Sage Journals: Discover world-class research

Abstract

Background:

Mental health and social exclusion are increasingly recognized as intertwined public health challenges. The aftermath of the COVID-19 underscores the need to integrate psychological well-being into national development and social inclusion agendas. Evidence and causal analyses from Southeast Asia remain sparse.

Design and methods:

This study combines longitudinal descriptive trends from the Indonesian Family Life Survey with an instrumental variable probit regression using cross-sectional data and examines the causal relationship between mental health and key dimensions of social exclusion: employment status and community participation. To address potential endogeneity and bidirectionality, “family mental health history” was employed as an instrumental variable, offering a theoretically plausible and statistically valid proxy for exogenous variation in individual mental health.

Results:

Descriptive findings showed that poorer mental health was associated with lower levels of employment and community participation. Instrumental variable results indicated that a one-point increase in the mental health index (worse mental health) reduced the probability of working by 25.3% and of participating in community life by 26.8%. The results support a robust negative association between mental health and social inclusion.

Conclusions:

This study makes a novel contribution by isolating the causal effect of mental health on social exclusion in an LMIC. Grounded in the capability approach, the findings illustrate how impaired mental health limits real freedoms and reinforces exclusion. The IFLS provides rare longitudinal, nationally representative evidence from Indonesia.

Public health implications:

Findings support integrated strategies to improve mental health and enhance inclusion—critical for LMICs navigating post-pandemic recovery and social transformations.

Keywords

mental health social exclusion indonesia capability approach instrumental variable analysis

Significance for public health

This study addresses the critical intersection of mental health and social exclusion within the context of Indonesia, a country facing significant public health challenges. By examining the relationship between mental health and social exclusion, this research highlights the pervasive impact public health outcomes such as mental health to social exclusion. The findings are particularly relevant in the wake of the COVID-19 pandemic which exacerbates mental health issues and social isolation. The study’s insights are crucial for informing public health strategies that aim to reduce social exclusion and improve mental health and vice versa, ultimately contributing to healthier, more resilient communities in Indonesia and similar low- and middle-income countries. This research underscores the need for integrated approaches in public health policy and intervention, particularly in the face of emerging global health crises.

Introduction

Mental health has become an increasingly critical public health issue globally, with significant implications for individuals’ ability to function and thrive in society. The COVID-19 pandemic has exacerbated mental health challenges, leading to a surge in cases of anxiety, depression, and social isolation globally, with studies indicating a 25% increase in the prevalence of these conditions worldwide.¹ The World Health Organization (WHO) identifies mental health disorders as leading causes of disability worldwide, with Indonesia facing similar challenges.² In low- and middle-income countries (LMICs) like Indonesia, the mental health landscape is shaped by a complex interplay of socio-economic factors, cultural beliefs, and limited access to healthcare services. Despite the growing awareness of mental health as a significant contributor to overall well-being, the provision of mental health services in Indonesia remains inadequate, particularly in rural areas where stigma and cultural barriers often hinder the utilization of available resources.³ The pervasive stigma associated with mental health conditions further exacerbates these issues, often leading to delayed or avoided treatment.⁴

This study is grounded in several interrelated theoretical frameworks. First, the social causation hypothesis suggests that exclusionary experiences such as unemployment, poverty, and social marginalization contribute to poor mental health by increasing stress and reducing protective social capital. Conversely, the social selection or drift hypothesis proposes that individuals with mental health conditions are more likely to experience downward social mobility or disengagement due to stigma, functional limitations, or behavioral challenges.^5,6 Both pathways are plausible and may operate simultaneously, creating a feedback loop that reinforces exclusion and distress.^7,8 A growing body of literature highlights the bidirectional relationship between mental health and social exclusion. Poor mental health can lead to labor market detachment and reduced social engagement through mechanisms such as cognitive impairment, stigma, and loss of motivation.^8,9 At the same time, exclusionary experiences—including unemployment, poverty, and social isolation—are themselves strong predictors of psychological distress, particularly in low- and middle-income contexts.^10,11

This bidirectional perspective is further supported by the capability approach, which emphasizes that mental health both affects and is affected by people’s ability to engage in work, form social relationships, and exercise agency in everyday life.^12,13 From this view, mental health is not simply a clinical outcome but a determinant of one’s freedom to function meaningfully within society. These frameworks provide the theoretical basis for the present study, which investigates not only how mental health affects social exclusion, but also recognizes that exclusionary conditions may reinforce psychological distress—a dynamic particularly relevant in low- and middle-income settings with constrained institutional support.^14,15

The selection of employment and community participation as key indicators of social exclusion in this study is based on Amartya Sen’s capability approach, which emphasizes individuals’ real freedoms to pursue lives they have reason to value. In this framework, social exclusion is conceptualized not merely as material deprivation but as the deprivation of basic capabilities—such as being able to work or engage meaningfully with others.¹² Employment and social participation were selected because they represent core functionings that enable individuals to exercise agency, attain dignity, and sustain social belonging. Prior work applying the capability approach in the mental health domain has underscored how mental health conditions can erode these functionings by limiting the ability to make choices, sustain relationships, or maintain productive roles in society.¹⁶ In turn, exclusion from employment or community life may reinforce psychological distress, further limiting capability expansion.¹⁷ These indicators are also particularly salient in the Indonesian context, where formal work and community involvement are deeply linked to social identity and access to support networks.¹⁸ This dynamic reinforces the need to consider mental health and social exclusion as mutually reinforcing processes.

