## insea2023007

## Source details

**Canonical URL:** [insea2023007](https://www.imf.org/-/media/files/publications/imf-notes/2023/english/insea2023007.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/imf-notes/2023/english/insea2023007.pdf.md)
- [Structured JSON version](/-/media/files/publications/imf-notes/2023/english/insea2023007.pdf.json)

---

### Overview
- Title: Expanding and Improving Social Safety Nets through Digitalization: Conceptual Framework and Review of Country Experiences.
- Authors: Nicolò Bird and Emine Hanedar.
- Publication: IMF Note NOTE/2023/007.
- Date: December 2023.
- Focus: How digital technologies can support expansion and improvement of social safety nets (SSNs) by enabling governments to identify, verify, and pay beneficiaries; includes six country case studies (Brazil, Democratic Republic of Congo, India, Pakistan, Togo, Türkiye) and broader regional analysis.

### Regional spending and disparities in SSNs
- Emerging and developing economies have significant room to strengthen SSNs; SSNs are key in addressing current and future poverty.
- Regional comparisons of government spending on SSNs (percent of GDP):
  - ED-Europe and advanced economies: average spending about 2 percent of GDP.
  - Lesotho: almost 5½ percent of GDP.
  - Cameroon, Guinea-Bissau, South Sudan: less than 0.2 percent of GDP.
  - Canada, Finland, France: almost 4 percent of GDP.
  - Lithuania, Korea, the Slovak Republic, United States: less than 1 percent of GDP.
- Regions with lowest regional spending on SSNs: ED-Asia and sub-Saharan Africa, followed by Latin America and the Caribbean; the Middle East, North Africa, Afghanistan, and Pakistan; and the Caucasus and Central Asia.

### Three assessment dimensions of SSNs
- Spending adequacy: whether spending levels are adequate to achieve poverty alleviation and human capital objectives (IMF 2019; IMF 2022).
- Spending efficiency: whether spending achieves objectives in a cost-effective manner; concerns include low targeting efficiency and leakages to wealthier income groups.
- Fiscal sustainability: whether SSNs are financed without undermining debt sustainability.

### Trifecta of SSNs: identify, verify, pay
- Identification systems: digital passports, biometric data, Social Security numbers, login credentials, bank account information.
- Verification (socioeconomic data): income, age, household composition, education level, occupation, assets, home ownership, ownership of durable goods.
- Payment delivery infrastructure: bank accounts, mobile money, e-wallets, digital vouchers, one-time passports, smart cards.
- Implications:
  - Adequacy and coverage improve when identification, socioeconomic data, and payment systems are in place.
  - Spending efficiency improves through better targeting; weak ID systems increase risks of beneficiary duplication, “ghost” beneficiaries, fraud, and misallocations.
- Digital public infrastructure investments provide platforms for identification, authentication, data exchange, and payments.

### Digitalization and SSN coverage during COVID-19
- Countries that used digital databases and trusted data sharing to identify beneficiaries reached on average 51 percent of the population.
- Countries that relied on collecting new information from beneficiaries reached on average 16 percent of the population (World Bank 2022c).
- Countries that leveraged digital systems for identification, registration, and payment processes—including Brazil, Togo, Namibia, South Africa, Thailand, Peru, Fiji, Brazil, Pakistan—reached on average 55 percent of the population.
- Thailand example: online application required only national ID numbers; system checked 20 digital databases linked by ID number to confirm eligibility; in spring 2020 Thailand approved about 23 million applications from informal sector workers and farmers—more than half of the working-age population.
- Many governments combined unique ID numbers and existing databases for eligibility determination, enabling digital service windows and mass registration.

### Payment mechanisms and regional patterns
- Most countries that used digital payments relied on either bank accounts or a combination of mobile money and bank accounts, except for countries in sub-Saharan Africa, which relied mainly on mobile accounts.
- Almost 70 percent of countries in sub-Saharan Africa relied on mobile accounts.
- In Latin America and the Caribbean and ED-Asia, more than half of countries relied on bank account payments.
- In the Middle East, North Africa, Afghanistan, and Pakistan, 60 percent of countries relied on bank and mobile account money.
- Most countries that used digital methods to deliver COVID-19 response payments had already implemented these to some extent before the pandemic.
- Several countries (Liberia, Angola, Cameroon) leveraged mobile money accounts and digital SSN payments for the first time during the pandemic.

### Safeguards, limitations, and implementation priorities
- Risks and limitations:
  - Digital divides can lead to exclusion errors whereby eligible beneficiaries do not receive benefits.
  - Digitalization can create new fraud opportunities; safeguards are needed for data quality, privacy, and security.
  - Investments in digital infrastructure (internet connectivity and mobile phone penetration) are key but results may not be uniform across and within countries.
- Implementation priorities and recommended investments:
  - Digital outreach campaigns and complementary efforts (community or in-person attendance, measures to support digital literacy).
  - Ensure privacy and security controls to mitigate fraud risks.
  - Build a clear SSN strategy with targeting criteria and delivery mechanism plans.
  - Invest in strong information systems as part of the broader Govtech agenda.
  - Focus on domestic implementation capacity and institutional arrangements.
  - Seek support from international partners where appropriate.
  - Active monitoring of beneficiary experiences to prevent exclusion of poor and vulnerable groups.

### Coverage gaps and regional challenges
- Sub-Saharan Africa lags in national ID ownership, mobile money, and financial account ownership relative to other regions.
- National ID ownership is high in all regions except sub-Saharan Africa; within sub-Saharan Africa, cross-country differences are larger than in other regions.
- Mobile money ownership in sub-Saharan Africa: lower than other regions’ averages but higher than financial account ownership, highlighting the importance of mobile phones for scaling SSNs in the region.

