Diagnostic and analytical tools to support the extension of social security Social Security Academy, ILO-ITC Turin Elective Course 6 Week 1, Monday to Wednesday September 2013, 14:00-17:00 ILO Social Protection Department, Geneva
Introduction Quick overview Objectives and structure of the seminar
Diagnostic and analytical tools to support the extension of social security: A quick overview Objective Introduction to diagnostic and analytical tools to support the extension of social security…... in order to.... better understand the range of techniques and tools available to support the extension of social security, and be aware of their potential and limitations. reflect on the linkages between social security policies and distribution outcomes be better equipped to make an informed choice among various policy options
Diagnostic and analytical tools to support the extension of social security: A quick overview Questions addressed How to map social protection systems: assess existing social protection benefits/schemes and identify coverage gaps? Which policy options could be considered for closing these gaps? How much would the realization of these policy options cost under different scenarios? What would be the impact of these policy options on the reduction of poverty? How can the results of these tools be used to support policy decisions with regard to the implementation of reforms (prioritization, sequencing, time frames), as well as the coordination of schemes and benefits within the overall social security system?
Diagnostic and analytical tools to support the extension of social security: What this course will achieve This course will Introduce you to some useful diagnostic and analytical tools to support the extension of social security Provide you with an understanding of what can be achieved with these methods and tools, and where are the limitations This course will not provide you with a full training regarding these tools (> separate courses and tailored technical assistance and capacity-building) provide you with an easy ready-made solution – most of the methods and tools will need be adapted for the respective country context
Outline of the course 1. Introduction and overview of the technical assessment process 2. Diagnostic: describe the present (& planned) social protection situation Mapping social protection systems: steps, data Identify legal/ design & implementation gaps: how to? Additional data to be considered as part of the diagnostic Time for practice 3. Analytical tools to support the extension of social security Assessment of the potential impact on poverty reduction of various benefits based on micro-data (ex-ante) Costing of policy options to close coverage gaps (short introduction only) 4. Decision-making with regard to the extension of social security Prioritization, time-frames, sequencing, coordination Day 1 Day 2 Day 3
1. Introduction and overview of the technical assessment process
The two-fold objective of the Recommendation: National social protection floors... 8 (a) Establish and maintain, as applicable, social protection floors as a fundamental element of their national social security systems (b) Implement social protection floors within strategies for the extension of social security that progressively ensure higher levels of social security to as many people as possible, guided by ILO social security standards nationally defined sets of basic social security guarantees which secure protection aimed at preventing or alleviating poverty, vulnerability and social exclusion Provides guidance to members to Horizontal dimension: Guaranteeing access to essential health care and minimum income security for all Vertical dimension: progressively ensuring higher levels of protection guided by C.102 and more advanced standards
... and national strategies for the extension of social security 9 (a) Establish and maintain, as applicable, social protection floors as a fundamental element of their national social security systems (b) Implement social protection floors within strategies for the extension of social security that progressively ensure higher levels of social security to as many people as possible, guided by ILO social security standards nationally defined sets of basic social security guarantees which secure protection aimed at preventing or alleviating poverty, vulnerability and social exclusion Provides guidance to members to Horizontal dimension: Guaranteeing access to essential health care and minimum income security for all Vertical dimension: progressively ensuring higher levels of protection guided by C.102 and more advanced standards Progressively build and maintain comprehensive and adequate social security systems
National social protection floors: At least four nationally-defined guarantees essential health care including maternity care basic income security for children providing access to nutrition, education, care and any other necessary goods and services basic income security for persons in active age unable to earn sufficient income basic income security for persons in old age NATIONAL SOCIAL PROTECTION FLOOR: nationally defined basic social security guarantees NATIONAL SOCIAL SECURITY SYSTEM Higher levels of protection
Extending social security and building comprehensive social security systems 11 individual/household income Social Protection Floor: Access to essential health care and basic income security for all Social security benefits of guaranteed levels Voluntary insurance under government regulation level of protection high low Horizontal dimension: Guaranteeing access to essential health care and minimum income security for all, guided by Recommendation No. 202 Vertical dimension: progressively ensuring higher levels of protection, guided by Convention No.102 and more advanced standards floor level Outcomes can be guaranteed through different means – there is no one-size-fits-all
National social security extension strategies: Formulation and implementation set objectives reflecting national priorities identify gaps in, and barriers to, protection seek to close gaps in protection through appropriate and effectively coordinated schemes specify financial requirements and resources, time frame and sequencing for the progressive achievement of the objectives Raise awareness about social protection floors and extension strategies... support growth of formal employment and the reduction of informality... be consistent with and conducive to the implementation of national social and economic development plans... support disadvantaged groups and people with special needs... complement social security by active labour market policies The strategies should...
Questions to address in each step Which are the elements of the social protection floor / social security system which are already in place? Inventory of schemes/programmes for the four garanties of the SPF, and elements of the wider social security system How far is the country advanced in building its social protection floor? Analysis & identification of protection gaps What are the policy options with regard to the establishment/completion of a social protection floor / social security system? Policy recommendations What would be the possible cost and poverty impact of the considered measures? Scenarios, costing and impact assessment
1. Introduction and overview of the technical assessment process Overview of the ILO technical assessment process (main steps)
PRELIMINARY ASSESSMENT OF COSTS AND IMPACTS ESTIMATE THE COSTS and the IMPACTS How much would it cost to introduce benefits that ensure to fill gaps ? What are the impacts on poverty? DIAGNOSTIC/ MAPPING -Mapping of Existing & planned schemes/programmes: qualitative & quantitative indicators -Gaps and needs assessments (coverage gap analysis, labour market analysis, poverty profile, etc,) COMPARE WITH THE FLOOR GUARANTEES & higher levels Do existing provisions meet the Recommendation principles? Meet income security guarantees? Meet higher levels? Basic income security for children, persons in active age, older persons & Access to essential health care for all IDENTIFY AND ASSESS GAPS and DEFINE OPTIONS for the extension: Design / policy and implementation gaps and options for extension ASSESSMENT MATRIX Diagnostic, assessment of costs & analysis of impacts NATIONAL DIALOGUE ACTUARIAL studies, etc. DESIGN OF NEW SCHEMES/PROGRAMMES Define priorities Fiscal space debate
SPF objectives Existing social protection provisions What is planned un the strategy GapsInvolved agencies Priorities Design gapsImple- mentation gaps Health Children Active age Older persons Diagnostic (2)| Rapid assessment matrix: structure Social protection floor: guarantees and objectives Describe the present and planned social protection situation, taking into account SP strategy objectives Identify design and implementation gaps Identify and share responsibility and activity among actors Priority policy options to be decided through national dialogue on assessment results
2. Diagnostic: describe the present (& planned) social protection situation Main questions to be answered & link to main tools
Mapping | Questions to ask about the overall social protection system Mapping national social protection system | Main questions Mapping national social protection system | Source and tool Coverage Who is covered? by population category: men/women, broad age groups, poor/non-poor, workers in formal/informal employment, other vulnerable groups? Which are the benefits provided (risks/contingencies covered) and which is the level of benefits? Which is the impact (potential/actual) of these benefits? Sending and financing Who is paying (financing sources); How much (how much for the benefits, how much for the administration)? How to? Inventory of existing social protection schemes & programmes in the country Identification, definition of schemes of benefits (categorisation of data) Data collection on delivered benefits, coverage, spending and financing Sources: administrative data from social security schemes as primary source complemented by survey data (coverage, impacts) Tool: ILO social security inquiry
Mapping | Questions to ask about the overall social protection system Estimate coverage gaps and identify coverage needs | Main questions Estimate coverage gaps & identify coverage needs Since the identification of gaps in coverage, together with the reasons for their existence & ways of filling them, are the main objectives, the following questions need to be answered: Who are those not currently covered but in need of coverage? What are their needs? What are the risks they are exposed to? What are the options for the coverage extension? What are their employment situation? In what measure can they contribute? What is the potential cost of the coverage extension? Mainly household surveys (household income & expenditure surveys, Labour force surveys) Considering those who are not or insufficiently covered: Identification of groups presenting different needs or features calling for different (adequate) options regarding social security extension Among key factors to consider: Age, area of residence Labour market structure Ability to work Status in employment Capacity to contribute; Poverty status; etc
2. Diagnostic: describe the present (& planned) social protection situation Mapping national social protection system | Inventory of schemes: main steps & data required
Categorisation of schemes & benefits Benefits featuresSchemes/ programmes definition SP system Older persons Old age Survivors* Family / child benefit Active age Sickness Disability/ Invalidity Unemployment Employment injury Maternity Survivors* ALMP Housing * Other social assistance* Health care benefit Function Cash In- kind Type of benefit periodic Lump-sum/ ad hoc periodic Lump-sum/ ad hoc Periodicity Statutory Non statutory Contributory Non contributory Contributory Non contributory Contributory Non contributory Contributory Non contributory Public Private Public Private Means-tested Not means tested Means-tested Not means tested Means-tested Not means tested Means- tested or not Other criteria Mandatory / voluntary benefit Complementary / main benefit Any other? Means-tested Not means tested
