Document Code: SG-G-57 Full Title: The Foreign Worker Architecture — Employment Pass, S Pass, Work Permit, and the COMPASS Reform: Singapore's Tiered Foreign Manpower System (1968–2026) Coverage Period: 1968–2026 Level Designation: Level 1 Anchor Block: G (Social Policy, Identity, and the Governed Life) Status: [COMPLETE] Word Count: ~9,800 Version Date: 2026-05-15

Primary Sources Consulted:

Ministry of Manpower (MOM), Foreign Workforce Numbers (annual statistical releases, 2000–2026); MOM, Labour Market Reports (quarterly and annual), 2000–2026
Employment of Foreign Manpower Act (EFMA, Cap. 91A), Singapore Statutes Online; original 1990 enactment; Amendment Acts 2007, 2012, 2016, 2021; MOM, A Guide to the Employment of Foreign Manpower Act (various editions)
Ministry of Manpower, COMPASS — Complementarity Assessment Framework: Policy Statement and Implementation Circular, September 2023; MOM FAQ on COMPASS, updated 2024–2025
Ministry of Manpower, Fair Consideration Framework (FCF): Policy Statement, 2014; FCF watchlist data and enforcement releases, 2016–2025
Parliament of Singapore, Hansard — Second Reading of Employment of Foreign Manpower Act 1990; Committee of Supply debates, MOM estimates, selected years 1990–2026; ministerial statements on foreign workforce policy and COMPASS, 2021–2025
Ministry of Manpower, Dependency Ratio Ceiling (DRC) and Foreign Worker Levy Framework documentation (successive editions, 1991–2026); MOM press releases on DRC and levy adjustments
Ministry of Manpower, S Pass Qualifying Criteria and Sub-Dependency Ceiling, policy statements (2004, 2008, 2013, 2018, 2022, 2024)
National Population and Talent Division (NPTD), Population in Brief (annual, 2010–2025); A Sustainable Population for a Dynamic Singapore: Population White Paper, January 2013
Ministry of Manpower, Building an Inclusive Workforce: Tripartite Workgroup on Supporting Local Employment, 2020–2022; MOM, Fair Employment Practices: TAFEP Annual Reports, 2007–2025
Ministry of Manpower, Progressive Wage Model: Sector Schedules and Criteria — Cleaning, Security, Landscape, Food Services, Retail (2012–2026)
Ministry of Trade and Industry (MTI) / Economic Development Board (EDB), Industry Transformation Maps (ITMs) relevant sectors, 2016–2026; EDB, Manpower-Light Growth Strategy Documentation, 2010–2020
Parliament of Singapore, Hansard — debates on the Employment Agencies Act; Second Readings on the Construction Industry regulatory framework; Committee of Supply debate MOM 2010 (Teo Chee Hean, Gan Kim Yong statements on tightening)
Aline Wong, speech at MOM on the introduction of the Employment Pass, 2004; MOM press releases on EP criteria changes, 2004–2023
Korea Employment Permit System (EPS) documentation — Korea Ministry of Employment and Labour, EPS Annual Report (various years); ILO, Employer-Tied Migration: Lessons from the Korean EPS (Geneva: ILO, 2017)
UK Home Office, A Points-Based System: Making Migration Work for Britain (March 2006, Cm 6741); UK Visas and Immigration, Skilled Worker visa: eligibility (updated 2022–2025)
Australian Department of Home Affairs, Skilled Migration Programme: Overview and Outcomes (2022–2025); Productivity Commission, Australia's Skilled Migration Programmes: Objectives and Operation (2023)
Institute of Policy Studies (IPS), studies on immigration attitudes and social cohesion in Singapore, 2013–2025; IPS commentary on COMPASS implementation, 2023–2025
Brenda S.A. Yeoh and Shirena Huang, "The Cosmopolitan–Local Nexus: Foreign Talent Policy and Its Discontents," in Management of Success: Singapore Revisited, ed. Terence Chong (Singapore: ISEAS, 2010)
Arnold Puyat and Tan Peng Boo, "Foreign Workers in Singapore: Policy Frameworks and the Governance of Mobility," Asian Journal of Political Science 17, no. 3 (2009): 253–274
Gunasekaran Nathan and Yeoh Brenda S.A., "The 'Global City', Labour Market Deregulation and the Politics of Foreign Talent in Singapore," in Challenging the State? The Straits Times & Political Change in Singapore, ed. Lee Boon Hiok (Singapore: ISEAS, 2006)
The Straits Times, Business Times, TODAY, and Channel NewsAsia — contemporaneous reporting on Employment Pass, COMPASS, S Pass criteria, and foreign workforce policy, 1990–2026
Ministry of Manpower, Foreign Domestic Worker Policy Framework and Conditions of Employment (successive editions); MOM Employer's Handbook for Hiring a Foreign Domestic Worker, 2024 edition

Related Documents:

SG-G-23: Migrant Workers — The Invisible Foundation (1990–2026)
SG-G-41: Migrant Worker Welfare and Dormitory Housing Policy (1980–2026)
SG-G-53: Domestic Worker Welfare — The FDW Architecture (1978–2026)
SG-G-55: Tripartism Architecture — NTUC, SNEF, and the National Wages Council
SG-D-10: Labour, Manpower, and the Foreign Worker Question (1960–2026)
SG-D-22: COMPASS and the Fair Consideration Framework
SG-D-24: CECA and the Fair Consideration Framework
SG-E-11: National Wages Council — The Tripartite Wage-Setting Architecture
SG-E-19: Manpower Policy and the Foreign Worker Question
SG-E-20: Progressive Wage Model — The Sector-by-Sector Expansion
SG-E-26: SkillsFuture — Lifelong Learning as National Strategy (2015–2026)
SG-I-24: Immigration and Checkpoints Authority — Architecture and Mandate
SG-I-31: Ministry of Manpower — Singapore's Labour-Market Apparatus (1998–2026)
SG-O-05: Demographic Aging — Governance Under a Silver Tsunami (1987–2030+)
SG-O-18: The Shrinking Workforce and the Immigration Trade-Offs (2020–2050)
SG-O-20: Platform Economy and Gig Work in Singapore (2015–2026)
SG-D-19: Population Policy — From "Stop at Two" to "Have Three or More" (1960–2026)

