RESEARCH DOSSIER
Population Well-Being Promotion, Policy, and Ethical Governance
Abstract. Population well-being promotion concerns the improvement and equitable distribution of opportunities, lived experiences, and psychological functioning through social and institutional change. This fundamental research monograph examines the definition and history of well-being governance, the conceptual differences among life evaluation, affect, functioning, and objective conditions, and the methods required to evaluate public policy. A critical narrative synthesis integrates original psychological measurement studies, randomized cash-transfer and community-prevention trials, intervention meta-analyses, international statistical standards, and health-promotion guidance. The review distinguishes individual efficacy from population effectiveness; experimental effects from observational patterns; and favorable averages from distributional, intergenerational, and rights-related consequences. It analyzes education, employment, poverty reduction, community systems, digital platforms, and the emerging Artificial Era. Its final framework specifies testable multilevel pathways and standards for non-coercive, privacy-protective, democratically accountable evaluation. This is a critical narrative review, not a preregistered systematic review, original clinical trial, or independent meta-analysis.
Keywords: population well-being; well-being policy; public policy evaluation; subjective well-being; health promotion; population psychology; equity and social determinants; life satisfaction; well-being statistics; causal inference; community prevention; workplace well-being; school belonging; economic security; ethical governance; autonomy; privacy; digital governance; Artificial Era
The Scientific Object of Population Well-Being Promotion
Population well-being promotion is the deliberate modification of social, institutional, economic, environmental, and informational conditions so that the distribution of favorable lived experience and functioning improves across a defined population. Its object is a population distribution rather than the therapeutic trajectory of an individual. Governance adds the rules by which authorities select goals, allocate resources, justify trade-offs, protect rights, evaluate consequences, and revise policies. This research dossier examines the intersection of psychological outcomes with public decision-making. It asks how collective conditions affect subjective experience, how interventions can be evaluated causally, and which ethical restrictions should govern the use of well-being measurements. The intellectual case for using well-being as a public outcome was developed by Diener & Seligman (2004) and refined in comparative policy measurement by OECD (2025) and UNECE (2025).
The relevant outcome set is multidimensional. Life evaluation records reflective judgments about a person's life. Experienced well-being concerns emotions, pain, and the quality of episodes as they unfold. Eudaimonic functioning concerns valued agency, meaningful participation, competence, and development. Material conditions, institutional fairness, safety, health, relationships, and ecological security are important in their own right and can also affect psychological outcomes. The OECD (2024) framework consequently reports current well-being across numerous dimensions and examines the resources sustaining future well-being. No single average life-satisfaction score can represent all these dimensions or substitute for people's rights. Population promotion becomes scientifically intelligible when its interventions, exposures, mechanisms, and outcomes are specified separately.
Three distinctions organize this monograph. Promotion establishes enabling conditions, prevention addresses foreseeable risks, and treatment responds to conditions already requiring care; their programs may overlap, but their causal objectives and evaluation populations differ. Policy is the authorized choice of instruments and priorities, whereas implementation is the actual delivery of those instruments through organizations and communities. Ethics concerns the legitimacy of how objectives are chosen and effects distributed, not simply whether a metric moves in the desired direction. The review uses a critical narrative synthesis of verified original studies, public methodological standards, experimental program evaluations, and systematic reviews. It conducts no new population survey or original meta-analysis. The resulting framework is analytical and its prospective proposals are explicitly identified as hypotheses.
From Health Promotion to Well-Being Government
The modern lineage of population well-being governance joins public health, social psychology, welfare economics, and political theory. The Ottawa Charter (1986) Ottawa Charter, adopted at the first International Conference on Health Promotion in Canada in November 1986, linked health promotion to the capacity of people to exercise control over the conditions of their lives. Its practical agenda included healthy public policy, supportive environments, community action, personal skills, and reoriented health services. The premise was institutional and ecological: health-related opportunities arise within living and working environments. Contemporary well-being strategies inherit this approach while adding direct measurement of felt experience, life evaluations, social participation, and intergenerational sustainability.
A later psychological contribution challenged the tendency to treat aggregate economic production as the dominant proxy for collective progress. Diener & Seligman (2004) argued that direct reports about people's evaluations and feelings contribute information unavailable from conventional economic indicators. The construction of subjective well-being measures, including the Satisfaction With Life Scale in Diener et al. (1985), made such outcomes more accessible to research. However, the emergence of measurable preferences did not make public welfare a purely psychological concept. Resources, institutional access, rights, and physical conditions remained independent policy considerations. The methodological challenge became how to incorporate reported quality of life into decision-making while retaining a public account of structural conditions and distributive consequences.
International governance subsequently expanded the framework. The Geneva Charter (2021) Geneva Charter for Well-being emphasized equitable economies, the common good, universal coverage, the digital transition, and ecological protection. The WHO (2024) global framework consolidated strategic directions for well-being societies, and WHO (2025) translated these commitments into policy pathways for country-level action. The OECD (2025) measurement guidance and UNECE (2025) recommendations address the statistical infrastructure needed to monitor outcomes. These institutional texts are normative and implementation guidance, not proof that every suggested policy will cause an improvement. Their historical significance is the explicit admission that well-being requires intersectoral governance, longitudinal monitoring, and attention to the distribution of resources.
Well-Being as a Psychological and Political Construct
The term well-being refers to partially overlapping constructs whose differences have direct policy consequences. Cognitive life evaluation asks respondents to assess their life overall; it is affected by aspirations, comparisons, recollection, and perceptions of opportunity. Affect measures the frequency or intensity of pleasant and unpleasant episodes. Eudaimonic measures examine meaningful activity, autonomy, competence, and socially embedded functioning. OECD (2025) explicitly differentiates these components, and its updated guidelines add stronger attention to pain and relationally oriented experimental indicators. An intervention that raises immediate positive affect may not change life evaluation, and a policy that improves economic security may change judgments before everyday emotions. Researchers must therefore match the measured outcome to the hypothesized mechanism rather than treating 'happiness' as a unitary dependent variable.
The psychological constructs acquire a political meaning when policy-makers decide whose experiences count and which dimensions should receive priority. A national average can rise while the lower tail of the distribution deteriorates. An employment program might raise average income while increasing insecurity for a subset of workers. An urban redevelopment project can improve reported neighborhood attractiveness while displacing residents unable to absorb higher housing costs. Thus aggregation requires an ethical choice about weights, affected populations, and losses. The descriptive question 'what changed?' differs from the evaluative question 'was the distribution of the change legitimate?'. Both questions are indispensable to responsible governance, and neither can be answered through scale scores alone.