### Country case studies — summary findings and key figures

- Brazil
  - Pre-pandemic: extreme poverty rose from 3.3 percent in 2014 to 5.4 percent in 2019; unemployment rose from 6.8 percent in 2014 to 12.1 percent in 2019.
  - Pandemic impact: shock led to a 9.2 percentage point increase in extreme poverty and a 5 percentage point increase in labor income inequality.
  - Digital assets: National Social Information Registry (CNIS) initially 17 administrative databases; eligibility later used more than 40 major government databases; CAIXA app and website for registration.
  - Program: Auxilio Emergencial (AEG) emergency cash transfer.
  - Coverage and financial inclusion:
    - AEG reached about 56 percent of the population.
    - Of about 68.2 million AEG beneficiaries during the first implementation cycle:
      - 19.5 million were Bolsa Familia Program (BFP) beneficiaries.
      - 10.5 million were already included in the social registry but did not receive BFP.
      - 38.2 million were online applicants who were not included in the CNIS.
    - Digital responses provided social savings accounts to the roughly 40 percent of AEG beneficiaries without access to a bank account before the pandemic.
  - Challenges: 21 percent of the Brazilian population without internet access experienced digital exclusion; eligibility loosened during AEG (more than 4 percent of GDP).

- Democratic Republic of the Congo (DRC)
  - Program: STEP-KIN all-digital mobile emergency cash transfer (ECT) for urban households.
  - Targeting and registration: geographical targeting using satellite imagery and telecom operator anonymized subscriber lists; whitelist substituted for a beneficiary registry.
  - Payments: telecom operators opened mobile money accounts; broad cash-out networks and dedicated customer services.
  - Coverage and outcomes:
    - In three months, identified, registered, and disbursed transfers to more than 100,000 individuals in 50 poor neighborhoods.
    - Expanded to 250,000 direct recipients, or approximately 1.3 million indirect beneficiaries in 114 poor neighborhoods.
    - STEP-KIN allocated 94 percent of total expenditure to transfers.
  - Transition challenge: moving from emergency ECT to a flagship program requires improved targeting and identification systems, institutional capacity, and government ownership.

- India
  - DBT and the JAM trinity: Jan Dhan, Aadhaar, and mobile communications.
  - Scale and components:
    - Jan Dhan: 470 million bank accounts opened.
    - Aadhaar biometric ID: 1.37 billion Aadhaar numbers issued from an estimated population of 1.42 billion; Aadhaar is a 12-digit unique identification number with demographic and biometric information (fingerprint and iris scan).
    - Mobile communications: more than 1.17 billion phones.
  - DBT coverage and delivery:
    - DBT coverage increased to cover 310 schemes under 53 ministries.
    - PDS accounted for 70 percent of all in-kind program beneficiaries and increased to 750 million beneficiaries in 2021–22.
    - DBT enabled rapid coverage expansion during the pandemic using Aadhaar and Aadhaar-seeded bank accounts.
  - Fiscal savings (cumulative up to March 2021) — selected entries:
    - Department of Fertilizers — Fertilizer: 100 (Indian Rupees, billions); 0.05 (Percentage of GDP)
    - Department of Rural Development — MGNREGS: 335 (Indian Rupees, billions); 0.17 (Percentage of GDP)
    - Ministry of Petroleum and Natural Gas — PAHAL: 729 (Indian Rupees, billions); 0.37 (Percentage of GDP)
    - Department of Food and Public Distribution — PDS: 1,016 (Indian Rupees, billions); 0.52 (Percentage of GDP)
    - Total savings: 2,230 (Indian Rupees, billions); 1.14 (Percentage of GDP)
  - Pandemic program PMGKY:
    - Cash transfers to 320 million beneficiary bank accounts in initial three-month phase.
    - Additional food rations to 236 million households enrolled in PDS database (nearly 72 percent of all households).
    - LPG subsidies to 80 million households.
    - Coverage: about 87 percent of poor households received at least one benefit under PMGKY; overall, 80 percent of all households received at least one benefit.
    - Household coverage breakdown (share of all households): Food only: 40 (All), 38 (Rural), 43 (Urban), 42 (Poor), 39 (Nonpoor); Cash only: 6 (All), 7 (Rural), 4 (Urban), 6 (Poor), 6 (Nonpoor); Food and cash: 34 (All), 40 (Rural), 22 (Urban), 40 (Poor), 32 (Nonpoor); At least one benefit: 80 (All), 85 (Rural), 69 (Urban), 87 (Poor), 76 (Nonpoor).
  - Limits and recommendations:
    - DBT lacks socioeconomic data and cannot by itself target households.
    - Exclusion sources: biometric recognition problems, limited internet connectivity, households lacking digital tools; about half of all women were excluded in some pandemic studies.
    - Recommended: alternative verification methods (one-time-passwords, QR code offline verification, passwords), invest in household digital access, literacy, and skills, leverage Aadhaar for socioeconomic data exchange.

- Pakistan
  - Preexisting systems: National Socio-Economic Registry (NSER) set up in 2011; Computerized National Identity Card (CNIC) biometric ID since 2000.
  - EEC approach:
    - Hybrid targeting combining NSER “known poor” and demand-led SMS application.
    - Eligibility used PMT techniques and CNIC cross-checks with NADRA data sources.
    - Biometric verification for payments; web portal for eligibility checks.
  - Complementary measures: free text messaging, allowing expired CNICs to apply, extended payment deadlines, more than 17,000 pay points, use of SIM registration for mobile money account opening.
  - Outcomes:
    - EEC reached 110 million direct and indirect beneficiaries (roughly half of the population).
    - Total disbursed: $1.2 billion.
  - Gender and delivery:
    - Evidence of gender disparity in receipt of support: 25 percent for women versus 63 percent for men (applying independently for emergency grants).
    - Predominant payment mode remained physical cash.

