Experiences in measuring living conditions and poverty from household - - PowerPoint PPT Presentation

Bosna i Hercegovina Agencija za statistiku Bosne i Hercegovine Bosna i Hercegovina Agencija za statistiku Bosne i Hercegovine

Experiences in measuring living conditions and poverty from household surveys in Bosnia and Herzegovina

Expert Workshop Research uses of high–precision census samples, Barcelona, Spain, 25‐27 June 2015 Edin Šabanović, The Agency for Statistics of Bosnia and Herzegovina

CONTENT

• Background
• Poverty analysis based on LSMS‐Living Standard Measurement

Survey

• Poverty analysis based on HBS‐Household Budget Survey
• New activities on poverty analysis: EU‐SILC introduction in

B&H

• Future work

2

BACKGROUND

• After the war in B&H, there was a lack of household and

population data

• Population data from the Census 1991 become out of date

because of significant demographic changes and population migrations during and after the war

• About 1 mil. people have left the country, thousands have

been displaced within the country, a lot of villages were devastated

• Register of spatial units has become completely unusable

3

POVERTY ANALYSIS BASED ON LSMS (1)

• First household survey was LSMS‐Living Standard Measurement

Survey 2001

• First sampling frame created by enumerating all households and

population in the sample of 25 municipalities (14 in Federation BiH and 11 in Republika Srpska entity)

• Brčko was excluded because of political problems
• LSMS net sample of size 5402 HHs was representative at state,

entity and urban/rural level (3002 HHS in FBiH, 2400 in RS)

4

POVERTY ANALYSIS BASED ON LSMS (2)

• Absolute poverty measure was applied on consumpion data
• LSMS continued as a panel survey called „Living in BiH“ in

2002‐2004 (n=3000 HHs)

• Full scale LSMS repeated only in 2004

5

POVERTY ANALYSIS BASED ON HBS(1)

• HBS was introduced in 2004 and repeated in 2007 and 2011
• HBS 2015 in process
• Three Master samples were created in order to establish

sampling frame for household based surveys

• MS 2003 (750 EAs; 40,000 HHs)
• MS 2006 (1500 EAs; 80,000 HHs)
• MS 2009 (1500 EAs; 80,000 HHs). It is still in use

6

POVERTY ANALYSIS BASED ON HBS(2)

• All samples selected from MS were representative at

country, entity and urban/rural level

• Net sample size of about 7500 HHs
• NUTS classification not defined and used as a temporary

version

• There is a need for surveys which are respresentative at

lower level of administrative organisation of the country (cantons/regions, municipalities, ethnic groups)

7

POVERTY ANALYSIS BASED ON HBS(3)

• Regional disparity studies made by UNDP
• Data at the level of 17 cantons/regions and 142 municipality
• Limited number of variables available
• Statistical offices do not produce such kind of data because
• f lack of population census data
• Population census was conducted in October 2013 but

results are still not available

8

NEW ACTIVITIES ON POVERTY ANALYSIS: EU‐ SILC INTRODUCTION IN BIH

• Pilot SILC conducted in 2015 within IPA 2012 programme
• Income approach to measure the poverty
• Better harmonisation to EU standards

9

FUTURE WORK

• Full scale BiH SILC as regular annual survey from 2017 (?)
• HBS BiH as regular survey in 5‐years periodicity
• Sample for household surveys selected from census sampling

frame

• NUTS classification finalized and implemented
• Possibility for data production at lower level of the

administrative/regional organisation of the country

10

Thank you for your attention!

Edin Šabanović edin.sabanovic@bhas.ba