4 DEFINITIONS AND STATISTICS
4.1 This section discusses the statistics used to identify women's participation in science, and identifies the patterns of under-representation of women in science, engineering and technology in Scotland and the UK.
Definitions of Science, Engineering and Technology ( SET)
4.2 When evaluating the statistics and policies on women in science it is important to understand what subjects have been included within definitions of SET.
4.3 Several classifications have been used by different groups (Glover and Fielding, 1999, Scottish Executive, 2007). These definitions agree on the inclusion of some subjects, such as engineering, physical science and biological sciences, within SET. However, other subjects are more difficult to categorise. Whether various different subjects allied to medicine, in particular nursing, are included within SET is a key point of difference between different sources (Glover and Fielding, 1999, UKRC 2009, DTI, 2003, 2006).
4.4 These different subjects have significantly differing gender balances: for example, engineering is strongly male dominated whereas nursing is strongly female dominated. This gendered split in subjects studied in higher education in also evident in Scotland, and this pattern has remained the same over the past ten years. In 2008-9, women made up 82.4% of those studying allied medicine, 65.6% of those studying biological sciences, 43.1% of those studying physical sciences and 14.5% of those studying engineering ( HESA, 1999-2009). In 1999-2000, female participation in the same series of subjects was 83.1%, 64.4%, 37.7% and 16.3% respectively ( HESA, 1999-2009). These different subjects are also of vastly differing numerical sizes (Scottish Funding Council, 2006). The combination of subjects included within a definition of SET can therefore significantly affect the apparent gender balance in SET as a whole.
4.5 Given that there were a range of classification systems available, choices had to be made about the ones that would be used in the project. It was decided that existing statistics, with their varying classification systems, would be used where appropriate. However, where original statistics were sought the STEM subject classification list produced by the UKRC ( UKRC, 2009) would be used in order to allow direct comparability with their UK wide statistics. The full UKRC subject classification is included on p.p. 47-8. The various different statistics on proportions of women in science may not show exactly the same patterns due to differences in definitions.
Cross national statistical patterns
4.6 The under-representation of women in science has been highlighted by many authors over a long time period and across different geographies. The key patterns and problems have been identified in many nations ( UNU- IAD, 2005, Bebbington, 2001). These patterns are explored in detail in European Commission research, which shows that women are in the minority in science across the EU (European Commission, 2008. It should, however, be highlighted that not all men with a SET degree progress on to employment within SET (Scottish Government, 2010, Fielding and Glover, 1999). Nevertheless, the attrition rates for men are generally significantly lower than those for women. Annual Population Survey data does not have sufficiently high numbers to allow us to examine the employment sectors to which women STEM graduates in Scotland progress later in life ( ONS, April 2009 - March 2010).
4.7 Different indicators and measures have been used to measure the participation of women in SET. The data needed to assess gender segregation has not always been routinely collected (Cronin and Roger, 1999). The proportion of students studying a particular subject who are female is commonly used. Nevertheless, changes in the proportion of women studying a subject could also be influenced by a decreasing number of men studying a subject as well as an increased number of women. This would indicate a decline of interest among men for certain subjects rather than an increase in interest among women (Cronin and Roger, 1999). The analysis in this report is limited by the types of data available and most of the data used concerns proportions of women in particular subjects or fields of employment.
4.8 Various patterns of under-representation can be identified. One key pattern is horizontal segregation, where women are concentrated in particular scientific subjects (Department of Trade and Industry, 2006). Another pattern is vertical segregation, where women are under-represented in more senior levels of organisations (Glover, 2000, UKRC 2010). These patterns have been a key focus of recent literature on this topic (Larios et. al., 2009).
4.9 These patterns have remained fairly constant over the past twenty years. There has been long term improvement in rates of female involvement in sciences from the 1960s onwards (Glover and Fielding, 1999). However, this improvement has not always been consistent or significant. From 1984 to 1994, women taking science degrees as a proportion of all women taking degrees remained stable, even though the proportion that was female of those taking science subjects increased. This shows that at least some of the increase in the proportion of women taking certain subjects was due to a decline in male uptake of these subjects (Glover and Fielding, 1999). In Scotland, higher education statistics reveal that gendered patterns remain ( HESA, 1999-2009).