Population well-being must also be distinguished from population mental health. Psychological distress, diagnosable disorders, and favorable functioning can be related without reducing one to the other. The Patel et al. (2018) Lancet Commission documented the importance of mental health to sustainable development and connected care, prevention, human rights, and social environments. Its subject includes clinical systems; the present dossier addresses the broader allocation and governance of well-being promotion. The remit includes healthy populations and populations experiencing significant disability, poverty, exclusion, or illness. A governance system should not define a desirable population by screening out difficult experiences or by treating the inability to report positive feelings as a deficiency in civic worth.
The Institutional Production of Psychological Opportunity
Well-being is shaped by a chain of opportunity. Public rules influence income stability, housing access, transport, exposure to pollution, institutional discrimination, caregiving burden, and access to health and education. These conditions affect the time, safety, social relationships, and perceived agency available in everyday life. At the psychological level, people appraise controllability, future possibilities, social recognition, and support. Their responses then influence functioning, participation, and life evaluation. The Lund et al. (2018) review of reviews links social determinants with mental disorders across demographic, economic, neighborhood, event-related, and sociocultural domains, underscoring the relevance of upstream conditions. Those findings concern risks for mental disorders and do not by themselves quantify every pathway to positive well-being; mechanism-specific research remains necessary.
The unit of analysis matters. Household income, school climate, neighborhood walkability, municipal services, and national welfare rules operate at different levels and timescales. A person's reported life satisfaction may depend on both personal circumstances and conditions shared with neighbors, coworkers, or household members. Multilevel models can partition associations across these levels, but such partitioning does not automatically identify intervention effects. Selection into neighborhoods, policy adoption in response to prior distress, and simultaneous macroeconomic changes can all confound an association. Researchers should prespecify the institutional exposure, identify who can access it, record uptake and dosage, and compare groups exposed to the same larger context whenever possible.
Institutional opportunity also depends on social recognition. The same service may be formally available but practically inaccessible because of language, geography, disability accommodations, immigration status, transportation, or distrust created by earlier experiences. These barriers affect both treatment assignment and the meaning of survey nonresponse. The ethical and causal implications are linked: excluding difficult-to-reach groups biases the estimated outcome distribution while withholding their voice from resource allocation. The WHO equity framework (2024) operational framework encourages cross-sector measurement of social determinants and health equity, and the present synthesis extends its methodological logic to multidimensional well-being outcomes. Claims about precise effects, however, require population-specific measurement and a valid comparison.
Promotion, Prevention, Treatment, and Universalism
Promotion should be defined through the positive outcome it seeks and the conditions it changes. A school-wide initiative may aim to increase belonging, agency, and safety, whereas an anti-bullying intervention can be described through prevention of harm. A cash transfer may improve material security and indirectly reduce distress; a counseling service treats individuals who seek care. Classifying a program by its promotional intent does not reveal whether it works. The appropriate evidence depends on whether the theory predicts a change in reach, process, symptoms, emotional experience, social functioning, or population distribution. Evaluation should include both expected benefits and burdens, including administrative demands, stigma, and displacement of alternative services.
Universal provision and targeted allocation represent distinct design strategies. Universal policies can avoid eligibility stigma and simplify access, while targeted policies may concentrate scarce resources among those with the most severe deprivation. Targeting requires accurate identification, and identification can create errors of inclusion or exclusion. A universal school program also has unequal effects if some students face barriers to participation. Public reasoning therefore needs three quantities: the effect among participants, the proportion reached among eligible people, and the distribution of outcomes among those never reached. An impressive standardized effect in volunteers can coexist with negligible population change if reach is small or if attrition is concentrated in disadvantaged groups.
The choice between universal and targeted action cannot be resolved solely by aggregate effectiveness. The decision should examine the baseline distribution of risks and opportunities, the feasibility of delivery, potential coercion or humiliation, administrative cost, and the counterfactual to the proposed program. Evidence from conditional and unconditional transfers in Malawi Baird et al. (2011) illustrates how program design produces differing education and reproductive outcomes. The lesson is methodological rather than a universal policy ranking: conditions can improve an intended behavioral outcome while imposing different opportunity constraints on people whose circumstances differ. Ethical governance should make those differing goals and effects explicit.
Population Measurement and Statistical Architecture
A population monitoring system begins with an explicit estimand: the statistical quantity the administration wants to learn. Examples include the average change in life evaluation among all adult residents, the share experiencing severe pain, the difference in flourishing between income groups, or the distribution of within-person change after a new entitlement. These quantities answer different questions. A survey of program participants estimates their experience, not the welfare of everyone eligible. A survey of current residents may omit people displaced from an area. A national average may conceal persistent deprivation. The OECD (2025) guidelines describe survey wording, sampling, context effects, and the complementary roles of life evaluation, affect, and eudaimonic experience. A credible public dashboard should preserve these distinctions in its data definitions and presentations.
Scale selection is a matter of construct validity rather than convenience. The Satisfaction With Life Scale Diener et al. (1985) measures global cognitive judgments; it does not itself measure chronic pain, freedom from discrimination, or public trust. The Day Reconstruction Method in Kahneman et al. (2004) reconstructs the preceding day's activities and feelings, reducing some burdens of continuous experience sampling while remaining dependent on retrospective reports. Daily or momentary assessments can reveal changes concealed by global judgments, yet intensive sampling can impose burdens and omit people without suitable devices. For large surveillance systems, modes of interviewing, translation, literacy, seasonality, and the social setting of answering a survey are part of the measurement process and can influence reported scores.
Policy dashboards should specify a conceptual and statistical hierarchy. Primary psychological outcomes may include life evaluation, experienced pain and affect, and felt agency. Complementary outcomes should include income and wealth distribution, health, education, safety, housing, civic access, time use, and environmental quality. Measures of future resources belong in a separate temporal layer because a high current score cannot demonstrate sustainability. The OECD (2024) multi-domain framework and UNECE (2025) guidance support this broader logic. Index construction is permissible only with explicit weighting assumptions, sensitivity analyses, and retention of the underlying component series. Decision-makers should be able to see whether an apparently favorable composite trend depends on changes to a narrow subset of dimensions.
Comparability, Inequality, and the Problem of Aggregation
Cross-cultural measurement requires more than translation. An item on autonomy can evoke individual choice, family responsibility, or freedom from coercive dependence in different settings. Ratings may also reflect response styles, expectations, reference groups, and local norms for expressing distress. Henrich et al. (2010) cautioned against generalizing psychological findings from Western, educated, industrialized, rich, and democratic samples to humanity at large. This challenge is especially serious for national policy comparisons because measurement differences can be confounded with the countries under comparison. Valid comparison requires cognitive interviews, locally grounded item development, analysis of differential item functioning, and careful assessment of measurement invariance. A uniform questionnaire is a starting point for comparability, not its proof.
