At the 14th Africa Evidence Summit held in Addis Ababa in July 2026, Isaac Obara of the Busara Center presented an experimental study entitled “Understanding Mental Health in the Workplace: An Experimental Investigation.” The presentation examined the relationship between mental health and workplace performance among young, educated, low-income jobseekers in Nairobi, Kenya. The study sought to understand not only whether mental health affects productivity and employment-related outcomes, but also the mechanisms through which psychological well-being influences workplace behaviour and decision-making.
Obara introduced the multidisciplinary research team involved in the study. The team brought together researchers and practitioners from the Busara Center and leading international universities, including the University of California, Berkeley; the University of California, Los Angeles; the University of Oxford; the University of Warwick; and other institutions.
The research team included Richard LeBeau, Director of the UCLA Psychology Clinic; Kate Orkin, Associate Professor at the Blavatnik School of Government at the University of Oxford; Ulrike Malmendier, Associate Professor at the University of California, Berkeley; Matthew Ridley, Professor of Economics at the University of Warwick; and Anasuya Narasimhan, Assistant Professor of Economics at the University of California, Berkeley. The Busara research team included Salome Njambi, Richa Mukerjee, Isaac Obara, Bihai Gao, Raphael Malenya, and other research assistants and scholars.
The multidisciplinary composition of the team reflected the study's effort to connect mental-health research, psychology, economics, labour-market behaviour, and experimental methods.
Obara explained that mental health was an increasingly important social and economic issue in Kenya, particularly among young people. He reported that more than 40 percent of participants in the study's pilot phase exhibited clinically significant symptoms of anxiety or depression. He also referred to national evidence indicating that approximately one in eight Kenyan adolescents experiences a mental-health condition each year.
The presenter further noted that mental disorders represented the second-leading cause of lost healthy life among Kenyan youth, demonstrating that mental health had consequences extending beyond individual well-being. Poor mental health could affect education, employment, productivity, interpersonal relationships, and broader economic participation.
He emphasized that mental health was closely connected to poverty and employment. Poverty could worsen anxiety and depression by exposing individuals to financial insecurity, unemployment, limited opportunities, and other forms of economic stress. At the same time, poor mental health could make it more difficult for individuals to secure and maintain employment or perform effectively at work. This created a reinforcing cycle in which poverty could contribute to poor mental health, while poor mental health could in turn limit economic opportunities.
The presenter referred to evidence from Kenya suggesting that cash-transfer programmes could reduce symptoms of depression, while treatment for mental illness could improve people's ability to work. However, he stressed that there remained a substantial treatment gap, with many people experiencing mental-health conditions still unable to access appropriate psychological or clinical care. This gap highlighted the need to identify scalable and practical interventions that could improve mental health while also addressing economic and employment-related challenges.
The study was motivated by the recognition that mental health could have important consequences for labour-market behaviour, but that the existing evidence base remained incomplete. Obara explained that relatively little research had examined how mental health affected productivity across different types of work tasks, particularly tasks involving social interaction, analytical thinking, and workplace decision-making.
The researchers were particularly interested in determining whether individuals experiencing anxiety or depression behaved differently when confronted with different workplace tasks or incentives. They wanted to understand whether mental health affected not only the amount of effort people exerted, but also their ability to communicate, cooperate, make decisions, demonstrate confidence, and respond to workplace situations.
The study also sought to contribute to the literature on mental-health interventions by developing and testing a work-focused cognitive behavioural therapy (WCBT) intervention. According to Obara, the intervention was specifically designed to address psychological challenges associated with work and job searching among economically active populations.
He explained that conventional cognitive behavioural therapy generally addresses mental-health symptoms, but the intervention in this study was tailored to the specific challenges faced by individuals seeking employment and participating in the labour market. These challenges included workplace confidence, job-search difficulties, interpersonal interactions, and responses to employment-related stress.
The researchers believed that this represented an important innovation because they were not aware of previous interventions in low- and middle-income countries that had specifically tailored cognitive behavioural therapy content to workplace and job-search problems.
Obara explained that the research was expected to make several contributions to both academic knowledge and public policy.
First, the study was expected to provide early evidence on the economic consequences of psychological interventions. By measuring both mental-health outcomes and workplace performance, the researchers hoped to establish whether improving psychological well-being could produce measurable economic benefits.
Second, the research was expected to advance understanding of the relationship between mental health and labour-market behaviour. Rather than treating mental health exclusively as a health issue, the study considered it an important factor influencing productivity, workplace behaviour, job search, and employment outcomes.
Third, the findings could inform the design of employment programmes, youth empowerment initiatives, and scalable mental-health interventions. If the intervention proved effective, it could potentially provide a model for integrating mental-health support into programmes targeting unemployed and economically vulnerable young people.
The presenter explained that the study was organized around three major questions.
The first question was whether work-focused cognitive behavioural therapy affected workplace outcomes. The researchers were particularly interested in productivity, assertiveness, attendance, and other indicators of workplace performance.
The second question concerned the mechanism of impact. If WCBT improved workplace outcomes, the researchers wanted to determine whether this occurred because the intervention reduced symptoms of depression, anxiety, and social anxiety.
The third question addressed the broader relationship between mental health and workplace behaviour. The researchers wanted to determine whether mental health could predict how individuals responded to different tasks and incentives and whether the experience of participating in work could itself influence mental health.
Obara explained that the research programme would eventually examine the longer-term effects of WCBT on earnings and employment, although these long-run outcomes were not the primary focus of the presentation.
The study was conducted among young, educated, low-income jobseekers in Nairobi. Participants were recruited from informal settlements in and around the city, including Kibera and Kawangware.
The researchers established several eligibility criteria. Participants had to be between 18 and 35 years of age, have completed at least high-school or vocational education, and be available to work full-time for six weeks. They also needed to report at least one symptom of depression or anxiety, although the symptom did not necessarily have to meet the threshold for a clinical diagnosis.
