Modelling the Effects of Public Support to Small Firms in the UK - Paradise Gained?



Modelling the Effects of Public Support to Small Firms in the UK - paradise Gained?

In general terms the estimated probit equations are similar to others of this type with
similar levels of correct predictions. The other main points arising from the estimation
exercise are:

No significant differences were identified between industries in the probability
of receiving BL assistance. The industry dummy variables were therefore
excluded from the final models

Similarly, no significant differences were evident between the geographical
clusters (rural, industrial city and London/Birmingham) in terms of the
probability of receiving BL support. The cluster dummies were also, therefore,
excluded from the estimation.

Firm characteristics had only weak effects on the probability of receiving BL
support. Partnerships were marginally more likely to receive support than
other types of firms as were older firms. Small firms were more likely to
receive support than larger businesses although again this effect was
statistically insignificant.

There was also no evidence that BL assistance was being targeted at firms
which had grown faster than average over the 1994-96 period. Both turnover
growth and employment growth were insignificant (and negative) over this
period.

Informational variables proved more significant. In particular, a ‘gossip effect’
from knowing other firms which had received BL support and a ‘mail-out’
effect when a firm received direct mail from BL both significantly increased
the probability that a firm would have had BL support. Other information
channels seemed largely insignificant.

Having the founder of a business still involved (and whether they had
significant equity in the business) proved unimportant in terms of the
probability of receiving BL support. There was evidence, however, that if the
owner-manager was willing to share power, had only school level
qualifications and experience of working in medium-sized firms this increased
the probability of receiving BL support.

Graduate owner-managers were also more likely to receive assistance than
other less well qualified groups as were those in their twenties and forties.
Neither of these effects was statistically significant, however.

Perhaps the key result here is that the probit models suggest no evidence that BL
support was targeted at firms that grew faster in terms of employment or sales in the
1994-96 period. In other words any attempt at 'targeting' assistance at faster growing
firms over this period was largely ineffective. Instead, the probability of support was
determined primarily by the willingness of the owner-manager of the business to
share power and whether or not they became aware of the potential benefits of BL
support. Particularly important in terms of awareness were the gossip effect and direct
mailings from BL.

Stephen Roper and Mark Hart

16




More intriguing information

1. The name is absent
2. The name is absent
3. Graphical Data Representation in Bankruptcy Analysis
4. Novelty and Reinforcement Learning in the Value System of Developmental Robots
5. The Composition of Government Spending and the Real Exchange Rate
6. The name is absent
7. 09-01 "Resources, Rules and International Political Economy: The Politics of Development in the WTO"
8. National curriculum assessment: how to make it better
9. DURABLE CONSUMPTION AS A STATUS GOOD: A STUDY OF NEOCLASSICAL CASES
10. PROFITABILITY OF ALFALFA HAY STORAGE USING PROBABILITIES: AN EXTENSION APPROACH