- Togo
  - Pre-pandemic: contributory social insurance covered approximately 8 percent of the population; mobile subscriptions rose from 2.6 million in 2010 to 6.1 million in 2018; approximately 90 percent of households had access to at least one mobile phone.
  - Novissi design:
    - Mobilized government IDs, phone-based open registration platform, mobile money platform.
    - Used satellite imagery and household consumption to select poorest villages; machine learning on mobile phone metadata and phone surveys to prioritize individuals.
    - USSD platform for on-demand registration developed in less than 10 days; immediate mobile money transfers upon approval.
  - Outcomes:
    - CFAF 13.3 billion (US$ 23 million) disbursed (almost 0.3 percent of GDP).
    - Covered nearly 820,000 beneficiaries (9.5 percent of the population).
    - Phase 1: 567,000 individuals in Great Lomé and Tchaoudjo areas.
    - Phase 2 and 3: reached 5,850 and 244,302 beneficiaries, respectively.
    - Five days elapsed between program announcement and first mobile transfers.
  - Inclusion measures: helplines, telephone registration, local actors/social workers for in-person registration, accessibility features (screen-reader compatibility, sign language).
  - Challenges: ensuring eligibility for those without IDs or SIM cards; exploring low-cost pay-as-you-go mobile phone subsidies.

- Türkiye
  - Preexisting assets: Integrated Social Assistance System (ISAS) consolidated 28 government databases; more than 10 million ISAS applications processed annually pre-COVID-19; national ID, national address system, Social Assistance and Solidarity Foundations (SASFs), Treasury Single Account (TSA).
  - COVID-19 response:
    - Economic Stabilization Support Program: one-time payments worth approximately $150 to approximately 4 million existing SSN beneficiaries (two phases).
    - Horizontal expansion increased eligible recipients from 4 million existing beneficiaries to 12 million households using ISAS and e-Government portal.
    - Payment channels: PTT Bank prepaid cards, bank accounts, home delivery; at-home payment delivery reached almost 50,000 households.
  - Governance and capacity: ISAS success depends on high-level decision-making, information security, staff continuity, and capacity investments.
  - Gender gap: as of 2017, 54 percent of women reported having a financial account compared with 83 percent for men.

### Cross-country takeaways: digitalization, MISs, and emergency SSNs
- Countries with preexisting digital infrastructures and MISs scaled coverage more rapidly and at larger scale than those without such systems.
- Common enablers for rapid expansion and reduced leakages:
  - National IDs and public registries (tax, property, mobile data).
  - Integrated social registries and interoperable databases.
  - Mobilization of existing programs for vertical and horizontal expansions.
- Emergency digital responses were possible even with limited prior MISs (examples: DRC and Togo) by leveraging mobile phone penetration and rapid digital platform development.
- Mitigating exclusion:
  - Active monitoring of beneficiary experiences.
  - Complementary local measures (helplines, in-person support).
  - Targeted efforts to include women, migrants, informal workers, and households without digital access.

### Program coverage and targeting (selected figures)
- STEP-KIN (DRC): 250,000 direct.
- India — DBT: 38 million direct in FY19/20; 17 million direct in FY20/21.
- Pakistan — EEC: 4.5 million households.
- Togo — Novissi: 567,000 direct.
- Türkiye — Emergency cash transfers: 4 million direct; 8 million direct.
- Brazil — AEG: about 68.2 million beneficiaries in first implementation cycle; AEG reached about 56 percent of the population; 19.5 million BFP beneficiaries among AEG recipients.

### Digital registry, enrollment, screening, and payment mechanisms (summary)
- Brazil: Social Registry Yes; Digital Registry and Enrollment Yes; Screening Using National ID Yes; Payment Mechanism: New bank accounts and existing channels.
- DRC: Social Registry No; Digital Registry and Enrollment Yes; Screening Using National ID No; Payment Mechanism: Mobile money.
- India: Social Registry No; Digital Registry and Enrollment No; Screening Using National ID Yes; Payment Mechanism: Bank accounts (almost all).
- Pakistan: Social Registry Yes; Digital Registry and Enrollment Yes; Screening Using National ID Yes; Payment Mechanism: Multipurpose (cash mainly).
- Togo: Social Registry No; Digital Registry and Enrollment Yes; Screening Using National ID Yes (voter ID); Payment Mechanism: Mobile money.
- Türkiye: Social Registry Yes; Digital Registry and Enrollment Yes; Screening Using National ID Yes; Payment Mechanism: Multiple.

