ARMENIA: POVERTY REDUCTION AND SHARED PROSPERITY

d Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized ARMENIA: POVERTY REDUCTION AND SHARED PROSPERITY South Caucasus Poverty team Poverty and Equity Global Practice October 4, 216

Acknowledgements This Poverty Assessment was prepared as part of the FY 16 South Caucasus Programmatic Poverty Assessment Technical Assistance (P156449). The Armenia Poverty Assessment team is comprised of Nistha Sinha (Task Team Leader), Moritz Meyer, Paul Andres Corral Rodas, Kadeem Ervyn Khan and Sasun Tsirunyan. The analytical work benefitted from discussions with Gohar Gyulumyan, Laura Bailey, Aleksan Hovhannisyan, detailed feedback from the Armenia Country Team. Arpine Azaryan and Armanda Carcani provided logistics and processing support. The report was written under the guidance of Mercy Miyang Tembon, Laura Bailey, Carolina Sanchez- Paramo, and Rashmi Shankar. The peer reviewers were Carolina Diaz-Bonilla and Congyan Tan. The team gratefully acknowledges all the comments and feedback received from the National Statistical Service of the Republic of Armenia (NSSRA), and participants of the consultation workshop held in Yerevan in May 216. 1 Armenia: Poverty Reduction and Shared Prosperity, South Caucasus Poverty team

Executive summary: patterns of poverty reduction and shared prosperity Economic growth translated into higher levels of consumption for all parts of the welfare distribution and lower poverty. Consumption growth for the bottom 4 was lower than for the total of the population. Between 21 and 214 poverty in Armenia declined. An international comparison shows that Armenia still has one of the highest poverty rates in Europe and Central Asia and only made limited progress after the global economic crisis hit the country in 29. Despite positive consumption growth for all households, (i) regional disparities between Yerevan, other urban areas and rural areas remain high, (ii) vulnerability to poverty persists, and (iii) non-monetary measures of welfare highlight development gaps. 2 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

Executive summary: drivers of income growth and poverty reduction Income growth and poverty reduction: Decline in poverty benefited from (i) higher employment rates and wages, as well as growth in agricultural sales, (ii) remittances and (iii) social expenditure such as pensions. Labor markets: growth of services and manufacturing translated into additional employment opportunities and higher wages. Growth of sales in the agricultural sector supported rural areas. Individuals in poor households (i) are more often unemployed, (ii) work fewer hours, (iii) are more often temporary or seasonal workers, and (iv) experience less protection due to different type of contracts. Other urban areas do not create sufficient employment opportunities in the non- agricultural private sector to accommodate decreasing employment in agricultural sector. Migration: international migration creates employment opportunities for a landlocked country. At the same time, remittances are subject to substantial fluctuations triggered by shifts in the economic environment. Fiscal activity: overall, taxes and expenditure reduce poverty and inequality; pensions are an important source of income and transfers through the family benefit program (FBP) reduce the intensity of poverty. Yet, poverty impact of FBP is small as budget is limited. 3 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

Executive summary: policy agenda Policy agenda: Patterns of poverty and shared prosperity suggest that policy agenda should focus on inclusiveness of economic growth: (i) Supporting growth, (ii) Investing in endowment, and (iii) Protecting the poor and vulnerable. 4 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

Agenda Armenia: Poverty reduction and shared prosperity World Bank Twin Goals in Armenia: an international perspective Recent trends: economic growth, poverty reduction and shared prosperity Picture of poverty: who are the poor, how do they experience poverty and where do they live Sources of income growth and poverty reduction: composition of income and changes over time Behind aggregate figures: three facts on drivers of income growth and challenges Policy agenda: support, invest and protect 5 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

World Bank Twin Goals in Armenia An international perspective Economic growth contributed to consumption growth for all households in Armenia and supported poverty reduction. An international comparison shows Armenia s poor performance on poverty and shared prosperity.

24 25 26 27 28 29 21 211 212 213 214 Poverty reduction: In 214, 26.3 percent of the population lived below the international poverty line of 2.5 USD PPP. Compared to 21, poverty has declined, but the share of poor households is still higher than before the global economic crisis which hit the country in 29. Economic growth and poverty in Armenia 6 6 5 5 GDP growth 4 4 3 2 3 2 national poverty 1 1 2.5 USD PPP 25-1 -1-2 -2 Poverty trend using the international poverty lines of 2.5 USD PPP, and the national poverty rate (yellow line), GDP growth in Armenia (bar graph). Note 1: World Bank staff calculation using a consumption aggregate drawn from ILCS data and harmonized for international comparison. Note 2: National consumption aggregate (using methodology adopted in 29). Source: World Bank s data base of harmonized consumption data (ECAPOV accessed 3/15/216). 7

An international comparison illustrates that Armenia has one of the highest poverty rates in Europe and Central Asia. Other countries with same level of GDP per capita report lower poverty: nature of economic growth and distribution of wealth differ from other countries in the region. 35 3 Poverty rate the international poverty lines of 2.5 USD PPP 26.3 35 3 GDP per capita and international poverty rate at 2.5 USD PPP Kyrgyz Republic (214) Georgia (214) 25 2 25 Armenia (214) 15 2 1 5 15 FYR Macedonia (213) Romania (212) 1 5 Albania (212) Turkey (213) Moldova (214) 5 1 15 2 25 Poverty rate the international poverty lines of 2.5 USD PPP (left) and relationship between GDP per capita in USD PPP and the international poverty rate at 2.5 USD PPP (right). Note: World Bank staff calculation using a consumption aggregate drawn from ILCS data and harmonized for international comparison. Source: World Bank s data base of harmonized consumption data (ECAPOV accessed 3/15/216). 8

Shared prosperity: Between 29 and 214 economic growth has contributed to consumption growth for households in the bottom 4. The distribution of growth has not been pro-poor: relatively poor households experienced larger welfare losses during crisis and slower growth thereafter. Shared prosperity growth rate of average per capita consumption expenditure 6 bottom 4 percent total population 4 2-2 3.38 3.42 3.58 1.64 2.35.69 29-214 29-21 21-212 212-214 -4-6 -5.47-8 -1-7.54 Trend in shared prosperity: growth rate of average per capita consumption expenditure (%). Note: World Bank staff calculation using a consumption aggregate drawn from ILCS data and harmonized for international comparison. Source: World Bank s data base of harmonized consumption data (ECAPOV accessed 3/15/216). 9

Montenegro (29-214) Serbia (28-213) Albania (28-212) Kyrgyz Republic (29-214) Armenia (29-214) Turkey (28-213) Ukraine (29-214) Georgia (29-214) Moldova (29-214) FYR Macedonia (29-213) Russian Federation (27-212) Belarus (29-214) Kazakhstan (29-213) An international comparison shows that during time period from 29 to 214, Armenia (1) belongs into the group of countries where consumption growth for the bottom 4 and the total population has been positive, but (2) the bottom 4 grew on average slower than the total population. 1 8 6 4 2-2 -4 Patterns of shared prosperity in Europe and Central Asia, circa 29-214 Bottom 4% Total Population Patterns of shared prosperity in Europe and Central Asia using the growth rate of consumption for the bottom 4 of the welfare distribution and the total population (circa 28 to 213). Note: World Bank staff calculation using a consumption aggregate drawn from ILCS data and harmonized for international comparison. Source: World Bank s data base of harmonized consumption data (ECAPOV accessed 3/15/216). 1

Recent trends Economic growth, poverty reduction and shared prosperity Economic recovery increased welfare for all households in Armenia and led to a decline in poverty. Consumption growth shows fluctuations and disparities across the welfare distribution. Dynamic patterns of poverty illustrate large number of vulnerable households.

