RESEARCH DOSSIER
China Basic Pension Participation and Hukou Portability
Abstract. China's basic pension system provides an unusually consequential setting for studying the interaction between social insurance enrollment, household registration, labor mobility, intergovernmental finance, and the continuity of accrued rights. By the end of 2025, the national statistical authority recorded 1,075.99 million basic pension participants, divided between 546.80 million in the urban employee system and 529.19 million in the urban–rural resident system. These large administrative stocks do not by themselves measure continuous contributions, the reliability of transferable records, eligibility for a particular retirement location, or the adequacy of benefits. This monograph examines how those dimensions interact under China's principal legal instruments and dated administrative reforms. The investigation distinguishes transfer within employee insurance, transfer of residence-linked resident accounts, and conversion between the employee and resident schemes. It examines the 2009 cross-province employee-transfer measures, the Social Insurance Law, both 2014 harmonization instruments, the 2021 flexible-employment guidance, national pooling introduced in 2022, and the pension eligibility changes scheduled from 2025 and 2030. Official statistical releases for 2024 and 2025 are reconciled without treating scheme membership as a coverage rate, and the migrant-worker survey series is used to specify the population whose mobility makes portability economically salient. Peer-reviewed work supplies evidence of heterogeneous enrollment and a connection between provincial pooling and intraprovincial migration, with explicit attention to identification limits. The central finding is that the legal preservation of contribution histories, the location in which those histories generate a pension, the conversion of histories between schemes, and the actual purchasing power of retirement income are four separately testable properties. China has substantially expanded the formal architecture of transfer and fund coordination, while hukou-linked residence rules, differing contribution histories, minimum vesting periods, and local administrative conditions continue to shape realized rights. The analysis generates a practical research agenda centered on person-level longitudinal reconciliation of work, hukou, account, contribution, transfer, and benefit records. No new proprietary theory, econometric estimate, or independently replicated causal effect is asserted.
Keywords: China; basic old-age insurance; pensions; hukou; household registration; labor migration; migrant workers; social insurance; pension portability; pension rights; urban employee pensions; urban–rural resident pensions; national pooling; social protection; actuarial adequacy; administrative data; retirement age.
Introduction
China's pension landscape combines a very large statutory enrollment base with a labor market characterized by geographically mobile employment histories. The 2025 national statistical communiqué records 1,075.99 million people participating in basic pension programs at the end of that year, while the 2025 migrant-worker monitoring survey reports 301.15 million migrant workers under a specific rural-hukou-based statistical definition. These two populations cannot be set against one another as a numerator and denominator: some migrant workers are resident pension participants, some belong to the employee system, some may have interrupted payment histories, and the statistical populations include different age and legal categories. Their coexistence nevertheless demonstrates why preserving a worker's past contributions across residence and employment changes is a first-order public insurance question.
The research problem is framed as a sequence of administrative and financial propositions. An individual can be permitted to enroll in a program yet encounter difficulty paying contributions regularly. Recorded contributions can be preserved during a move yet be evaluated against a retirement-claiming location determined by prior years and household registration. Funds can be transferred between administrations while contribution years are converted differently across statutory schemes. A beneficiary can receive a legally correct pension whose amount reflects a markedly different contribution base, residence location, or benefit formula. A coherent assessment of portability therefore follows the person across each institutional transition and identifies exactly which property is observed at each point.
The legal analysis uses China-specific instruments in their date sequence. The 2009 employee-transfer measures entered into force on January 1, 2010. Article 19 of the Social Insurance Law sets a general continuity rule when people work across pooling jurisdictions. The 2014 resident-system reform merged the previous rural and urban resident pension arrangements, while the 2014 scheme-connection rules specified movement between resident and employee insurance. Subsequent guidance for new forms of employment addressed hukou restrictions for flexible workers. National employee fund pooling began in 2022 according to the ministry's implementation statement. A September 2024 legislative decision subsequently changed the retirement timeline and scheduled higher contribution-minimum requirements for employee pension claims from 2030.
The evidentiary method combines close reading of full official Chinese legal texts, national and migrant-worker statistical releases, methodological definitions from the International Labour Organization, and critical reading of original academic publisher records. For official legal documents, the analysis reports the operative provision and the effective date where provided. For government statistics, it preserves the precise reporting year, units, and population definition. For research articles, it distinguishes full original legal texts and statistical tables from an accessible original abstract or extended publisher preview. This distinction is important in the recent causal literature: evidence favoring the effect of provincial pooling on migration does not identify the effect of every pension rule on all cross-provincial migrants.
The parent scientific domain is Social Insurance and Public Protection. This dossier examines social risk allocation and benefit continuity within that domain, while giving priority to the mechanics of enrollment and portable accrued rights. It concerns the pension system of mainland China under the instruments cited here; specific Hong Kong, Macao, and Taiwan arrangements and international totalization agreements require separate legal treatment. The inquiry is written as international comparative insurance science because it addresses general empirical questions concerning covered populations, contributory careers, fiscal coordination, and old-age income security through carefully dated national evidence.
Conceptual and Historical Foundations
Pension participation as a stock of administrative statuses
The word 'participation' refers to at least four different observations: a person listed as insured, a person making a current contribution, a person with a preserved but temporarily inactive account, and a person already receiving a benefit. National statistical communication often aggregates categories relevant to fund management, including workers and retirees, while an international active-contributor indicator is designed around a working-age or labor-force denominator. The ILO distinction between contributors and beneficiaries and its technical indicator definitions show why a national administrative enrollment stock must not be described as a percentage of effectively protected mobile workers without matching person-level data. This conceptual distinction controls the interpretation of every aggregate in the dossier.
The legal unit is a participant's pension relationship and recorded history, rather than a generic national insurance card. Article 19 of China's 2018-consolidated Social Insurance Law provides that a basic old-age insurance relationship moves with a person who works across pooling regions, contribution years are accumulated, and the pension at statutory retirement is calculated in segments and paid in a unified manner according to State Council rules. The rule establishes legal continuity; its individual application depends on identifying the relevant pooling jurisdiction, the worker's scheme, the transfer procedure, and the place where the benefit will be claimed. Those subordinate elements were developed in administrative instruments, including Document No. 66 of 2009.