Obara noted that more than 90 percent of participants had at least one mental-health symptom, illustrating the significant psychological challenges experienced by the study population.
The selection of young, educated, low-income jobseekers was deliberate because this group faced the intersection of several important challenges: limited employment opportunities, financial constraints, psychological stress, and the need to transition successfully into the labour market.
The presenter described the study as an experimental investigation implemented through a structured multi-phase process.
Approximately two weeks before each new study phase, the research team conducted recruitment and screening. The team used a randomized list generated through field recruitment to identify and invite eligible participants for a particular study phase.
On Day 1, participants arrived in groups of approximately 30–50 people and completed a baseline survey. Following the baseline assessment, approximately 60 participants were selected to proceed to the workplace-task component of the study.
From Day 2 through Day 30, participants took part in a structured work-task environment lasting approximately 29 days. The programme incorporated individual and social tasks, work-related activities, and intervention workshops.
The intervention was conducted during the first four weeks, while Weeks 5 and 6 focused primarily on measuring outcomes. The researchers therefore had an opportunity to observe participants both during the intervention and after the intervention period.
On Days 29 and 30, participants took part in midline activities. These included work tasks and a midline survey. Individuals who had not initially been selected for the work-task component were also invited back to participate in the relevant survey activities.
The study subsequently included endline surveys three months and 12 months after the intervention workshops. These follow-up assessments were designed to determine whether any effects observed immediately after the intervention persisted over time.
Obara explained that the research relied extensively on digital systems to manage recruitment, data collection, experimental tasks, and participant payments.
The research system was used to manage participant recruitment databases and mobile-money payments. During recruitment drives, the team also collected GPS information and audio recordings, allowing the researchers to document field recruitment activities and maintain accurate records.
The digital platform hosted the baseline, midline, and endline surveys, as well as surveys associated with the experimental work tasks. This allowed the research team to maintain a standardized data-collection process throughout the study.
The platform also supported laboratory-based work tasks involving both individual and team activities. In addition, workshop sessions and mock interviews were recorded to facilitate quality control and assess implementation consistency.
The presenter emphasized that the integration of digital research tools made it possible to manage a complex longitudinal experiment involving repeated measurements of psychological, behavioural, and workplace outcomes.
A major component of the study involved having participants work for approximately 30 days in a computer laboratory. The purpose was to create a controlled environment in which researchers could observe workplace behaviour and objectively measure different dimensions of performance.
Obara explained that the work-task component incorporated multiple types of activities designed to measure different dimensions of productivity and workplace behaviour.
One of the tasks involved data entry, which was described as a routine effort task. Participants were required to perform data-entry activities using a computer interface, allowing researchers to measure indicators such as effort, persistence, accuracy, and productivity.
The broader task environment also included individual and social activities. These allowed the researchers to examine how participants responded to different workplace situations and whether mental-health characteristics influenced performance differently depending on whether tasks required routine effort, analytical reasoning, social interaction, or decision-making.
The presenter emphasized that a central innovation of the study was its effort to examine mental health through an economic and workplace lens.
Rather than simply asking whether individuals were experiencing anxiety or depression, the researchers sought to determine how these psychological conditions translated into observable workplace behaviour.
For example, an individual experiencing anxiety might perform differently on a routine data-entry task than on a task requiring interpersonal interaction. Similarly, someone experiencing social anxiety might respond differently to teamwork, negotiation, interviews, or communication-based activities than to individual work.
The study therefore sought to identify whether mental health affected specific dimensions of productivity rather than assuming that psychological distress had the same effect on all forms of work.
The WCBT intervention was designed around the idea that psychological treatment could be adapted to address the practical challenges encountered by people who were actively seeking employment or participating in work.
The presenter explained that the intervention incorporated cognitive behavioural approaches but tailored them to employment-related concerns. The objective was not only to reduce psychological symptoms but also to help participants develop ways of managing workplace stress, job-search challenges, social interactions, confidence, and other barriers to effective labour-market participation.
The researchers intended to determine whether such an intervention could simultaneously improve mental health and workplace performance.
Although the presentation concentrated on experimental workplace outcomes, Obara explained that the research formed part of a broader programme examining the long-term economic consequences of mental-health interventions.
The researchers planned to examine whether improvements generated by WCBT could eventually translate into higher employment rates and increased earnings. The three-month and 12-month follow-up surveys were therefore important for determining whether any improvements in mental health or workplace behaviour persisted beyond the immediate intervention period.
The longer-term objective was to establish whether mental-health interventions could generate economic returns by improving individuals' ability to obtain employment, remain employed, and perform productively in the workplace.
The presentation suggested that the study could have important implications for how governments and development organizations approach both mental health and employment programming in Kenya and other African countries.
If the intervention demonstrated that improving mental health increased productivity, assertiveness, attendance, or other workplace outcomes, employment programmes could potentially incorporate mental-health screening and psychological support as part of their service packages.
Similarly, youth empowerment programmes could benefit from integrating mental-health interventions into skills development and job-placement initiatives. Rather than treating psychological well-being as separate from economic empowerment, programmes could address the two challenges simultaneously.
The research could also contribute to the development of scalable mental-health interventions for low-income populations, particularly where access to conventional clinical services remained limited.
In concluding the presentation, Isaac Obara emphasized that mental health should be understood not only as a health concern but also as an important determinant of economic participation, productivity, workplace behaviour, and employment outcomes.
The experimental investigation in Nairobi sought to provide rigorous evidence on how anxiety and depression affected workplace performance and whether a work-focused cognitive behavioural therapy intervention could improve both psychological well-being and labour-market-related outcomes.
The study's distinctive contribution lay in bringing together mental-health research, behavioural economics, workplace experimentation, and employment policy. By observing participants performing different individual and social tasks in a controlled work environment and following them over several months, the researchers aimed to identify the mechanisms connecting mental health with workplace performance.