4.10 Between 1971 and 1991, the gap between men and women's post- SET degree employment in SET has narrowed but not closed (Bebbington, 2001). The increase in representation of women in SET has been very slow compared to changes in law and medicine (Glover, 2000). Due to small samples, Higher Education Statistics Authority and Annual Population Survey data cannot provide us with data regarding the longer term destinations of graduates in Scotland ( HESA, 1999-2009 & ONS, April 2009 - March 2010).
4.11 "Time lags" have been suggested as one cause of vertical segregation. One American study identified that the proportion of physics professors who were female was in line with the recruitment of women into physics in the age groups from which most professors were drawn (Committee on Science, Engineering and Public Policy, 2007). However, they also acknowledged that this did not fully explain all vertical segregation, and in any case did not explain ongoing causes of segregation (Blackwell, in Glover, 2000).
Current situation in the UK
4.12 In 2009, girls made up 42.4% of GCE A-Level students in STEM subjects. This conceals significant differences between subjects, as women are 7.5% of all computer science students and 21.4% of all physics students ( UKRC, 2009, p1).
Women made up 33.5% of all higher education students in SET disciplines. There is considerable horizontal segregation, with women making up 15.9% of technology students and 21.4% of computer science students ( UKRC, 2009, p1).
Women represented 18.5% of SET employees ( UKRC, 2009, p1).
Women hold 9.0% of Directorships in the UKFTSE 100 companies in SET sectors ( UKRC, 2009, p1).
8.0% of all SET professors are female. Women make up less than 5% of the professorship in many subject areas including physics, mathematics, and various subsets of engineering ( UKRC, 2009, p1).
UKRC estimated that at any point in time 50,000 female SET graduates are economically inactive ( UKRC, 2010). Of those who re-entered the workplace after a career break, only 8,000 went back into jobs that utilised their qualifications and expertise.
Projecting these figures based on past trends in England and Wales, female and male students will be equally represented at A Level in 2058, in SET disciplines in higher education in 2069, and female SET employment will not reach parity within the 21 st century ( UKRC, 2009, p4).
The current situation in Scotland
4.13 In 1995-6, women accounted for 32% of university students in SET disciplines (Cronin and Roger, 1999). This rose to 37% by 2002-3, and 39% in 2008-9 ( HESA, 1999-2009). There was considerable variation between disciplines in 1995-6, ranging from 14% in engineering and technology to 62% in biological sciences (Cronin and Roger, 1999). This pattern was consistent across the next decade, so that in 2008-9, 65.6% of those studying biological sciences were female compared to 14.5% studying engineering ( HESA, 1999-2009).
4.14 These patterns were persistent, with three broad groups of subjects, namely engineering, physical sciences and mathematical sciences, having 60-90% male students. Biological sciences and subjects allied to medicine showed a predominance of women, with women making up over 65% and 85% of students respectively (Scottish Funding Council, 2006).
4.15 In Scotland there are 209,200 people who hold a STEM degree. Of these, 71% are male: this is higher than the overall figures for graduates in Scotland, where 50% are male ( ONS, April 2009-March 2010).
4.16 Data from the Annual Population Survey for April 2009 - March 2010 gives information about the employment of STEM graduates in Scotland. The employment rate for STEM degree holders is 83.8%, which is higher than for people in Scotland with no degree (71.2%), but lower than for those with a degree in any discipline (85.5%). Within the STEM group the employment rate for women is lower (80.2%) than for men (85.3%) ( ONS, April 2009-March 2010). This means that women with STEM degrees are less likely to be in employment than men with STEM degrees ( ONS, April 2009-March 2010). Lack of data from Higher Education Statistics and the Annual Population Survey means that we cannot determine graduates' career destinations and whether STEM graduates progress onto STEM related employment ( ONS, April 2009-March, 2010).
4.17 According to the Scottish Resource Centre (2010), despite this difference in gendered participation rates, only 29% of female SET graduates in Scotland are working in the sector in which they are qualified, compared to 52% of male graduates (Scottish Resource Centre for Women in SET, 2010, p6).
4.18 Data on workplace participation from 2006 shows that men dominate certain industries, making up 89% of the workforce in construction, and 73% in manufacturing, whereas women make up the majority in sectors such as public administration, teaching and health, where they are 72% of the total workforce (Scottish Executive, 2007, p103). Trends in the distribution of men and women in employment show little overall change between 1996 and 2004 (Scottish Executive, 2007, p104).