The evolution of resident pensions followed another institutional route. Earlier rural and urban resident arrangements were formally integrated by State Council Document No. 8 of 2014, which described coverage of people generally aged 16 and above, excluding students and people already within employee pension coverage, and tied enrollment to the registered household location. That instrument combined individual payments, collective support where available, and government subsidies, with a basic pension and an individual-account pension component. It gave resident insurance a recognizable lifetime account structure while preserving a central role for registered domicile in entry and in cross-regional transfers during the contribution phase.
The two statutory pathways were then connected through the 2014 intersystem connection measures. These were necessary because a person whose career includes rural residence, urban employment, self-employment, and a later return to the registered community may accumulate rights within separate legal programs. A transfer from one account system to the other is a legally specified conversion, and the conversion rules determine whether previous years qualify toward the employee pension's minimum contribution requirement. The institutional history is therefore best read as successive solutions to distinct coordination problems: cross-province employee transfers, resident system integration, cross-scheme conversion, mobility-related admission, and national funding equalization.
Hukou and place-based administration
Hukou denotes household registration, including its jurisdictional location and registration classification. Its significance in pension governance varies by rule. It can establish where a resident is initially entitled to enroll in resident insurance; it can determine the fallback location of employee pension claims; and it can affect whether flexible workers without a local employment relationship can enter the employee program in the city where they work. Those effects are legally and temporally specific. The 2021 guidance on flexible employment explicitly directs regions to relax hukou restrictions on enrollment at the place of work, while permitting gradual implementation in certain very large cities. The existence of this instruction indicates both a national reform direction and the need to verify local implementation at the observation date.
A central economic consequence is that labor mobility and pension mobility need not coincide. Wage employment may begin in one province, shift between employers, and later return to a home area while an individual's enrollment classification, contribution base, or age-related portability procedure changes. Earlier work by Lin (2015) on pension fragmentation explains why decentralized administration can create incentives and costs distributed unevenly between sending and receiving regions. Its analysis describes a historical institutional setting; later provincial centralization and national pooling must be treated as intervening reforms rather than assumed absent throughout the present.
Four dimensions of continuity
A scientifically useful descriptive distinction separates record continuity, qualifying-period continuity, financial transfer, and benefit continuity. Record continuity asks whether the years and monetary balances are attributable to the same person after a move. Qualifying-period continuity asks which contribution years count toward the pension being claimed. Financial transfer asks which account balances or pooling-fund amounts are moved between institutions. Benefit continuity asks whether a legally eligible person actually receives an appropriate periodic pension in the designated location without interruption. These dimensions come directly from separate legal clauses and observable administrative operations; they are analytical categories for reading the sources, not a newly proposed theory.
Under the employee transfer procedure, accrued employee contribution periods are accumulated and individual-account balances preserved. Under the intersystem rules, the treatment of years differs by transfer direction. Under the resident pension framework, post-claim movement has different consequences from movement during contribution. Economic measurement must therefore retain the original scheme and transaction context. Counting each registered transfer as an equal unit of portable benefit protection would erase the most consequential institutional heterogeneity.
Enrollment Tiers
Employee insurance and resident insurance
China's two dominant basic pension classifications are urban employee basic pension insurance and urban–rural resident basic pension insurance. The employee framework applies to workers within employment relationships and includes routes for certain self-employed and flexible workers under the Social Insurance Law. It relies on employment-related contributions and an earnings-sensitive calculation, together with pooled financing and individual accounts. Resident insurance serves people outside that employee framework and, under State Council Document No. 8, uses flexible annual contribution tiers, an individual account, and a publicly subsidized basic pension. These are different financing and benefit structures even though both are described within national headline basic pension totals.
The distinction matters most for career interruptions. An employee relationship often brings a legally expected employer payment, documented earnings, and administrative wage reporting. A flexible worker may need to organize contributions personally under local administrative rules. A resident participant may pay lower annual amounts and receive a pension partly financed by government support. Therefore a year of recorded participation in one program is not financially or legally equivalent to a year of contributory service under the other. The 2014 connection rules make the unequal treatment explicit rather than leaving it as a merely economic inference.
In official figures, the 2024 national communiqué recorded 534.49 million participants in urban worker basic pension insurance and 538.30 million in urban–rural resident basic pension insurance. The 2025 communiqué recorded 546.80 million and 529.19 million, respectively. Its total for both categories was 1,075.99 million. The increase in employee participation and decline in resident participation can be described directly from the published stocks, but national stock changes alone do not establish the number of individuals who moved from one scheme to the other, the share newly employed, or the number who ceased paying. Such decompositions require linked enrollment and transaction data.
The two 2024 scheme figures sum to 1,072.79 million. The 2025 communiqué's reported increase of 3.16 million from the previous year implies a 2024 total of 1,072.83 million. That difference of 0.04 million is a cross-release reconciliation issue, possibly involving source revisions or category conventions; the inspected communiqués do not themselves document a bridge permitting its definitive explanation. The dossier therefore reports each series exactly as published and excludes the difference from any substantive demographic inference. A putative 2024–2025 net shift computed by subtracting categories is likewise a change in participant stocks, not a demonstrated individual transition flow.
Migrant-worker denominators and the enrollment question
The migrant-worker population is shaped by survey definitions. According to the 2025 migrant-worker survey, the national count was 301.15 million, with 180.06 million working outside their registered township and 67.65 million moving across provinces; the interprovincial component was 37.6 percent of the out-of-town category. The survey defines migrant workers through rural household registration and at least six months of nonagricultural employment locally or away from home during the year. Its counterpart 2024 survey recorded 299.73 million migrants, of whom 68.40 million crossed provincial borders. These are measures of labor movement under a particular statistical convention, not a register of pension transfers or a list of currently uncovered workers.
