1. Introduction

The specific aims of the present NACE coder project are technical in nature. We aim to produce a high-quality NACE concordance allowing the user to make maximum use of the available NACE information when consulting our database, and giving a high degree of transparency about the comparability of UK and German data cut according to different NACE revisions. We would also like to hand database extracts of harmonized UK and German NACE-specific data to BERR and other researchers or policy bodies, and there the computer-literate user may wish to find out NACE codes for particular data items. This suggests the provision of a NACE code lookup facility and various directories of NACE-related data items, which will be substantially aided by use of the coder.

The EU itself is a “natural” unit for comparative research because of the common policy environment and the high degree of economic interdependence. Hence, we are interested in preparing an EU-wide comparison of firm-level data in the longer term. The main project now planned will focus only on the UK and Germany, but efforts will be made to ensure that as far as possible the data protocols and methods used are extendable to other EU countries. To this end, we would be interested in establishing contact with similar projects using microdata public use data from establishment or firm-level surveys in other countries. Establishing affiliations with Eurostat and with EU research networks and initiatives would also be a long-term goal.

The Lisbon theme has resurfaced recently in the UK under the heading of “better regulation,” encompassing a range of policy initiatives designed to enhance the competitiveness and productivity of UK business. Germany has faced similar problems on competitiveness and productivity, and indeed the UK policy initiatives in this area have at times been greeted in Germany as unwelcome instances of an “Anglo-Saxon” attack on the Rhenish model of capitalism! This is therefore a propitious moment for undertaking a comparison of the UK and Germany in terms of the structure and performance of business, including the effectiveness of policies. The availability of a common policy initiative at the EU level is often an occasion for policy “spillovers” and mutual learning among EU member countries. For the UK, this was seen most starkly in the implementation of competition law following the Single European Act, which forced a review and liberalization of UK laws and policies on competition in many areas. A good understanding of the structural differences between UK and German business might help to avoid misunderstandings and policy conflicts that can arise when countries attribute observed differences in firm behavior and performance to cultural stereotypes, rather than to rational responses to differences in market incentives and institutional constraints.

1.1 Purpose of the Study The purpose of this study is to provide the Department for Business, Enterprise and Regulatory Reform (BERR) with a detailed assessment of the scope and compatibility of the data available in the UK and Germany for microdata comparisons of firm-level performance. The UK has a tradition of policy evaluation using secondary data on establishment and firm-level performance from a variety of sources, but the increasing availability of comparable data in electronic form from administrative sources has raised methodological and technical possibilities for more ambitious work in this area. At the same time, the advent of the EU Lisbon agenda and especially the renewed self-conscious emphasis on making the EU “the most competitive economy in the world” has increased the interest in policy evaluation and cross-national learning both within member countries and at the level of EU comparisons.

1.1. Purpose of the Study

This study aims to develop a framework for comparing enterprise size and a model of size data collection and analysis. This will be done using UK and German micro firm data. The terms of reference ask that specific consideration be given to MODINPSI’s research project on integrated firm and establishment size data. Some external analysis of UK and German NACE concordance and differences will also be considered. The practical motivation for this project comes from problems faced in previous and ongoing research as discussed with Jonathan Haskel. In some recent work with Stephen Roper, Alex Kats and Ciaran Mac an Bhaird of Warwick and Aston Universities we have encountered difficulties when comparing UK and German firm sizes using public domain UK datasets and the Mannheim ZEW innovation survey micro file. This work with Stephen Roper is part of an ongoing research collaboration between NIESR and the Enterprise Research Centre at the University of Warwick. In previous joint research, Jonathan has also encountered difficulties when trying to compare UK and German firm and establishment sizes which were not fully resolved. In a related matter, Jonathan’s discussions with Ciaran Mac an Bhaird about employing the micro file for a comparison of family firm and non-family firm innovation activity revealed that the German researchers were encountering some problems with the manner of defining and identifying German micro firms in the ZEW survey dataset hardcore. These experiences suggest that more detailed knowledge of the span and structure of enterprise size in the two countries is needed. Previous NIESR and ZEW joint research has highlighted the potential usefulness of harmonising UK/German micro data sources and sharing research experiences. With this in mind, a steering committee has been formed to facilitate collaboration and knowledge sharing in enterprise research between the Enterprise Research Centre and ZEW. This project has already been discussed in principle with Michael Anyadike-Danes of ERC and Dr. Muhamed Kudic and hopes to contribute to a greater overall understanding of the UK and German enterprise system, which the NACE concordance project is also part of. With these various interests in mind, the proposal has been adapted to make the comparison of UK and German enterprise size the central focus of a micro data expertise project. A working paper series to disseminate findings to other researchers is intended, and potential exists for future paper presentation at seminars with German research partners. This groundwork will open up in terms of further contact and collaboration with ZEW and those in the ERC.