Implications of these patterns
4.19 Under-representation is concerning for several reasons. The literature suggests that these concerns centre round three main themes: equalities, culture, and business and economic. The following section therefore sets out the main arguments in each theme.
4.20 The unequal representation of women in scientific education and careers can be framed as an equality issue. This foregrounding of equality can be linked to an overall aim of fairness in liberal society. However, in recent academic debates justifications based solely on equality have become less common and concerns based on economic factors have become more commonly cited. Some authors are concerned that justifications based on equality are being displaced by economic concerns (Phipps, 2008).
4.21 Authors have suggested that greater female involvement in science, in particular academic science, could act to change the culture of science for the better (Shiebinger, 2000). This argument emerged from various discussions from the 1970s onwards about the masculine nature of science and the practice of science (Phipps, 2008). Some argued that higher proportions of women could lead to changes in the way science was conducted. However, the extent to which science would be fundamentally changed by the involvement of more women has been questioned by some authors (Phipps, 2008). There are few investigations of this theme in the literature on the UK (Larios et. al., 2009).
Business and Economic
4.22 The idea that gender segregation is detrimental to business is also highlighted as a major concern. Businesses with more diverse staff and leadership tend to perform more positively on financial measures ( UKRC, 2010). This argument has been adopted by many of the key organisations working in this field including the UKRC. The UKRC argues that companies images and respected status, for instance as an employer of choice, will be affected by whether they have a diverse workforce ( UKRC, 2010). A diverse workforce will have other benefits in increasing productivity by recruiting from the whole talent pool, and profitability by being able to identify with their entire range of clients (Bagilhole, Powell and Dainty, 2008).
4.23 Costs to businesses are also highlighted. The loss of skilled and productive women from SET employment is represented as having a financial impact on business due to the costs of replacing staff and the loss of skills and expertise then experienced staff leaves ( UKRC, 2010). The UKRC has also emphasised the potential costs to businesses of ignoring equalities issues or getting them wrong, highlighting the potential for high compensation settlements if equalities issues are ignored ( UKRC, 2010).
4.24 Skills shortages are also cited as a reason for concern. Projections predicted that by 2010, white able bodied males would be in the minority in the labour force, meaning that companies that recruit primarily from this group would need to look more widely or risk being caught up in a skills shortage ( UKRC, 2010). This could be placed in a science specific context: as the uptake of science subjects by men decreases, the problems of uptake and retention of women in science become more visible and acute leading to more concern about attracting women to SET (Glover and Fielding, 1999). Statistics from the Labour Force Survey reveal that white males make up just under half (48.3%) of all those in employment in the UK, which constitutes a larger proportion of the workforce than any other group. The statistics are not disaggregated by able and non able bodied ( ONS, July-Sept 2010).
4.25 The business and economic related arguments for retaining women in SET careers have, however, been criticised by some authors as inadvertently reinforcing some existing discriminatory ideas about women (Phipps, 2008). Some authors have claimed that the recent primary focus on the economic benefits of women in science could perpetuate problems for women in science by singling them out as different from other scientists, and thus perpetuating cultures which see women as different and potentially unsuited to scientific careers (Phipps, 2008). By presenting women as the answer to a skills shortage there is also danger that they could be seen as the "last resort" and not as a normal group to recruit from (Bagilhole et. al., 2008). This would mean if skill shortages abated, this rationale for recruiting women would then disappear. An historical example of this is the employment of women to fill temporary skill shortages in engineering during World War Two. When the skill shortages no longer existed, women's employment in engineering declined. The business case could therefore undermine the long term progress of women in science.
4.26 The business and economic argument also does not deal with some factors affecting women's retention in science. Attempting to recruit women as a way to avoid skills problems or maximise profit can co-exist with a lack of attempts to change potentially harmful organisational cultures and workplace conditions that worked against the retention of women in science. Existing research shows that the idea of "critical mass", whereby increasing representation of women in particular sectors leads to changes in the conditions in these sectors, does not seem to strongly improve the position of women in science (Etzkowitz et. al., 1994). For these reasons, the recent focus on the business case has been seen by some as providing an incomplete view of under-representation of women in science. All of the impacts described in this section must be taken into account if the full significance of failing to tackle under-representation of women in science is to be understood.