The 2025 cross-provincial count illustrates the scale of careers potentially touched by the 2009 employee portability rules. It does not imply that 67.65 million people initiated a pension transfer in 2025. Some may have maintained preexisting coverage in the destination, some were covered in resident insurance, some may have become noncontributors, and others may have changed jobs within a multi-year migration episode. The correct administrative outcome denominator would identify persons entering a new pooling jurisdiction with preserved histories, and distinguish those already claiming pensions from those still accumulating contributions. Survey definitions and administrative definitions must be reconciled at the individual level before calculating transfer use.
Socioeconomic selection and scheme choice
Labor contracts, earnings, education, employer type, and local registration all affect which pension route is feasible and attractive. Using China Household Income Project data from 2007 and 2013, Wang (2019) examined migrant pension selection through logit and multinomial probit specifications, finding positive associations of education, income, labor contracts, state enterprise employment, and local hukou with membership and the choice of employee insurance. The publication is valuable because it isolates scheme choice rather than collapsing all participation into a single binary outcome. Its observational design does not identify whether changing hukou alone would causally produce the reported choices; omitted job quality, institutional enforcement, and household resources can influence both hukou status and pension participation.
Selection matters for comparing benefit adequacy later in life. People enrolled as employees may accumulate higher wage-based rights because they have more stable employment and incomes. People with fragmented or lower-paid careers may use resident insurance or switch programs. A comparison of mean pensions between the two categories therefore reflects program rules, work histories, contributions, labor-market sorting, and local transfer decisions. It cannot be interpreted solely as an effect of hukou policy. Gender, age, employment status, and urban or rural labor-market access should be documented before a statistical analysis attributes differences to formal enrollment provisions.
Contribution persistence and the actuarial base
An actuarial account of coverage requires observing payments over time. Define, descriptively, a person's calendar-year contribution record as the sum of months legally recognized under the relevant scheme, with a separate field for the contribution wage base or annual resident contribution tier. An employee who records five years, interrupts payments for three years, and then returns for another ten years has a different qualifying history from a continuously insured worker, even when both appear in the same year-end registration stock. Under the employee transfer measures, employment interruption does not by itself extinguish preserved records; the subsequent adequacy of an eventual pension depends on cumulative recognized periods and benefit rules.
The distinction is especially important as demographic policy changes. The retirement-age decision of 2024 schedules a gradual increase in the employee minimum contribution period from fifteen to twenty years beginning in 2030, in increments of six months per year, alongside gradual statutory retirement-age increases beginning in 2025. The reform's transition and the rules for postponing or otherwise completing required contributions must be applied to actual cohorts, not retroactively inserted into earlier studies or mechanically extended to every resident benefit. A person with intermittent contributions may face a materially different probability of qualifying for an employee pension under a later retirement cohort even if transfer administration becomes faster.
Hukou Mobility
Residence, work, and pension location
Household registration, current employment location, current insurance relationship, and ultimate pension-claiming location can point to four different places. Chinese law addresses this configuration by allowing accumulated employee contributions to follow employment while retaining rules about where an eventual pension will be administered. Article 19 of the Social Insurance Law supplies general continuity, and the 2009 implementation measures make operational distinctions. A migrant's physical arrival in a receiving province does not automatically settle where the final employee pension will be claimed; that determination depends on the relationship's location and accumulated service histories when conditions for benefits are met.
The 2009 employee regulations state that, when the basic pension relationship is located in the hukou province, that province administers benefits. When it is elsewhere and the relevant contribution record reaches ten years, the corresponding province can become the benefit location. If the present province falls short of ten years, the rules examine the previous province with ten or more accumulated years. If none meets that threshold, the relationship and associated funds return to the hukou province for claiming. This mechanism gives legally meaningful weight to residence registration even after many years of work elsewhere.
Another provision regulates new destination accounts for older migrants. In the same 2009 measures, men aged at least fifty and women aged at least forty moving to non-hukou provinces are, subject to specified exceptions, assigned temporary contribution accounts in the new location while preserving the earlier pension relationship. Contributions and interest in the temporary account are subsequently consolidated into the original or benefit-claim location. That rule provides continuity of money and records while limiting the legal form of the newly established relationship. Researchers who code every move as a normal new province-level pension enrollment could misclassify its administrative consequences.
The resident scheme follows another geographic logic. During the contribution phase, a participant whose household registration moves may seek transfer to the new hukou jurisdiction; the full individual-account balance is moved and years are accumulated according to the destination framework. Once resident pension benefits have begun, the pension relationship is not transferred under that clause, irrespective of later hukou movement. A study of transfers therefore needs to segment resident contributors before retirement from resident beneficiaries already in payment, rather than assuming the same portability route applies throughout life.
Migrants with fragmented careers
Consider a hypothetical worker born in a rural-registered county who first contributes as a resident, then works for several years in a coastal province under employee insurance, returns home temporarily, and later works in a second coastal province. This is an explanatory example, not an observed person. Her history can contain resident individual-account contributions, employee contribution months, years in temporary or permanent employee relationships, and periods with no payments. She must eventually determine the employee benefit-claim location under the ten-year and hukou rules before evaluating a possible resident-to-employee or employee-to-resident interface. The 2014 scheme-connection procedure places the location determination ahead of intersystem conversion precisely because the receiving institution is relevant to the ensuing transaction.
In this case, a residence-based classification is an insufficient measure of actual labor integration. A worker may hold a rural hukou and maintain an employee contribution record, or hold a local urban hukou and have no recent contributions. Conversely, the same nominal pension account balance can carry different prospects for periodic benefits depending on recognized employee years and local earnings histories. These distinctions are observable in linked administrative records, but they cannot be established through the single binary question 'Did this person join a pension?' An effective empirical design would record the full employment and registration sequence, contribution dates, wage bases, original location, receiving location, and claims outcome.
