participants were asked to hire graduate students to participate in a business pitch competition, in which their bonus depended on the performance of the candidates they hired. In Study 4B, participants were asked to recommend speakers for an academic conference.
One limitation of these studies is that the researchers only examined settings where women were underrepresented. In their online supplemental materials, the researchers also present a study in which they examined the same procedures in conditions where men were underrepresented. They replicate their findings and find in the supplemental study that participants were more likely to hire men (the underrepresented group) in the collective choice condition.