One of the drawbacks of the programme was its cost (£4,500 per start up course training place). Since 1988 the volume of subsidy reduced as national unemployment declined (for a period) and government questioned the comparatively high cost per participant. In 1991, with the demise of the Training Agency, GEP as a national model ceased. In England and Wales it was transferred to the DTI, where the awareness raising campaign continued for a short period, some providers continued the start up training with some support from the TECs. In Scotland, ten LECs funded places on the 1991 start up course. Scottish Enterprise and LECs supported the 1992 awareness raising campaign, where participants were directed to local start up training. Since then the training provision for graduates has become fragmented. In Scot-land a GEP Providers manual “Graduates into Business” (Smith and Middleton, 1993) was produced which has been used in some regions to support graduates locally. In 1995 phase I awareness events and workshops were relaunched nationally as part of the Scottish Enterprise Business Birth Rate Strategy (Scottish Enterprise, 1993a) to increase the number of young people setting up in business and of academic spin outs. However this was part of a wider initiative to promote entrepre-neurship education in the universities.

benefit. Surveys have shown (Brown, 1992) that cumulative profits of GEP businesses have exceeded costs of the programme by the third year of business. Each £1 of training investment has been backed by £2 of debt and equity investment. For a full discussion of the programme see Rosa and Smith (1994).

The problem with these national evalua-tions and other such studies (Westhead, 1998) is their concern with hard-core economic indicators and the lack of “micro” evaluation such as assessment of development of skills and competencies.

Aim of study

This paper presents the findings of a research study to review the outcomes of the partici-pants of the Scottish GEP business start up courses. Given the concern over Scotland’s low business birth rate (Scottish Enterprise, 1993b) where young people and women are identified as under represented in business ownership, the aim of the study is to:

There has been an increase in the number of graduates entering self-employment since the introduction of GEP. From 1983 to 1990 University First Destination Statistics show the annual number has almost doubled (Brown, 1992). Unfortunately nearly all this increase has been accounted for by the fact student numbers have increased proportion-ately over the same period. This is reflected by the fact there has only been marginal increases in the rate of self-employed gradu-ates (at between 1.1 per cent and 1.5 per cent it is still lower than the USA or Japan). Any marginal increase could also be accounted for by other factors, such as increased govern-ment support for new businesses and increas-ing unemployment.

Nevertheless, research at Cranfield (on behalf of the Employment Department) has shown that GEP has been a positive factor in this process by encouraging trainees to at least accelerate and bring forward the timing of their plans to run a business (Brown and Burnett, 1990). Where participants in a work-shop had not been selected for the start up course, the majority accepted the decision had been correct and nearly all had received some

The study was carried out during 1994, following on previous research as part of longitudinal research on the programme participants (Rosa and Fletcher, 1997). First, a profile of total GEP participants was carried out through desk research. The process of tracing and updating current activity records was carried out by telephone and mail shot to last known, or parents address.

Second, those traced were then asked to take part in the survey by completing a detailed questionnaire, either by attending a social evening or by post. The questionnaire was piloted in advance by ten participants. The questionnaire method was chosen to enable all participants to be included in the research, interviews would have been too time consuming. It avoided interviewer bias