Flexible work and the changing enrollment frontier
The legal frontier for flexible workers changed with the 2021 multi-ministry guidance. Its paragraph 8 calls on local governments to remove hukou restrictions for eligible flexible workers seeking employee basic pension or medical insurance where they are employed. It specifically recognizes that certain very large cities might require phased implementation. Paragraph 12 calls for more convenient procedures for contribution payment, rights inquiry, benefit collection, and social insurance transfers. Therefore the correct 2026 research question is whether and to what extent particular jurisdictions implemented the specified measures, and how eligibility actually changed by cohort, employer status, and city, rather than stating that a uniform nationwide relaxation operated identically from the date of the guidance.
The guidance also distinguishes forms of platform-related work in which an employment relationship is legally established from more autonomous forms of work. For the former, employer obligations remain pertinent; for the latter, flexible-employee participation or resident insurance may be the available route. A platform-worker category is therefore not homogeneous in pension financing. Legal classification, contribution burden, and local service accessibility influence persistence even when initial enrollment is technically allowed. Administrative records of applications denied by reason, accepted enrollments, and months paid would be stronger evidence of enforcement than general announcements that enrollment was encouraged.
Mobility as a jointly determined labor and insurance outcome
A household moves for wages, family formation, childcare, housing, health services, and other reasons as well as for pension access. Pension portability can reduce one anticipated cost of moving, but the observed decision emerges from several simultaneous constraints. Earlier institutional analysis by Lin (2015) identified the importance of interregional fund incentives and decentralized governance for migrant entitlements. More recently, Lu, Yang, and Zhong (2026) studied a staggered provincial pooling reform and reported an increase of approximately 4.5–4.6 percentage points in intraprovincial migration, with evidence concerning job matching and provincial outcomes. Its treatment concerns improved portability across cities within provinces over the 2014–2020 observation window. It is therefore informative about one mobility mechanism rather than proof of an equivalent effect for cross-provincial pension transfers or all hukou groups.
The study's reported gains in job matching and income must also be evaluated alongside its statement that life satisfaction did not improve comparably. This difference is consistent with remaining nonpension integration constraints, including residence-linked access to services, but the result does not uniquely isolate a hukou policy effect. The research implication is that mobility and welfare should be measured separately. A reform can influence geographic job matching without delivering equal improvements in pension adequacy, social inclusion, or subjective well-being. Evaluating only the migration rate could overstate the scope of protection delivered by pension administration.
Benefit Portability
Portability within employee insurance
Employee pension portability has an explicit legal structure. The 2009 measures require transfer of the recorded individual-account balance and specify transfer of pooled-fund financing calculated from twelve percent of the individual's post-1998 annual actual contribution wage base, prorated for shorter periods. This financial transfer percentage is an administrative financing rule, not a direct statement that twelve percent of the wage becomes an individual's personal pension wealth. Individual-account assets and pooled-fund financial allocations perform different functions, and neither should be equated without qualification with the present value of future lifetime benefits.
The measures also provide for cumulative recognition of employee contribution years and prohibit ordinary termination and cash-out of the pension relationship before retirement, subject to specified separate rules. During interruption or return migration, the contribution history and individual account are preserved. Administrative processing originally involved successive time limits for new, original, and receiving agencies of up to fifteen working days each in relevant transfer steps. Legal deadlines provide an observable service benchmark; they do not show the fraction of cases completed within that period. Actual performance needs an application-level dataset with timestamped requests, response categories, and reconciled balances.
The ultimate benefit uses locally administered calculation rules, historically involving a retirement location and wage histories under the 2009 employee framework. The place in which a pension is claimed can affect the wage reference and institutional allocation. Thus 'all contribution years are portable' is compatible with 'the resulting pension level is sensitive to the legally determined claiming place.' This is a central difference between a rights ledger, in which years remain recorded, and the actuarial result, in which those years interact with a benefit formula. A study alleging portability failure must specify whether the failure is an erased contribution year, a delayed institutional transfer, a different legally applied formula, or an inability to qualify for payment.
Portability across the two basic pension schemes
The most important asymmetry appears in the 2014 connection measures. A participant satisfying the employee pension contribution requirement may transfer the resident individual-account balance into the employee account, but prior resident contribution years are not added or converted into qualifying employee pension years. In the reverse direction, when an employee transfer proceeds into resident insurance under the applicable eligibility framework, the employee individual-account balance transfers and employee contribution years count as resident contribution years. The rule preserves financial amounts in both directions but does not grant equal qualifying-time conversion.
This asymmetric treatment follows the distinct funding logic of the two programs. Resident contribution years can be based on annual tier payments and public subsidies of different size and incidence from employee payroll contributions. The asymmetry therefore has a legal and fiscal rationale: program membership and account balance remain portable, while the calculation of a qualifying employee career requires employee-recognized service. The distributional effect can nevertheless be substantial. Individuals with several early years of resident insurance and an incomplete later employee record may have less recognized employee tenure than their total number of years paying into public pension arrangements. They can face the choice of additional employee payments or a resident pension outcome where law permits.
The 2014 rules prohibit simultaneous employee and resident basic pension receipt and address duplicate participation periods. These provisions seek to prevent double counting and overlapping benefits. A researcher observing two administrative enrollment entries for one individual should not automatically infer two valid future pensions. Similarly, a successful transfer of the resident account into the employee system does not imply that resident years satisfied the employee minimum contribution threshold. The distinction is legally consequential for estimating future liabilities, measuring retirement poverty risk, and evaluating the probability of a worker with a mixed career meeting qualification criteria.
Residence portability, beneficiary relocation, and national pooling
For resident contributors, the unified resident regulations allow the relationship and individual-account balance to move when hukou changes during the accumulation phase, with contribution years accumulated. Beneficiaries already receiving resident pension benefits are handled differently: the relationship does not transfer under that rule after benefit commencement. This is best understood as a distinction between the jurisdiction managing a new or accumulating relationship and the continuing payment obligation of an established benefit. Research on post-retirement residential moves should accordingly collect payment-responsibility and receipt data rather than counting only transfers of the underlying insurance relationship.