The distribution of well-being is an independent object of study. Means, medians, variances, lower-tail deprivation, and gaps among population groups convey different information. A movement in the average should be accompanied by estimates for disadvantaged and underrepresented groups, and uncertainty intervals should reflect the sampling design. Equity analysis examines both outcomes and access to the programs that produce them. Where a group is too small to support reliable public reporting, privacy safeguards and appropriate pooling methods are necessary; broad suppression of its experience can also erase the very inequality a monitoring system claims to assess. Research should document the reason for any suppression and whether the information remains available for protected policy evaluation.
An ethical problem arises when an administratively easy statistic becomes a substitute for a legitimate political objective. Aggregating across people requires a defensible account of whose benefits offset whose harms. The same increase in mean life satisfaction could result from a small benefit distributed widely or a substantial benefit restricted to a relatively privileged minority. Such outcomes are descriptively distinguishable even before a formal social welfare function is chosen. A defensible public dashboard therefore presents the whole distribution alongside summary measures, records the policy's foreseeable disparate effects, and acknowledges the value judgment embedded in any conversion of several dimensions into one rank. Measurement transparency is part of democratic accountability.
Causal Identification in Policy Evaluation
An observed improvement after a policy is introduced is evidence of temporal change, not yet evidence that the policy caused it. Economic cycles, shifting population composition, concurrent reforms, and regression toward the mean may account for the pattern. When feasible, randomized assignment of comparable units supports a causal interpretation by balancing potential confounders in expectation. Haushofer & Shapiro (2016) used household- and village-level randomization to investigate unconditional transfers in rural Kenya, including distinct assignments of transfer recipient, schedule, and amount. This permits questions about the particular cash-transfer design rather than relying on a simple comparison of recipients and nonrecipients. However, the study's setting, program, and time horizon also define the scope of its generalization.
Policy interventions frequently operate above the individual. Randomizing neighborhoods, schools, hospitals, or municipalities can capture shared institutional exposure, but analyses must account for cluster dependence and often a small number of assignment units. Contamination can occur when families or organizations interact across treated and comparison areas. Interference can also be the intervention's intended mechanism, as when a safer public space benefits residents who never directly participate. The standard individual-level assumption that one person's treatment does not affect another's outcome may then be implausible. A clear causal framework should describe exposure, spillovers, implementation fidelity, and what a comparison group actually experiences, rather than assuming all untreated people form an isolated control.
Where experiments are infeasible, interrupted time series, difference-in-differences, regression discontinuity, synthetic control, instrumental variables, and natural experiments may provide informative quasi-experimental estimates. Each method requires assumptions. A difference-in-differences estimate rests on an appropriate parallel-trends counterfactual, not on matching preintervention averages alone. Regression discontinuity yields local effects around an allocation threshold and requires protection against manipulation. Instrumental variables identify effects only under strong exclusion and relevance conditions. Synthetic controls require credible untreated donor units and defensible pre-policy fit. Transparency about such assumptions prevents observational associations from being converted into causal claims by rhetorical certainty. Researchers should specify the estimand, time horizon, analysis population, and testable diagnostics before evaluating a policy.
Longitudinal Designs, Adaptation, and Policy Durability
Population welfare changes at several temporal scales. A sudden income transfer can relieve urgent stress; a school climate reform may alter relational security over semesters; housing policy can shape developmental outcomes across years; ecological degradation can create consequences for later generations. Evaluations should distinguish immediate affect, medium-term functioning, and sustained life evaluation. A favorable result at three months does not establish a stable effect at three years. Attrition is especially consequential where people experiencing worse outcomes are more likely to migrate, stop responding, or leave a program. Intention-to-treat analyses preserve the assigned comparison, while longitudinal designs must also disclose differential follow-up and assumptions about missing data.
Panel studies can distinguish change within individuals from differences between people, although time-varying confounding remains a difficulty. Experience sampling captures fluctuations and contextual sensitivity, whereas annual surveys are better positioned to track broad trends with lower participant burden. The Day Reconstruction Method Kahneman et al. (2004) sits between these strategies by reconstructing sequences of daily activities. The optimal schedule depends on the expected latency of a mechanism: perceived economic security may change quickly, community trust over months, and childhood educational opportunities over longer periods. A policy evaluation that measures only one domain or one timepoint may misrepresent the magnitude, durability, or interpretation of an effect.
The observation that people adapt to circumstances cannot justify withholding improvements in material living conditions. Adaptation is neither uniform nor complete, and some harms accumulate rather than diminish. Income research underscores why the measured construct matters: Kahneman & Deaton (2010) found different income patterns for life evaluation and reported emotional experience in a large US survey, whereas a subsequent adversarial collaboration Killingsworth et al. (2023) refined the picture with more granular experience data and attention to heterogeneity. The proper inference is that income–well-being relationships depend on measurement, model specification, and subgroup, rather than a fixed universal threshold. Public policy should treat adaptive trajectories as empirical questions and track hard conditions alongside psychological evaluations.
Cost, Opportunity Cost, and Well-Being Policy Appraisal
Governments allocate finite funds, labor, land, political attention, and administrative capacity. Evaluation therefore asks what population would experience under the best feasible alternative use of those resources. Cost-effectiveness analyses can compare program costs with prespecified improvements in a well-being measure; cost-benefit analyses sometimes translate outcomes into monetary terms. Both procedures incorporate normative choices concerning the valuation of time, uncertainty, future effects, and benefits to different groups. A single well-being-year calculation can communicate scale, but it does not automatically establish equivalent meaning across populations, dimensions, or contexts. The broader welfare literature in Diener & Seligman (2004) and the multi-domain analysis of OECD (2024) show why decisions require both subjective and objective information.
An intervention should be appraised across reach, effectiveness, adoption, implementation, maintenance, and opportunity cost. If beneficiaries face travel expenses, unpaid time, privacy loss, or eligibility hurdles, those costs should enter the analysis. If a small social program creates improvement among volunteers but displaces a more effective housing measure, the local program's positive result may coexist with a poor system-level allocation. Research should also examine whether public resources reinforce existing inequalities through higher uptake by better-connected groups. Equity-weighted appraisal can make these considerations visible, provided the choice of weights is published and subjected to public deliberation. Cost and equity need not be collapsed into one opaque score.
Policy evaluation is ultimately sequential. Authorities first define outcomes and protected constraints; researchers then identify credible counterfactuals; evaluation reveals variation in effects; institutions revise their decisions in response to the evidence. Periodic monitoring supports accountability only if the findings can influence funding and implementation. A dashboard that never changes decisions has limited policy value regardless of statistical sophistication. To avoid metric gaming, the evaluation system should separately report exposure, implementation quality, participant outcomes, population outcomes, distributional change, complaints, and unintended consequences. These elements make it more difficult to manufacture an apparent success by improving the reported index while displacing harm outside the measurement frame.
