آیا عوامل مختلف نرخ خود اشتغالی دختر و پسر را توضیح می دهید؟

This article challenges the assumption that the factors associated with the self-employment choices of women differ from those of men; specifically, we test the extent to which women are influenced by standard economic factors compared with family and social issues. We find that economic factors influence the self-employment choices made by men and by women in the long and short-run. Although some findings were sensitive to the chosen self-employment measure our short-run findings, in particular, are at variance with the interpretation that self-employed women are less likely to be influenced by economic factors than their male counterparts. Consequently, we argue that gender-based explanations have exaggerated the importance of social factors in the self-employment choices made by women.

During the last four decades women's formal labour force participation has grown substantially (Cloin et al., 2011, DiCecio et al., 2008 and Gutierrez-Domenech and Bell, 2004). Reflecting this trend, more women are also pursuing careers in self-employment, but men continue to be more than twice as likely to be self-employed (Díaz García and Welter, 2011 and OECD, 2009). In addition, prevailing evidence suggests men and women prioritise differing factors within self-employment; so, for the former, self-report data suggests economic factors such as the state of the economy and access to credit are more influential (DeMartino and Barbato, 2003 and Kepler and Shane, 2007). In contrast, for women it is assumed that social factors such as work–life balance (Shelton, 2006), parenthood (Georgellis and Wall, 2005), childcare concerns (Kirkwood and Tootell, 2008) and esteem issues (Taylor and Newcomer, 2005) have a greater impact. Thus, reflecting gender stereotypical assumptions, women are considered to prioritise socially-related issues when considering self-employment and, are less concerned with economic factors than their male counterparts. The arguments and evidence we present in this article do not support this characterisation; rather, they suggest the gendered differences between reported attitudes to, and actual participation in, self-employment is not that simplistic or clear-cut. Specifically, we challenge the presumption that when entering self-employment, women are more likely to value and prioritise social rather than economic factors. We acknowledge that our findings are limited to one country, the United Kingdom (UK) and to one measure of enterprise — self-employment. Nevertheless, this study differs from the extant research as it does not rely upon self-report data, the key limitations of which are the difficulty of verifying the answers given and of generalising from the small, and often unrepresentative, samples from which they are derived. We also avoid the use of cross-sectional frameworks (Georgellis et al., 2005) and instead undertake time-series modelling over a time span of more than 30 years using official data. Unlike cross sectional studies, this approach allows us to analyse long-run patterns and dynamics associated with self-employment decisions by gender in relation to economic variables whilst at the same time, it mitigates statistical issues related to non-stationarity and endogeneity (see the discussion by Parker, 2004, pp. 29–30). Our novel contribution lies within an examination of factors influencing self-employment rates amongst men and women over both the short and the long-run. Linking long and short-run social and economic changes with self-employment rates leads us to dispute the presence of clear gendered differences. Consequently, we challenge assumptions that women's self-employment choices reflect a gendered bias towards social factors. To support these arguments, we examine the role of alternative employment options, personal circumstances, macro-economic conditions and government policies on self-employment. We develop the earlier work of Parker (1996) and Cowling and Mitchell (1997) in five ways. First, we extend the time period by more than a decade; second, we include new variables to capture social, as well as economic influences; third, we estimate separate models for men and women to allow for gender differences in the determinants of self-employment; fourth, we use the vector autoregressive (VAR) methodology to capture immediate (short-run) and permanent (long-run) changes and deal with statistical issues related to non-stationarity and co-integration. Finally, we provide a robustness check to examine the sensitivity of the results to different definitions of the self-employment rate. Drawing from this analysis, we find that in the short-run, there appear to be few gender-related differences in terms of the economic factors influencing self-employment. In the long-run, women continue to be strongly influenced by economic considerations, but additionally they are more strongly influenced by marital factors and family-related obligations. We therefore, question the simplistic notion that the enterprise decisions of women are primarily influenced by social and particularly, family-based motives whereas those of men are more driven by economic factors. To explore and illustrate these arguments, the discussion is organised as follows: Section 3 explores how the choice between waged work and self-employment differs according to gender; from this analysis, two hypotheses are derived. The first is that although both males and females are influenced in their self-employment choice by marital and family-related obligations, the association is stronger for women. The second hypothesis is that over time, economic factors are of less importance for women in explaining changes in self-employment. Section 4 describes the data whilst Section 5 presents the statistical model. Section 6 provides the results, distinguishing between short and long-run effects, with the final section exploring the implications of the results and highlighting issues for future development.

