A Brief History of the CAED, 1995-2006

Contributed by Eric Bartelsman, Mark Doms, and Seppo Laaksonen

The analysis of firm and establishment level micro data has become a common tool in many areas of economics and the technical issues involved continue to keep statisticians busy. Further, top-level theoretical work now takes as given the stylized facts that have emerged from two decades of empirical work from around the globe. These stylized facts include the problematic nature of the assumption of the representative firm and the importance of processes of creative destruction for a well functioning economy. Reading current journals such as Quarterly Journal of Economics, Review of Economics and Statistics, and Journal of Industrial Economics, it is hard to remember the long and difficult route followed by researchers since the first publications based on longitudinal micro data in the late 1980s.

By the early 1990s, a number of national statistical agencies had made some of their proprietary survey data of businesses available to economic researchers. Also during the early 1990s, there was a coincidental increasing interest in relationships between new technologies and economic performance, whether that performance was productivity, wages or growth. Aggregate data was of limited use in examining these relationships because of poor measures of technology coupled with a very limited time series. To better understand how new technologies were affecting the economy, a number of researchers in a wide variety of countries coupled special surveys on technology use with data on firm performance.

In the spring of 1995, the U.S. Department of Commerce in conjunction with the National Academy of Science and the Organization for Economic Cooperation and Development organized a conference highlighting research using these data. Papers were delivered using data from the U.S., Japan, France, Canada, the U.K, Finland, the Netherlands, Denmark, and several others. This conference highlighted the commonalities and the differences across countries in the data available and in research methods. As a result of this conference, the sharing of research methods across countries became commonplace. Additionally, the conference produced a large number of results showing how improved technology affects the economy, usually by increasing productivity and wages.

The first Comparative Analysis of Enterprise (Micro) Data conference was organized by Statistics Finland and held in Helsinki in June 1996. A source of inspiration for bringing together researchers from the academic and statistical community was the conference held in May 1995 at the National Academy of Sciences in Washington DC. Eurostat also conducted an International Workshop on Enterprise Panels held in Luxembourg 21-23 February 1994. The proceedings entitled "Techniques and uses of enterprise panels" was published by Eurostat in 1995 (edited S. Laaksonen and E. Ojo). While the Washington conference was geared toward research on the effects of technology on firm performance, the scope of CAED was wider and included any economic or statistical topics that could be addressed with micro-level data. These conferences formed the beginning of a cycle of conferences held at irregular intervals and organized on an ad-hoc basis by disparate researchers, statisticians and policy makers. To date, there have been six CAED conferences:

Looking back at the program for the conference in Helsinki, the main themes were productivity, firm demography, worker flows, and statistical measurement. These themes continue to be at the core of more recent conferences. While work was presented in Helsinki based on firm-level data for a dozen countries, only a handful of countries had comprehensive, longitudinally, linked enterprise data. Further, even for these countries the datasets generally were limited to the manufacturing sector and rarely spanned the length of a business cycle. Nonetheless, even at this early stage in firm-level research some examples of things to come were on the program. A few presentations were made of research using data on individual wage and worker characteristics linked to firm-level indicators and one paper explored the links between exports and productivity.

Since 1996, the size of the conference has continued to expand, with many of the researchers present in Helsinki becoming continuers and with new practitioners entering the fray. In Bergamo, the first presentations were made of research comparing results from multiple countries. Further, statistical work was presented on the difficulties of international comparisons and on the plans of Eurostat to compile comparable indicators of firm dynamics. At this point, the themes also expanded, with papers exploring the effects financing and corporate governance or of competition on firm performance.

In the 1999 CAED conference, many new countries had micro-level data available for researchers, resulting in presentations of work from Russia, China and Ivory Coast. Further, the range of themes continued to expand, with papers considering such macroeconomic questions as investment of the business cycle, or management issues as the effects of worker training or HRM practices on firm performance. By 2001, the availability of linked employer-employee data in more countries, and the ability to link in firm-level data from separate surveys, allowed much more delving into the firm policies that promote success and the role of hitherto unobserved worker heterogeneity. Also, many of the papers at the conference were able to extend the analysis of previous conferences to firms in business services, retail, and other non-manufacturing areas of the economy.

The previous two conferences in the U.K., in 2003 and 2005, truly show that the field has become mature. The research methods for much work is becoming more standardized, following the iterative process of testing, critique, and improvements made through healthy scientific dialog. Much of the work in these conferences is now being conducted by well established researchers with high quality publications records, or by their students and recent PhDs. And best of all, more and more countries are opening up the treasures buried in their statistical agencies so that the prospect of true "Comparative Analysis" is coming into sight.