Ultimately, the presentation highlighted the possibility that addressing mental-health challenges among economically vulnerable young people could generate benefits extending beyond improved psychological well-being. Effective mental-health interventions could potentially improve productivity, confidence, workplace participation, employment prospects, and ultimately earnings, making mental health an important component of broader strategies for poverty reduction and youth economic empowerment in Kenya and other African countries.
Continuing the presentation, Isaac Obara described the additional workplace tasks that were incorporated into the experimental study to capture different dimensions of productivity, social interaction, analytical ability, teamwork, leadership, and workplace confidence. He explained that the researchers deliberately designed a diverse set of tasks because mental health might influence different types of workplace behaviour in different ways. Rather than relying on a single measure of productivity, the study therefore exposed participants to routine, social, analytical, managerial, and communication-based activities.
One of the principal tasks was a customer-service exercise involving both social interaction and analytical skills. Obara explained that participants completed an AI-powered simulated customer-engagement exercise designed to reproduce some of the challenges encountered in a real customer-service workplace.
Participants were compensated according to the quality of the service they provided during each customer conversation. Their payment rate depended on how well they performed in dealing with the simulated customer. The quality of their performance was assessed according to several dimensions, particularly accuracy, helpfulness, and politeness.
The task therefore enabled the researchers to measure not only whether participants could complete a particular task correctly but also how effectively they could communicate with another person while solving a problem. The use of an AI-powered simulated customer environment also allowed the researchers to standardize interactions across participants and expose them to different customer scenarios.
Obara explained that this task was especially relevant to the study because it combined cognitive performance with interpersonal behaviour. It could therefore help the researchers examine whether anxiety, depression, social anxiety, or psychological confidence influenced participants' ability to perform jobs requiring communication with customers.
In addition to the two principal workplace tasks—data entry and customer service—the researchers introduced several variations and additional exercises. These activities were intended to capture a wider range of workplace capabilities and behavioural outcomes.
One variation involved cooperative data entry, in which participants sometimes completed the data-entry activity collectively rather than individually. This allowed the researchers to examine whether participants performed differently when productivity depended partly on cooperation with others.
Another activity was a promotion interview. In this exercise, participants had their customer-service skills assessed directly by an enumerator who acted as a customer. The enumerator evaluated the participant's performance, and the best-performing participants were offered the opportunity to be "promoted" to a more advanced AI-powered customer-service role. In this role, participants were required to deal with more difficult customers but received higher rewards for successful performance.
Obara explained that the promotion exercise allowed the researchers to examine ambition, confidence, performance under evaluation, and willingness to take on more demanding responsibilities. It also provided a realistic simulation of an important workplace process in which employees are assessed before being given greater responsibility.
The study also included a presentation or public-speaking task. Participants were required to speak publicly, and their performance was evaluated by an audience consisting of other participants and members of the Busara research staff.
The task was designed to measure communication and presentation ability as well as confidence in public settings. The researchers were particularly interested in whether the intervention affected participants' willingness and ability to express themselves in front of others.
This task was particularly relevant to the study's investigation of assertiveness and social confidence, because individuals experiencing anxiety or low confidence might be less willing to speak publicly or put themselves forward in professional environments.
Another important component was a management task based on the managerial exercise developed by Weidmann and colleagues in 2024. Participants were organized into teams of three, with one participant assigned the role of manager.
The manager was responsible for allocating team members to different work activities. Managers could monitor the progress of workers in real time and reallocate individuals between tasks according to different strategies.
Obara explained that this exercise allowed researchers to observe leadership, decision-making, delegation, monitoring, and managerial behaviour. It also provided an opportunity to examine whether mental-health characteristics influenced individuals' willingness to take responsibility and their effectiveness when placed in positions of authority.
The task was important because many workplace outcomes depend not only on an individual's ability to complete assigned tasks but also on their capacity to organize other people, make decisions, respond to changing circumstances, and manage team performance.
The researchers also conducted a teamwork exercise in which participants were placed into groups and asked to solve a "tower-building" challenge. Using ordinary office supplies, each team had to construct the tallest possible tower within a specified time limit.
The team that constructed the tallest tower received a monetary reward. Obara explained that the exercise was intended to provide a controlled measure of teamwork, cooperation, problem-solving, communication, leadership, and collective performance.
Because participants had to work together under a time constraint and with a common financial incentive, the exercise created an environment in which interpersonal relationships and individual behaviour could affect group outcomes. It therefore allowed researchers to examine whether psychological characteristics translated into differences in collaborative workplace behaviour.
A further activity was the "speaking up" exercise, which was designed to measure participants' willingness to express ideas and make suggestions for organizational improvement.
Participants took part in group meetings during which they were encouraged and financially incentivized to suggest practical improvements that could make the laboratory more productive. Participants received rewards when they proposed ideas that could actually be implemented.
Enumerators recorded the number of times each participant contributed to the discussion. Obara explained that this provided a behavioural measure of voice, assertiveness, confidence, participation, and willingness to communicate ideas to an organization.
The exercise was particularly important because workplace success often depends on employees being willing to raise concerns, propose improvements, communicate with supervisors, and advocate for their own ideas. Such behaviours can be difficult for people experiencing social anxiety or low psychological confidence.
A central component of the experiment was the Work-Based Cognitive Behavioural Therapy (WCBT) intervention. Obara explained that the intervention had been developed and adapted for the Kenyan context by the study's U.S.-based psychology co-principal investigators, drawing on established cognitive behavioural therapy techniques and adapting them to challenges associated specifically with employment and job searching.
The intervention applied core principles of cognitive behavioural therapy to problems that economically active individuals could encounter when seeking or performing work. Rather than focusing exclusively on clinical symptoms, the programme sought to help participants understand and change the ways in which they thought about and responded to workplace-related challenges.
