Individual Motives for Choosing Self-employment in the UK: Does Region Matter?

Abstract

Dawson C., Henley A. and Latreille P. Individual motives for choosing self-employment in the UK: does region matter?, Regional Studies. Regional entrepreneurship policy is often framed in terms of spatial shortcomings in entrepreneurial culture. However, differences in why individuals choose self-employment may reflect structure rather than culture. This paper investigates UK data for 1999–2001 on the reported motives for choosing self-employment. After controlling for individual characteristics and industrial structure, some regional differences persist. These are largely for men and are quantitatively small. Northern Ireland stands out, reflecting the different composition of its self-employed. Conclusions for the emphasis of regional policy and further research are discussed.

Dawson C., Henley A. and Latreille P. 在英国选择从事自由职业的个人动机：有关区域吗？区域研究。区域的创业政策经常为因应创业文化在空间上的短处而制定，但个人选择从事自由职业的原因差异可能反映了结构，而非文化问题。本文调研英国自 1999 年至 2001 年间有关选择从事自由职业之动机的报告资料，在控制个人特质与产业结构的变因后，仍然存在一些区域性的差异。这些差异多半存在于男性中，且为数不多。北爱尔兰的案例特别突出，反映出该地自由职业的不同组成。结论部分将以讨论强调区域政策与未来之研究做结。

Dawson C., Henley A. et Latreille P. Les motivations individuelles de choisir le travail indépendant au R-U: la région, importe-t-elle?, Regional Studies. La politique régionale en faveur de l'esprit d'entreprise s'exprime en termes des lacunes géographiques de la culture entrepreneuriale. Cependant, les motivations individuelles de choisir le travail indépendant pourraient refléter la structure plutôt que la culture. Cet article cherche à examiner des données pour le R-U de 1999 à 2001 sur les motivations exprimées de choisir le travail indépendant. Une fois contrôlé des caractéristiques individuelles et de la structure industrielle, quelques écarts persistent. Ces derniers se rapportent en grande partie aux hommes et sont quantitativement petits. Il ressort l'Irlande du Nord, ce qui reflète la structure différente de ses travailleurs indépendants. On discute des conclusions quant à l'orientation de la politique régionale et à la recherche future.

Dawson C., Henley A. und Latreille P. Individuelle Motive für die Entscheidung zu einer freiberuflichen Tätigkeit in Großbritannien: Kommt es auf die Region an?, Regional Studies. Die Politik für regionales Unternehmertum wird oft in einen Rahmen der räumlichen Defizite hinsichtlich der Unternehmenskultur gestellt. Die unterschiedlichen Gründe von Einzelpersonen für die Entscheidung zu einer freiberuflichen Tätigkeit sind jedoch eventuell eher strukturell als kulturell begründet. In diesem Beitrag werden britische Daten des Zeitraums von 1999 bis 2001 über die genannten Motive für die Entscheidung zu einer freiberuflichen Tätigkeit untersucht. Auch bei Berücksichtigung von individuellen Merkmalen und der Branchenstruktur bleiben einige regionale Unterschiede bestehen. Diese Unterschiede gelten in erster Linie für Männer und fallen in quantitativer Hinsicht geringfügig aus. Einen Sonderplatz nimmt Nordirland ein, worin sich die unterschiedliche Zusammensetzung seiner Freiberufler widerspiegelt. Wir erörtern die Schlussfolgerungen für die Schwerpunkte in der Regionalpolitik und weiteren Forschung.

Dawson C., Henley A. y Latreille P. Motivos individuales para elegir el empleo autónomo en el Reino Unido: ¿importa la región?, Regional Studies. La política empresarial regional muchas veces se enmarca en términos de deficiencias espaciales en la cultura empresarial. Sin embargo, los diferentes motivos por los que las personas eligen el empleo autónomo podrían deberse a factores estructurales más que culturales. En este artículo analizamos los datos del Reino Unido para el periodo entre 1999 y 2001 sobre los motivos informados de elegir el empleo autónomo. Después de considerar las características individuales y la estructura industrial, persisten algunas diferencias regionales. Estas diferencias son válidas sobre todo para hombres y son pequeñas desde un punto de vista cuantitativo. Irlanda del Norte destaca al reflejar la composición diferente de sus empleados autónomos. Para terminar, sacamos conclusiones para los enfoques en la política regional y las futuras investigaciones.

• Self-employment
• Self-employment motives
• Regional entrepreneurship policy

• 自由职业
• 自由职业的动机
• 区域创业政策

• Travail indépendant
• Motivations du travail indépendant
• Politique en faveur de l'esprit d'entreprise

• Freie Berufe
• Motive für freiberufliche Tätigkeit
• Regionale Unternehmenspolitik

• Empleo autónomo
• Motivos del empleo autónomo
• Política empresarial regional

JEL classifications:

• J24
• M13
• R58

Acknowledgements

Chris Dawson's contribution to this research was supported through a PhD studentship from the UK Economic and Social Research Council (ESRC). The authors are grateful to three referees and the journal Editors for their valuable comments.