Period of poverty reduction between 21 and 214 happens in a context of sluggish economic growth: following the global economic crisis which hit Armenia in 29, the country is now growing at a much lower rate than during the pre-crisis period. 2 Growth rate of GDP per capita 15 14.3 13.7 14.3 1 5 1.8 7.4 2.3 4.5 6.8 3. 2.9 1.7 2.4 2.8 2.9-5 24 25 26 27 28 29 21 211 212 213 214 215 216 217 218-1 -15-13.9-2 Growth rate of GDP per capita. Note: Growth rates are based on year on year changes of real GDP (at constant market prices). Source: World Development Indicators published by the World Bank. Forecast for 215 to 217 published by the World Bank (March 216). 12

Poverty in Armenia declined from 35.8 percent in 21 to 3. percent in 214 (national poverty line). Over the same time period, poor households experienced positive consumption growth and moved closer to poverty line. But this pace of poverty reduction is slower than in pre-crisis years (24-27). 6 National poverty rate in Armenia between 24 and 214 poverty headcount poverty gap squared poverty gap 5 4 3 2 1 24 25 26 27 28 29 21 211 212 213 214 National poverty rate in Armenia between 24 and 214. Note: National consumption aggregate (using methodology adopted in 29). Source: Social Snapshot and Poverty 215. 13

All parts of the welfare distribution benefited from most recent growth period and managed to increase their consumption levels. However, reduction in poverty could have been bigger if poor households would have grown at an equal pace as relatively richer households. 8 6 4 2-2 -4-6 -8-1 -12 Datt Ravallion decomposition for Armenia, 21 to 214 poverty rate poverty gap Percentage point change Change due to growth Change due to redistribution Growth incidence curve (left) and Datt Ravallion decomposition (right) for Armenia, 21 to 214. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214 14

Moreover, growth rates of consumption for different deciles of the welfare distribution show (1) large fluctuations over time and (2) substantial disparities between different groups of the welfare distribution. 25 2 Growth rates of consumption for different deciles of welfare distribution (from the poorest to the richest). 15 1 5-5 29 21 211 212 213 214-1 -15 Growth rates of consumption for different deciles of welfare distribution (from the poorest to the richest). Note: National consumption aggregate (using methodology adopted in 29); year on year changes between 28 and 214. Source: World Bank staff calculations based on Armenia ILCS 28 to 214 15

Gini coefficient of.277 in 214 indicates moderate to low levels of inequality: (1) inequality increased mildly between 28 and 214; (2) levels and change in inequality appear to be biggest for Yerevan; (3) inequality mostly within regions not too much inequality between regions. Table X: Consumption inequality in Armenia Table X: Consumption inequality in Armenia, Gini coefficient Gini coefficient Theil (α=-1) Yerevan Other urban Rural Year Year 28.242.97 28.247.246.222 29.257.111 29.269.258.23 21.265.115 211.266.117 212.269.123 213.271.124 214.277.129 21.297.241.234 211.35.248.27 212.279.271.248 213.289.286.231 214.341.244.215 Patterns of inequality on the national level and by location. Note: National consumption aggregate (using methodology adopted in 29). Gini and Theil coefficient range from to 1, and Gini coefficient of (Theil of ) means perfect equality. Source: World Bank staff calculations based on Armenia ILCS 28 to 214. 16

Vulnerability to poverty persists, and reflects idiosyncratic and aggregate shocks. Between 21 and 214, 1.6 percent were poor in both periods and only 53.8 percent were not affected by poverty as defined by upper poverty line. Simulated transition between poverty and non-poverty between 21 and 214 53.8 23.8 1.6 11.7 POOR - POOR POOR- NON POOR NON POOR - POOR NON-POOR Simulated transition between poverty and non-poverty between 21 and 214. Note: Patterns of economic mobility are estimated on a sample that is restricted to households with household head between 25 and 55 years. National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 28 to 214. 17

Picture of Poverty Who are the poor, how do they experience poverty and where do they live Poor households differ in demographics and endowment from non-poor households. Poor households experience worse housing conditions and lack access to public services also regional variation. Poor households are concentrated in and around Yerevan; regional poverty rates are highest in other urban areas.

Who are the poor: differences in demographics and gaps in the endowment such as health and education across the welfare distribution. Poor households have higher dependency ratios and are more often female headed; multiple deprivation related to basic education and health status. 3 25 2 15 1 Dependency ratios 6 5 4 3 2 No household member has more than secondary education 25 2 15 1 At least one household member had to interrupt daily routine because of health problems. 5 1 5 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Quintile 1 Quntile 2 Quintile 3 Qunitile 4 Quintile 5 Quintile 1 Quntile 2 Quintile 3 Qunitile 4 Quintile 5 4 3 2 1 Share of female headed households 12 1 8 6 4 2 At least one child between 6 and 17 did not attend school. 25 2 15 1 5 Household was not able to make use of health services because of lack of money. Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Quintile 1 Quntile 2 Quintile 3 Qunitile 4 Quintile 5 Quintile 1 Quntile 2 Quintile 3 Qunitile 4 Quintile 5 Deprivations linked to different dimensions of multidimensional poverty in Armenia. Note: Source: World Bank staff calculations based on Armenia ILCS 214 19

How do they experience poverty: non-monetary measures of welfare highlight development gaps, which are persistent over time. Poor households report worse housing conditions and inferior access to public services; differs systematically across five quintiles of the welfare distribution and between locations. 45 4 35 3 25 2 15 1 5 No access to adequate housing conditions year 214 year 21 6 5 4 3 2 1 No access to healthy heating (gas or electricity) year 214 year 21 7 6 5 4 3 2 1 Living in a place which is considered overcrowded year 214 year 21 8 7 6 5 4 3 2 1 Not having access to hot running water year 214 year 21 4 3 2 1 Adequate housing 1 8 Yerevan 6 Other urban 4 Rural 2 Healthy heating 6 5 4 Yerevan 3 Other urban 2 Rural 1 Overcrowding 1 8 Yerevan 6 Other urban 4 Rural 2 Hot running water Yerevan Other urban Rural 21 214 21 214 21 214 21 214 Deprivations linked to housing and infrastructure for different quintiles of the welfare distribution for 21 and 214 (top) and by location (bottom). calculations based on Armenia ILCS 21 and 214. Source: World Bank staff 2

