Authors: Ana Gugushvili, Noa Fernandez, Riccardo Curti, Zere Aktay, Sofia Pickert
Key Points:
1. Research investigating the effect of diversity on job performance should always be analysed critically to assess the research methods used, as the common use of subjective performance ratings can lead to biased – and therefore misleading – results.
2. Subjective evaluations, especially from external raters, tend to overestimate the benefits of job-related diversity and underestimate the value of demographic diversity due to unconscious bias.
3. Job-related diversity is more beneficial for complex and innovative tasks, while demographic diversity does not negatively affect performance. However, neither type of diversity has a universal, large-scale impact on performance, highlighting the need for nuanced evaluation.
Diversity is a buzzword in workplace conversations – but do we truly understand its impact on performance? Since the start of 2025, there has been an increase in the number of conversations about Diversity, Equity, and Inclusion (DEI) within the workplace. You might also note that these conversations often drift into the realm of politics, overshadowing evidence-based insights. Alongside this trend, the persistent belief that demographic diversity undermines employee performance has resurfaced. However, this claim was already critically examined by Van Dijk et al. (2012), whose insights appear to be more relevant than ever.
Demographic diversity – referring to a variety of differences in characteristics among a group of people, such as ethnicity, age, gender, or educational level among others – does not inherently decrease job performance. Instead, biases in evaluation often leads to an underestimation of its value while exaggerating the benefits of job-related diversity – differences in aspects such as career or professional tenure. Imagine a demographically diverse marketing team presenting a campaign idea. Their manager, who subconsciously favors familiar perspectives, rates a more homogeneous team, one with less cultural differences and similar profile in terms of age and gender among the team members, as ‘more aligned with company culture,’ even if the diverse team’s idea is equally strong if not stronger. This bias skews workplace evaluations, reinforcing false beliefs about diversity’s impact by preventing the objective and accurate analysis of the effect that employee diversity has in organizations.
Where We’re At – Current Conventional Wisdom
Many researchers and managers share a common misconception: demographic diversity (i.e., age, ethnicity, gender, etc) harms performance, while job-related diversity (i.e., functional background, work experience) enhances it. A leading cause of this belief is biased research methods. Most studies rely on subjective performance ratings instead of objective data, distorting the real impact of diversity.
Subjective measures of performance include qualitative performance ratings, such as reflections on the workers from team members or team leaders. Objective group measures would include quantitative evaluations, such as financial performance (production figures) or number of ideas generated by a team. Additionally, many studies fail to specify which aspect of job performance – such as innovation or task complexity – they assess, which leads to the loss of valuable insights.
Research on diversity and job performance has revealed that depending on the type of diversity that exists within a group, performance will be affected differently. For example, diversity may enhance the quantity of information available to members, thereby providing more resources for problem-solving. On the other hand, less diverse groups may work more cohesively due to greater harmony caused by their perceived similarities. Taking the different types of diversity into account, people often assume that demographic diversity harms job performance, as members of the group may be judged prematurely or subconsciously. In contrast, job-related diversity is believed to solely increase job performance by improving information elaboration without introducing many physical differences which could trigger that bias. This assumption has shaped workplace policies, sometimes leading organizations to prioritize professional diversity while overlooking the benefits of demographic diversity. Let’s see how this belief has come to be so well known by focusing on the key factor that led to this scientific assumption in the first place: bias.
What Data Doesn’t Tell You
Rater bias refers to how certain prejudices or stereotypes affect how an employee is being evaluated, usually by judging their performance to be worse than it is. In demographically diverse teams, this bias is more impactful as more minority groups are represented in those teams. Furthermore, with demographically diverse teams, there is a lower chance of similarity between the performance evaluator and the subjects. Since demographic similarity is generally linked with a more positive evaluation of a group, the evaluator may be subconsciously favouring homogeneous groups that they feel they belong to. Lastly, the optimistic organisational attention that job-related diversity has received is also likely to cause managers to draw early conclusions and forget to judge the trustworthiness of data. Overall, these biases lead to an underestimation of the effect of demographic diversity, and an overestimation of professional diversity on performance.
Van Dijk et al. (2012) confirmed this using Pearson’s correlation coefficient r. In subjective performance assessments, demographic diversity had a small negative effect (r = -0.05), while job-related diversity showed a slight positive effect (r = 0.04). Such differences were not found in objective measures of performance (e.g. financial performance, questions answered correctly, ideas generated by a team). Since the effect sizes for all studies are very small, any changes have a dramatic effect on percentage change, which is why it should be kept in mind that most effect sizes remain negligible. Interestingly, the effect sizes for demographic diversity change from r = -0.01 to r = -0.14 (ethnic diversity) when switching from objective evaluations to subjective ones respectively. In fact, during the objectively measured research reports, it was found that for basic tasks, neither type of overall diversity – demographic or job-related – affect performance at all. This further highlights how subjective measurements strongly differ from objective findings when it comes down to ethnic differences.
Overall, this shows that subjective ratings might distort performance evaluations, making them less reliable substitutes for objective measures. So, how do different evaluation methods impact our understanding of diversity’s role in performance?