From data to indicators | Older persons* receiving an old age pension above statutory pensionable age Benefits featuresSchemes/ programmes definition SP system Older persons Old age Survivors* Family / child benefit Active age Sickness Disability/ Invalidity Unemployment Employment injury Maternity Survivors* ALMP Housing * Other social assistance* Health care benefit Function Cash In- kind Type of benefit periodic Lump-sum/ ad hoc periodic Lump-sum/ ad hoc Periodicity Means-tested Not means tested Means-tested Not means tested Means-tested Not means tested Means- tested or not Other criteria Mandatory / voluntary benefit Complementary / main benefit Any other? Statutory Non statutory Contributory Non contributory Contributory Non contributory Contributory Non contributory Contributory Non contributory Public Private Public Private Means-tested Not means tested
Diagnostic | Data collection: : define & categorize schemes/ programmes and benefits Step 1 | Inventory of existing programmes/schemes and define/characterise each of them Contributory or non-contributory Public or private Statutory scheme or programme Mandatory or voluntary scheme Universal scheme or not Step 2 | For each scheme/programme: identification of benefits provided Function: Risk or contingency covered (medical care, old age, invalidity, survivors, employment accidents, unemployment family/children, maternity, poverty and social exclusion) => link with SPF basic social security guarantees Benefits in cash or in kind Periodic benefits vs. ad-hoc/one-off benefits Means-tested or non-means-tested benefits “Supplementary” or “basic” benefit
Diagnostic | Data collection: financial data Steps 3 & 4 | Data collection at scheme and benefit levels Financial data: expenditure and revenue at the scheme level Expenditure at a scheme level Benefits (in cash or in kind), Administrative costs, Transfer to other schemes, Other spending Revenue at a scheme level Social contributions (employers, employees/protected persons); Contributions from the government; Transfers from other schemes; Transfers from the rest of the world; Other receipts Financial (expenditure) data at the benefit level Benefit expenditure Minimum benefit if exists Average benefit
Diagnostic | Data collection: coverage data Steps 3 & 4 (cont’d.) | Data collection at scheme and benefit levels Data on coverage Information on the scheme level Target groups of the scheme Statutory coverage (Qualitative & Quantitative) Protected persons (scheme level) Number of persons affiliated by age groups and sex Number of active contributors by age groups and sex Information on benefit level Actual beneficiaries of the benefit (by sex & age) Benefit amount (by sex & age) « Scope of coverage» and « Extent of statutory» coverage Effective coverage « Extent of effective coverage» (1) Effective coverage « Extent of effective coverage » (2) « Level of coverage»
2. Diagnostic: describe the present (& planned) social protection situation Mapping national social protection system | A tool: the ILO social security inquiry
ILO Social Security Inquiry as a tool for mapping social protection systems Covers both statutory schemes and some non-statutory schemes Covers both private and public schemes Type of data | From administrative sources: institutions which administer the scheme. Household survey data if available may be used as well Functions covered Old age; disability; survivors; sickness and health; unemployment; employment injury & occupational disease; family/children, and maternity, housing; other income support and assistance (n.e.c.) /social exclusion. Can be expanded Scope of data & indicators
Screen shots| A “guided tour in the inquiry” Levels of information & sources Inventory of schemes & benefits: qualitative mapping of social protection provision Inventory of schemes & benefits: qualitative mapping of social protection provision Data collected at the scheme and benefit levels on: Data collected at the scheme and benefit levels on: Expenditure and financing Coverage (protected persons & beneficiaries) and level of benefit Coverage (protected persons & beneficiaries) and level of benefit Scheme level National aggregate indicators
Step 1 | List of existing national social protection schemes and characteristics of each of them Contributory, not contributory or universal Public versus private schemes, Statutory scheme or not Compulsory or voluntary Screen shots | A “guided tour in the inquiry” Armenia - An tentative Inventory of schemes Step 1 | Inventory of social security schemes & definition of each scheme
Screen shots | A “guided tour in the inquiry” Armenia - An tentative Inventory of schemes For each scheme First pat | Textual information & direct links Description of the scheme incl. information from SSPTW Step 2 Data collection of coverage data and Financial data at the scheme level Step 3 | Identify and define benefits provided Covered risk or function (health, old age, disability, survivor, work injury, unemployment, family/children, maternity, poverty and social exclusion) => can be linked with the SPF guarantees And collect for each of them data on Expenditure by benefit Beneficiaries Level of benefits Example: expenditure data for the State Social Security Service (SSSS) | State pension in ArmeniaState Social Security Service (SSSS) | State pension Benefit expenditure = Sum individual benefit expenditure
Benefit level | Armenia– Old age pension data collected 1. | Definition of benefit: function, periodicity, cash or in- kind, means tested of not A benefit is defined by: The function covered: old age, survivors, maternity, etc. Basic or supplementary benefit The periodicity: (periodic or not) The type: cash or in kind Beneficiaries: individuals or households Means-tested or not 2. | Expenditure at the benefit level Social protection expenditure by function, as a percentage of Total social protection expenditure GDP 3. | Beneficiaries, if possible by sex & age group Effective coverage « Extent of effective coverage» [2] Actual beneficiaries 4. | Benefit level, if possible by sex & age group Effective level of coverage: average level of benefit related to average wage, minimum wage, poverty line …
Results| Aggregate indicators online Indicators can be consulted from 1. the indicators section (mainly tables) Consultation of all indicators for a given country Depending on availability, data available from 2000 to 2012 for expenditure and coverage related indicators Consultation of All countries for a given indicator 2. From the from page: maps for a selected number of indicators Old age pension beneficiaries as a proportion of persons above retirement age Unemployed receiving unemployment benefits (periodic cash benefits)
2. Diagnostic: describe the present (& planned) social protection situation Identify legal/ design & implementation gaps and provides indicators regarding scope and extent of coverage
How to measure coverage ? | the multiple dimensions of coverage For a given function, coverage is a multidimensional concept with at least three elements: Scope. Measured by the range (number) and type of social security branches / functions to which the population of the country has access by law. Extent. Percentage of persons covered within the whole population or the target group, by social security measures in each specific social security function. Level. This refers to the adequacy of coverage by a specific social security function : Can be measured by the level of cash benefits provided: either absolute or relative to selected benchmark values such as previous incomes, average incomes, the poverty line, and so on. Mainly statutory information Both statutory & effective
The multiple dimensions of coverage Coverage Level of statutory coverage Level of effective coverage Persons protected or ‘potential’ beneficiaries Actual beneficiaries Mainly statutory Statutory versus effective Number & type of social security functions covered by law Extent of statutory coverage Extent of effective coverage Persons legally protected or ‘potential’ beneficiaries (by law) Average amount (effective) of benefit Min, max, basic amount (or percentage) of benefit as stated by law Multiple dimensions Scope Extent Level
Scope and extent of statutory coverage How to? Indicators?
The multiple dimensions of coverage Coverage Level of statutory coverage Level of effective coverage Persons protected or ‘potential’ beneficiaries Actual beneficiaries Mainly statutory Statutory versus effective Number & type of social security functions covered by law Extent of statutory coverage Extent of effective coverage Persons legally protected or ‘potential’ beneficiaries (by law) Average amount (effective) of benefit Min, max, basic amount (or percentage) of benefit as stated by law Multiple dimensions Scope Extent Level
Scope of statutory coverage | Number and type of social security branches / functions covered by law Resident citizens of Zambia: health care as a basic human right guaranteed and accessible to every citizen. Health Social insurance scheme, 1996 law implemented in Employees persons, incl. agricultural workers, household workers in urban areas, apprentices. Voluntary coverage for self-employed. Exclusions: Workers earning less than 15,000 kwacha a month. Old age, survivors, invalidity Employer-liability system, involving compulsory insurance with a public carrier. Employees, incl. casual workers, household workers, apprentices, and public-sector employees. Exclusion: self- employed Work injury No statutory provision Sickness No statutory provision (for cash periodic benefits) Maternity No statutory provision Family allowances No statutory provision Unemployment Zambia In Zambia, 5 (or 4 if excluding health) out of the 9 social security functions of Convention 102 provide by law some social security to specific groups of the population
Scope of statutory coverage | Percentage of countries with statutory provision by branch (basis 178 countries) Scope of social protection | Statutory information Very limited provision: 4 out of the 8 branches considered here Usually Old age, survivors, disability and employment injury Limited legal provision (5/6 branches) Usually no legal provision for social security in case of unemployment, sickness or no family/child benefit Semi- comprehensive 7 out of 8 branches No family benefits Comprehensive social security provision (8/8)
The multiple dimensions of coverage Coverage Level of statutory coverage Level of effective coverage Persons protected or ‘potential’ beneficiaries Actual beneficiaries Mainly statutory Statutory versus effective Number & type of social security functions covered by law Extent of statutory coverage Extent of effective coverage Persons legally protected or ‘potential’ beneficiaries (by law) Average amount (effective) of benefit Min, max, basic amount (or percentage) of benefit as stated by law Multiple dimensions Scope Extent Level Extent of statutory coverage Persons legally protected or ‘potential’ beneficiaries (by law) Extent
Extent of statutory coverage | Objective and examples of indicators Extent of statutory coverage | How to? Estimates of the extent of statutory coverage use both: Information on the groups covered by statutory schemes for a given social security function in national legislation, & Available statistical information quantifying the number of persons concerned at the national level. The statutory extent of coverage rate for a given social security function = Estimated number of people covered by law as appropriate Total number of employees, Total employment; EAP or working age or total population
Extent of statutory coverage: proportion of the working age covered by law for old age pensions Social insurance scheme Coverage according to the law (statutory coverage) Employees, including agricultural workers, household workers in urban areas, apprentices. Voluntary coverage for self-employed and some categories of informal-sector workers who were previously covered for at least 60 months. Exclusions: Workers earning less than 15,000 kwacha a month; armed forces personnel. Old age, survivors and invalidity more Zambia | Example of extent of statutory coverage Source: social security programs throughout the world
Extent of statutory coverage: proportion of the working age covered by law for old age pensions Zambia | Example of extent of statutory coverage Source: LCMS 2010
Effective social protection: extent and level of effective coverage & expenditure Main principles and examples Measurements of effective coverage should reflect how in reality the statutory provisions are implemented.
