Abstract:
During the last two decades gender equality has come to be seen increasingly as equality of outcome. Anne Phillips argued in 2004 that inequality of outcome is invariably a result of inequality of opportunity, and this position seems to have been so widely accepted that skewed gender statistics are now generally interpreted as manifestations of gender inequality. In the present paper I show that Phillips ignored the time aspect of change, and I present empirical evidence that contradicts the philosophical arguments put forward in defence of equality of outcome as the touchstone of gender equality. Inequality of outcome can indeed co-exist with equality of opportunity. This has consequences for affirmative action. The distinction between equality of opportunity and equality of outcome must be maintained, lest affirmative action be misdirected to persons who are not in need of help and those whose opportunities are lacking, be ignored.
Keywords: equality of opportunity, equality of outcome, gender equality, affirmative action
Introduction
Gender equality is increasingly seen as equality of outcome: Even when equal pay is paid for equal work, the fact that women earn less on average is taken as an indication of gender inequality; It may be true that there are no formal hindrances for women becoming professors or CEOs of companies, but there are few female professors and CEOs, thus not de facto equality, etc.
This is in contrast to the classical concept of gender equality as equality of opportunity which was prevalent up to the late 20th century. How and why the latter position was largely abandoned is not entirely clear – failed expectations that equality of outcome would result from equality of opportunity have undoubtedly contributed. Alice Eagly, phrased the expectation like this in 2004:
”… demise of many sex differences with increasing gender equality is a prediction of social role theory that will be more adequately tested to the extent that societies produce conditions of equality or near-equality between women and men” (A. H. Eagly et al., 2004: 289)
Equality of outcome did, however, not materialized despite more equality of opportunity, and the emphasis shifted to equality of outcome. The feminist critique of the notion of equality of opportunity rested
“on the argument that opportunities for men and women are very seldom equal, since men have a privileged position in society, which implies that there is no real equality of opportunity” (Dahlerup, 2007).
It may be noted that this argument acknowledges equality of opportunity as the touchstone for gender equality, in that it argues that the problem lies in there not being real equality of opportunity.
Equally importantly, the argument acknowledges that there could be instances of women actually having the same opportunities as men, although this is assumed to be “very seldom”, implicitly so seldom that for practical purposes one need not worry about it. This may have been a reasonable position in the past, but the possibility must be considered that a time could come when this is no longer so. When this happens, the matter of greatest interest will be to determine specifically when real equality of opportunity exists and when it does not. I will give an example below of a setting where real equality of opportunity exists even though inequality of outcome is observed.
In parallel, I shall address a paper by Anne Phillips, professor of Gender Theory at London School of Economics and Political Science, (Phillips, 2004) which claims that inequality of outcome is invariably the result of inequality of opportunity. Obviously, if Phillips is right, there is no need to ask in concrete instances of inequality of outcome, whether inequality of opportunity is at play or not. In the following I will examine Phillips’ arguments, and show 1) that she ignored the time aspect of change, and 2) that 15 years of empirical research has largely eroded the basis for her conclusions. New empirical evidence overthrows the philosophical arguments.
Phillips’ defence of equality of outcome
In the paper entitled “Defending equality of outcome” Anne Philips (2004) argues that there is no difference between equality of outcome and equality of opportunity. Thus, gender equality is in effect defined by equality of outcome.
Phillips presents three possible ways that inequality of outcome can develop: individual choice, discrimination, and structural constraints.
The first, individual choice, cannot explain inequality of outcome in society, says Phillips, for
”When dealing with social groups, the variations are more easily averaged out, and it becomes inherently suspicious to attribute systematic differences in outcome to the different mind-sets of different groups: this begs too many questions about why the members of one group might have ended up with a radically different set of preferences to another“ (Phillips, 2004: 15)
We must thoroughly understand this premise because it seems to be valid. It is true that application of principles of individual actions to groups may lead one astray. It can be argued that some individuals are more energetic and talented than others, and some make choices the consequences of which can lead to very different social positions. On the individual level, we may argue that Hans and Jens had the same opportunities, but Hans worked hard to get an education and ended up in a well-paid job, whereas Jens squandered his youth away on partying, and ended up jobless. They had different initiative and made different choices, and it is just that Hans fared better than Jens. We could tell more elaborate stories about the reasons behind their choices, but these are part of their individual stories and have no general applicability. The choices of Hans and Jens are free in this sense.