Bias Within Subjective Measurements
There are generally two ways that subjective measurements of team performance are done. Either performance ratings come from external evaluators – such as team supervisors or managers – or from internal evaluators – such as team members or internal team leaders. Internal evaluators are expected to have spent more time with the team, therefore learning to look past the superficial composition of the team and being able to notice additional attributes of individuals, beyond only the qualities that make them diverse. This means that internal evaluators are less likely to fall for rater bias against demographically diverse teams. On the other hand, external evaluators, who do not have the additional insights into the teams, are expected to show stronger bias against demographically diverse teams.
A closer look at how performance is assessed reveals a key distinction: whether the evaluation comes from within the team or from an external observer. Van Dijk et al., (2012) revealed a small negative bias (r = -0.06) in how external evaluators rated diverse teams. In contrast, internal evaluators, who work alongside their team members daily, do not show the same negative bias. As an example; in a diverse marketing team, an external manager might unconsciously rate a homogeneous group as more ‘efficient,’ even if both teams deliver similar results.
When assessing job-related diversity, external team leaders reported a small but statistically significant positive correlation with job performance (r = 0.09). However, internal team members did not observe this, or any other significant, relationship. This was similar to the results found from the objective measurements, adding credibility to their assessment. Here’s an example:
A multinational firm might have two project teams – one composed of employees with diverse professional backgrounds (e.g., engineers from different fields and number of years working) and another with high demographic diversity (e.g., employees from different cultural backgrounds, genders, and ages). An external manager assessing these teams without direct involvement might assume that the functionally diverse team performs better, even if both teams deliver comparable results as subjective evaluations can be shaped by subconscious prejudice or bias, rather than actual performance outcomes.
Complex Tasks, Diverse Teams. What works?
Complex tasks that require deep knowledge processing and novel solutions benefit from teams with diverse skills and experience. In contrast, having a mix of ethnicities or backgrounds doesn’t automatically boost performance – it depends on how well the team collaborates. Looking only at complex tasks, research found no significant link between demographic diversity and job performance, while job-related diversity showed a small but positive effect (r = 0.06). For these calculations, it is not clear whether only objective, subjective, or a mix of both evaluation methods was used. Regardless, this overall trend suggests that job-related diversity supports complex tasks, whereas demographic diversity has no significant impact. For routine tasks, neither type of diversity influenced performance.
What About Innovation?
Innovative performance refers to tasks that require creative use of team resources and is opposed to in-role performance. For in-role performances, members are simply asked to complete tasks aligned with their normal job requirements. Team innovation requires greater out-of-the-box and critical thinking, two processes diversity is often assumed to enhance. Therefore, both demographic and professional diversity are expected to positively influence innovative performance.
Following the framework of componential theory, which outlines that three factors – relevant knowledge, cognitive ability, and intrinsic motivation – are needed to allow for the expression of creativity,job-related diversity is expected to have a stronger positive impact on innovative performance out of the two diversity categories. This is because job-related diversity strengthens the first two out of the three components needed to enable greater innovation. Furthermore, domain-specific knowledge can increase confidence, and thereby comfort to explore novel solutions.
In the study, demographic diversity had a small, positive effect size of r = 0.02 for innovative tasks, and job-related diversity boosted an effect size of r = 0.09 across the different diversity clusters. Interestingly, diversity within educational levels (referring to the different tiers of formal education, such as bachelor’s or master’s degrees) led to the highest effect size of r = 0.20, though it should be mentioned that this specific characteristic was only investigated in three papers.
Speaking of Reliability, Are These Findings Noteworthy?
It can be seen that all effect sizes for different diversity clusters remained small. This does not however mean that effect sizes for all teams and in all organisations will be small. Instead, the role of different diversity moderators and different tasks should be considered. Additionally, integration of diverse teams may play a role in how diversity affects job performance.
Takeaways for your practice
The relationship between diversity and job performance is not as clear-cut as conventional wisdom suggests. While job-related diversity can offer advantages in complex and innovative tasks, demographic diversity is not inherently harmful – its perceived negative impact often stems from evaluator bias rather than actual performance differences. In general, diversity does not seem to be related with job performance, therefore, there is no need to abandon DEI programs as they can still help to avoid discriminatory practices and promote equal opportunities.
• Diversity on its own should not be treated as a fix-all solution, nor as a deterrent of job performance. Instead, organizations should focus on reducing evaluation biases and creating inclusive environments where diverse teams can thrive.
• If your organization relies on performance evaluations, prioritize objective metrics over external subjective ratings, as the latter may unintentionally disadvantage demographically diverse teams.
• Ultimately, diversity’s impact depends on how organizations foster collaboration and ensure fair evaluation practices, because diversity alone isn’t enough; successful inclusion further drives success.
Did you like this article? share it with your network by clicking on the buttons below!
Follow us on LinkedIn, Twitter and subscribe to our newsletter to receive all the quality of scientific research curated by us for you in a clear, engaging language.
Trustworthiness score:
This meta-analysis studied and combined the results of 146 carefully selected research papers, ensuring that each of them had a computable effect size representing the relationship between work diversity and job performance. As the studies included in the analysis consist of controlled before-after studies, among others, the trustworthiness score is 90%. This means that there is only a 10% chance that the found effect is due to other factors.
References
van Dijk, H., van Engen, M. L., & van Knippenberg, D. (2012). Defying conventional wisdom: A meta-analytical examination of the differences between demographic and job related diversity relationships with performance. Organizational Behavior and Human Decision Processes, 119(1), 38–53. ScienceDirect. https://doi.org/10.1016/j.obhdp.2012.06.003
You can find the original article here!