The multiple dimensions of coverage Coverage Level of statutory coverage Level of effective coverage Persons protected or ‘potential’ beneficiaries Actual beneficiaries Mainly statutory Statutory versus effective Number & type of social security functions covered by law Extent of statutory coverage Extent of effective coverage Persons legally protected or ‘potential’ beneficiaries (by law) Average amount (effective) of benefit Min, max, basic amount (or percentage) of benefit as stated by law Multiple dimensions Scope Extent Level Persons protected or ‘potential’ beneficiaries Actual beneficiaries Extent of effective coverage Extent
Diagnostic (1)| Coverage indicators and coverage gaps | main principles (1) The measure of social protection coverage (and resulting coverage gaps) should follow a set of principles Principle 1 | Coverage indicators by branch or social security function Social security coverage can be directly measured only separately for each of the specific risk or social security function such as health care, old age or unemployment People may have access to health care, but not to old age pensions, or vice versa. A differentiated approach to measuring coverage is necessary. Note: Aggregate coverage measures such as the ADB Social Protection Index can be built only by aggregating the separate coverage indicators for all social security functions. The resulting index loose however some concrete reference to reality. Principles
Diagnostic | Coverage indicators and coverage gaps | main principles (2) Principle 2 | Coverage by social security schemes against specific social risks and contingencies can be understood in two ways “Potential” coverage, measured by the number of persons “potentially” protected if a given contingency occurs (for example, those covered by social insurance schemes, or contributors to such schemes), Reference population: economically active population, employed, working age or total population, specific target groups, etc. Numerator: Active contributors or total persons insured by the scheme or groups entitled to an non-contributory benefit (when needed) Actual coverage, the concept of ‘actual recipients’ or ‘actual beneficiaries’ relates to those receiving benefits at a certain point in time Population de reference: Target population for a given risk or contingency ‒ Ex. Population aged 60 and above (or above statutory retirement age) for old age pensions; Unemployed for unemployment benefits Numerator: Persons who actually receive the benefit These two concepts are complementary to each other and should be assessed separately Principles
Diagnostic | Coverage indicators and coverage gaps | main principles (4) Principle 3 | The « double counting issue» of beneficiaries (relevant in the case of administrative data) Situation | Treatment of double counting for a given risk or function when there is a possibility for one person to be covered and / or beneficiary of several of these benefits (provided by the same or different schemes) Solutions Make the distinction between basic and supplementary benefits and do not count people benefiting in addition to the basic scheme, from supplementary (or top-up) benefits When estimating coverage (& coverage gaps) ‘by function’ : limits the risks of double counting between several functions (ie: pension and health; family allowance and sickness benefits) Household survey data is a way to overcome this risk of double counting associated to administrative data. It allows in addition to analysis of beneficiaries of multiple benefits. Principles
Diagnostic | Coverage indicators and coverage gaps | main principles (4) Principle 3 (cont.)| The « double counting issue» of beneficiaries (relevant in the case of administrative data) Example of Senegal | Old age pension schemes IPRES (General retirement scheme) Deliver some periodic benefits as well as lump sum benefit for those not entitled for periodic pensions. The Old age pension considered as basic benefit for old age coverage An non contributory old age pension Supplementary pension delivered by the «Régime complémentaire des cadres (IPRES) » in addition to the general pension to some sub- groups. These beneficiaries should be counted two time on the numerator FNR (scheme of civil servants) Old age pension The indicator should be clear and consistent about the level of protection provided. Focus on cash periodic benefits Not included | Already counted once Not included | Non periodic cash benefits
Diagnostic | Coverage indicators and coverage gaps | main principles (3) Principle 4 | Choice of the numerator and denominator or the general question of the reference population Ideally, absolute number of persons covered against a specific risk should be set in relation to the group of the population that are addressed by the benefit Examples Number of pensioners can be related to the number of older persons or both numerator and denominator can be restricted to a given limit of age (65+ or legal retirement age) Main questions for discussion Trade–off between reflection of national circumstances (and relevance of the indicator at the national level regarding for example the retirement age) and cross national comparability Question of quality and availability of the required detailed information Questions: Is there a single statutory pensionable age? What about early retirement age and the proportion of beneficiaries below the statutory pensionable age? Principles
Effective social protection: extent and level of effective coverage & expenditure | Indicators focusing on income security of elderly Access to essential health care for all Income security for the elderly Proportion of the labour force contributing to an old age pension scheme (decent work indicator) Proportion of the population above statutory pensionable age receiving a pension ( contributory, non contributory pension, means-tested or not) Average level of old pension (from contributory and non contributory) as a percentage of average or minimum wage Social security expenditure targeting elderly as a proportion of GDP Income security for the elderly Protected persons Actual beneficiaries
Example of social protection benefits for elderly Cash periodic benefits (scope of the graphs presented on coverage) Old age pension from contributory schemes (or survivors pension) Non contributory old age pension (‘social pension’) either universal (Old-age pension from social security fund in Seychelles or State old age pension in Bostwana) or means tested (old age allowance in Bangladesh; old age grant in South Africa) Lump sum cash benefits: Withdrawals from individual savings account; One off payments (ex. Savings bonus; death grant) Benefits in kind Residential care / Home-help services for services (Caring services for elderly in South Korea)
Worldwide 29.9% of the working age, contribute to an old age pension scheme (entitled to pension cash benefits when reaching pensionable age) 54.6% of the elderly above retirement age receive an old age pension (45.1% without China) Extent of effective coverage | Old age pensions Protected persons (or potential beneficiaries) & actual beneficiaries Two dimensions of coverage: Persons protected (grey) (future beneficiaries of an old age pension) and the actual beneficiaries (red)
Income security of persons in old age | Proportion of older persons in the labour force related to the proportion of elderly receiving a pension Old age pensioners as a % of the population above statutory pensionnable age Labour force to population ratio among people aged 65 and over Low coverage High labour force participation (65+) High coverage Low labour force participation (65+)
Coverage by non contributory pensions | what level of benefits? High coverage but relatively low pension level... ... Income security for elderly or the impact of poverty reduction or prevention can be limited In Nepal, in the group of elderly aged 70 years old and over, single women and senior citizen allowances are “responsible” for a decrease in poverty rates of 3 points (3.8 points if considering elderly aged 70 and over in rural areas).
Elderly| Distribution of the population aged 65 and over and distribution of public SP expenditure for elderly In Asia 52% of the world population aged 65 and over & 23 % of the world public social protection expenditure for elderly (in 2010) In Western Europe 14% of the world population aged 65 and over & 35 % of the world public social protection expenditure for older persons (in 2010) In North America, less than 9 % and close to 24% respectively.
