4. Entrepreneurship in European Regions

See for instance the most recent Dutch government treaty (Dutch Coalition Agreement 2007) and the German “Ich-AG” (BMWi 2005).

According to Pasinetti (1993), an economy that does not increase the variety of industries over time will suffer from structural unemployment and will ultimately stagnate. In this view, the development of new industries in an economy is required to absorb labor that has become redundant in pre-existing industries. This labor has become redundant due to a combination of productivity increases and demand saturation in pre-existing industries, characterizing the product lifecycle dynamics in each sector.

Even when controlled for recent macroeconomic growth and time lags of the effect on economic growth (see Thurik et al. 2008).

In what Audretsch and Fritsch (2002) call “revolving door” regimes: inefficient entrants, which exit soon after entry will not make a valuable contribution to the economy.

The studies of Audretsch and Keilbach find no (2005), or only very weak (Audretsch and Keilbach, 2004) associations of new firm formation in general- and labor-productivity growth. Only specific forms of entrepreneurship, like new firm formation in high-tech or ICT industries (i.e., technology start-ups) have strong positive associations with labor productivity growth.

The Global Entrepreneurship Monitor makes a distinction between “necessity entrepreneurship,” which is having to become an entrepreneur (often “self-employed”) because you have no better option, and “opportunity entrepreneurship,” which is an active choice to start a new enterprise based on the perception that an unexploited or underexploited business opportunity exists. Analyzing data in 11 countries, Acs and Varga (2005) found that effects on economic growth and development of necessity and opportunity entrepreneurship vary greatly: necessity entrepreneurship has no effect on economic development while opportunity entrepreneurship has a positive and significant effect. They also found that the ratio of opportunity to necessity entrepreneurship in a country is positively related to GDP per capita.

See Reynolds et al. (2005) for a detailed description of the GEM methodology.

NUTS stands for Nomenclature of Territorial Units for Statistics. The Eurostat introduced the standard European NUTS classification. In this selection we have indices for 125 regions corresponding to the classification used by ESRI. This classification comprises of NUTS1 levels for Belgium, France, Germany, Greece, Ireland, the Netherlands and the United Kingdom. NUTS 2 levels are applied for Croatia, Denmark, Finland, Hungary, Norway, Portugal, Slovenia and Sweden an2d a combination of NUTS1 and NUTS 2 for Italy, Spain and Switzerland.

The abstracted regions are Antwerp and Ghent (Belgium); Aarhus (Denmark); Helsinki (Finland); Duisburg-Essen, Düsseldorf, Köln, Rhein-Main, Stuttgart and Munich (Germany); Budapest (Hungary); Dublin (Ireland); Amsterdam, Rotterdam, The Hague and Utrecht (Netherlands); Barcelona, Valencia, Seville and Malaga (Spain).

This issue is not relevant for our empirical analysis since it is based on the individual-level observations constituting the regional aggregates shown in Figs. 4.2–4.5.

This region includes the Sophia-Antipolis cluster.

In line with Davidsson (1991) one could wonder why individual level perceptions of ability and opportunities to start firms are not included in our analyses. Indeed, Arenius and Minniti (2005) find a strong relationship between individuals’ perceptions to entrepreneurship and their involvement in nascent entrepreneurship. However, we feel that the data poses methodological restrictions to do so, since perceived ability, opportunities, and fear of failure are posed directly after questions on involvement in entrepreneurial activity. One would not expect many people involved in early-stage entrepreneurial activity to say that they do not have the skills to start a business or that they do not see opportunities to start a business. Let alone that people already making actual preparation to start a firm will answer that fear of failure would prevent them from starting a business.

The general idea of multilevel analysis is that individuals in the same social context show similar progressive behavior. The most researched cases are within educational studies on school performances: students learn by individual and class influences (Raudenbusch and Bruyk 2002).

We apply Stata’s gllamm procedure (see Rabe-Hesketh and Skrondal 2005), using the logit link from the binomial family.