Almudena Fernandez Luis Felipe Lopez-Calva UNDP Regional Bureau for Latin America and the Caribbean London, 9 November 2009 Transitory Shocks, Permanent.

Almudena Fernandez

Luis Felipe Lopez-Calva

UNDP Regional Bureau for Latin America and the Caribbean

London, 9 November 2009

Transitory Shocks, Permanent Effects: Impact of the Economic Crisis on the Well-

Being of Households in Latin America and the Caribbean

Content

2

1. Objective

2. Conceptual Framework

3. Methodology

4. Empirical Evidence

5. Final Remarks

Objective

3

Results from the project The Effects of the Economic Crisis on Household’s Well-being in Latin America and the Caribbean conducted by UNDP’s Regional Bureau for Latin America and the Caribbean.

Aims to understand how transitory macroeconomic downturns have permanent effects on households’ well-being in Latin America and the Caribbean, as measured by various social indicators.

Developed a methodology to measure the effects of past crises on health, education and poverty and hired local teams to implement it in five countries

Conceptual Framework

4

Crises affect the well-being of households in the long-term depending on how they affect their stock or accumulations of income-generating assets1.

Economic crises can: decrease the stock of income-generating assets held by

households (using savings, selling physical assets);decrease the rate at which assets are used

(unemployment);decrease the market value of assets (wages, prices);affect the accumulation of income-generating assets for

future generations (education, health).

1 Attanasio, Orazio and Miguel Szekely (1999)

Conceptual Framework II

5

How much households choose to reduce their income-generating assets depends on1: Income effectsSubstitution effects Heterogeneity according to household characteristics

Which of these effects dominates determines the extent to which households forgo future consumption vis-á-vis current consumption.

1 Ferreira, F. and N. Schady (2008)

Methodology

6

The empirical analysis was conducted using data from past crises, and various methodologies: difference-in-difference fixed effects instrumental variables

“Economic crisis” were defined as an aggregate negative economic shock characterized by a sharp decline of the GDP per capita.

Outcomes of interest: health, education, and poverty and which is studied in each country depends on the availability of data at the country level.

Empirical Evidence: Poverty

7

8

A look at the data hints at poverty increases during economic downturns.

Evolution of Poverty in ArgentinaPoverty and GDP Per Capital in LAC

9

Country

Indicator

Results

Argentina

Overall poverty

A one percent decrease in per capita GDP is associated with an increase in poverty of 0.24 percentage points.

Child poverty

A one percent decrease in per capita GDP is associated with an increase in child poverty of 0.42 percentage points. When times of growth and contraction are differentiated a one percent decrease in GDP is significantly associated with an increase in child poverty by 0.44 percentage points while a one percent increase in per capita GDP is significantly associated with a reduction in child poverty by 0.40 percentage points.

Brazil Overall Poverty

A one percent decrease in GDP is associated with an increase in poverty by 0.113 percentage points.

Empirical Evidence: Health

10

11

Country

Indicator

Results

Mexico Child mortality

For children up to one year of age, a one percent decrease in GDP is significantly associated with an increase of 1.97 female deaths and 2.58 male deaths per thousand live births. When taking account the population at risk this implies 640 additional deaths for this age group for every one percent decrease in GDP.

Total Mortality

A one percent decrease in GDP is significantly associated with an increase of 17.51 female deaths and 18.42 male deaths for every hundred thousand in the population. This implies 2,381 additional deaths per hundred thousand in the population for every one percent decrease in GDP.

Peru Child mortality

A one percent decrease in GDP per capita is associated with an increase in child mortality of 0.27 deaths per thousand live births. This implies an elasticity of mortality for children of -0.39. These effects vary significantly according to mother education levels and do not vary significantly when distinguished by gender.

12

Country

Indicator

Results

Argentina

Maternal mortality

A one percent decrease in per capita GDP is associated with an increase of 0.33 deaths per 10,000 live births.

Child mortality

A one percent decrease in per capita GDP is associated with an increase of 0.41 deaths per 10,000 live births. When growth and contraction periods are differentiated, a one percent decrease in GDP is significantly associated with an increase in child mortality by 1.1 deaths per 10,000 live births while a one percent increase in per capita GDP is not significantly associated with a reduction in child mortality.

Empirical Evidence: Education

13

Country

Indicator

Results

Brazil School delay(7-14)

Boys that attended school during a crisis episode were 9 percentage points more likely to be delayed in school by at least one year than those that attended school before/after crisis. Girls that attended school during a crisis episode were 6.9 percentage points more likely to be delayed in school by at least one year than those that attended school before/after crisis.

School enrollment (7-14)

Boys that were between 7 and 14 years of age during a crisis episode were 11.5 percentage points more likely to be enrolled in school than those in the same age group before/after crisis. Girls that were between 7 and 14 years of age during a crisis episode were 11.4 percentage points more likely to be enrolled in school than those in the same age group before/after crisis.

Jamaica

Primary school attendance

A one percent decrease in GDP per capita growth rate is strongly associated with a 1.45 percentage point increase in primary school attendance, although results are not significant when growth rates are instrumented.

Secondary school attendance

A one percentage point decrease in GDP per capita growth rate is strongly associated with a 1.48 percentage point increase in primary school attendance, although results are not significant when growth rates are instrumented.

14

Summary of Results

15

  Argentina Brazil Jamaica Mexico Peru

POVERTY

Overall Poverty (-) (-)Child Poverty  (-)EDUCATION

Enrollment Primary Ambigous (+)1

Enrollment secondary Ambigous (-)Attendance primary (+) (-)Attendance secondary  (-) (-)School delay primary (-) (-)School delay secondary (-) (-)HEALTH

Child mortality (-) (-) (-)Maternal mortality  (-) (-) (-)2

Total mortality (-)HAZ (+)Illness Rate (-)

1 Using lagged growth rates2 Using prenatal care as proxy

Entries in red are statistically significant.

Final Remarks

16

Economic downturns have significant impacts in long-term households’ well-being.

In all countries studied, economic downturns have a negative impact on child mortality or child health.

For those countries in which the effects of the crises on poverty are examined, there is also strong evidence that recessions are associated with increases in child and overall poverty.

Results on education are ambiguous.

17

Policies to mitigate the effects of crises should move beyond programs to establish a comprehensive social protection system that protects the income of households during economic downturns.

Maintain or expand the provision of public services, mainly education and health

Provide incentives that prevent reversals in asset accumulation in order to reduce the vulnerability of households’ well-being to temporary shocks.

Almudena Fernandez

Luis Felipe Lopez-Calva

London, 9 November 2009

Transitory Shocks, Permanent Effects: Impact of the Economic Crisis on the Well-

Being of Households in Latin America and the Caribbean