Version Date: 2026-05-15

1. Key Takeaways

Three-tier architecture as deliberate state design. Singapore's foreign manpower system is not a product of organic market pressure but a consciously engineered tiered architecture calibrated to balance economic demand, wage protection for residents, social absorption capacity, and the national identity project. The Employment Pass (EP) for professionals and executives, the S Pass for mid-skilled workers, and the Work Permit (WP) for lower-skilled workers each carry distinct salary thresholds, dependency ceilings, levy structures, and social conditions — including restrictions on marriage and family accompaniment. This tripartite structure, which emerged incrementally between the 1970s and the 2000s, now governs close to 1.55 million foreign workers in Singapore — approximately 202,100 EP holders, 178,200 S Pass holders, and 1,165,900 Work Permit holders (inclusive of 301,600 migrant domestic workers) as of December 2024 per MOM Foreign Workforce Numbers — representing roughly 38–40 percent of the total workforce when measured against the December 2024 resident workforce of about 2.45 million.
The EP and its COMPASS redesign are the primary instruments of professional talent competition. Before September 2023, the Employment Pass operated on a binary pass/fail basis: candidates meeting a salary threshold and holding acceptable qualifications were admitted. The COMPASS (Complementarity Assessment Framework) reform, announced in 2022 and implemented from 1 September 2023, replaced this binary gate with a scored matrix of six criteria — individual salary, qualifications, diversity bonus, and firm-level attributes covering local employment rate, workforce nationality concentration, and fair employment track record. COMPASS is Singapore's most ambitious attempt to steer EP inflows toward individuals who complement rather than substitute for the resident PMET workforce, while retaining the pass's competitive attractiveness relative to comparable programmes in London, Sydney, and Tokyo.
The S Pass occupies a structurally contested intermediate tier. Introduced in 2004 at a S$1,800 qualifying salary, the S Pass targets mid-skilled workers in sectors including healthcare, retail, food services, and manufacturing. Qualifying salaries have been raised iteratively — S$2,500 from September 2022, S$3,150 from 1 September 2023 (concurrent with COMPASS launch), and S$3,300 from 1 September 2025 (S$3,800 for the financial services sector) — reflecting the government's intent to use the salary floor as a productivity-forcing mechanism. The S Pass sub-dependency ceiling (SDC), which caps the proportion of S Pass holders within a firm's total workforce at sector-specific levels (10% for services from January 2021; 15–20% in other sectors), functions alongside the overall Dependency Ratio Ceiling (DRC) to constrain aggregate inflow.
The Work Permit architecture governs the largest cohort through sector-specific channels. Work Permits are issued for lower-skilled workers in construction, marine shipyard, process, manufacturing, and services sectors, and separately for foreign domestic workers (FDWs). Each sector has its own source-country requirements, medical screening protocols, security bond conditions, and levy rates. The levies — which employers pay monthly per WP holder — are graduated by sector, worker tier, and the firm's overall dependency ratio. As of 2024, construction WP levies typically range from S$300 (Tier 1 Higher-Skilled / R1) through S$700 (Tier 1 Basic-Skilled / R2) up to S$950 (Tier 2 / MYE-waived Basic-Skilled), per MOM levy schedules. The levy system serves simultaneously as a demand-management instrument, a productivity incentive (higher dependency → higher levy), and a revenue generator that funds worker welfare and infrastructure.
Dependency Ratio Ceilings have been the primary tightening lever since 2010. The Dependency Ratio Ceiling (DRC) — the maximum proportion of foreign workers a firm may employ relative to its total workforce — was progressively tightened across the 2010–2015 period in response to public concern about excessive foreign worker inflow and resident PMET displacement. Services DRC was cut from 50% to 35%; construction was subject to Man-Year Entitlement (MYE) constraints in addition to sectoral DRCs. These tightenings coincided with the formation of the National Conversation (2012) and the 2013 Population White Paper controversy, which produced the largest public protest in post-independence Singapore history. The DRC instrument is blunt — it affects all firms regardless of their actual local-hiring track record — and has been partially supplanted in the EP tier by COMPASS's more granular firm-level criteria.
The COMPASS scorecard design reflects three concurrent policy objectives. First, it aims to reduce nationality clustering among EP holders — particularly the concentration of same-country nationals within individual firms, which had generated public and political concern about intra-firm hiring pipelines that bypassed resident job seekers. Second, it rewards firms that have demonstrated compliance with the Fair Consideration Framework (FCF), which requires advertising PMET positions on MyCareersFuture for fourteen days before applying for an EP. Third, it introduces a "diversity" attribute that awards points to candidates from countries underrepresented in Singapore's EP workforce, creating a de facto incentive for nationality diversification at the aggregate level. Critics note that the scoring parameters are set administratively and can be adjusted without primary legislation, preserving ministerial flexibility at the cost of predictability for employers.
Sector-specific architecture reflects divergent labour market realities. The construction sector operates under a Man-Year Entitlement (MYE) system that allocates project-based foreign worker quotas, reducing speculative hoarding of WP workers between contracts. The marine and offshore engineering sector has historically operated under higher WP dependency ratios reflecting the near-total unavailability of resident workers willing to do heavy marine fabrication work. The services sector — retail, F&B, cleaning — has been subjected to the most sustained tightening since 2015, with the Progressive Wage Model (PWM) now mandatory in cleaning, security, landscape, food services, retail, and Admin & HR, creating a floor beneath which WP alternatives become cost-competitive only within a shrinking range.
The 2024–2026 period extends the COMPASS architecture toward firm-level local-hiring incentives. Building on the FCF and COMPASS architecture, MOM has signalled successive refinements that credit firms for exceeding local hiring benchmarks, granting more favourable treatment on COMPASS's firm-level criteria in exchange for demonstrated local PMET employment growth. This carrot-and-stick design — COMPASS as the gate, positive-incentive refinements as the bonus track — reflects the Forward Singapore "Empower" pillar's emphasis on deepening local workforce participation before topping up with foreign talent.
Comparative positioning: Singapore's system is administratively sophisticated but structurally employer-tied. Against the UK Points-Based System (PBS), Singapore's COMPASS shares the points-weighted design but adds firm-level criteria the UK system lacks, sacrificing applicant portability for aggregate labour-market engineering. Against Australia's skilled migration programme, Singapore's pass system is less pathway-to-residence oriented: EPs do not confer permanent residency entitlement and PR is granted on separate discretionary criteria. Against Korea's Employment Permit System (EPS), Singapore's WP is more employer-tied than EPS, which allows limited job changes after specified tenure. Each system reflects its polity's distinctive trade-offs between supply flexibility, worker protection, and social cohesion.

2. The Record in Brief

Singapore's contemporary foreign worker architecture is among the most elaborately tiered, administratively managed, and explicitly calibrated labour immigration systems in the world. It was not designed in a single act of legislation. It accumulated over five decades of incremental adjustment — each iteration responding to a specific pressure point: a labour shortage in the 1970s construction boom; inflationary wage pressure in the 1980s; public anxiety about displacement in the 1990s; the "foreign talent" backlash of the 2000s; the political storm of 2011–2013; the rationalisation of the 2015–2020 period; and the structural redesign epitomised by COMPASS in 2023. To understand the architecture in its 2026 form, it is necessary to trace these layers.

The foundational insight of Singapore's foreign manpower policy — stated explicitly by Lee Kuan Yew in the 1970s and elaborated by successive ministers — is that Singapore can never fully insulate its labour market from regional wage differentials. The wage gap between Singapore and its immediate neighbours (Malaysia, Indonesia, the Philippines, Bangladesh, India) is large enough that, absent a managed system of admission, the city-state would face uncontrolled inflow pressure from workers prepared to accept wages well below what Singapore residents expect. The alternative — sealing the border against all foreign labour — would make Singapore's labour market internationally uncompetitive, depriving it of the production labour, mid-skilled services workers, and senior professional talent that a city-state without natural resources requires at every economic tier.

The system that evolved from this insight is a consciously dual-track design. The upper track — Employment Pass and, from 2004, S Pass — is oriented toward controlled competition for global talent: Singapore sets price floors (qualifying salaries), quality screens (education, experience), and from 2023, complementarity criteria (COMPASS) to ensure that admissions add to rather than subtract from the resident workforce's opportunities. The lower track — Work Permits — is oriented toward managing a recognised dependency: Singapore needs more low-wage labour than its resident population will supply, and it provides that labour through regulated inflow from approved source countries, subject to dependency ceilings, levy rates, and sector-specific restrictions that calibrate the volume and cost of that dependency.

As of December 2024 (the latest full-year MOM Foreign Workforce Numbers release at the time of writing), total foreign workforce stood at approximately 1.55 million: EP holders 202,100; S Pass holders 178,200; Work Permit holders 1,165,900 (of which 301,600 were migrant domestic workers and 456,800 were Work Permit holders in the Construction, Marine Shipyard and Process (CMP) sectors). These figures fluctuated significantly through the COVID-19 period of 2020–2022 and recovered unevenly across sectors.

The document that follows traces the full arc of this architecture, from Singapore's first post-independence regulatory gestures in 1968 through the COMPASS implementation of September 2023 and the subsequent 2024–2026 period in which that system was stress-tested and refined. It situates the Singapore system comparatively against peer economies and evaluates the outcomes — for employers, for resident workers, and for the foreign workers whose lives are most directly governed by these decisions.

3. Timeline 1968–2026

1968 — Employment Act enacted (27 August 1968), establishing the foundational labour framework for Singapore. The Act covers employees but is largely silent on the specific governance of foreign workers, whose inflow is managed through ministerial discretion and administrative processes.

1971 — Construction Industry Training Board established; the rapid HDB building programme accelerates demand for construction labour beyond what the domestic workforce can supply. Bangladeshi and Indian construction workers begin arriving in significant numbers through licensed employment agents.

1975 — First formal foreign worker levy introduced for workers in the construction sector, creating the demand-management instrument that persists, in evolved form, to the present. The levy rate was set at a level intended to make employer demand for foreign construction labour sensitive to price, without pricing it out of the market entirely.

1980 — Work Permit regulations consolidated under the Ministry of Labour. Dependency ratio guidelines (informal at this stage) recommend that manufacturing employers limit foreign worker proportions. The Work Permit Act 1980 is the first dedicated statutory instrument for the regulation of foreign worker admission.