### Key statistics and figures cited
- IMF Note number: NOTE/2023/007.
- Publication date: December 2023.
- Regional average SSN spending in ED-Europe and advanced economies: about 2 percent of GDP.
- Lesotho SSN spending: almost 5½ percent of GDP.
- Cameroon, Guinea-Bissau, South Sudan SSN spending: less than 0.2 percent of GDP.
- Canada, Finland, France SSN spending: almost 4 percent of GDP.
- Lithuania, Korea, the Slovak Republic, United States SSN spending: less than 1 percent of GDP.
- Thailand approved about 23 million applications in spring 2020.
- Countries that used database linkages for registration reached on average 51 percent of the population; those collecting new information reached on average 16 percent of the population.
- India: Jan Dhan 470 million bank accounts; Aadhaar 1.37 billion numbers issued from an estimated population of 1.42 billion; more than 1.17 billion phones.
- India fiscal savings cumulative up to March 2021: Total savings 2,230 (Indian Rupees, billions); 1.14 (Percentage of GDP).
- Brazil AEG beneficiaries first cycle: about 68.2 million; 19.5 million BFP beneficiaries among them.
- DRC STEP-KIN: expanded to 250,000 direct recipients; approximately 1.3 million indirect beneficiaries; 94 percent of expenditure allocated to transfers.
- Togo Novissi disbursed CFAF 13.3 billion (US$ 23 million) (almost 0.3 percent of GDP); covered nearly 820,000 beneficiaries (9.5 percent of the population).
- Pakistan EEC: reached 110 million direct and indirect beneficiaries; total disbursed $1.2 billion.
- Türkiye COVID one-time payments: approximately $150 to approximately 4 million existing beneficiaries; horizontal expansion to 12 million households.
- Sub-Saharan Africa: almost 70 percent of countries relied on mobile accounts for digital payments.

*Source: IMF Note NOTE/2023/007 — Expanding and Improving Social Safety Nets through Digitalization: Conceptual Framework and Review of Country Experiences (Nicolò Bird and Emine Hanedar, December 2023).*

### References .............................................................................................................

### insea2023007 - References

### Overview
- Title: Expanding and Improving Social Safety Nets through Digitalization: Conceptual Framework and Review of Country Experiences.
- Authors: Nicolò Bird and Emine Hanedar.
- Publication: IMF Note NOTE/2023/007.
- Date: December 2023.
- Focus: How digital technologies can support expansion and improvement of social safety nets (SSNs) by enabling governments to identify, verify, and pay beneficiaries; includes six country case studies (Brazil, Democratic Republic of Congo, India, Pakistan, Togo, Türkiye) and broader regional analysis.

### Regional spending and disparities in SSNs
- Emerging and developing economies have significant room to strengthen SSNs; SSNs are key in addressing current and future poverty.
- Regional comparisons of government spending on SSNs (percent of GDP) noted:
  - ED-Europe and advanced economies: average spending about 2 percent of GDP.
  - Sub-Saharan Africa example extremes: Lesotho spends almost 5½ percent of GDP; Cameroon, Guinea-Bissau, and South Sudan spend less than 0.2 percent of GDP.
  - Advanced economies example extremes: Canada, Finland, and France spend almost 4 percent of GDP; Lithuania, Korea, the Slovak Republic, and the United States spend less than 1 percent of GDP.
- Regions with lowest regional spending on SSNs cited: ED-Asia and sub-Saharan Africa, followed by Latin America and the Caribbean; the Middle East, North Africa, Afghanistan, and Pakistan; and the Caucasus and Central Asia.

### Three assessment dimensions of SSNs
- SSNs can be assessed across three dimensions:
  - Spending adequacy: whether spending levels are adequate to achieve poverty alleviation and human capital objectives (IMF 2019; IMF 2022).
  - Spending efficiency: whether spending achieves objectives in a cost-effective manner; concerns include low targeting efficiency and leakages to wealthier income groups.
  - Fiscal sustainability: whether SSNs are financed without undermining debt sustainability.

### Trifecta of SSNs: identify, verify, pay
- Digitalization strengthens SSNs by improving:
  - Identification: ability to accurately identify individuals or households (examples include digital passports, biometric data, Social Security numbers, login credentials, bank account information).
  - Verification (socioeconomic data): required for poverty-targeted programs; examples of socioeconomic information include income, age, household composition, education level, occupation, assets, home ownership, ownership of durable goods.
  - Payment delivery infrastructure: inclusive, safe, and transparent delivery mechanisms (examples include bank accounts, mobile money, e-wallets, digital vouchers, one-time passports, smart cards).
- Adequacy and coverage can be substantially improved when an identification system, socioeconomic data, and a payment system are in place.
- Spending efficiency improves through better targeting using socioeconomic data and ID systems; weak ID systems increase risks of beneficiary duplication, “ghost” beneficiaries, fraud, and misallocations.

### Forms of the trifecta and digital public infrastructure
- Identification systems should be universal, reliable, and based on principles of identification for sustainable development.
- Socioeconomic data should be integrated with information systems and kept updated to determine eligibility for targeted benefits.
- Delivery mechanisms must consider households’ ability to use the payment system.
- Investments in digital public infrastructure are crucial to provide platforms for identification, authentication, data exchange, and payments.

### Country experiences and digital adaptations
- During the COVID-19 pandemic, many governments combined unique ID numbers and existing databases for eligibility determination, enabling digital service windows and mass registration.
  - Examples of countries that used database linkages for registration: India, Togo, Namibia, South Africa, Thailand, Peru, Fiji, Brazil, Pakistan.
  - Thailand example: online application required only national ID numbers; system checked 20 digital databases linked by ID number to confirm eligibility; in spring 2020 Thailand approved about 23 million applications from informal sector workers and farmers—more than half of the working-age population.
- Case studies illustrate:
  - Innovative digital technologies can overcome lack of government capacity to implement SSNs, even in countries lacking digital infrastructure or capacity.
  - Countries with stronger preexisting digital infrastructure or investments in SSNs before COVID-19 were able to complement existing policies to reach more people and provide stronger responses than countries without preexisting SSN frameworks.

