Measuring Remittances using Multi-Purpose Household Surveys: the LSMS experience Gero Carletto Development Research Group The World Bank (with Claudia Rodriguez)
Feb 07, 2016
Measuring Remittances using
Multi-Purpose Household Surveys:
the LSMS experience
Gero CarlettoDevelopment Research Group
The World Bank
(with Claudia Rodriguez)
What is an LSMS?
DISTINCTIVE FEATURESMulti-Topic QuestionnaireMultiple InstrumentsConsumption-based welfare measure
Quality ControlDemand-driven
Main objectives of LSMS Surveys
Measuring and Monitoring Welfare
Poverty and policy analysisCapacity buildingMethodological validation (LSMS
IV)Consumption, subjective poverty,
migration, …Income (remittances)
Review: main methodologies
Transfer/remittances moduleHousehold (total transfer; by
recipient) Individual (remitter)
Other incomeHH levelClosed vs. open
Individual migrant’s rostersCombination
Individual level transfers (Nepal 2002/03)
Other Income, HH level (Vietnam 2004)
Individual migrant’s rosters
GHANA LSS 2005/06, ALBANIA LSMS 2005, BULGARIA MTHS 2007 Current household members
Present (past migration) Absent
Former Household members Other individuals
No. of surveys
With remittance information68
Unable to differentiate bet/w domestic and int'l remittances
19
No remittance information4
Total91
Most survey contain some remittance info …
Large heterogeneity in methods …
Method No. %
Transfers/ Remittances (Indiv. level) 34 50.0
Transfers/ Remittances (HH level) 8 11.8
Individual level (migrant roster) 4 5.9
Other income 19 27.9
Other (social protection) 3 4.4
TOTAL 68 100.0
Some differences in reference periods
No. of surveys %
Last 12 months 49 72.1
Last year 3 4.4
Last 14 months 1 1.5
Last 6 months 3 4.4
Last 3 months 1 1.5
Last 30 days 1 1.5
Multiple 5 7.4
Flexible 5 7.4
Total 68 100.0
Country X, 2004
Country X, 2004 (cont’d)
Country Y, 1997
Any improvements over time?
Albania 2002-2005 Bulgaria 1995-2007 BiH 2001-2004 Ghana 1987-2005 Nicaragua 1993-2001 Vietnam 1993-2004
Country Methodology% HHs w/
remittances
Vietnam 1992Individual level remittances 5.0
Vietnam 1997Individual level remittances 6.0
Vietnam 2004Other income (HH level) 5.0
BoP vs. survey estimates
Figure 1. Remittances According to National Accounts and Household Surveys (Millions of US$)
0
500
1,000
1,500
2,000
ALBANIA BULGARIA NICARAGUA JAMAICA GUATEMALA NEPAL
National Accounts Household Survey
Bulgaria MTHS 2007 (Sample: 4,300 HHs)
Current HH
members(1)
Former HH members
(2)
Other income module
(3)
(2)+(3)
HH with migrants 205* 458
HH with migrants sending remittances 64 223 138 334
Average remittances received by HH (yearly) ??? U$2,186 US$779 US$1,993
Conclusions
Most surveys contain info on remittances but … Little consistency in methodology Often different components and groups of
individuals included Ambiguities and double counting “Large” underestimation? Lack of benchmark!
Some progress, but … Limited to few countries: demand-driven?? Shortage of best practices
Conclusions (cont’d)
Reference period: 12 months (?) Importance of placement (?) Individual vs. household-level HH membership Distinguish transfers from abroad Multiple modules Temporary migrants and returnees: info (last 12
months) on country of work, length of spell abroad, occupation, wages
De-coupling remittances from migration???