Economic Security, Poverty Reduction, and Mental Experience
Material security is a major candidate mechanism for population well-being improvement because it affects consumption, exposure to shocks, perceived control, and the options available for education, work, care, and social participation. International observational evidence Deaton (2008) shows systematic relationships between income and life evaluation across countries, but such comparisons cannot isolate the effects of a specific fiscal policy. In randomized work, Haushofer & Shapiro (2016) studied transfers to poor households in rural Kenya and reported effects across economic and psychological outcomes. The original experiment is informative precisely because treatment assignment and transfer design were observed; it cannot establish that all unconditional transfers, financed and administered differently, will generate identical effects in every population.
The multifaceted graduation experiments in six countries Banerjee et al. (2015) tested combinations of productive assets, consumption support, training, and other assistance for people in extreme poverty. Their design demonstrates the potential of interventions that alter several constraints together. Yet a bundled treatment cannot by itself tell policy-makers which component is essential or whether the same combination is cost-effective under a different labor market. In Malawi, Baird et al. (2011) compared conditional and unconditional cash-transfer arrangements and reported heterogeneous consequences for schooling and reproductive outcomes. These experiments reveal both the causal relevance of material conditions and the limits of reducing policy appraisal to one favored metric. A child's future choices, a household's short-term insecurity, and a young person's control over time all warrant separate consideration.
Economic interventions raise questions of funding, distribution, and political durability. Transfer programs depend on administrative reach and fiscal continuity; the comparison of treated and control households may not capture broad price changes or shifts in community relations if scaled nationally. Psychological outcomes should therefore be interpreted together with income, debt, food security, health, education, and social participation. Investigators need to track who receives payments, who is excluded, and how household members share the resources. Positive average psychological effects should be regarded as evidence for a specific intervention under measured conditions, with explicit attention to duration and institutional context. The policy conclusion is to invest in credible comparative evaluation and transparent allocation, not to treat a single trial as a universal blueprint.
Community Systems, Social Relations, and Collective Capacity
Community belonging and social trust are simultaneously outcomes of social conditions and possible mechanisms through which institutions affect individuals. Neighborhood organization may change perceived safety, access to help, mutual recognition, and willingness to participate in decisions. Promotion strategies can therefore target public spaces, associations, school-community cooperation, and accessible opportunities for voluntary interaction. Evaluating these programs requires a distinction between participating in a community activity and experiencing durable belonging. A program can create frequent attendance while leaving exclusionary norms intact; another may change safety or trust among nonparticipants through spillovers. Measures should include perceived agency, reciprocal support, participation barriers, and subgroup differences, while distinguishing the social climate of a place from the private dispositions of residents.
The Communities That Care prevention system provides an instructive example of cluster-level experimentation. Hawkins et al. (2009) reported results from a community-randomized translational study designed to prevent adolescent substance use and delinquency by organizing local prevention systems. The research tests specified youth behavioral outcomes, not a complete index of population flourishing. Its relevance to this dossier lies in institutional coordination, theory-based selection of practices, and the evaluation of communities as units of change. It also raises questions about fidelity, local capacity, and the time required for social systems to affect individuals. Such community prevention evidence should not be reclassified as direct proof that general happiness rose without measuring happiness.
Community governance becomes ethically significant when residents influence the definition and interpretation of outcomes. A top-down well-being initiative that rewards visible participation can marginalize privacy-seeking residents or those working multiple jobs. Participatory design can identify transportation barriers, stigmatizing language, inaccessible facilities, and priorities that survey designers have missed. Community feedback should be assessed alongside representative measurement because public meetings often overrepresent people with time, confidence, and political access. The methodological solution is plural: population surveys, qualitative inquiry, administrative data, accessible grievance channels, and protected participation by underrepresented groups. Governance is improved when disagreement changes the model of success rather than being treated as resistance to a predefined positive-psychology program.
School Policy, Development, and the Ecology of Well-Being
Schools shape population well-being through compulsory or near-universal exposure during developmental periods. Their influence includes learning opportunities, peer safety, teacher relationships, inclusive procedures, institutional demands, and access to support. A school-level intervention can improve belonging or autonomy without producing immediate changes in standardized achievement, and the reverse is equally possible. Evaluation should identify distinct outcomes: distress and safety, social connection, academic engagement, self-efficacy, life evaluation, and later educational participation. Programs promoted as 'well-being curricula' should be tested against credible alternatives, including changes to disciplinary policy, workload, school climate, or staffing. A classroom lesson and a school-wide anti-exclusion policy involve different mechanisms and costs.
Population inference from school trials demands attention to clustering, developmental timing, family context, and selection into participation. School attendance itself is unequal, so a policy evaluated only among regular attendees may miss the very students who face greater insecurity. Repeated measurement can distinguish short-lived enthusiasm from longer-term changes in functioning. Data collection must guard against teachers identifying individual pupils through sensitive questionnaire responses and against incentives to inflate favorable school ratings. Public reporting should respect children's confidentiality and include an accessible procedure for responding to adverse experiences. When psychological screening is part of an initiative, referral pathways and informed consent procedures require their own clinical and legal governance.
The public role of schools includes supporting agency, the freedom to disagree, and access to developmentally appropriate assistance. Ethical promotion does not require children to affirm that they are happy or to adopt a prescribed outlook toward adversity. It provides conditions in which diverse emotional experiences can be communicated safely. The experimental literature on psychological interventions van Agteren et al. (2021) can inform which components warrant evaluation, but the meta-analysis pools interventions across clinical and nonclinical populations and cannot be substituted for local school policy trials. Structural changes to safety, belonging, and resources require direct tests. Public investment should be judged by outcomes across the full enrolled population, including pupils who receive few direct services.
Employment Institutions, Time, and Organizational Responsibility
Workplace well-being arises from remuneration, predictable hours, autonomy, job security, interpersonal treatment, physical safety, and the organization of tasks. Policies focused solely on individual resilience can miss preventable exposures created by schedules, workload, discrimination, and inadequate staffing. A responsible occupational strategy defines employer-controlled risk factors and assesses the change in working conditions alongside employee-reported outcomes. Evaluations should include turnover, sickness absence, job satisfaction, time conflict, harassment exposure, and the distribution of burdens across temporary and lower-paid workers. These outcomes may respond differently: a workplace could report increased productivity while employees experience worse fatigue. Measures should be designed to detect such divergence.
Evidence synthesis from organizational settings, including the 2024 systematic review Aust et al. (2024) of mental-health and well-being interventions among healthcare workers, supports careful comparison of organizational intervention types. Applicability depends on sector, workforce composition, implementation fidelity, and the nature of comparison conditions. The organizational setting of healthcare is particularly demanding and should not be assumed to represent other workplaces. Where interventions are rolled out by team or department, interference, supervisor behavior, and organizational change complicate attribution. Evaluation should measure the environmental change actually delivered rather than infer exposure from a policy memo announcing new intentions.