The Indonesian Family Life Survey (IFLS) has provided valuable insights into the mental health status of the Indonesian population. Findings indicate a high prevalence of mental health issues, particularly among vulnerable groups such as women and the elderly.¹⁹ These mental health challenges are closely linked with socio-economic factors, including income and education levels, highlighting the need for comprehensive public health strategies that address these underlying determinants.²⁰

By leveraging data from IFLS, this study employs a mixed-methods approach, including descriptive analysis of longitudinal data to explore trends over time and instrumental variable probit regression analysis using cross-sectional data to investigate causal relationships. The descriptive analysis provides insights into the evolving patterns of mental health and social exclusion across different demographic groups in Indonesia, while the instrumental variable probit regression analysis attempts to find a causal relationship between mental health and social exclusion by determining whether or not mental health affects social exclusion using family mental health history as the instrumental variable. The instrumental variable probit analysis also addresses endogeneity caused by the possible bidirectional relationship between mental health and social exclusion indicators such as employment and community participation. These analytical methods allow for a comprehensive examination of how mental health influences social exclusion, providing crucial evidence to inform public health strategies.

Although a substantial body of research has examined the links between mental health and social exclusion, most studies are situated in high-income countries and rely on cross-sectional data, limiting the ability to infer causal relationships.^10,21,22 Furthermore, the bidirectional nature of this relationship—in which poor mental health may lead to exclusion, and exclusion may worsen mental health—is often under-theorized or neglected in empirical analyses.^5,6,8 There is also a notable lack of research from LMIC contexts, particularly in Southeast Asia, where both social exclusion and mental health burdens are rapidly evolving.^14,15,23 In Indonesia specifically, the literature remains sparse, with most existing studies descriptive in nature and lacking methodological rigor to account for endogeneity. To our knowledge, no prior study in Indonesia has employed instrumental variable techniques or nationally representative panel data to explore the mental health-social exclusion relationship. This study seeks to address these critical gaps by using data from the IFLS.

Design and methods

This section details the research framework, combining descriptive analysis of longitudinal panel data from IFLS with an instrumental variable probit regression using cross-sectional data. It outlines the study’s data sources, key variables (mental health and social exclusion indicators), and statistical techniques, ensuring robust analysis and credible findings to inform public health strategies.

Study design

This study employs a mixed-methods approach to examine the relationship between mental health and social exclusion in Indonesia. Specifically, it integrates a descriptive analysis of longitudinal data with an instrumental variable probit regression analysis using cross-sectional data. This approach introduces several methodological limitations that should be acknowledged. First, the descriptive trend analysis and the instrumental variable probit regressions are derived from partially overlapping but not identical samples. While both datasets draw from the same nationally representative panel, differences in inclusion criteria (e.g. age coverage, availability of mental health, and control variables) could introduce sample selection bias or limit generalizability of findings across time.^24,25 Second, the panel data capture intra-individual changes in mental health and social exclusion, while the cross-sectional analysis relies on point-in-time estimates using instruments (e.g. family mental health history) assumed to be stable. This methodological shift may lead to inconsistencies in interpretations if unobserved time-varying confounders are not fully accounted for, potentially weakening causal inference.^8,26 Third, there is a risk of recall bias and measurement non-equivalence over time, especially regarding mental health indicators and household conditions.

To minimize such effects, consistent variable definitions, harmonized coding, and robust checks for missingness were employed.²⁷ Even so, caution is warranted in interpreting the relationship between mental health and social exclusion as stable across years, given the dynamic and context-sensitive nature of both phenomena.^7,14 Despite these caveats, the complementary strengths of both data types—temporal depth from the panel and instrument-based causal identification from the cross-section—allow for a more nuanced understanding of the interplay between mental health and exclusion in the Indonesian context.

Data source

The data utilized in this study comes from IFLS, a longitudinal survey renowned for its extensive coverage and rich datasets on individual, household, and community-level variables across Indonesia.¹⁹ The IFLS, conducted in multiple waves since 1993, provides a unique opportunity to track changes over time and analyze the factors contributing to mental health outcomes and social exclusion.²⁰ This article utilizes waves 4 and 5 conducted in 2007 and 2014 respectively.

Descriptive analysis method

The descriptive analysis leverages longitudinal data from the IFLS to explore trends and patterns in mental health and social exclusion over time. This analysis focuses on identifying correlations and changes within individuals and demographic groups, such as age, gender, and socio-economic status, across different waves of the survey. By examining these trends, the study provides insights into the evolving relationship between mental health and social exclusion in the Indonesian context.

