Indonesia can increasingly answer the question of how many older people live in each part of the country. The more difficult question is what support those people need, whether that support is reaching them, and whether it is helping them remain healthy, independent and connected to their communities. Those distinctions will become increasingly important as population aging shifts from a demographic trend into an operational reality for health services, local government, families and emerging long-term care systems.
The Indonesia Aging, Long-Term Care & Community Support Knowledge Hub has examined the country’s aging transition through financing, workforce, family caregiving, primary care, dementia, rehabilitation, rural access, safeguarding and technology. Data connects all of those subjects. Without a stronger picture of need, access, function, outcomes and unpaid care, Indonesia risks knowing that its population is aging without knowing sufficiently well how that aging is being experienced locally.
The country already has substantial statistical and digital assets. BPS publishes dedicated statistics on older people; the 2025 Intercensal Population Survey provides updated demographic evidence; the Ministry of Health has established population targets for health programs for 2026–2030; Integrated Primary Care is expanding; and SATUSEHAT is creating a larger national health-data ecosystem. The next challenge is not simply collecting more information. It is building a measurement architecture capable of converting those data into decisions about prevention, service capacity, long-term care, geographic inequality and outcomes that matter to older people themselves.
Indonesia Now Has a Much Clearer Demographic Baseline
Reliable population denominators are a foundation for aging policy. The 2025 Intercensal Population Survey, or SUPAS 2025, reported that people aged 60 and over represented 11.97 percent of Indonesia’s population. BPS described this as confirmation that Indonesia is in an aging-population phase.
Bappenas has separately cited around 34.7 million older people, or 12.33 percent of the population, for 2025 and projects the proportion to rise to 20.31 percent by 2045. Differences between published figures can reflect different projection bases, survey approaches or reference datasets; they should not be casually treated as contradictions. For operational planning, the important requirement is to use a clearly defined official denominator appropriate to the program being measured.
The Ministry of Health took a significant step in this direction with Minister of Health Decree No. HK.01.07/MENKES/920/2025. The decree establishes target-population data for health development programs from 2026 to 2030, based on BPS population projections and disaggregated by age, sex and kabupaten/kota. It is intended as a single reference for planning, target setting and program monitoring across levels of government.
For older-person services, this matters because percentages without reliable denominators can be misleading. A district may report that screening increased by 20 percent, but leaders also need to know what proportion of the eligible older population was actually reached. A rapidly aging district may appear to be increasing activity while still falling further behind population need.
This is where population needs assessment becomes more than a planning exercise. Demographic intelligence should establish the size, distribution and likely trajectory of the population before service activity is interpreted.
Population Counts Do Not Measure Long-Term Care Need
Age is an important planning variable, but it is a weak proxy for dependency on its own.
Two people aged 78 can have completely different support requirements. One may live independently, work informally, travel regularly and manage several stable long-term conditions. Another may have advanced frailty, dementia, impaired mobility and require substantial daily assistance from family members.
A long-term care system therefore needs information about function as well as age.
Relevant domains can include mobility, cognition, ability to perform everyday activities, continence, nutrition, sensory impairment, communication, social participation and the availability of reliable informal support. The precise assessment mechanism may differ between health, social welfare and emerging community-care models, but the underlying principle is consistent: service need should reflect what a person can do and what assistance is required, not simply their birthday or diagnostic list.
This creates an important future data requirement for Indonesia. BPS provides strong population-level evidence about older people’s demographic, health, economic and social circumstances. Clinical systems capture diagnoses and encounters. Community programs may identify people requiring practical support. The missing connection is often a consistent view of functional need across these environments.
That gap matters particularly if community-based long-term care expands. Without functional information, local government may know how many older people live within its jurisdiction but remain unable to estimate how many require intermittent assistance, intensive family support, rehabilitation or ongoing personal care.
BPS Provides a Broad View of Aging That Health Data Alone Cannot Supply
BPS’s Statistics of Aging Population 2025 publication illustrates why national aging intelligence needs to remain broader than medical information. It draws on sources including the National Socio-Economic Survey, or Susenas, the National Labor Force Survey, Sakernas, population projections and secondary government data. Its scope includes demographics, education, health, economic activity, social conditions and access to protection and empowerment.
That breadth is important because later-life outcomes are shaped by much more than disease.
Income affects whether transport and privately purchased assistance are affordable. Living arrangements influence who is available when support is needed. Employment matters because many older Indonesians continue economic activity. Education can affect access to information. Geography shapes the distance to services. Social participation can influence isolation and mental wellbeing.
