## 1. Consumption Growth, Associated with Decline in GDP Growth

## Source details

**Canonical URL:** [1. Consumption Growth, Associated with Decline in GDP Growth](https://www.imf.org/-/media/files/publications/cr/2021/english/1sgpea2021002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2021/english/1sgpea2021002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2021/english/1sgpea2021002.pdf.json)

---

### Consumption and GDP dynamics — key empirical findings
- A decline in annual GDP growth by 1 ppt was associated with a simultaneous decline in consumption growth by about 0.7 ppt on average for 1980-2019 among advanced economies (AE).
- The impulse response of consumption growth to a shock to GDP growth remains negative for another two years after the shock; e.g., consumption growth following a recession year would be about 0.4 ppt lower than the rate following a normal year (small negative point estimates for T+1).
- Consumption growth is generally less volatile than GDP growth, reflecting policy stabilization of household disposable income and consumption-smoothing behavior by households.
- During past major pandemic episodes in 1980-2019 in AEs (SARS, H1N1, MERS, Ebola, and Zika), consumption growth declined by about 1 ppt with each 1 ppt fall in GDP growth (compared to the overall average response of 0.7 ppt).
- Pandemics with greater health risks (higher infection and mortality rates) are associated with stronger and more persistent negative impacts on consumption growth.
- Consumer confidence:
  - After the global financial crisis (GFC) consumer confidence among AEs remained below the pre-GFC peak for a prolonged period even as business confidence fully recovered.
  - After the peak of COVID-19, consumer confidence recovered only half of its pre-COVID level by 2021Q1, while business confidence fully recovered.
- OECD consumer confidence index benchmark used: average 1980-2020 = 100 (amplitude adjusted).

### Singapore-specific observations and recent experience
- Historical pattern:
  - In Singapore, consumption slumps in recent major recessions tended to be less pronounced but lasted longer than GDP slumps.
  - After the Asian Financial Crisis (AFC) consumption trough and recovery were rapid; after the dot-com and GFC episodes consumption recovery to pre-crisis peak took longer than GDP recovery.
- COVID-19 peak impact in Singapore:
  - Private consumption contracted by a record 29 percent, y/y, at the peak in 2020Q2.
  - The economy contracted by 13 percent, y/y, in 2020Q2.
  - Singapore implemented a nationwide partial lockdown (Circuit Breaker) from April 7 to June 1, 2020.
- Recovery trajectory:
  - Consumption rebounded in 2020Q3 with the end of the Circuit Breaker, but the pace of recovery eased thereafter, likely reflecting lingering health risks and subdued consumer sentiment.

### Mechanisms and interpretation
- Dampened consumption response relative to GDP may reflect:
  - Policy measures that stabilize household disposable income relative to GDP.
  - Consumption-smoothing motives where households dissave during income drops and save during income rises.
- Pandemics affect consumption beyond income channels: health risks and public-health measures limit spending opportunities and raise uncertainty about the future, weakening consumer sentiment and reducing consumption even as GDP begins to recover.

### Policy-relevant implications and buffers
- High savings in Singapore, together with policy support for households, can buffer consumption compression during the crisis and recovery phase.
- Even with policy support, consumption recoveries tend to be slower than GDP recoveries after major shocks — suggesting targeted measures to restore consumer confidence and safe spending opportunities may be needed to accelerate consumption normalization.

*Prepared by IMF staff (Jiae Yoo, APD); IMF staff estimates and figures as presented in the chapter.*

### Consumption levels reflect a variety of socioeconomic conditions that underly consumer behavior — key determinants
- Precautionary saving: Greater uncertainty about future income raises aggregate saving (Carroll et al, 2011; Mody et al, 2012).
- Life-cycle demographics: Households save during working years and dissave during retirement (Modigliani and Brumberg, 1954 and 1980; Modigliani, 1986).
- Other structural factors: Income level, credit availability, and interest rates help explain cross-economy variation in aggregate consumption rates.

### Singapore structural profile — key statistics and structural drivers
- GNDI and income:
  - Gross national disposable income (GNDI) is about 86 percent of GDP.
  - Singapore’s per capita income is among the highest in the world in both nominal and PPP terms.
- Demographics:
  - Population is young but aging fast; current old age dependency (OAD) ratio is relatively low, while aging speed (projected change in OAD in 20 years) is among the fastest.
- Labor share and corporate sector:
  - Labor share in income is among the lowest in the world.
  - Total corporate sector assets amounted to 1,828 percent of GDP in 2019.
  - Share of foreign enterprises is about 63 percent in terms of total value added in 2019.
- Income uncertainty:
  - Higher income uncertainty, measured by forecast error of real per capita GDP growth, contributes modestly to lower consumption rates.

### Cross-country regression (sample: 41 AEs, 1980–2019) — selected coefficient findings
- Income uncertainty: -0.0662* (0.0309) in column (1); up to -0.142*** (0.0351) in column (4).
- Population growth: -0.410** (0.157) to -1.362*** (0.153).
- Old-age dependency ratio: 0.211*** (0.0385) to 0.0807* (0.0386).
- Aging speed: -0.238*** (0.0320) to -0.0697* (0.0279).
- Labor share: 0.347*** (0.0288) to 0.331*** (0.0307).
- Per capita income, PPP, lagged in log: -2.831*** (0.566) and -4.595*** (0.659) in alternative specs.
- Private sector credit: 0.0170*** (0.00332) in specification (6).
- Sample sizes and fit: N. Obs. vary (e.g., 1302, 1821, 2829, 209); R-sq. range from 0.117 to 0.393 across columns.
- Interpretation: Demographics, labor share, and real per capita income are important in explaining Singapore’s relatively low consumption share of GNDI.

### Crisis interactions — selected regression insights (Annex Table 1.2)
- Crisis dummy: defined as year when real GDP growth falls more than one standard deviation below historical average.
- Selected coefficients:
  - Income uncertainty: around -0.0784** (0.0275) to -0.106** (0.0342).
  - Population growth: about -1.252*** (0.155) to -1.278*** (0.160).
  - OAD: positive and significant, e.g., 0.0874*** (0.0262).
  - Aging speed: negative and significant, e.g., -0.0561* (0.0240) to -0.0847*** (0.0255).
  - Labor share: 0.321*** (0.0276).
  - Per capita income, PPP, lagged in log: -1.982*** (0.519) to -4.216*** (0.603).
- Crisis interaction terms (examples):
  - Crisis × OAD: 0.0932** (0.0310) in column (5).
  - Crisis × Aging speed: 0.0803** (0.0302) in column (6).
  - Crisis × Income level: 0.954* (0.443) in column (7).
- Crisis main effects examples: Crisis = -2.310** (0.749) in column (3); Crisis = -2.767*** (0.779) in column (5); Crisis = -10.37* (4.704) in column (7).
- N. Obs. examples: 1262 and 894; R-sq. range: 0.220 to 0.333.
- Implications for Singapore:
  - Negative relationship between consumption share and income level is more muted during crises, suggesting consumption compression could be smaller for high-income economies like Singapore.
  - Low current OAD implies greater consumption compression during a crisis (younger populations reduce consumption more).
  - Fast aging (high aging speed) implies smaller consumption compression during a crisis (stronger saving motives for future old-age consumption).
  - For Singapore, the mitigating effect of fast aging outweighs effect of a low current OAD.

### Household-level expenditure patterns (SG-HES vs U.S. and U.K.)
- Overall: Singapore’s consumption expenditure as a share of gross household income is relatively low compared to the U.S. and the U.K.
- Composition:
  - Non-durables and services (food, health, education, personal care, apparel, recreation) shares are similar across Singapore, U.S., and U.K.
  - Main gaps driven by housing and transport expenditures.
- Transport:
  - Owned vehicle purchase and maintenance spending in Singapore is about half or less than in the U.S. and U.K.; higher public transport spending more than offsets lower vehicle spending.
  - Certificate of Entitlement (COE) requirement leads to low car ownership.
- Housing:
  - Housing- and shelter-related consumption expenditure in Singapore is about half the level in the U.S. and U.K., with rented housing expenditure especially low on average.
  - Homeownership: 90 percent in 2019, with over 80 percent of Singapore Residents living in public housing.
- Survey measures:
  - SG consumption measures are based on income excluding employer CPF contributions.
  - Some imputed rental or mortgage items are not included in cross-country comparisons.

### COVID-19 effects and near-term outlook
- Pandemic impact:
  - Consumption contracted by 14 percent, y/y, in 2020 despite sizable fiscal support.
- Policy implications:
  - Income subsidies or cash grants may deliver higher consumption as restrictions are eased and households feel more secure.
  - Continued support for job creation and a robust labor market will help near-term recovery and boost consumption over the medium- to long-term.
- Structural resilience:
  - Singapore’s structural drivers (high income, fast aging, demographic saving motives) can help alleviate consumption compression during crises.
- Digitalization and e-commerce:
  - Pandemic accelerated online shopping; e-commerce in Singapore remains below most peer economies.
  - Further digitalization and expansion of e-commerce could create new demand beyond substitution from offline to online and could boost consumption over the medium-term.

### Policy-relevant takeaways on consumption
- Structural factors explaining low consumption-to-GNDI in Singapore: high per capita income (PPP), fast aging, low labor share, and modestly higher income uncertainty at the aggregate level.
- Targeted policies to support consumption recovery:
  - Income support (subsidies, cash grants) timed with easing restrictions to restore consumer confidence.
  - Support for job creation and a robust labor market.
  - Policies facilitating digitalization and e-commerce expansion.
- Analytical priorities:
  - Incorporate alternative uncertainty measures more directly related to household disposable income.
  - Use granular household survey analysis to identify item-level drivers and the role of public policies.

### Survey comparison methodology (SG-HES, US-CEX, UK-LCF) — harmonization highlights
- Surveys:
  - SG-HES 2017/18; US-CEX 2018; UK-LCF 2018.
- Harmonization steps:
  - For SG-HES use income excluding employer CPF contributions to align with US-CEX and UK-LCF.
  - Separated pensions, social security, mortgage- and property tax-related payments to align concepts across surveys.
- Mapping of consumption categories provided (food; alcohol & tobacco; health; transport; housing and utilities) with noted definitional differences.

### Labor market impact of COVID-19 — key statistics (Singapore)
- Pre-pandemic unemployment: "a little above 2 percent" for many years; non-residents represented 29.3 percent of total employment (excluding migrant domestic workers) in December 2020.
- Unemployment dynamics:
  - Overall unemployment peaked at 3.5 percent in September 2020.
  - Average annual unemployment rate in 2020 was 3.0 percent.
- Sectoral employment changes in 2020:
  - Manufacturing lost 37,800 workers.
  - Construction lost 51,800 workers.
  - Services lost 91,100 workers.
  - Employment in construction declined by 11.3 percent in 2020.
  - Employment in manufacturing declined by 7.8 percent in 2020.
  - Employment in services declined by 3.2 percent in 2020.
- Non-resident impacts:
  - Non-resident retrenchment rate: 1.6 percent in 2020.
  - Resident retrenchment rate: 1.1 percent in 2020.
  - Non-resident employment share fell to 29.3 percent in December 2020 from 33.1 percent in December 2019.
  - Non-resident employment loss in 2020: 13.7 percent.
- Self-employed: taxi and private hire car drivers, freelancers, vendors, direct sellers were severely impacted.