Instrumental variable probit regression method

To address potential endogeneity issues and the bidirectional nature of the relationship between mental health and social exclusion, the study employs a two-stage instrumental variable (IV) probit regression analysis. The two-stage IV probit analysis is a statistical technique used to estimate causal relationships in situations where the independent variable of interest is endogenous, meaning it is correlated with the error term, potentially leading to biased estimates in a standard probit model.²⁸ Endogeneity can arise due to various factors such as omitted variable bias, measurement error, or reverse causality.²⁶ In this case, reverse causality might occur between mental health and social exclusion. To address this, IV probit analysis employs an instrumental variable—an external variable that is correlated with the endogenous independent variable but uncorrelated with the error term—to correct for endogeneity.²⁹ In this study, family history of mental health issues serves as the instrumental variable to predict individual mental health status. This approach helps isolate the exogenous variation in mental health that can be used to estimate its causal impact on social exclusion indicators, specifically employment status and community participation.²⁸ This method is widely used in economics, epidemiology, and social sciences to estimate causal effects when simple regression models might produce biased results.^30,31

This study employs “family mental health history” as an instrumental variable. This choice is grounded in two key assumptions: first, relevance—family history is a well-established predictor of individual mental health due to genetic vulnerability and shared environmental factors; and second, exogeneity—it is unlikely to influence current employment or social participation directly, except through its effect on mental health. Prior studies have supported this approach, using family mental health history as an exogenous source of variation in psychological distress. For example, Boardman et al.³² treated familial mental illness as a proxy for genetic predisposition to assess causal mechanisms in mental health outcomes, reinforcing its validity as an instrument independent of current social conditions. This strategy enhances causal inference and strengthens the internal validity of the instrumental variable model used in this study.

The validity of the instrument is assessed through standard tests to ensure that it meets the relevance and exogeneity criteria, providing reliable estimates of the causal effects under study. The use of IV probit regression allows for a nuanced understanding of how mental health influences social exclusion in a cross-sectional context, complementing the longitudinal insights from the descriptive analysis.

Variables

The primary independent variable is the mental health index, derived from the IFLS data, which quantifies the severity of mental health issues based on a standardized scale ranging from 0 to 30.³³

The mental health index was constructed based on the 10-item Center for Epidemiologic Studies Depression Scale (CES-D), which is widely used to screen for depressive symptoms in epidemiological studies. The index was made following standard scoring procedures, including reverse coding for two positive-affect items prior to summation. The primary endogenous variable was the total mental health index score.

Regarding psychometric validation, the CES-D scale has been previously validated in multiple LMIC contexts, including Indonesia. For example, Surya et al.¹⁵ reported acceptable reliability and construct validity of the CES-D when used among Indonesian populations. Further, the CES-D items have been included in national surveys in Indonesia, such as the IFLS, from which the present data are derived.

In terms of internal consistency within the current study sample, the 10-item CES-D index demonstrates high reliability, with a Cronbach’s alpha exceeding 0.80, indicating strong internal coherence among items. Therefore, the mental health index used is both psychometrically sound and appropriate for use in the Indonesian context.

Data for the instrumental variable (which is family history of mental health) is generated from the 2007 survey in the fourth wave of IFLS. Individuals were classified as having a family history of mental health problems if any co-residing household member recorded an elevated CES-D score (>8) during the 2007 survey. The dependent variables are binary indicators of social exclusion: employment status (whether the individual is working) and community participation (whether the individual is actively involved in community activities) in 2014. Control variables include a range of demographic and socio-economic factors, such as age, gender, marital status, education level, urban/rural residence, and access to essential resources like clean water and electricity.^12,34

The selection of control variables is theoretically guided by established frameworks on the structural and social determinants of mental health and social exclusion. These variables were included to account for contextual, demographic, and infrastructural factors that may influence both mental health outcomes and individuals’ ability to participate in society, as emphasized in the capability approach.¹²

Age and gender are critical demographic controls, given their strong associations with both the prevalence of mental health conditions and patterns of labor market participation and social engagement.^35,36 Marital status is included as a proxy for social support and emotional stability, factors frequently shown to buffer against psychological distress and influence social inclusion.³⁷

Years of schooling (education) serves as an indicator of human capital and agency, both of which are central to the capability approach’s emphasis on functioning and choice. Higher education is generally associated with better mental health literacy and broader access to social and economic opportunities.³⁸

Urban-rural residence and Java versus non-Java location reflect geographic inequalities in Indonesia, where infrastructure, healthcare access, and economic prospects are highly uneven.^39,40 These spatial controls also align with prior research demonstrating that urban settings often confer both risks (e.g. stress, social fragmentation) and advantages (e.g. service access) to mental health outcomes.⁴¹

Household access to electricity and distance to water sources act as proxies for basic living standards and infrastructural deprivation. Poor access to utilities and water increases time burdens (particularly on women), reduces comfort and safety, and is often linked to poorer health and psychosocial stress, all of which can constrain social and economic participation.^42,43

Including these variables ensures that the estimated effects of mental health on social exclusion outcomes are not confounded by underlying socioeconomic or geographic conditions, thereby improving the robustness of the model’s causal interpretation.

Statistical analysis

All statistical analyses are conducted using STATA, ensuring precision and reliability in the results. Descriptive statistics provide an overview of trends and patterns, while the results of the IV probit regression are presented as marginal effects, making the findings more interpretable for policy, and public health implications.

The statistical significance of the findings is assessed, and robustness checks, including sensitivity analyses, are performed to ensure the validity of the results across different subpopulations, such as urban versus rural residents and Java versus non-Java regions.³⁰

Missing values were handled conservatively by restricting the analysis to individuals with complete information across all variables of interest. This resulted in a final analytical sample of 20,712 individuals, compared to the initial sample of 21,196 observations. This conservative approach to handling missing data minimized potential biases arising from imputation or misclassification. To capture the working population of Indonesia, the sample was restricted to age 15 years or older.