The first was cognitive restructuring, through which participants learned to identify and reconsider unhelpful ways of thinking and develop alternative ways of interpreting and responding to difficult situations.
The second was behavioural practice or exposure therapy. Participants were given opportunities to practice situations that they might normally avoid. Through role-play and repeated practice, they could gradually become more comfortable with situations such as speaking publicly, negotiating, interacting with others, or putting themselves forward in professional settings.
Obara explained that the WCBT intervention consisted of eight two-hour group workshops delivered over four weeks, with two sessions conducted each week.
The researchers also established a placebo or comparison intervention consisting of workshops focused on environmental sustainability. The placebo workshops were designed to have the same group size and approximately the same duration as the WCBT sessions. This enabled the researchers to distinguish the effects of the psychological intervention from the effects of simply participating in group workshops.
The design therefore allowed the research team to determine whether changes in participants' behaviour resulted specifically from WCBT rather than from spending time in a structured group environment or interacting with other participants.
The presenter explained that the WCBT workshops were designed to create an interactive and supportive environment. The sessions emphasized practice rather than perfection, encouraging participants to experiment with new ways of responding to challenging situations.
Another important element was gradual exposure to situations that participants might normally avoid. Instead of forcing participants immediately into highly uncomfortable situations, the intervention allowed them to develop confidence progressively through repeated practice.
The programme also respected participants' personal boundaries. Obara emphasized that there was no requirement for participants to disclose personal experiences beyond what they felt comfortable sharing.
The overall objective was to move participants from simply understanding their psychological experiences toward actively changing how they responded to workplace and job-search challenges.
At the time of the presentation, Obara reported that the study was still ongoing. Recruitment was being conducted on a phase-by-phase basis, while baseline, midline, and endline data collection was continuing.
The laboratory-based workplace tasks, particularly data entry and AI-powered customer service, were also being implemented. At the same time, the WCBT intervention was being rolled out according to the schedule established for each study phase.
Because the study was still in progress, the presenter emphasized that the results being discussed were interim findings rather than final conclusions.
Obara then presented preliminary findings based on approximately 55 percent of the expected final sample. He cautioned that data collection was still underway and that the results could change once the complete dataset became available.
The preliminary evidence suggested that WCBT had a meaningful effect on participants' presentation and assertiveness. Participants who received the intervention appeared to become better at presenting themselves and were more willing to speak, negotiate, and put themselves forward generally.
According to the presenter, treated participants appeared more willing to engage in activities that required them to demonstrate themselves publicly or advocate for themselves. This finding was particularly relevant to job seekers because successful job search often requires individuals to communicate confidently during interviews, negotiations, presentations, and professional interactions.
However, the preliminary results did not show a large improvement in overall mental-health measures. Obara explained that WCBT did not appear to substantially improve measures of depression and anxiety in the interim data.
Another notable finding was that the intervention's behavioural benefits did not appear to be restricted to participants who began the study with poor mental health. Participants with relatively good baseline mental health appeared to benefit from the intervention to a similar degree.
This finding suggested that WCBT might have effects that extend beyond the treatment of mental illness. Rather than functioning only as a clinical intervention for individuals experiencing significant psychological distress, the intervention might also help people develop general workplace skills, confidence, assertiveness, and professional behaviour.
The presenter then turned to a second area of preliminary findings: the broader relationship between mental health and workplace participation.
Based on approximately 55 percent of the data, the researchers found suggestive evidence that participating in the work environment itself could have a positive effect on mental health. Participants who were offered the opportunity to participate in the work activities appeared to experience lower levels of depression and anxiety than a comparison group that was not offered work.
However, the researchers also found that depression and anxiety were associated with higher rates of dropout and non-attendance from the work programme.
Obara explained that this created a potentially important paradox. Work could improve mental health, but poor mental health could simultaneously make it more difficult for individuals to participate consistently in work.
The presenter described this pattern as suggestive of a possible "trap" in which poor mental health and employment difficulties reinforce one another.
On one side of the relationship, participation in decent and structured work appeared capable of improving psychological well-being. Work could provide income, social interaction, structure, purpose, confidence, and a sense of achievement, all of which could contribute positively to mental health.
On the other side, individuals experiencing depression or anxiety could find it more difficult to attend work consistently, maintain concentration, interact with colleagues or customers, and meet workplace expectations. Consequently, poor mental health could reduce their ability to benefit from the very work environment that might otherwise improve their psychological well-being.
The presenter stressed that this interpretation remained speculative and should not yet be treated as a definitive causal conclusion, because the relevant analysis was still based on interim data and some of the evidence was non-randomized.
Nevertheless, the finding raised an important policy question: if work improves mental health but poor mental health makes it harder for people to remain in work, then employment programmes may need to address mental-health barriers at the point of entry rather than waiting until individuals have successfully entered the labour market.
The emerging evidence presented by Obara suggested that the relationship between mental health and employment was more complex than a simple relationship in which psychological distress reduces productivity.
The study's wide range of experimental workplace tasks—including data entry, AI-powered customer service, promotion interviews, public presentations, management, teamwork, and speaking-up exercises—was designed to identify the specific workplace behaviours through which mental health matters.
The preliminary WCBT results suggested that psychological interventions could improve assertiveness, presentation, willingness to negotiate, and the ability to put oneself forward, even when substantial improvements in clinical mental-health symptoms were not immediately observed.
At the same time, the preliminary evidence suggested that participating in work itself could improve mental health, while depression and anxiety could make sustained participation more difficult. This reciprocal relationship highlighted the possibility of a self-reinforcing cycle between employment and psychological well-being.
The presentation therefore underscored the importance of integrating mental-health support into employment and youth empowerment programmes, particularly for young people from economically disadvantaged backgrounds. It also demonstrated the potential of combining AI-powered workplace simulations, behavioural experiments, digital data collection, and psychologically informed interventions to generate more precise evidence about how mental health affects economic participation.