Employee consent and workplace hierarchy create distinct ethical risks. A survey administered by an employer may generate pressure to report high morale, particularly where employees fear being identified or losing advancement opportunities. Data on distress or life satisfaction should not be used to rank individuals for performance management. Public procurement and workplace regulation can establish accessible reporting, safe schedules, and accountability for employer-controlled exposures without requiring disclosure of intimate psychological information. Employee representatives should be included in outcome selection and interpretation. The test of organizational responsibility is whether institutions can demonstrate fair conditions and credible improvement without converting well-being measurement into a tool for surveillance or compliance.
Psychological Programs, Preventive Services, and Implementation
Individual-level psychological interventions can contribute to population promotion, particularly when access is broad, voluntary, and supported by appropriate referral. However, trial-level efficacy must be distinguished from population effectiveness. In their synthesis of 419 randomized trials with 53,288 participants, van Agteren et al. (2021) found evidence that several classes of interventions can improve mental well-being, while rating the overall quality of evidence as generally low to moderate and identifying variation by type and intensity. A pooled effect is informative about studies included in a review, not a guarantee for a nationally scaled program. Participants who volunteer for a mindfulness course may be systematically different from residents with the greatest material barriers or severe symptoms.
Reanalysis offers another essential constraint. White et al. (2019) argued that earlier meta-analytic estimates for positive psychology interventions were inflated by methodological and analytic problems. The lesson is not to reject psychological interventions wholesale; it is to ask whether a claimed benefit survives examination of control-group quality, attrition, selective reporting, outcome duplication, and small-study effects. Governments should favor preregistered protocols, valid active comparisons where feasible, independent follow-up, and adverse-event reporting. Particular care is needed when well-being services interact with established medical care, when group participation could reveal sensitive information, and when an intervention claims to prevent illness without evidence on that endpoint.
Implementation is a scientific variable. Program reach, staff training, cultural adaptation, fidelity, uptake, and discontinuation all influence whether trial results travel into public services. An intervention that works in a specialist research center can lose effectiveness under staffing shortages, language barriers, or fragmented administration. Conversely, adaptation that improves access may change the treatment components and require fresh evaluation. Policy-makers should monitor who declines participation and why, whether psychological burden increases for some participants, and which existing services are displaced. Clinical claims about diagnosis, treatment, and patient safety require their appropriate professional review; the present dossier evaluates promotion governance rather than providing medical treatment instructions.
Ethical Governance and the Authority to Define Well-Being
Well-being policy invokes both empirical judgments and legitimate choices about collective ends. Scientific evidence can establish that a program changes a measured outcome under defined conditions; science alone cannot determine whose interests deserve priority, which trade-offs are permissible, or whether a beneficial intervention is compatible with individual rights. Defining a public goal requires open justification, especially when authorities choose among happiness, autonomy, economic security, health, and ecological sustainability. The Geneva Charter (2021) charter and WHO (2024) framework provide international normative commitments to equity and sustainability. Such commitments establish an agenda for public deliberation; they do not exempt agencies from demonstrating causal benefit or from explaining why a chosen instrument imposes acceptable costs.
A fundamental risk is paternalism through psychological optimization. Governments and organizations may be tempted to treat an upward shift in reported happiness as authorization to shape citizens' values or discourage criticism. Yet a person's dissatisfaction can be an intelligible response to exploitation, injustice, grief, or political exclusion. A legitimate policy should preserve the freedom to experience and express unfavorable emotions without penalty. It should support access to resources and participation rather than make cheerful self-presentation a condition for services. When an institution rewards positive survey responses, the measure can become vulnerable to strategic answers and conceal the adversity the policy was meant to address.
Ethical authority therefore rests on procedural and substantive safeguards. Goal selection should involve affected communities; programs should be contestable; data collection should have a clear purpose and proportionality; and harms should be reported alongside benefits. Public justification should include alternative policy options and explain which constraints are treated as nonnegotiable, such as nondiscrimination and due process. Governments may reasonably use subjective well-being as one central indicator while acknowledging that rights and living conditions retain their importance regardless of measured feelings. This is an implication of the multidimensional evidence assembled in OECD (2025), UNECE (2025), and WHO equity framework (2024), together with the normative requirement for accountable institutions.
Autonomy, Consent, and Non-Coercive Promotion
In social research, voluntariness is not secured merely by displaying a consent box. People receiving public benefits, attending school, or working under supervisory authority can face implicit pressure to participate in an intervention or reveal information. The ethical evaluation must consider the consequences of refusal, the separation of service eligibility from research participation, and the availability of meaningful alternatives. Consent must be understandable to different language and literacy groups and appropriate to legal capacity and developmental stage. Research-grade informed consent procedures and public-service duties can operate together when data use is minimized, withdrawal is feasible, and care or benefits are not made conditional on agreeing to unrelated measurement.
Behavioral public policy introduces a further problem: interventions sometimes change choice architecture without the recipient's detailed awareness. A default or message may influence behavior, but its acceptability depends on purpose, transparency, reversibility, and whether it targets an independently justified public interest. Methods that exploit fear, vulnerability, or manipulated social approval demand special scrutiny even if a short-term satisfaction measure rises. A well-being objective does not license deception, emotional steering, or the suppression of dissent. Evaluation should therefore include autonomy indicators and procedural complaints as outcomes, not merely total participation rates or the desired behavior.
Non-coercive promotion also respects plural conceptions of a good life. Some individuals prioritize security, others creativity, religious practice, family care, political agency, solitude, or community service. Standardized instruments can provide comparability without establishing that a single lifestyle is universally optimal. Policy-makers can test whether conditions enable people to pursue their own valued goals, whether services can be accessed without ideological conformity, and whether people retain practical exit options. The psychological study of agency and meaning supports the relevance of valued activity, while the OECD (2025) guidelines caution that distinct subjective dimensions should be measured distinctly. Respect for diversity improves both ethics and construct validity.
Data Privacy, Surveillance, and Administrative Power
Population monitoring collects information that can reveal vulnerability, mental health history, household stress, social isolation, political distrust, and perceived institutional mistreatment. Combining surveys with health records, employment files, location traces, or digital behavioral logs can create greater inferential power but also increase reidentification and function-creep risks. A data set originally gathered to estimate population well-being may later be used to decide eligibility, detect dissent, or evaluate employees. Data governance must specify purposes, authorization, storage limits, access controls, retention periods, release thresholds, and independent oversight. Statistical utility does not cancel the obligation to protect people from foreseeable institutional misuse.
Public dashboards should publish distributions at an aggregation level that protects participants. Small cells can expose personal circumstances in sparsely populated areas or minoritized groups. Suppressing all small-group reporting, however, can make severe inequality invisible. A defensible system can use restricted access for qualified researchers, differential privacy where appropriate, pooling across suitable time windows, and co-designed dissemination protocols. The choice depends on the structure of the data, likely attackers, and whether the published statistic could alter the circumstances of the people it describes. Privacy guarantees and uncertainty from protective transformations should be disclosed rather than treated as invisible technical adjustments.