Where do they live: poverty rates are highest in other urban areas, and lowest in the capital city Yerevan. Spatial disparities between marzes in Armenia are high. 5 Poverty rates by location differences between Yerevan, other urban areas and rural areas 5 Regional poverty rates for eleven different marzes in Armenia 45 4 rural areas 45 4 35 35 3 25 2 other urban areas 3 25 2 15 1 Yerevan 15 1 5 5 28 29 21 211 212 213 214 Poverty headcount differences between Yerevan, other urban areas and rural areas (left) and differences on the marz level (right). Note: National consumption aggregate (using methodology adopted in 29). Source: Social Snapshot and Poverty 215. 21

The concentration of population around capital city influences the distribution of poor households across marzes: 41 percent of the poor live either in Yerevan or Kotayk. Outside Yerevan, number of poor households is large in other urban areas in the country. Distribution of poor across marzes Sjunik Aragatsotn Vayots Dzor Tavush Gegharkuni Yerevan Ararat Armavir Kotayk Lori Shirak Distribution of poor across marzes (left) and number of poor households by administrative district (left). Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 214 and ILCS 211 (for poverty map). 22

The poor tend to (1) live in bigger households with larger number of dependents, (2) reside in other urban areas of the country, (3) have lower education, (4) show lower employment rates, and (5) are less likely to receive (attract) international remittances. Household characteristics which Rural Urban Baseline: capital city Yerevan Household income: remittances greater than mean Percent of adults employed Number of children Number of people 65 and over in HH Number of adults College High school or vocational Less than secondary Education of household head - baseline: primary education decrease probability of being poor increase probability of being poor Household head age (squared) Household head age Household head is female -4% -3% -2% -1% % 1% 2% 3% Household characteristics which decrease or increase probability of being poor. Note: Figure shows marginal effects estimated from a probit model with dependent variable being one if the household is poor. Results are obtained from repeated cross section between 21 and 214. Source: World Bank staff calculations based on Armenia ILCS 214. 23

Sources of income growth and poverty reduction Composition of income and changes over time Income from labor markets and pensions are the biggest income share for poor and non-poor households. Composition differs between poor and non-poor households; lower share of labor income for the poor. Income growth and poverty reduction benefited from labor markets, agricultural sales, remittances and pensions.

Both, poor and non- poor households experienced growth of income from 21 to 214. Composition differs systematically between poor and non-poor households: poor have lower levels and shares of (1) labor income and (2) remittances, and depend more on (3) pension income and (4) public transfers. 7, Income from different sources of income 6, 5, 4, 3, 2, 1, 25.5 49.6 7.1 29.6 47.3 9. 6. 21.6 56.4 1.4 7.6 2.3 56.6 Assets Private transfers Public transfers Remittances Agriculture Pension Labor poor 21 poor 214 non-poor 21 non-poor 214 Income from different sources of income (left scale constant 214 AMD, per capita); labels show relative composition. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 25

Regional patterns in the composition of income highlight dominant role of labor earnings and pensions. Also: (1) sales from agricultural production and labor market earnings in agriculture in rural areas, (2) remittances in other urban areas, and (3) private transfers (internal migration) in Yerevan. 8, Income from different sources of income - by location 7, 7.3 Assets 6, 2.7 Private transfers 5, 4, 3, 2, 1, 7.8 16.5 3.3 35.1 9.5 9.5 18.4 26.7 38.7 2.5 59.6 11.4 6.4 21.6 57.7 17.8 68.2 66.6 Public transfers Remittances Agriculture Pension Labor rural 21 rural 214 urban 21 urban 214 Yerevan 21 Yerevan 214 Income from different sources of income (left scale constant 214 AMD, per capita); labels show relative composition. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 26

Income growth and decline in poverty between 21 and 214 driven by (1) higher employment and wages, (2) increasing agricultural output but also employment in rural areas, and (3) additional remittances and (4) private transfers; also, (5) pensions helped households to escape poverty. Drivers of poverty reduction between 21 and 214 Employment -3.3 Labor income -3.8 Agricultural income -1.3 Remittances -2.6 Private transfer.2 Pensions -2.5 Public transfer.5 Property.2 Dependency rate.6-4.5-4. -3.5-3. -2.5-2. -1.5-1. -.5..5 1. Drivers of poverty reduction to the change in the poverty rate between 21 and 214. Note: Negative estimated coefficients describe by how many percentage points corresponding factor has contributed to poverty reduction (and vice-verse). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 27

Behind the aggregate figures Three facts on drivers of income growth and challenges

Cheat Sheet to Three Facts Fact 1 Structural transformation and labor markets Fact 2 Migration Fact 3 Fiscal and social policy Private sector growth has raised incomes for all households. How will structural transformation reshape the economy? Internal and external migration have extended domestic labor markets. How to explore opportunities and mitigate challenges? Social expenditures have protected poor and vulnerable households. 29 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

Fact 1 Structural transformation and labor markets Private sector growth has raised incomes for all households. Growth of services and manufacturing translated into new employment opportunities and higher wages. Increasing agricultural sales, high productivity services in Yerevan and limited job creation in other urban areas. Labor market outcomes for poor households: lower quality of employment and lower productivity of labor.

Structural transformation: growth of services and manufacturing (excluding construction) translated into new employment opportunities and higher wages. Reallocation between different sectors of the economy. GDP production, by sector 214 21 2,5, Total employment, by sector 214 21 6, Monthly wages, by sector 214 21 25, 2,, 5, 2, 1,5, 4, 15, 1,, 3, 2, 1, 5, 1, 5, GDP production (constant 214, million AMD), by sector (left) total employment, by sector (middle) and monthly wages (constant 214, AMD), by sector (right). Note: Sector by NACE 2 classification. Source: Central Bank of Armenia (accessed 3/15/216). 31

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Patterns of GDP growth describe differential performance across sectors of the economy. Expansion of agricultural sector and growth of industry and services; construction sector decreased substantially. GDP production, by sector 214 21 2,5, GDP production, only service sector 214 21 6, 2,, 5, 4, 1,5, 3, 1,, 2, 5, 1, Structure of GDP production (constant 214, million AMD), by sector (left) and more disaggregated view into the service sector (right). Note: GDP production (NACE 2 classification) at current prices, million Armenian drams Source: Central Bank of Armenia (accessed 3/15/216). 32

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Total employment in Armenia decreased marginally between 21 and 214 also driven by a decline of the working age population due to demographic shifts and migration. However, analysis by sector of employment illustrates that structural transformation led to substantial shifts between sectors. Total employment, by sector 214 21 6, 5, Total employment, only service sector 214 21 14, 12, 4, 1, 3, 2, 8, 6, 4, 1, 2, Services Construction Industry Agriculture Structure of total employment, by sector (left) and more disaggregated view into the service sector (right). Note: Total number of employed. Source: Central Bank of Armenia (accessed 3/15/216). 33