7.1. Key findings and implications This article has explored the factors influencing self employment rates amongst men and women in the UK. This discussion was contextualised through an exploration of women's increasing economic participation in waged employment within developed economies over four decades. However, this propensity to enter waged employment has not been matched by enterprise creation rates where, in Europe as a whole, men are almost twice as likely to be self-employed as women (Díaz García and Welter, 2011 and OECD, 2009). The growing literature exploring the impact of gender upon entrepreneurial ambitions and behaviours frequently ascribes such differences to stereotypical gendered assumptions articulated within the context of self-employment (Greene et al., 2011). In effect, it is argued that, given their association with the domestic sphere and related responsibilities, there is a gendered imperative for women to prioritise social motives. Accordingly, it is argued that women are more likely to begin, and then manage, their enterprises to reflect social priorities and, in addition, for those with caring demands, that such responsibilities can be flexibly accommodated (Duberley and Carrigan, 2012 and Wellington, 2006). Generic assumptions persist that women place greater weight upon non-economic factors such as family considerations or esteem when making a self-employment choice. In contrast, men are argued to be more influenced by financial/economic considerations. To test this, we take time-series data for the UK over four decades and utilise advanced macro-econometric methods to deal with issues related to nonstationarity and co-integration. We also distinguish clearly between both short-run and long-run effects of social and economic factors on self-employment by gender, whilst avoiding the problems associated with qualitative analysis such as the recall bias noted by Fraser et al. (2007) and measurement error. Finally, we use two metrics to capture self-employment rates, one normalised by population and the other by labour force. Our conclusion is clear: there are strong statistical associations between changing economic factors and both the short and long-run changes in female self-employment. Accordingly, it is valid to infer that women are significantly influenced in their self-employment choices by economic considerations, such as the state of the economy and access to finance. Indeed, GDP and house prices appear to have a more powerful long-run impact upon women than men. A second critical outcome of this discussion is that social, as well as economic, factors influence self-employment. However, it is incorrect to infer that these social factors exclusively, or even primarily, influence women's choices. Instead, we find that social factors influence both men and women in a broadly similar fashion. The two key exceptions are that (1) unemployment increases self-employment rates — but only for men, and (2) in the long run, divorce lowers self-employment — but only for women. Our third notable result is that conditions amongst wage earners exert a strong influence upon self-employment choices; improved conditions amongst wage earners – whether male or female – reduce self-employment rates. We suspect that, post-1997, the efforts of UK Labour governments to improve working conditions for low income wage workers may, in part, explain the fall in self-employment rates during its term of office — but this effect is observed only for males. Consequently, we make the case that time-series data are more appropriate for describing long-run patterns and dynamics between self-employment rate and macroeconomic variables than survey data and micro-based approaches. Specifically, by observing and analysing a sequence of data ordered in time, rather than at one point, we can observe how the variables move, identify structural changes and patterns of co-movement and importantly, study the dynamics associated with the self-employment decision. Illustrating this argument, our findings are clearly incompatible with the view that the self-employment decisions made by women are not influenced by standard economic factors. In effect, our results expose the problem of building theory upon what people say they do (self-report data) rather than what they actually do over time according to more encompassing official data. 7.2. Limitations and directions for future research Our study is limited by its focus upon one economy — the UK. Further time-series work, on other countries, is important to investigate whether similar or different patterns emerge. Of particular interest would be the results for countries which initially had low levels of self-employment, but where public programmes were put in place to address gender imbalances. In addition, we strongly suspect a different picture would emerge, particularly for women, if a distinction were made between full and part-time activity. However, the UK data is too short a time-series for this currently to be satisfactorily undertaken. The implementation of recently developed (micro or macro) panel data techniques may also provide further insights, especially concerning variables with high local variation (e.g. demographic variables) and models that control for the effect of omitted variable bias and allow for the inclusion of a wider range of covariates. A number of co-integration statistics for panel data have also been proposed (Kao, 1999, Pedroni, 1999 and Pedroni, 2004) to study long-term economic relations, and there is now a growing empirical literature regarding this area (for review, see Georgellis et al., 2005). In short, the statistical techniques for such work are now available. Less tractable however, is the availability of time-series datasets that capture the slippery concepts of family, social context and human capital (e.g. education) which we would like to have included in this study. For example proxies for changes in the overall ‘social dynamics of the household’, reflecting the point in the life course of the business owner when the self-employment choice is exercised would be most helpful (Davis and Shaver, 2012). Another deficiency is our inability to examine the role played by the state system of financial support for both men and women and how this influences self-employment choices. The plea of researchers for more and better data to address omitted variable bias is familiar but, even when such data are available they frequently fail to support often well-established theorising. Of some surprise, at least to us, was the absence in our work of any statistical relationship with immigration, taxation or the age structure of the population. Nevertheless, what remains clear is that there is considerable interest in the influence of gender upon propensity towards, and experiences of, self-employment. Regarding our contribution to this debate, our data and analysis present robust analytical insights which challenge persistent assumptions concerning gender stereotypes and self-employment. Our contention that within the household, women are as, if not more, influenced than men by economic considerations — is at variance with those who primarily examine social and family-related factors. We do not dismiss the influence of the prevailing socio-economic context; clearly differentiated gendered effects will shape the self-employment choices/experiences of men and women given life-course events, sectoral channelling and unequal caring/domestic responsibilities. However, our evidence does not imply that women essentially prioritise the social sphere. In effect, within developed economies, given the educational attainments of contemporary women, pressures towards economic participation for all social actors, falling fertility rates and that the ideal of the ‘male breadwinner’ is a thing of the past (Vogler, 2005) it would actually be surprising if self-employed women were not as motivated towards economic factors as their male counterparts.