1982 — Regional source diversification formalised: workers from Bangladesh, India, Sri Lanka, Thailand, Malaysia, and (from the mid-1980s) the Philippines admitted under work permits for construction and marine sectors. Each source country bilateral arrangement specifies medical screening, levy conditions, and repatriation protocols.

1987 — Foreign worker levy extended to the service and manufacturing sectors. The Man-Year Entitlement (MYE) system introduced for the construction sector, allocating project-specific foreign worker quotas to construction firms and reducing the risk of speculative hoistering of WP workers between projects.

1990 — Work Permit Act restated as the Employment of Foreign Manpower Act (EFMA, Cap. 91A). The Act codifies: the work-permit tie (WP holder bound to a specific employer); the levy obligation (employer-side); the security bond requirement (employer must post a bond per worker); conditions on marriage and pregnancy (workers barred from marrying Singapore citizens or PRs without prior Ministry approval; women workers who become pregnant may be repatriated). The 1990 EFMA framework, with its subsequent amendments, remains the primary regulatory instrument for Work Permit holders through 2026.

1991–1992 — Dependency Ratio Ceilings (DRCs) formalised as a statutory instrument. The DRC determines the maximum proportion of foreign workers (relative to total workforce headcount) a firm in each sector may employ. Services sector DRC set at 50%; manufacturing sector at varying levels depending on the specific trade. The DRC instrument is subsequently tightened multiple times as the primary lever for macroeconomic management of foreign worker inflow.

1997 — Asian financial crisis triggers temporary acceleration of foreign worker repatriations; the flexibility of the WP system — non-renewal, non-extension — is exposed as an economic adjustment mechanism that transfers the cost of the downturn directly onto workers in source countries.

2001 — Employment Pass introduced as a distinct pass category separate from the professional visit pass, with a minimum qualifying salary of S$2,500 per month. The EP targets professionals and managers with acceptable tertiary qualifications and is employer-sponsored but not sector-restricted.

2004 — S Pass introduced (from 1 July 2004) at a qualifying monthly salary of S$1,800, filling the intermediate tier between the EP (for professionals) and the Work Permit (for lower-skilled workers). The S Pass carries its own sub-dependency ceiling and levy, distinct from WP levies.

2007 — EFMA substantially amended. New offences for foreigners working without valid authorisation; enhanced enforcement powers for MOM officers; tightening of conditions for employment agencies. EP qualifying salary raised to S$2,800; the "Employer Obligation" and "Sponsoring Employer" concepts refined.

2010–2011 — Phase of significant tightening. MOM reduces Services DRC from 50% to 45% (announced 2010, phased 2010–2013). Manufacturing DRC reduced. EP qualifying salary raised to S$2,800 (effective 1 January 2011). Growing public and political concern about foreign worker and foreign talent inflows following GE2011, in which the PAP's vote share fell to its post-independence low of 60.1% amid intense debate on housing affordability, transport congestion, and PMET displacement.

2012 — EP qualifying salary raised to S$3,000 (effective 1 January 2012). Fair Consideration Framework development begins; National Conversation initiated. Further DRC tightening announced for services sector (from 45% to 40%, phased to 2015–2016). EFMA amended to enhance employer accountability for upfront costs and recruitment fees charged in source countries.

2013 — Population White Paper (January 2013) triggers large public protest (Hong Lim Park, 16 February 2013). Government announces a moderate recalibration — slowing the pace of foreign workforce growth — but does not reverse the structural dependency. The "Quality" framework for EP applications tightened. EP qualifying salary subsequently raised to S$3,300 with effect from 1 January 2014.

2014 — Fair Consideration Framework (FCF) formally launched (August 2014). Firms hiring PMETs must advertise on Jobs Bank (later MyCareersFuture) for a minimum of 14 days before applying for an EP. FCF watchlist established for firms with disproportionate concentrations of same-nationality foreign nationals.

2017 — Mandatory progressive wages for the cleaning sector begin enforcement. S Pass qualifying salary raised to S$2,200 from 2018. MOM strengthens EFMA enforcement against salary norm non-compliance and moonlighting.

2020 — COVID-19 dormitory crisis (see SG-G-41): mass outbreak in WP dormitories. By mid-December 2020, MOH reported 54,505 PCR-confirmed cases among dormitory residents (out of 58,320 total Singapore cases); serology testing eventually showed ~47% of the ~320,000 dormitory population had been infected, equating to approximately 152,794 PCR-positive, serology-positive, or both. The WP architecture's welfare gaps were catastrophically exposed. Simultaneously, COVID-19 disrupts EP and S Pass inflows significantly; some EP holders depart Singapore as multinational employers scale back regional offices.

2021–2022 — Post-COVID recalibration. S Pass qualifying salary raised to S$2,500 (September 2022) for non-financial-services. EP qualifying salary raised to S$5,000 (September 2022) for most sectors, S$5,500 for financial services. MOM announces the COMPASS framework (Budget / COS 2022), signalling the most significant redesign of the EP system since the pass was created.

2023 — COMPASS (Complementarity Assessment Framework) implemented on 1 September 2023, replacing the binary EP assessment with a points matrix structured around four foundational criteria (salary, qualifications, diversity, support for local employment) plus two bonus criteria (shortage occupation skills bonus, strategic economic priorities bonus); applicants must score at least 40 points. The change applies to new EP applications; renewals transition from 1 September 2024. S Pass qualifying salary raised to S$3,150 (S$3,500 for financial services) from 1 September 2023.

2024–2026 — Post-COMPASS consolidation. EP renewal applications come under COMPASS from 1 September 2024. S Pass minimum qualifying salary raised from S$3,150 to S$3,300 (S$3,800 for financial services) from 1 September 2025. Construction sector quota ratio reduced from 1:7 to 1:5 from 1 January 2024. Sector-specific progressive wages continue to raise the floor beneath WP-alternative cost comparisons.

4. The Pre-1990 Architecture — Open Inflow, Casual Workers

Singapore's pre-independence economy was powered by migrant labour. The Straits Settlements port economy of the nineteenth and early twentieth centuries drew Hokkien, Teochew, Cantonese, and Hakka Chinese labour for commerce, construction, and trade; Tamil labour for public works; Malay and Bugis sailors and artisans. By the time of independence in 1965, Singapore's population was already the product of wave upon wave of migration, and the colonial labour regime had normalised the idea of a workforce significantly composed of people who were not born in Singapore.

What changed after 1965 was not the fact of migrant labour but its governance logic. The PAP government's primary manpower concern in the first years of independence was not immigration management but employment generation for its own citizens. The formal labour framework established in 1968 — the Employment Act and, crucially, the 1968 Industrial Relations (Amendment) Act, which stripped trade unions of the right to bargain on many categories of employment condition — was directed at the resident workforce. Foreign workers were a secondary concern, admitted through largely informal administrative channels when demand exceeded resident supply.

The 1970s construction boom created the first sustained demand pressure that could not be met domestically. The Housing Development Board's ambitious building programme — launched to address the appalling housing conditions documented in the 1950s and accelerated through the 1970s — required a scale of construction labour that Singapore's small citizen and permanent resident workforce simply could not supply, even at the wage rates the government considered appropriate. Construction wages, unchecked, would have risen to levels that made HDB's per-unit costs prohibitive. The solution — importing Bangladeshi and Indian construction labour, initially through licensed employment agencies and bilateral informal arrangements — was economically rational and consistent with the government's approach to other resource constraints: when a domestic supply is insufficient, import from the cheapest available international source.

This early inflow was governed by administrative discretion rather than statutory architecture. The Ministry of Labour issued work passes on the recommendation of the Construction Industry Training Board and the Housing Development Board; conditions were set by ministerial circular rather than legislation; enforcement was sporadic. Housing for these early workers was provided by contractors in whatever form was cheapest — kampong-adjacent sheds, shophouse attics, converted factory dormitories — with no regulatory standards.

The 1975 foreign worker levy for construction was the first policy acknowledgement that the inflow needed to be managed, not merely admitted. The levy served a dual purpose: it generated revenue, and it created a price signal that would theoretically induce employers to seek productivity improvements rather than endlessly expanding their foreign worker headcount. Both purposes were real; neither fully achieved the intended effect in the short run, since the productivity improvements required capital investment that small construction sub-contractors could not easily make, and the levy rates were set at levels that remained competitive against domestic wage expectations in the source countries.