Effective social protection: extent and level of effective coverage & expenditure | Global indicators of expenditure Access to essential health care for all Global indicators: mainly expenditure and financing. e.g Public social protection expenditure as a proportion of GDP or as a proportion of total government expenditure (ILO decent work indicator) Distribution of social protection expenditure compared to distribution of the population (worldwide) Proportion of expenditure going to contributory versus non contributory benefits Social contributions in total social protection revenue... Inter-generational indicators
Public social protection expenditure Public social protection expenditure as a percentage of GDP reflects the social spending effort in a given country relative to the size of its economy. Total annual public social security expenditure = sum of expenditures (including benefit expenditure administration costs) of all existing public social security/social protection schemes or programmes in the country. Scope: the scope of C102, which establishes nine classes of benefits: medical care, sickness benefit, unemployment benefit, old-age benefit, employment injury benefit, family benefit, maternity benefit, invalidity benefit and survivors’ benefit, plus other income support and assistance programmes, including CCTs not included under the classes above
Public social protection expenditure (% GDP) Social protection expenditure = health care expenditure + non- health expenditure ( pensions, maternity, sickness, employment injury, unemployment…) The share of health expenditure in total SP expenditure decreases with the increase in the level of income at the national level Low income countries spend on average 3.6% of GDP on social protection ( data) This proportion increases over time (1.7% in 1995) Medium income countries spend on average 6.9% of GDP on social protection ( data) increases over time with the development of social security systems (4.1% in 1995) High income countries: the increase in the last 15 years is not so significant, except in the post 2007 crisis. From 19.2% in 2007 to 22.2% in
Distribution of global social security expenditure by region In Asia Nearly 60% of the world population & 20 % of the world public social protection expenditure (in 2010) Increase in this proportion from 13% to 20% in 10 years In Western Europe 6% of the world population & 34 % of the world public social protection expenditure (in 2010) Decrease in this proportion from 43% to 34% in 10 years In Latin America 8.6% of the world population & 3.8 % of the world public social protection expenditure (in 2010) Significant increase in this proportion from 5.7% to 8.9 in 10 years In Africa 15 % of the world population & 2.1 % of the world public social protection expenditure (in 2010) Increase in this proportion from 1.1% to 2.1 in 10 years
2. Diagnostic: describe the present (& planned) social protection situation Main sources of information Administrative data versus households surveys
National data sources (1) Administrative data and household survey data Administrative data Primary or “traditional” source Data are (or: should be) regularly collected and published by the institutions administering social security schemes/programmes Information on beneficiaries, benefits and persons covered: indispensable for the administration, monitoring and evaluation of the scheme/programme Not explicitly collected for research purposes and largely unexploited by research Principal advantages: ideally... Complete information on persons protected and actual beneficiaries (break-down by sex and age) Little additional cost: data are collected for the regular functioning of the scheme/programme
National sources (2)| Administrative data (2) Disadvantages with regard to coverage measurement Administrative data usually contain ample information on those groups of the population that are covered but not on those who are NOT covered Eligible non-recipients usually are not captured While administrative data can be used to estimate the extent of coverage, they usually do not provide any insights on the causes and effects of non-coverage. Double counting possible in the case of beneficiaries receiving various benefits AND in many developing countries, often poor quality and availability of such data in absence of properly managed records and information system Difficult to capture schemes that are less visible Fragmentation of sources (multiple schemes and programmes) and lack of coordination at the national level Need for Household survey data
Complement data from administrative records: ideally both data sources should be combined Overcome some of the problems of administrative data (double counting, fragmentation, etc.) Appropriate data source for identifying and analysing gaps in coverage Information on potential beneficiaries and their characteristics Analyses of the causes and effects of non-coverage Information on the distribution of specific risks among the population Constitutes key data source for the evaluation on policy options for the extension of coverage Importance of survey data National data sources (3) |Survey data Appropriate tool for the analysis on the effects of social protection on the distribution of household incomes (analysis of the impact on poverty and income inequality reduction)
Limitations Few countries include a sufficient number of questions on social protection in regular survey instruments and no standardized method available to do so... And even fewer countries systematically analyse and use this information if available Costs: household surveys which are representative at the national level are costly Solution: module of questions included in existing surveys Sampling methods and size can limit the value of the results extrapolated from the sample National data sources (3) |Survey data (2)
66 National sources (3) | Household survey data A limited set of questions to be included in regular household surveys Deal with primary data collection via household surveys: Promoting the use of a module of questions to include in regular surveys What for? Better coverage of the social protection dimension in regular surveys: Labour force surveys; surveys on household budget and expenditure and assimilated surveys, FS, HH budget Surveys and possibly demographic and health surveys Fill the information gap in particular on the issue of social protection coverage Overcome or compensate limitations of administrative data What? 4 main questions or types of questions Provide some estimates & facts at the national level … elements which could call for the need for an independent more detailed survey Contribute to awareness raising of the need for and usefulness of social protection statistics
2. Diagnostic: describe the present (& planned) social protection situation Example from Cameroun
Diagnostic (1)| Main objectives A reference frame The assessment matrix Data and statistics : On existing social security schemes and programmes On additional dimensions to characterize each “group” along the life cycle approach (children, older persons, persons in active age) poverty, labour market participation, composition of the household, etc.
Diagnostic| Rapid assessment matrix: structure & objectives Social protection floor: guarantees, national priorities and objectives Describe the present (& planned) social protection situation, taking into account SP strategy objectives. Identify and describe existing schemes and programmes Identify legal/ design and implementation gaps and provides indicators regarding scope and extent of coverage
Diagnostic – Existing provisions Cameroon| Active age population (1) Income security for active age population Statutory provision (some of the design gaps) Most existing social protection provision specifically targets employees. Own account workers & unpaid family members are excluded by law from the coverage by existing social security schemes Lack of unemployment and sickness social security provision Indicators of effective coverage (some of the implementation gaps) CNPS & Civil servant, military schemes cover employees in formal employment Coverage: 5,2% (8,7% men/ <2% women) of the working age and 6,9% of the economically active population (10,2% men/ 3,2% women in 2011). Functions: occupational injuries, maternity, old age*, disability, survivors pensions and family allowances* Expenditures: without pensions, represents less than 0.5% of GDP Other in-kind benefit programmes Work in exchange for food (PAM): 16,590 beneficiaries in 2010 (less than 0.1 % of rural underemployed population) / 0.15% GDP in 2010 Yaoundé sanitation project (PAD-Y) only in Yaoundé (MINEPAT and AfDB) 0.12% of urban underemployed population Lack of specific provision for informal workers Active age
Diagnostic – Existing provisions Cameroon| Active age population (2) Active age persons: some figures Size and composition The represent 50.8% of total population 80,5 of the are active (either employed or unemployed). Situation on the labour market (15-59 years old) 75.7 per cent have a job (poverty rate 38.6%) 4.7% are unemployed (poverty rate 12.6%) 19.5% are inactive (poverty rate 20.9%) Employment status (15-59) Employees: 17.1% Self-employed workers: 55 % Unpaid family contributors: 25.7 % 57.3% work in agricultural sector; 36.5% in private sector (without agriculture) and 5.4% in the public sector Close to 90 % of people in employment are in informal employment More than 60 % of paid employees are not in fully formal employment Active age
Diagnostic – Existing provisions Cameroon| Active age population (3) Active age persons: some figures from household surveys Affiliation to social security Less than 40% of paid employees declared being covered for an old age pension: 72.1% of public sector employees (15- 59) and 22,1% in the private sector (among paid employees) Only 3,1% of all workers in the private sector are affiliated (1,1% women and 5,1 men) One of the obstacles to achieving greater social protection coverage may be the fact that nearly half (47%) say either that they do not have a written contract with their employer or they don’t know In the private sector affiliation is limited to paid employees in the formal sector, especially in large enterprises (> 80% affiliation in enterprises of more than 100 employees). Less than 1 % of people working in agriculture are affiliated to any social security scheme (19% in non agriculture employment) Active age | from household surveys
Diagnostic - Existing provisions Country Cameroon|Children Income security for children: no universal benefit for children CNPS: Child benefits, prenatal benefits, allowances for Newborn Children Covers employees in formal employment in the private sector (financed by employers) and public sector employees (financed by the State) Children up to 14 (conditional to children education from 6) Level of benefit: 1800 CFA Francs per month/ child (< 8% of poverty line; 6,4% of minimum wage) Coverage: 307,102 children in 2009 (3.6 % of targeted group) Expenditures: less than 0.1% of GDP Other in-kind benefit programmes Nutritional/ school feeding programmes managed by MINEDUB; PAM, UNICEF, NGOs: 3.1% of covered children / 0.03% GDP Exemption from tuition fees managed by MINEDUB 1.2% of children / 0.04% of GDP Existing social protection provisions
Diagnostic – Existing provisions Cameroon| Children Children: figures Size & composition Children under the age of 15 represent 43% of population Poverty Poverty rate are higher than the average among children: 47% of children are poor against 40 % for the entire population In rural areas, 64% of children are poor (58% for the total rural population) Group characterization Proportion of children in national population Poverty rate higher among children
Diagnostic – Existing provisions Cameroon |Older people Income security for older people CNPS and Civil servant and military schemes Coverage: 15 per cent of aged 60 and over receive a pension Expenditure: less than 0.5 per cent of GDP Pension beneficiairies Private secteur (CNPS) Civil servants Total Population Ratio12,513,614,9 Lack of non contributory pension scheme capable of covering older people who were not affiliated: 85 per cent of older people aged 60 and over Older persons
Diagnostic – Existing provisions Cameroon |Older people Income security for older people : some information Representation of this group in total population Persons aged 60 and over represent 4.5% of total population (proportion of 3% for 65+) Household composition Almost half of older persons (aged 65 and over) live in households composed of three generations (57%) 16% of older persons live alone or with another person of the same generation This household composition means a significant role of pension redistribution on younger generations Labour force participation 69 % of people aged 60 and over are in employment (74% of men; 64% of women), of which 85% in agriculture Poverty Poverty rate 41.5% Among the minority who receive a pension, the poverty rate is 17% (42% among non pensioners) Older persons
Diagnostic – Existing provisions Cameroon | Health Access to essential health goods and services An unsatisfying situation despite the efforts of Health ministry and private initiatives No health insurance system in Cameroon (formal and informal sector) To fill this lack of insurance, development of mutual health insurance schemes as well as several fragmented programmes across the country 158 mutual health insurance scheme in 2009 (against 9 in 2000) Coverage: 2% of population (2009) Additional initiatives & programmes with limited coverage and covering specific diseases. Expenditure: According to WHO data, public health expenditures per inhabitants were just above 17 dollars in 2010 Health
2. Diagnostic: describe the present (& planned) social protection situation This is time for practice
Based on these guidelines It is now time for you to assess and suggest ways to improve the situation of the population from Nemoland regarding social security coverage. We count on you to provide — through a in-depth assessment of the current social security system & the identification of possible coverage gaps (design & implementation gaps) — the necessary elements which will contribute to the definition of realistic options for the extension of social security to meet the the Social protection Floor guarantees. This is your task Time for practice
A joint task force composed of members of the Ministry of Social Development and the Ministry of Labour is in charge to undertake the gaps analysis, taking into account government priorities The objectives of this task are notably the identification and assessment of existing social protection schemes/ programmes and the identification of gaps the analysis of population needs regarding social security coverage (including highlighting critical situations of some particular groups of the population through the analysis of poverty and labour market data). the comparison of actual social security provision in terms of statutory and effective coverage with the four guarantees of the SPF; and Preliminary suggestions of options for extension We have also with us today a the statistician from the NSO of Nemoland to help us, as far as he can, if there is any critical information missing to carry out the assessment.