Obara concluded that, while the study was still ongoing and the interim results required further validation with the complete dataset, the emerging evidence pointed toward an important policy lesson: improving employment outcomes may require addressing psychological barriers alongside conventional employment and skills interventions, while employment itself may be an important pathway for improving mental well-being.
Obara introduced the multidisciplinary research team involved in the study. The team brought together researchers and practitioners from the Busara Center and leading international universities, including the University of California, Berkeley; the University of California, Los Angeles; the University of Oxford; the University of Warwick; and other institutions.
The research team included Richard LeBeau, Director of the UCLA Psychology Clinic; Kate Orkin, Associate Professor at the Blavatnik School of Government at the University of Oxford; Ulrike Malmendier, Associate Professor at the University of California, Berkeley; Matthew Ridley, Professor of Economics at the University of Warwick; and Anasuya Narasimhan, Assistant Professor of Economics at the University of California, Berkeley. The Busara research team included Salome Njambi, Richa Mukerjee, Isaac Obara, Bihai Gao, Raphael Malenya, and other research assistants and scholars.
The multidisciplinary composition of the team reflected the study's effort to connect mental-health research, psychology, economics, labour-market behaviour, and experimental methods.
Obara explained that mental health was an increasingly important social and economic issue in Kenya, particularly among young people. He reported that more than 40 percent of participants in the study's pilot phase exhibited clinically significant symptoms of anxiety or depression. He also referred to national evidence indicating that approximately one in eight Kenyan adolescents experiences a mental-health condition each year.
The presenter further noted that mental disorders represented the second-leading cause of lost healthy life among Kenyan youth, demonstrating that mental health had consequences extending beyond individual well-being. Poor mental health could affect education, employment, productivity, interpersonal relationships, and broader economic participation.
He emphasized that mental health was closely connected to poverty and employment. Poverty could worsen anxiety and depression by exposing individuals to financial insecurity, unemployment, limited opportunities, and other forms of economic stress. At the same time, poor mental health could make it more difficult for individuals to secure and maintain employment or perform effectively at work. This created a reinforcing cycle in which poverty could contribute to poor mental health, while poor mental health could in turn limit economic opportunities.
The presenter referred to evidence from Kenya suggesting that cash-transfer programmes could reduce symptoms of depression, while treatment for mental illness could improve people's ability to work. However, he stressed that there remained a substantial treatment gap, with many people experiencing mental-health conditions still unable to access appropriate psychological or clinical care. This gap highlighted the need to identify scalable and practical interventions that could improve mental health while also addressing economic and employment-related challenges.
The study was motivated by the recognition that mental health could have important consequences for labour-market behaviour, but that the existing evidence base remained incomplete. Obara explained that relatively little research had examined how mental health affected productivity across different types of work tasks, particularly tasks involving social interaction, analytical thinking, and workplace decision-making.
The researchers were particularly interested in determining whether individuals experiencing anxiety or depression behaved differently when confronted with different workplace tasks or incentives. They wanted to understand whether mental health affected not only the amount of effort people exerted, but also their ability to communicate, cooperate, make decisions, demonstrate confidence, and respond to workplace situations.
The study also sought to contribute to the literature on mental-health interventions by developing and testing a work-focused cognitive behavioural therapy (WCBT) intervention. According to Obara, the intervention was specifically designed to address psychological challenges associated with work and job searching among economically active populations.
He explained that conventional cognitive behavioural therapy generally addresses mental-health symptoms, but the intervention in this study was tailored to the specific challenges faced by individuals seeking employment and participating in the labour market. These challenges included workplace confidence, job-search difficulties, interpersonal interactions, and responses to employment-related stress.
The researchers believed that this represented an important innovation because they were not aware of previous interventions in low- and middle-income countries that had specifically tailored cognitive behavioural therapy content to workplace and job-search problems.
Obara explained that the research was expected to make several contributions to both academic knowledge and public policy.
First, the study was expected to provide early evidence on the economic consequences of psychological interventions. By measuring both mental-health outcomes and workplace performance, the researchers hoped to establish whether improving psychological well-being could produce measurable economic benefits.
Second, the research was expected to advance understanding of the relationship between mental health and labour-market behaviour. Rather than treating mental health exclusively as a health issue, the study considered it an important factor influencing productivity, workplace behaviour, job search, and employment outcomes.
Third, the findings could inform the design of employment programmes, youth empowerment initiatives, and scalable mental-health interventions. If the intervention proved effective, it could potentially provide a model for integrating mental-health support into programmes targeting unemployed and economically vulnerable young people.
The presenter explained that the study was organized around three major questions.
The first question was whether work-focused cognitive behavioural therapy affected workplace outcomes. The researchers were particularly interested in productivity, assertiveness, attendance, and other indicators of workplace performance.
The second question concerned the mechanism of impact. If WCBT improved workplace outcomes, the researchers wanted to determine whether this occurred because the intervention reduced symptoms of depression, anxiety, and social anxiety.
The third question addressed the broader relationship between mental health and workplace behaviour. The researchers wanted to determine whether mental health could predict how individuals responded to different tasks and incentives and whether the experience of participating in work could itself influence mental health.
Obara explained that the research programme would eventually examine the longer-term effects of WCBT on earnings and employment, although these long-run outcomes were not the primary focus of the presentation.
The study was conducted among young, educated, low-income jobseekers in Nairobi. Participants were recruited from informal settlements in and around the city, including Kibera and Kawangware.
The researchers established several eligibility criteria. Participants had to be between 18 and 35 years of age, have completed at least high-school or vocational education, and be available to work full-time for six weeks. They also needed to report at least one symptom of depression or anxiety, although the symptom did not necessarily have to meet the threshold for a clinical diagnosis.
Obara noted that more than 90 percent of participants had at least one mental-health symptom, illustrating the significant psychological challenges experienced by the study population.