A national aging measurement system should therefore resist becoming dominated by whichever data are easiest to extract electronically.
Clinical records may be increasingly structured, but this does not make them a complete representation of aging. The person who never reaches a health facility may disappear from clinical datasets despite having substantial unmet need.
Good data collection and data quality therefore begin with clarity about what each source represents. Coverage, missingness and selection effects are part of the data, not technical inconveniences that can be ignored.
Longitudinal Evidence Adds the Dimension of Change
Cross-sectional statistics tell leaders what a population looks like at a particular point. Aging policy also needs to understand how people change over time.
The Indonesia Longitudinal Aging Survey 2023 provides an important additional evidence source. More than 4,000 people aged over 45 across nine regions were interviewed, creating information on demographic circumstances, health, economic conditions, living arrangements, use of aged-care services, technology, financial inclusion and social engagement.
Its longitudinal design is especially relevant to long-term care because dependency usually develops as a trajectory rather than an instantaneous event.
A person may move from full independence to mild difficulty, intermittent family assistance, more substantial functional limitation and eventually intensive daily support. If policy sees only the end point, opportunities for prevention, rehabilitation or caregiver support can be missed.
Longitudinal information can help answer questions such as whether mobility decline predicts later service use, how employment changes as people age, whether family structure affects care availability, and which groups experience accelerated economic or health vulnerability.
That does not mean one survey should become the sole planning dataset. Its value is different. Population statistics establish scale. Administrative information shows service interaction. Longitudinal research helps explain trajectories and relationships.
A mature aging-data architecture uses those sources together rather than expecting one system to answer every question.
Primary-Care Reform Is Creating a New Local Intelligence Layer
Indonesia’s Integrated Primary Care transformation creates perhaps the most important operational opportunity for better aging intelligence.
By July 2026, the Ministry of Health reported that around 9,000 Puskesmas, or 87.4 percent, had implemented Integrasi Layanan Primer. The model is organized around the life course and strengthens relationships between Puskesmas, Pustu, Posyandu and community-level activity.
Digitalization is also supporting local area monitoring. The Ministry has described the use of village health situation dashboards and the integration of Puskesmas information systems with SATUSEHAT.
This can make a crucial shift possible: from counting people who present to facilities toward understanding the health needs of defined local populations.
For older people, a Puskesmas could increasingly examine not only how many consultations took place, but whether priority groups received screening, whether people with identified risks completed follow-up, where functional deterioration is emerging and which villages have unusually low access.
But this requires a disciplined distinction between activity and coverage.
Five hundred older-person contacts may represent strong reach in one kecamatan and weak reach in another. The denominator, age structure, repeat visits and level of need all change the interpretation.
Organizations building similar performance systems can use the Quality Dashboard Builder to structure measures, trends and escalation thresholds. Its relevance is methodological rather than regulatory: it helps clarify what a measure is intended to show and what management response should follow.
Scenario: A district discovers that rising activity is masking falling coverage
A kabupaten sees a 15 percent annual rise in older-person health contacts through its Puskesmas network. At first glance, leaders regard this as evidence that expanded community outreach is succeeding.
When the figures are compared with updated target-population data, the picture changes. Several kecamatan have experienced rapid growth in the number of residents aged 60 and over. In those areas, activity has increased, but the proportion of older residents receiving the intended screening and follow-up has actually fallen.
The district then disaggregates the data by village, age band and sex. It finds that people aged 75 and over in several rural communities are particularly underrepresented. Staff discussions identify likely explanations: travel difficulty, reduced mobility and reliance on relatives for transport.
The operational response is therefore different from simply asking Puskesmas to conduct more appointments. Outreach and home-contact capacity are targeted to the communities with the largest gap, while the district monitors whether coverage improves among the older age groups previously missed.
The scenario illustrates why updated denominators matter. Activity was increasing, but access was not keeping pace with demographic need. Better data altered the decision.
SATUSEHAT Can Strengthen Measurement, but Integration Is Not the Same as Completeness
SATUSEHAT is progressively creating a national infrastructure for health-data exchange. Ministry dashboards already monitor implementation by health facilities, electronic medical-record data flows and morbidity information derived from transactions submitted through the platform.
For aging policy, the potential is significant. More connected health data can make patterns visible across facilities, reduce reliance on isolated organizational records and support analysis of disease burden, service use and transitions.
Yet integration needs to be interpreted carefully.
