ANNEX A. Sources of data on timely indicators of entrepreneurship
This Annex presents the sources used to develop the OECD Timely Indicators of Entrepreneurship database, which contains sub-annual data on enterprise creations and bankruptcies. The database is available on http://stats.oecd.org//Index.aspx?QueryId=72208.
Data on enterprise creations are sourced from a variety of administrative and statistical sources (Table A.1), whose definitions and coverage vary significantly by country, and may differ from the concepts and coverage of the benchmark definitions of births provided by the Eurostat-OECD Manual on Business Demography Statistics. For instance, enterprise creations may include new enterprises created via mergers, break-ups, split-offs as well as re-activations of dormant enterprises, in addition to pure births.
The underlying administrative and statistical data can vary significantly by country, with differences in the population of enterprises covered, such as types of legal form (e.g. corporates only), sectors of activity (e.g. coverage of agriculture or education) or enterprises below a certain turnover or employment threshold. For example, the underlying administrative data for Spain exclude natural persons and sole proprietors, and data for the United Kingdom exclude non-incorporated companies.
The underlying data can also be volatile as the scope enterprises covered may change over time. For example, for the United Kingdom, data from 2009 on also include Northern Ireland; and for Sweden, methodological changes were introduced in 2010. Changes in policies towards particular forms of enterprises (in particular legal status) can also have a considerable impact on the raw data, particularly if the policy favours a change in legal form towards enterprises covered in the raw administrative data away from legal forms not covered (or indeed vice-versa). For example in France, a new individual enterprise status (régime de l’auto-entrepreneur) was implemented in January 2009.
Data on bankruptcies (Table A.2) are sourced from raw administrative sources. Whenever possible the raw data are adapted to ensure that the sectoral coverage reflects the standard used in this publication, i.e. only the business economy is considered. Bankruptcy is based on the legal and institutional frameworks in place. A key difference with the enterprise death measure discussed in this publication is that a ‘bankrupt’ firm may continue to operate.
Data on bankruptcies are affected by differences in national legislation. In some countries a declaration of bankruptcy means that the enterprise must stop trading immediately, and so is more closely aligned with the concept of enterprise death used in this publication. In other countries however enterprises are able to continue trading with receivers in operational control even after a formal declaration of bankruptcy. Indeed, some of those firms declaring themselves bankrupt may eventually recover. The proportion of bankruptcy procedures that end up in actual liquidations (deaths) of the companies, and not in reorganisations, varies across countries depending on the bankruptcy code. Of additional note in relation to comparisons with enterprise deaths is that not all firms file for bankruptcy in advance of closure (death).
Because of these comparability challenges, international comparisons of new creations and bankruptcies data focus on changes in levels rather than levels per se. Trends are computed for both new creations and bankruptcies; specifically, the trend-cycle reflects the combined long-term (trend) and medium-to-long-term (cycle) movements in the original series (see http://stats.oecd.org/glossary/detail.asp?ID=6693).