But we cannot conclude about groups based on such choices. If a group in a society end up in a worse position than another group, one may put forward an explanation based on different choices and talent, but the story cannot end there. Differences in choice and talent must be systematic in some way in order to show up as group differences, they cannot be more or less random as in the case of Hans and Jens. The question why one group has more talent and makes different choices from another remains. We have not provided a group-answer.
Here is a simple thought experiment: The Danish population consists of around 5.8 million individuals. Some are talented, full of energy, and make wise choices. They fare better than the lazy and untalented. Now we let a computer draw a ballot to divide the population randomly into 2 groups of equal size. This could be done by an ordinary laptop in a few seconds. The result would be two groups of around 2.9 million each. These groups will be very close to identical with regard to age, height, weight, gender, social class, educational level, income, marital status, health, etc. We need not make the experiment, we know that the law of large numbers will apply and result in equal groups. There will be no trace of differences in talents and choices. Like everything else they will be equally distributed over groups.
Now divide the Danish population in two groups of almost equal size based on gender, and we will see the marked differences in labour statistics, pay gaps, CEOs, etc. that we know so well. If we want to explain these differences we cannot just claim that this has to do with people having different energy levels, ambitions, talents and making different choices. We are forced to assume marked, systematic, differences in ambitions, talent, and choice as explanations for the differences in outcome for the genders. We have in effect explained nothing by referring to ambitions, talent and choices, since we still have not explained how these systematic differences between the groups have come about.
If groups have equal opportunities, an explanation based in individual choice, talent, ambition is indeed suspect (Phillips, 2004). Thus far we can follow Phillips’ argument.
If not individual differences then what?
In discussing the two remaining possibilities, discrimination and structural constraints, Phillips begins from the simple case, political representation. In a democratic system it may be assumed that citizens wish to be represented by politicians of their own gender, ethnic group, race etc. As the saying goes, no one knows where the shoe pinches, but he who wears it. Unequal representation must therefore be due to obstacles: “the only explanation for underrepresentation is that something is blocking the way” (Phillips, 2004: 8), either discrimination understood as actions to keep others out, or as structural constraints. Both are inequalities of opportunity:
” a disparity arising from overt discrimination should not be regarded as significantly different from one that arises out of structural constraints. There is nothing particularly mysterious about the under-representation of women in politics, for in societies still shaped by a male breadwinner model and still requiring women to shoulder the bulk of care responsibilities, it is entirely predictable that more men will be available for a full-time political career.” (Phillips, 2004: 7)
Phillips is admirably clear about the nature of the structural constraints. Women must make themselves available as carers for the children. No society has yet removed the structural constraints related to childcare:
“Somebody has to devote her life to these activities that are incompatible with reaching the top of a large corporation, somebody has to make sure that the children are fed and clothed and get their homework done in time, and since no society has yet organised itself so that these activities can be shared equally between women and men, it is a fortunate fact of life that many women actively volunteer.” (Phillips, 2004: 12)
Since discrimination as well as structural constraints are inequalities of opportunity, Phillips can make the general conclusion:
”It makes sense to start from the expectation that all groups would normally be distributed in roughly equal proportions along all measures of social activity: to expect, therefore, an equality of outcome, and to take any divergence from this as a reasonably safe indication that opportunities are not yet equal.” (Phillips, 2004: 18), my emphasis.
In short, Phillips’ argument is that inequalities of outcome may be due to discrimination or structural constrains, both of which are inequalities of opportunity. The commonplace reference to differences in choices and preferences is invalid, because such individual differences will tend to equal out between groups.
What is implied here, is that there are no other possible explanation of inequalities of outcome than these three: discrimination, structural constraints, and personal choice. But as often happens, reality turns out to be more complex. Phillips has neglected at least two issues: Time and the “equality paradox”.
Time
Inequalities of outcome are usually not seen in a historical perspective. Below I give an example of a case of inequality of outcome that appears to be a time-lagged consequence of inequalities of the past: the low proportion of women among professors in Danish universities. The obvious inequality of outcome is that only around 20% of professors are women, whereas 60% of the students are women. This inequality is often presented as a leaky pipeline, a term coined by Alper (1993) and typically illustrated as seen in Figure 1 (European Commission, 2019; Gvozdanovic and Maes, 2018) which shows data for Danish universities in 2013 (Uddannelses- og Forskningsministeriet, 2017). It is argued that women are lost along the way in the career path.