### Policy response and programs — design, coverage, and budgets (selected)
- Jobs Support Scheme (JSS):
  - Period: Oct. 2019 to Sept. 2021 (extended Apr 2021 to Sep 2021 with reduced co-funding).
  - Budget: S$30 bn (6.3 percent of GDP) in FY2020 and FY2021; S$26.9 bn in FY20.
  - Coverage: More than 140,000 firms employing more than 2 million local workers benefited.
  - Co-funding initially: 25 to 75 percent of the first S$4,600 of gross monthly wages per local employee for 11 months (up to Aug 2020); Sep 2020–Mar 2021: 0-50 percent; Apr–Sep 2021: 10-30 percent for all firms (Tier 3 firms not covered under extended JSS from either June or September 2021).
- Wage Credit Scheme (WCS):
  - Over S$2 billion in wage credits disbursed during the pandemic; over $1 billion disbursed in 2020 to more than 95,000 employers.
- Foreign Worker Levy (FWL) rebates and waivers (Apr 2020 to Dec 2021):
  - Received S$2.3 bn budget in FY20; more than 62,000 firms benefited.
- SEP Income Relief Scheme (SIRS): S$2.0 bn budget in FY20; more than 200,000 self-employed people benefited.
- Temporary Relief Fund (Apr 2020): Budget S$0.2 bn; served about 450,000 people.
- COVID-19 Support Grant (May–Dec 2020): Received S$0.2 bn in FY20; more than 98,000 workers benefited.
- COVID-19 Recovery Grant (CRG) (Jan–Dec 2021): Open since January 18, 2021; about 10,500 applications approved as of March 22, 2021.
- Enhanced Workfare Special Payment (2020 and 2021): Eligible low-income workers and self-employed persons have been receiving S$3,000 each.
- Jobs Growth Incentive (JGI): Government allocated S$6.2 billion (S$1 billion in FY2020 and S$5.2 billion in FY2021).
- SGUnited Jobs and Skills Package (May 2020 to Mar 2022): Nearly 93,000 applicants placed in jobs and skills opportunities as of end-Feb 2021; funds set aside in 2021 to provide 35,000 additional opportunities.

### Policy implications and outcomes — labor market
- Targeting and distribution:
  - Policy support concentrated on residents; mitigated resident unemployment but shifted burden of job losses to non-residents.
- Labor market resilience:
  - Despite the shock, Singapore’s unemployment outcomes remained relatively favorable in cross-country comparison (average annual unemployment rate in 2020 was 3.0 percent).
  - Resident employment grew by 14,900 in 2020.
  - Ministry of Finance reported the Gini coefficient for residents, after transfers and taxes, fell to 0.375 in 2020 from 0.398 in 2019.
- Gaps and affected groups:
  - Non-resident workers and self-employed individuals experienced larger adverse impacts and faced more limited support coverage.

### JSS adjustments during May–June 2021 (Phase 2 Heightened Alert)
- Temporary increases in JSS support (May 16, 2021 to June 13, 2021):
  - Food and beverage, performing arts, and arts education: raised from 10 to 50 percent.
  - Gyms, fitness studios and other sports facilities: raised from zero to 50 percent.
  - Indoor playgrounds and other family entertainment centers, and affected personal care services: raised from zero to 30 percent.
  - Retail, cinema operators, and museums, art galleries and historical sites: increased from 10 to 30 percent.
- JSS payout schedule (selected months/payout ratios):
  - Apr. 2020 (Oct.-Dec. 2019): Tier 1 75%, Tier 2 50%, Tier 3A 25%, Tier 3B 25%.
  - Mar 2021 (Sep.-Dec. 2020): Tier 1 50%, Tier 2 30%, Tier 3A 10%, Tier 3B 10%.
  - Jun 2021 (Jan.-Mar. 2021): Tier 1 50%, Tier 2 30%, Tier 3A 10%, Tier 3B 0%.
  - Sep. 2021 (Apr.-Jun. 2021): Tier 1 30%, Tier 2 10%*, Tier 3A 0%**, Tier 3B 0%**.
  - Dec. 2021 (Jul.-Sep. 2021): Tier 1 10%, Tier 2 0%, Tier 3A 0%, Tier 3B 0%.
- Authorities estimate fiscal policy support measures, among which JSS is the largest single measure, have prevented the resident unemployment rate from rising by 1.7 percentage points in 2020.
- Fiscal measures have helped save or create about 155,000 jobs during 2020–2021 (up to January 2021).

### E-payments during the COVID-19 pandemic and beyond — key trends
- Payment infrastructure milestones: credit cards in the 1970s; GIRO and EFTPOS in the 1980s; contactless EZ-Link cards in 2002; FAST instant payment in 2014; PayNow in 2017; SGQR codes in 2017.
- FAST and PayNow uptake and 2020 changes:
  - FAST volume increased on average exceeding 50 percent per year since 2016.
  - FAST monthly transaction volume rose from 15.5 million to 24.5 million in 2020.
  - FAST average transaction size declined from S$1,592 to S$1,434 in 2020.
  - PayNow mobile phone number users increased by 30 percent, to around 2.8 million users in 2020.
  - PayNow national ID number users grew by 80 percent (YoY), to around 2 million users.
  - Unique entity numbers for companies doubled in 2020 to around 250,000 users.
  - Individual PayNow transaction volumes almost doubled (YoY) to 125 million in 2020.
  - PayNow Corporate volumes increased six times (YoY) to 19 million in 2020.
  - SGQR deployments: more than 120,000 SGQR labels deployed during 2020.
- Household e-payment adoption (2017/18 Household Survey):
  - 36.7 percent of household expenditure is made through e-payments.
  - Share of e-payment expenditure by category (2017-18): Others 30.5; Food 25.8; Transport 44.7; Clothing & Footwear 48.9; Communication 63; Educational Services 64.6; Accommodation Services 76.8.
- Cash and ATM statistics:
  - Over 220 retail bank branches and more than 2,660 ATMs across Singapore.
  - ATM density stable at 59 per 100,000 residents from 2012 to 2019.
  - Cash withdrawals per month fell sharply during the COVID pandemic.
  - Cash estimated to be used in around 60 percent of all retail transactions in Singapore (KPMG, 2016).
  - Currency in circulation around 10 percent of GDP in Singapore and relatively stable over past five years.
  - Cross-country: Sweden 1 percent of GDP; Japan over 20 percent of GDP; BIS CPMI average around 8 percent of GDP.

### Cash use determinants — panel regression (20 economies, 2012-2019)
- Model: CiC regressed on GDP per capita (GDP_pc), financial and macroeconomic uncertainty index (FCI), short-term interest rate (Int), average age (Age), digitalization proxy (Dgt, IMD Digital Competitiveness Index).
- Selected regression results (Table 2; dependent variable cic):
  - GDP_pc coefficient: -2.26*** (0.58)
  - Fci coefficient: 0.02 (0.14)
  - Int coefficient: 0.004 (0.02)
  - Age coefficient: 0.80*** (0.18)
  - Dgt coefficient: -1.52 (5.61)
  - Observations: 130
  - R2: 0.34
  - Adjusted R2: 0.19
  - F Statistic: 10.77*** (df=5; 106)
  - Note: **p<0.1; **p<0.05; ***p<0.01.
- Interpretation:
  - Average age positively related to demand for cash.
  - Currency in circulation inversely related to GDP per capita.
  - Interest rate, cross-country digitalization, and macroeconomic uncertainty not significant in this specification.

### Policies to decrease cash use and foster e-payments — recommendations
- Build inclusive e-payment infrastructure:
  - Ensure basic digital skills and enhance financial literacy for older and poorer segments.
  - MAS “E-Pay Okay” public awareness campaign (March 2021) to promote safe, simple and speedy e-payments.
  - “Seniors Go Digital” program helps senior citizens learn basic digital skills including e-payments.
  - Consider CBDC options to ensure e-payment services for underserved populations (Armelius et al, 2020 a and b).
- Introduce common standards for interoperability and competition:
  - Allow users to carry data across platforms while protecting privacy.
  - MAS Project Ubin explored blockchain-based settlement and a multi-currency settlement prototype.
- Domestic and international policy collaboration:
  - MAS collaborates with IMDA and Enterprise Singapore to encourage e-payments among merchant segments, e.g., SMEs.
  - MAS linked with Bank of Thailand (launched end-April 2021) to connect national retail fast payments.
  - BIS Innovation Hub and MAS joint project to enhance cross-border payments including multi-CBDC transactions.

### Digitalization, innovation, and productivity — Singapore evidence
- Digitalization of production:
  - Operational stock of robots in manufacturing: about 600 robots in 2008; above 17,000 robots in 2018.
  - Robot density: about 1 operating robot per 1,000 employees in 2008; 45 operating robots per 1,000 employees in 2018.
  - Semiconductors accounted for 70 percent of all industrial robots in Singapore in 2018.
- Digitalization of consumption (e-commerce):
  - Business-to-consumer e-commerce sales: 0.3 percent of GDP in 2017; about 0.7 percent of GDP in 2020.
  - E-sales grew by 32 percent in Singapore in 2020.
  - Rakuten Insight survey: close to 80 percent of respondents in Singapore stated they would continue to buy products online even when businesses reopen.
  - E-commerce in Singapore remained under 1 percent of GDP in 2020, below most peer Asian and advanced economies.
  - More than 80 percent of firms had adopted e-payments by 2019.
- Innovation, patents, and R&D:
  - ICT-related patents represent close to 40 percent of total patents (IP5 patent families, 2012–15).
  - Singapore’s share of “triadic” patent families is relatively low compared to peers.
  - R&D expenditure relatively low and broadly constant in recent decades.
  - Global competitiveness (2019): Singapore ranked first overall; ranked 21st in research and development and 23rd in scientific publications.
- Empirical findings on productivity:
  - Increase in e-sales by 1 percentage point raises firm-level labor productivity growth by 0.3 percentage points on average (Falk and Hagsten, 2015).
  - Sector-level analysis (22 AEs, 2000-2019) finds:
    - Larger share of e-commerce to GDP associated with higher labor productivity growth.
    - Higher robot density associated with higher labor productivity growth.
    - Higher business R&D spending associated with higher labor productivity growth.
    - Results robust; no evidence digitalization-productivity association larger in more recent years.

### Policy directions to foster a smarter recovery (digitalization & R&D)
- Key implication: Boosting e-commerce and R&D would support productivity growth, the nascent recovery, and transformation toward a smarter economy.
- Rationale: Singapore has scope for significant expansion of e-commerce and larger R&D, both associated with higher labor productivity growth.
- Existing initiatives noted:
  - IMDA’s SMEs Go Digital program (Start Digital Pack, Industry Digital Plan, Grow Digital, SME Digital Tech Hub, Productivity Solutions Grant).
  - IMDA Open Innovation Platform.
  - Multi-prong 5G development approach (5GSG).

