8
Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Embed Size (px)

Citation preview

Page 1: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Impact of hidden methodological differences

NESIS-workshop in Rome26-27.6.2003Mikael ÅkerblomStatistics Finland

Page 2: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Contents

International efforts to harmonise data

’Hidden differences Effects of globalisation Linguistic bias Institutional factors Different data sources and ways of collecting data

What can we do?

Page 3: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

International efforts to harmonise data

EU legislation Other standards

United Nations OECD and other international organisations

Gentlemen agreements Benchmarking exercises Ad hoc data collection

Page 4: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Effects of globalisation Role of multinationals has strongly increased during the last

years (financial sector, telecom, pulp&paper, car manufacturing)

Decision making, strategic planning, cost accounting, R&D and innovation on a group or division of group level

Therefore basic statistical units on the national level, like enterprises and establishments are less relevant to describe activities of global enterprises

Statistics dependent on how multinationals are able to break down the figures between for them sometimes irrelevent units. Intergroup often non-monetary transactions may distort national aggregates

One or few multinationals may completely dominate the picture for especially smaller countries, which makes problems of interpretation

Page 5: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Cultural and linguistic bias

Attitudes towards surveys differ Translation of concepts and definitions might cause inconsistencies

ExamplesR&DTechnological innovationTechnicianScience

Page 6: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Institutional factors

Organisations of enterprises influence all industrial breakdowns

Combination of enterprises into groupsSplitting up of enterprises into smaller establisment units

Educational systems influence comparability of education indicators

Degree structures are differentOrganisation of units (university systems differ between countries)

Page 7: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

Different sources and ways of collecting data

Different sources; registers and other administrative data, surveys (censuses, large or small scale samples), estimations

Different survey methods and routines like face to face or phone interviews, mail surveys

Different registers over various kind of units

Page 8: Impact of hidden methodological differences NESIS-workshop in Rome 26-27.6.2003 Mikael Åkerblom Statistics Finland

What can we do

Need to develop and implement indicators on globalisation

Pay attention to careful translation of definitions and concepts to avoid misunderstandings

The abovementioned reasons for hidden differences should be more explicitly incorporated in metadata