Data gaps in international databases

1
Data gaps in international databases
2007 International Conference on Millennium
Development Goals Statistics Manila, 1 3
October 2007

• Francesca Coullare
• United Nations Statistics Division

2
Overview
1

• Global monitoring and the Inter-agency and Expert
  Group on MDGs indicators
• Working modalities,
• Millennium Development Goals Indicators database
  mdgs.un.org

2

• How international agencies adjust country data
  to obtain regional estimates and/or address data
  gaps issues

3

• Current mechanisms/initiatives to improve
  international data series used to monitor
  progress towards the MDGs

3
1

• Global monitoring of MDGs indicators

4
International Monitoring Efforts
1

• The Inter-Agency and Expert Group (IAEG) on MDG
  Indicators (2 meetings per year)
• Coordinated by UN Statistics Division/DESA
• Composed of representatives from
• 25 specialized agencies,
• regional commissions,
• NSOs

• Thematic sub-groups of the IAEG
• Gender
• Employment
• Health
• Poverty and hunger
• Environment
• Slums

5
International Monitoring Efforts

• IAEG is responsible for
• compiling data and undertaking analysis to
  monitor progress towards the MDGs at the global
  and regional levels
• reporting on status of annual progress through
  printed reports, progress charts, CD-roms and
  internet
• reviewing and preparing guidelines on
  methodologies and technical issues related to the
  indicators
• helping define priorities and strategies to
  support countries in data collection, analysis
  and reporting on MDGs.

6
Type of indicator/series Type of indicator/series
Agency MDGs Others Total
FAO 2 1 3
ILO 4 9 13
IPU 1 3 4
ITU 3 3 6
OECD 8 7 12
UNAIDS 1 4 5
UNEP-Ozone 1 1 2
UNEP-WCMC 1 1 2
UNESCO 9 7 16
UNFCCC (CDIAC) 2 2 4
UN-HABITAT 1 1 2
UNICEF 17 8 25
UNPD 2 2 3
WB 7 4 11
WHO 8 1 9
WTO 5 0 5
17 72 54 126

1. compiling data for the global/regional monitoring
   of MDGs

7
(a) Data compilation data flow
International agency country office
Agency Headquarterse.g. UNICEF

• MDG Indicators database
• 48 indicators
• 192 Member States
• 1990-2006
• mdgs.un.org

Agency Headquarterse.g. UNESCO
Line Ministry in country
Agency Headquarterse.g. ILO
National Statistical Office in country
8

• Adjustment of country data by international
  agencies to ensure international comparability
  and address data gaps

2
9
Data gaps for MDG 3 in international databases
Percentage of countries with at least 2 data
points since 1990 (excluding modeled data), by
indicator and MDG region
Ind. 9 Enrolment Ind. 9 Enrolment Ind. 9 Enrolment Ind. 10 Ind. 11 Ind.12
Primary Secondary Tertiary Youth Literacy Employment Parliament
Developing Regions 86 85 68 - 55 81
Northern Africa 83 83 50 - 67 67
Sub-Saharan Africa 94 92 82 - 28 96
Latin America Caribbean 83 83 57 - 76 72
Eastern Asia 83 83 83 - 83 67
Southern Asia 89 89 78 - 67 100
South-eastern Asia 82 82 82 - 73 82
Western Asia 100 100 93 - 80 87
Oceania 65 60 30 - 25 60
Source UNSD-MDGs database, access on June 2007
10
The example of UNESCO

• Indicator 6. Net enrolment ratio in primary
  education

• UNESCO Steps
• An adjustment to account for over- or
  under-reporting
• To include enrolments in private schools and/or
  geographical areas left out
• To exclude pupils of other programmes than
  primary (i.e. adult education)
• An estimate of the number of enrolments in the
  official age group for primary education
• (when only total enrolments in primary education
  is reported, using reliable source for age
  distribution)

11
The example of UNESCO

• Indicator 6. Net enrolment ratio in primary
  education

• UNESCO Steps (cont.)
• A redistribution of enrolments of unknown age
  (across known ages - only if more than 5 of tot.
  enrolments)
• An estimate of the population in the official age
  group for primary education (if neither UNPD nor
  the country itself can provide estimates of their
  own)

Treatment of missing values When missing data
for a variable, use (a) previous years
submissions, (b) other correlated variable or
(c) similar countries (never published-only
used in regional aggregates)
12
The example of ILO-Gender

• Indicator 11. Share of women in wage employment
  in the non-agricultural sector

ILO-Gender Estimated values vs. Predicted values

• Estimations based on auxiliary variables
• Total paid employment
• Total employment in non-agriculture
• Employees
• Total employment
• Economically Active Population in non-agriculture
• Empirical analysis shows that strong correlation
  exits between the indicator and the auxiliary
  variable.

13
The example of ILO-Gender

• Indicator 11. Share of women in wage employment
  in the non-agricultural sector

ILO-Gender Estimated values vs. Predicted values

• Predictions based on statistical models
• Only for producing regional and global aggregates
  
• Separate two-level models developed for each of
  the 5 regions, considering
• between-countries variation over time,
• within-country variation over time.
• Based on the assumption that available data are
  representative of a countrys deviation from the
  average trend in its region, across time .

14
3

• Improving international data series used to
  monitor progress towards the MDGs

15
Strengthening country statistical capacity

• 2006 ECOSOC Resolution
• 2004 Marrakech Action Plan for Statistics
• PARIS21 Partnership in Statistics for
  Development in the 21st Century

• Renewed commitment on the importance of sound
  statistical systems to produce evidence-based
  policies

• Blue print identifying 6 steps for achieving
  better statistics for better monitoring policies
• 1. NSDS national strategy
• 2. Increased budget allocated to Statistics
• 3. 2010 Round of Population and Housing Census
• 4. Better support for Household Surveys - (IHSN)
• 5. Quick and better data in key areas such as
  MDGs (ADP)
• 6. Increased accountability and better
  coordination among international statistical
  partners

• Promoting a culture of Evidence-based decision
  making and implementation

16
(b) Improving mechanisms for data transfer and
consultation with countries

• Within countries among different stakeholders
  producing data in the national statistical system
  
• Between countries and international agencies
• Role of Regional Commissions
• Establishing a central repository of data
• Between international agencies and UNSD
• SDMX initiative
• in pilot in 3 SADC countries
• Work in progress in IAEG on MDGs indicators

17
(c) Enhance transparency in MDGs Global database

• UNSD MDG database to present metadata information
  at the cell level for country-level estimates
• Showing data source, reference period, ,
  pointing out possible discrepancies between
  international and national figures

18
(c) Enhance transparency in MDGs Global database
Revised metadata for MDG Indicators in the IAEG
MDG Database
CONTACT POINT in international agency
DEFINITION
METHODS OF COMPUTATION
COMMENTS AND LIMITATIONS
SOURCES OF DISCREPANCIES BETWEEN GLOBAL AND NATIONAL FIGURES
PROCESS OF OBTAINING DATA
TREATMENT OF MISSING VALUES
DATA AVAILABILITY
REGIONAL AND GLOBAL ESTIMATES
EXPECTED TIME OF RELEASE

• UNSD MDG database to present more detailed
  indicator-level metadata
• Explaining in details methodology used to
  calculate indicators and presenting contact
  details for users to contact to obtain additional
  information

19
http//mdgs.un.org