By the mid-1980s, Singapore had an estimated 100,000–200,000 foreign workers in construction and manufacturing, with no precise public census and significant undercounting; the exact figure remains contested in the historical literature. Throughout the late 1980s, recurring concerns about communicable disease, fire safety, and welfare conditions in unregulated employer-provided dormitories — documented in Straits Times reporting and Ministry of Labour press statements of the period — highlighted the absence of any binding housing standards for this population. The ministerial response was a series of voluntary codes of conduct and circulars for employer-provided housing, the immediate precursor to the more formal EFMA architecture that followed in 1990.

The Work Permit Act 1980 had already established the structural logic of the post-1990 system: a pass tied to a specific employer, in a specific sector, from an approved source country, for a maximum duration (typically two years, renewable), with mandatory repatriation on expiry or earlier termination. What the 1990 EFMA added was codification of employer obligations, statutory offences for breach, and — critically — the extension of the framework to cover the growing services sector alongside construction and manufacturing. The 1990 legislation marked the transition from informal administrative management to a recognisably modern regulatory architecture, though the gaps in welfare standards, enforcement capacity, and worker protection rights would remain substantially unremedied for another three decades.

5. The 1990s Work Permit Architecture and Foreign Worker Levy

The Employment of Foreign Manpower Act 1990 did not emerge from a single policy crisis. It was the legislative crystallisation of lessons accumulated across two decades of ad hoc management — lessons about what happens when employer demand for cheap labour, left institutionally unmediated, produces social costs that the state bears but does not control. Three categories of lesson drove the 1990 codification: fiscal (the informal levy regime leaked revenue through evasion and under-declaration); operational (the absence of statutory employer obligations produced frequent abandonment of injured or ill workers, whose repatriation and medical costs then fell to the government); and political (the growing visibility of foreign worker welfare conditions was generating uncomfortable media coverage that the government preferred to manage proactively rather than reactively).

The EFMA 1990 established the architecture that has governed Work Permit holders through to the present. Its core elements were: the employer-sponsorship tie (a WP holder was bound to a named employer; changing employers required cancelling the existing pass and applying afresh, giving employers substantial leverage over workers); the security bond (employers were required to post a security deposit per worker, forfeited if the worker absconded or was not repatriated at permit expiry); the levy obligation (employers were required to pay a monthly per-worker levy to the government, with rates set by ministerial gazette and adjustable without parliamentary amendment); the source-country restriction (WP workers must come from MOM-approved source countries, which varied by sector); and the social condition covenants (WP holders were forbidden from marrying Singapore citizens or permanent residents without prior Ministry approval, and were subject to mandatory pregnancy testing, with positive results triggering repatriation).

The levy structure was the Act's most economically consequential feature. Unlike a simple headcount cap, the levy creates a continuous financial incentive for employers to manage their foreign worker dependency, since every additional worker costs more per month the closer the employer approaches the Dependency Ratio Ceiling. The levy schedule is graduated: within the DRC, workers are classified as "basic tier" or "higher tier" depending on the employer's aggregate dependency ratio, with higher-dependency firms paying levies at a higher rate. This tiered-within-ceiling design means that an employer at 35% foreign dependency faces a lower per-worker cost than one at 50% dependency, creating a gradient rather than a cliff.

Throughout the 1990s, the DRC was the primary instrument of macroeconomic demand management for foreign labour. Manufacturing DRCs were set sector by sector through subsidiary legislation — typically in the 30–40% range for most trades. The services DRC was set at 50% in 1991, reflecting the government's then-more-permissive view of services sector dependency. MOM's policy preference in this period was still primarily for quantity management rather than quality management: the question was how many foreign workers to admit, not whether the specific skills profile of admits matched the economy's needs. Quality management — salary thresholds, skills testing, qualification verification — would come later, primarily through the EP and S Pass architecture.

The 1990s also produced the first systematic bilateral labour agreements that characterised the WP system's source-country dimension. The Philippines–Singapore bilateral arrangement formalised conditions for Overseas Filipino Workers (OFWs), largely FDWs but also a growing number of services and process workers. The Bangladesh–Singapore arrangement formalised the existing construction worker channel. Each bilateral agreement specified: the list of approved recruitment agencies in the source country; the maximum recruitment fees that workers could be charged (often exceeded in practice); medical clearance requirements; insurance obligations; and repatriation protocols. These bilateral frameworks gave source country governments limited but not negligible leverage over conditions — a factor that would become more significant from 2010 onward when both the Philippines and Indonesia periodically threatened to suspend worker deployments over welfare incidents.

The 1997 Asian financial crisis was the first major stress test of the WP architecture as an economic shock absorber. When construction projects were cancelled or deferred and manufacturing orders collapsed, Singapore employers did not retain surplus WP workers: they let permits expire, declined renewals, and in some cases terminated employment mid-tenure and repatriated workers. An estimated several tens of thousands of foreign workers left Singapore between 1997 and 1999 as construction and manufacturing employers declined renewals; precise crisis-period attrition figures are not in MOM's publicly released Foreign Workforce Numbers series, which begins systematic publication only from 2000. This absorption function — transferring the labour-surplus adjustment cost to workers in source countries rather than to the resident workforce or the state — was never explicitly stated as an architectural objective, but it was implicitly understood by policymakers and is now well-documented in the academic literature on Singapore's macroeconomic flexibility. The price paid was borne by workers who had paid significant recruitment fees in their home countries and returned without the savings they had anticipated.

6. The 2004 Employment Pass / S Pass Architecture

The Employment Pass had a prehistory before its 2001 formalisation. From the late 1980s, Singapore had been admitting foreign professionals under a Professional Visit Pass and, from 1992, under an Employment Pass that was largely indistinguishable in practice from the PVP except in duration. The 2001 reform created the EP as a distinct statutory category with a minimum qualifying salary — S$2,500 per month — and the requirement of an acceptable tertiary qualification or its equivalent in specialised skills. The EP targeted professionals, managers, executives, and technicians (PMETs), and was employer-sponsored: an employer needed to make the application on the worker's behalf, and the pass was not portable without a new employer application.

The introduction of the S Pass in July 2004 was the more architecturally significant development. The S Pass responded to a recognised gap: there existed a category of foreign worker — skilled enough to command a salary above the Work Permit range, but not holding the degree qualifications the EP required — for whom neither the WP nor the EP was appropriate. Technicians, experienced supervisors, skilled tradespeople in specialised manufacturing or marine engineering, some categories of healthcare para-professional: these workers were being admitted either through Work Permits (undervaluing their skills) or through EP applications that required qualification inflation or employer misclassification. The S Pass created a middle tier with its own qualifying salary (S$1,800 per month at introduction), its own sub-dependency ceiling, and its own levy.

The S Pass sub-dependency ceiling (SDC) was designed to prevent the S Pass from becoming a bulk-import vehicle for mid-skilled workers in the way that the WP was used for lower-skilled sectors. The SDC limited the proportion of a firm's total workforce that could be S Pass holders, initially set at 15% for most sectors and lower for others. Employers who wished to employ S Pass workers beyond the SDC had to reduce their total foreign dependency or seek exceptional approval. The SDC has been tightened iteratively. For the services sector, it dropped from 15% to 13% (1 January 2020) and then to 10% (1 January 2021). For most other sectors, it has held at 15–20%, with the overall S Pass quota reduced from 25% to 20% in 2024 in selected sectors — in parallel with successive reductions in the overall DRC.

Between 2001 and 2014, both the EP and S Pass qualifying salaries were progressively raised. The EP threshold went from S$2,500 (2001) to S$2,800 (1 January 2011), then to S$3,000 (1 January 2012), and S$3,300 (1 January 2014). The S Pass threshold went from S$1,800 (2004) and was raised in stages through the 2010s — including to S$2,200 by 2017–2018 — before the post-2020 increases described below. These increases served a dual purpose: they kept EP and S Pass admissions tied to market-competitive wage levels (ensuring that the resident PMET workforce did not face foreign competition at below-market wages) and they applied implicit pressure on employers to seek higher-productivity foreign hires rather than substituting cheaper foreign workers for more expensive residents at the professional tier.

