Poverty Diagnosis in Syria

1
Poverty Diagnosis in Syria
2
Questions

• How have the gains from recent growth affected
  the welfare of the Syrian people?
  
• Who are the poor, where do they live?
• What factors most affect whether a person or
  family is poor?

3
Main findings

• Poverty in Syria fell from 14.4 percent to 11.4
  percent.
  
• The poor accounted to 11.4 percent of the
  population (2.02 million people) in 2003-2004,
  using lower poverty line 30.12 percent were
  poor, using upper poverty line.
• Poverty incidence is higher in rural areas , but
  regional differences are more pronounced.
  
• Highest poverty rates are found in North-Eastern
  regions.
  
• The strongest correlate of poverty was
  education.
  
• Poverty correlates with unemployment, employment
  in private informal sector, agricultural
  activities and large households.

Percentage of the poor
4
Data Sources

• The principle survey instrument Household Income
  and Expenditure Survey conducted by CBS in
  1996/97 and 2003/2004
  

5
Identification of the poor

• What is the standard of living indicator to
  measure welfare? Total expenditure.
  
• How to distinguish between the poor and
  non-poor? Draw a poverty line
  
• How can poverty levels be expressed in one
  indicator? Headcount, poverty gap and severity of
  poverty indices
  

6
Conceptual Issues Income versus expenditure

• Favoring expenditure, as opposed to income,
• All income is not consumed
• Income fluctuation is greater
• Difficult to measure incomes of people who
  operate their own business records of family
  businesses are often not kept
  
• Respondents may be more willing to reveal their
  consumption patterns rather than their income.
  

7
Methodology for Constructing Poverty Lines

• Traditional methods in estimating poverty lines ,
  including 1 and 2 suffer from problems
  
• If calculated for Syria as a whole, they ignore
  differences in consumption patterns and prices
  across regions
  
• May not account for the differing basic needs
  requirements of different household members
  
• For example, young versus old, male versus
  female
  
• Usually ignore economies of scale within
  households
  
• Non-food items can be shared among household
  members (one person using it does not decrease
  the consumption of another).
  
• So, living in a larger household can result in
  lower per-person expenditures to maintain a given
  standard of living.

8
Household specific poverty line
characteristics

• the cost of basic needs methodology is used.
• the food bundle is consistent with the
  consumption behavior of the poor.
  
• bundle is evaluated using prices in each region
  and at each date.
  
• The poverty lines can be interpreted as
  Laspeyres cost-of-living numbers.
  
• Food poverty line is augmented by an allowance
  for expenditure on essential non-food goods.
  

9
Household specific poverty line
Characteristics, cont.

• Differences in poverty lines reflect variations
  in the food and non-food prices across the seven
  regions.
  
• They also incorporate household differences in
  the size, age and gender composition.    
  

10
Household specific poverty line Construction

• Stage 1
• Estimate minimum caloric requirements for
  different types of individuals (using tables from
  WHO , caloric needs are separately specified for
  urban and rural individuals, by sex and 13 age
  categories).
• Obtain calories requirement for each household.

11
Household specific poverty line
Construction, cont.

• Stage 2 Cost of the minimum level of calories
• Average quantities of all food items of the
  second quintile, reflecting consumption
  preferences of the poor, is obtained.
  
• Total calories generated by this bundle is
  calculated using calories contents in every food
  items.
  
• Bundle is augmented/ deflated to meet food
  requirements for each household.
  
• Bundle is priced using prices in each region to
  obtain household specific food poverty line.

12
Household specific poverty lineConstruction,
cont.

• Stage 3
• Food shares are regressed against logarithm of
  total household expenditure, logarithm of
  household size, share of small and older
  children, share of adult males and females, and
  share of elderly, thus non-food allowance for
  each household is estimated.
• Non food share differ according to region,
  household size, the share of children, elderly,
  adult males and adult females
  
• Households with the same gender and age
  decomposition in each region have the same
  poverty lines.
  
• This approach takes into account location, age,
  gender decomposition and the sharing behaviors
  among household members.

13
Household Specific Poverty Lines

• For each household in the sample, the study
  constructs its own food poverty line, which
  satisfies the households minimum nutritional
  requirements depending on its age, gender
  composition and location.
• This is then augmented by a non-food component to
  devise the overall poverty line.
  
• Although there are multiple poverty lines for
  Syria, depending on age and location, the average
  for southern urban is SL1,664, for Eastern
  region is SL 1454 and overall is SL1458.