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles All sectors of the economy show a positive wage growth between 21 and 214. However, large gaps between sectors (in services) which link to differential labor productivity. Low productivity in agricultural sector. Monthly wages, by sector 214 21 25, 2, Monthly wages, only service sector 214 21 4, 35, 3, 15, 25, 2, 1, 15, 1, 5, 5, Services Construction Industry Agriculture Monthly wages (constant 214, AMD), by sector (left) and more disaggregated view into the service sector (right). Note: Monthly wages in AMD. Source: Central Bank of Armenia (accessed 3/15/216). 34

Regional view: Reallocation of employment for Yerevan, other urban areas and rural areas. Job creation in other urban areas to provide employment opportunities for people leaving agriculture. Sector of employment (and employment rates), by location 6, 5, 6.7 62.1 7% 6% 4, 3, 2, 39.8 41.8 37.1 41.8 5% 4% 3% 2% 1, 1% Yerevan 21 Yerevan 214 Other urban 21 Other urban 214 Rural 21 Rural 214 annualized change in labor income agriculture manufacturing construction service % Sector of employment 21 and 214, by location (left), annualized change in labor income (right). Note: Employment rates (on top of each bar) calculated based on a working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 35

Total Agriculture, hunting, forestry and fishing Mining and quarrying Manufacturing Electricity, gas, steam and air conditioning supply Water supply, sewerage, waste management and Construction Wholesale and retail trade; repair of motor vehicles, Тransportations and warehouse economy Accommodation and food service activities Information and communication Financial and insurance activities Real estate activities Professional, scientific and technical activities Administrative and support service activities Public administration Education Human health and social work activities Arts, entertainment and recreation Other service activities Activities of private households as employers Where are the jobs? Agriculture is concentrated in rural areas; 85 percent of service sector is in Yerevan. In times of declining employment in agriculture and construction, an intermediate sector in other urban areas could create new opportunities for additional employment yet, geographic mobility matters. Employment by location, different sectors of the economy 1 9 8 7 6 5 4 3 2 1 Yerevan other urban rural Employment by location, different sectors of the economy. Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 214. 36

Distributional view: Structural transformation also works for the poor. Helped to reduce poverty between 21 and 214. Yet, compared to the total population, poor experienced lower growth of labor income. 2, 18, 16, 14, 12, 1, 8, 6, 4, 2, Sector of employment for the poor (and employment rates), by location Yerevan 21 31.2 37.3 Yerevan 214 31.5 Other urban 21 34.6 Other urban 214 57.7 57.5 agriculture manufacturing construction service Rural 21 Rural 214 annualized change in labor income 7% 6% 5% 4% 3% 2% 1% % Sector of employment for the poor in 21 and 214, by location (left), annualized change in labor income (right). Note: Employment rates (on top of each bar) calculated based on a working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 37

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Construction Water supply, sewerage and waste management Electricity, gas and steam supply Manufacturing Mining and quarrying Agriculture, hunting, forestry and fishing Patterns of employment across sectors of the economy document: (1) no major gaps in the sector of employment between workers from poor and non-poor households; (2) all parts of the welfare distribution have experienced shifts in sector of employment. Sector of employment (poor and non poor) in 21 and 214 214 poor 21 poor 214 non poor 21 non poor 5 45 4 35 3 25 2 15 1 5 Sector of employment (poor and non poor) in 21 and 214. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 38

Status of employment: Poor are less attached to labor markets and work fewer hours. Higher unemployment rates among individuals from poor households and higher levels of underemployment in rural areas. Labor market status Hours worked per week 1 1 9 8 7 6 5 4 3 23.6 29.9 14.3 16.2 5.7 6.4 2.4 13.2 11.1 7.7 Hired employee Own account worker Contributing family worker Unemployed Pensioner Student 9 8 7 6 5 4 3 18.5 19.3 39. 45.4 13.2 8.7 32.2 54.2 15.1 13.2 28.9 58.5 14.7 19.3 5 hours plus 4-49 hours 3-39 hours 2-29 hours 2 1 6.6 6.7 18.2 2. poor 214 non-poor 214 Out of labor force 2 1 7. 16.9 11.7 8.9 4.7 23.9 12.4 14.7 6.7 1.4 poor non poor Yerevan urban rural 1-19 hours Labor market status for individuals from poor and non-poor households (left) and educational attainment for individuals from poor and non-poor households (right). Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 39

Quality of employment: Poor have less protection and worse contracts. Literature links both to vulnerability of employment (and welfare) and highlights relationship between informality and low productivity. Type of worker Type of contract 1 9 8 7 16. 4.5 14. 2.8 3. 6.1 1.5 3.5 24. 1.2 occasional 1 9 8 7 41.4 5.9 6.7 23. 6.1 written contract verbal agreement 6 seasonal 6 78.8 employer 5 4 3 2 1 88.9 82.1 84.1 77.2 74.1 poor non poor Yerevan urban rural temporary permanent 5 4 3 2 1 12.6 41.8 6.1 24.4 21.9 11.1 4.5 12.1 8.1 7.9 7.2 24.5 11.5 13.1 12.2 1.2 3.9 poor non poor Yerevan urban rural own account in agriculture other own account contributing family member Status of employment (left) and type of contract (right) by location and poverty status of households. Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 4

Productivity of employment: Poor have less education which then translates into lower wages. Higher levels of education allow for productivity gains if demand and supply of labor markets is well-aligned. Male Baseline: Other Industry Services Baseline: capital city Yerevan Urban Rural Baseline: primary education Less than secondary HS or vocational College decrease in labor income increase in labor income Baseline: agegroup 2 to 24 years agegroup 25 to 29 years agegroup 3 to 34 years agegroup 35 to 39 years agegroup 4 to 44 years agegroup 45 to 49 years agegroup 5 to 54 years agegroup 55 to 59 years agegroup 6 to 64 years -3% -2% -1% % 1% 2% 3% 4% 5% 6% Individual characteristics which decrease or increase labor income earned from employment outside agricultural sector. Note: Figure shows estimated coefficients from a Mincer labor income regression where the dependent variable is the logarithm of labor income in constant AMD 214. Results are obtained from repeated cross section between 21 and 214. Source: World Bank staff calculations based on Armenia ILCS 214. 41

Fact 2 Migration Internal and external migration extend domestic labor markets Opportunity for migration expands domestic labor market and generates additional income from remittances. Role of internal and external migration as an equalizer between locations (poverty in other urban areas). Sustainability: dependence on remittances also increases vulnerability through contagion of external shocks.

remittances 21 remittances 214 private transfers 21 private transfers 214 Labor mobility characterizes small open economy: across all parts of the welfare distribution, 15.9 percent of households receive remittances from international migration and 3.5 percent of households receive private transfers from within the country. 25 Share of households receiving remittances and private transfers by location 1,, Composition of income by quintile of welfare distribution 2 Yerevan 9, 8, Assets Private transfers 15 1 5 urban rural 7, 6, 5, 4, 3, Public transfers Remittances Agriculture Pension 2, Labor 1, 1 2 3 4 5 Share of households receiving remittances and private transfers by location (left) and composition of income by quintile of welfare distribution (constant 214 AMD per capita, right). Source: World Bank staff calculations based on Armenia ILCS 21 to 214. 43

