OECD OCDE International Data on Business Start-ups: Factors Affecting Comparability Steven VALE, OECD / ONS
The Issue Increasing political and academic interest in business demography and entrepreneurship Accurate measures of business start-ups are important indicators for both OECD project on international comparability of start-up rates Funded by the International Consortium for Dynamic Entrepreneurship Benchmarking, led by FORA
Current Position Inventory of 48 Internet data sources from 26 countries Statistical Business Register is the most common source Examination of metadata to try to understand the differences in sources Typology of factors affecting comparability Tested on countries with multiple sources
Factors Affecting Comparability Reasons why data are not comparable Genuine variation Methodological differences Factors provide a framework to help understand methodological differences Start-up rates = New Businesses Population Numerator and denominator factors
Numerator Factors Source – register, census or survey? Units – enterprise, establishment,...? Scope – inclusions / exclusions? Threshold – small units? Time – when to measure? Purity – births or other events?
Denominator Factors Population – businesses or people? Coverage – inclusions / exclusions? Temporal basis – “point in time” or “live during period”?
United States Data – 4 Sources
Reasons for Differences If data are for one country, differences must be entirely methodological Most important factors: Scope (numerator) Coverage (denominator) Source (numerator) Purity (numerator) Temporal basis (denominator)
Conclusions The charts clearly show why start-up rates need to be broken down into their components Each component is affected by a number of factors A framework to systematically assess these factors helps to make better informed and more reliable comparisons
What Next? Short-term: Long-term: Apply framework to data from different countries Standard conversion methods / ratios to adjust for methodological differences? Look for “quick wins” – improving comparability for negligible cost Long-term: What do data users really want? “Gold-standard” methodology as a target to work towards
Thank-you for listening. Any Questions? steven.vale@oecd.org steven.vale@ons.gov.uk