The growth of automated decision-making intensifies these concerns. Algorithms used to target well-being services may learn from historically unequal access, label reluctant participants as low priority, or recommend interventions using opaque proxies for distress. The UNESCO (2021) Recommendation on the Ethics of Artificial Intelligence provides a rights-oriented global governance reference for assessing AI's effects on human life and social institutions. In this dossier, algorithmic systems are studied as institutional instruments affecting Homo populations, while the broader Artificial Era provides a context for examining the consequences of new nonbiological systems. Claims about distinct artificial psychological organization belong to their own research objects and require separate operational validation.
Digital Platforms and the Artificial Era
Digital environments can modify daily experience through social comparison, exposure to hostility, access to community, attention capture, and algorithmic selection of content. These are policy-relevant exposures when participation is widespread and the same platform can affect different demographic groups in different ways. Population monitoring should therefore distinguish total screen time, type of activity, interpersonal quality, platform affordances, and the accessibility of non-digital alternatives. The World Happiness Report (2026) report investigates the relation of social media to well-being across populations and documents a field of competing explanations. Cross-sectional comparisons cannot establish whether platform exposure causes low well-being, whether vulnerable people seek different content, or whether third factors affect both.
The challenge for governance is to specify interventions with clear counterfactuals. A platform warning, default setting, age restriction, moderation change, or algorithmic feed adjustment may affect safety, engagement, autonomy, and social belonging differently. Experiments should disclose exposure changes, measure outcomes beyond short-term clicks, and examine unintended substitutions to other platforms. School-based digital policies should also consider opportunities for students excluded from mainstream social networks. Restricting an exposure can produce direct benefits while limiting valuable access for isolated groups. Causal attribution requires measuring the relevant change in digital experience, not relying on an undifferentiated measure of technology use.
The Artificial Era adds interactive artificial systems that can participate in communication, educational support, companionship, and advice. A well-being evaluation should ask whether these systems improve capability and reliable access to information, change human social ties, create dependency or privacy burdens, or redistribute institutional responsibilities. Public procurement and platform governance require transparent evaluation of intended use, error rates, accessible contestation, and routes to human assistance. The UNESCO (2021) framework supports human-rights attention to consequential AI deployment, while Geneva Charter (2021) explicitly identifies digital transformation as a domain of societal well-being governance. These instruments are governance sources, not experimental evidence that all AI applications improve or harm well-being.
Intergenerational Justice, Ecology, and Democratic Accountability
A policy can increase measured satisfaction in the present while undermining future access to clean environments, stable housing, healthy public institutions, or financial resources. The distinction between current well-being and resources for future well-being in OECD (2024) makes this temporal problem visible. Intergenerational assessment considers environmental stocks, public debt, infrastructure resilience, educational investment, and the transmission of social trust. Today's survey respondents cannot fully represent future persons, so political deliberation needs commitments and evidence about consequences that extend beyond current subjective reports. The Geneva Charter (2021) and WHO (2024) documents place sustainability within the definition of a well-being society.
Ecological policies create distributional dilemmas in the present as well. A transition to cleaner energy may improve long-run health yet impose short-term costs on households with limited alternatives. Spatial planning that expands green space can improve neighborhoods while increasing property values and displacement risk. An adequate appraisal evaluates both the environmental process and the distribution of access to its benefits. Reported satisfaction among remaining residents may rise after poorer households have moved away, creating a misleading apparent success. Population definitions should therefore track displacement, migration, and the experiences of affected people beyond new administrative boundaries when feasible.
Accountability requires institutions that can receive unfavorable evidence and change course. Participatory budgeting, independent evaluation, complaint mechanisms, and accessible scientific publication make policy revision possible. Periodic reporting should publish preregistered primary outcomes where feasible, reveal departures from the original evaluation plan, and separate political claims from scientific estimates. Governance fails when the measure of success becomes a target insulated from criticism. The best statistical infrastructure cannot compensate for an institution that refuses to reconsider its policy when harms are documented. An ethical well-being strategy combines measurement with legitimate oversight, clear responsibility, and rights-respecting mechanisms for correction.
Evidence Strength, Replication, and Transportability
The evidence base for population well-being promotion contains distinct grades of inference. Randomized trials can support claims about assigned interventions under their designs. Cluster experiments test some institutional mechanisms but may face contamination and limited numbers of independent units. Quasi-experiments can estimate effects when assumptions are defensible, while observational associations principally inform surveillance and hypothesis development. Meta-analyses aggregate studies with different interventions, populations, controls, and outcomes; their pooled estimates inherit the limitations of the included studies. The randomized findings from cash transfer experiments Haushofer & Shapiro (2016), multisite graduation programs Banerjee et al. (2015), and community prevention systems Hawkins et al. (2009) thus cannot be treated as interchangeable measures of one universal psychological benefit.
Evidence about psychological interventions also demands critical attention. The broad synthesis van Agteren et al. (2021) found benefits across multiple intervention categories but described generally low-to-moderate quality of evidence. The reanalysis White et al. (2019) reported that effects in earlier positive-psychology meta-analyses were smaller than previously suggested. These findings must be read together: intervention efficacy is plausible and supported in some settings, yet effect sizes, comparators, and generalizability differ substantially. Policy makers should not extrapolate efficacy in motivated trial volunteers to national population outcomes without measuring program reach and uptake. They should treat adverse response, attrition, and effects on socially disadvantaged groups as central evidence, not supplementary notes.
Transportability concerns the relationship between a studied setting and an intended new population. Cultural meanings, baseline resources, available services, political institutions, language, cost structures, and organizational competence can alter intervention outcomes. Henrich et al. (2010) illustrates the wider problem of sampling from populations that are not globally representative. A strong policy trial should report the context of treatment, moderators that may explain heterogeneous responses, and the range of settings to which its estimated effect plausibly applies. When an intervention enters a new jurisdiction, researchers should consider a fresh pilot, transparent adaptation record, and independent replication before claiming equivalent benefits. Evidence quality is not a permanent label attached to a program name; it is a property of a particular claim under a specified design.
Competing Explanations and Limits of Well-Being Metrics
Several theoretical models compete to explain why societies differ in reported well-being. Economic resources can enable consumption and reduce insecurity. Social integration may provide support and recognition. Institutional trust can affect perceptions of fairness and control. Cultural response styles and expectations can shape survey answers. Health and safety create both direct experiences and constraints on valued activity. These mechanisms coexist, and their relative contributions vary by population and context. Observational findings such as Deaton (2008) reveal associations across countries, whereas randomized interventions test narrower mechanisms. A persuasive synthesis asks which variables were manipulated, which mediators were observed, and whether rival explanations were excluded.