Individuals living in households which receive remittances or private transfers, are on average (1) better educated and have (2) lower employment rates than households which do not receive remittances. Educational attainment of individuals living in households receiving Labor market status of individuals living in households receiving... 1 9 8 7 6 5 4 3 2 1 26.4 24. 42.7 36.6 24.8 34.2 29. 27. 37.2 6.9 4.4 6.9 neither remittances nor private transfers private transfers remittances Bachelors and Post Grad College/vocational High school General/primary or less 1 9 8 7 6 5 4 3 2 1 32.1 16.4 6.8 14.5 8.4 6.5 15.4 neither remittances nor private transfers 24.8 21.4 7.6 14.8 3.6 19.1 18.5 16.8 12.8 1.2 8.2 2.5 2.7 private transfers remittances Own account worker Contributing family worker Unemployed Non-participants Pensioner Student Out of labor force Educational attainment (left) and labor market status (right) of individuals living in households receiving either private transfers, remittances or none of both. Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 214. 44

Large majority of international migrants leaves the country to work in Russia. Domestic migration with movements between Yerevan and other parts of the country majority moves to Yerevan to study or leaves the household for family reasons. Destination for internal and external migrants Migrant activity conditional on destination 1 1 9 9 8 7 Russia 8 7 other, mostly family reasons, business or hospital 6 5 4 3 other, mainly: Europe and USA Armenia Yerevan 6 5 4 3 Study Work 2 2 1 1 21 211 212 213 214 Yerevan Armenia Russia Elsewhere Share of internal and external migrants by destination (left) and Population temporarily as migrants by reason (right). Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 to 214. 45

31/3/27 3/6/27 3/9/27 31/12/27 31/3/28 3/6/28 3/9/28 31/12/28 31/3/29 3/6/29 3/9/29 31/12/29 31/3/21 3/6/21 3/9/21 31/12/21 31/3/211 3/6/211 3/9/211 31/12/211 31/3/212 3/6/212 3/9/212 31/12/212 31/3/213 3/6/213 3/9/213 31/12/213 31/3/214 3/6/214 3/9/214 31/12/214 31/3/215 3/6/215 3/9/215 31/12/215 Remittances from abroad increased in 214 and supported consumption growth for all parts of the welfare distribution; fluctuations reflect shifts in economic conditions in sending countries and a increasing number of return migrants imposes pressure on domestic labor markets. 7, Quarterly inflow of non commercial flows (nominal) 6, 5, 4, 3, 2, 1, Quarterly inflow of non commercial flows (nominal, in million AMD). Note: Dotted line shows trend between 27 and 215. Source: Central Bank of Armenia (accessed 3/15/216). 46

Fact 3 Fiscal and social policy Social spending protect poor and vulnerable households Fiscal activity contributed to lower poverty and inequality. Pensions and family benefit program are important to lift households out of poverty.

Market Income Market Income plus Pensions Net Market Income Disposable Income Consumable Income Final Income Market Income Market Income plus Pensions Net Market Income Disposable Income Consumable Income Final Income Fiscal policy redistributes income in Armenia (status quo): (1) reduces poverty, (2) decreases inequality, and (3) expenditures on social protection, health and education are progressive. 35% 3% 25% 2% 15% 1% 5% Poverty: poverty headcount.4.35.3.25.2.15.1.5 Inequality: Gini coefficient % Plus pension Less direct taxes Plus direct transfers Less indirect taxes Plus in kind benefits Plus pension Less direct taxes Plus direct transfers Less indirect taxes Plus in kind benefits Findings for the fiscal incidence analysis in Armenia 214: poverty head count (left) and Gini coefficient (right). Note: Armenia, Fiscal Incidence Analysis 214. Source: Armenia ILCS 214. 48

Equity of expenditure and taxes depends on progressivity and size of transfer. Progressivity of taxes and spending in 214 (status quo): Near neutral (neither progressive nor regressive) for taxes; progressive for social spending. Progressivity: Kakwani index 214 All Social Protection Family Benefit Marginal effect of each transfer to inequality All Social Protection Family Benefit In-kind Education Benefits In-kind Education Benefits In-kind Health Benefits In-kind Health Benefits Direct Taxes Indirect Taxes -.5.5 1 Direct Taxes Indirect Taxes -.5.5.1 Findings for the fiscal incidence analysis in Armenia 214: Kakwani index 214 and marginal effects. Note: Armenia, Fiscal Incidence Analysis 214. Source: Armenia ILCS 214. 49

Total Quinitle 1 Quintile 2 Quinitle 3 Quintile 4 Quintile 5 Extreme poor Poor Non-poor Total Quinitle 1 Quintile 2 Quinitle 3 Quintile 4 Quintile 5 Extreme poor Poor Non-poor Social pensions and old age benefits are an important source of income: (1) more than half of households receive pension income; (2) targeting accuracy reflects that pensions are part of social insurance and not assistance; (3) generosity shows how important income from pensions is. 8 7 6 5 Coverage for pensions 8 6 4 2 Distribution of benefits (targeting accuracy) for pensions 4 3 2 1 6 5 4 3 2 1 Generosity for pensions Key indicator on family benefit program in Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 5

Total Quinitle 1 Quintile 2 Quinitle 3 Quintile 4 Quintile 5 Extreme poor Poor Non-poor Total Quinitle 1 Quintile 2 Quinitle 3 Quintile 4 Quintile 5 Extreme poor Poor Non-poor The family benefit program supports poor and vulnerable: (1) coverage decreases for higher quintile of the welfare distribution; (2) targeting accuracy suggests that large parts of the money go to the poor; (3) generosity illustrates how important these benefits are for the bottom of the welfare distribution. 4 35 3 25 Coverage for family benefit 6 4 2 Distribution of benefits (targeting accuracy) for family benefit 2 15 1 5 5 4 3 2 1 Generosity for family benefit Key indicator on family benefit program in Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 51

Policy Agenda

Basic policy agenda Context: Structural transformation of the economy and poverty reduction, but challenges to shared prosperity. Environment with global slowdown (in the short term), limited fiscal space and demographic transition (in the long term). 1 Supporting growth Deepen structural transformation and promote productivity and employment growth Raise agricultural productivity and non-agriculture private sector growth in other urban areas 2 Investing in endowment Expand asset endowment (skills) and improve access to and quality of basic services Enhance connectivity (infrastructure and people) 3 Protecting the poor and vulnerable Strengthen safety nets (family benefit program) to reduce vulnerability to shocks and reforms (energy prices) while being mindful of fiscal sustainability reforms 53 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

BACKGROUND INFORMATION ON DATA AND METHODOLOGY SOUTH CAUCASUS POVERTY TEAM POVERTY AND EQUITY GLOBAL PRACTICE

Data: Integrated Living Conditions Survey The National Statistics Service of the Republic of Armenia (NSS RA) conducts the Integrated Living Conditions Survey (ILCS) on an annual basis. The survey is the official source to monitor poverty and measure social indicators in the country. Analyses based upon household data inform decision makers in Armenia and serve as primary source of information on living standards for international organizations. Consequently, these data are heavily relied upon by the government. The NSSRA produces an annual publication - Social Snapshot and Poverty in Armenia - presenting analysis of poverty along with employment and other social indicators using the ILCS. This report is used by line ministries (such as the Ministry of Labor and Social Affairs) and members of the public. The ILCS data and documentation is available through the website www.armstat.am. 55 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

