An analysis of self-employment subsidies with duration models

Abstract

The determinants of continuity of subsidized self-employment activities are studied by means of duration models in Asturias (Spain). Variables related to socio-economic characteristics of the self-employed and to subsidized businesses are included in the models. Moreover, to consider the economic situation, the regional unemployment rate has been taken into account. A general survival analysis is complemented with a competing risk model that considers two types of exits from self-employment (failure versus other reasons). The significant variables that explain survival are age, industry, whether the business is the main source of family income or not and the unemployment rate.

Notes

¹ Blau, 1987; Bögenhold and Staber, 1991; Meager 1992a, 1992b, 1993; Robson, 1998and Staber and Bogenhold (1993).

² Blanchflower and Meyer, 1994; Evans and Leighton, 1989; Georgellis and Wall, 2000; Kuhn and Schuetze, 2001; Moore and Mueller, 2002; Taylor, 1999and Uusitalo, 2001.

³ Changes in this direction pushed up survival rates in five points in UK and eleven points in Australia (Wilson and Adams, 1994).

⁴ With respect to employability, O’Leary (1999) obtains that self-employment programmes increase the probability of getting a non-subsidized job by 14–29%.

⁵ Since 1992 this option is limited to unemployed people who join cooperatives.

⁶ Boletín Oficial del Principado de Asturias (09/04/96; 29/04/97; 19/03/98; 12/04/99; 08/10/99 and 21/03/00).

⁷ Grouping the sample by frequencies under the National Classification of Economic Activities (Clasificación Nacional de Actividades Económicas) businesses have been divided in eight branches: Industry, Construction, Retail trade, Restaurants, Transport, Hairdresser, Business activities and Other services.

⁸ Occupation was codified in five categories according to the National Classification of Occupations (Clasificación Nacional de Ocupaciones): Manager, Professional, Technician, Workers in Services and Non-skilled workers.

⁹ Taylor (1999) finds that the majority of self-employment spells finished with a transition into paid employment (48% of men and 36% of women, 1979 data). Johansson (2000) estimates exits to paid employment at about 39% with data referred to the period 1988–1993.

¹⁰ In the previous section we divided the analysis by sex, but in this case some groups are small so grouping the entire sample seems more appropriate.