Time for practice In hands... A description of the situation in Nemoland (word document) A set of tables prepared by the National statistical office providing information regarding basic socio demographic characteristics of the population, the situation on the labour market a simple poverty profile social security coverage A frame: empty matrix by social protection floor guarantee
2. Diagnostic: describe the present (& planned) social protection situation Additional information related to diagnostic: Socio-demographic & poverty profile
Additional data as part of the diagnosis … as a minimum inputs to discuss options for extension As a minimum… Distribution of the population by age and household structure To assess possible redistribution effects; magnitude of target population by age, etc… Situation on the labour market Proportion of employed population by age groups Proportion of unemployed Magnitude of informal employment (in total employment, among employees) Status in employment: relevant options for employees are not necessarily relevant for own account workers Gaps of coverage by status in employment, etc…. Poverty profile
Nemoland in figures… Distribution of the population & household size Nearly 55% of the population is less than 20 years old 5% is aged 60 or over
Nemoland in figures… Situation on the labour market 80% of the active age (15-59) are in employment; 75,5 % of older persons aged 60 and over and Almost one fourth of the children aged less than 15 (39% when considering 5-19)
Nemoland in figures… Situation on the labour market Child labour is mainly composed of unpaid family workers; the same status prevails among older workers Among the 15-59: the majority are own account workers and more than one third of women are unpaid family workers
Nemoland in figures… Social protection coverage 6,7 per cent of the people aged are contributing to a pension scheme 1,4 % to the public pension scheme 5,3% to the private non statutory pension scheme Percentage of people in employment contributing to a pension scheme …. by area of residence …. by sex 1,3 % in rural areas / 12,2 per cent in urban areas 3,9 % of women and 9,8% of men
Poverty profile as a complement to coverage gap analysis Objectives Bring one much closer to answering the more fundamental questions regarding the determinants of poverty. Describing the characteristics of those in poverty: Analysing the relationship between poverty rates and household’s or individual’s socio-demographic characteristics Demographic characteristics, size and composition of households, area of residence, situation in the labour market and employment status; housing conditions and access to basic services, etc. Developing a picture of who are the most exposed to poverty by comparing poverty rates and poverty gaps between different groups of the population
Poverty profile | Poverty rates by age and area of residence Nemoland U-cruved with higher poverty rate among children and older persons Poverty rate of 51.2 among children aged less than 15 53% among children aged 5-14 years old 38.3% among the years old the highest poverty rates are among elderly: 57.8 % among the % if considering the 55+ Poverty rates Significantly higher in rural areas whatever the age group 17% of the population living in urban areas are poor 66% in rural areas: close to four times more
Poverty profile | Poverty rates by size of households Nemoland Poverty rate increases when size of household increase
Poverty profile | Poverty rates & socio- demographic characteristics in Nemoland Highest poverty rates among children & older persons Highest poverty rates among workers than non working Highest poverty rate among unpaid family workers (65.7%) Unpaid family workers represent 27.3% of total employment among the (and 39.7% of total employment regardless of age groups) Own account workers which represent 52% of total employment among the have in Nemoland a lower poverty rate than average (44.6%) Only 13.1 % of paid employees are poor and less than 1% of employees in the public sector
Poverty profile | Kinds of household older persons live in and poverty rate Nemoland The majority of older persons lives among multi-generational families, either with a active age generation (20.5%) or within a three generation households (56.5%). One older person out of 6 lives alone or with a person from the same generation Poverty rates are significantly higher among 3 generation households and most vulnerable household
Poverty profile | Poverty rates among children by area of residence Nemoland Poverty rates among children are higher than average The worst situation is among the 5-14 years old with a poverty rate of 53.3% 73.9% in rural areas 20.8% in urban areas
Poverty profile | Poverty rates among children by household size Nemoland
3. Analytical tools to support the extension of social security
PRELIMINARY ASSESSMENT OF COSTS AND IMPACTS ESTIMATE THE COSTS and the IMPACTS How much would it cost to introduce benefits that ensure to fill gaps ? What are the impacts on poverty? DIAGNOSTIC/ MAPPING -Mapping of Existing & planned schemes/programmes: qualitative & quantitative indicators -Gaps and needs assessments (coverage gap analysis, labour market analysis, poverty profile, etc,) COMPARE WITH THE FLOOR GUARANTEES & higher levels Do existing provisions meet the Recommendation principles? Meet income security guarantees? Meet higher levels? Basic income security for children, persons in active age, older persons & Access to essential health care for all IDENTIFY AND ASSESS GAPS and DEFINE OPTIONS for the extension: Design / policy and implementation gaps and options for extension ASSESSMENT MATRIX Diagnostic, assessment of costs & analysis of impacts NATIONAL DIALOGUE ACTUARIAL studies, etc. DESIGN OF NEW SCHEMES/PROGRAMMES Define priorities Fiscal space debate
3. Analytical tools to support the extension of social security Assessment of the potential impact on poverty reduction of various benefits: Using micro-data for an assessment of the impact on poverty
Evaluating the effects of social transfers on living standards Ex-post assessment: Analysis of existing programmes /schemes or social security systems Effectiveness: Does the programme reach its objectives? Efficiency: Does the programme achieve its objectives in an optimal way? Ex-ante assessment: Preparing future reforms Adjusting parameters of existing programmes Designing new social security benefits/schemes Enhancing the coordination between different schemes within the social security system DIAGNOSIS
Evaluating the effects of social transfers on living standards: Ex-post assessment (1) Ex-post assessment (part of diagnosis): Analysis of existing schemes/programmes Effectiveness: Does the programme reach its objectives? Coverage: e.g. which proportion of elderly receives an old age pension? Benefit levels and effects: e.g. effects of old age pensions on living standards of pensioners; effects of social assistance on overall poverty reduction DIAGNOSIS
Evaluating the effects of social transfers on living standards: Ex-post assessment (2) Efficiency: Does the programme achieve its objectives in an optimal way? Coverage: Are benefits going to those who should receive them? (inclusion/exclusion errors) Benefit levels and effects: Are benefit levels adequate to reach objectives? Costs: Does the programme reach its objectives in a cost-effective way? DIAGNOSIS
Evaluating the effects of social transfers on living standards Ex-post assessment: Analysis of existing programmes /schemes or social security systems Effectiveness: Does the programme reach its objectives? Efficiency: Does the programme achieve its objectives in an optimal way? Ex-ante assessment: Preparing future reforms Adjusting parameters of existing programmes Designing new social security benefits/schemes Enhancing the coordination between different schemes within a social security system SIMULATION
Evaluating the effects of social transfers on living standards: Ex-ante assessment (1) Ex-ante assessment: Preparing for future reforms Adjusting parameters of existing programmes How can the coverage of a social insurance scheme can be extended to still uncovered groups of workers? Which benefit levels should be provided in order to effectively prevent poverty? Designing new social security benefits/schemes How could a new social pension programme be designed? What is the most effective way to introduce child benefits? Enhancing the coordination between different benefits within a social security system Identifying undue cumulation of benefits and coverage gaps SIMULATION
Evaluating the effects of social transfers on living standards: Ex-ante assessment (2) Relevance of ex-ante assessments for extending social security and building national social protection floors Identifying effective ways to close coverage gaps Assessing the possible impacts of a set of basic social security guarantees on poverty reduction Assessing the interplay between different programmes and identifying gaps Assessing several options for benefit levels and their impact on poverty reduction Assessing the cost of a proposed reform and its possible impact SIMULATION
Ex-ante assessment of impacts (1) Ex-ante assessment of the impact of cash transfers on the reduction of poverty is.... a static micro-simulation of the (direct) impact of transfers on individual/household income/expenditure and poverty status And is not.... a full assessment of the impact which would take into account changes in people’s behaviour and further socio-economic effects a Poverty Impact Assessment / Poverty and Social Impact Assessment (WB, OECD and others) SIMULATION
Ex-ante assessment of impacts (2) Why conducting ex-ante assessments of impacts? Estimating the likely effects of future reforms on people’s incomes Testing different reform options... without hurting anyone... Reflecting the interplay of different programmes with household structures, employment, wages etc. Enhancing the effectiveness of future reforms Preconditions Know how.... Availability of good survey data (see next section...) Being aware of limitations SIMULATION
Total population (51) Poor (21) Non poor (30) Poverty rate 21/(30+21) = 41,2% Gap Indi = 270 Total aggregate poverty gap c<z Gap Indi = 3970 Poverty impact | Overview Initial situation Poverty gap ratio (3970/500)/51 =15.6% Poverty line = Consumption expenditure per capita
Non poor (30) Poverty impact | Overview Post benefit Total population (51) Poor (21) Social cash benefit provided: means-tested cash benefit of 300 to all poor children aged Poverty rate 17/(30+21) = 40,5% Poor (17) New Non poor (34) Poverty line = 500 New Total aggregate poverty gap c<z Gap Indi = 3740 Poverty gap ratio (3970/500)/51 =7.5%
Poverty impact analysis | Steps 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit(s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) “ex-post” income or consumption expenditure at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit Pre- requisite Initial situation & diagnosis Test pre-defined options & possibly new ones Situation post benefit Go through main hypothesis (and their limitations) Providing the fact that we will test mainly ‘individual’ benefits | quick insight in the aggregation of individual benefits to household level & the redistribution
Poverty impact analysis | Steps Pre- requisite 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit (s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) income or consumption expenditure « ex-post » at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit
Pre-requisite | Ready to use data file (1) Micro data file Microdata is data on the characteristics of units of a population, such as individuals, households, or establishments collected by a census, a survey, or experiment Dealing with microdata requires good knowledge of statistics and statistical software packages (SPSS, SAS, Stata, etc.).