The selection of young, educated, low-income jobseekers was deliberate because this group faced the intersection of several important challenges: limited employment opportunities, financial constraints, psychological stress, and the need to transition successfully into the labour market.
The presenter described the study as an experimental investigation implemented through a structured multi-phase process.
Approximately two weeks before each new study phase, the research team conducted recruitment and screening. The team used a randomized list generated through field recruitment to identify and invite eligible participants for a particular study phase.
On Day 1, participants arrived in groups of approximately 30–50 people and completed a baseline survey. Following the baseline assessment, approximately 60 participants were selected to proceed to the workplace-task component of the study.
From Day 2 through Day 30, participants took part in a structured work-task environment lasting approximately 29 days. The programme incorporated individual and social tasks, work-related activities, and intervention workshops.
The intervention was conducted during the first four weeks, while Weeks 5 and 6 focused primarily on measuring outcomes. The researchers therefore had an opportunity to observe participants both during the intervention and after the intervention period.
On Days 29 and 30, participants took part in midline activities. These included work tasks and a midline survey. Individuals who had not initially been selected for the work-task component were also invited back to participate in the relevant survey activities.
The study subsequently included endline surveys three months and 12 months after the intervention workshops. These follow-up assessments were designed to determine whether any effects observed immediately after the intervention persisted over time.
Obara explained that the research relied extensively on digital systems to manage recruitment, data collection, experimental tasks, and participant payments.
The research system was used to manage participant recruitment databases and mobile-money payments. During recruitment drives, the team also collected GPS information and audio recordings, allowing the researchers to document field recruitment activities and maintain accurate records.
The digital platform hosted the baseline, midline, and endline surveys, as well as surveys associated with the experimental work tasks. This allowed the research team to maintain a standardized data-collection process throughout the study.
The platform also supported laboratory-based work tasks involving both individual and team activities. In addition, workshop sessions and mock interviews were recorded to facilitate quality control and assess implementation consistency.
The presenter emphasized that the integration of digital research tools made it possible to manage a complex longitudinal experiment involving repeated measurements of psychological, behavioural, and workplace outcomes.
A major component of the study involved having participants work for approximately 30 days in a computer laboratory. The purpose was to create a controlled environment in which researchers could observe workplace behaviour and objectively measure different dimensions of performance.
Obara explained that the work-task component incorporated multiple types of activities designed to measure different dimensions of productivity and workplace behaviour.
One of the tasks involved data entry, which was described as a routine effort task. Participants were required to perform data-entry activities using a computer interface, allowing researchers to measure indicators such as effort, persistence, accuracy, and productivity.
The broader task environment also included individual and social activities. These allowed the researchers to examine how participants responded to different workplace situations and whether mental-health characteristics influenced performance differently depending on whether tasks required routine effort, analytical reasoning, social interaction, or decision-making.
The presenter emphasized that a central innovation of the study was its effort to examine mental health through an economic and workplace lens.
Rather than simply asking whether individuals were experiencing anxiety or depression, the researchers sought to determine how these psychological conditions translated into observable workplace behaviour.
For example, an individual experiencing anxiety might perform differently on a routine data-entry task than on a task requiring interpersonal interaction. Similarly, someone experiencing social anxiety might respond differently to teamwork, negotiation, interviews, or communication-based activities than to individual work.
The study therefore sought to identify whether mental health affected specific dimensions of productivity rather than assuming that psychological distress had the same effect on all forms of work.
The WCBT intervention was designed around the idea that psychological treatment could be adapted to address the practical challenges encountered by people who were actively seeking employment or participating in work.
The presenter explained that the intervention incorporated cognitive behavioural approaches but tailored them to employment-related concerns. The objective was not only to reduce psychological symptoms but also to help participants develop ways of managing workplace stress, job-search challenges, social interactions, confidence, and other barriers to effective labour-market participation.
The researchers intended to determine whether such an intervention could simultaneously improve mental health and workplace performance.
Although the presentation concentrated on experimental workplace outcomes, Obara explained that the research formed part of a broader programme examining the long-term economic consequences of mental-health interventions.
The researchers planned to examine whether improvements generated by WCBT could eventually translate into higher employment rates and increased earnings. The three-month and 12-month follow-up surveys were therefore important for determining whether any improvements in mental health or workplace behaviour persisted beyond the immediate intervention period.
The longer-term objective was to establish whether mental-health interventions could generate economic returns by improving individuals' ability to obtain employment, remain employed, and perform productively in the workplace.
The presentation suggested that the study could have important implications for how governments and development organizations approach both mental health and employment programming in Kenya and other African countries.
If the intervention demonstrated that improving mental health increased productivity, assertiveness, attendance, or other workplace outcomes, employment programmes could potentially incorporate mental-health screening and psychological support as part of their service packages.
Similarly, youth empowerment programmes could benefit from integrating mental-health interventions into skills development and job-placement initiatives. Rather than treating psychological well-being as separate from economic empowerment, programmes could address the two challenges simultaneously.
The research could also contribute to the development of scalable mental-health interventions for low-income populations, particularly where access to conventional clinical services remained limited.
In concluding the presentation, Isaac Obara emphasized that mental health should be understood not only as a health concern but also as an important determinant of economic participation, productivity, workplace behaviour, and employment outcomes.
The experimental investigation in Nairobi sought to provide rigorous evidence on how anxiety and depression affected workplace performance and whether a work-focused cognitive behavioural therapy intervention could improve both psychological well-being and labour-market-related outcomes.
The study's distinctive contribution lay in bringing together mental-health research, behavioural economics, workplace experimentation, and employment policy. By observing participants performing different individual and social tasks in a controlled work environment and following them over several months, the researchers aimed to identify the mechanisms connecting mental health with workplace performance.
Ultimately, the presentation highlighted the possibility that addressing mental-health challenges among economically vulnerable young people could generate benefits extending beyond improved psychological well-being. Effective mental-health interventions could potentially improve productivity, confidence, workplace participation, employment prospects, and ultimately earnings, making mental health an important component of broader strategies for poverty reduction and youth economic empowerment in Kenya and other African countries.