### Coverage gaps and regional challenges
- Sub-Saharan Africa lags in national ID ownership, mobile money, and financial account ownership relative to other regions.
  - National ID ownership: high in all regions except sub-Saharan Africa; within sub-Saharan Africa, cross-country differences are larger than in other regions.
  - Mobile money ownership in sub-Saharan Africa: lower than other regions’ averages but higher than financial account ownership, highlighting the importance of mobile phones for scaling SSNs in the region.

### Key statistics and figures cited in the text
- IMF Note number: NOTE/2023/007.
- Publication date: December 2023.
- Regional average SSN spending in ED-Europe and advanced economies: about 2 percent of GDP.
- Lesotho SSN spending: almost 5½ percent of GDP.
- Cameroon, Guinea-Bissau, South Sudan SSN spending: less than 0.2 percent of GDP.
- Canada, Finland, France SSN spending: almost 4 percent of GDP.
- Lithuania, Korea, the Slovak Republic, United States SSN spending: less than 1 percent of GDP.
- Thailand approved about 23 million applications in spring 2020.

*Source: IMF Note NOTE/2023/007 — Expanding and Improving Social Safety Nets through Digitalization: Conceptual Framework and Review of Country Experiences (Nicolò Bird and Emine Hanedar, December 2023).*

### 1. National ID Coverage 2. Mobile Money Ownership 3. Financial Account Ownership

### 1. National ID Coverage 2. Mobile Money Ownership 3. Financial Account Ownership

### Digitalization and SSN coverage during COVID-19
- Countries that used digital databases and trusted data sharing to identify beneficiaries reached on average 51 percent of the population.
- Countries that relied on collecting new information from beneficiaries reached on average 16 percent of the population (World Bank 2022c).
- Countries that leveraged digital systems for identification, registration, and payment processes—including Brazil, Guatemala, Iran, Namibia, Mauritius, Pakistan, South Africa, Thailand, Togo, and Zimbabwe—reached on average 55 percent of the population.
- Several countries lacking data to scale up cash transfer coverage were still able to make digital payments quickly by leveraging mobile money markets, notably in sub-Saharan African countries such as Malawi, Nigeria, Ethiopia, Mozambique, Rwanda, Benin, Ghana, Madagascar, Kenya, and Sierra Leone; this was also the case in Haiti and China.

### Payment mechanisms and regional patterns
- Most countries that used digital payments relied on either bank accounts or a combination of mobile money and bank accounts, except for countries in sub-Saharan Africa, which relied mainly on mobile accounts.
- Almost 70 percent of countries in sub-Saharan Africa relied on mobile accounts.
- In Latin America and the Caribbean and ED-Asia, more than half of countries relied on bank account payments.
- In the Middle East, North Africa, Afghanistan, and Pakistan, 60 percent of countries relied on bank and mobile account money.
- Most countries that used digital methods to deliver COVID-19 response payments had already implemented these to some extent before the pandemic.
- Multiple countries—e.g., Liberia, Angola, and Cameroon—leveraged mobile money accounts and digital SSN payments for the first time during the pandemic, facilitating a long-term shift to a modern and comprehensive SSN framework.

### Safeguards, limitations, and implementation priorities
- Digitalization can help reach eligible households but can create a digital divide and new fraud opportunities; safeguards are needed for data quality, privacy, and security.
- Digital divides can lead to exclusion errors whereby eligible beneficiaries do not receive benefits.
- Investments in digital infrastructure (for example, internet connectivity and mobile phone penetration) are key but results may not be uniform across and within countries.
- Digital outreach campaigns and complementary efforts (for example, community or in-person attendance, measures to support digital literacy) are needed to increase coverage and guarantee social inclusion.
- Governments should ensure privacy and security controls to mitigate fraud risks.
- Gains from digitalization depend on institutional and economic conditions and on the objectives underpinning reforms (Gelb, Mukherjee, and Navis 2020).
- Countries need to invest in information systems and administrative and institutional capacity to strengthen SSNs, including:
  - Building a clear SSN strategy with targeting criteria and delivery mechanism plans.
  - Investing in strong information systems as part of the broader Govtech agenda.
  - Focusing on domestic implementation capacity and institutional arrangements.
  - Seeking support from international partners where appropriate.

### Country case studies
- General context:
  - The COVID-19 shock increased adequacy needs of existing SSN beneficiaries and the number of newly vulnerable groups.
  - Responses took the form of both vertical and horizontal expansions of SSNs.

- Brazil
  - Pre-pandemic trends:
    - Extreme poverty rose from 3.3 percent in 2014 to 5.4 percent in 2019.
    - Unemployment rose from 6.8 percent in 2014 to 12.1 percent in 2019.
    - The shock led to a 9.2 percentage point increase in extreme poverty and a 5 percentage point increase in labor income inequality (Al Masri, Flamini, and Toscani 2021).
  - Digital and administrative assets used:
    - Leveraged the National Social Information Registry (Cadastro Nacional de Informações Sociais, CNIS) initially composed of 17 administrative databases; eligibility verification later used more than 40 major government databases (Brollo and others, forthcoming).
    - Caixa Econômica Federal (CAIXA) created a website and an app for registration for those not in CNIS.
  - Program design and delivery:
    - Emergency cash transfer Auxilio Emergencial (AEG) focused first on existing SSN beneficiaries and expanded horizontally to new beneficiaries.
    - Innovations included simplified free all-digital savings accounts through CAIXA and use of Pix (fast payment system launched November 2020) for payments.
    - Demand-led registration campaigns, assisted registration at post offices, and apps allowing application status checks and additional information submissions.
  - Coverage and financial inclusion outcomes:
    - AEG reached about 56 percent of the population.
    - Of about 68.2 million AEG beneficiaries during the first implementation cycle:
      - 19.5 million were Bolsa Familia Program (BFP) beneficiaries.
      - 10.5 million were already included in the social registry but did not receive BFP.
      - 38.2 million were online applicants who were not included in the CNIS.
    - Digital responses provided social savings accounts to the roughly 40 percent of AEG beneficiaries without access to a bank account before the pandemic.
  - Challenges and risks:
    - Digital registration raised barriers for the 21 percent of the Brazilian population without internet access, resulting in digital exclusion among poorer groups.
    - Eligibility criteria were considerably loosened during AEG, reaching middle income families and triggering discussion on generosity and cost (more than 4 percent of GDP).
    - The government switched from individual to household-level concepts, increasing single-person households and total household counts; data verification and ex-post audits were used to address fraudulent declarations.