Income research provides a useful demonstration of the evolving scientific record. Kahneman & Deaton (2010) distinguished reflective and emotional well-being in a large US data set. Jebb et al. (2018) estimated income satiation patterns with a global sample under particular models. The later collaboration Killingsworth et al. (2023) examined disagreements between earlier income findings and emphasized heterogeneity across the distribution of happiness. These studies should not be reduced to a single universally valid income threshold. They vary in measures, sample composition, modeling, and inferential aims. Policy should focus on the concrete question of how changing income or insecurity affects specific outcomes in a target population and not infer a complete psychological theory from a cross-sectional curve.
The transformation of an indicator into a performance target creates a further rival explanation for favorable trends: reporting behavior may change while actual opportunity does not. Agencies can encourage favorable survey responses, selectively enroll participants, or redefine the denominator of a program. Such gaming is more likely where a single index controls funding or reputation. A governance system should preserve independent fieldwork, publish instrument changes, cross-validate with objective conditions, and provide a channel for dissenting interpretations. The possibility of strategic reporting is an empirical threat to construct validity and a political threat to accountability. It is especially important when respondents depend on the same institutions whose performance they are asked to rate.
A Multilevel Evaluation Architecture for Public Policy
The reviewed literature supports a practical research architecture with six connected levels. The first records social and material exposures: income security, housing, employment, safety, educational access, environment, and digital conditions. The second describes institutional instruments, including eligibility rules, service quality, and implementation fidelity. The third examines relational and cognitive mechanisms such as perceived control, belonging, trust, and opportunity. The fourth measures psychological outcomes across experienced affect, life evaluation, pain, and valued functioning. The fifth evaluates distributional effects by baseline disadvantage, age, disability, gender, geography, and other relevant categories, subject to privacy protection. The sixth monitors persistence, spillovers, public accountability, and future resources. This arrangement synthesizes the measurement logic of OECD (2025) and OECD (2024) with the upstream determinants described in WHO equity framework (2024).
The architecture is an integrative analytic model rather than an independently validated causal theory. Its proposed testable implication is that policies changing reliable structural opportunity will produce well-being effects partly mediated by perceived agency and relational security, with moderation by baseline deprivation. A second hypothesis is that improvements in average psychological outcomes will be overstated when program reach is measured only among compliant participants. A third is that accounting for displacement and future environmental resources will change the ranking of otherwise attractive municipal initiatives. The hypotheses require prospective designs, independent measurement, and policy-specific causal identification. Their value is to clarify what must be observed and what could falsify an explanatory claim.
A complete protocol should publish the target population, eligibility process, pre-policy baseline, primary and secondary outcomes, relevant subgroup definitions, measurement translations, consent procedures, evaluation horizon, and analysis plan. The statistical design should address clustering, longitudinal attrition, multiple outcomes, and missingness. A mixed-methods component can examine unintended consequences and explain apparent heterogeneity, but qualitative testimony and causal estimates serve different inferential functions. Independent researchers should be able to inspect data dictionaries, assumptions, and reproducible analyses within privacy restrictions. Public summaries should communicate practical magnitude, uncertainty, and equity rather than rely solely on statistical significance.
Unresolved Problems and Priorities for New Research
The central unresolved scientific problem is connecting system-level decisions with individual psychological mechanisms across time. Many well-being datasets are observational, and many intervention trials measure short horizons in selective participant groups. Complex reforms combine several policy instruments while macroeconomic and institutional conditions change simultaneously. Methodological progress requires credible natural experiments, cluster trials where ethical and feasible, harmonized longitudinal measurement, and preplanned examination of who is excluded. A second problem is the alignment of measures: researchers need to know when life evaluations, daily emotions, and experienced functioning diverge and how those divergences change policy conclusions. Work on measurement invariance, response shift, and survey nonresponse is particularly important for international comparison.
A third frontier concerns governance itself as a causal exposure. Few standard well-being indicators describe whether affected populations participated meaningfully in choosing outcomes, whether complaints were acted upon, or whether public agencies changed course after adverse findings. Research can experimentally or quasi-experimentally compare participatory and conventional decision processes while measuring both procedural fairness and welfare outcomes. A fourth frontier involves digital mediation. Platform policies and artificial systems can influence what information people receive, which relationships remain available, and whether decisions are contestable. This requires longitudinal measurement, independent access to relevant exposure data, and protection against intrusive tracking. The World Happiness Report (2026) discussion of social media and well-being illustrates the importance of competing hypotheses and policy translation.
A fifth problem is the ethical evaluation of trade-offs. Policy science needs reporting standards that place changes in average outcomes beside disparities, privacy costs, autonomy effects, implementation failures, and effects on future resources. The current international tools of WHO (2025), OECD (2025), and UNECE (2025) provide useful foundations, while actual instruments require local validation and public deliberation. Research must also confront limits of quantification: important civic values, cultural commitments, and rights cannot always be represented through commensurable scores. Scientific progress depends on recognizing those limits and making the inferential and normative structure of every policy claim inspectable.
Conclusion
Population well-being promotion is a multilevel program of scientific inquiry and public governance. Its object is the distribution of opportunities, experiences, and functioning created by social conditions and institutional choices. Historical work in health promotion established the importance of enabling environments, while psychological measurement introduced direct evidence about people's evaluations and lived experiences. Contemporary international frameworks extend the agenda toward equity, digital governance, and ecological sustainability. The empirical record demonstrates that material and institutional interventions can change consequential outcomes in defined settings, and that psychological programs have evidence of efficacy under conditions qualified by study design, selection, and bias.
Four principles follow from the synthesis. First, policy should measure several outcome dimensions and the conditions that produce them. Second, causal claims must be tied to a specified intervention, comparison, population, and time horizon. Third, averages must be accompanied by distributional analysis, including access barriers, nonresponse, and unintended displacement. Fourth, governance must preserve autonomy, privacy, nondiscrimination, democratic participation, and the capacity to revise decisions. Together these principles connect research quality to public legitimacy and make it possible to distinguish a genuine improvement in people's lives from an improvement in institutional performance statistics.
The continuing task is to connect individual psychological mechanisms to community, economic, educational, organizational, technological, and ecological change through designs capable of testing causal explanations and exposing unequal consequences. In the Artificial Era, this includes studying how artificial systems reshape the conditions under which Homo populations learn, relate, decide, and exercise agency. Such investigations require careful operational definitions and direct evidence. The research dossier establishes a framework for evaluating population well-being as a public outcome with measurable mechanisms, explicit normative commitments, and permanent accountability to the people whose lives are affected.