For employee insurance, the move toward provincial centralization and later nationwide pooling changed intergovernmental fund coordination. The official policy history reports provincial pooling of receipts and expenditure by the end of 2020 and national pooling beginning January 2022. The 2022 ministerial statement specifies the purpose as adjusting current interprovincial surpluses and shortfalls and sharing central and local spending responsibilities. This improves the financial capacity to sustain pensions across areas with different aging and employment patterns. It should not be mechanically described as the immediate disappearance of hukou-related benefit-location rules, individual record-transfer procedures, or all subnational differences in pension entitlements.
A public pension system can be nationally coordinated at the fiscal level while retaining person-level qualifying rules and separate local service arrangements. In actuarial terms, intergovernmental cash equalization addresses solvency and liquidity risk across provinces, while portability addresses continuity of individual recognition and benefit claims. The former may support the latter operationally but does not prove every administrative transfer has succeeded. The 2026 provincial-pooling study suggests that more unified local arrangements can relax mobility friction; evaluation of the nationwide pooling stage would require its own post-2022 identification design and outcome series.
Adequacy, pension timing, and lifetime risk
Portability is economically valuable because retirement income is an annuity-like protection against longevity risk and income exhaustion, not simply the repayment of a participant's recorded contributions. The resident benefit described in Document No. 8 of 2014 combines a publicly financed basic amount with an individual-account pension, originally calculated for the latter with a monthly divisor of 139. The document's specific original contribution tiers and benefit rules are historical legal terms; current local rates should be established from dated provincial rules before an actual benefit is computed. Employee pensions use a different wage-related structure, and therefore a raw comparison of account balances is an inadequate comparison of expected retirement income.
The 2024 retirement legislation adds important cohort timing. It gradually raises the statutory retirement age for specified employee groups from January 2025 and schedules a progressive rise in the minimum contribution period required for employee monthly pension entitlement beginning January 2030. The minimum does not increase immediately to twenty years for everyone in 2025, and the rule is not a basis for automatically rewriting the resident scheme's eligibility. An actuarial projection must use the applicable worker category, birth cohort, effective year, planned service history, and the precise transition arrangement. Failure to do so could materially overstate or understate eligibility risk among workers with fragmented migration histories.
A useful descriptive measure of portability quality is the proportion of people whose correctly recorded pre-move qualifying employee periods remain fully recognized after a transfer, reported separately from the proportion of eligible claimants who receive their first payment on time. Another is the difference between legally projected and actually paid monthly benefits for otherwise comparable movers, after applying destination-specific rules. Such outcomes require administrative microdata, wage histories, and lawful comparison groups. The absence of those data in the present dossier prevents numerical estimation of a portability-loss rate. It does not prevent a detailed legal determination of where that loss could arise and what evidence would establish it.
Evidence Verification and Sensitivity
Primary source hierarchy and legal dating
The highest-priority materials for legal propositions are consolidated law and formally issued national instruments. Article 19 in the Social Insurance Law establishes general transfer and accumulated-year principles. The 2009 State Council measures specify cross-province employee machinery and place-of-benefit rules. The two resident harmonization and intersystem linkage instruments specify other legal transactions. The 2021 guidance is a reform instruction whose city-level implementation may vary, while the 2024 legislative decision controls defined later cohorts. The distinction between a statute, a binding detailed rule, a policy implementation statement, and an administrative statistical report is preserved throughout: they establish different kinds of proposition.
Every claim about a rule is indexed to its date and the population it covers. The 2009 employee measures were issued in December 2009 and took effect in January 2010. The intersystem connection instrument took effect in July 2014. The 2021 guidance openly contemplated phased implementation in some megacities. National pooling was introduced in January 2022, while the retirement-age decision was adopted in September 2024 with staged implementation from January 2025 and a separate contribution-period change from 2030. An apparent contradiction between a 2015 paper describing limited portability and a 2026 paper examining improved pooling becomes intelligible once reform timing and geographic unit are specified.
The reviewed legal texts do not provide a complete inventory of all province- and city-specific implementation measures as of November 2026. This limits the dissertation-like claim one could make about uniform operational conditions. The monograph makes national-rule claims where the national primary text supports them and treats local enforcement as an empirical research question. It does not infer a concrete municipal rule from another city's practice or attribute to a particular claimant a specific eligibility outcome without birth date, hukou location, contribution records, and competent agency.
Statistical reconciliation
The two annual national communiqués offer a simple but important audit example. In 2024, the national release listed 534.49 million employee and 538.30 million resident participants. In 2025, the next release listed 546.80 million and 529.19 million, for an announced total of 1,075.99 million. The figures use millions of persons, represent year-end stocks, and should retain the wording of their own publication year. The earlier two scheme figures add to 1,072.79 million; the later communique's total less its reported increase gives 1,072.83 million. The 40,000-person difference is disclosed. It is too small to alter the substantive scale of the system, and it is not explained here by an invented correction factor.
The observed employee stock rose by 12.31 million between the two publications, while the resident stock fell by 9.11 million. These arithmetic differences do not define program-to-program migration. A stock can change because of deaths, retirements, new registrations, employment reclassification, transfers, record cleanups, or changed statistical coverage; the supplied tables do not separately identify these causes. The appropriate test for a claim of transfer-induced enrollment would use a transition matrix linking each participant's previous and subsequent scheme, rather than calculating a transition share from net changes across two unrelated reporting stocks.
The 2024 migrant survey and 2025 migrant survey show why denominators require care. The total increased from 299.73 to 301.15 million, while interprovincial migrant counts changed from 68.40 to 67.65 million. Those observations describe rural-hukou-defined employment geography under the survey conventions. They do not provide the number eligible to transfer employee insurance or the share that actually applied. A pension portability rate with the entire migrant-worker total as denominator would mix people who did not cross the relevant insurance boundary with people who crossed it years earlier. A legally aligned denominator would be the number of insured persons with a qualifying movement between the pertinent administrative jurisdictions during a defined period.