### Green finance and sustainability — overview and key initiatives
- Market scale and growth:
  - GSIA: at least US$30.7 trillion in sustainable/green funds in early 2018.
  - ESG assets under management in Singapore grew to about US$700 billion as of 2019 (about 28 percent of total).
  - Global green debt issuances cumulatively US$2.2 trillion in 2020 (from around US$31 billion in 2012).
  - APAC and ASEAN green debt increased to around US$433 billion and US$56 billion, respectively.
  - Singapore accounted for 55 percent of ASEAN issuances and about 1.4 percent of global green debt issuances in 2020.
- MAS actions and programs:
  - Guidelines on Environmental Risk Management for financial institutions issued December 2020.
  - Complementary GFIT handbook launched January 2021.
  - Stress testing: insurers subjected to extreme flooding scenario in 2018; broader climate risks to be incorporated by end-2022.
  - Sustainable Bond Grant (launched 2017; upgraded 2020); by end-2020 about S$11 billion in green, social and sustainability bonds had been issued in Singapore.
  - Green and Sustainability-Linked Loan Grant Scheme launched in 2020.
  - US$2 billion Green Investments Program (GIP) launched in 2019; US$100 million placed in BIS Green Bond Investment Pool (GBIP).
  - Government identified green infrastructure projects worth S$19 billion (US$14 billion) for green bond funding.
  - MAS is founding member of NGFS and active in international sustainable finance fora.
- Private sector and market developments:
  - ABS guidelines for transparent ESG disclosures and responsible financing.
  - SGX mandatory sustainability reporting for listed entities for financial years ending on or after 31 Dec 2017.
  - GFIT public consultation on taxonomy for green/transitioning activities in early 2021.
  - Climate Impact X (CIX) carbon exchange announced May 2021.
- Banks and disclosures:
  - DBS, OCBC, UOB adopting transition- and physical-risk assessments and integrating ESG into credit and risk frameworks.
  - Banks provided about US$25.5 billion in green and sustainability-linked loans in 2018-2020.
  - DBS committed to zero thermal coal exposure by 2039; UOB prohibited project financing of greenfield thermal coal mines since 2019.
- Green bond listings on SGX:
  - Grew to 103 listings valued at US$44 billion in 2020 from one listing valued at US$500 million in 2013.
  - Green bond listings on SGX largely denominated in US dollars and issued by foreign entities.

### Concluding policy implications on green finance
- Growing green finance market presents opportunity for Singapore to support climate-friendly capital allocation.
- MAS can facilitate trends by enforcing prudent climate risk management and disclosure and supporting development of green finance solutions.
- Private sector role: shift investment priorities by incorporating climate and sustainability considerations.

*Prepared by IMF staff (Jiae Yoo, APD); IMF staff estimates and figures as presented in the chapter.*

### 1. Consumption Growth, Associated with Decline in GDP Growth ____________________ 5

### 1. Consumption Growth, Associated with Decline in GDP Growth

### Key empirical findings on consumption and GDP dynamics
- A decline in annual GDP growth by 1 ppt was associated with a simultaneous decline in consumption growth by about 0.7 ppt on average for the period 1980-2019 among advanced economies (AE).
- The impulse response of consumption growth to a shock to GDP growth remains negative for another two years after the shock; for example, the consumption growth rate following a recession year would be about 0.4 ppt lower than the rate following a normal year (small negative point estimates for T+1).
- Consumption growth is generally less volatile than GDP growth, reflecting policy stabilization of household disposable income and consumption-smoothing behavior by households.
- Pandemics amplify consumption declines: during past major pandemic episodes in 1980-2019 in AEs (SARS, H1N1, MERS, Ebola, and Zika), consumption growth declined by about 1 ppt with each 1 ppt fall in GDP growth (compared to the overall average response of 0.7 ppt).
- Pandemics with greater health risks (higher infection and mortality rates) are associated with stronger and more persistent negative impacts on consumption growth.
- Consumer confidence trends differ from business confidence: following the global financial crisis (GFC) consumer confidence among AEs remained below the pre-GFC peak for a prolonged period even as business confidence fully recovered. After the peak of COVID-19, consumer confidence recovered only half of its pre-COVID level by 2021Q1, while business confidence fully recovered.
- The OECD consumer confidence index is shown with average 1980-2020 = 100 (amplitude adjusted).

### Singapore-specific observations and recent experience
- Historical pattern: in Singapore, consumption slumps in recent major recessions tended to be less pronounced but lasted longer than GDP slumps. After the Asian Financial Crisis (AFC) consumption trough and recovery were rapid, but in other episodes (dot-com, GFC) consumption recovery to pre-crisis peak took longer than GDP recovery.
- COVID-19 peak impact in Singapore:
  - Private consumption contracted by a record 29 percent, y/y, at the peak in 2020Q2.
  - The economy contracted by 13 percent, y/y, in the same quarter.
  - Singapore implemented a nationwide partial lockdown known as Circuit Breaker from April 7 to June 1, 2020 (included closures of all non-essential workplaces, compulsory mask wearing, movement restrictions while testing, and health monitoring).
- Consumption rebounded in 2020Q3 with the end of the Circuit Breaker, but the pace of recovery eased thereafter, likely reflecting lingering health risks and subdued consumer sentiment that can continue to weigh on consumption in the near term.

### Mechanisms and interpretation
- The dampened consumption response relative to GDP may reflect:
  - Policy measures that stabilize household disposable income relative to GDP.
  - Consumption-smoothing motives where households dissave during income drops and save during income rises to stabilize consumption.
- Pandemics affect consumption beyond income channels: health risks and public-health measures limit spending opportunities and raise uncertainty about the future, weakening consumer sentiment and reducing consumption even as GDP begins to recover.

### Policy-relevant implications and buffers
- High savings in Singapore, together with policy support for households, can buffer consumption compression during the crisis and the recovery phase.
- Even with policy support, consumption recoveries tend to be slower than GDP recoveries after major shocks — suggesting targeted measures to restore consumer confidence and safe spending opportunities may be needed to accelerate consumption normalization.

*Prepared by IMF staff (Jiae Yoo, APD); IMF staff estimates and figures as presented in the chapter.*

### 7.      Consumption levels reflect a variety of socioeconomic conditions that underly

### 1sgpea2021002 - 7.      Consumption levels reflect a variety of socioeconomic conditions that underly

### Key determinants of consumption and saving
- Precautionary saving: Greater uncertainty about future income incentivizes precautionary saving and increases the aggregate saving rate (Carroll et al, 2011; Mody et al, 2012).
- Life-cycle demographics: Households save during working years and dissave during retirement (Modigliani and Brumberg, 1954 and 1980; Modigliani, 1986). Aggregate consumption rates tend to be higher where the share of working-age population is lower; faster aging increases current saving incentives.
- Other structural factors: Income level, credit availability, and interest rates also help explain cross-economy variation in aggregate consumption rates.

### Singapore-specific structural profile
- Income and GNDI:
  - Gross national disposable income (GNDI), which remains within Singapore for consumption and investment, is about 86 percent of GDP.
  - Economies with higher income tend to have a lower consumption rate; Singapore’s per capita income is among the highest in the world in both nominal and PPP terms.
- Demographics:
  - Singapore’s population is young but aging fast. Current old age dependency (OAD) ratio is relatively low, while aging speed (projected change in OAD in 20 years) is among the fastest. Both factors contribute to a lower consumption rate.
- Labor share and corporate sector:
  - Singapore’s labor share in income is among the lowest in the world.
  - Total corporate sector assets amounted to 1,828 percent of GDP in 2019, and the share of foreign enterprises is about 63 percent in terms of total value added in 2019.
  - A large corporate/foreign enterprise presence implies a relatively small share of national income ultimately available for household consumption, so consumption as a share of national income may understate households’ propensity to consume.
- Income uncertainty:
  - Higher income uncertainty in Singapore, measured by the forecast error of real per capita GDP growth, contributes modestly to lower consumption rates. The aggregate uncertainty measure likely captures a limited aspect of household-level uncertainties.

### Empirical analysis: cross-country regression (sample: 41 AEs, 1980–2019)
- Estimated baseline relationship (consumption = % of GNDI) includes: income uncertainty; demographics (OAD & aging speed); labor share; controls (income level, growth, credit, real deposit rates); year and country fixed effects.
- Selected regression findings from Annex Table 1.1 (coefficients shown as reported; standard errors in parentheses):
  - Income uncertainty: -0.0662* (0.0309) in column (1); ranges across specifications (e.g., -0.142*** (0.0351) in column (4)).
  - Population growth: -0.410** (0.157) to -1.362*** (0.153).
  - Old-age dependency ratio: 0.211*** (0.0385) to 0.0807* (0.0386).
  - Aging speed: -0.238*** (0.0320) to -0.0697* (0.0279).
  - Labor share: 0.347*** (0.0288) to 0.331*** (0.0307).
  - Per capita income, PPP, lagged in log: -2.831*** (0.566) and -4.595*** (0.659) in alternative specs.
  - Private sector credit: 0.0170*** (0.00332) in specification (6).
  - Sample sizes and fit: N. Obs. vary (e.g., 1302, 1821, 2829, 209), R-sq. range from 0.117 to 0.393 across columns.
- Interpretation: Demographics, labor share, and real per capita income are important in explaining Singapore’s relatively low consumption share of GNDI.

### Crisis interactions: how structural drivers affect consumption compression
- Crisis dummy: defined as year when real GDP growth falls more than one standard deviation below historical average.
- Key findings from Annex Table 1.2 (selected coefficients as reported; standard errors in parentheses):
  - Income uncertainty: around -0.0784** (0.0275) to -0.106** (0.0342) across models.
  - Population growth: about -1.252*** (0.155) to -1.278*** (0.160).
  - Old-age dependency ratio (OAD): positive and significant, e.g., 0.0874*** (0.0262).
  - Aging speed: negative and significant, e.g., -0.0561* (0.0240) to -0.0847*** (0.0255).
  - Labor share: positive and significant, e.g., 0.321*** (0.0276).
  - Per capita income, PPP, lagged in log: -1.982*** (0.519) to -4.216*** (0.603).
  - Crisis dummy interactions:
    - Crisis × OAD: 0.0932** (0.0310) in column (5).
    - Crisis × Aging speed: 0.0803** (0.0302) in column (6).
    - Crisis × Income level: 0.954* (0.443) in column (7).
  - Crisis main effects vary in significance and magnitude across specifications (examples: Crisis = -2.310** (0.749) in column (3); Crisis = -2.767*** (0.779) in column (5); Crisis = -10.37* (4.704) in column (7)).
  - N. Obs. example: 1262 and 894; R-sq. range: 0.220 to 0.333.
- Implications for Singapore:
  - The negative relationship between consumption share and income level is more muted during crises, suggesting consumption compression could be smaller for high-income economies like Singapore.
  - Demographics have mixed effects in crises:
    - Low current OAD implies greater consumption compression during a crisis (younger populations reduce consumption more).
    - Fast aging (high aging speed) implies smaller consumption compression during a crisis (stronger saving motives for future old-age consumption remain salient).
  - For Singapore, the mitigating effect of fast aging outweighs the effect of a low current OAD, so demographic characteristics and high income can help mitigate consumption compression during a crisis.