Methodology: national poverty measurement methodology for Armenia Official poverty estimates for Armenia are based on the "Cost of Basic Needs" approach which determines three different poverty lines: (1) the food poverty line which uses the minimum required level of calories; (2) the lower poverty line which refers to the "Consumption Basket Method;" and (3) the upper poverty line which makes use of the "Food Expenditure Method. In 214, the most recent year of published figures, the three different poverty headcount rates were 2.3, 1.9 and 3. percent, respectively (exploring a consumption aggregate which is corrected using an adult equivalence scale). The World Bank estimates a harmonized consumption aggregate from ILCS data which is then used for international comparisons (such as international poverty rates). The welfare aggregate includes food and nonfood consumption, durables and health expenditure. For more information on the micro data, see http://ecadataportal/. National poverty measurement methodology for Armenia Metric and concept Absolute poverty using consumption expenditure Welfare aggregate includes food and nonfood consumption Durables and own produced goods Price adjustments: regional prices Scaling: per adult equivalent Poverty line (revised in 29) Cost of basic needs: 2,232 calories Consumption patterns of the 2 nd to 4 th decile of the welfare distribution 56 Armenia: Poverty Reduction and Shared Prosperity, WB South Caucasus Poverty team

APPENDIX South Caucasus Poverty team Poverty and Equity Global Practice

APPENDIX: Recent trends Poverty trend using the international poverty lines of 2.5 USD PPP and 5. USD PPP Exchange rate movements between Armenian Dram and Russian Rubel, US Dollar and EURO Loans to enterprises and households Agricultural sector between 21 and 215 Income growth for bottom 4 and top 6

Poverty trend using the international poverty lines of 2.5 USD PPP and 5. USD PPP. 9 8 7 6 5 Poverty trend using the international poverty lines of 2.5 USD PPP and 5. USD PPP 83.3 76.1 8.6 84.5 82.1 79.6 78.4 75.9 2.5 USD PPP 25 4 3 2 34.4 26. 29.8 35.9 32.6 3.1 3.2 26.3 5. USD PPP 25 1 27 28 29 21 211 212 213 214 Poverty trend using the international poverty lines of 2.5 USD PPP and 5. USD PPP 25. Note: World Bank staff calculation using a consumption aggregate drawn from ILCS data and harmonized for international comparison. Source: World Bank s data base of harmonized consumption data (ECAPOV accessed 3/15/216). 59

11/1/21 5/5/21 26/8/21 17/12/21 2/4/211 16/8/211 7/12/211 6/4/212 6/8/212 28/11/212 4/4/213 1/8/213 22/11/213 26/3/214 23/7/214 13/11/214 17/3/215 13/7/215 11/1/21 2/5/21 28/9/21 1/2/211 22/6/211 1/11/211 19/3/212 31/7/212 7/12/212 3/4/213 11/9/213 29/1/214 11/6/214 17/1/214 5/3/215 16/7/215 11/1/21 12/5/21 9/9/21 14/1/211 19/5/211 19/9/211 25/1/212 31/5/212 2/1/212 8/2/213 18/6/213 17/1/213 25/2/214 1/7/214 29/1/214 9/3/215 1/7/215 Exchange rate movements between Armenian Dram and Russian Rubel, US Dollar and EURO. 16 14 12 1 8 6 4 2 Rubel exchange rate 6 5 4 3 2 1 USD exchange rate 7 6 5 4 3 2 1 EUR exchange rate Exchange rates in Armenia. Note: Source: Central Bank of Armenia (accessed 3/15/216). 6

11/1/21 12/4/21 13/7/21 8/1/21 12/1/211 14/4/211 15/7/211 14/1/211 19/1/212 19/4/212 25/7/212 24/1/212 31/1/213 7/5/213 8/8/213 6/11/213 13/2/214 2/5/214 19/8/214 14/11/214 23/2/215 26/5/215 24/8/215 21 - Jan 21 - May 21 - Sep 211 - Jan 211 - May 211 - Sep 212 - Jan 212 - May 212 - Sep 213 - Jan 213 - May 213 - Sep 214 - Jan 214 - May 214 - Sep 215 - Jan 215 - May 215 - Sep 216 - Jan In November 214, Armenia saw a substantial adjustment in the exchange rate between Armenian Dram and US Dollar: however, inflation did not increase beyond the inflation target; still, it affected households and firms through the real value of loans which are often denominated in US Dollar. 6 5 4 3 2 1 Exchange rate Armenian Dram and US Dollar 115 11 15 1 95 9 85 8 75 7 65 Inflation in Armenia CPI Food Services Exchange rates between Armenia Dram and US Dollar (left) and Inflation in Armenia (right). Source: Central Bank of Armenia (accessed 3/15/216). 61

31/12/25 3/6/26 3/12/26 1/6/27 1/9/27 29/2/28 31/8/28 1/2/29 1/8/29 1/2/21 1/8/21 28/2/211 31/7/211 31/1/212 31/7/212 3/12/212 31/5/213 31/1/213 31/3/214 3/9/214 31/3/215 3/9/215 31/12/25 3/6/26 3/12/26 1/6/27 1/9/27 29/2/28 31/8/28 1/2/29 1/8/29 1/2/21 1/8/21 28/2/211 31/7/211 31/1/212 31/7/212 3/12/212 31/5/213 31/1/213 31/3/214 3/9/214 31/3/215 3/9/215 Loans to enterprises and households denominated in Armenian Dram and Foreign Exchange. 12 Loans to enterprises 6 Loans to households 1 5 8 4 6 3 4 2 2 1 Loans to enterprises in AMD Loans to enterprises in FX Loans to households in AMD Loans to households in FX Loans to enterprises (left) and loans to households (right). Note: Source: Central Bank of Armenia (accessed 3/15/216). 62

Increase in agricultural output with lower levels of employment reflects higher productivity which led to a positive supply shock. In consequence: (1) lower producer and consumer prices, (2) lower food inflation and (3) increasing exports with limited diversification. Output of plant-growing and animal husbandry in million AMD (left), and price index for agricultural products (right) Exports from Armenia: Food and live animals - trade volume in US Dollar animal husbandry (left) Rest of the World plant-growing (left) Georgia 1,2, Price indices compared to december of the previous year (right) 15 12,, Russian Federation 14 1,, 13 1,, 8, 12 8,, 11 6, 1 6,, 9 4, 8 4,, 2, 7 2,, 6 21 211 212 213 214 215 5 Output of plant-growing and animal husbandry in million AMD (left scale), and price index for agricultural products (right scale) (left panel), and Exports from Armenia: Food and live animals - trade volume in US Dollar (right). Note: National consumption aggregate (using methodology adopted in 29). Source: Central Bank of Armenia and UN Comtrade data 21-215. 21 211 212 213 214 215 63