A facility being technically connected does not mean every relevant encounter is complete, timely or standardized. The Ministry’s own SATUSEHAT monitoring architecture distinguishes different levels of data submission and resource completeness. That is important because the absence of information can otherwise be mistaken for the absence of need.
Data governance should therefore monitor not only whether a facility is connected, but whether the information required for a particular purpose is being submitted reliably.
The issue becomes even more important if long-term care information is eventually linked across sectors. Health systems, social welfare services and community organizations may collect information for different purposes. Integration should not mean copying all data into one repository. It should mean allowing necessary information to move safely enough to support coordination and policy.
This is the practical territory of interoperability and data-exchange workflows: defining what information needs to move, between whom, at what point and with what safeguards.
Indonesia Needs Outcomes That Reflect Life, Not Only Service Use
Aging services can produce impressive activity statistics while changing very little for the person receiving support.
The number of screenings, referrals, home visits, consultations or community activities can demonstrate implementation. They cannot on their own demonstrate that people are maintaining independence or experiencing better quality of life.
A stronger measurement framework therefore needs outcomes frameworks and indicators that connect activity with meaningful change.
For older-person support, relevant outcomes might include maintaining or improving functional ability, remaining safely at home where that is the person’s preference, avoiding preventable deterioration, improving symptom control, sustaining caregiver capacity, participating in community life or experiencing smoother transitions between services.
Not every outcome should be expected to improve indefinitely.
For somebody with progressive dementia or advanced frailty, maintaining function for longer, reducing distress or supporting a family to continue safely may represent a strong outcome. Measurement systems that reward only improvement can undervalue good care for people whose underlying condition is deteriorating.
This distinction is important for the development of Indonesian long-term care. An older person’s blood pressure may be well controlled while their ability to eat, wash or move independently is declining. A clinical measure and a care outcome can therefore move in different directions.
Performance intelligence needs both.
Community-Based Long-Term Care Creates a Measurement Opportunity
Indonesia’s community-based long-term care development provides a practical setting in which broader outcome measurement can be tested.
ADB-supported community care hubs in Yogyakarta and Bali have used case management and community support to connect older people with services. The pilot has also demonstrated variation between sites in quality, coverage and engagement.
That variation is valuable evidence.
A pilot should not be evaluated only by asking whether the model operated. Leaders need to understand which functions worked, for whom, under what local conditions and with what level of workforce and community participation.
Measures might therefore combine reach, assessment, continuity, functional outcomes, caregiver experience, unmet need and the timeliness of referrals. Qualitative evidence should sit alongside quantitative indicators because it can explain why similar activity levels produce different experiences.
The Community Impact Report Builder can help organizations examining comparable community models connect service activity with outcomes, lived experience and wider local value. It does not provide an Indonesian evaluation standard; its usefulness lies in forcing a clearer distinction between what a service did and what changed because of it.
Scenario: Two community-care sites report similar activity but different outcomes
Two local community-aging programs each report that approximately 300 older people received support during the year. Both appear equally productive when performance is judged by reach alone.
A deeper review identifies very different service patterns.
In the first area, assessments are followed by coordinated action. Older people with mobility decline receive rehabilitation input, families receive practical guidance, unresolved referrals are followed up and functional status is reviewed. In the second area, many people attend activities and receive initial advice, but complex needs are referred onward without systematic confirmation that support was obtained.
Both programs can legitimately count 300 participants. Their impact is not equivalent.
The local government therefore introduces a small outcome set: change or maintenance in priority functional goals, completed referrals, unresolved high-risk needs, caregiver confidence for people receiving substantial family support and the older person’s experience of coordination.
The intention is not to punish the weaker site. The data allow leaders to see what additional workforce, referral relationships or supervision may be required. Performance measurement becomes a mechanism for adaptation rather than a league table.
Family Care Is a Major Part of the System but Remains Difficult to Measure
One of the largest gaps in conventional service data is unpaid family caregiving.
Indonesia’s long-term care system continues to depend heavily on relatives and community networks. More than half of older participants in the Indonesia Longitudinal Aging Survey lived in multigenerational households, illustrating how strongly later life remains connected to family structures.
But living with relatives does not tell policymakers how much care is being provided.
A family member may prepare an occasional meal, or may provide continuous supervision, intimate personal care, medication support, transfers and nighttime assistance. Those situations have completely different implications for caregiver wellbeing, employment, gender equality and the sustainability of aging in place.
If policy measures only formal services, a substantial part of the care economy remains invisible.