### Household-level expenditure patterns (SG-HES compared to U.S. and U.K.)
- Overall: Singapore’s consumption expenditure as a share of gross household income is relatively low compared to the U.S. and the U.K. (Figure 12).
- Composition highlights:
  - Non-durables and services (food, health, education, personal care, apparel, recreation) share: similar across Singapore, U.S., and U.K.
  - Main gaps driven by housing and transport expenditures (Figure 13).
- Transport:
  - Owned vehicle purchase and maintenance spending in Singapore is about half or less than in the U.S. and U.K., more than offsetting higher public transport spending.
  - Policy drivers: Certificate of Entitlement (COE) requirement for car purchases leads to low car ownership; high-quality, widely available public transport makes public transport affordable.
- Housing:
  - Housing- and shelter-related consumption expenditure in Singapore is about half the level in the U.S. and U.K., with rented housing expenditure especially low on average.
  - Housing policy drivers: public housing widely available at relatively affordable price.
  - Homeownership: 90 percent in 2019, with over 80 percent of Singapore Residents living in public housing.
- Notes on survey measures:
  - SG consumption measures are based on income excluding employer CPF contributions.
  - Some imputed rental or mortgage items are not included in cross-country comparisons as noted in figures.

### COVID-19 effects and near-term outlook
- Pandemic impact:
  - Consumption contracted by 14 percent, y/y, in 2020 despite sizable fiscal support, due to mobility and containment-related restrictions and heightened uncertainties.
  - Recovery expected to be gradual as pandemic-related uncertainties and lingering health risks weigh on consumer sentiment.
- Policy support and near-term drivers:
  - Income subsidies or cash grants may deliver higher consumption as restrictions are eased and households feel more secure about future income prospects.
  - Continued support for job creation during recovery and a robust labor market will help near-term recovery and boost consumption over the medium- to long-term.
- Structural resilience:
  - Singapore’s structural drivers (high income, fast aging, demographic saving motives) can help alleviate consumption compression during crises.
- Digitalization and e-commerce:
  - Pandemic accelerated online shopping and e-commerce globally; e-commerce in Singapore remains below most peer economies.
  - Further digitalization and expansion of e-commerce could create new demand beyond substitution from offline to online and could boost consumption over the medium-term (Kinda, 2019).

### Policy-relevant takeaways
- Structural factors explaining low consumption-to-GNDI in Singapore:
  - High per capita income (PPP), fast aging, low labor share, and modestly higher income uncertainty at the aggregate level.
- Targeted policies to support consumption recovery:
  - Income support (subsidies, cash grants) timed with easing restrictions to restore consumer confidence.
  - Support for job creation and a robust labor market to sustain consumption recovery.
  - Policies facilitating digitalization and e-commerce expansion to unlock medium-term consumption growth.
- Analytical priorities:
  - Incorporate alternative uncertainty measures more directly related to household current and future disposable income to better identify precautionary saving drivers.
  - Use granular household survey analysis to identify item-level drivers (e.g., housing, transport) and the role of public policies (housing and transport) in shaping consumption patterns.

*Italic: Source—IMF staff estimates and analysis as presented in the chapter "Consumption levels reflect a variety of socioeconomic conditions that underly consumer behavior."*

### 2.      We compare the SG-HES key indicators to the corresponding indicators from the U.S.

### 1sgpea2021002 - 2.      We compare the SG-HES key indicators to the corresponding indicators from the U.S.

### Survey comparison and methodology
- Comparative surveys:
  - SG-HES (Singapore Household Expenditure Survey) 2017/18.
  - US-CEX (U.S. Consumer Expenditure Survey) 2018.
  - UK-LCF (U.K. Living Costs and Food Survey) 2018.
- Commonalities:
  - Surveys are broadly consistent in concepts of income and expenditure and categorization of consumption items.
  - Income is defined as gross income before personal taxes across surveys, including employment, business, investment, pension and government transfers.
- Harmonization steps taken for comparability:
  - For SG-HES use income excluding employer CPF contributions where possible to align with US-CEX and UK-LCF (which do not include employers’ contribution to employees’ savings or social securities).
  - Separated pensions and social security contributions out of US-CEX consumption expenditure because SG-HES excludes households’ contribution to Central Provident Fund; UK-LCF also excludes contributions to pension funds.
  - Separated mortgage- and property tax-related payments on owned dwellings from US-CEX’s and UK-LCF consumption expenditure because SG-HES does not include loan payments including mortgage interest.
- Reference to detailed mappings:
  - Annex Tables 2.1 & 2.2 provide detailed comparisons of corresponding concepts and consumption categories across the surveys.

### Conceptual differences in income and expenditure (Annex Table 2.1 highlights)
- Income:
  - SG-HES, 2017/18: Income from all sources, recurrent and regular income from employment, business, as well as income from investment, rental and other sources such as cash contribution from relatives/friends, pension and regular government transfers. Not included: irregular receipts or one-off payments (sale of properties, capital gains, windfalls, lump sum CPF withdrawals).
  - US-CEX, 2018: Money income before taxes, including wages and salaries; self-employment income; social security, private & government retirement; interests, dividends, rental income, and other property income; unemployment and workers' compensation and veterans' benefits; public assistance, supplemental security income, and food stamps; regular contributions for support, and other income. Not included: net changes in assets, lump-sum payments, refunds from estate or trust, insurance.
  - UK-LCF, 2018: Receipts (either monetary or in kind) that are received on a regular basis and are available for current consumption.
- Expenditure:
  - SG-HES: Value of consumer goods and services acquired, used or paid for by a household for the satisfaction of needs and wants. Not included: non-consumption expenditure such as loan repayments, income taxes, purchase of house.
  - US-CEX: Transaction costs of goods and services acquired during the recordkeeping period, including gifts. Not included: purchases directly assignable to business purposes; periodic or installment payments on goods and services already acquired; Mortgage principal paid on owned property.
  - UK-LCF: Total expenditure represents current expenditure on goods and services. Not included: savings or investments (national savings certificates, life assurance premiums, contributions to pension funds), income tax payments, national insurance contributions, mortgage capital repayments and other payments for major additions to dwellings.

### Mapping of consumption categories (Annex Table 2.2 highlights)
- Food:
  - SG-HES: Food at home; Food away from home.
  - US-CEX: Food and non-alcoholic beverages; Food serving services.
  - UK-LCF: Food at home; Food away from home; Catering services.
- Alcohol & tobacco:
  - SG-HES: Alcoholic beverages and tobacco.
  - US-CEX: Alcoholic beverage.
  - UK-LCF: Alcoholic drink, tobacco and narcotics; Tobacco products and smoking supplies.
- Health:
  - SG-HES: Health products; Health services; Health insurance.
  - US-CEX: Medical products, appliances and equipment; Outpatient services; Hospital services; Misc. goods & services: Insurance connected with health.
  - UK-LCF: Drugs; Medical products, appliances and equipment; Medical services; Hospital, convalescent and rehabilitation services; Health insurance; Medical insurance premiums.
- Transport:
  - SG-HES: Vehicle purchase; Maintenance; Public transport.
  - US-CEX: Purchase of vehicles; Operation of personal transport equipment; Gasoline, other fuels, and motor oil; Land transport services.
  - UK-LCF: Vehicle purchases; Operation of personal transport; Other vehicle expenses (incl. insurance); Public and other transportation; Transport services.
- Housing and utilities:
  - SG-HES: Shelter (owned-dwelling, rented-dwelling, other lodging); Utilities; Maintenance and repair of dwelling; Insurance connected with dwelling.
  - US-CEX and UK-LCF contain comparable categories but differ on inclusion of mortgage interest and property taxes; SG-HES excludes loan payments including mortgage interest.

### Labor market impact of COVID-19 (Key findings)
- Pre-pandemic characteristics:
  - Singapore had low unemployment "a little above 2 percent" for many years and a high share of non-resident workers.
  - Non-residents represented 29.3 percent of total employment (excluding migrant domestic workers) in December 2020.
- Unemployment dynamics:
  - Overall unemployment peaked at 3.5 percent in September 2020.
  - Average annual unemployment rate in 2020 was 3.0 percent.
- Sectoral employment changes in 2020:
  - Manufacturing lost 37,800 workers.
  - Construction lost 51,800 workers.
  - Services lost 91,100 workers.
  - Employment in construction declined by 11.3 percent in 2020.
  - Employment in manufacturing declined by 7.8 percent in 2020.
  - Employment in services declined by 3.2 percent in 2020.
- Non-resident impacts:
  - Non-resident retrenchment rate: 1.6 percent of non-resident employees were retrenched in 2020.
  - Resident retrenchment rate: 1.1 percent for residents.
  - Non-resident employment share fell to 29.3 percent in December 2020 from 33.1 percent in December 2019.
  - Non-resident employment loss in 2020: 13.7 percent.
- Self-employed:
  - Self-employed groups (taxi and private hire car drivers, freelancers, vendors, direct sellers) were severely impacted.

