Social Science Microdata in India

1
Social Science Microdata in India

George Smith, Sony Pellissery, Sweta Rajan
Sylvie Dubuc University of Oxford
2
Outline

• Context
• Some key organisations
• Methods and limitations
• Types of Data
• Key Data by Functional Areas
• Demographic data
• Health
• Education
• Crime
• Election studies
• Conclusions and Observations
• The Report

3
Context India

• Population 1.1 billion
• Population density 324 per sq km
• 32 aged 0-14 (median age 24.4)
• Life expectancy 64 years (Males 63, Females 65)
• Sex ratio 933 females per 1000 males (ranging
  from 821/1000 in Delhi to 1058/1000 in Kerala)
• Literacy rate of 64.8 (females 53.7) ranges
  from 47 in Bihar, to 90.9 in Kerala (2001
  Census).
• Federal structure Central union government, 29
  states and 6 union territories 603 local
  districts
• National statistics key part of central
  information and planning

(Source S.Raghavendra (2004) Map 2004 GIS
Conference)
4
Indian Data Management Organisations
Central Statistical Organisation (CSO)
National Statistical Commission (NSC)
National Sample Survey Organisation (NSSO)
Statistical Units (Various Ministries)
National
Other Govt Institutions Census, Reserve Bank
India, Election Commission
Other National
Academic institutions generating national
surveys
Multilateral Agencies supporting studies and
holding data
Private agencies collating data from different
sources data on request
State / District level Economics and Statistical
units
State/District level NSSO
State Level units of Statistics wing of
Ministries
State / District
5
Methods and Limitations

• Phase 1 discovering datasets currently in active
  use by UK academics in India, and Indian research
  contacts. Both micro and macro.
• Phase 2 Two sets of visits to national
  organisations in Delhi and Mumbai, and state
  organisations in Mumbai.
• Phase 3 Review of material, including datasets
  collected
• Limitations
• selective coverage
• national rather than state level
• limited contact with university or other research
  organisations or the private sector
• originally aimed at UK based researchers
• A Rough Guide!

6
Type of Data and Key Organisations 1

• Indian censuses
• Decennial population census
• Economic census
• Agricultural and livestock census
• Survey Data
• The National Sample Survey Organisation (60
  rounds of national sample survey since 1950)
• Other survey organisations
• National Council of Applied Economic Research
  (NCAER)
• Indian Institute of Populations Studies (IIPS)
• Centre for the Study of Developing Societies
  (CSDS)

7
Type of Data and Key Organisations 2

• Administrative Data pyramid principle - issues
  of reliability, consistency from one area to
  another and accuracy
• District information system for education (DISE)
• Corrected and consistent time series (SL Shetty
  at EPW Foundation)
• Boundary data, geocoding and the use of remote
  sensing, GIS systems etc
• Digitised boundaries widely available
• Movement towards a unified geocoding system
• National Remote Sensing Agency

8
(No Transcript)
9
Key Datasets by Function Demographic data
10
Key Datasets by Functional Area Health Data
11
Key Datasets by Functional Area Education
12
Female literacy in India (2001)
13
Key Datasets by Functional Area Crime
Key Datasets by Functional Area Election Data
14
Some Key Indian Microdatasets
15
Conclusions

• India has a long and distinguished history in the
  collection of social and economic statistics
  (first full census, 1881)
• Increasing moves since 1999 from the highest
  level to making data accessible and available
  (Right to Information Act, 2005)
• Almost all documentation of national level data
  is in English
• Good quality microdata, based on cross-sectional
  surveys (single and regularly repeated series)
  widely available
• Limited panel and longitudinal data sets
• Best data is of excellent all round quality some
  issues about relatively conservative strategies
  used in sampling, classification, documentation
  in some studies.
• Large scale administrative datasets, built on the
  pyramid system, but issues of quality and
  variability
• Some nodal data access points (eg the Computer
  Centre/NSSO, the Reserve Bank of India, Indian
  Council of Social Science Research, NCERT/NIEPA
  in education, NCAER for economic development,
  CSDS for election studies but no single data
  archive

16
Some Observations

• Given the easily accessible and high quality
  data, apparently limited use is made of some of
  this data for more complex analyses groups such
  Indicus Analytics, are developing this area.
• Advantage of central access to data through
  single data archive or gateway?
• Possibility of a sample of anonymised records
  from the census?
• More panel and longitudinal surveys?
• Victim or crime surveys to complement recorded
  crime data?
• Improvements in quality of administrative data,
  along the lines pioneered by DISE
• 40 Years On!

17
The ESRC Report Indian Microdata Scoping Study

• Methods
• India Data Organisations
• Key Data Sets by Data Type
• Indian censuses
• Survey Data
• Administrative Data
• Boundaries, geocoding and remote sensing data
• Data Sets by Functional Areas
• Demographic data
• Economic data
• Housing and Access to Services
• Health and Family Welfare
• Education
• Infrastructure and Transport
• Crime
• Election studies
• Data Quality and Access

18
Contact details

• Website http//www.apsoc.ox.ac.uk
• Department of Social Policy and Social Work
• Barnett House
• 32 Wellington Square
• OXFORD OX1 2ER
• UK
• George Smith George.smith_at_socres.ox.ac.uk