The methodological framework of the Sectoral e-Business Watch builds upon the methodology established in 2002. However, while a common methodology was applied to all sector studies conducted until 2006, the methodology was revised in 2007 in order to support the shift from monitoring "e-readiness" and "e-activity" to the evidence-based analysis of "e-impact". In particular, econometric modelling was introduced to study ICT impacts on value added growth, productivity, employment and energy consumption.
Furthermore, studies conducted since 2007 are more flexible (and therefore less homogeneous) in their choice of the methodological approach. The focus of the research activities is no longer on benchmarking the degree of e-business adoption across different sectors by conducting large-scale surveys (primary data collection). By contrast, since 2007, each study has its own, specific research agenda; the approach has to be adapted accordingly, and data collection is often conducted for the specific sector or issue only.
Thus, main characteristics of the recent studies are the extended use of analytical statistical methods and of external data sources for studying ICT impact, for example multiple regression models and dynamic econometric models.
The following statistical methods have been used for the economic analyses of ICT impact in most of the studies since 2007.
Growth accounting. The growth accounting framework, based on regression models, was pioneered by Solow (1957). This methodology was applied in sector studies since 2007, using data from the EU KLEMS Productivity and Growth Accounts database. Growth accounting can be used to asses the contribution of ICT capital to economic growth and to estimate the development of total factor productivity (TFP). In a growth accounting framework, the growth rate of real output is equal to the weighted growth rates of labour input and real capital input, plus growth in TFP. As to capital input, a distinction can be made between the contribution of ICT capital and of other, non-ICT capital to output. This enables an analysis of the ICT contribution to productivity growth. The method has been used to assess productivity change in a number of industrial sectors.
Dynamic econometric models. Some studies find that in the context of ICT and economic transformation the concept of time plays an important role, i.e. the returns to ICT investments usually occur with a significant time lag. Thus, instead of constructing econometric models in static terms, it is worthwhile to formulate a model that incorporates past values (lags) of explanatory variables in the estimation.
Panel data models. Panel data models are considered to be the most efficient analytical method in handling econometric data. It allows the inclusion of a number of cross-sections and time periods. The combined panel matrix set consists of a time series for each cross-sectional member in the data set, and offers a variety of estimation methods. Examples of different methods of panel data estimations include the common constant method, the fixed effects method and the random effects method, among others. The advantage of using panel methods has already been proven in a number of empirical studies analysing the impact of ICT.
Multiple regression models. ICT diffusion is not the only factor influencing firm performance and the ongoing economic transformation. Thus, whenever assessing the drivers of innovation patterns or changes in employment, not only ICT-related variables should be included in the econometric model. Multiple regression models account for various factors affecting the dependent variable and assess the strength of their impact and were used, for example, in the cross-sector study on the economic impact of ICT (2008).
Studies are based on a mix of primary data collection and secondary sources. The main primary data sources for many of the studies are:
The main secondary data sources are:
Most of the data about ICT and e-business adoption presented in Sectoral e-Business Watch studies are based on representative decision-maker surveys in European enterprises, conducted with CATI method (computer-assisted telephone interviews). Researchers can request to receive survey data (case level) for their own research activities.
The first e-Business Survey was carried out in June 2002 and included more than 9,000 firms from 15 sectors, with interviews in all 15 EU Member States of 2002.
The second e-Business Survey was carried out in March and November 2003, and covered more than 10,000 firms from 10 sectors, with interviews in 25 EU Member States (those of May 2004) and Norway.
The third e-Business Survey took place in February 2005, with a scope of more than 5200 interviews with companies from 10 sectors and 7 countries.
The fourth e-Business Survey was conducted in spring 2006 and covered about 14,000 enterprises of 10 sectors from 25 EU Member States and EEA and Candidate Countries.
In 2007, four e-Business Surveys were conducted, each with a different questionnaire adapted to the specific sectors and topics:
For 2009, two sector surveys are planned: (i)in the glass, cement and ceramic industry and (ii) in the energy supply industry.
All these surveys are based on CATI (computer assisted telephone interviews), with average interview lengths of 15-20 minutes per interview (8-10 minutes for the special topic surveys on RFID and IP).
Methodology reports about the surveys and the questionnaires used can be downloaded (see table below). More detailed background information about the surveys (e.g. response rates), if required, can be obtained upon request from the Sectoral e-Business Watch (info at ebusiness-watch dot com ).
|Surveys 2007||Methodoloy Report - Sectors
|Methodology Report - RFID and Intellectual Property Rights||310KB|
|Questionnaire - Manufacturing||513KB|
|Questionnaire - Ratail / Transport & Logistics||504KB|
|Questionnaire - RFID||468KB|
|Questionnaire - Intellectual Property Rights||494KB|
|Survey 2006||Methodology Report||114KB|
|Survey 2005||Methodology Report||278KB|
|Survey 2003||Methodology Report||338KB|
|Questionnaire – Part I (Mar. 2003)||187KB|
|Questionnaire – Part II (Nov. 2003)||181KB|
|Survey 2002||Methodology Report||262KB|