### Policy response and programs (design, coverage, and budgets)
- Overall approach:
  - Government programs initially focused on preserving jobs, later on employment generation and upskilling; measures largely tilted toward low-income residents and residents who lost job or income.
  - Active labor market policies (Workforce Singapore, NTUC’s Employment and Employability Institute) provide job search assistance, Career Conversion Programs, individualized career coaching; Workfare Income Supplement (WIS) supports bottom 20 income percentile.
- Major programs and budgets:
  - Jobs Support Scheme (JSS):
    - Period: Oct. 2019 to Sept. 2021 (extended Apr 2021 to Sep 2021 with reduced co-funding).
    - Budget: S$30 bn (6.3 percent of GDP) in FY2020 and FY2021; S$26.9 bn in FY20.
    - Coverage: More than 140,000 firms employing more than 2 million local workers benefited.
    - Co-funding initially: 25 to 75 percent of the first S$4,600 of gross monthly wages per local employee for 11 months (up to Aug 2020).
    - Subsequent 7-month period (Sep 2020 to Mar 2021): co-funding reduced to 0-50 percent of the first S$4,600.
    - Extended Apr–Sep 2021: co-funding further reduced to 10-30 percent for all firms; Tier 3 firms not covered under extended JSS from either June or September 2021.
    - Firms categorized into tiers with differing support: Tier 1 (aviation, aerospace, tourism, and others), Tier 2 (retail, arts and culture, food services, built environment and others), Tier 3 (biomedical sciences, precision engineering, electronics, financial services, ICT, media, postal and courier, online retail, supermarkets and convenience stores, and others).
  - Wage Credit Scheme (WCS) (2019, 2020, and 2021):
    - Over S$2 billion in wage credits disbursed during the pandemic; over $1 billion disbursed in 2020 to more than 95,000 employers.
  - Foreign Worker Levy (FWL) rebates and waivers (Apr 2020 to Dec 2021):
    - Received S$2.3 bn budget in FY20; more than 62,000 firms benefited (including 15,000 firms in Construction, Marine Shipyard and Process sectors).
  - SEP Income Relief Scheme (SIRS) (Apr 2020 to Dec 2020):
    - Received S$2.0 bn budget in FY20; more than 200,000 self-employed people benefited.
  - SEP Training Support Scheme (Apr 2020 to Mar 2021):
    - NTUC-run; Government allocated S$84 million.
  - Temporary Relief Fund (Apr 2020):
    - Budget S$0.2 bn; served about 450,000 people.
  - COVID-19 Support Grant (May 2020 to Dec 2020):
    - Received S$0.2 bn in FY20; more than 98,000 workers benefited.
  - COVID-19 Recovery Grant (CRG) (Jan to Dec 2021):
    - Open since January 18, 2021; about 10,500 applications approved as of March 22, 2021.
  - Enhanced Workfare Special Payment (2020 and 2021):
    - Eligible low-income workers and self-employed persons have been receiving S$3,000 each.
  - Courage Fund (started Apr 2020, end date not declared):
    - Supports families of healthcare workers, frontline workers, community volunteers, and other workers who contracted COVID-19.
  - Incentives for new hiring (Apr–Aug 2020):
    - Salaries for new hires co-paid up to a maximum 40 percent for eligible new hires; subsequently replaced by Jobs Growth Incentive (JGI).
  - Jobs Growth Incentive (JGI) (Sept 2020 to Sept 2021):
    - Benefitted about 130,000 new local hires in its first three months (Sep - Nov 2020).
  - SGUnited Jobs and Skills Package (May 2020 to Mar 2022):
    - Nearly 93,000 applicants placed in jobs and skills opportunities as of end-Feb 2021.
    - Funds set aside in 2021 to provide 35,000 additional opportunities.

### Policy implications and outcomes
- Targeting and distribution:
  - Policy support was concentrated on residents; this mitigated resident unemployment but shifted burden of job losses to non-residents.
  - Measures were largely tilted toward low-income and vulnerable groups, with multiple programs providing cash grants, wage support, incentives for new hires, and training/upskilling.
- Labor market resilience:
  - Despite the shock, Singapore’s unemployment outcomes remained relatively favorable in cross-country comparison (average annual unemployment rate in 2020 was 3.0 percent).
  - Non-resident employment declines helped cushion impact on resident labor market and overall unemployment rate.
- Gaps and affected groups:
  - Non-resident workers and self-employed individuals experienced larger adverse impacts and faced more limited support coverage relative to residents and employees covered by employer schemes.

*Italic: Source — 1sgpea2021002 - 2. We compare the SG-HES key indicators to the corresponding indicators from the U.S.*

### 9.      In response to recent increases in community infections and a move to Phase 2

### 1sgpea2021002 - 9.      In response to recent increases in community infections and a move to Phase 2

### Jobs Support Scheme (JSS): adjustments and schedule
- Between May 16, 2021 and June 13, 2021 (Phase 2 Heightened Alert until June 13, 2021), JSS support was increased for several sectors affected by tightened social distancing measures:
  - Food and beverage, performing arts, and arts education: raised from 10 to 50 percent.
  - Gyms, fitness studios and other sports facilities: raised from zero to 50 percent.
  - Indoor playgrounds and other family entertainment centers, and affected personal care services: raised from zero to 30 percent.
  - Retail, cinema operators, and museums, art galleries and historical sites: increased from 10 to 30 percent.
- Table 2 (JSS Schedule and Payout Ratio) payout ratios by month of payout for wages paid in specified periods:
  - Apr. 2020 (Oct.-Dec. 2019): Tier 1 75%, Tier 2 50%, Tier 3A 25%, Tier 3B 25%.
  - Jul. 2020 (Feb.-Apr. 2020): Tier 1 75%, Tier 2 50%, Tier 3A 25%, Tier 3B 25%.
  - Oct. 2020 (May-Aug. 2020): Tier 1 75%, Tier 2 50%, Tier 3A 25%, Tier 3B 25%.
  - Mar 2021 (Sep.-Dec. 2020): Tier 1 50%, Tier 2 30%, Tier 3A 10%, Tier 3B 10%.
  - Jun 2021 (Jan.-Mar. 2021): Tier 1 50%, Tier 2 30%, Tier 3A 10%, Tier 3B 0%.
  - Sep. 2021 (Apr.-Jun. 2021): Tier 1 30%, Tier 2 10%*, Tier 3A 0%**, Tier 3B 0%**.
  - Dec. 2021 (Jul.-Sep. 2021): Tier 1 10%, Tier 2 0%, Tier 3A 0%, Tier 3B 0%.
- Actual JSS payout (in S$ bn) to the tiers (up to December 2020): 2, 5, 16.
- Notes that define tiers and special adjustments:
  - Tier 1 includes Aviation and Aerospace, Tourism, Hospitality, Conventions and Exhibitions.
  - Tier 2 includes Food Services, Retail, Arts and Entertainment, Land Transport, Marine and Offshore.
  - Tier 3B includes Biomedical Sciences, Precision Engineering, Electronics, Financial Services, Information and Communications Technology (ICT) and Media; Tier 3A includes others.
  - * During 16 May 2021 to 13 June 2021, increased support ratios applicable at 50 percent for food and beverage, and performing arts and arts education, and 30 percent for retail, cinema operators, and museums, art galleries and historical sites.
  - ** During 16 May 2021 to 13 June 2021, increased support ratios applicable at 50 percent for gyms, fitness studios and other sports facilities, and 30 percent for indoor playgrounds and other family entertainment centers and affected personal care services.

### JSS outcomes and labor-market impact
- The JSS had strong take up and helped achieve significant job retention for residents:
  - Authorities estimate fiscal policy support measures, among which JSS is the largest single measure, have prevented the resident unemployment rate from rising by 1.7 percentage points in 2020.
  - Fiscal measures have helped save or create about 155,000 jobs during 2020–2021 (up to January 2021).
- Distributional effect on job losses:
  - Tier 1 and 2 firms, thanks to JSS support, accounted for 22 percent of overall job losses during April-July 2020, compared to 42 percent during January-March 2020.

### Wage Credit Scheme (WCS) enhancements
- Government co-funding rates adjusted in response to the pandemic:
  - The Government’s share in co-funding for WCS for 2019 and 2020 raised to 20 percent and 15 percent from earlier declared 15 percent and 10 percent, respectively.
  - FY2021 budget extended the scheme up to 2021 with a co-funding rate of 15 percent.
  - Qualifying gross wage ceiling raised from S$4000 to S$5000 for the period 2019-2021.
- Purpose: support workers’ job retention and income growth and preserve employment relationships.

### COVID-19 Support Grant (CSG) and COVID-19 Recovery Grant (CRG)
- CSG (May-December 2020):
  - Eligible applicants aged 16 years or older.
  - Up to S$800 per month for three months in case of job loss or involuntary no-pay leave for three consecutive months.
  - Up to S$500 per month for three months in case of salary loss of at least 30 percent for three consecutive months.
  - As of end-2020, around S$215 million were disbursed to around 98,000 people under the CSG program.
- CRG (January-December 2021) replacing CSG:
  - Applicant age floor raised to 21 years and eligibility extended to self-employed persons.
  - Support payment for job loss reduced to S$700 per month.
  - Added condition that applicants must have actively participated in job search and training efforts.
  - A new Covid-19 Recovery Grant (Temporary) payout between May 16, 2021 and June 30, 2021:
    - Up to S$700 in case of involuntary no pay leave for a month.
    - Up to S$500 in case of 50 percent or more income loss for a month.

### Foreign Workers Levy (FWL) waiver and rebates
- Waiver and rebate program supported employers’ adjustment and worker upkeep during the Circuit Breaker and thereafter:
  - Waiver initially extended for two months for businesses unable to resume operations; subsequently extended till December 2021 for employers in construction, marine shipyard and process sectors.
  - Amount of support gradually reduced as the economy improved; 100% waiver applicable for periods when workers are under Stay-Home Notice (SHN).
- Table 3 (FWL Waiver and Rebate Payments) — Payment due in / Rate of Waiver / Rebate Amount per worker*:
  - June 2020: 100% / S$750
  - July 2020: 100% / S$375
  - August 2020: 100% / S$375
  - September 2020: 100% / S$375
  - October 2020: 75% / S$90
  - November 2020: 50% / S$90
  - December 2020: 25% / S$90
  - January-April 2021: 0%** / S$90
  - May-September 2021: 0%** / S$250
  - October-December 2021: 0% / S$250
- *Workers include S Pass holders (mid-level skilled staff earning at least S$2,500 a month) and Work Permit Holders (semi-skilled foreign workers).
- **However, 100% waiver of FWL is applicable for the period during which workers are under Stay-Home Notice (SHN).

### SGUnited Jobs and Skills (SGUJS) and Jobs Growth Incentive (JGI)
- SGUJS (introduced May 2020):
  - Curated close to 100,000 jobs and skills opportunities, of which more than 40,000 were to be jobs.
  - As of end-December 2020, nearly 76,000 placements into SGUJS opportunities, of which close to 80 percent (59,400) were placed into jobs, exceeding the original target of 40,000 jobs.
  - As at end-February 2021, nearly 93,000 applicants have been placed in jobs and skills upgrade opportunities.
  - Skills opportunities under SGUJS include SGUnited Traineeships (Government co-funds 80 percent of training allowances up to S$2,500), SGUnited Mid-Career Pathways (training allowances up to S$3,800 per month with Government co-funding up to 90 percent for attachments; monthly training allowance of S$1,500 for company training), and SGUnited Skills (monthly training allowance of S$1,200).
  - In 2021, budget set aside to provide 35,000 traineeships, attachments and training opportunities; SGUJS extended through March 31, 2022 with higher support for jobseekers aged 40 years and above.
- JGI (introduced August 2020):
  - Provides up to 18 months of salary support, with enhanced support for workers aged 40 years or more, persons with disabilities, and ex-offenders.
  - Government allocated a total of S$6.2 billion (S$1 billion in FY2020 and S$5.2 billion in FY2021) to the JGI.
  - Phase 1 (eligible hires between September 2020 to February 2021 inclusive):
    - Conditions: An increase in overall local workforce size in the firm and an increase in local workforce size earning ≥S$1,400/month, compared to August 2020.
    - Benefits: For people below 40 years of age: Up to 25 percent of first S$5,000 for 12 months. For people aged 40 years or more, PwDs, and ex-offenders: Up to 50 percent of first S$5,000 for 18 months. For wages paid from March 2021 onwards, up to 50 percent of first $6,000 of wages.
  - Phase 2 (eligible hires between March 2021 to September 2021 inclusive):
    - Conditions: An increase in overall local workforce size in the firm and an increase in local workforce size earning ≥S$1,400/month, compared to February 2021.
    - Benefits: For people below 40 years of age: Up to 25 percent of first S$5,000 for 12 months. For people aged 40 years and above, PwDs, and ex-offenders: Up to 50 percent of first S$6,000 for 18 months.