The strength and validity of the instrumental variable were assessed using the first-stage F-statistic, which consistently exceeded the threshold of 10, with F-statistics of 36.34 observed across specifications.

This methodological approach combines rigorous statistical techniques with a comprehensive analysis of both longitudinal and cross-sectional data, providing a robust framework for understanding the complex relationship between mental health and social exclusion in Indonesia.

Results and discussion

The analysis revealed significant relationships between mental health and social exclusion in Indonesia. The findings are presented in two parts: (1) descriptive statistics that illustrate trends over time, and (2) results from the instrumental variable probit regression that estimate the causal effects of mental health on social exclusion indicators such as employment status and community participation.

Descriptive analysis

Descriptive results from Table 1 indicate a general improvement in employment and community participation across all mental health categories between 2007 and 2014. For instance, among individuals classified as mentally healthy (MHI score 0–8), employment increased from 72.08% in 2007 to 78.17% in 2014, while participation in community activities rose from 69.25% to 79.48%. Notably, employment gains were also observed among those with mild and severe mental health issues, with employment among individuals with severe symptoms increasing from 58.06% to 77.66%. Paired t-tests and McNemar’s tests confirm that these changes are statistically significant at the 1% level.

Table 1.

Employment and community participation of Indonesians in various mental health conditions based on IFLS.

Mental health index score	Employment		Community participation		Employment change(p-Value)	Community participation change(p-Value)
	2007 (%)	2014 (%)	2007 (%)	2004 (%)	Employment change(p-Value)	Community participation change(p-Value)
Mentally Healthy (MHI score 0–8)	72.08	78.17	69.25	79.48	<0.01	<0.01
Mild mental health issue (MHI score 9–15)	67.04	79.22	67.26	77.54	<0.01	<0.01
Severe mental health issue (MHI score 16–30)	58.06	77.66	53.23	71.12	<0.01	<0.01

Following established practice and prior research, the sample was stratified into three mental health categories from the range of 0–30 on the 10-item version of the CES-D, with higher scores indicating more severe depressive symptomatology:

Healthy (0–8): Individuals scoring below the commonly used CES-D screening threshold of 10 are considered to have minimal depressive symptoms. A slightly more conservative cutoff of 8 has been used in studies to reduce false positives, especially in community settings.^33,44

Mild (9–15): This range corresponds to subclinical or moderate depressive symptoms. It captures individuals above the minimal threshold but below the level considered clinically significant.

Severe (16 and above): A score of 16 or more on the full 20-item CES-D scale has historically been used as the threshold for probable clinical depression.²⁹ On the shortened 10-item version, a score ≥ 16 has similarly been used to indicate more severe depressive risk.⁴⁵

These cutoffs are consistent with empirical use in previous population health research, including studies conducted in Southeast Asia and low- and middle-income countries.^15,23 They allow for meaningful categorization of mental health status while maintaining comparability with global epidemiological benchmarks.

While these patterns suggest simultaneous improvements in employment and social engagement, they coincide with a documented decline in average mental health scores over the same period. This apparent paradox can be contextualized within broader macroeconomic and social changes in Indonesia, including post-crisis economic growth, rapid urbanization, and labor market shifts that expanded employment opportunities across demographic groups.³¹ In such settings, increased labor absorption may occur even among individuals experiencing psychological distress, particularly in the absence of comprehensive mental health support systems.

Moreover, individuals facing deteriorating mental health may have remained in or entered the workforce out of economic necessity, driven by financial pressures and the limited availability of formal social protection mechanisms.⁴⁶ This aligns with findings in other LMICs, such as India and Brazil, where labor force participation often persists despite worsening mental health due to strong economic incentives or social expectations.^21,47 Similarly, higher community participation—typically viewed as protective—may introduce additional social demands, role expectations, or interpersonal tensions that compound psychological strain for vulnerable individuals.⁴⁸ Cultural values in Indonesia that emphasize diligence, family duty, and communal engagement may further encourage participation even among those experiencing mental health difficulties, potentially masking underlying vulnerabilities.

These findings emphasize the need to interpret descriptive patterns within structural and cultural contexts. From a policy standpoint, they point to the importance of integrating mental health considerations into employment and community development programs. Promoting mental health literacy, expanding access to mental health services at the community level, and designing inclusive labor policies that account for psychosocial well-being are essential for ensuring that economic and social participation does not come at the cost of worsening mental health outcomes.

Instrumental variable probit regression analysis

Table 2 presents the instrumental variable estimation. The cross-sectional instrumental variable probit regression for social exclusion indicators which include working status and community participation uses data from 2007 for family history of mental health condition only, and the rest of the variables are from 2014. The instrumental variable probit regression allows for the exogenous variable to be continuous, hence the mental health index score was used. The outcomes for working status and community participation are binary.

Table 2.

Result from instrumental variable probit regression for social exclusion variables.