The EP architecture in its pre-COMPASS form was notable for what it did not measure. It assessed individual credentials (salary, qualifications, age, and employer track record) but was blind to the aggregate labour-market consequences of the pass: it could not distinguish between an employer who was genuinely unable to find a qualified Singaporean for a specialised role and one who simply preferred to hire from a particular nationality for network or cost reasons. The Fair Consideration Framework, introduced in 2014, added the 14-day advertising requirement to close this gap partially, but compliance was unverifiable without employer data that MOM did not systematically collect until the FCF watchlist programme began publishing anonymised firm-level statistics in 2016.

The COMPASS reform of 2023 addressed the EP architecture's blindness to aggregate effects by adding firm-level criteria to the individual-level assessment. But the S Pass in 2004–2023 also reflected the same blindness: it assessed individuals rather than the firm-level pattern of hiring. The progressive salary increases for the S Pass in the post-2020 period — S$2,500 from September 2022, S$3,150 from September 2023, and S$3,300 (S$3,800 for financial services) from September 2025 — represent an attempt to use the salary floor as a blunter but administratively simpler substitute for the kind of firm-level complementarity assessment that COMPASS provides for the EP tier.

7. The 2010–2014 Tightening — Dependency Ratios and the Tier System

The 2011 General Election was the proximate political trigger for the most aggressive tightening of the foreign worker architecture since the EFMA's introduction. The PAP's vote share fell to 60.1%, the lowest since independence, with the loss of Aljunied GRC to the Workers' Party in the most symbolically significant electoral shift of the post-Lee Kuan Yew era. Exit polling and post-election analysis consistently identified foreign worker and foreign talent policy as a top-three voter concern — specifically, the perception that the government had prioritised economic growth through foreign inflows over the quality of life, job security, and cultural coherence that resident citizens valued. The housing affordability crisis (closely linked to population growth through immigration) and transport congestion (linked to the same) reinforced the political salience of the foreign worker question.

The MOM response in 2011–2014 was the most systematic use of the DRC instrument in the system's history. The services sector DRC was reduced from 50% to 45% (announced 2010, implemented 2011), then further to 40% (announced 2012, phased to 2014), then to 38% (effective 1 January 2020) and to 35% (effective 1 January 2021). Manufacturing DRCs were reduced in parallel. Each step was accompanied by a phase-in period of two to three years to allow firms to adjust, and by workforce transition support from MOM and Workforce Singapore — funded internship programmes, on-the-job training subsidies, and hiring incentives for disadvantaged local job seekers.

The "Tiered System" within the DRC was refined in this period to create sharper differentiation between firms with different dependency profiles. The key insight behind the tier system is that a blanket DRC — a single ceiling applied uniformly — cannot distinguish between firms in labour-scarce industries where local supply is genuinely insufficient and firms in industries with adequate local supply but a preference for cheaper foreign workers. The tier system addresses this by varying the levy rate based on the employer's position within the DRC band: the closer to the ceiling, the higher the levy per worker. Employers who stay well below the DRC pay lower levies, preserving their cost advantage and reducing the incentive to substitute foreign for local workers in positions where locals could be attracted with modest wage improvements.

The Fair Consideration Framework (FCF), formally launched in August 2014, added a parallel quality-management dimension to the quantity management that DRCs provided. The FCF required employers wishing to hire EP holders for PMET positions to advertise those positions on the Jobs Bank (later MyCareersFuture) for at least 14 days before submitting an EP application, allowing MOM to verify that a genuine local job search had been conducted. The FCF was enforced through the EP application process: applications from firms suspected of discriminatory hiring were escalated for enhanced scrutiny, and firms that attracted multiple discrimination complaints could be placed on the FCF watchlist — a public listing that named firms with disproportionate nationality concentrations relative to peers in the same industry.

The watchlist mechanism — introduced in 2016 — was designed to deploy reputational pressure as a complement to administrative sanctions. Watchlisted firms faced enhanced EP application scrutiny, potential denial of new EP and S Pass applications until they demonstrated improved local hiring, and the reputational cost of public listing in a business community where talent recruitment depends on employer brand. By August–September 2020, MOM had placed approximately 400 firms on the FCF watchlist (with a further 47 added in August 2020); of those 47, more than 60% were in the financial services and professional services sectors where EP dependency was highest and nationality concentration most visible. Specific cases highlighted by MOM included a wealth management firm whose PMETs were ~75% of a single nationality and a bank where same-nationality foreigners made up roughly two-thirds of the PMET headcount.

The 2013 Population White Paper deserves specific attention because it exposed the structural tension within the government's own position. The White Paper projected Singapore's total population at 6.5–6.9 million by 2030 (from approximately 5.3 million in 2013), with continuing substantial foreign worker and foreign talent inflows as a structural necessity given the declining local birth rate. The political reaction was the largest public protest since the Hepatitis B vaccination controversies of the 1980s: a 4,000–5,000 person rally at Hong Lim Park on 16 February 2013 — a modest number by international standards, but exceptional in Singapore's controlled civic environment. The government did not withdraw the White Paper but re-framed its projections as scenarios rather than targets, and announced a package of additional DRC tightenings, EP quality improvements, and local hiring support measures that partially addressed public concerns without resolving the underlying structural dependency.

8. The 2023 COMPASS Implementation — Six-Criteria Scorecard

COMPASS — the Complementarity Assessment Framework — was the product of a prolonged policy development process that began informally with the post-2014 FCF experience and was formalised in a September 2022 public consultation paper. The consultation paper named the precise problem that the binary EP gate had failed to solve: even with the FCF's advertising requirement, MOM had limited visibility into whether individual EP applications, considered in aggregate, were complementing or substituting for the resident PMET workforce. The FCF addressed the front end (was the role advertised?) but not the outcome (is the firm's workforce profile consistent with genuine local hiring effort?). COMPASS was designed to solve the outcome problem.

The framework operates as a points matrix applied to each EP application. An applicant must score at least 40 points (out of 100) to qualify. Points are awarded across four individual-level criteria and two firm-level criteria. The individual criteria are: Salary (compared to age-specific local PMET benchmarks — the single most heavily weighted criterion, able to deliver up to 20 points); Qualifications (degree from a highly ranked university delivers maximum points; lower-ranked or unverified institutions deliver fewer, with MOM maintaining a tiered list of institutions); Diversity (applicant nationality relative to the prevailing distribution of EP holders — applicants from countries underrepresented in Singapore's EP workforce receive a bonus, creating aggregate diversity incentive); and Skills Bonus (applicable to applicants in shortage occupations on the Skills in Demand list, or to recent local-graduate sponsors). The firm-level criteria are: Local PMET Share (the proportion of PMETs in the firm who are Singapore citizens or PRs relative to industry benchmarks — firms below the benchmark score zero on this criterion; firms above it score up to 10 points); and Nationality Spread (the share of the firm's existing EP workforce from the applicant's nationality — firms with high same-nationality concentration score zero; firms with diverse EP workforces receive points for new applications).

The scoring parameters — the benchmarks, the thresholds, the exact point allocations — are set by ministerial gazette and can be adjusted without primary legislation. This was a deliberate design choice, allowing MOM to calibrate the instrument in response to labour market shifts without the delay of parliamentary amendment. Critics, including the Singapore National Employers Federation (SNEF) and several multinational employer associations, raised predictability concerns: an employer planning a multi-year expansion cannot reliably project its EP entitlement if the benchmarks are subject to annual administrative revision. MOM's response was that the direction of change would always be announced with at least six months' lead time, and that the alternative — a legislatively fixed system unable to respond to fast-moving labour market conditions — would be worse for Singapore's overall competitiveness.

The transition arrangements for COMPASS reflected the scale of the disruption to existing EP holders. For new applications from 1 September 2023, COMPASS applied immediately. For renewal applications, the framework applied from 1 September 2024, giving holders of existing EPs one renewal cycle to adjust. Firms that were on the FCF watchlist before COMPASS implementation had their eligibility for points on the Local PMET Share criterion frozen at zero until they were removed from the watchlist, creating an additional incentive for watchlisted firms to accelerate local hiring improvements.