Better evidence could include who provides support, what tasks are performed, the approximate intensity of care, whether the caregiver has alternative support and whether caregiving affects employment or health. Such information does not need to become an intrusive national registry of family relationships. Periodic surveys and appropriately designed assessments can provide substantial intelligence.
This is also a measurement issue of equity. Women frequently provide a disproportionate share of unpaid care. A service model that appears inexpensive to government can therefore create significant hidden household and labor-market costs.
Equity Requires Data to Be Disaggregated
National averages are particularly dangerous in a country as geographically and socially diverse as Indonesia.
Access and outcomes can vary by province, kabupaten/kota, rurality, income, sex, age, disability, education and connectivity. Older people in remote islands may experience a completely different service environment from those living close to major urban hospitals.
Data-led equity planning therefore requires disaggregation rather than simply larger datasets.
Suppose 70 percent of older residents receive a particular preventive intervention. That figure can look strong nationally. If coverage is 90 percent in urban communities and 35 percent in remote areas, the national average conceals the operational problem.
Disaggregation also needs appropriate denominators. A small district can experience substantial year-to-year variation in absolute numbers, while a densely populated city may contain very different neighborhoods within a single administrative average.
Good governance therefore looks for persistent patterns rather than reacting automatically to every fluctuation.
The stronger question is not simply, “Which area is below average?” It is, “Is the difference persistent, material, explainable and actionable?”
Missing Data Can Be an Indicator of Missing Access
In aging services, the absence of records deserves analytical attention.
People who encounter services frequently generate data. Those who cannot reach them may generate very little.
This produces a form of visibility bias. A person with several hospital admissions can appear highly complex in administrative data, while an isolated older adult with severe mobility limitations but no recent contact may be almost invisible.
Indonesia’s extensive Posyandu and community-cadre networks offer one way of reducing that blind spot because community contacts can reveal people who are not regularly attending formal facilities. Integrated Primary Care can strengthen this population orientation when local services use denominators and outreach rather than waiting exclusively for attendance.
But the data need to retain the distinction between “no identified problem” and “not assessed.”
Blank fields, missed screening and incomplete follow-up should not automatically be treated as negative findings. Good data governance and information accountability make missingness visible so that leaders understand how confident they can be in the conclusions drawn.
Performance Dashboards Need an Operating Rhythm
Dashboards are useful only if somebody uses them to make decisions.
A local aging dashboard containing dozens of measures can create the appearance of sophisticated governance while producing little operational change. The stronger approach is a smaller set of measures linked to an explicit review rhythm.
Different information should operate at different frequencies. Immediate safety events may require rapid escalation. Referral backlogs could be reviewed weekly or monthly. Population coverage and geographic variation may be better understood over longer periods. Functional outcomes may require sufficient time for meaningful change.
This is the principle behind a dashboard operating rhythm and performance cadence. Data become useful when teams know:
- which measures require routine review;
- what level of variation warrants investigation;
- who is responsible for interpreting the finding;
- what contextual information is needed before action;
- how agreed interventions are followed up; and
- when a persistent local problem should reach higher-level governance.
The purpose is not to make local services spend more time reporting. It is to ensure that the information already being collected creates a feedback loop between delivery and decision-making.
Scenario: A dashboard detects a referral problem that activity reports missed
A city monitors older-person screening through its primary-care network. Screening performance is high and the indicator remains green for several months.
A new measure is introduced showing what happens after an identified problem requires referral. Leaders discover that completion rates for rehabilitation referrals are substantially lower than completion rates for medical referrals.
Discussion with Puskesmas teams shows that older people frequently receive a recommendation for rehabilitation but face transport barriers or uncertainty about where the service is available. Some families assume that a written referral itself means the process is complete.
The city does not respond by lowering the screening target. It separates the pathway into stages: need identified, referral initiated, service reached and follow-up completed.
This exposes where people are being lost.
Over subsequent months, local teams test more active follow-up and clearer navigation. Governance then reviews whether completion improves, particularly among people with reduced mobility.
The original screening indicator was technically accurate. It simply measured too early in the pathway to reveal whether the intervention actually reached the person.
Measurement Must Connect Funding With Need and Outcomes
As Indonesia considers how to finance a larger long-term care system, data will become central to resource allocation.
Health financing through Jaminan Kesehatan Nasional provides substantial protection for covered health services, while local government budgets, national programs, social assistance, household spending and unpaid care all contribute to the wider support environment. Long-term care does not yet sit within one unified funding mechanism.