Income growth for bottom 4 and top 6 benefited from (1) higher levels of employment, (2) positive growth of wages, and (3) expansion of agricultural activity; (4) remittances contributed relatively more to income growth for the top 6, whereas (5) pensions raised incomes for the bottom 4. Contribution of different factors to income growth for the bottom 4 and the top 6 of the welfare distribution Per capita remittances income Per capita asset income Per capita public transfer income Per capita private transfer income Per capita pension income Per capita agriculture income Labor income per employed adult Employment rate Dependency rate Correction (ad eq. to pc) Consumption to income ratio -2.1-1.7 -.5 -.5 -.5 -.7 -.3. -.1 -.1.1.2.7.8 1.1 1.1 1.7 1.6 1.7 1.6 2. Bottom 4 Top 6 2.5-3. -2. -1.. 1. 2. 3. Contribution of different factors to income growth for the bottom 4 and the top 6 of the welfare distribution. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 64

APPENDIX: Structural transformation and labor markets Labor market status by location Education attainment and labor market status by gender Sector of employment (capital city Yerevan) in 21 and 214 by poverty status Sector of employment (other urban areas) in 21 and 214 by poverty status Sector of employment (rural areas) in 21 and 214 by poverty status

Labor market outcomes differ systematically across locations: labor force participation and employment rates are lowest in other urban areas, whereas the share of self-employed and other employed is highest in rural areas. 1 9 8 7 6 5 4 3 2 1 16.2 18. 26.3 18.2 15.2 1.6 1.3 4.5 6.7 5.7 28.8 6.9 Labor market status by location 29.1 3. 24.5 18.5 18.9 1.9 11. 13. 7.7 6.4 9.5 8. 17.5 18.6 19.9 22.3 18.8 17.8 1.1 rural 21 rural 214 urban 21 urban 214 Yerevan 21 35.1 38.1 4.3 5.9 17.4 12.5 Yerevan 214 Hired employee Own account worker Contributing family worker Unemployed Pensioner Student Out of labor force Labor market status by location. Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 66

Heterogeneity in education and labor markets does not only exist between poor and non-poor: even though women obtain more education than men (27.6 percent versus 24.7 for tertiary education), labor force participation rates and employment status are worse for women. 1 Education attainment by gender 1 Labor market status by gender 9 8 7 6 5 4 3 2 1 22.4 24.7 24.8 27.6 2.9 46.3 21.2 45.9 27. 4.4 27.4 39.9 1.3 8.2 7.9 5.1 male 21 male 214 female 21 female 214 Bachelors and Post-Grad College/vocational High school General/primary or less 9 8 7 6 5 4 3 2 1 33.3 31.9 15.1 18.3 5. 4. 15.9 14.6 7.7 7.3 7.4 6.3 15.6 17.6 male 21 male 214 2.2 11.1 8.9 19.2 1.7 24.8 13.2 8.2 15.6 9.9 8.4 6.9 21.5 21.2 female 21 female 214 Hired employee Own account worker Contributing family worker Unemployed Pensioner Student Out of labor force Education attainment by gender (left) and labor market status by gender (right). Note: Sample is restricted to individuals in the working age population between 15 and 75 years of age. Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 67

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Construction Water supply, sewerage and waste management Electricity, gas and steam supply Manufacturing Mining and quarrying Agriculture, hunting, forestry and fishing More than one third of the population lives in Yerevan and 29 percent of all jobs is concentrated in the capital city. 54 percent of output in construction, 42 percent in industry and 85 percent in the service industry are produced in Yerevan. Sector of employment (Yerevan) in 21 and 214 214 Yerevan 21 Yerevan 25 2 15 1 5 Sector of employment (capital city Yerevan) in 21 and 214. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 68

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Construction Water supply, sewerage and waste management Electricity, gas and steam supply Manufacturing Mining and quarrying Agriculture, hunting, forestry and fishing Labor markets in other urban areas saw a decline in jobs in the construction sector. Public sector and wholesale and retail trade offer additional employment opportunities. Still, other urban areas outside Yerevan do not create sufficient number of jobs to absorb surplus workers leaving agricultural sector. Sector of employment (other urban areas) in 21 and 214 214 other urban 21 other urban 18 16 14 12 1 8 6 4 2 Sector of employment (other urban areas) in 21 and 214. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 69

Undifferentiated production and services of private households Other services Arts, entertainment and recreation Human health and social work Education Public administration Administration and support services Professional, scientific and technical activities Real estate Finance and insurance Information and communication Accommodation and food services Тransportations and warehouse economy Wholesale, retail trade and repair of motor vehicles Construction Water supply, sewerage and waste management Electricity, gas and steam supply Manufacturing Mining and quarrying Agriculture, hunting, forestry and fishing In rural areas, 7 percent of employment in 214 was in the agricultural sector with a declining trend compared to 21. Even though welfare increased through higher productivity, increased incomes and improved quality of employment, less employment opportunities create challenges for rural areas. Sector of employment (rural areas) in 21 and 214 214 rural 21 rural 8 7 6 5 4 3 2 1 Sector of employment (rural areas) in 21 and 214. Note: National consumption aggregate (using methodology adopted in 29). Source: World Bank staff calculations based on Armenia ILCS 21 and 214. 7

APPENDIX: Fiscal and social policy Social expenditure: pensions and family benefit program Taxes: PIT progressive, VAT regressive Spending: (absolutely) progressive Key indicators on social expenditure: Pension, Family Benefit Program and Child Benefit Transfers from within Armenia and abroad

Social expenditure created favorable environment for private consumption growth: (1) pensions reduced poverty the incidence of poverty but grew at a lower pace than other income (increasing gaps by demographic groups); (2) family benefit program reduced the severity of poverty. 28 29 21 211 212 213 214 215 Total number of pensioners 469,747 467,555 465,84 454,488 452,55 453,917 458,569 462,539 Monthly average pension per individual, AMD constant 25 19,272 21,529 21,879 2,363 21,617 2,246 22,857 24,665 Change in average pension to previous year, inflation adjusted 11.7% 1.6% -6.9% 6.2% -6.3% 12.9% 7.9% Basic pension, AMD constant 25 5,81 6,61 8,19 7,449 8,993 8,51 8,89 9,722 Value of one year of service, AMD 45 45 5 Total number of FBP beneficiary households 121,16 17,493 15,5 91,575 96,36 12,57 14,13 15,3 Monthly average FBP spending per household, AMD constant 25 Change in average FBP to previous year, inflation adjusted 16,938 18,683 18,745 18,732 19,82 18,366 18,5 17,792 1.3%.3% -.1% 5.8% -7.3% -2.% -1.2% 72