- Democratic Republic of Congo (DRC)
  - Context and constraints:
    - Low economic and human development, limited SSN delivery systems, lack of management information systems, and a weak financial ecosystem.
  - Program design and partnerships:
    - Solidarité par Transferts Economiques contre la Pauvreté à Kinshasa (STEP-KIN) was an all-digital mobile emergency cash transfer (ECT) for urban households affected by lockdown measures.
    - Capacity developed through public–private partnerships and external partners including GiveDirectly and Ideas42.
  - Identification and registration approach:
    - Geographical targeting of poor neighborhoods using satellite imagery and flood-prone cartography; all people in selected areas deemed eligible.
    - Telecom operators provided anonymized lists of mobile phone subscribers in targeted areas via cell tower mapping; the whitelist substituted for a beneficiary registry and excluded smartphone users, using mobile phone data as a proxy for poverty status.
    - Self-registration enabled through automated data collection; bulk text/audio messaging and radio campaigns supported outreach.
  - Payments and complementary measures:
    - Telecom operators opened mobile money accounts for beneficiaries and initiated social transfers; broad cash-out networks and dedicated customer services ensured access.
    - Complementary measures: 24/7 hotline grievance mechanism; postdistribution monitoring surveys; Central Bank special accounts for bank deposits or electronic money for eligible beneficiaries without mobile phones.
  - Coverage, cost-effectiveness, and outcomes:
    - In three months, STEP-KIN identified, registered, and disbursed transfers to more than 100,000 individuals in 50 poor neighborhoods.
    - The program expanded to 250,000 direct recipients, or approximately 1.3 million indirect beneficiaries in 114 poor neighborhoods (Bermeo and Bance 2022).
    - STEP-KIN was largely cost effective, with 94 percent of total expenditure allocated to transfers.
    - Payments were used primarily on food needs, health services, education spending, reinvestment in livelihoods, and rent payments.
  - Transition challenges:
    - Moving from an emergency cash transfer to a flagship program requires improved targeting and identification systems, institutional capacity, and government ownership to cover a greater share of poor and vulnerable households.

- India
  - DBT and the JAM trinity:
    - Direct Benefit Transfer (DBT) strategy aims to consolidate delivery of existing SSN schemes based on JAM: Jan Dhan, Aadhaar, and mobile communications (Gelb, Mukherjee, and Navis 2020).
  - Components and scale:
    - Jan Dhan: 470 million bank accounts were opened in both urban and rural areas.
    - Aadhaar biometric ID: A total of 1.37 billion Aadhaar numbers have been issued from an estimated population of 1.42 billion in India; Aadhaar provides each citizen with a 12-digit unique identification number with demographic and biometric information (fingerprint and iris scan).
    - Mobile communications: low-cost mobile communication is provided to more than 1.17 billion phones in India, serving as an information exchange platform especially important for rural areas.

*Source: IMF staff synthesis of World Bank 2022c and IMF staff calculations as presented in the referenced IMF Note.*

### 1. Cash Schemes                  2. In-Kind Schemes

### 1. Cash Schemes                  2. In-Kind Schemes

### Direct Benefit Transfer (DBT) — India: coverage, delivery, and fiscal effects
- DBT coverage increased to cover 310 schemes under 53 ministries.
- Largest-number beneficiaries among cash schemes:
  - Pratyaksh Hanstantrit Labh (PAHAL, LPG fuel subsidy)
  - Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS)
- Public Distribution Scheme (PDS) share among in-kind program beneficiaries:
  - PDS accounts for 70 percent of all in-kind program beneficiaries and increased to 750 million beneficiaries in 2021–22.
- In 2021–22, the largest DBT payments were delivered to MGNREGS and Pradhan Mantri Gramin Awas Yojana (PMAYG).
- DBT mechanisms that reduce leakages:
  - Aadhaar identification numbers linked to beneficiaries to prevent duplication and remove “ghost” beneficiaries.
  - Elimination of middlemen by transferring subsidy value directly to Aadhaar-seeded bank accounts, removing dual pricing mechanisms.
- DBT enabled rapid coverage expansion during the pandemic using Aadhaar and Aadhaar-seeded bank accounts.