EP applications in the first year of COMPASS operation declined in volume relative to 2022 — reflecting both the tighter qualifying conditions and a recalibration of employer appetite as HR teams assessed which roles would clear the new hurdles . MOM's stated objective was not to reduce the total EP workforce but to improve its composition: the same volume of EP holders, drawn from a wider range of nationalities, with a higher salary and qualifications profile, in firms that could demonstrate genuine local hiring effort alongside their foreign talent requirements. Whether the first years of COMPASS operation achieved this objective is a question that the MOM's post-implementation evaluation data, when published, will answer more precisely than the available contemporaneous data permits .

9. The Sector-Specific Architecture — Construction, Marine, Manufacturing, Services

The generic three-tier architecture (EP/S Pass/WP) describes the pass categories. The sector-specific architecture describes the real operating conditions for the 900,000-plus Work Permit holders who constitute the bulk of Singapore's foreign workforce. Each major sector — construction, marine and offshore engineering, process industry, manufacturing, and services — operates under a distinct overlay of source-country restrictions, dependency ceilings, levy schedules, and worker welfare obligations that the pass category alone does not capture.

Construction is the sector where the tension between economic dependency and policy aspiration is starkest. The construction sector operates at WP dependency ratios that would be considered extraordinary in any other developed economy: the effective ratio of foreign to local workers on large project sites routinely exceeds 80–90%, since the resident population largely refuses construction site labour regardless of wage levels. The sector's foreign worker governance instrument is not merely the WP or the DRC but the Man-Year Entitlement (MYE) system, which allocates project-specific foreign worker quotas based on project value and duration. Introduced in 1987, the MYE prevents firms from speculative hoarding of WP holders between projects and forces allocation decisions to align with actual project labour requirements. The levy schedule for construction WP holders is graduated by tier: Basic Tier workers (whose employer is at lower dependency ratios) attract lower levies than Higher Tier workers at more dependency-intensive firms. As of 2024, construction WP levies ranged from approximately S$300 (Tier 1 Higher-Skilled / R1) through S$700 (Tier 1 Basic-Skilled / R2) to S$950 (Tier 2 / MYE-waived Basic-Skilled) per worker per month. Approved source countries for construction WP holders fall into three groups per MOM: Traditional Sources — Malaysia and the People's Republic of China; Non-Traditional Sources (NTS) — India, Sri Lanka, Thailand, Bangladesh, Myanmar, and the Philippines; and North Asian Sources — Hong Kong (HKSAR passport), Macau, South Korea, and Taiwan. From 1 January 2024, the construction sector quota ratio was reduced from 1:7 to 1:5 (i.e. five WP holders per local employee earning the Local Qualifying Salary).

Marine and offshore engineering has historically operated under even higher effective WP dependency than construction, reflecting the near-total unavailability of Singapore citizens willing to perform heavy marine fabrication, painting, sandblasting, and outfitting work in the demanding conditions of a shipyard environment. The COVID-19 dormitory crisis of 2020 — which centred on WP dormitories in the Jurong Island and western industrial belt — disproportionately affected the marine sector workforce, which constituted a significant share of the 52 large foreign worker dormitories then regulated under the Foreign Employee Dormitories Act (FEDA) 2015. Post-crisis dormitory standards under FEDA, including minimum living space requirements (4.5 square metres per resident), medical facilities, recreation amenities, and food provision standards, imposed significant cost increases on dormitory operators and, ultimately, on marine sector employers.

Manufacturing covers the most heterogeneous WP workforce: from wafer fabrication technicians (who may qualify for S Pass or even EP) to electronics assembly operators (who work under WP, typically from Malaysia, Bangladesh, or the Philippines) to pharmaceutical manufacturing process workers (governed under the process sector DRC rather than manufacturing). The manufacturing DRC has held at 60% in the post-2010 period (with sub-sector variation), with the principal tightening lever applied instead through progressive levy increases and the S Pass / SDC framework rather than the headline manufacturing DRC — accompanied by sustained automation investment, including MOM's investment grant schemes for foreign worker levy-funded productivity improvements. The levy revenue is partially recycled through these grant schemes, creating a self-funding logic: higher-dependency firms pay more, and those funds partly subsidise automation investments that reduce future dependency.

Services — encompassing retail, food and beverage, cleaning, security, accommodation, and a range of consumer services — is the sector where the political salience of foreign worker policy is highest and the DRC tightening has been most sustained. The services DRC trajectory from 50% (1991) to 35% (2016–2017) is the most dramatic single-sector reduction in the architecture's history. The reason for the particular focus on services is political as much as economic: services sector foreign workers — cashiers, kitchen staff, cleaners, security guards — are the foreign workers most visible to the Singaporean public, and their perceived presence and wage competition most directly animates the political concerns about displacement and identity that have driven electoral outcomes since 2011.

The Progressive Wage Model (PWM) has restructured the services sector's effective labour cost architecture since its introduction in cleaning (2014, enforcement of mandatory licensing under the Environmental Public Health Act), security (2016, mandatory under the Private Security Industry Act), and landscape (2016, mandatory under the Public Sector Standard Conditions). Sectoral expansion followed in retail (1 September 2022), food services (1 March 2023), waste management (July 2023), and selected occupational categories including administrators and drivers from 2023–2024. By setting mandatory wage floors significantly above the WP-tier wage equivalents in each sector, the PWM reduces the cost arbitrage that made WP substitution for local workers attractive and pushes sector wages toward levels at which locals are more willing to accept employment. The PWM and the DRC tightening are complementary instruments: the DRC limits the volume of WP inflow, and the PWM raises the price floor beneath which WP workers provide no labour-cost advantage to employers.

10. The 2024–2026 Local Inclusion Multiplier (LIM) Era

The period from 2024 to 2026 represents the maturation phase of the COMPASS architecture, as MOM moved from the threshold instrument (COMPASS as the EP gate) toward a positive incentive mechanism designed to reward firms that meaningfully exceeded their local hiring obligations rather than merely meeting them. The corpus uses "Local Inclusion Multiplier" / "LIM" as a working label for the bundle of post-COMPASS positive-incentive refinements that MOM has signalled across 2024–2026 COS factsheets. In substance, the mechanism is intended to grant qualifying firms more favourable treatment on COMPASS's firm-level criteria — and potentially augmented EP or S Pass quota entitlements — in exchange for demonstrated exceeding of local PMET hiring benchmarks over a defined reference period.

The design logic of the LIM reflects a broader shift in the MOM policy philosophy under the Forward Singapore framework, which Lawrence Wong's government formally released in October 2023 and which underpins the social compact renegotiation of the post-COVID era. The Forward Singapore "Empower" pillar explicitly frames the foreign worker question not as a zero-sum competition between residents and foreigners but as a complementarity problem: Singapore needs both an internationally competitive talent-attraction system and a demonstrably fair local first-hiring culture. The LIM is intended to make these two objectives mutually reinforcing: firms that invest in developing local PMET talent and can demonstrate sustained above-benchmark local hiring are rewarded with greater flexibility to bring in the specific foreign talent that complements their local workforce.

The LIM operates alongside — not instead of — COMPASS. A firm seeking to hire an EP holder still needs the individual application to clear the COMPASS points threshold. But a firm with LIM-certified status may receive a higher sub-quota allocation (the maximum number of EP holders relative to its workforce) or a more favourable treatment of its firm-level COMPASS criteria (specifically the Local PMET Share criterion), effectively allowing it to hire more EP holders per qualifying local PMET than a non-LIM firm. The incentive is therefore structured to create a virtuous cycle: hire more locals, earn more EP flexibility, use that EP flexibility to bring in complementary talent that lifts the productivity of the entire team, which in turn creates conditions in which more locals can be promoted and hired at PMET level.

The practical impact of these post-COMPASS positive-incentive refinements in their first operating years (2024–2026) is not yet fully documented in the public record. The technology and financial services sectors — which had the highest FCF watchlist exposure and the most acute COMPASS compliance challenges — are the intended primary beneficiaries, since these sectors have both the appetite for EP talent and, increasingly, the talent development infrastructure (Graduate Development Programmes, structured mentorship, and local PMET pipeline investments) to qualify under the COMPASS firm-level criteria.