Variable	(1)	(2)	(3)
	First stage: mental health index	Second stage: work status	Second stage: community participation status
Family MH history	0.204*** (0.022)
Urban 2014	0.019 (0.032)	−0.244*** (0.023)	−0.159*** (0.022)
Age 2014	−0.019*** (0.007)	0.108*** (0.005)	0.053*** (0.005)
Age 2014 squared	0.000** (0.000)	−0.001*** (0.000)	−0.000*** (0.000)
Gender 2014	0.075** (0.030)	1.140*** (0.025)	0.194*** (0.021)
Years of schooling 2014	−0.034*** (0.004)	−0.002 (0.004)	0.023*** (0.004)
Married 2014	−0.323*** (0.040)	−0.181*** (0.038)	0.221*** (0.035)
Java 2014	−0.104*** (0.030)	−0.050** (0.024)	0.245*** (0.022)
Constant	8.547*** (0.152)	−1.031 (0.678)	−0.964 (0.634)
Observations	20,712	20,712	20,712

Note. Robust standard errors are reported in parentheses. Column (1) reports the first-stage regression predicting the mental health index. Columns (2) and (3) present second-stage probit regressions predicting employment and participation, respectively. The control variables include urban status, age, age squared, gender, years of schooling, marital status, and Java region. Sample restricted to individuals aged 15 years and older. Significance levels: ***p < 0.01, **p < 0.05.

Table 3 presents the average marginal effects estimated from the instrumental variable probit models predicting employment status and participation in social activities. The marginal effects represent the change in the probability of the outcome associated with a 1-unit change in each explanatory variable, holding other variables constant.

Table 3.

Marginal effects of instrumental variable probit regression for social exclusion variables.

Variable	(1)	(2)
	Marginal effect (work status)*	Marginal effect (community participation status)*
Mental health index	−0.253 (0.167)	−0.268 (0.176)
Residual	−0.010 (0.019)	−0.004 (0.020)
Urban 2014	0.015 (0.019)	0.005 (0.020)
Age 2014	−0.060*** (0.006)	−0.044*** (0.006)
Age 2014 squared	0.027*** (0.001)	0.015*** (0.001)
Gender 2014	−0.000*** (0.000)	−0.000*** (0.000)
Years of schooling 2014	0.280*** (0.006)	0.054*** (0.006)
Married 2014	−0.001 (0.001)	0.006*** (0.001)
Java 2014	−0.045*** (0.009)	0.061*** (0.010)

Note. Marginal effects are calculated at the means after two-stage IV-Probit models. Robust standard errors are reported in parentheses. Sample restricted to individuals aged 15 years and older. Significance levels: ***p < 0.01.

Mental health status, as measured by the constructed index, was negatively associated with both employment and participation. Despite being large in magnitude, the marginal effects were not significant. Poorer mental health was associated with a 25.3 percentage point lower probability of employment and a 26.8 percentage point lower probability of participating in social activities, but neither effect reached conventional levels of statistical significance.

Among the control variables, urban residence was associated with a significantly lower probability of both employment and participation. Each additional year of age increased the probability of employment and participation, while the negative marginal effect of age squared indicated a diminishing return with age. Being male was associated with a significantly higher probability of employment and participation. Higher levels of education were positively associated with participation but had no significant effect on employment. Being married increased the likelihood of participation but did not significantly affect employment. Residing in Java was associated with higher participation rates.

Overall, the marginal effects analysis corroborates the findings from the IV-Probit coefficient estimates, further highlighting the importance of sociodemographic factors in shaping employment and social inclusion outcomes in the Indonesian context.

Although the estimated effects of mental health status on employment and participation do not reach statistical significance, the negative direction of the marginal effects remains consistent with prior findings. This trend aligns with literature indicating that psychological distress may impair cognitive, motivational, and interpersonal functioning, thus limiting individuals’ ability to engage productively in labor markets or community life. For instance, studies in LMICs such as Nigeria and Vietnam have shown that poor mental health reduces labor market performance and social participation, especially in informal sectors where social capital and functional ability are critical for access and productivity.^49–51

Despite the small marginal effects (–0.253 for employment and –0.268 for participation), the consistency in direction suggests a structural vulnerability among those with lower mental health scores, even if behavioral adaptation or economic necessity compels them to remain economically active. In contexts like Indonesia, where mental health services and disability protections are minimal,⁵² individuals with poor mental health may continue to work or engage socially out of survival rather than capability.

From a policy perspective, this suggests the urgent need for community-level mental health integration within employment support and social services. For example, public employment programs targeting vulnerable populations should include psychosocial support, while social capital interventions (e.g. community mental health groups) may improve inclusion outcomes beyond economic returns.²¹ Localized mental health screening tools integrated into primary health services could help identify at-risk individuals early, particularly in urban settings where the analysis suggests a negative association between urban residence and both inclusion outcomes.

Importantly, the analysis of control variables provides insight into key social determinants. Being male increases the probability of both employment and participation, reflecting persistent gender disparities in labor and social life across Southeast Asia.⁵³ Educational attainment, while not significantly influencing employment, significantly boosts participation, possibly due to the role of education in fostering civic engagement and confidence in social settings.⁵⁴ Urban residency’s negative correlation with participation and work may reflect the atomization of urban communities or competitive labor markets with higher exclusion risks, as observed in metropolitan centers in Thailand and the Philippines.⁵⁵

Comparative studies in Kenya and Bangladesh have similarly noted that urban poverty often interacts with exclusion, creating invisible barriers to inclusion even among the economically active.⁵⁶ These findings suggest that public health policies must target urban-specific drivers of social exclusion, including limited social cohesion and lack of inclusive urban planning. Additionally, the quadratic age effects highlight the importance of age-sensitive policies that support inclusion among both youth and the elderly, who may face distinct structural barriers.

