A model for accessing international firm-level data

1
A model for accessing international firm-level
data

• Eric J. Bartelsman
• Vrije Universiteit Amsterdam
• and Tinbergen Institute
• Prepared for OECD/Eurostat Conference
• Luxembourg October 26, 2006
• This work is partially funded by the European
  Commission, Research Directorate General as part
  of the 6th Framework Programme, Priority 8,
  "Policy Support and Anticipating Scientific and
  Technological Needs".

2
Overview of Presentation

• What International comparisons of micro-based
  indicators
• Why Meets urgent policy needs
• How New statistical regulations remote
  execution remote access
• Who Networked researchers, NSOs, Trusted third
  parties
• Discussion pros/cons costs/benefits

3
Using Microdata for Analysis

• Policy analysis
• SNA / Final Expenditures GDP demand management
  policy
• Indicators for structural policy and policy
  evaluation microdata
• Academic research
• Estimation of behavioral responses
• Microdata at NSOs
• Improve quality of output
• Reduce response burden
• Provide facilities for outsiders

4
International Comparisons

• Not possible to stack data from all countries
• Cross-country variation in intl microdata
  research provides
• valuable lessons for policy makers
• identification of effects for academics

5
Available Data Sources
Longitudinal Micro Data
National Accounts Industry Data
Macro and Sectoral Timeseries
Surveys, Business Registers
Single country

• SC LMD

EUKLEMS
N.A.
Multiple countries
6
Recent Intl Microdata Research

• There is demand for international comparisons of
  micro-based indicators
• Firm-level projects for OECD, WB, IADB, Eurostat
• Intl Wage flexibility project (FRB/ECB)
• IPUMS
• Luxembourg Income Studies

7
Firm-Level Projects

• Distributed micro-data analysis
• Harmonized collection of indicators from
  longitudinal micro-level business datasets
• Firm Demographics Entry/Exit, Survival
• Productivity higher moments, conditional
  moments, special tabulations (by size,
  ownership status, etc)

8
Firms produce not countries or industries

• Variation in firm-level productivity within
  industry or country
• A country could have a long tail problem
• Or a lack of world class firms

country1
long tail

• Mean productivity may not be a sufficient policy
  indicator

country2
Global frontier
9

10
(No Transcript)
11
Productivity and market contestability
12
(No Transcript)
13
How to generate indicators?

• Eurostat regulation
• After international debate on definitions, NSOs
  must supply the requested indicators
• Distributed micro-data analysis
• Networked collection, through remote execution
  (or remote access)

14
Distributed micro data research
Policy Question Research Design
Researcher
Program Code
Publication
Metadata
Cross-country Tables
Network
Network members
Provision of metadata. Approval of
access. Disclosure analysis
NSOs
15
Network for International Microdata

• Remote execution using meta-data at center, and
  network of NSO contacts
• Coordination issues
• Secure remote access to confidential data at
  trusted center
• Technical issues
• Legal issues

16
Participants in Network

• Policy analysts and academics
• Answered research questions
• Spillovers from knowledge launching customer
• NSOs
• Meet user needs fits within organizational
  goals learn from best practice improve
  reputation
• Provide facility Provide expertise and
  experience
• OECD
• Improved comparability of stats answered policy
  questions
• Provide institutional support contribute to
  analytical capabilities

17
Issues for Discussion

• What are dangers to NSOs and how to minimize
• Confidentiality
• Provider of undisputed data
• Costly sink of resources
• What are benefits to community and how to
  maximize
• Low marginal costs for research output
• Turnaround time lowered
• Learning from broad-based experience