The LIM era also coincided with a broader recalibration of Singapore's talent attraction architecture. The Overseas Networks and Expertise (ONE) Pass, introduced in January 2023, created a premium EP-equivalent category for high earners (S$30,000 per month or above) and individuals with exceptional achievements, with five-year validity and permission for the holder's spouse to work in Singapore without a separate work pass. The ONE Pass, like the Personalised Employment Pass (PEP) before it, represents the top tier of a de facto four-tier system: ONE Pass / PEP (premium talent, maximum flexibility), EP (professional/managerial tier, COMPASS-gated), S Pass (mid-skilled tier, salary-floored and quota-controlled), and WP (lower-skilled, sector-specific and employer-tied). The LIM inserts a firm-level qualification dimension across the EP and S Pass tiers, making the effective quota available to each firm a function of its local hiring track record rather than a fixed proportion of its headcount.

11. Comparative Lens — Singapore vs UK Points-Based, Australia Skilled, Korea EPS

Singapore's foreign manpower architecture is frequently positioned in policy discourse as a globally competitive talent attraction system. Comparative analysis against three peer systems — the UK Points-Based System (PBS), Australia's skilled migration programme, and Korea's Employment Permit System (EPS) — reveals the architecture's distinctive design choices and the trade-offs they embody.

United Kingdom Points-Based System. The UK PBS, introduced in 2008 and substantially reformed following Brexit (from 1 January 2021), operates as a sponsorship-based scored system for overseas nationals seeking to work in the UK. The Skilled Worker visa — the primary work route — requires a sponsoring employer, a job at RQF Level 3 or above (roughly A-level equivalent), a minimum salary (£26,200 per year from April 2024, with occupation-specific going rates), and English language proficiency. Like COMPASS, the UK system is points-weighted: candidates can trade off salary against shortage occupation status. Unlike COMPASS, the UK system contains no firm-level criteria: there is no equivalent of the Local PMET Share or Nationality Spread components. An employer with 100% of its skilled workforce from a single nationality would face no particular UK PBS disadvantage, as long as each individual application met the individual-level thresholds. Singapore's COMPASS is therefore architecturally more ambitious than the UK PBS in its attempt to steer aggregate workforce composition, but more opaque in its scoring parameters and less portable from the individual applicant's perspective (a COMPASS score is employer-specific, not individual-portable). The UK PBS also provides a clearer pathway to settlement (Indefinite Leave to Remain after five years in the Skilled Worker route), whereas Singapore's EP does not confer any PR entitlement and the EP-to-PR conversion rate is administratively discretionary and not publicly published.

Australia Skilled Migration Programme. Australia's skilled migration architecture differs from Singapore's in its explicit orientation toward permanent residence. The primary skilled migration channels — the Employer Nomination Scheme (ENS) and the General Skilled Migration (GSM) programme — are designed to select individuals with the skills and age profile suited for long-term economic contribution, and most subclasses carry a direct pathway to Australian permanent residency. The Subclass 482 (Temporary Skill Shortage) visa provides a temporary skilled worker channel analogous to the EP, with similar salary floors (the Temporary Skilled Migration Income Threshold, TSMIT, raised to A$73,150 per year from July 2023). Australia's system includes a Labour Market Testing (LMT) requirement — advertising the position to Australian residents before lodging a 482 application — analogous to Singapore's FCF advertising requirement, though enforcement mechanisms differ. Singapore's WP has no Australian equivalent at the lower-skilled tier: Australia's Pacific Australia Labour Mobility (PALM) scheme for Pacific Island workers is structurally similar in providing a managed low-skilled worker inflow, but the PALM scheme is smaller, more development-partnership-oriented, and provides significantly more worker mobility rights (including the right to change employers within the scheme). Singapore's WP employer-tie remains more restrictive than any comparable system in an advanced economy.

Korea Employment Permit System. Korea's EPS, introduced in 2004, provides the most instructive comparison for the WP tier. Korea shares with Singapore a manufacturing and construction economy that cannot be fully staffed by domestic workers at acceptable wage levels, a reluctance to extend permanent residence rights to low-skilled labour migrants, and a preference for managed bilateral arrangements with sending countries. The EPS admits workers from sixteen sending countries (as of mid-2024; Tajikistan was added via MOU signed 31 October 2024, expanding the list to seventeen, with first deployments in 2025), assigns them to specific employers and sectors under conditions broadly similar to Singapore's WP, and requires workers to return to their home country after a maximum of five years (with a brief home-country stay required before re-application). Where the EPS differs from Singapore's WP is in the degree of worker mobility: EPS workers who have been employed for one year may change employers (with the employer's consent or in defined hardship cases without consent), and this job-mobility right has been expanded in successive EPS reviews. Singapore's WP holder cannot change employers without the current employer's agreement and a new permit application, giving Singapore employers more leverage over WP workers than Korean employers have over EPS workers. The ILO has documented the link between employer-tied systems and vulnerability to exploitation; Singapore's 2020 COVID-19 dormitory crisis provided a dramatic illustration of the welfare consequences when that structural vulnerability is compounded by inadequate welfare infrastructure.

12. Outcomes Through 2026

Evaluating the foreign worker architecture's outcomes requires separating three distinct populations — EP/S Pass holders, WP holders in industry sectors, and FDWs — since the architecture operates differently for each and the welfare and economic outcomes diverge correspondingly.

For EP and S Pass holders, the architecture in its 2026 form provides a relatively well-functioning professional labour market, subject to the constraints of employer sponsorship. EP holders enjoy the same employment rights as Singapore citizens and PRs under the Employment Act (including protection against unfair dismissal, statutory leave entitlements, and CPF contributions for those who qualify), though CPF obligations apply only to PRs and citizens, not to EP holders. The main vulnerability for EP holders is the pass's dependence on a single employer sponsor: a retrenchment terminates the EP, and the worker has a defined short-term period to find new employment or depart. Post-COMPASS, EP approval rates have become more variable across firms — firms with poor FCF records or high nationality concentration face higher hurdles — but the system remains competitive with peer economies at the high end.

For WP holders, the outcomes picture is more mixed. On wages, real wages for construction and marine WP holders have risen substantially since 2010, driven by progressive levy increases, the labour shortage in the sector, and the post-COVID tightening of foreign worker inflows. MOM's sector-specific Foreign Worker Levy data and the National Wages Council's recommendations have collectively pushed median construction worker wages upward; precise median construction WP wage figures are not separately published in headline MOM Labour Market Reports, and detailed wage-trend numbers should be sourced from sector reports rather than asserted here. On welfare, the COVID-19 dormitory crisis exposed catastrophic gaps in living conditions, disease management, and mental health support for the WP population housed in large dormitories. The Foreign Employee Dormitories Act (FEDA) 2015 and its post-COVID revisions established minimum standards, but advocacy organisations including Transient Workers Count Too (TWC2) and HOME have documented ongoing compliance gaps in smaller, non-FEDA-regulated accommodation. On legal protection, the EFMA amendments of 2012 and 2016 strengthened protections against salary non-payment and the imposition of recruitment fees, but enforcement against overseas recruitment agents in sending countries remains structurally difficult, and pre-departure fee deductions — technically illegal under Singapore's employer obligation framework — persist in some bilateral corridors.

For the resident PMET workforce, the system's stated objective is complementarity — ensuring that EP and S Pass admissions add productive capacity without displacing qualified Singaporeans. The evidence on this is contested. MOM's own data show that PMET employment among Singapore citizens and PRs has grown consistently through the 2010–2026 period, even as total EP numbers rose. Critics — including academic economists and opposition MPs — argue that this growth is consistent with a scenario in which both local and foreign PMET employment expanded, and does not prove that local PMETs were unaffected by the competition. The COMPASS reform's focus on firm-level Local PMET Share benchmarks represents MOM's acknowledgement that the pre-COMPASS system did not adequately ensure complementarity at the firm level, even if aggregate local PMET numbers were growing.

The aggregate labour market outcome as of December 2024 is a foreign workforce of close to 1.55 million against a resident workforce of approximately 2.45 million — i.e. roughly four in ten workers in Singapore are non-residents (a higher share than the often-cited "30%" figure, which excludes WP holders from some older counting conventions). The economy has achieved near-full resident employment, with resident unemployment in the low 2-percent range in 2024–2025 , at the cost of significant social complexity around housing, transport, public space, and national identity that the foreign worker architecture continues to generate even as it enables the economic performance that funds Singapore's extensive public services.