Ultimately, while the primary relationship between mental health and social exclusion is not statistically significant in this analysis, the observed patterns reinforce the theory of bidirectionality. Mental health may both be a cause and a consequence of exclusionary experiences, a dynamic well-documented in the LMIC literature.⁷ The small marginal effects found here may underestimate longer-term effects, especially without longitudinal modeling of feedback mechanisms. These findings emphasize the need for multi-sectoral policy strategies that address the structural roots of exclusion while enhancing individual resilience and well-being.

Breaking the cycle between mental health and social exclusion in Indonesia

Addressing the complex and cyclical relationship between mental health and social exclusion in Indonesia requires an integrated policy response that spans the health, social, and economic sectors. Research has consistently demonstrated that mental health challenges can contribute to exclusion from employment, education, and community participation.^8,21 At the same time, exclusion itself—through mechanisms such as unemployment, poverty, and social isolation—can act as a powerful driver of psychological distress.^10,57 Breaking this feedback loop necessitates interventions that address both directions of influence simultaneously.

First, mental health services must be expanded and decentralized to reach populations outside of major urban centers. Integrating mental health into primary care—a recommendation supported by the WHO Mental Health Gap Action Program (mhGAP)—can improve accessibility and reduce stigma associated with specialized mental health treatment.⁵⁸ In Indonesia, where shortages of trained mental health professionals remain a significant barrier, task-shifting strategies, and community-based care models have shown promise in improving coverage and responsiveness.¹⁵

Second, employment policies must incorporate mental health considerations. This includes promoting inclusive hiring practices, enforcing anti-discrimination laws in the workplace, and embedding mental health support within job placement and training programs. Social protection mechanisms—including conditional cash transfers, disability benefits, and community reintegration programs—can also buffer against the exclusionary impacts of poor mental health, especially in vulnerable populations.^50,59

Third, public education campaigns should move beyond clinical awareness to address the societal and economic drivers of poor mental health. Campaigns should emphasize the legitimacy of psychological distress as a public health concern while challenging the structural conditions that exacerbate exclusion, such as gender inequities, informal labor markets, and community disconnection. Programs that foster peer-led support, participatory civic engagement, and intergenerational solidarity can promote recovery and re-engagement.^7,13

Finally, poverty alleviation and inclusive development strategies must incorporate psychosocial components. Evidence from low- and middle-income countries shows that mental health is both a cause and consequence of economic hardship, and interventions that ignore one side of this equation are unlikely to produce sustainable change.^8,14 Policies that integrate mental health into broader frameworks for social protection, education, and employment are essential for disrupting the cycle of disadvantage and promoting inclusive, community-based recovery in the Indonesian context.

Conclusion

This study contributes to the growing literature on the relationship between mental health and social exclusion by advancing both theoretical framing and methodological rigor in a low- and middle-income country (LMIC) context. Drawing on the capability approach and dual pathway theories of social exclusion—namely, the social causation and social drift hypotheses—this study conceptualizes mental health not merely as a clinical state but as a determinant of individuals’ capacity to engage meaningfully in work and community life. This perspective aligns with Amartya Sen’s notion of social exclusion as deprivation of capabilities and recognizes the bidirectional interplay between mental health and social functioning.

Methodologically, the study addresses potential endogeneity and reciprocal causality by applying an instrumental variable probit model using family mental health history as a theoretically motivated and statistically validated instrument. By combining longitudinal descriptive analysis and cross-sectional regression using nationally representative Indonesian data, the research offers a nuanced examination of evolving mental health patterns and their social consequences. The study is among the first in Indonesia to apply this level of econometric adjustment to the mental health–social exclusion nexus using a large-scale survey, thereby filling a critical gap in both local and global mental health research.

The findings indicate that poor mental health is modestly but significantly associated with reduced probability of employment and community participation. These associations remain robust even after accounting for potential endogeneity, reinforcing the need to recognize mental health as a structural barrier to full social inclusion. Subsample analyses further highlight greater vulnerability in rural and non-Java regions, pointing to geographic inequities in access to care and support.

From a policy perspective, these results imply that mental health cannot be siloed from broader social development efforts. In Indonesia, improving access to affordable and decentralized mental health services, especially in rural and outer-island regions, will be crucial. Integrated policies that simultaneously target stigma reduction, promote inclusive labor market opportunities, and support community engagement for those with mental health conditions are vital. Comparisons with other LMICs—such as India and South Africa, where mental health programs have been incorporated into community-based care models—suggest that intersectoral strategies are feasible and beneficial. As such, Indonesia could benefit from adopting similar localized, culturally sensitive public mental health interventions.