References
- Aust, B., Leduc, C., Cresswell-Smith, J., O’Brien, C., Rugulies, R., Leduc, M., Ní Dhálaigh, D., Dushaj, A., Fanaj, N., Guinart, D., Maxwell, M., Reich, H., Ross, V., Sadath, A., Schnitzspahn, K., Tóth, M. D., van Audenhove, C., van Weeghel, J., Wahlbeck, K., ... Greiner, B. A. (2024). The effects of different types of organisational workplace mental health interventions on mental health and wellbeing in healthcare workers: A systematic review. International Archives of Occupational and Environmental Health, 97(5), 485–522. https://doi.org/10.1007/s00420-024-02065-z
- Baird, S., McIntosh, C., & Özler, B. (2011). Cash or condition? Evidence from a cash transfer experiment. The Quarterly Journal of Economics, 126(4), 1709–1753. https://doi.org/10.1093/qje/qjr032
- Banerjee, A., Duflo, E., Goldberg, N., Karlan, D., Osei, R., Parienté, W., Shapiro, J., Thuysbaert, B., & Udry, C. (2015). A multifaceted program causes lasting progress for the very poor: Evidence from six countries. Science, 348(6236), Article 1260799. https://doi.org/10.1126/science.1260799
- Deaton, A. (2008). Income, health, and well-being around the world: Evidence from the Gallup World Poll. Journal of Economic Perspectives, 22(2), 53–72. https://doi.org/10.1257/jep.22.2.53
- Diener, E., & Seligman, M. E. P. (2004). Beyond money: Toward an economy of well-being. Psychological Science in the Public Interest, 5(1), 1–31. https://doi.org/10.1111/j.0963-7214.2004.00501001.x
- Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment, 49(1), 71–75. https://doi.org/10.1207/s15327752jpa4901_13
- Haushofer, J., & Shapiro, J. (2016). The short-term impact of unconditional cash transfers to the poor: Experimental evidence from Kenya. The Quarterly Journal of Economics, 131(4), 1973–2042. https://doi.org/10.1093/qje/qjw025
- Hawkins, J. D., Oesterle, S., Brown, E. C., Arthur, M. W., Abbott, R. D., Fagan, A. A., & Catalano, R. F. (2009). Results of a type 2 translational research trial to prevent adolescent drug use and delinquency: A test of Communities That Care. Archives of Pediatrics & Adolescent Medicine, 163(9), 789–798. https://doi.org/10.1001/archpediatrics.2009.141
- Helliwell, J. F., Layard, R., Sachs, J. D., De Neve, J.-E., Aknin, L. B., & Wang, S. (Eds.). (2026). World Happiness Report 2026. Wellbeing Research Centre, University of Oxford. https://www.worldhappiness.report/ed/2026/
- Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2–3), 61–83. https://doi.org/10.1017/S0140525X0999152X
- Jebb, A. T., Tay, L., Diener, E., & Oishi, S. (2018). Happiness, income satiation and turning points around the world. Nature Human Behaviour, 2, 33–38. https://doi.org/10.1038/s41562-017-0277-0
- Kahneman, D., & Deaton, A. (2010). High income improves evaluation of life but not emotional well-being. Proceedings of the National Academy of Sciences, 107(38), 16489–16493. https://doi.org/10.1073/pnas.1011492107
- Kahneman, D., Krueger, A. B., Schkade, D. A., Schwarz, N., & Stone, A. A. (2004). A survey method for characterizing daily life experience: The day reconstruction method. Science, 306(5702), 1776–1780. https://doi.org/10.1126/science.1103572
- Killingsworth, M. A., Kahneman, D., & Mellers, B. (2023). Income and emotional well-being: A conflict resolved. Proceedings of the National Academy of Sciences, 120(10), Article e2208661120. https://doi.org/10.1073/pnas.2208661120
- Lund, C., Brooke-Sumner, C., Baingana, F., Baron, E. C., Breuer, E., Chandra, P., Haushofer, J., Herrman, H., Jordans, M., Kieling, C., Medina-Mora, M. E., Morgan, E., Omigbodun, O., Tol, W., Patel, V., & Saxena, S. (2018). Social determinants of mental disorders and the Sustainable Development Goals: A systematic review of reviews. The Lancet Psychiatry, 5(4), 357–369. https://doi.org/10.1016/S2215-0366(18)30060-9
- Organisation for Economic Co-operation and Development. (2024). How’s Life? 2024: Well-being and resilience in times of crisis. OECD Publishing. https://doi.org/10.1787/90ba854a-en
- Organisation for Economic Co-operation and Development. (2025). OECD guidelines on measuring subjective well-being (2025 update). OECD Publishing. https://doi.org/10.1787/9203632a-en
- Patel, V., Saxena, S., Lund, C., Thornicroft, G., Baingana, F., Bolton, P., Chisholm, D., Collins, P. Y., Cooper, J. L., Eaton, J., Herrman, H., Herzallah, M. M., Huang, Y., Jordans, M. J. D., Kleinman, A., Medina-Mora, M. E., Morgan, E., Niaz, U., Omigbodun, O., ... Unützer, J. (2018). The Lancet Commission on global mental health and sustainable development. The Lancet, 392(10157), 1553–1598. https://doi.org/10.1016/S0140-6736(18)31612-X
- United Nations Economic Commission for Europe. (2025). Guidelines on measurement of well-being. United Nations. https://unece.org/info/publications/pub/407877
- United Nations Educational, Scientific and Cultural Organization. (2021). Recommendation on the ethics of artificial intelligence. UNESCO. https://www.unesco.org/en/legal-affairs/recommendation-ethics-artificial-intelligence
- van Agteren, J., Iasiello, M., Lo, L., Bartholomaeus, J., Kopsaftis, Z., Carey, M., & Kyrios, M. (2021). A systematic review and meta-analysis of psychological interventions to improve mental wellbeing. Nature Human Behaviour, 5, 631–652. https://doi.org/10.1038/s41562-021-01093-w
- White, C. A., Uttl, B., & Holder, M. D. (2019). Meta-analyses of positive psychology interventions: The effects are much smaller than previously reported. PLOS ONE, 14(5), Article e0216588. https://doi.org/10.1371/journal.pone.0216588
- World Health Organization. (1986). Ottawa Charter for Health Promotion. First International Conference on Health Promotion, Ottawa, Canada. https://www.canada.ca/en/public-health/services/health-promotion/population-health/ottawa-charter-health-promotion-international-conference-on-health-promotion.html
- World Health Organization. (2021). The Geneva Charter for Well-being. World Health Organization. https://www.who.int/publications/m/item/the-geneva-charter-for-well-being
- World Health Organization. (2024). Achieving well-being: A global framework for integrating well-being into public health utilizing a health promotion approach. World Health Organization. https://www.who.int/publications/i/item/9789240084858
- World Health Organization. (2024). Operational framework for monitoring social determinants of health equity. World Health Organization. https://www.who.int/publications/b/67490
- World Health Organization. (2025). Implementing the global framework on well-being at the country level: Policy pathways. World Health Organization. https://www.who.int/publications/i/item/9789240112773