A further denominator problem concerns active contributions. Under ILO monitoring methodology, the share of labor-force members actively contributing and the share of persons past retirement age actually receiving pensions are distinct effective-coverage indicators. China-wide administrative participant figures should be interpreted alongside, and not substituted for, those concepts. The ILO global figures in the report are international aggregates and are not presented here as China's national rates. Without matched active-contributor or pension-recipient denominators for the relevant Chinese cohorts, the dossier cannot claim a verified universal-effective-coverage percentage.
Empirical studies and causal identification
The scholarship provides complementary forms of evidence. Wang (2019) analyzes earlier survey associations between characteristics and employee-versus-rural pension participation. Lin (2015) examines decentralized governance incentives through a political economy lens. Huang and Zhang (2021) study the rollout of China's New Rural Pension Scheme and report effects on several welfare outcomes among eligible older people. The latter is evidence about pension availability and welfare from a historical resident-program rollout, not direct evidence that a transfer mechanism changed migrant mobility. These studies address different treatments and populations and therefore should not be pooled into one undifferentiated estimate of the 'effect of portability.'
The most directly related recent study, Lu, Yang, and Zhong (2026), uses staggered differences in provincial pension pooling and longitudinal information from the 2014–2020 period to analyze migration across cities within a province. Its publisher record reports a migration effect around 4.5–4.6 percentage points, with mechanisms connected to job matching. The reported result is specific to that reform, geography, period, sample, and identifying assumptions. Policy timing may correlate with provincial economic trends or other local changes; staggered-policy studies need transparent counterfactual assumptions, pretrend diagnostics, appropriate comparison cohorts, and robustness to treatment timing. The present dossier critically uses the original reported result and does not claim to have independently replicated the data, code, or standard errors.
One important external-validity test is whether the reform's effects are strongest for low-skilled workers, as reported in the study. If so, average effects may depend on the skill composition of the mobile population. Effects estimated for intraprovincial moves need not apply to cross-provincial moves, where different benefit-location criteria may be decisive. A second test is whether higher employment earnings translate into improved pension qualifying years and later actual benefit receipt. The study's immediate labor-market outcomes are distinct from such longitudinal pension outcomes. A third test concerns simultaneous hukou reform or welfare-service changes, which might affect migrants independently of pension pooling.
Sensitivity to cohort, location, and definition
At minimum, empirical results should be recomputed across four boundaries: resident-versus-employee scheme, intra- versus interprovincial movement, hukou registration versus actual residence, and contribution accumulation versus benefit receipt. Each boundary corresponds to distinct law or statistical data. Within employee insurance, separate normal account transfers from temporary accounts triggered by age and location under the 2009 measures. Across schemes, mark the direction of conversion because the 2014 year-credit rules are asymmetric. Across cohorts, apply the future contribution-minimum changes of the retirement transition only where legally relevant.
Missing administrative data may be nonrandom. Migrants with fragmented work histories may be less likely to retain consistent local records or participate in household surveys; temporary workers may leave a city before a follow-up interview; informal workers may have periods without employer reporting. If missingness correlates with tenure, hukou, sex, or job quality, simply restricting an analysis to complete records can bias estimated portability success upward. Robustness analysis should describe missing-record rates, matching procedures, pre-transfer characteristics, and the difference between unresolved cases and legally denied transfers. It should also distinguish person identifiers from household identifiers and account identifiers to avoid double counting.
An actuarial sensitivity exercise would project benefits across documented wage bases, future contribution years, permitted retirement dates, and the legally determined claim location. These are scenario calculations rather than new forecasts in the present monograph. To make a numerical projection valid, the researcher must first obtain dated wage indices, actual local benefit formula parameters, the applicable retirement coefficient, and the worker's verified contribution ledger. Absent those inputs, a precise replacement rate would create the appearance of measurement without a source. This dossier confines its numerical conclusions to verifiable published aggregates and clearly described published study estimates.
Source-access controls and bibliographic integrity
The legal and statistical basis was checked in original official web publications. The 2019 Wang article record was inspected on its original J-STAGE publisher page. For Lin's 2015 article, the 2026 pooling study, and the rural-pension experiment, source identities and substantive statements were confirmed from original indexed publisher records and available abstracts or excerpts; unrestricted direct full-text readback was not obtained for every article. The evidentiary ledger records these differences so that an abstract-based conclusion is not represented as an independently reanalyzed full article. The 2026 study's original publisher page returned a direct-fetch restriction even though extensive publisher-indexed text was available.
There is no quantitative meta-analysis or independent causal replication in the present work. The strongest original contribution is a disciplined comparison between the legal transactions and the empirical quantities used to describe them, a transparent reconciliation of official statistics, and an agenda specifying what person-level evidence would decide remaining questions. Citation links have been placed on short contextual titles and directly supported phrases. The complete bibliography records every substantive external source used; source identifiers are retained only inside the machine-readable evidence ledger.
Scientific Synthesis and Research Agenda
What the legal sequence establishes
China's policy architecture shows a progressive differentiation of problems once gathered under the phrase 'pension portability.' The 2009 employee measures made cross-province record and financing transfers operational and assigned benefit-claim locations through hukou and contribution-history rules. The 2014 resident unification created a nationwide resident framework with residence-linked entry and transfer during accrual. The 2014 scheme connection defined asymmetric time-credit conversion between employee and resident insurance. The 2021 guidance sought to open employment-location enrollment for flexible workers who previously faced local registration barriers. Finally, national fund pooling from 2022 strengthened redistribution of cash resources across provinces. These measures address connected problems, but no individual measure subsumes the others.
The result is a layered public insurance landscape in which a worker can have a valid recorded right but face a different contribution-year test or claiming location after moving. The employee system treats accumulated years as an important portable entitlement, and the legal rule intends continuity across pooling regions. The resident system protects a different class of contribution and subsidy histories. The interface between them transfers balances and applies distinct time-credit rules. The relevant question for public protection is therefore which right the individual has at each stage and how it is translated into a periodic pension at the legally designated time and place.