### Other targeted support and recovery measures
- Temporary Relief Fund: one-off S$500 cash assistance (April 2020) to lower- and middle-income Singapore Citizens/Permanent Residents affected by COVID-19.
- Courage Fund: one-time lump sum of up to S$30,000 to households with at least one Singapore Citizen or PR member who contracted or succumbed to COVID-19 or were on SHN/LOA/HQO and suffered loss of at least 10% household income.
- Enhanced Workfare Special Payment (WSP): S$3,000 paid over two payments of S$1,500 each for eligible recipients; extension and inclusion of lower-wage workers aged 35 and above in 2020 and 2021.
- Support for Self-employed Persons (SEPs):
  - SEP Income Relief Scheme: Cash payment of S$1,000 per month for nine months to eligible Singaporeans (April 2020 to March 2021).
  - SEP Training Support Scheme: Enhanced training allowance at S$10 per hour (raised from S$7.50), plus course fee subsidies up to 90%.
- Hiring Incentive and Enhanced Hiring Incentive (part of SkillsFuture Mid-Career Support Package):
  - Hiring Incentive (1 Apr 2020 — 26 May 2020) and Enhanced Hiring Incentive (27 May 2020 to 31 Aug 2020).
  - Employers hiring a local worker aged 40 and above: salary support of 40% for 6 months, capped at S$12,000.
  - Employers hiring a local worker aged below 40: salary support of 20% for 6 months, capped at S$6,000.
- COVID-19 Driver Relief Fund: S$133 million allocated in 2021 budget to support taxi and private hire car drivers (targeted Jan-June 2021 for taxi main hirers; Jan-July 2021 for private hire car drivers).

### Concluding assessment and labour-market outcomes
- Singapore’s comprehensive labour market policies have been effective during the COVID-19 shock:
  - Programs have helped keep the unemployment rate low and provided targeted support to the most affected sectors and lower income people.
  - Resident employment grew by 14,900 in 2020.
  - Ministry of Finance report indicated the Gini coefficient for residents, after transfers and taxes, fell to 0.375 in 2020 from 0.398 in 2019.
- Overall policy design: programs described as well-targeted and progressive, preserving employment relationships to capture growth opportunities when the pandemic eases.

*Source: IMF staff summary of Singapore labor market and COVID-19 support measures (content unit 1sgpea2021002).*

### 1.      A generally agreed-upon definition of the digital economy is yet to emerge. A narrow

### 1.      A generally agreed-upon definition of the digital economy is yet to emerge. A narrow

### Definition and scope
- Narrow definition: information and communication technology (ICT) sector, including telecommunications, the Internet, ICT services, hardware and software.
- Broad definition: ICT sector plus parts of traditional sectors integrated with digital technology.
- This annex focuses on two areas of digitalization due to definitional diversity and data sparsity:
  - Digitalization of production (robotics).
  - Digitalization of consumption (e-commerce).
  - Broader innovation that supports the supply of digitalization.

### Singapore’s digital usage and stage
- Singapore ranks high on the digital user’s index relative to countries in Asia (Asia has the highest dispersion in digital technologies adoption).
- Singapore is at a mature stage of digital momentum, characterized by relatively smaller increases in the digital usage index.

### Digitalization of production — industrial robots
- Singapore’s operational stock of robots in manufacturing:
  - about 600 robots in 2008.
  - above 17,000 robots in 2018.
- Robot density in manufacturing:
  - about 1 operating robot per 1,000 employees in 2008.
  - 45 operating robots per 1,000 employees in 2018.
- Sector concentration:
  - Semiconductors accounted for 70 percent of all industrial robots in Singapore in 2018.
- Context from literature:
  - Graetz and Michaels (2015): robots may have increased productivity growth by more than 15 percent and account for about one-tenth of the increases in GDP between 1993 and 2007.

### Digitalization of consumption — e-commerce
- Business-to-consumer e-commerce sales in Singapore:
  - 0.3 percent of GDP in 2017.
  - about 0.7 percent of GDP in 2020.
- E-sales growth in 2020:
  - e-sales grew by 32 percent in Singapore in 2020.
- Consumer behavior during/after COVID-19:
  - A Rakuten Insight survey found close to 80 percent of respondents in Singapore stated they would continue to buy products online even when businesses reopen and social distancing is lifted.
- E-commerce readiness and payments:
  - E-commerce in Singapore remained under 1 percent of GDP in 2020, below most peer Asian and advanced economies.
  - Singapore ranks among countries with the highest readiness to e-commerce (indicators include secure internet services, postal reliability, and access to financial accounts).
  - More than 80 percent of firms had adopted e-payments by 2019.

### Innovation, patents, and R&D
- Digital component (ICT sector) is relatively large in Singapore.
- ICT-related patents:
  - Patents in ICT represent close to 40 percent of total patents (IP5 patent families, 2012–15).
- Triadic patent families:
  - Singapore’s share of “triadic” patent families is relatively low compared to peer Asian and advanced economies.
- R&D expenditure:
  - Singapore’s R&D expenditure has been relatively low and broadly constant in recent decades.
  - Contrasts with rapid growth and higher R&D spending in economies such as Korea, Switzerland, and Taiwan Province of China between 2000 and 2018.
- Global competitiveness rankings (2019):
  - Singapore ranked first in the World Economic Forum’s Global Competitiveness Overall Index in 2019.
  - Singapore ranked 21st in research and development and 23rd in scientific publications.

### Digitalization, innovation, and productivity growth — empirical findings
- Literature evidence:
  - Falk and Hagsten (2015): an increase in e-sales by 1 percentage point raises firm-level labor productivity growth by 0.3 percentage points on average in 14 European economies.
  - Yang, Shi, and Yan (2017): e-commerce participation can positively impact firm productivity in China.
  - World Bank (2016): firms using e-commerce in Vietnam had on average 3.6 percentage points higher total factor productivity growth than firms not using e-commerce.
  - Kinda (2019): Asian firms engaged in e-commerce have on average 30 percent higher total factor productivity than other firms.
- Data and sample for sector-level analysis:
  - Labor productivity proxy: value added per worker by industry (OECD database; Singapore data from Singapore Department of Statistics).
  - Sample covers 22 advanced economies during 2000-2019.
  - Sector categorization: 6 sub-sectors — manufacturing; construction; wholesale and retail trade; information and communication; finance and insurance; and business services.
- Sectoral productivity patterns in Singapore:
  - Manufacturing experienced robust labor productivity growth since 2015.
  - Most service sectors (except financial services and information and communications) experienced limited to no labor productivity growth in the same period.
- Empirical strategy (econometric specification):
  - ΔLPijt = α + γ Digijt + δ Xit-1 + φ bi + η Ij + θtt + εit
  - ΔLPijt: annual labor productivity growth rate of country i in sector j at time t.
  - Digijt: digitalization or innovation proxy (e-commerce, robot density, business R&D).
  - Xit-1: controls including sectoral inward FDI as % of GDP and trade to GDP ratio.
  - bi, Ij, tt: country, sector, and time fixed effects.
- Estimation results:
  - Larger share of e-commerce to GDP is associated with higher labor productivity growth.
  - Higher robot density is associated with higher labor productivity growth.
  - Higher business R&D spending is associated with higher labor productivity growth.
  - Results are robust to a variety of tests.
  - No evidence found that digitalization’s association with productivity growth has been larger in more recent years.

### Spillovers and broader effects
- Evidence (from literature) of productivity spillovers:
  - Javorcik (2004): positive productivity spillovers from highly productive foreign firms to local suppliers in Lithuania.
  - Kinda (2012): local firms supplying the most efficient firms have higher productivity, indicating vertical spillovers through backward linkages.
- While difficult to quantify due to data limitations, productivity spillovers from highly digitalized sectors to other sectors may support economy-wide productivity.

### Implications for Singapore — policy directions to foster a smarter recovery
- Key implication:
  - Boosting e-commerce and R&D would support productivity growth, the nascent recovery, and transformation toward a smarter economy.
- Rationale:
  - Singapore has scope for significant expansion of e-commerce and larger R&D, both associated with higher labor productivity growth.
  - Further digitalization has the potential to boost aggregate productivity growth and lift medium-term growth prospects.
- Existing Singapore initiatives to support digitalization (examples included in the source):
  - Infocomm Media Development Authority’s (IMDA) SMEs Go Digital program with components such as Start Digital Pack, Industry Digital Plan, Grow Digital, SME Digital Tech Hub, Productivity Solutions Grant (PSG).
  - IMDA programs to accelerate digital innovation (Open Innovation Platform).
  - Multi-prong approach to facilitate 5G development (5GSG): facilitating applications in selected industries, supporting a regulatory framework, and grooming talent.

*Source: 1sgpea2021002*

### 16.      In addition to boosting productivity growth, advancing the digitalization of the

### E-PAYMENTS DURING THE COVID-19 PANDEMIC AND BEYOND

### Digitalization, productivity, and distributional challenges
- Advancing the digitalization of the economy can increase tax revenues; the digital tax base is still modest in most countries, including Singapore, but the rapid growth of the sector presents an opportunity to raise additional revenues in the medium to long-term if a tax system commensurate to the digital economy is in place. (Paragraph 16)
- Digitalization brings productivity opportunities and challenges, including labor displacements and possible rise in inequality, calling for complementary labor market policies such as skills upgrades and training. Singapore is noted to be at the forefront in designing such policies. (Paragraph 17)

### A. Cash Use and E-payments in Singapore — key trends and statistics
- Historical infrastructure milestones:
  - Credit cards in the 1970s; GIRO and EFTPOS in the 1980s; contactless EZ-Link cards in 2002; FAST instant payment in 2014; PayNow in 2017; SGQR codes in 2017. (Paragraph 1)
  - Currently, 23 banks participate in FAST, of which nine offer PayNow. (Paragraph 1)

- Household e-payment adoption (2017/18 Household Survey):
  - 36.7 percent of household expenditure is made through e-payments. (Paragraph 2)
  - Share of e-payment expenditure by category (2017-18): Others 30.5; Food 25.8; Transport 44.7; Clothing & Footwear 48.9; Communication 63; Educational Services 64.6; Accommodation Services 76.8. (Paragraph 2; Figure 2)