Nonetheless, some limitations should be acknowledged. First, the IV analysis is based on 2014 cross-sectional data, which does not capture post-COVID-19 dynamics or longer-term mental health trajectories. Second, while family mental health history met statistical relevance and passed the instrument strength test, residual unobserved confounding cannot be completely ruled out. Third, all key variables were self-reported, which may introduce bias through cultural stigma or social desirability. Despite these limitations, this study offers important theoretical and empirical insights into the pathways linking mental health and social exclusion in LMIC contexts. It supports the need for integrated policy solutions that view mental health not only as a health issue but also as a driver of inclusion, equity, and national development.

Footnotes

ORCID iD

Gerardia Gadisvania Sibbald

Ethical considerations

Not applicable.

Consent to participate

Not applicable.

Author contributions

The authors contributed equally from the conception, design, providing the study materials, and writing the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data used in this research is publicly available at .

References

World Health Organization. COVID-19 pandemic triggers 25% increase in prevalence of anxiety and depression worldwide. WHO, 2022. Available at: https://www.who.int/news/item/02-03-2022-covid-19-pandemic-triggers-25-increase-in-prevalence-of-anxiety-and-depression-worldwide

World Health Organization. Depression and other common mental disorders: Global health estimates. WHO, 2017.

Minas

Diatri

Pasung: physical restraint and confinement of the mentally ill in the community. Int J Ment Health Syst 2008; 2(1): 8.

Thornicroft

Brohan

Rose

, et al. Global pattern of experienced and anticipated discrimination against people with schizophrenia: a cross-sectional survey. Lancet 2009; 373(9661): 408–415.

Dohrenwend

Levav

Shrout

, et al. Socioeconomic status and psychiatric disorders: the causation-selection issue. Science 1992; 255(5047): 946–952.

Muntaner

Eaton

Miech

, et al. Socioeconomic position and major mental disorders. Epidemiol Rev 2004; 26(1): 53–62.

Sayce

Social inclusion and mental health. Psychiatr Bull 2001; 25(4): 121–123.

Lund

Breen

Flisher

, et al. Poverty and common mental disorders in low and middle income countries: a systematic review. Soc Sci Med 2010; 71(3): 517–528.

Wilton

Putting policy into practice? Poverty and people with serious mental illness. Soc Sci Med 2004; 58(1): 25–39.

10.

Paul

Moser

Unemployment impairs mental health: meta-analyses. J Vocat Behav 2009; 74(3): 264–282.

11.

Dalgard

Mykletun

Rognerud

, et al. Education, sense of mastery and mental health: results from a nation wide health monitoring study in Norway. BMC Psychiatry 2007; 7: 20.

12.

Sen

Development as freedom. Oxford University Press, 1999.

13.

Ware

Hopper

Tugenberg

, et al. Connectedness and citizenship: redefining social integration. Psychiatr Serv 2007; 58(4): 469–474.

14.

World Health Organization and World Bank. Out of the shadows: Making mental health a global development priority. WHO, 2016.

15.

Surya

Jaffri

Stilwell

, et al. The state of mental health and psychosocial support in Indonesia: a review of research and practice. Int J Ment Health Syst 2017; 11: 30.

16.

Ware

Hopper

Tugenberg

, et al. A theory of social integration as quality of life. Psychiatr Serv 2007; 58(4): 469–474.

17.

Mitra

The capability approach and disability. J Disabil Policy Stud 2006; 16(4): 236–247.

18.

Ibrahim

Alkire

Agency and empowerment: a proposal for internationally comparable indicators. Oxford Dev Stud 2007; 35(4): 379–403.

19.

Strauss

Witoelar

Sikoki

The fifth wave of the Indonesia Family Life Survey (IFLS5): overview and field report. RAND Corporation, 2016.

20.

Frankenberg

Karoly

Gertler

, et al. The Indonesian Family Life Survey (IFLS): study design and results from Waves 1 and 2. RAND Corporation, 2000.

21.

Patel

Saxena

Lund

, et al. The Lancet Commission on global mental health and sustainable development. Lancet 2018; 392(10157): 1553–1598.

22.

Shields

Price

SW.

Exploring the economic and social determinants of psychological well-being and perceived social support. Health Econ 2005; 14(10): 971–988.

23.

Liao

Wang

Fan

, et al. Measuring depressive symptoms in rural China using the CES-D: implications for prevention and treatment. Soc Psychiatry Psychiatr Epidemiol 2019; 54(9): 1121–1131.

24.

Caruana

Roman

Hernández-Sánchez

, et al. Longitudinal studies. J Thorac Dis 2015; 7: E537–E540.

25.

Vandenbroucke

von Elm

Altman

, et al. Strengthening the reporting of observational studies in Epidemiology (STROBE): explanation and elaboration. Epidemiology 2007; 18(6): 805–835.

26.

Angrist

Pischke

JS.

Mostly harmless econometrics: an empiricist’s companion. Princeton University Press, 2009.

27.

De Leeuw

. To mix or not to mix data collection modes in surveys. J Off Stat 2005; 21(2): 233–255.

28.

Wooldridge

JM.

Econometric analysis of cross section and panel data. 2nd ed. MIT Press, 2010.

29.

Stock

Watson

MW.

Introduction to econometrics. 4th ed. Pearson, 2019.

30.

Cameron

Trivedi

PK.

Microeconometrics: methods and applications. Cambridge University Press, 2005.