Notes

The choice of the time period 1999–2001 is constrained by the availability of the question on motivation for becoming self-employed. This question has been asked only periodically in the QLFS and was dropped after 2001. This is unfortunate as it would be desirable to see how motives might have changed over a longer period of time, and in particular to assess the impact of recent economic recession on the patterns of motives.

Each member of the QLFS sample is interviewed for five consecutive quarters in order to provide a rotating longitudinal element to the survey. This means that the spring quarter files for 2000 and for 2001 included two observations on those who were self-employed in each year, and therefore a duplicate (although potentially inconsistent) response to the question on reasons for becoming self-employed. To avoid duplicate observations in the analysis, those individuals in the spring quarter 2000 sample who were also included in the spring quarter 1999 sample, and those in the spring quarter 2001 sample who were also included in the spring quarter 2000 sample were deleted from the analysis. In principle, the first rather than the second duplicate observation could have been deleted. Both methods were investigated, and it was found that the results of the secondary analysis in each case were almost identical. Within the QLFS a significant proportion of respondents are by proxy (36.9% of the self-employed). The proportion of the self-employed who respond by proxy is higher than for the employed. In order to maintain sample size proxy responses were retained, but a control variable was included in the regression analysis.

It would be desirable to investigate this correlation using a more disaggregated level of geography. However, this is not possible with the publically released QLFS datasets.

A model for the final option shown in Table 3 (‘Other reasons’) is not reported as the motivations here may be highly heterogeneous and idiosyncratic.

The cross-tabulation of responses where more than one is coded suggests that the ordering does contain information, and it is not the case that interviewers coded those motives which appear higher on the grid as higher in importance. This does mean that a number of second, third and fourth responses are discarded in order to avoid over-weighting in the analysis. Dawson and Henley (2012) investigated in further detail the information that may be contained within these combinations of response. Regression results are available from the authors on request, which include these second, third and fourth motives, as well as results which restrict the sample to those who only report one motive. The broad conclusions from these are very similar to those reported in the paper. This is perhaps not surprising given that 86% of respondents do in fact only provide one motive. One advantage of this method over other approaches, which ask individuals to score a long list of possible motivations, is that irrelevant alternatives are ignored.

One consideration here is that regression errors across the equations for each motive may not be independent. This possibility was investigated using a multivariate probit estimator allowing for non-zero cross-equation covariances. However, results obtained were little changed from those reported. A further consideration is the extent to which the sample of self-employed within the QLFS is itself a non-random sample of the working population. Research on the modelling of the self-employment decision (Le, 1999; Parker, 2009) has addressed this question. In the present context it is not easy to conceive of appropriate identifying instruments which would identify separately the choice to become self-employed from the motivation for that choice. Unpublished work by the authors has attempted to investigate this issue by employing an identification strategy based on regression functional form.

Parker (2009) has provided a comprehensive and succinct summary of arguments and evidence concerning this relationship. For recent British evidence, see Clark and Drinkwater (2000).

A total of 37% of responses was provided by proxy from another household member.

This suggestion is supported by further model estimation in which the sample was split into those who had been self-employed for fewer than four years and those for four or more years – a distinction that roughly corresponds to the Global Entrepreneurship Monitor's definition of early stage entrepreneurial activity (Kelley et al., 2011). Approximately 32% of the sample has been self-employed for fewer than four years. These results are not reported in this paper, but are available from the authors on request. The main findings are that there is little observed difference between the two groups in the regional effects for opportunity motivations. However, other things being equal, there is significantly higher regional variation in family/lifestyle self-employment for the experienced self-employed than for the recently self-employed, with the former significantly less likely to report family/lifestyle motives in more prosperous regions. The experienced self-employed in prosperous regions are also significantly less likely to report necessity motives. These individuals therefore appear more likely, with the passage of time, to justify their choice of self-employment in terms of personal fulfilment.

The reference region for the analysis is the East Midlands, chosen because of its geographical centrality along the Humber–Severn ‘north–south’ axis and because it might be viewed as an example of a well-performing region away from the London and South East England.

Although patterns of employment and experience of employment discrimination between Protestant and Catholic communities in Northern Ireland are very different, rates of self-employment are only slightly higher for Catholics (Northern Ireland Statistics and Research Agency, 2010, table A7.6).

The US Panel Survey of Entrepreneurial Dynamics (Gartner et al., 2004) is the main notable example of an attempt to construct such a data source.

A deeper level of policy localism appears likely in England as the regional development agencies are now replaced by local enterprise partnerships. Some have expressed scepticism that policy ‘localism’ reflects a genuine ‘bottom-up’ attempt to build local capacity in policy instruments; implementation of policy may be devolved, but models of intervention remain laterally integrated (Gibb, 2000). Variation in the scale, if not the type of provision across the UK, has also reflected the use of significant levels of European Structural Funds in peripheral regions to support small and medium-sized enterprise (SME) start-up and development activity. Localism in policy has been seen as a means to address concerns about both the efficiency of delivery and the level of take-up amongst the self-employed and business owners (Curran and Blackburn, 1994).