- FAST and PayNow usage and pandemic effects:
  - FAST volume increased on average exceeding 50 percent per year since 2016. (Paragraph 3)
  - FAST transaction monthly volume rose from 15.5 million to 24.5 million in 2020. (Paragraph 3)
  - FAST average transaction size declined from S$1,592 to S$1,434 in 2020. (Paragraph 3)
  - PayNow user counts and volumes in 2020:
    - Mobile phone number users increased by 30 percent, to around 2.8 million users. (Paragraph 3)
    - National ID number users grew by 80 percent (YoY), to around 2 million users. (Paragraph 3)
    - Unique entity numbers for companies doubled in 2020 to around 250,000 users. (Paragraph 3)
    - Individual PayNow transaction volumes almost doubled (YoY) to 125 million in 2020. (Paragraph 3)
    - PayNow Corporate volumes increased six times (YoY) to 19 million in 2020. (Paragraph 3)
  - SGQR deployments: more than 120,000 SGQR labels deployed during 2020. (Paragraph 3)
  - Inter-bank GIRO volumes remained broadly stable in 2020; cheque clearing volumes continued structural decline. (Paragraph 3)

- ATM and cash use:
  - Over 220 retail bank branches and more than 2,660 ATMs located across Singapore. (Paragraph 4)
  - ATM density stable at 59 per 100,000 residents from 2012 to 2019. (Paragraph 4)
  - Cash withdrawals per month fell sharply during the COVID pandemic. (Paragraph 4; Figure 6)
  - Cash is estimated to be used in around 60 percent of all retail transactions in Singapore (KPMG, 2016). (Paragraph 6)
  - Currency in circulation as proxy for cash use is around 10 percent of GDP in Singapore and has been relatively stable over the past five years. (Paragraph 6)
  - Cross-country comparisons: Sweden 1 percent of GDP; Japan over 20 percent of GDP; BIS CPMI average around 8 percent of GDP. (Paragraph 6; Figure 8)

- Automated Clearing House payment system values (selected table entries):
  - Singapore Dollar Cheque Clearing:
    - Volume (mn): 2018: 52; 2019: 45; 2020: 30.
    - Value (SGD mn): 2018: 589,606; 2019: 505,603; 2020: 381,012.
    - Average Size: 2018: 11,339; 2019: 11,236; 2020: 12,700.
  - Inter-bank GIRO:
    - Volume (mn): 2018: 109; 2019: 112; 2020: 109.
    - Value (SGD mn): 2018: 418,592; 2019: 451,438; 2020: 453,509.
    - Average Size: 2018: 3,840; 2019: 4,031; 2020: 4,161.
  - FAST:
    - Volume (mn): 2018: 60; 2019: 93; 2020: 147.
    - Value (SGD mn): 2018: 108,232; 2019: 148,055; 2020: 210,795.
    - Average Size: 2018: 1,804; 2019: 1,592; 2020: 1,434.
  - (Source: CEIC and Fund Staff Calculations; Table 1)

### B. What explains cash use? — empirical model and results
- The paper models currency in circulation (CiC) for 20 advanced and emerging market economies (BIS CPMI participants) as a function of:
  - GDP per capita (GDP_pc), financial and macroeconomic uncertainty index (FCI), short-term interest rate (Int), average age of population (Age), and digitalization proxy (Dgt, IMD Digital Competitiveness Index). The model is estimated on annual data, 2012-2019. (Paragraph 8)

- Key panel regression results (Table 2: Dependent variable cic):
  - GDP_pc coefficient: -2.26*** (0.58)
  - Fci coefficient: 0.02 (0.14)
  - Int coefficient: 0.004 (0.02)
  - Age coefficient: 0.80*** (0.18)
  - Dgt coefficient: -1.52 (5.61)
  - Observations: 130
  - R2: 0.34
  - Adjusted R2: 0.19
  - F Statistic: 10.77*** (df=5; 106)
  - Note: **p<0.1; **p<0.05; ***p<0.01. (Paragraph 8; Table 2)

- Interpretation of results:
  - Average age of the population is positively related to demand for cash. (Paragraph 9)
  - Currency in circulation is inversely related to GDP per capita. (Paragraph 9)
  - Interest rate, digitalization across countries, and the measure of macroeconomic uncertainty are not significant in this specification. (Paragraph 9)

### C. Policies to decrease cash use and foster e-payments
- Potential productivity benefits:
  - Wider adoption of e-payments can reduce costly cash processing, enable faster payments, and provide data on payment behaviors that support productivity gains. (Paragraph 10)

- Policy directions to promote e-payments while ensuring inclusion, competition, and innovation:
  - Building an inclusive e-payment infrastructure:
    - Ensure basic digital skills and enhance financial literacy for older and poorer segments of the population.
    - MAS launched a public awareness campaign in March 2021, “E-Pay Okay”, to promote safe, simple and speedy e-payments. (Paragraph 11)
    - “Seniors Go Digital” program helps senior citizens learn basic digital skills, including e-payments. (Paragraph 11)
    - Some countries explore Central Bank Digital Currencies (CBDC) partly to ensure e-payment services for underserved populations (Armelius et al, 2020 a and b). (Paragraph 11)

  - Introducing common standards for interoperability and competition:
    - Singapore’s payments infrastructure historically had non-interoperable systems, slowing e-payment uptake (KPMG, 2016).
    - Allowing users to carry data across platforms while protecting data privacy can provide the “public good” of common standards to foster competition. (Paragraph 11)
    - MAS Project Ubin explored blockchain-based settlement systems and developed a blockchain-based prototype for digital multi-currency settlement. (Paragraph 11)

  - Domestic and international policy collaboration:
    - MAS collaborates with Infocomm Media Development Authority (IMDA) and Enterprise Singapore (ESG) to encourage e-payments among specific merchant segments, e.g., SMEs. (Paragraph 11)
    - MAS pioneered a link with the Bank of Thailand (launched end-April 2021) to connect national retail fast payments, reducing international transfer times to minutes rather than one-to-two days. (Paragraph 11)
    - A joint project by the BIS Innovation Hub and MAS aims to enhance cross-border payments, including transactions using multi-CBDCs. (Paragraph 11)

*International Monetary Fund — Excerpt from "E-Payments During the COVID-19 Pandemic and Beyond" (Singapore country chapter).*

### References

### 1sgpea2021002 - References

### Reference list
- Adrian, T. (2020), “Evolving to Work Better: Public Private Partnerships for Public Payments”, July.  
- Armelius, H., G. Guibourg, A. Levin, and G. Soderberg (2020a), “The Rationale for Issuing E-Krona in the Digital Era”, Sveriges Riksbank Economic Review, November 2020.  
- Armelius. H. G. Guibourg, S. Johansson, J. Schmalholz (2020b), “E-Krona Design Models: Pros and Cons and Trade Offs” Sveriges Riksbank Economic Review, November 2020.  
- Baumol, W. J. (1952) “The Transactions Demand for Cash: An Inventory Theoretic Approach.” Quarterly Journal of Economics 66 (4): 545–56. 
- Bech, M., U. Furuqui, F. Ougaard, and C. Picillo (2018), “Payments are a-changin’ but cash still rules”, BIS Quarterly Review, March.  
- IMF (2020), “Digital Money Across Borders: Macrofinancial Implications”, IMF Staff Report, November. 
- IMF (2017), “Is Growth Bank?” Global Financial Stability Report, October.  
- Keynes, J (1936): “The general theory of employment, interest and money”, Macmillan. 
- KPMG (2016), “Singapore Payments Roadmap: Enabling the Future of Payments”, August. 
- Menon, R. (2016), “An Electronic Payments Society”, Speech at Sim Kee Boon Institute Conference on FinTech and Financial Inclusion, Singapore, August. 
- Monetary Authority of Singapore (2021), “Reply to Question on Installation of ATMs in Non-Mature Estates”, https://www.mas.gov.sg/news/parliamentary-replies/2021/reply-to-parliamentary-question-on-installation-of-atms-in-non-mature-estates 
- Png, I. and C. Tan (2020), “Privacy, Trust in Banks and Use of Cash”, MAS Macroeconomic Review, April.  
- Rogoff, K (2015), “Costs and Benefits to Phasing Out Paper Currency”, NBER Macroeconomics Annual, Vol. 29(1), pp. 445–456. 
- Singapore Statistics (2019), “Prevalence of E-Payment Transactions – Findings from the Household Expenditure Survey 2017/18, Singapore Statistics Newsletter”, Issue 2. https://www.singstat.gov.sg/-/media/files/publications/households/ssn219-pg1-5.pdf 
- Williams, J and C Wang (2017): “Reports of the death of cash are greatly exaggerated”, San Francisco Fed blog, 20 November 2017. 
- World Bank (2021), “Automated Teller Machines (ATM) per 100,000 people”, https://data.worldbank.org/indicator/FB.ATM.TOTL.P5

### Overview of Green Finance in Singapore
- Policymakers, investors, and firms in Singapore have started to act on climate change risks, including the mainstreaming of green financial solutions.
- Green finance market characteristics:
  - Modest in absolute value but growing fast.
  - Singapore accounts for about half of cumulative ASEAN green debt issuances.

### A. Green Finance Globally — Key findings
- ESG and sustainable finance integration:
  - ESG (Environmental, Social, Governance) considerations are increasingly integrated in decision-making by investors, firms and policymakers.
  - Environmental drivers: rising concerns about climate-related physical risks and awareness of transition risks.
- Role of financial system:
  - Financial system mobilizes resources needed for climate mitigation and adaptation.
  - Investors can allocate capital toward lower carbon activities, engage with company management, advocate for low-carbon strategies, and lend to firms committed to sustainability.
- Market scale and growth:
  - GSIA estimates at least US$30.7 trillion of funds held in sustainable or green investments in early 2018, up by 34 percent from 2016 in the five largest markets.
  - As of 2019, ESG assets under management in Singapore grew to about US$700 billion (about 28 percent of total).
  - Global green debt issuances grew cumulatively to US$2.2 trillion in 2020, compared with around US$31 billion in 2012.
  - Green debt in APAC and ASEAN increased to around US$433 billion and US$56 billion, respectively, from US$6 billion and US$1 billion.
  - Singapore accounted for 55 percent of ASEAN issuances, and its share relative to global green debt issuances stood at around 1.4 percent in 2020.