Taxes: PIT progressive, VAT regressive Kakwani Index 214: PIT Personal income tax: PIT, self-employment PIT, passive income PIT, wages -.2.2.4.6.8 Three brackets; only.2 percent of households pay highest tax rate effective tax rate 25.6 percent. Informal employment, yet many formal employees directly below poverty line. Kakwani Index 214: VAT and Excises Value added tax: Alcohol excises Petroleum excises Import Duties VAT Tobacco excises Formally 2 percent with some exemptions effective tax rate around 1.67 (211 data). -.2.2.4.6.8 Findings for the fiscal incidence analysis in Armenia 214: Kakwani index for PIT (top) and Kakwani index for VAT and excises (bottom). Note: Armenia, Fiscal Incidence Analysis 214. Source: Armenia ILCS 214. 73

Spending: Even though transfers are (absolutely) progressive, limited size of social budget constraints impact on poverty. Kakwani Index 214: Social Protection Family Benefit Social Pension Subsidized Rent (Housing) Child Care Benefit -.2.2.4.6.8 Kakwani Index 214: Education (in-kind benefits) Social protection FBP targeted towards poor households (non) contributory pensions Education Impact depends on level of education Secondary School Pre-school -.2.2.4.6.8 Kakwani Index 214: Health In-patient Maternal Care -.2.2.4.6.8 Health benefits Relative progressive Findings for the fiscal incidence analysis in Armenia 214. Note: Armenia, Fiscal Incidence Analysis 214. Source: Armenia ILCS 214. 74

Key indicators on social expenditure: Pension Quintiles of per ae consumption, net of all SA transfers Poverty Status Pension Total Q1 Q2 Q3 Q4 Q5 XP MP NP Average Transfer Value, Per Capita 7,352.9 6,762. 7,518.6 7,481.2 7,265.4 7,736.7 5,85.6 7,269.2 7,434.5 Average Transfer Value, Per Capita, Beneficiary Households Of Indicated Transfer Only 13,316.1 1,89.7 12,447.1 13,911.9 14,633.5 15,563.1 8,717.1 11,931.2 14,137.6 Coverage 55.2 62.6 6.4 53.8 49.6 49.7 67.1 6.9 52.6 Distribution of Beneficiaries 1. 22.6 21.9 19.5 18. 18. 2.8 3.4 66.8 Distribution of Benefits (Targeting Accuracy) 1. 18.4 2.5 2.3 19.8 21. 1.8 27.3 7.9 Relative Incidence 15. 28.6 22.6 17.3 13.7 8.4 38.4 27.3 12.6 Generosity 28.6 45.5 36.9 31.5 27.3 18.1 56.9 44.4 24.9 Key indicator on Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 75

Key indicators on social expenditure: Family benefit Quintiles of per ae consumption, net of all SA transfers Poverty Status Family benefit Total Q1 Q2 Q3 Q4 Q5 XP MP NP Average Transfer Value, Per Capita 853.1 1,533.3 829. 823.4 69.4 39.1 2,77.6 1,294.3 639.8 Average Transfer Value, Per Capita, Beneficiary Households Of Indicated Transfer Only 6,434.6 5,834.4 6,155.3 6,767.2 7,263.4 7,99.5 5,642.3 6,18.1 6,917.8 Coverage 13.3 26.3 13.5 12.2 9.5 4.9 36.8 21.5 9.2 Distribution of Beneficiaries 1. 39.6 2.3 18.4 14.3 7.4 6.3 44.8 48.9 Distribution of Benefits (Targeting Accuracy) 1. 35.9 19.5 19.3 16.2 9.1 5.5 41.9 52.6 Relative Incidence 1.7 6.5 2.5 1.9 1.3.4 13.6 4.9 1.1 Generosity 18.5 26.9 19.5 16.6 14.7 11.1 41. 24.4 14.7 Key indicator on Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 76

Key indicators on social expenditure: Child benefit Quintiles of per ae consumption, net of all SA transfers Poverty Status Child benefit Total Q1 Q2 Q3 Q4 Q5 XP MP NP Average Transfer Value, Per Capita 47. 38.2 15.3 48.8 39.3 93.2. 31.3 54.6 Average Transfer Value, Per Capita, Beneficiary Households Of Indicated Transfer Only 3,615.5 3,176.3 2,897.5 3,592.2 3,81.2 3,93.3 n.a. 3,91.5 3,759.2 Coverage 1.3 1.2.5 1.4 1. 2.4. 1. 1.5 Distribution of Beneficiaries 1. 18.5 8.1 2.9 15.9 36.5. 21.5 78.5 Distribution of Benefits (Targeting Accuracy) 1. 16.3 6.5 2.8 16.8 39.7. 18.4 81.6 Relative Incidence.1.2..1.1.1..1.1 Generosity 7.2 13.6 9.4 9.1 7.3 5.4 n.a. 12.8 6.6 Key indicator on Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 77

Transfers from within Armenia Quintiles of per ae consumption, net of all SA transfers Poverty Status Transfers from within Armenia Average Transfer Value, Per Capita Average Transfer Value, Per Capita, Beneficiary Households Of Indicated Transfer Only Total Q1 Q2 Q3 Q4 Q5 XP MP NP 666. 227.3 256.5 555.4 545.9 1,744.5 18.4 29.2 831.9 19,199.6 9,194.5 11,582.7 14,54.2 13,926.8 35,475.2 5,384.1 1,719.9 21,82.5 Coverage 3.5 2.5 2.2 3.8 3.9 4.9 2. 2.7 3.8 Distribution of Beneficiaries 1. 14.2 12.8 22. 22.6 28.4 1.3 21.5 77.1 Distribution of Benefits (Targeting Accuracy) 1. 6.8 7.7 16.7 16.4 52.4.4 12. 87.6 Relative Incidence 1.4 1..8 1.3 1. 1.9.7 1.1 1.4 Generosity 29.3 38.1 34.3 31.4 25.7 28.5 28.3 39.3 28.3 Key indicator on Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 78

Transfers from abroad Quintiles of per ae consumption, net of all SA transfers Poverty Status Transfers from abroad Total Q1 Q2 Q3 Q4 Q5 XP MP NP Average Transfer Value, Per Capita 5,451.6 2,19.2 3,15.8 4,611. 6,564.3 1,819.8 853.9 2,681.1 6,69.5 Average Transfer Value, Per Capita, Beneficiary Households Of Indicated Transfer Only 34,225.4 19,653.7 22,87.9 29,71.7 32,612.8 56,525.9 28,936.4 21,376.4 37,839.7 Coverage 15.9 1.7 13.8 15.9 2.1 19.1 3. 12.5 17.7 Distribution of Beneficiaries 1. 13.5 17.3 19.9 25.3 24..4 21.7 77.9 Distribution of Benefits (Targeting Accuracy) 1. 7.7 11.6 16.9 24.1 39.7.4 13.6 86.1 Relative Incidence 11.1 8.9 9.5 1.7 12.4 11.8 5.6 1.1 11.4 Generosity 65.2 81.7 69.3 68.2 6.9 63.1 179.3 8.2 63.1 Key indicator on Armenia. Note: Quintiles of per ae consumption, net of all SA transfers. Source: Armenia ILCS 214. 79

THANK YOU