This fragmentation makes measurement particularly important.
Policymakers need to understand not only expenditure but what that expenditure is purchasing and which needs remain outside funded pathways. A district may spend relatively little on older-person support because need is low, because families absorb most care, or because services simply do not exist. Expenditure alone cannot distinguish those explanations.
Similarly, low service use can indicate independence or unmet access.
As funding models develop, measures should therefore connect population need, service volume, unit cost where appropriate, outcomes and household burden. This supports a more meaningful conversation about aging outcomes and system value.
Value should not be reduced to the cheapest service model. A community intervention that costs more upfront but maintains function, supports a family caregiver and delays avoidable dependency may have greater long-term value than a cheaper intervention that records activity without changing the trajectory.
Workforce Data Need to Show Capability as Well as Headcount
Long-term care workforce planning cannot rely solely on the number of people employed.
Indonesia will need to understand where different skills are available, how workers are distributed geographically, which roles support community care and what level of supervision exists. Primary-care professionals, rehabilitation practitioners, social workers, care workers, community cadres and family caregivers all contribute differently.
Headcount data can hide significant capability differences.
Two districts may employ the same number of relevant workers, but one may have stronger rehabilitation skills, better supervision or more effective community outreach. Turnover and vacancy patterns can also alter practical capacity even when funded posts remain unchanged.
Future workforce intelligence should therefore connect supply to population need.
For example, areas with rapidly growing numbers of people aged 75 and over may require different skill mixes from areas with younger populations. Communities with limited family availability may need greater formal support. Rural areas may require broader generalist capability and different use of remote specialist advice.
Data should help identify those differences before they become visible as service failure.
Governance Needs to Distinguish Signal From Reporting Noise
More data can create more opportunities to overreact.
A monthly decline in one indicator may reflect random variation, seasonal patterns, data-entry delays or a genuine operational problem. Leaders need sufficient statistical and contextual literacy to distinguish these possibilities.
The same applies in the opposite direction. A small improvement should not automatically be attributed to a new initiative without understanding whether the change is sustained and plausible.
Organizations examining similar governance challenges can use the Governance Maturity Assessment to test whether performance information is connected to clear responsibility, escalation and oversight rather than merely circulated. The relevant lesson for Indonesia is that evidence becomes governance only when decision rights and accountability are visible.
This matters in decentralized systems. Kabupaten/kota need room to understand local conditions, while provincial and national levels need enough standardization to compare access and identify persistent variation.
Uniform data definitions and locally adaptable responses are not contradictory. They are often the combination required for effective decentralized governance.
Predictive Planning Should Be Used Carefully
Better demographic and service data also make scenario modelling increasingly possible.
Indonesia can estimate where the older population will grow, examine changing age structures, model different assumptions about disability or family availability and test what those assumptions might mean for community-care capacity.
This is particularly valuable because building a long-term care workforce and service infrastructure takes time. Waiting until demand becomes visible through crisis-driven hospital use would produce a reactive system.
The Digital Twin Scenario Modeler provides a practical example of how leaders can test alternative demand, workforce and capacity assumptions. It does not predict Indonesia’s future or replace official demographic modelling; its value is in demonstrating how assumptions can be made explicit and stress-tested.
Scenario models should always show uncertainty.
A projection that assumes a fixed proportion of older people will require formal care may be misleading if future health, family structure, housing, technology or prevention change the pattern. Models are most useful when they compare plausible scenarios rather than present one future as inevitable.
Scenario: A province models demand before building services
A provincial planning team expects substantial growth in its population aged 75 and over during the next decade. Rather than translating that demographic increase directly into a fixed number of residential beds, it models several possible pathways.
One assumes current patterns of family care remain broadly stable. Another assumes fewer working-age relatives are available for intensive unpaid care. A third tests the effect of stronger rehabilitation, prevention and community support on the proportion of people requiring higher-intensity assistance.
The scenarios produce different workforce and funding implications.
Instead of selecting one projection as “correct,” the province identifies investments that remain valuable across several futures: better functional assessment, stronger community rehabilitation, workforce development, caregiver support and data systems capable of detecting whether dependency is rising faster than expected.
Annual evidence is then compared with the assumptions.
This creates adaptive planning. Data are not being used to claim certainty about 2040; they are being used to make current decisions more resilient to uncertainty.
Older People Need to Influence What Counts as a Good Outcome
A technically sophisticated measurement system can still measure the wrong things.