14
Poverty Measures in 2003/2004

• Overall poverty in Syria was 11.4 in 2003/2004.
  
  
• Almost 2.02 million people could not obtain their
  basic food non-food needs.
  
• Poverty was relatively shallow and not very
  severe.

15
Regional Variations in Poverty Incidence,
2003-2004 (P0, of individuals)
16
Regional Distribution of Poor Households,
2003-2004
Rural
Urban
17
Incidence of poverty Using Different Poverty Lines
18
Incidence of Poverty by Governorate, 2003/2004
19
Regional Changes in Poverty, 1996-97 to 2003-04
Urban
Rural

• The incidence of poverty has increased in rural
  areas of North-Eastern and Coastal regions
  
• All other regions experienced declines in their
  poverty measurements

20
Growth and Distribution Across Regions

• All regions except Urban Coastal and Rural
  Southern, Coastal and North-Eastern regions
  showed growth in expenditures with increased
  inequality
• Growth in Urban Coastal region and Rural Southern
  region were pro-poor, as inequality decreased.
  
• Both low growth and increased inequality
  contributed to poverty growth in Rural Coastal
  and North-Eastern regions

21
Poverty- Growth Nexus

• According to the HIES, Poverty declined due to an
  annual increase in Real Per Capita Expenditure of
  2, that outweighed the worsening distribution.
  
• Yet according to National Accounts per capita GDP
  growth was less than 1 yearly in 1996-2002, and
  salaries adjusted for inflation also grew of 0.8
  yearly between 1997 and 2001.

The inconsistency can be explained by the
important role played by INFORMAL INCOME
GENERATION MECHANISMS (informal employment and
workers remittances)
22
Sources of Growth, 1996 - 2002
Growth contribution by Expenditure
Growth contribution by Sector
The reliance on the oil sector is reflected on
the big contribution to growth given by Net
Export (demand side) and Mining Manufacturing
(supply side). The investment component instead i
s alarmingly low, suggesting that return to
investment is insufficient to prompt capital
accumulation, especially in the manufacturing
sector.
23
Growth Elasticity

• Poverty is relatively shallow in Syria, which
  implies that even small changes in growth may
  have very important effects on poverty numbers.
  
• Poverty in Rural Northern-Eastern region and
  the Rural Coastal region are considerably less
  sensitive to growth

24
Who are the Poor in Syria? Education and Poverty
Poverty Rates by Educational Attainment

• Low education is the strongest correlate of
  poverty
  
• 81 of the poor were illiterate or have primary
  education
  
• Poverty incidence, depth and severity decreased
  as education increased
  

25
Who are the Poor in Syria?Employment and
Poverty

• The poor had lower labor participation rates
  overall than did the non- poor
  
• Unemployment rates were correlated with poverty
  for Syria as a whole
  
• Unemployment rate for the poor reached 12 percent
  in both urban and rural areas, in 2003-2004,
  
• For the non-poor, it ranged from 7.4 percent in
  urban areas and 9 percent in rural areas

Incidence of Poverty by Employment Status
26
Who are the Poor in Syria?Sector of Employment

• While employment was dominated by the private
  sector in 2003-2004, especially in rural areas,
  private sector workers were also most likely to
  be poorer than public workers.
• Employment in the government or in a
  government-owned corporation (public sector)
  exhibits a clear correlation with welfare levels.
  Only 6.4 percent of individuals employed in the
  government are found in poor households.
• Individuals employed in formal private sector
  were distributed equally among the two poverty
  groups.
  
• Informal private sector seems to be the only
  sector of employment for the poor, as it provides
  jobs to the uneducated and unskilled workers,
  which are the main characteristics of the poor.

27
Incidence of Poverty by Sector of Employment
28
Who are the Poor in Syria?Type of Employment

• The poor were over-represented in agriculture,
  construction activities.
  
• Individuals with permanent jobs are less
  represented among the poor compared to non poor
  (by 15 percentage points in urban areas and 12
  percentage points in rural areas).
• Casual workers constitute 22 percent of poor
  employed persons in urban areas and 14 percent in
  rural areas, they are more represented in the
  poor groups specially in urban areas.

29
Household Size, Composition and Poverty

• Poor people typically live in larger households
  than non-poor, the difference accounting for 2.45
  persons per household.
  
• Rural areas has the largest household size for
  both poor and non poor, and the poverty gap is
  wider than urban/rural gap.
  