31.

Wooldridge

JM.

Introductory econometrics: a modern approach. 5th ed. South-Western Cengage Learning, 2013.

32.

Boardman

Alexander

Stallings

, et al. Genetic risk for mental illness and family-level stress: testing gene-environment interactions. Soc Sci Med 2011; 72(9): 1321–1329.

33.

Andresen

Malmgren

Carter

, et al. Screening for depression in well older adults: evaluation of a short form of the CES-D. Am J Prev Med 1994; 10(2): 77–84.

34.

Marmot

Allen

Goldblatt

, et al. Fair society, healthy lives: strategic review of health inequalities in England post-2010. The Marmot Review, 2010.

35.

Kieling

Baker-Henningham

Belfer

, et al. Child and adolescent mental health worldwide: evidence for action. Lancet 2011; 378(9801): 1515–1525.

36.

WHO. World mental health report: transforming mental health for all. World Health Organization, 2022.

37.

Umberson

Montez

JK.

Social relationships and health: a flashpoint for health policy. J Health Soc Behav 2010; 51(Suppl): S54–S66.

38.

Cutler

Lleras-Muney

. Education and health: Evaluating theories and evidence. NBER Working Paper No. 12352. 2006.

39.

Tjiptoherijanto

Infrastructure development in Indonesia. J Indones Econ Bus 2014; 29(3): 198–210.

40.

World Bank. Indonesia Economic Prospects, July 2020: The long road to recovery. 2020. Available at: https://www.worldbank.org/en/country/indonesia/publication/july-2020-indonesia-economic-prospect. (accessed 24 September 2025).

41.

Gruebner

Rapp

Adli

, et al. Cities and mental health. Deutsches Ärzteblatt International, 2017; 114(8), 121–127.

42.

Foster

Briceño-Garmendia

Africa’s infrastructure: a time for transformation. World Bank, 2010.

43.

UNICEF. Progress on drinking water, sanitation and hygiene 2019. Available at: https://www.unicef.org/media/55276/file/Progress%20on%20drinking%20water,%20sanitation%20and%20hygiene%202019%20.pdf. (accessed 24 September 2025).

44.

Björgvinsson

Kertz

Bigda-Peyton

, et al. Psychometric properties of the CES-D-10 in a psychiatric sample. Assess 2013; 20(4): 429–436.

45.

Radloff

LS.

The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977; 1(3): 385–401.

46.

Puspitasari

Idrus

Bass

, et al. Exploring the perceived challenges and support needs of Indonesian mental health stakeholders: A qualitative study. Int J Ment Health Syst 2021; 15(1): 1–13.

47.

Diniz

Almeida

Souza

, et al. Employment status and depression among Brazilian adults during economic recession. Rev Saúde Pública 2020; 54: 52.

48.

Dalgard

Bjørk

Tambs

Social support, negative life events and mental health. Br J Psychiatry 1995; 166(1): 326–334.

49.

Odejide

AO.

Mental health care in Africa: signs of progress. World Psychiatry 2006; 5(2): 92–95.

50.

Lund

De Silva

Plagerson

, et al. Poverty and mental disorders: breaking the cycle in low-income and middle-income countries. Lancet 2011; 378(9801): 1502–1514.

51.

Jenkins

Lewis

Bhat

, et al. Integration of mental health into primary care in Vietnam: a randomised controlled trial. Br J Psychiatry 2011; 199(6): 444–450. https://doi.org/10.1192/bjp.bp.110.086405

52.

Basrowi

Yuwono

Wagner

, et al. Exploring mental health issues and priorities in Indonesia. Front Public Health 2023; 11: 11748059.

53.

ILO. Women and men in the informal economy: a statistical picture. 3rd ed. International Labour Office, 2018.

54.

Berkman

Glass

Brissette

, et al. From social integration to health: Durkheim in the new millennium. Soc Sci Med 2000; 51(6): 843–857.

55.

Rahman

Islam

Urban exclusion and informal labour in Bangladesh. Dev Policy Rev 2019; 37(S2): O275–O295.

56.

Kabiru

, et al. Transition into adulthood: vulnerability among Nairobi’s youth. J Youth Stud 2013; 16(8): 1024–1040.

57.

Dalgard

Dowrick

Lehtinen

, et al. Negative life events, social support and gender difference in depression: a multinational community survey with data from the ODIN study. Soc Psychiatry Psychiatr Epidemiol 2006; 41(6): 444–451.

58.

WHO. Mental Health in development: stigma and discrimination. World Health Organization, 2010.

59.

World Bank. Indonesia’s rising divide: poverty, inequality, and the role of public policy. World Bank, 2016.

Mental health and social exclusion in Indonesia: A public health perspective

Abstract

Background:

Design and methods:

Results:

Conclusions:

Public health implications:

Keywords

Significance for public health

Introduction

Design and methods

Study design

Data source

Descriptive analysis method

Instrumental variable probit regression method

Variables

Statistical analysis

Results and discussion

Descriptive analysis

Instrumental variable probit regression analysis

Breaking the cycle between mental health and social exclusion in Indonesia

Conclusion

Footnotes

ORCID iD

Ethical considerations

Consent to participate

Author contributions

Funding

Declaration of conflicting interests

Data availability statement

References