The 2025 participation figures demonstrate the population scale of this architecture, and the migrant survey shows the geographic scale of relevant labor movements. They do not establish an actual national portability success rate. The 2026 empirical portability study provides more specific evidence that reducing provincial administrative fragmentation can increase migration across cities within provinces. Its identification scope supports the economic importance of the mechanism but does not eliminate the need for evidence on cross-province mobility, the hukou-based retirement location, and benefit payment outcomes. A coherent scientific conclusion accepts these findings together without extending any source beyond its observed population.
Analytical implications for equity and fiscal balance
The transfer of an individual-account balance represents a different transaction from the movement of an actuarially sufficient amount to fund a lifetime pension. The 2009 rule's pooled-fund allocation and the subsequent national equalization sequence are financial instruments for coordinating liabilities across administrative regions. Neither constitutes an individual marketable pension asset. The government remains responsible for benefit rules, payment administration, and, where applicable, financial adjustment. Measurement of public pension obligations should therefore distinguish intergovernmental fund transfers, individual account balances, recognized service years, and the present value of prospective benefits.
The scheme-conversion asymmetry has direct distributional consequences. Workers whose career includes many resident-program years followed by an incomplete employee history may carry balances forward without equivalent employee qualifying time. More mobile workers with stable employee coverage can benefit from cumulative employee service periods. If both groups are described as covered in national enrollment counts, that statistical sameness conceals considerable differences in prospective old-age income. The migrant scheme-choice evidence is consistent with the importance of labor contracts and income for access to the employee pathway; determining a causal hukou contribution requires more demanding data and identification.
The rural pension rollout evidence additionally reminds us that even comparatively small publicly supported pensions can change living conditions and labor allocation for eligible older people. Adequacy and portability therefore interact but remain separately measurable. A preserved low-benefit resident entitlement can provide meaningful income support while remaining inadequate relative to needs, and a fully recognized employee career can produce different outcomes under different contribution bases. International monitoring as explained by the ILO pension chapter makes entitlement, effective contribution, and pension receipt distinct outcomes for precisely this reason.
Priorities for a research-ready data architecture
The first priority is a deidentified longitudinal person-level history with unique IDs across hukou location, actual residence, employment status, employer classification, pension scheme, contribution months, contribution wage or resident tier, account balances, and local government jurisdiction. Each move should have a date, source location, receiving location, temporary-account flag, and the legal regime applicable at that date. The administrative transfer record should contain application date, acceptance or refusal, reason code, fund amount, credited periods, completion date, and later claim adjudication. Without these variables, the difference between nominal portability and actual transfer completion remains obscured.
The second priority is a statutory concordance of national and local legal instruments, retaining their promulgation dates, effective periods, affected categories, and supersession relationships. This is essential to evaluate the 2021 flexible-employment guidance and the 2025–2030 retirement transition. A change in a city's enrollment practice, a move of a worker, and a contribution-year threshold may occur in the same observation window. Correctly coding the rule in effect for each event is necessary before attributing a subsequent outcome to pension portability.
The third priority is comparative outcome evaluation. For cross-provincial employee moves, compare legally transferable records with records recognized after the move and measure successful pension claim location and benefit receipt. For resident transfers, compare pre- and post-hukou relocation during accrual, separately from people already receiving resident pensions. For intersystem transfers, evaluate the employee-to-resident and resident-to-employee directions separately. For newly eligible flexible workers, measure approved enrollment and contribution persistence, including initial denials and appeals. Finally, connect pension results with labor earnings, job mobility, household consumption, and later retirement poverty using transparent privacy and identification safeguards.
A fourth priority is fiscal incidence. Sending and receiving jurisdictions can experience different short-run and long-run pension fund implications because contribution revenues, accrued rights, and beneficiaries are geographically separated. The transition from local pooling toward national equalization changes the intergovernmental distribution of these risks, as described in the 2022 launch account. Future actuarial study should calculate how much of an individual's expected pension expenditure is financed by current contributions, transfers, subsidies, and pooled reserves under each regime. It should avoid treating an administrative transfer amount as the full actuarial liability unless a defensible valuation establishes that relationship.
Empirical propositions for future testing
The first testable proposition is that, conditional on prior contributions, legally portable employee records should suffer fewer missing-period discrepancies after successful computerized transfer than under older fragmented administrative arrangements. The relevant outcome is an audited ledger discrepancy, not a survey respondent's subjective impression. Testing requires matched pre- and post-transfer data and a control for changes in the reporting system. A second proposition is that local implementation of relaxed hukou restrictions should increase enrollment among eligible nonlocal flexible workers, with effects on actual contribution persistence differing by income, contribution burden, and city. A staggered implementation design could be informative if policy timing and parallel-trend assumptions are supported.
A third proposition is that intersystem transitions should show different distributions of qualifying employee years according to conversion direction, as the 2014 rules predict. Here the legal mechanism is established, and the empirical question concerns frequency and ultimate benefit adequacy. A fourth proposition is that improvements in intraprovincial portability will have a different labor-market effect from the relaxation of cross-provincial hukou-linked benefit-location constraints. The former has some published support from Lu, Yang, and Zhong (2026), while the latter needs separate evidence. These are research hypotheses grounded in the analyzed laws and observations; they are not asserted as new doctrines, confirmed effects, or a proprietary conceptual system.
The final proposed comparison concerns older cohorts. Workers approaching retirement after 2030 may need more recognized employee contribution years than comparable earlier cohorts. The staged statutory increase provides clear legal timing, but estimating its effect on pension claiming will require cohort-specific administrative outcomes and consideration of adjustments to retirement ages. If reforms simplify transfers while the minimum contribution period rises, a study may find improved record continuity but heterogeneous entitlement attainment. This interaction is a legitimate scientific question whose answer cannot be read directly from aggregate participation statistics.
Conclusion
China's basic pension arrangements have moved through several legally distinct reforms: employee pension relationship transfers from 2010, integrated urban–rural resident insurance in 2014, explicit intersystem conversion in the same year, direction to relax hukou limits for flexible employment in 2021, and nationwide employee fund pooling from 2022. The retirement-age and contribution-minimum transitions approved in 2024 introduce further cohort-specific requirements. A correct appraisal of these reforms requires attention to their dates, statutory populations, and administrative transaction types.