• Poor households have a relatively larger number
  of children and fewer members in the working age
  than the national average (average number of
  children in poor households living in urban areas
  is 3.25 versus 1.9 for non poor households).

30
Children in Poverty

• In Syria, a households poverty level was
  strongly correlated with the proportion of
  working children in the household. Percentages of
  working children are higher in rural regions than
  in urban regions, specially within the age group
  15-17 years.
• Child labor is the highest among boys in poor
  households in urban areas (9.6 percent among
  poor boys aged 6-14 years).
  
• Among poor individuals, 1.97 percent of males of
  age 10-15 years and 2.54 percent of females were
  illiterate, while the corresponding proportions
  for non-poor children were half these ratios.

31
non enrollment among children aged 6-14
Illiteracy rate among children aged 10-14
32
Gender and Poverty

• The majority of female-headed poor households, by
  far, were headed by widows. Widows who had
  children were even more likely to be poor.
  
• There was a gender gap in education, but this was
  primarily a rural phenomenon.
  
• The illiteracy rate for girls of age 12-15 years
  was almost twice that of boys. This gap held
  whether the households were poor or not.

33
Housing Conditions
Connected to Piped water

• The poor suffer from deteriorated housing
  conditions.
  
• Almost 90 percent of urban population and 70
  percent of rural was connected to the public
  water system.
  
• In rural areas, 71 percent of the non-poor had
  access to public water system, 62 percent of the
  poor did the corresponding gap in urban areas is
  only 3 percentage points.
• For sewerage, the gap was larger, at 28
  percentage points.

34
Vulnerability to Poverty
Poverty profiles are a useful way of summarizing
information on the levels of poverty and the
characteristics of the poor in a society.
Multivariate analysis is used to assess the
impact of a change, in a particular factor,
would have on the probability of an individual
being poor, if all other factors are kept
constant.
We relied on a two-step approach.
First log per capita expenditure is regressed on
several socio-economic variables
ln(Ci)Xi??i, Then the probability of household
i to be poor is estimated Viprob(ln(Ci))?((ln(zi)- Xi?)/?),
35
Consumption regression results

• Main Findings
• Household size has a significant and negative
  effect on the level of household per capita
  consumption in all seven regions.
  
• Controlling for household size, the presence of
  children age 0 to 6 has the strongest negative
  effect on household consumption.
  
• Larger shares of children age 7-15, the elderly
  and adult females also decrease household
  consumption.
  
• Households with a larger proportion of literate
  members, especially household members with
  university degrees or higher, have significantly
  higher per capita consumption levels.
•  Households with the less educated heads have
  significantly lower level of per capita
  consumption.
  
• Households that have heads employed in
  agricultural activities have a lower per capita
  consumption level.

36
Simulations

• A newborn child increases risks of poverty in all
  regions of Syria
  
• When a family member loses a job the household
  risk of poverty increases substantially.
  
• Households with the head working in blue color
  occupation are more likely to be poor than the
  households with white color occupation.
  
• Employment in the formal private sector
  substantially reduces the risk of poverty
  compared to employment in informal. private
  sector.
• A heads education level has greater impact on
  household poverty vulnerability in urban areas.
  Relative to households with illiterate heads, the
  probability of being poor is about 7.3 percent
  lower for households with literate heads.

37
Simulations Cont.,
Impact of changes in household characteristics on
poverty. Percent change.

38
Towards a Poverty Reduction Strategy for Syria

• The diagnosis suggests 5 main pillars for the
  strategy
  
• Job growth Set the economic foundations for
  more sustainable growth in jobs, productivity and
  incomes for the poorest groups.
  
• Education Ensure better incomes and
  opportunities for the future, for both men and
  women.
  
• Regional balance Ensure that growing regional
  disparities in incomes, opportunities and
  services are reversed.
  
• Safety net Better protect the most vulnerable
  in society, especially those who are unable to
  obtain sufficient incomes through the labor
  market.
• Monitoring Improve the quality and frequency of
  data collection and monitoring outcomes,
  especially at the regional level, to update and
  adapt the strategy.

39
Targeting Mechanisms

• GEOGRAPHIC TARGETING it is simple to use and
  promising, especially when used in combination
  with other criteria. Targeted areas should
  include
• Northern-Eastern region, where 58 of the poor
  live
  
• Rural areas, 61 of the poor live there
• Informal Housing Areas
• CHARACTERISTIC TARGETING significant household
  characteristics include
  
• Level of Education (particularly of the Head of
  Household)
  
• Economic activity (55.9 of poor work in
  agriculture and construction)