### Estimated fiscal savings from DBT and related governance reforms (cumulative up to March 2021) — Table 1
- Department of Fertilizers — Fertilizer: 100 (Indian Rupees, billions); 0.05 (Percentage of GDP)
- Department of Rural Development — MGNREGS: 335 (Indian Rupees, billions); 0.17 (Percentage of GDP)
- Department of Rural Development — NSAP: 5 (Indian Rupees, billions); 0.00 (Percentage of GDP)
- Ministry of Women and Child Development — Others: 15 (Indian Rupees, billions); 0.01 (Percentage of GDP)
- Ministry of Petroleum and Natural Gas — PAHAL: 729 (Indian Rupees, billions); 0.37 (Percentage of GDP)
- Department of Food and Public Distribution — PDS: 1,016 (Indian Rupees, billions); 0.52 (Percentage of GDP)
- Ministry of Minority Affairs — Scholarships: 14 (Indian Rupees, billions); 0.01 (Percentage of GDP)
- Department of Social Justice and Empowerment — Scholarships: 3 (Indian Rupees, billions); 0.00 (Percentage of GDP)
- Various — Others: 12 (Indian Rupees, billions); 0.01 (Percentage of GDP)
- Total savings: 2,230 (Indian Rupees, billions); 1.14 (Percentage of GDP)
- Note: Government estimate attributed most savings to elimination of duplicate and ineligible beneficiaries and termination of the PAHAL dual pricing mechanism.

### Pandemic response using DBT: Pradhan Mantri Garib Kalyan Yojana (PMGKY)
- PMGKY combined preexisting cash and food programs; not a new scheme.
- Initial three-month phase:
  - Cash transfers to 320 million beneficiary bank accounts.
  - Additional food rations to 236 million households enrolled in PDS database, covering nearly 72 percent of all households.
  - LPG subsidies provided directly to bank accounts of 80 million households.
- Coverage statistics during the first months of the pandemic:
  - About 87 percent of poor households received at least one benefit under PMGKY.
  - Overall, 80 percent of all households received at least one benefit under PMGKY.
  - Breakdown of household coverage (share of all households):
    - Food only: 40 (All), 38 (Rural), 43 (Urban), 42 (Poor), 39 (Nonpoor)
    - Cash only: 6 (All), 7 (Rural), 4 (Urban), 6 (Poor), 6 (Nonpoor)
    - Food and cash: 34 (All), 40 (Rural), 22 (Urban), 40 (Poor), 32 (Nonpoor)
    - At least one benefit (food or cash): 80 (All), 85 (Rural), 69 (Urban), 87 (Poor), 76 (Nonpoor)

### Limits, exclusion risks, and recommended system improvements (India)
- Current DBT lacks socioeconomic data and cannot by itself target households.
- Exclusion error sources:
  - Biometric (fingerprint) recognition problems and limited internet connectivity.
  - Households lacking access to digital tools or insufficient digital adaptation.
  - Pandemic response evidence: about half of all women were excluded in some studies.
- Recommended measures:
  - Establish appropriate digital networks (power, internet, mobile connectivity).
  - Implement alternative verification methods (one-time-passwords, QR code offline verification, passwords) when biometric verification fails.
  - Invest in household digital access, literacy, and skills.
  - Active monitoring of beneficiary experiences to prevent exclusion of poor and vulnerable groups.
  - Streamline and redesign social schemes to increase flexibility and include migrants, informal workers, and the urban poor.
  - Leverage Aadhaar identifiers to enable socioeconomic data exchange across schemes.

### Pakistan: digitalization and Ehsaas Emergency Cash (EEC)
- Preexisting features:
  - National Socio-Economic Registry (NSER) set up in 2011 (not updated at EEC rollout).
  - Computerized National Identity Card (CNIC) biometric ID in place since 2000.
  - Prior digitalization: biometric payment system, demand-led SMS campaign, wealth-profiling big data analytics.
- EEC approach:
  - Hybrid targeting: covered “known poor” in NSER and used demand-led SMS application for others.
  - Eligibility used proxy-means testing (PMT) techniques and CNIC cross-checks with NADRA data sources (property registries, travel logs, tax returns).
  - Biometric verification for payments; web portal for eligibility checks.
- Complementary inclusion measures:
  - Free text messaging.
  - Allowing expired CNICs to apply.
  - Extended payment deadlines for biometric verification issues.
  - Increased availability of more than 17,000 pay points (retail shops, biometric ATMs, bank branches).
  - Use of SIM registration information to enable mobile money account opening.
- Outcomes:
  - EEC reached 110 million direct and indirect beneficiaries (roughly half of the population).
  - Total disbursed: $1.2 billion.
- Gender and delivery considerations:
  - Evidence of gender disparity in receipt of support: 25 percent for women versus 63 percent for men (applying independently for emergency grants).
  - Predominant payment mode remained physical cash; digital payment adoption is an avenue for future strengthening.

### Togo: Novissi mobile emergency cash transfer (ECT)
- Pre-pandemic context:
  - Small, fragmented national social protection schemes; contributory social insurance covered approximately 8 percent of the population.
  - Mobile subscriptions rose from 2.6 million in 2010 to 6.1 million in 2018.
  - Approximately 90 percent of households had access to at least one mobile phone.
  - A biometrically deduplicated voter ID existed (from 2011 census efforts) and included occupational status.
- Novissi design and rollout:
  - Mobilized government IDs and introduced a phone-based open registration platform and a mobile money platform.
  - Used high-resolution satellite imagery and household consumption to select poorest villages; machine learning using mobile phone metadata and phone surveys to prioritize individuals.
  - USSD platform for on-demand registration developed in less than 10 days; immediate mobile money transfers upon approval.
- Outcomes and scale:
  - CFAF 13.3 billion (US$ 23 million) disbursed (almost 0.3 percent of GDP).
  - Covered nearly 820,000 beneficiaries (9.5 percent of the population).
  - Phase 1: directly benefitted 567,000 individuals in Great Lomé and Tchaoudjo areas.
  - Phase 2 and 3: reached 5,850 and 244,302 beneficiaries, respectively.
  - Five days elapsed between program announcement and first mobile transfers.
- Inclusion and complementary measures:
  - Helplines and telephone registration; local actors/social workers mobilized for in-person registration.
  - Accessibility features: screen-reader compatibility and sign language options.
  - Challenges: ensuring those without IDs or SIM cards are eligible; exploring low-cost pay-as-you-go mobile phone subsidies to overcome exclusion.