The primary legal evidence establishes that employee contribution years can be combined across pooling regions and that benefit-claim location still depends on legally specified relationships among hukou, contribution history, and qualifying location. Resident insurance permits account transfer during the contribution phase when hukou relocates, but its post-benefit rule differs. Cross-scheme conversion is asymmetric: resident contribution years do not turn into employee qualifying years when a resident account is transferred upward into the employee system, whereas recognized employee years can contribute to resident insurance in the reverse direction under the 2014 rule. These distinctions directly affect the chances of qualifying for a retirement pension after a fragmented career.
The official 2025 administrative participation total of 1,075.99 million is evidence of system scale, not a national percentage of actively insured migrants or of fully transferable pensions. Published statistics for the two schemes and for migrant workers use different denominators and are not a person-level transfer register. The reported 2024–2025 figures also contain a small, explicitly disclosed cross-release arithmetic discrepancy. These issues affect measurement discipline rather than the legal analysis. International monitoring similarly separates active contributors, legally covered people, and actual pension recipients.
Peer-reviewed research supports the importance of institutional variation in scheme choice and the potential for improved provincial portability to facilitate intraprovincial labor mobility, with the 2026 study reporting an effect of roughly 4.5–4.6 percentage points in its defined setting. That effect cannot automatically be transported to cross-provincial moves, later cohorts, or actual pension adequacy. The original contribution of this monograph is an integrated, source-bounded interpretation of the distinct legal, demographic, fiscal, and empirical meanings of pension portability and a specified longitudinal research program capable of evaluating them. The resulting conclusion is substantive: effective old-age protection for mobile workers depends on the combined preservation, recognition, conversion, and eventual payment of their accrued rights across a changing labor career.
References
Standing Committee of the National People's Congress. (2010; amended 2018). 中华人民共和国社会保险法 [Social Insurance Law of the People's Republic of China], especially Articles 10–22 and 58. Official consolidated Chinese text, State Administration for Market Regulation republication. Original source.
General Office of the State Council. (2009). 国办发〔2009〕66号: 城镇企业职工基本养老保险关系转移接续暂行办法 [Interim Measures for Transfer and Continuation of Basic Old-Age Insurance Relationships of Urban Enterprise Employees]. Issued December 28; effective January 1, 2010. Official Beijing government reproduction. Original source.
State Council. (2014). 国发〔2014〕8号: 国务院关于建立统一的城乡居民基本养老保险制度的意见 [Opinions on Establishing a Unified Basic Old-Age Insurance System for Urban and Rural Residents]. February 21. Ministry of Finance official text. Original source.
Ministry of Human Resources and Social Security and Ministry of Finance. (2014). 人社部发〔2014〕17号: 城乡养老保险制度衔接暂行办法 [Interim Measures for Connecting Urban Employee and Urban–Rural Resident Pension Systems]. February 24; effective July 1, 2014. Official Ministry of Finance text. Original source.
Ministry of Human Resources and Social Security et al. (2021). 人社部发〔2021〕56号: 关于维护新就业形态劳动者劳动保障权益的指导意见 [Guiding Opinions on Protecting the Labor and Social Security Rights of Workers in New Forms of Employment]. July 16, 2021. Official Ministry of Emergency Management text. Original source.
Ministry of Human Resources and Social Security. (2022). 企业职工基本养老保险全国统筹启动实施 [Launch of National Pooling for Enterprise Employees' Basic Old-Age Insurance]. February 23. Original source.
Ministry of Human Resources and Social Security / China Employment Network. (2024). 社会保障制度稳健运行 [The Social Security System Operates Steadily]. July 9. Original source.
Standing Committee of the National People's Congress. (2024). 全国人民代表大会常务委员会关于实施渐进式延迟法定退休年龄的决定 [Decision on Implementing a Gradual Delay of Statutory Retirement Ages]. September 13, 2024; implementation January 1, 2025. Official reproduced decision and measures. Original source.
National Bureau of Statistics of China. (2025). Statistical Communiqué of the People's Republic of China on the 2024 National Economic and Social Development. February 28, 2025. Original source.
National Bureau of Statistics of China. (2026). Statistical Communiqué of the People's Republic of China on the 2025 National Economic and Social Development. February 28, 2026. Original source.
National Bureau of Statistics of China. (2025). 2024年农民工监测调查报告 [2024 Migrant Worker Monitoring Survey Report]. April 30, 2025. Original source.
National Bureau of Statistics of China. (2026). 2025年农民工监测调查报告 [2025 Migrant Worker Monitoring Survey Report]. April 30, 2026. Original source.
Wang, W. (2019). Analysis of Chinese migrant worker’s pension participates: Choice between “Worker’s Basic Pension Insurance” and “New Rural Social Pension Insurance.” Journal of Chinese Economy & Management Studies, 3(2), 37–54. https://doi.org/10.50865/jcems.3.2_37. Original source.
Lin, J. (2015). Labor mobility without pension portability: Migrant workers’ endangered pension entitlement in China. Asian Social Work and Policy Review, 9(3), 269–281. https://doi.org/10.1111/aswp.12067. Original source.
Lu, M., Yang, X., & Zhong, H. (2026). Pension portability, labor mobility, and regional convergence: Evidence from China. Journal of Asian Economics, 104, Article 102158. https://doi.org/10.1016/j.asieco.2026.102158. Original source.
Huang, W., & Zhang, C. (2021). The power of social pensions: Evidence from China’s New Rural Pension Scheme. American Economic Journal: Applied Economics, 13(2), 179–205. https://doi.org/10.1257/app.20170789. Original source.
International Labour Organization. (2024). World Social Protection Report 2024–26: Universal Social Protection for Climate Action and a Just Transition, Chapter 4, section 4.3, pensions and other non-health benefits. Original source.
International Labour Organization. (2024). World Social Protection Report 2024–26, Annex: Definitions, data and estimation methods for SDG indicator 1.3.1. Original source.