Census Micro Data

1

• Census Micro Data
• A Discussion
• Prof. Ben Kiregyera
• Director
• African Centre for Statistics
• August 2007

2

• COVERAGE
• Value of data and information
• Issues arising from presentations
• Conclusion

COVERAGE
3
I. VALUE OF DATA AND INFORMATION

• data and information now universally
• recognized as

• part of the enabling environment for
• development
• a priority for results-based agenda
• (PRSPs, MDGs, etc) (e.g. Marrakech
• Roundtable, 2004)

• data and information have no intrinsic value
  
• they are not gold or silver. They have
• extrinsic value which lies in their power
  to
• inform processes e.g. policy debate
• design, planning, monitoring, etc.
• their value lies in fact that

• they can reach those who need them,
• can be easily understood
• are usable and actually used

4
LIS-based research has catalyzed changes in
national policies (LIS paper)
5
DATA CYCLE
Planning
Stage 1
Stage 2
Implementation
Dissemination
Feedback
Stage 3
Reporting
Processing
Analysis/Interpretation
6

• More and more data
• Need information from this
• these data

7
Policy-related information
8
Data, Information, Knowledge, Actions
Informed Actions/ decisions
Knowledge
Value addition
Information
Data
9

• Who should do data analysis?
• Preliminary or general analysis
• statisticians/data producers (NSO)
• Definitive or in-depth or detailed analysis
• subject-matter specialists
• researchers
• Importance of involving subject-matter
  specialists and
• researchers in analysis
• enriches analysis by adding subject-matter
• perspectives
• possibility to link policy variables to
  micro-level
• outcomes (paper on LIS)
• enhances collaboration (advancing from
• coordination to collaboration)
• spreads ownership of statistical products
• feedback and advocacy

10
Basic principle of data analysis
Once
Data Collection
Feedback
Many times
Data Analysis

• NSO
• Policy analysts
• Researchers
• Students
• etc.

11
II. ISSUES ARISING FROM PRESENTATIONS

• All presentations make a case for enhancing data
  management including
• Country level we see great need for
• data archiving (IPUMS-International)
• creation of user-friendly and accessible
• databases (documentation crucial -
  metadata)
• Benins Jupiter
• ensuring that databases are not empty boxes or
  
• have garbage
• encouraging researchers to do more detailed
  data
• analysis (Benin example shows need for
  this)
• a lot of value in doing broad range of
  analysis
• (through space and time)

12

• ISSUES ARISING FROM PRESENTATIONS (ctd)
• Challenges
• i) appreciation of value of data (by data
  producers and users)
• Users use data for evidence-based policy
  (debate, design)
• decision-making invest more to build
  statistical capacity
• development
• Producers ensure data quality
  relevance better data
• management
• inconsistencies through time caused by changes
  in
• definitions(1992 2002 Benin censuses)
• - special attention to harmonization to
  permit comparability
• technical capacity (Benin paper mutually
  beneficial TA)
• Statistics Acts which are not congenial
• Other laws (e.g. designating areas as rural or
  urban)

13

• International level we see increasingly
• construction of cross-national micro databases
  to
• enhance research infrastructure
• Integrated European Census Microdata (IECM)
• Luxembourg Income Study (LIS)
• Redatam Software for micro data dissemination
  
• analysis
• IPUMS-International initiative archiving,
• integration and disseminate high-density
  census
• micro data samples
• defining features of these databases include
• research centres not only involved in data
• analysis but also in coordination,
  dissemination
• and harmonization activities (IECM)
• interest and cooperation of NSOs
• emphasis on integration documentation
• - IECM major part of project
• - IPUMS International
  (post-harmonization)

14
Creating databases become a driver for data
harmonization and integration
15

• Challenges
• reliability and comparability of data sources
  across
• countries
• lack of consistency, terminology,
  classifications,
• question numbers and wording (all papers)
• confidentiality (strict conditions set by
  countries more)
• ownership (IPUMS - International)
• dissemination
• Redatam developed by CELADE, Popn. Division
• of ECLAS (UN) allows on-line processing of
  any
• database over Internet
• LIS uses remote-access system (primary means
• to access data)
• IPUMS posts census documents on Internet, etc

16

• Other issues
• improved analytical capacity at NSOs
• training institutions should strike a
  productive
• balance between data demand and data
• supply in training programmes
• bigger dose of data analysis in training
  centres

17
III. CONCLUSION

• Case made by all papers about need to add
  value to data
• through more detailed analysis
• Papers urge NSOs to move away from risk
  avoidance to
• risk management (European
  statisticians)
• Enhance research infrastructure by creating
  user-friendly
• databases including micro databases
• Issues to address include integration,
  harmonization,
• confidentiality

18
Thank You