Continuing the presentation, Isaac Obara described the additional workplace tasks that were incorporated into the experimental study to capture different dimensions of productivity, social interaction, analytical ability, teamwork, leadership, and workplace confidence. He explained that the researchers deliberately designed a diverse set of tasks because mental health might influence different types of workplace behaviour in different ways. Rather than relying on a single measure of productivity, the study therefore exposed participants to routine, social, analytical, managerial, and communication-based activities.
One of the principal tasks was a customer-service exercise involving both social interaction and analytical skills. Obara explained that participants completed an AI-powered simulated customer-engagement exercise designed to reproduce some of the challenges encountered in a real customer-service workplace.
Participants were compensated according to the quality of the service they provided during each customer conversation. Their payment rate depended on how well they performed in dealing with the simulated customer. The quality of their performance was assessed according to several dimensions, particularly accuracy, helpfulness, and politeness.
The task therefore enabled the researchers to measure not only whether participants could complete a particular task correctly but also how effectively they could communicate with another person while solving a problem. The use of an AI-powered simulated customer environment also allowed the researchers to standardize interactions across participants and expose them to different customer scenarios.
Obara explained that this task was especially relevant to the study because it combined cognitive performance with interpersonal behaviour. It could therefore help the researchers examine whether anxiety, depression, social anxiety, or psychological confidence influenced participants' ability to perform jobs requiring communication with customers.
In addition to the two principal workplace tasks—data entry and customer service—the researchers introduced several variations and additional exercises. These activities were intended to capture a wider range of workplace capabilities and behavioural outcomes.
One variation involved cooperative data entry, in which participants sometimes completed the data-entry activity collectively rather than individually. This allowed the researchers to examine whether participants performed differently when productivity depended partly on cooperation with others.
Another activity was a promotion interview. In this exercise, participants had their customer-service skills assessed directly by an enumerator who acted as a customer. The enumerator evaluated the participant's performance, and the best-performing participants were offered the opportunity to be "promoted" to a more advanced AI-powered customer-service role. In this role, participants were required to deal with more difficult customers but received higher rewards for successful performance.
Obara explained that the promotion exercise allowed the researchers to examine ambition, confidence, performance under evaluation, and willingness to take on more demanding responsibilities. It also provided a realistic simulation of an important workplace process in which employees are assessed before being given greater responsibility.
The study also included a presentation or public-speaking task. Participants were required to speak publicly, and their performance was evaluated by an audience consisting of other participants and members of the Busara research staff.
The task was designed to measure communication and presentation ability as well as confidence in public settings. The researchers were particularly interested in whether the intervention affected participants' willingness and ability to express themselves in front of others.
This task was particularly relevant to the study's investigation of assertiveness and social confidence, because individuals experiencing anxiety or low confidence might be less willing to speak publicly or put themselves forward in professional environments.
Another important component was a management task based on the managerial exercise developed by Weidmann and colleagues in 2024. Participants were organized into teams of three, with one participant assigned the role of manager.
The manager was responsible for allocating team members to different work activities. Managers could monitor the progress of workers in real time and reallocate individuals between tasks according to different strategies.
Obara explained that this exercise allowed researchers to observe leadership, decision-making, delegation, monitoring, and managerial behaviour. It also provided an opportunity to examine whether mental-health characteristics influenced individuals' willingness to take responsibility and their effectiveness when placed in positions of authority.
The task was important because many workplace outcomes depend not only on an individual's ability to complete assigned tasks but also on their capacity to organize other people, make decisions, respond to changing circumstances, and manage team performance.
The researchers also conducted a teamwork exercise in which participants were placed into groups and asked to solve a "tower-building" challenge. Using ordinary office supplies, each team had to construct the tallest possible tower within a specified time limit.
The team that constructed the tallest tower received a monetary reward. Obara explained that the exercise was intended to provide a controlled measure of teamwork, cooperation, problem-solving, communication, leadership, and collective performance.
Because participants had to work together under a time constraint and with a common financial incentive, the exercise created an environment in which interpersonal relationships and individual behaviour could affect group outcomes. It therefore allowed researchers to examine whether psychological characteristics translated into differences in collaborative workplace behaviour.
A further activity was the "speaking up" exercise, which was designed to measure participants' willingness to express ideas and make suggestions for organizational improvement.
Participants took part in group meetings during which they were encouraged and financially incentivized to suggest practical improvements that could make the laboratory more productive. Participants received rewards when they proposed ideas that could actually be implemented.
Enumerators recorded the number of times each participant contributed to the discussion. Obara explained that this provided a behavioural measure of voice, assertiveness, confidence, participation, and willingness to communicate ideas to an organization.
The exercise was particularly important because workplace success often depends on employees being willing to raise concerns, propose improvements, communicate with supervisors, and advocate for their own ideas. Such behaviours can be difficult for people experiencing social anxiety or low psychological confidence.
A central component of the experiment was the Work-Based Cognitive Behavioural Therapy (WCBT) intervention. Obara explained that the intervention had been developed and adapted for the Kenyan context by the study's U.S.-based psychology co-principal investigators, drawing on established cognitive behavioural therapy techniques and adapting them to challenges associated specifically with employment and job searching.
The intervention applied core principles of cognitive behavioural therapy to problems that economically active individuals could encounter when seeking or performing work. Rather than focusing exclusively on clinical symptoms, the programme sought to help participants understand and change the ways in which they thought about and responded to workplace-related challenges.
The first was cognitive restructuring, through which participants learned to identify and reconsider unhelpful ways of thinking and develop alternative ways of interpreting and responding to difficult situations.
The second was behavioural practice or exposure therapy. Participants were given opportunities to practice situations that they might normally avoid. Through role-play and repeated practice, they could gradually become more comfortable with situations such as speaking publicly, negotiating, interacting with others, or putting themselves forward in professional settings.