Older people may value remaining in their own home, being able to attend religious or community activity, continuing productive work, maintaining relationships or avoiding dependence on their children. These outcomes do not always appear naturally in clinical or administrative datasets.
Outcome design therefore requires lived experience.
Periodic surveys, structured feedback, community discussion and qualitative research can reveal whether services are achieving what older people themselves consider important. Family caregivers should also contribute, while remaining distinct from the voice of the older person.
This can expose differences between professional and personal definitions of success.
A service may regard risk reduction as the primary outcome while the person values mobility and freedom. A family may prioritize safety while the older person prioritizes participation. Measurement cannot resolve those tensions automatically, but it can ensure they are visible.
This is especially important for people with cognitive impairment, disability or communication needs. Difficulty giving conventional feedback should not become exclusion from outcome evaluation.
A National Aging Data Architecture Should Connect, Not Centralize Everything
Indonesia does not need every piece of information about an older person to sit in one national database.
Different datasets serve different legitimate purposes. BPS provides population intelligence. Health systems require clinical information. Local governments need planning data. Social welfare services may hold information about protection or assistance. Community programs need enough information to coordinate support.
The strategic objective should be coherence rather than indiscriminate centralization.
That means common definitions where comparison matters, interoperable identifiers where lawful coordination requires them, transparent governance over access and an ability to combine de-identified information for population planning.
Privacy becomes increasingly important as datasets become more linkable. Functional dependency, cognitive impairment, financial circumstances and family relationships are sensitive information. Better analytics do not remove the requirement to collect only what is justified and control who can see it.
Trust is therefore infrastructure.
If people believe community assessment automatically exposes private information widely across government, they may become less willing to disclose the very circumstances needed for effective care planning.
From Reporting Performance to Learning From Performance
The final step is cultural rather than technical.
Performance systems can encourage defensive behavior if every variation is interpreted as failure. Local services may then focus on protecting indicators rather than understanding people.
The stronger model treats evidence as a learning mechanism.
If one Puskesmas achieves much higher older-person follow-up than similar areas, governance should ask what operational practice explains the difference. If a community-care pilot produces better functional outcomes in one village, leaders should examine whether workforce continuity, local leadership, referral relationships or family engagement contributed.
Similarly, repeated underperformance should lead to investigation and support before it becomes normalized.
This is the distinction between collecting performance information and creating a performance system.
The former produces reports. The latter connects data to curiosity, accountability, adaptation and resource decisions.
International Learning: Measure the Care System People Actually Experience
Many countries face the same measurement problem as Indonesia: health data are often more developed than long-term care data because clinical services have established administrative systems, payment mechanisms and professional documentation.
The transferable lesson is not that every country should create one standardized global aging dataset.
It is that measurement needs to match the system people actually experience.
Where families provide substantial care, family capacity needs visibility. Where care is community-based, measures should capture function and participation. Where geographic access varies, national averages require disaggregation. Where several agencies share responsibility, referral completion and continuity matter as much as individual service activity.
Indonesia also demonstrates the value of building data infrastructure while long-term care policy is still evolving. That creates an opportunity to avoid embedding a narrow institutional model into measurement from the outset.
If data architecture develops around function, independence, caregiver sustainability and community participation, it can support a more person-centered care system as formal provision expands.
Conclusion
Indonesia no longer lacks evidence that population aging is happening. BPS, SUPAS 2025, national population projections, longitudinal research, Ministry of Health target-population data and an expanding digital-health ecosystem provide an increasingly sophisticated picture of who is aging, where people live and how they interact with health and social systems.
The strategic challenge is now to measure what demographic statistics alone cannot reveal. Indonesia needs stronger visibility of functional need, unpaid caregiving, unmet access, continuity between services, geographic inequality and the outcomes people experience after support is provided. Those measures will become increasingly important as community-based long-term care moves from pilots and local models toward wider system development.
Good measurement will require national consistency without suppressing local variation. Central government can establish definitions, demographic baselines and interoperable infrastructure; provincial and kabupaten/kota leaders need to interpret those data through local geography, workforce and service capacity. Frontline services need information that improves decisions rather than simply increasing reporting burden.
The most important shift is conceptual. Indonesia should not judge its response to aging principally by how much activity the system produces. It should increasingly be able to show whether older people maintain function, receive support when needs change, experience equitable access, remain connected to their communities and avoid unnecessary dependency where possible. When measurement begins to answer those questions, data cease to be a description of population aging and become part of the infrastructure for managing it.