Conclusion

Singapore's foreign worker architecture — the Employment Pass, S Pass, Work Permit, and the COMPASS reform that restructured the EP tier in September 2023 — is the product of five decades of incremental adjustment to a structural condition that the city-state cannot escape and has chosen not to conceal. Singapore is a small, densely populated city-state without the domestic labour force to staff all sectors of its economy at the productivity and wage levels it requires. The architecture that has evolved to manage this condition is, by the standard of comparable systems, administratively sophisticated, legally codified, and continuously refined in response to economic, political, and social pressures.

The system's enduring tension is between two objectives that are genuinely difficult to reconcile: attracting and retaining the foreign talent that keeps Singapore competitive, and protecting the employment prospects, wages, and social environment of the resident population. The pre-2011 architecture tilted toward the first objective at the cost of the second; the 2011–2023 period of sustained tightening tilted in the other direction, with sustained DRC reductions, FCF introduction, and progressive wage floors; the COMPASS and LIM innovations of 2023–2026 represent the most ambitious attempt yet to pursue both simultaneously, by replacing blunt quantity controls with a granular complementarity framework that rewards firms for achieving local hiring outcomes rather than merely tolerating a fixed foreign worker ceiling.

The WP architecture for lower-skilled workers remains structurally less reformed than the EP tier. The employer-tie, the social condition covenants, the source-country restrictions, and the levy system have been retained in essentially their 1990 form, with welfare additions layered on top. The 2020 dormitory crisis demonstrated that this architecture's welfare gaps could become a public health and reputational crisis at national scale. The post-crisis investments in dormitory standards, mental health services, and worker welfare infrastructure represent a genuine improvement; whether they are sufficient is a question that will be tested when the next large-scale welfare incident occurs, as it inevitably will given the structural conditions of housing 300,000-plus workers in industrial corridor dormitories.

Singapore's foreign worker architecture is, in the end, a mirror of the city-state's governing philosophy: pragmatic, technocratic, willing to tolerate complexity and hierarchy in pursuit of economic functionality, attentive to political feedback when that feedback is sufficiently strong, and persistently reluctant to resolve structural tensions through the kind of normative commitments — permanent residence pathways for long-serving workers, portable rights, or full labour market equalisation — that would limit future policy flexibility. The architecture is continuously under construction, and that, by design, is how it will remain.

Spiral Index

The COMPASS framework introduces firm-level criteria that no comparable OECD work-pass system has deployed at scale. What does the five-year post-implementation record reveal about whether firm-level scoring actually changes aggregate nationality concentration in Singapore's EP workforce, or whether employers game the scoring by selectively identifying which applications to submit?
The WP employer-tie has been the most durable and most criticised structural feature of Singapore's lower-skilled foreign worker governance. How has Singapore's willingness to retain the employer-tie — in contrast to Korea's partial liberalisation through the EPS — shaped the wage outcomes, welfare conditions, and bargaining position of WP holders, and what are the political economy constraints that have prevented reform?
The 2013 Population White Paper projected Singapore's population at 6.5–6.9 million by 2030 and was met with the city-state's most significant political protest since independence. How has the subsequent decade of tightening, COMPASS reform, and demographic pressure altered the trajectory against those projections, and what does the divergence between projection and outcome reveal about the limits of technocratic population management?
The Progressive Wage Model was initially limited to a small number of cleaning sector workers and has expanded to cover six major services sectors by 2025. What are the structural incentives and political constraints that have shaped the pace and scope of PWM expansion, and how does the PWM interact with the WP levy system in determining the effective labour cost floor that governs employer substitution decisions?
Singapore's foreign worker architecture has been described both as a model of efficient governance of labour mobility and as a system that extracts economic value from a structurally subordinated migrant labour class. Evaluate the evidence for each characterisation with specific reference to: (a) the construction sector WP architecture post-FEDA; (b) the COMPASS scorecard's firm-level incentive design; and (c) the LIM framework's early outcomes.

Sources and References

Ministry of Manpower (MOM), Foreign Workforce Numbers (annual statistical releases, 2000–2026); MOM, Labour Market Reports (quarterly and annual), 2000–2026.
Employment of Foreign Manpower Act (EFMA, Cap. 91A), Singapore Statutes Online; original 1990 enactment; Amendment Acts 2007, 2012, 2016, 2021; MOM, A Guide to the Employment of Foreign Manpower Act (various editions).
Ministry of Manpower, COMPASS — Complementarity Assessment Framework: Policy Statement and Implementation Circular, September 2023; MOM FAQ on COMPASS, updated 2024–2025.
Ministry of Manpower, Fair Consideration Framework (FCF): Policy Statement, 2014; FCF watchlist data and enforcement releases, 2016–2025.
Parliament of Singapore, Hansard — Second Reading of Employment of Foreign Manpower Act 1990; Committee of Supply debates, MOM estimates, selected years 1990–2026; ministerial statements on foreign workforce policy and COMPASS, 2021–2025.
Ministry of Manpower, Dependency Ratio Ceiling (DRC) and Foreign Worker Levy Framework documentation (successive editions, 1991–2026); MOM press releases on DRC and levy adjustments.
Ministry of Manpower, S Pass Qualifying Criteria and Sub-Dependency Ceiling, policy statements (2004, 2008, 2013, 2018, 2022, 2024).
National Population and Talent Division (NPTD), Population in Brief (annual, 2010–2025); A Sustainable Population for a Dynamic Singapore: Population White Paper, January 2013.
Ministry of Manpower, Building an Inclusive Workforce: Tripartite Workgroup on Supporting Local Employment, 2020–2022; MOM, Fair Employment Practices: TAFEP Annual Reports, 2007–2025.
Ministry of Manpower, Progressive Wage Model: Sector Schedules and Criteria — Cleaning, Security, Landscape, Food Services, Retail (2012–2026).
Ministry of Trade and Industry (MTI) / Economic Development Board (EDB), Industry Transformation Maps (ITMs), relevant sectors, 2016–2026; EDB, Manpower-Light Growth Strategy Documentation, 2010–2020.
Parliament of Singapore, Hansard — debates on the Employment Agencies Act; Second Readings on the Construction Industry regulatory framework; Committee of Supply debate MOM 2010 (Teo Chee Hean, Gan Kim Yong statements on tightening).
Ministry of Manpower, press releases and ministerial statements on Employment Pass criteria changes, 2004–2025; MOM, One Pass — Overseas Networks and Expertise Pass: Policy Statement and Criteria, January 2023.
Korea Ministry of Employment and Labour, Employment Permit System Annual Report (various years); ILO, Employer-Tied Migration: Lessons from the Korean EPS (Geneva: ILO, 2017).
UK Home Office, A Points-Based System: Making Migration Work for Britain (March 2006, Cm 6741); UK Visas and Immigration, Skilled Worker visa: eligibility (updated 2022–2025).
Australian Department of Home Affairs, Skilled Migration Programme: Overview and Outcomes (2022–2025); Productivity Commission, Australia's Skilled Migration Programmes: Objectives and Operation (2023).
Institute of Policy Studies (IPS), studies on immigration attitudes and social cohesion in Singapore, 2013–2025; IPS commentary on COMPASS implementation, 2023–2025.
Brenda S.A. Yeoh and Shirena Huang, "The Cosmopolitan–Local Nexus: Foreign Talent Policy and Its Discontents," in Management of Success: Singapore Revisited, ed. Terence Chong (Singapore: ISEAS, 2010).
Arnold Puyat and Tan Peng Boo, "Foreign Workers in Singapore: Policy Frameworks and the Governance of Mobility," Asian Journal of Political Science 17, no. 3 (2009): 253–274.
Gunasekaran Nathan and Yeoh Brenda S.A., "The 'Global City', Labour Market Deregulation and the Politics of Foreign Talent in Singapore," in Challenging the State? The Straits Times & Political Change in Singapore, ed. Lee Boon Hiok (Singapore: ISEAS, 2006).
The Straits Times, Business Times, TODAY, and Channel NewsAsia — contemporaneous reporting on Employment Pass, COMPASS, S Pass criteria, and foreign workforce policy, 1990–2026.
Ministry of Manpower, Foreign Domestic Worker Policy Framework and Conditions of Employment (successive editions); MOM, Employer's Handbook for Hiring a Foreign Domestic Worker, 2024 edition.

SG-G-57: The Foreign Worker Architecture — EP, S Pass, Work Permit, and the COMPASS Reform (1968–2026)