Obara explained that the WCBT intervention consisted of eight two-hour group workshops delivered over four weeks, with two sessions conducted each week.
The researchers also established a placebo or comparison intervention consisting of workshops focused on environmental sustainability. The placebo workshops were designed to have the same group size and approximately the same duration as the WCBT sessions. This enabled the researchers to distinguish the effects of the psychological intervention from the effects of simply participating in group workshops.
The design therefore allowed the research team to determine whether changes in participants' behaviour resulted specifically from WCBT rather than from spending time in a structured group environment or interacting with other participants.
The presenter explained that the WCBT workshops were designed to create an interactive and supportive environment. The sessions emphasized practice rather than perfection, encouraging participants to experiment with new ways of responding to challenging situations.
Another important element was gradual exposure to situations that participants might normally avoid. Instead of forcing participants immediately into highly uncomfortable situations, the intervention allowed them to develop confidence progressively through repeated practice.
The programme also respected participants' personal boundaries. Obara emphasized that there was no requirement for participants to disclose personal experiences beyond what they felt comfortable sharing.
The overall objective was to move participants from simply understanding their psychological experiences toward actively changing how they responded to workplace and job-search challenges.
At the time of the presentation, Obara reported that the study was still ongoing. Recruitment was being conducted on a phase-by-phase basis, while baseline, midline, and endline data collection was continuing.
The laboratory-based workplace tasks, particularly data entry and AI-powered customer service, were also being implemented. At the same time, the WCBT intervention was being rolled out according to the schedule established for each study phase.
Because the study was still in progress, the presenter emphasized that the results being discussed were interim findings rather than final conclusions.
Obara then presented preliminary findings based on approximately 55 percent of the expected final sample. He cautioned that data collection was still underway and that the results could change once the complete dataset became available.
The preliminary evidence suggested that WCBT had a meaningful effect on participants' presentation and assertiveness. Participants who received the intervention appeared to become better at presenting themselves and were more willing to speak, negotiate, and put themselves forward generally.
According to the presenter, treated participants appeared more willing to engage in activities that required them to demonstrate themselves publicly or advocate for themselves. This finding was particularly relevant to job seekers because successful job search often requires individuals to communicate confidently during interviews, negotiations, presentations, and professional interactions.
However, the preliminary results did not show a large improvement in overall mental-health measures. Obara explained that WCBT did not appear to substantially improve measures of depression and anxiety in the interim data.
Another notable finding was that the intervention's behavioural benefits did not appear to be restricted to participants who began the study with poor mental health. Participants with relatively good baseline mental health appeared to benefit from the intervention to a similar degree.
This finding suggested that WCBT might have effects that extend beyond the treatment of mental illness. Rather than functioning only as a clinical intervention for individuals experiencing significant psychological distress, the intervention might also help people develop general workplace skills, confidence, assertiveness, and professional behaviour.
The presenter then turned to a second area of preliminary findings: the broader relationship between mental health and workplace participation.
Based on approximately 55 percent of the data, the researchers found suggestive evidence that participating in the work environment itself could have a positive effect on mental health. Participants who were offered the opportunity to participate in the work activities appeared to experience lower levels of depression and anxiety than a comparison group that was not offered work.
However, the researchers also found that depression and anxiety were associated with higher rates of dropout and non-attendance from the work programme.
Obara explained that this created a potentially important paradox. Work could improve mental health, but poor mental health could simultaneously make it more difficult for individuals to participate consistently in work.
The presenter described this pattern as suggestive of a possible "trap" in which poor mental health and employment difficulties reinforce one another.
On one side of the relationship, participation in decent and structured work appeared capable of improving psychological well-being. Work could provide income, social interaction, structure, purpose, confidence, and a sense of achievement, all of which could contribute positively to mental health.
On the other side, individuals experiencing depression or anxiety could find it more difficult to attend work consistently, maintain concentration, interact with colleagues or customers, and meet workplace expectations. Consequently, poor mental health could reduce their ability to benefit from the very work environment that might otherwise improve their psychological well-being.
The presenter stressed that this interpretation remained speculative and should not yet be treated as a definitive causal conclusion, because the relevant analysis was still based on interim data and some of the evidence was non-randomized.
Nevertheless, the finding raised an important policy question: if work improves mental health but poor mental health makes it harder for people to remain in work, then employment programmes may need to address mental-health barriers at the point of entry rather than waiting until individuals have successfully entered the labour market.
The emerging evidence presented by Obara suggested that the relationship between mental health and employment was more complex than a simple relationship in which psychological distress reduces productivity.
The study's wide range of experimental workplace tasks—including data entry, AI-powered customer service, promotion interviews, public presentations, management, teamwork, and speaking-up exercises—was designed to identify the specific workplace behaviours through which mental health matters.
The preliminary WCBT results suggested that psychological interventions could improve assertiveness, presentation, willingness to negotiate, and the ability to put oneself forward, even when substantial improvements in clinical mental-health symptoms were not immediately observed.
At the same time, the preliminary evidence suggested that participating in work itself could improve mental health, while depression and anxiety could make sustained participation more difficult. This reciprocal relationship highlighted the possibility of a self-reinforcing cycle between employment and psychological well-being.
The presentation therefore underscored the importance of integrating mental-health support into employment and youth empowerment programmes, particularly for young people from economically disadvantaged backgrounds. It also demonstrated the potential of combining AI-powered workplace simulations, behavioural experiments, digital data collection, and psychologically informed interventions to generate more precise evidence about how mental health affects economic participation.
Obara concluded that, while the study was still ongoing and the interim results required further validation with the complete dataset, the emerging evidence pointed toward an important policy lesson: improving employment outcomes may require addressing psychological barriers alongside conventional employment and skills interventions, while employment itself may be an important pathway for improving mental well-being.