Pre-requisite | Ready to use data file (1) A micro data file is A « flat file», a table with: in columns: variables in lines: individuals, households or any other unit of observation In case of survey data, this is associated with a questionnaire basis for variable creation and essential for the understanding of possible filters and reference population for each question a variable dictionary (code book) Ideally a survey methodology (description of sampling, the weighting variable, etc..) and measurement of poverty if familiar... a source of freedom … it is not …...a black box...necessarily transparent not always one single file but a collection of files.... to merge not always in line with official results presented in national reports.. So it is sometimes … a puzzle in Vietnamese or Nepalese to “reconstruct” on the basis of associated documents … to come to this stage of « ready for use » can sometimes be the hardest and longest part of the process
Step 1 | Which data? which surveys? Officially used data: Do not look for one data file and any variable on expenditure or income, but look for the data file used at the national level to carry out the “official” poverty analysis and the variable(s) used to determine the poverty rate at the national level Application-oriented and not (purely academic) research Should work with national data and an initial situation regarding poverty indicators and other basic indicators and data commonly used and shared in the country Types of surveys National income and expenditure survey or equivalent (contact: national statistical office) LSMS or similar surveys (but often limited)
Step 1 | Minimum set of data and information required General information to get from the National Statistical Office (NSO) …. of Nemoland Poverty line or poverty line(s) for the year for which data are available Nemoland Poverty line per capita and per month 85,553 nemos per month and per capita A minimum of methodological information such as the basis for redistribution from household to individual Does the country use a per capita basis (total expenditure or income divided by the size of the household) or a per equivalent adult basis If per capita: who is included as part of ‘household members’ If per equivalent adult: which is the equivalent scale used? The National Statistical Office of Nemoland uses a per capita adult
Step 1 | Minimum set of data and information required From the data file Individual characteristics used to define eligibility to benefits Age Sex Number of household members (equivalent adult when relevant) Area of residence Any other variables needed to identify beneficiary and to define benefits (criteria used for proxy means tested, labour market indicators, etc) Main variable(s) used to calculate poverty related indicators Depending on the variable used for poverty calculation Household consumption per capita (or per equivalent adult) Household income per capita (or per equivalent adult) Actual beneficiaries of various relevant benefits Sample weights (individual/household level)
One line = one individual Random sample from the population of Nemoland (representative of the total population) Total population = 360,103 Representative sample of 5,233 Step 1 | Excel file — at the individual level Variables to identify individuals and households # Ind_ID Unique ID for each individual # HH_IDUnique Household ID Link with the HH datasheet for the aggregation of each individual income/expenditure at the household level From the sample to the total population WEIGHT Adjusting for non-response Adjusting for external information notably some proportions known in the total population for key variables selected when defining the sampling (e.g distribution of the population by region, district obtained from a recent census or an extrapolation of the most recent one #J – R: basic socio-demo characteristics and limited information on actual beneficiaries from existing schemes or programmes Total annual expenditure per capita in national currency (Nemoland: nemos) Initial Poverty rates and gaps, quintiles are calculated on an individual basis based on this variable The individual data file
Step 1 | Excel file — at the household level The household data file One line = 1 household # HH_ID Unique Household ID Link with the Individual datasheet for the aggregation of each individual income/expenditure at the household level. Example: household n° 1 is composed of 8 household membres (8 lines in the individual data file but one single line when considering household as a unit of reference). #E: Household size: total number of household members Per capita basis calculation #F: Annual total consumption expenditure at the household level = SUM of individual per capita expenditure pre-benefit for all household members Starting #H : Automatic calculation of the sum of individual consumption expenditure post benefit to obtain the total aggregate household consumption expenditure and allow the redistribution among all HH members
Step 2 | Analysis of the initial situation 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit (s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) income or consumption expenditure « ex-post » at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit … taking stock of the work done during the diagnostic phase …. identification of existing social security schemes and benefits provided The comparison with the social protection floor guarantees The assessment of the coverage gap and possible deficiencies of the existing social protection system
Poverty impact analysis | Step 3 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit (s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) “ex post” income or consumption expenditure at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit
Step 3 | Test previously defined options & new ones taking into account the results of step 2) (1) Questions : Which benefit(s) for whom? First | For whom? Define and identify target groups Individuals or households? By age group, situation on the labour market or status in employment, ability to work or not, etc This should be based on the analysis of the initial situation Identification of needs and gaps The poverty profile analysis Priorities in the country and ideally as part of gradual extension process Any other?
Step 3 | Define and identify target groups for the new benefit(s) (2) How to fix eligible age? General Based on detailed analysis of poverty profile of potential target group (children, elderly) Cost versus impact on poverty reduction Specifically for pensions Legal retirement age for existing contributory pension or higher age as a starting point with a planned gradual decrease over a given period (consistency of the pension system) Existing pension used as a basis for extension Specifically for child benefits Associated policy objectives, e.g. promoting access to education (> school-age children), nutrition (>starting at early age) Examples?
Step 3 | Define and identify target groups for the new benefit (s) (3) Some particular cases … Change in an existing benefit, notably Change in the level of an existing benefit Example: increase in the level of the non contributory pension Change in the eligibility criteria Extension to groups of the population at present uncovered Identification of actual beneficiaries for this benefit: Important in countries where the proportion of the population covered by existing social security provision represents a significant proportion (more than 10 per cent). How to? (with survey data) Ideal case: the identification variable is available in the survey (rare) Alternative: find some proxy variables to identify this group of the population
Step 4 | Which benefit? (1) Question : Which benefit(s) for whom? Second| For whom? Define benefit or benefits Type of benefit and amount Cash or in in-kind —If benefit in-kind, estimate of corresponding cash value of a benefit in-kind Relative or absolute value —If relative value: expressed as a proportion of the poverty line; of minimum wage or any other value of reference in the exercise: 3 options provided relative to poverty line or minimum wage or absolute value — Value depending on household structure?