1.2. Research Questions

What are the research questions which will be addressed in this report? Using a variety of qualitative and quantitative techniques, we will provide a detailed analysis of the size and turnover of firms in the UK and Germany. This has been stimulated by recent and fundamental changes to European Union (EU) competition and industrial policies. These policies have been driven by the Lisbon European Council in 2000, which stated that: “the overall aim of creating a more competitive and dynamic knowledge-based economy, capable of sustainable economic growth with more and better jobs and greater social cohesion” (European Commission, 2003). This policy and the broader changes within the world economy towards more globalized and service/knowledge-based industries have large implications for the makeup and competitiveness of different industries and markets. We are concerned with all firms in the UK and Germany, with a firm defined as an individual unit of production at a single locational site. The first research question is “How big are firms in the UK and Germany, and how does the size distribution vary between industries and over time?”

1.3. Scope and Limitations

There are aspects of the NACE coding method that may lead to measurement error in the treatment variable. There is evidence to suggest that corporate groups can strategically split/divide subsidiary firms to take advantage of more favourable state aids and this could be in the form of re-defining firm size to meet certain criteria. This would lead to a differentiation of firms with the same characteristics if one has received state aids and the other has not. The methods that state aids affect firm behaviour or how they are obtained have never been perfectly observed and third sector distortive state aids are also difficult to quantify. High quality data on NACE codes of firms involved in state aids is not publicly available so we do not know which firms have actually received state aids. As a result, there is no direct measure of the treatment variable and we need to find an IV.

We are only able to compare the two different cross-sectional data sets from UK and Germany fiscal organisations. The primary reason for this is because the calculator readily available for measurement error in an IV-style system for endogenous treatment effect cases is only applicable to cross-sectional data. The methods used to identify small and medium sized enterprises serve to provide a robustness check of results since previous studies have not found consistent results when comparing to large firms. A similar check will be performed by comparing the turnover results for size categorisation, since it is expected that larger firms will have much higher turnover than the cut-off points used to define size categories.

2. Literature Review

2.1. Definition of NACE Coder

2.2. Importance of NACE Coder in Business Analysis

2.3. Previous Studies on NACE Coder and Firm Size/Turnover Analysis

3. Methodology

3.1. Data Collection

3.2. Selection of UK and German Firms

3.3. Application of NACE Coder

4. Analysis of UK Firms’ Size and Turnover

4.1. Overview of UK Firms’ Size Distribution

4.2. Analysis of UK Firms’ Turnover Distribution

4.3. Comparison of UK Firms’ Size and Turnover by NACE Code

5. Analysis of German Firms’ Size and Turnover

5.1. Overview of German Firms’ Size Distribution

5.2. Analysis of German Firms’ Turnover Distribution

5.3. Comparison of German Firms’ Size and Turnover by NACE Code

6. Comparison of UK and German Firms’ Size and Turnover

6.1. Overall Comparison of Size and Turnover

6.2. Comparison by NACE Code

6.3. Factors Influencing Differences in Size and Turnover

7. Conclusion

7.1. Summary of Findings

7.2. Implications for Business Analysis

7.3. Recommendations for Future Research