### B. Sustainable Finance Initiatives and Regulatory Support — MAS actions and programs
- MAS policy and guidance:
  - MAS issued the Guidelines on Environmental Risk Management for financial institutions (FIs) in December 2020.
  - Guidelines cover banks, insurers, and asset managers and expect FIs to:
    - establish a framework and build capacity to identify, measure, monitor and manage environmental risks beyond just climate change, including through scenario analysis and stress testing;
    - engage with riskier customers and investee firms to transition to more sustainable business practices over time;
    - disclose their approach to environmental risk management in accordance with well-regarded international reporting frameworks, such as TCFD recommendations.
  - Complementary handbook launched in January 2021 by the MAS-convened Green Finance Industry Taskforce (GFIT).
- Stress testing:
  - MAS subjected insurers in 2018 to an extreme flooding scenario.
  - By end-2022, MAS will incorporate a broader range of climate risks in stress tests for the financial industry.
- Grants and market support:
  - Sustainable Bond Grant (launched 2017) subsidizes external review costs for green, social and sustainability bonds; expanded in 2020 to include sustainability-linked bonds.
    - By end 2020, about S$11 billion in green, social and sustainability bonds had been issued in Singapore.
  - 2018 grant scheme to subsidize issuance of insurance-linked securities.
  - 2020 grant scheme for green and sustainability-linked loans to support corporates of all sizes by defraying expenses of engaging independent service providers to validate credentials.
- Reserve management and investments:
  - MAS aims to build a climate–resilient portfolio; analysis with GIC showed negative impacts mitigated by well-diversified portfolio with fixed income instruments largest allocation.
  - Actions include integration of ESG in investment process, overlay strategy within equity assets via benchmark customization, and exclusion of companies susceptible to sharp permanent value drops in transition.
  - MAS announced a US$2 billion Green Investments Program (GIP); US$100 million has been placed in the Bank for International Settlements' Green Bond Investment Pool (GBIP).
- Building capabilities and centers of excellence:
  - MAS facilitated the Singapore Green Finance Centre in October 2020.
  - Sustainable Finance Institute Asia (SFIA) established headquarters in Singapore in March 2021.
  - MAS providing incentives to anchor sustainability teams, external reviewers, and rating agencies.
- Government green bond pipeline:
  - Government identified green infrastructure projects worth S$19 billion (US$14 billion) for which it plans to raise funds through green bonds.
- International cooperation:
  - MAS is a founding member of the Network for Greening the Financial System (NGFS).
  - MAS leads NGFS workstream on micro-prudential practices.
  - MAS is a member of the Sustainable Insurance Forum (SIF) and IOSCO Sustainable Finance Taskforce (STF).
  - Singapore joined the International Platform on Sustainable Finance (IPSF) in June 2020.
- Private sector initiatives:
  - Association of Banks in Singapore (ABS) published guidelines for transparent ESG disclosures and minimum standards on responsible financing practices.
  - Singapore Exchange (SGX) introduced mandatory sustainability reporting for listed entities for financial years ending on or after 31 Dec 2017, with reporting on 5 primary components.
  - GFIT launched a public consultation in early 2021 on a taxonomy to identify activities that can be considered green or transitioning to green.
  - In May 2021, Singapore announced a carbon exchange, Climate Impact X (CIX).

### Table summary (MAS' Green Finance Action Plan—Selected Initiatives) — key program details
- Environmental Risk Management Guidelines: Issued in December 2020; standards for governance, risk management and disclosures of environmental risk for banks, asset managers and insurers.
- Stress Tests for Climate-related Risks: 2018 insurer flood scenario; broader climate risks to be incorporated by end-2022.
- Green and Sustainability-Linked Loan Grant Scheme:
  - Defrays 100 percent of expenses of engaging independent service providers to develop green and sustainability frameworks and targets, obtain external reviews, and report on sustainability impact of the loan, capped at S$100,000 per loan.
  - Supports banks by defraying up to 90 percent of costs, capped at S$180,000 per loan.
  - Launched in 2020.
- Sustainable Bond Grant Scheme:
  - Defrays 100 percent of pre-and post-issuance external review costs for green, social, sustainability and sustainability-linked bonds, subject to a cap of S$100,000.
  - Introduced in 2017 and upgraded in 2020.
- Insurance-linked Securities (ILS) Grant Scheme:
  - Funds up to 100 percent of upfront issuance costs of ILS bonds in Singapore.
  - Introduced in 2018 and valid until December 2022.
- Green Investments Program:
  - US$2 billion program launched in 2019 to place funds with public market investment strategies with a strong green focus.
  - US$100 million placed in the BIS Green Bond Investment Pool (GBIP).
- Green Finance Capabilities:
  - Support establishment of sustainability teams, external reviewers, rating agencies, and Centers of Excellence in Singapore.
- Green FinTech:
  - MAS has earmarked S$50 million out of FSTI to support Green FinTech solutions.

### C. Singapore Banks and Environmental Sustainability — actions and disclosures
- Assessment and disclosure:
  - Methodologies to quantify environmental risks are evolving; disclosures are at a nascent stage.
  - Transition risk disclosures:
    - DBS and UOB have disclosed assessments of climate-related risks through scenario analysis of carbon prices.
    - OCBC launched a pilot project focusing on carbon price as a key transition risk and pledged to release TCFD-aligned report in 2021.
  - Physical risk disclosures:
    - DBS launched a 2019 pilot assessment of physical risk for a sample of customers using three climate change impact scenarios on the energy and mining and metals sectors.
    - UOB intends to assess physical risks in 2022.
- Bank strategies and commitments:
  - Domestic banks have adopted responsible financing practices based on ABS guidelines and pledged to cease financing of new coal power plants.
  - UOB prohibited project financing of greenfield thermal coal mines since 2019.
  - DBS committed in April 2021 to reach zero thermal coal exposure by 2039.
  - ESG considerations incorporated across credit lending approvals, capital market activities, risk management and analysis.
  - Banks have developed green finance solutions and reduced their own carbon footprints.
- Market impact and volumes:
  - In 2018-2020, the banks provided about US$25.5 billion in green and sustainability-linked loans.
  - Banks have provided such loans particularly to local real estate developers developing properties with green certification.
- Indicators of increased ESG focus:
  - Two Singaporean banks adopted the Equator Principles.
  - Rising mentions of ‘climate change’ and ‘ESG’ in banks’ annual and sustainability reports.
  - Adoption of green/sustainable bond frameworks.

*International Monetary Fund. 1sgpea2021002 - References*

### 15.      Sustainability became an integral part of investment strategies for many Singapore

### 15. Sustainability became an integral part of investment strategies for many Singapore asset managers, including sovereign wealth funds.

### Adoption by asset managers and sovereign wealth funds
- The number of Singapore asset manager signatories to the United Nations Principles for Responsible Investment (UN PRI) has risen quickly (cumulative 2009–2020; Figure referenced).
- Total assets of ESG-listed funds available in Singapore stood at around US$120 billion as at end-2020, with equity funds accounting for more than 80 percent of total assets.
- ESG fixed income funds in Singapore are small at around 10 percent but are benefitting from an increased recognition of the relevance of ESG in fixed income.
- Temasek Holdings aimed at closing 2020 with carbon neutrality and committed to halve the net greenhouse gas emissions of its portfolio by 2030.
- GIC is incorporating climate change into its portfolio by evaluating how long-term capital market assumptions are affected by climate change drivers and under different scenarios.

### Banks’ approaches to responsible financing and climate risk assessments
- DBS, OCBC, and UOB have integrated ESG into lending and risk management (Table 2 summarizing responsible financing approaches).
- DBS practices and findings:
  - Group Core Credit Risk Policy incorporates principles and approaches to managing ESG issues in lending practices; due diligence primarily by relationship managers and reviewed by credit risk managers.
  - Sector Lending Guides for agricultural commodities, palm oil, chemicals, oil & gas, mining & metals, power generation, infrastructure, animal husbandry & feed, apparel, footwear & textiles.
  - Financing prohibited for activities violating laws or involving forced/child labor, illegal logging, military goods production.
  - Transition risk analysis approaches:
    - Bottom-up assessment assuming a carbon price increase for a sample of listed companies in carbon-intensive sectors (about 9.6 percent of the Institutional Banking Group’s total exposure, covering 60 percent of companies in five sectors).
    - Top-down, portfolio level approach for the rest of the entire corporate lending portfolio.
    - Under the bottom-up approach, DBS found that credit ratings of 40 out of over 400 customers would be reduced by at least one notch.
- UOB findings:
  - Based on a sample of large corporates and SMEs covering about 5 percent of total non-bank loans, under a high carbon price scenario credit rating of companies included in the sample would deteriorate by two notches.
  - Most carbon-intensive sectors like building materials would be hit the most.
  - Similar assessment for the real estate portfolio (about 6 percent of non-bank loans) suggested negligible deterioration of credit ratings.
  - UOB concluded that the resulting credit risk for the bank would be immaterial, while acknowledging methodological and data limitations.
- OCBC practices:
  - Integrates ESG Risk Assessment Framework in lending and capital markets activities; developed exclusion lists and sector policies; provides green financing solutions linked to Sustainable Development Goals.
- Equator Principles noted as a risk management framework adopted by financial institutions for environmental and social risks in large-scale development projects.

### Corporate environmental disclosures and reporting
- SGX-listed companies must prepare a yearly sustainability report (¶12).
- Compliance:
  - 99.5 percent of companies required to report by December 31, 2020 issued sustainability reports, either standalone or as part of annual reports.
- Climate change materiality and reporting framework usage:
  - One third of listed issuers mentioned climate change as material in their 2020 reports, up from around 6.5 percent in 2018.
  - Only 2 percent of listed issuers used climate change specific reporting frameworks such as TCFD, Carbon Disclosure Project, or Science Based Targets Initiative.
- Disclosure quality metric:
  - The median Bloomberg Environmental Disclosure Score of the Straits Times Index constituents rose to 40 in 2019 from 12 in 2012.
  - Singaporean corporates’ environmental disclosures are comparable to US levels but below EU peers.
- SGX is looking to include recommendations by the TCFD within its guidelines to help listed companies with climate-related financial disclosures.
- Global challenge: lack of consistent methodologies and reporting standards makes it challenging for investors to adopt ESG considerations in their investment process.

### Green, sustainable, and insurance-linked markets
- Green bond listings on the SGX:
  - Cumulative number of green bond listings on the SGX grew to 103 valued at US$44 billion in 2020 from just one listing valued at US$500 million in 2013.
  - Green bond listings on SGX are small compared to Frankfurt and Luxembourg exchanges.
  - Green bonds listed on the SGX were largely denominated in US dollars and issued by foreign entities mainly from Asia and other regions.
  - The average issuance size was about US$200-800 million.
- Sectoral coverage:
  - The sectoral coverage expanded over time in all segments of green and sustainable finance; real estate projects are present in all types of funding.
- Insurance-linked securities and catastrophe bonds:
  - There have been 14 insurance-linked securities launched in Singapore.
  - The first catastrophe bond (cat bond) in Singapore was issued in February 2019 by Insurance Australia Group.
  - Two World Bank Philippines cat bonds were listed on the SGX on November 2019 (the first ever on SGX and on an Asian exchange); the World Bank bonds finance up to US$225 million of protection against earthquakes and cyclones in the Philippines over the next three years.

### Conclusion and policy implications
- The growing green finance market provides an opportunity for Singapore to support climate-friendly capital allocation.
- As a well-developed financial hub, Singapore is already attracting a good proportion of green finance related flows in Asia.
- Monetary Authority of Singapore (MAS) can facilitate these trends by:
  - Enforcing prudent climate risk management and disclosure.
  - Supporting the development of green and sustainable finance solutions.
- The private sector has a key role to play in shifting investment priorities by incorporating climate and sustainability considerations in strategies.

*Source: IMF staff compilation from the provided chapter content.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1sgpea2021002.pdf_