Step 4 | Which benefit? (2) Question : Which benefit(s) for whom? Type of benefit and amount (Cont.) Means-tested benefit or universal —If means-tested: perfect targeting of poor population or simulation of an imperfect targeting more close to reality such as proxy means-tests Suppose availability of appropriate indicators: type of housing, equipment, etc.) as part of the survey Do we consider the consumption expenditure, the income (with associated difficulties when implementing)? Do we consider assets? Conditional ot not? Direct and indirect costs of conditionalities
Poverty impact analysis | Steps 4, 5 & 6 First main assumption 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit (s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) income or consumption expenditure « ex-post » at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit
Step 4 | Adding the new benefit amount to the consumption expenditure per capita ex-ante Providing the benefit to all eligible individuals Hypothesis 1 « unrealistic »: assume perfect targeting of eligible individuals Applies for universal benefit and even more in case of means tested benefits (which requires the identification of the poor) What can be done? Simulate an imperfect targeting by including For means-tested benefits, inclusion and exclusion errors in the selection of beneficiaries (random selection of non targeted people in a certain range of expenditure/ income outside the scope) Exclusion errors in the case of universal benefits Not taken into account in our exercise to come, so be aware that direct impacts are probably over estimated and should be considered as the maximum possible impacts/ effects However, there may be indirect effects (e.g. multiplier effects at the community level) (also not taken into account) which may offset some of these limitations
Step 4 | Adding the new benefit amount to the consumption expenditure per capita ex-ante Adding the amount of the benefit to the original value of consumption expenditure per capita (for eligible individuals) Hypothesis 2: Total consumption of the benefit received
Poverty impact analysis | Steps 5&6 1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Define and identify target groups for the new benefit (s) and define benefit (type and amount) 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) income or consumption expenditure « ex-post » at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit Second main assumption
Step 5 | Aggregation of individual (or equivalent) consumption expenditure « ex-post » - household level 6) Aggregation of ex-post consumption expenditure at the household level Household level From individual expenditure to household aggregate expenditure Hypothesis 2: distribution of the individual benefit for a shared and equal consumption among all household members New individual expenditure (at the individual level) (INDexp_post1 = INDexp_0 + Benef1) At the household level: Sum the new individual expenditure = HHexp = ∑ INDexp_post1 (of each household member) Calculate consumption expenditure post benefit per capita New consumption expenditure per capita = ∑ INDexp_post1 / household size (or sum equivalent adult values) Consequences A benefit provided to all elderly will benefit — in theory and possibly in practice but not necessarily — to other household members, including children and working age population (shared equally among them) Any ‘primary’ beneficiary (e.g. elderly receiving a pension) will not get the full amount of the benefit but an amount divided by the number of household members
Step 5 | Aggregation of individual expenditure « ex- post » - household level 6) Aggregation of consumption expenditure at the household level. How to? Household level < Raw data Individuals < Household level (“HHBasis”) = Sum of individual expenditure in the household HHId « 9 » 10,283,874 = 3407, , ,958 With one child less than 15 receiving the benefit (before redistribution)
Step 6 | Redistribution of this total amount among household members < Household level (“HHBasis”) Raw data Individuals Redistribution of expenditure post benefit per capita (calculated for each household) to each household member (same value for each in the individual sheet). New basis for the calculation of poverty indicators post benefit
1) Data file ready for use and key variables identified 2) Analysis of the initial situation (focus on initial poverty) 3) Test previously defined options (and possibly new ones taking into account the results of step 2). 4) Add, for target groups, the new benefit amount to the income or consumption expenditure ‘before benefit’ 5) Aggregate individual (or equivalent adult) income or consumption expenditure « ex-post » at the household level 6) Redistribution of this amount ( per capita or per equivalent adult ) among household members 7) Calculation of distribution and poverty indicators after provision of the new benefit Poverty impact analysis | Step 7 Starting from the NEW value of consumption expenditure per capita « « « post benefit »
Ex-ante impact assessment through microsimulations Microsimulations are a useful tool to feed in the national dialogue on the extension of social security and on building national social protection floors Reflect the combined effects of a set of benefits on the realities of households Allow for interaction with household structures and employment patterns But: need to be aware of its limitations Possibility of transforming assessment methodology into policy monitoring tools to support implementation Extremely useful if done seriously
Examples of typical poverty indicators
Simulation | Examples of typical poverty indicators Poverty related indicators Poverty rates, poverty gap and squared poverty gap post benefits Reduction in the number of poor: absolute figure and relative (as a percentage reduction) Total aggregate gap Reminder: The aggregate poverty gap shows the cost of eliminating poverty by making perfectly targeted transfers to the poor. This total cost can be related to GDP. Cost related indicators Total cost (including administrative costs) in absolute numbers Total cost (including administrative costs) as a percentage of GDP or government expenditure These are broad estimates not replacing a proper costing exercise Cost over gap reduction (number of units cost per 1 unit reduction of aggregate poverty gap)
Simulation | Main indicators to consider.... Cost over gap reduction Comparison between expenditure to provide the benefit and the reduction of poverty gap Total COST (benefit + admin costs) Ratio = Poverty gap reduction = (initial aggregate poverty gap – new aggregate poverty gap) A ratio of 1 is not possible, as it would require Perfect targeting Level of benefit equal to the difference between each individual’s income (or expenditure per capita) and the poverty line No other cost than benefit (and notably no administration costs) Even in the case of a means-tested benefit of a given value but perfectly targeted to the poor, the ratio is necessary superior to 1 Limitations
Impacts of universal old age pension on poverty reduction | Cameroon Base: SPSS database| Calculation of the impact based on the entire database Poverty rate reduction Initial poverty rate
Impacts of benefit package on poverty rate reduction | Cameroun Base: SPSS database| Calculation of the impact based on the entire database Poverty rate reduction Initial poverty rate Post package universal Post package targeted on poverty
Policy implications for building national social protection floors Assessments are a useful tool to feed in the national dialogue on the extension of social security and on building national social protection floors Reflect the combined effects of a set of benefits on the realities of households Allow for interaction with household structures and employment patterns But: need to be aware of its limitations Possibility of transforming assessment methodology into policy monitoring tools to support implementation Extremely useful if done seriously
Moving forward... Some policy implications and challenges 139 Financing of social security, particularly with respect to national Social Protection Floors ● Identifying and extending fiscal space within current budgets (ILO-IMF collaboration) ● Challenge: Identifying possible options for more effective collection of taxes and other sources of revenue Effective policy dialogues at national level ● Identification of priority benefits, priority target groups and sequencing in gradual implementation processes ● Dynamics of national policy dialogues ● Including estimates on direct impact on poverty Strengthening national institutional capacities ● Starting from existing institutional frameworks where possible ● Building up sustainable institutional capacities and ensuring their effectiveness ● Importance of monitoring and policy planning
3. Analytical tools to support the extension of social security Estimating the cost of various policy options in order to close coverage gaps
Cost estimate of policy options to close coverage gaps ILO has developed modelling techniques to estimate the cost of policy options aimed at building national social protection floors Types of tools/protocols Generic tools for rough estimates of the cost of a package of basic social protection benefits (“early generation”) Country-specific modelling protocol, tailored to country context for the estimation of the cost of policy options These tools are designed to support national policy dialogues, but they may need to be complemented by more detailed financial feasibility/actuarial studies
ILO tools to estimate the cost of the extension of coverage The Rapid Assessment Protocol (RAP) is a simple projection model that aims at estimating the future costs of introducing one or more cash transfer elements of the Social Protection Floor. At the same time, the RAP provides a general picture of the country's main indicators: demographics, labour market, macroeconomic and the development of the general government's balance sheet.
GGO (BS) GENERAL GOVMNT. OPS. BS EAP ECON. ACTIVE POPULATION ECO MACRO- ECONOMIC FRAMEWORK GGO (SQ) GENERAL GOVMNT. OPS. SQ POP POPULATION BY AGE AND SEX AR ACTIVITY RATES BY AGE AND SEX README INFORMATION INSTRUCTION S BEN (n) BENEFIT – MODELING EXERCISE BEN (...) BENEFIT – MODELING EXERCISE BEN (2) BENEFIT – MODELING EXERCISE BEN (1) BENEFIT – MODELING EXERCISE Cost estimation: Core elements of modelling tool (1)
Cost estimation: Core elements of modelling tool (2)
3. Analytical tools to support the extension of social security Estimating the cost of policy options to close coverage gaps: Examples from Cameroon & Nemoland
Scenarios Cameroon| Children Recommendation and scenario Low coverage and low benefit level CoverageAgeMeans testBenefit level Scenario 1Universal14No 1,800FCFA/month Scenario 2Universal10No Scenario 3Poor children14 Yes (poverty threshold) Scenario 4 New beneficiaries: poor children Current beneficiaries: Increased benefit level (poor or non poor) 14 Yes (poverty threshold) Only for new beneficiaries 25 per cent of poverty threshold in 2012 = 6500 FCFA per month and per child) Scenario 510 First 3 scenarios: amount equal to benefits delivered by CNPS Universal Benefit level 3.6 times higher
Cameroon | Additional costs generated to fill protection gaps: children Additional costs as a share of GDP for the 5 extension scenarios 2.3% 0.6% Universal
Scenarios Cameroon| Active age population Recommendation and scenario Propose a basic income security for poor workers Difference between Persons able to work Persons highly handicapped unable to work Persons able to work Public works programmes : 100 guaranteed employed days per household and per year Salary based on poverty threshold Alternative: 50 guaranteed employed days Persons highly handicapped unable to work Periodic in-cash benefit: 50 per cent of poverty threshold Concerns 5.2 per cent inactive person of working age and 1 per cent of employed persons
Cameroon | Additional costs generated to fill protection gaps: active age population Additional costs as a share of GDP for the 5 extension scenarios <0.7% 50 days <1.1% 100 days
Scenario Cameroon| Older persons Recommendation and scenario Extend coverage to the 85 per cent of non covered older persons CouvertureAgeMeans testBenefit level Scenario 1Universal 60No 40 per cent of poverty threshold Scenario 2Universal 65No Scenario 3 Universal70 Yes (poverty threshold) Scenario 4 Means-tested65 Yes (poverty threshold) 40 per cent of poverty threshold Scenario 5Means-tested65 Yes (poverty threshold) 75 per cent of poverty threshold 40% of poverty threshold Higher amount: pertinence to be discussed Universal Eligibility age
Cameroon | Additional costs generated to fill protection gaps: older persons Additional costs as a share of GDP for the 5 extension scenarios < 0.7% < 0.15% Universal
Scenarios Cameroon| Health Recommendation Propose universal access to health WHO recommendation in terms of public health expenditures In order to provide basic health services in low income countries, public health expenditure is estimated to 51$ per person and per year in 2012 60$ per person and per year in This estimation includes the cost to extend health services The scenario refers to this estimation To be considered as a basis (Cameroon moderate income country) Only the gap between those values and current health expenditure in Cameroon are taken into account for the calculation of additional costs associated to the floor health composant.
