International and National Data Differences Alison Kennedy Chief of Statistics Human Development Rep

1
International and National Data
DifferencesAlison KennedyChief of Statistics
Human Development Report Office, New York
HDRO/RBA Regional Technical Workshop on Measuring
Human Development Nairobi September - 2007
2
Purpose of international data

• Preferred for cross-national comparisons because
  standardised
• Agreed definitions
• Same coverage (temporal, spatial, sectoral)
• Standard reporting frameworks (classifications)
• Independent quality assurance

3
National data

• Preferred for in-country analyses
• More detailed
• Reflect national specificities
• Greater levels of disaggregation
• More timely
• More pertinent

4
Sources

• Most international data come originally from
  national sources and are collected by national
  authorities (eg NSOs, line Ministries)
• Raw data (not indicators) are reported by various
  national bodies to numerous multilateral and
  international agencies

5
Adjustments

• International agencies
• May correct errors in nationally reported data
• May estimate or impute for entirely missing
  national data or for partially complete data
• Calculate indicators combining data from
  differnet sources (eg GER uses enrolments from
  UNESCO and population from UNPD GDP per capita
  uses GDP from World Bank and, usually, population
  from UNPD)

6
Adjustments

• National authorities
• Also may make adjustments
• Need to transform national data to meet
  international definitions and to fit
  international reporting frameworks
• Many need to adjust the coverage of national data
  before reporting them to international agencies
• Thus making national data less recognisable

7
Data sources for HDI

• International data
• Life expectancy UN Population Division
• Adult literacy UIS
• Gross enrolment UIS ( UNPD)
• GDP per capita World Bank (UNPD)

8
Data sources for HDI

• National data
• Life expectancy NSO (Censuses, births,
  deaths)
• Adult literacy NSO (household surveys,
  censuses)
• Gross enrolment Ministries of Ed, NSO
• GDP per capita Ministry of Finance, NSO

9
Life expectancy

• NSOs report Census data either to UN Statistical
  Division or to UN Regional Commissions (eg
  ECALAC)
• NSOs may report births, deaths etc to WHO
• UNICEF/WHO may gather birth, death rates from
  household surveys (MICS, DHS)

10
Life expectancy

• UN Population Division, WHO, UNICEF, World Bank
  agree on estimates of infant and child mortality
  rates (which feed into UNPDs models)
• UN Population Division/UNAIDS (and others?) agree
  on HIV/AIDS prevalence, etc
• UNPD revises its estimates and projections (for
  1950-2050) every two years

11
Combined gross enrolment ratio

• Enrolments (of all ages) in primary, secondary
  and tertiary education
• Population of the official age for education in
  the given country (eg 6-22 6-10 (p), 11-17 (s),
  18-22 (t))

12
Combined gross enrolment ratio

• Enrolments
• Usually administrative data (pupil counts
  reported by schools and universities to
  Ministries of Education or NSOs)
• Occasionally, enrolment (or attendance) ratios
  can be estimated from household surveys
  (collected by NSOs)

13
Combined gross enrolment ratio

• Population data
• Derived from Censuses of Population (collected by
  NSOs) with inter-censal projections/estimates
• School-age range determined by the official
  structure of education system (defined typically
  by Ministries of Education)

14
Combined gross enrolment ratio

• Enrolments
• Ministry of Education (or NSO) reports enrolments
  to UNESCO
• Ministry of Education may need to gather data
  from other Ministries first
• Population
• NSOs report to UNSD or UN Regional Commissions
  (as for life expectancy)

15
Combined gross enrolment ratio

• UNESCO
• Checks nationally reported data
• May make adjustments to ensure data are reported
  according to standard framework (ISCED)
• May make estimates for missing or incomplete data
• Combines the enrolment data with UNPD population
  data to calculate the GER

16
Differences with national data

• Causes
• Standardisation
• Timing
• Imputation/estimation (by international agencies)
• Coverage
• Definitions
• Errors (inc compounded errors when 2 or more
  sources combined)

17
Example

• National GER is greater than the international
  one
• National GER is for primary education only
• National GER is for schools covered by Min of Ed
  only
• National GER uses same enrolments as
  international GER but different population
• National GER is for a more recent year
• International GER is incorrect

18
What to do?

• Contact HDRO (or UNDP CO)
• Explain the problem and ask for explanation
• HDRO can
• Ask international data provider for explanation
• Put country in direct touch with data provider
• Act as go-between
• Virtually (by video conference, Email, telephone)
• Face-to-face (in New York with video link to
  Montreal and Washington)

19
Other data sources

• HDRO does use alternative data sources
• HDRO occasionally makes estimates
• GER (Bhutan, Ecuador, Haiti, Turkmenistan)
• PPP exchange rates (Palestine)
• Almost never uses national data
• Cuba (GDP per capita)
• Timor-Leste (literacy, GER)
• BiH, Luxembourg, Singapore (GER)