### Türkiye: integrated systems and rapid horizontal expansion
- Preexisting SSN and MIS assets:
  - Integrated Social Assistance System (ISAS) developed since 2010; integrated social registry, beneficiary registries, and payment/monitoring systems.
  - ISAS consolidated 28 government databases (land, auto, income tax, social insurance registries) and standardised paper-based procedures to electronic systems.
  - More than 10 million ISAS applications processed annually pre-COVID-19.
  - National ID system, national address system, Social Assistance and Solidarity Foundations (SASFs), and Treasury Single Account (TSA) used for identification, local intake/grievances, and payment delivery.
- COVID-19 response:
  - Economic Stabilization Support Program included one-time payments worth approximately $150 to approximately 4 million existing SSN beneficiaries (two phases).
  - Horizontal expansion increased eligible recipients from 4 million existing beneficiaries to 12 million households using ISAS and the e-Government portal for applications and text-message notifications.
  - Payment delivery channels: PTT Bank prepaid cards, bank accounts, home delivery; at-home payment delivery reached almost 50,000 households.
  - Digital adoption encouraged online shopping and digital bill payments to reduce in-person interactions.
- Governance and capacity considerations:
  - ISAS success depends on high-level decision-making, information security, staff continuity, and investments in capacity.
  - Potential efficiency improvement: route funds directly from TSA into recipient accounts instead of separate program accounts.
- Gender gap in financial inclusion:
  - As of 2017, 54 percent of women reported having a financial account compared with 83 percent for men.

### Cross-country takeaways: digitalization, MISs, and emergency SSNs
- Countries with preexisting digital infrastructures and MISs scaled coverage more rapidly and at larger scale than those without such systems.
- Common enablers for rapid expansion and reduced leakages:
  - National IDs and public registries (tax, property, mobile data).
  - Integrated social registries and interoperable databases.
  - Mobilization of existing programs for vertical and horizontal expansions.
- Emergency digital responses were possible even with limited prior MISs (examples: DRC and Togo) by leveraging mobile phone penetration and rapid digital platform development.
- Mitigating exclusion:
  - Active monitoring of beneficiary experiences.
  - Complementary measures at local levels (helplines, in-person support).
  - Targeted efforts to include women, migrants, informal workers, and households without digital access.

*Source: IMF Note (content unit insea2023007).*

### 19.5 million direct 49.7 million

### insea2023007 - 19.5 million direct 49.7 million

### Program coverage and targeting
- Democratic Republic of the Congo — Social Transfers in Kinshasa (STEP-KIN)
  - Modality: Horizontal
  - Targeting: Urban poor determined by geographical and mobile phone data
  - Social registry: None
  - Coverage reported: 250,000 direct
- India — Direct Benefit Transfer (DBT)
  - Modality: Horizontal and vertical
  - Targeting: Chronic and temporary poor, with schemes targeted to women
  - Coverage reported: 38 million direct in FY19/20; 17 million direct in FY20/21
- Pakistan — Ehsaas Emergency Cash (EEC) Program
  - Modality: Horizontal and vertical
  - Targeting: Chronic and temporary poor based on PMT and social registry data
  - Coverage reported: 4.5 million households
- Togo — Novissi cash transfer
  - Modality: Horizontal
  - Targeting: Informal workers determined by occupation and place of residence
  - Social registry: None
  - Coverage reported: 567,000 direct
- Türkiye — Emergency cash transfers (ECTs)
  - Modality: Horizontal and vertical
  - Targeting: Chronic and temporary informal poor based on social registry data
  - Coverage reported: 4 million direct; 8 million direct

### Digital registry, enrollment, screening, and payment mechanisms (Table 4)
- Brazil
  - Social Registry: Yes
  - Digital Registry and Enrollment: Yes
  - Screening Using National ID: Yes
  - Payment Mechanism: New bank accounts and existing channels
- Democratic Republic of the Congo
  - Social Registry: No
  - Digital Registry and Enrollment: Yes
  - Screening Using National ID: No
  - Payment Mechanism: Mobile money
- India
  - Social Registry: No
  - Digital Registry and Enrollment: No
  - Screening Using National ID: Yes
  - Payment Mechanism: Bank accounts (almost all)
- Pakistan
  - Social Registry: Yes
  - Digital Registry and Enrollment: Yes
  - Screening Using National ID: Yes
  - Payment Mechanism: Multipurpose (cash mainly)
- Togo
  - Social Registry: No
  - Digital Registry and Enrollment: Yes
  - Screening Using National ID: Yes (voter ID)
  - Payment Mechanism: Mobile money
- Türkiye
  - Social Registry: Yes
  - Digital Registry and Enrollment: Yes
  - Screening Using National ID: Yes
  - Payment Mechanism: Multiple

### Key statistics and figures
- Header numeric identifiers: 19.5 million direct 49.7 million
- Democratic Republic of the Congo: 250,000 direct
- India: 38 million direct in FY19/20; 17 million direct in FY20/21
- Pakistan: 4.5 million households
- Togo: 567,000 direct
- Türkiye: 4 million direct; 8 million direct

*Sources: IMF staff based on information compiled from the case studies.*

---


_Source: https://www.imf.org/-/media/files/publications/imf-notes/2023/english/insea2023007.pdf_