Cameroon | Additional costs generated to fill protection gaps: « packages » Universal Package Children: universal benefit for children aged < 15 : 1800 FCFA per month (scenario 1); Older persons: universal old-age pension for persons aged 60 or more which amount equivalent to 40 per cent or poverty threshold (scenario 2) For active age population A public works programme : 100 guaranteed employed days per household and per year at poverty threshold level In-cash benefit whose amount is half the poverty threshold for active age highly handicapped person Health: basic health benefit on a universal basis: total amount of public health expenditures per year and per inhabitants : 51 dollars in 2012 and 60 dollars in 2015
Cameroon | Filling the gaps: how much would it cost to combine scenarios: « packages » Package focused on poverty Children: means-tested benefits for children under 15 living under the poverty threshold amounting to 1800 FCFA per month (scenario (3)) Older persons: universal old-age pension for older persons aged 65 or more living under the poverty threshold amounting to 40 per cent of poverty threshold (scenario (4)) For active age population (aged 15-64) A public works programme: 50 guaranteed employed days per household and per year at poverty threshold level Same as “universal package” Health: Same as “universal package”
Cameroon | Filling the gaps: how much would it cost to combine scenarios: universal package Inferior to 4.5% of GDP
Cameroon | Filling the gaps: how much would it cost to combine scenarios: Package focused on poverty Inferior to 3 per cent of GDP
Cameroon | Filling the gaps: how much would it cost to combine scenarios: Package focused on poverty Inferior to 16 % of consolidated government expenditures
Impacts of various options on poverty reduction | Nemoland Poverty rate reduction Initial poverty rate
NemoLand | Additional estimated costs generated to fill protection gaps: children Additional costs as a share of GDP for several options and impacts on poverty reduction
NemoLand | Additional estimated costs generated to fill protection gaps: older persons Additional costs as a share of GDP for several options and impacts on poverty reduction
Monitoring: Importance of regular monitoring of schemes/programmes and the social security system as a whole Nationally-defined monitoring mechanisms and national consultations Collection, compilation and publication of social security data, statistics and indicators Legal framework to protect private individual information Taking into consideration relevant guidance provided by ILO and other international organisations (in particular resolution of 9 th ICLS) Exchange of information, experience and expertise among members and with ILO Technical assistance from ILO and others 161 involving social partners and, as appropriate, other stakeholders See ILO Social Protection Floors Recommendation, paras Regular and ongoing diagnosis and analysis of social security schemes/prog rammes and the system as a whole
4. Decision-making with regard to the extension of social security Prioritization, time-frames, sequencing, coordination
PRELIMINARY ASSESSMENT OF COSTS AND IMPACTS ESTIMATE THE COSTS and the IMPACTS How much would it cost to introduce benefits that ensure to fill gaps ? What are the impacts on poverty? DIAGNOSTIC/ MAPPING -Mapping of Existing & planned schemes/programmes: qualitative & quantitative indicators -Gaps and needs assessments (coverage gap analysis, labour market analysis, poverty profile, etc,) COMPARE WITH THE FLOOR GUARANTEES & higher levels Do existing provisions meet the Recommendation principles? Meet income security guarantees? Meet higher levels? Basic income security for children, persons in active age, older persons & Access to essential health care for all IDENTIFY AND ASSESS GAPS and DEFINE OPTIONS for the extension: Design / policy and implementation gaps and options for extension ASSESSMENT MATRIX Diagnostic, assessment of costs & analysis of impacts NATIONAL DIALOGUE ACTUARIAL studies, etc. DESIGN OF NEW SCHEMES/PROGRAMMES Define priorities Fiscal space debate
Decision-making with regard to the extension of social security Prioritization of various policy options Most urgent needs? Resources (available fiscal space and possibilities of extending fiscal space) Sequencing Sequence of implementation of policy options: Enhancing the existing provision, and adding new elements Coordination between different elements of the social security system Time frames Progressive implementation with defined timeframes and targets
Role play National dialogue to develop a strategy for the extension of social security for Nemoland Objective: Decide on strategy and implementation plan for the period Actors: Government: various ministries, statistical office, planning commission Social partners: Employers and trade unions Other national stakeholders: several NGOs representing various groups of the population Development partners: UN, IMF, World Bank, donors It’s your turn!
Government Ministry of Labour and Social Insurance (team) Ministry of Social Development (team) Ministry of Finance (team) Ministries responsible for education, health, gender etc. Statistical Office Planning Commission Social partners: Employers and trade unions (teams) Other national stakeholders: several NGOs representing various groups of the population Development partners: e.g. UN, IMF, World Bank, donors Actors represented in national dialogue on social protection in Nemoland
Lessons from role play? Prioritization of various policy options Sequencing Time frames Other issues
Annex: useful references Survey data, other useful data sources & references Examples from Viet Nam: assessment and estimates of costs and simulation of impacts on poverty reduction How to measure poverty rates and poverty gap?
Microdata | Sources: which surveys? Do not look for ONE data file and ANY variable on expenditure or income, but for THE data file used at the national level to carry out the poverty analysis and the variable(s) used to determine the poverty rate at the national level Application-oriented and not (purely academic) research Work with national data and initial statistics that are used at the national level Population: level and distribution by age and geographic areas Initial poverty rate and other related indicators used and shared in the country Types of surveys Income and expenditure surveys or equivalent
Microdata | Some examples of typical surveys to be used for impact analysis
Microdata | Where to find them? (1) National institute of statistics More and more countries developed a data catalog with sometimes a direct access to micro data in SPSS or STATA format Links to national statistical offices: International sources (1) ISHN | International Household Survey Network Survey catalog ( Survey questionnnaires ( Recently: Household survey data catalog ( )
Microdata | Where to find them? (2) International sources (2) World Bank Central Microdata Catalog The Central Microdata Catalog is a portal for all datasets held in catalogs maintained by the World Bank and a number of contributing external repositories Europe: Luxembourg Income Study LIS is home to the Luxembourg Income Study Database and the Luxembourg Wealth Study Database: harmonised microdata from high- and middle-income countries around the world. ( ) From GESS | Statistics page Social protection in regular national household surveys: questionnaires, types of questions and links to resources (including microdata when possible) ttp:// ttp://
GESS | Social protection in regular national household surveys: questionnaires, types of questions and links to resources Information organized by region country Access to main resources questionnaires survey reports microdata when available webspace dedicated to the survey updated on a regular basis
Guided visit| Information and data sources 1. Texte générique… à reformuler par pays 2. Disponibilité des données dans SSI pour le pays considéré et principales limites 3. Sites des principales institutions de sécurité sociale, les bureaux de statistiques, profils pays dans Gess, etc. What? Economic and financial information Population & Employment Denominators of the indicators Expenditure and revenues (IMF and ESSPROS) Sources Internatoinal data sourcesutomatique) Sources nationales National level Données principales dans SSI Institutions de sécurité sociale | données des régimes Institutions de sécurité sociale | données des régimes Saisie possible en ligne Niveau régime Calcul automatique d’indicateurs agrégés au niveau national
ILO Social Security Inquiry database
Assessing impact on poverty reduction | some references (1) ILO studies Cichon, M.; Bonnet, F.; Schmitt, V.; Galian, C.; Mazeikaite, G. : Analysis of the Viet Nam National Social Protection Strategy ( ) in the context of social protection floors objectives. A rapid assessment (Geneva: ILO) Bonnet, F.; Hagemejer, K.; Wiechers, T.; Yamabana, H., 2012: Zambia. Introduction of a Universal Pension Scheme. Section 5. Impact of the universal old age pension on poverty reduction Schmitt, V.; Satriana, S., 2012: Social protection assessment based national dialogue. Towards a nationally defined social protection floor in Indonesia (Bangkok: ILO) Bonnet, F.; Brand-Weiner, I.; Butare, T.; Nkanagu, T., 2012: Réalisation des garanties du socle de protection sociale au Cameroun. Evaluation rapide des coûts sur la période et estimation de l'impact sur la réduction de la pauvreté (ILO, Draft for discussion) Gassmann, F. and Behrendt, C., 2006: Cash benefits in low-income countries: Simulating the effects on poverty reduction for Senegal and Tanzania, Issues in Social Protection Discussion Paper (Geneva: ILO), Broader Poverty impact assessment (PIA) Promoting Pro-Poor growth: A Practical Guide to ex-ante Poverty Impact Assessment | Access to resources / download guideAccessdownload guide
Assessing impact on poverty reduction | some references (2) The Poverty impact “webspace” (Gess) Access to main resources associated to this course: Excel files, documents, guides and examples Page.do?pid= Page.do?pid=1235