The term Leaky pipeline is, however, a misnomer: a pipeline is a structure with a horizontal flow, but what is depicted in Figure 1 is a cross-sectional analysis of the population of students and researchers at a point in time. Since it takes time to become qualified for late career stages, the “pipeline” will look bad when preceded by a period where many women have entered the early career stages. Thus, when things are going well in terms of many women getting into academia, the pipeline seems to be leaking.
In Denmark where the mean age of new professors is 48 years (Ståhle, 2007), it will take 35 years before an increase in the female proportion among college students is fully implemented in the professor pool. The interpretation of the leaky pipeline as a career path is an extreme example of ignored time effects.
Figure 1. University students and scientific staff at Danish universities 2013, redrawn from (Uddannelses- og Forskningsministeriet, 2017)
Addition of data about new hires (Table 1) gives an understanding of the dynamics of the population. The female proportion of professors hired in the period 2015-2017 was 30% which was close to the female representation in the preceding career step, tenure, of 32%. Likewise, the female proportion of newly hired tenure was 39%, which was close to the female proportion in the preceding career step, postdoc, of 40%. Thus, there is no leak in the pipeline between tenure and full professor or between postdoc and tenure.
Table 1. Proportion of females among academic staff on Danish Universities 2015-2017. Existing pool and new hires (Uddannelses- og Forskningsministeriet, 2018: 14)
| Postdoc | Tenure | Full professor | |||
| New hires | Existing pool | New hires | Existing pool | New hires | Existing pool |
| 39% | 40% | 39% | 32% | 30% | 19% |
The findings of Table 1 are supported by statistics about hires to research positions at Danish universities. It has been known since the late 1990ies that when there were both male and female applicant to research positions, the women have higher chances of getting the job than expected (Ståhle, 1999). As seen in Figure 2, this applied to all types of academic positions. Specifically for postdocs the success rate was 47% for qualified women and 41% for qualified men in 1995-1997. Since then the overall chance of getting the job has declined for both genders and the relative chance for women over men has further increased so that in the period 2015-2017 the chance of getting the job as postdoc was 8% for a qualified woman and 4% for a qualified man (Uddannelses- og Forskningsministeriet, 2018: 24).
Figure 2. Success rates for men and women applying for academic positions at all Danish universities, 1995-1997, 2004-2006, 2007-2009, 2015-2017. Positions with at least one qualified application from each gender. Data extracted from (Ståhle, 1999, 2007, 2011; Uddannelses- og Forskningsministeriet, 2018)
It has been claimed that discrimination in funding is hampering women’s careers in academia (Watson and Hjorth, 2015), but statistics about grants by gender from the Independent Research Fund Denmark (Uddannelses- og Forskningsministeriet, n.d.) show no difference in the success rates of men and women of applications or amounts of funding obtained in recent years.
Taken together, the statistics do not support the interpretation that the low proportion of women among professors is due to inequalities of opportunities for women when applying for professorates in Danish academia. This notwithstanding the Danish Parliament has repeatedly passed legislation to increase the chances of women advancing from tenure to full professor.
This does not rule out other inequalities of opportunity, especially early in the career path and related to pregnancy and childbirth – indeed I think there are some. My point here is that a simplistic interpretation of all inequalities of outcome as consequences of inequalities of opportunity can very easily be misleading. As a consequence of such simplistic thinking, affirmative action may be misdirected – as in the case above where the female scientists who would manage on their own since they already obtained tenure, are being lifted by affirmative action, whereas the young female scientists who must combine an academic career with raising children and having a family and who do need help for advancement of their career, are not.
Paradoxes of equality
In 2004 Alice Eagly derived from social role theory that the differences between genders would diminish as equality of opportunity between genders increased, and she expected future research to prove this prediction. Research from the last 15 years concludes otherwise. Adhering to convention I am going to call the findings paradoxes of equality, although there is nothing paradoxical about research disproving predictions from theory. The feeling of paradox stems from exaggerated confidence in the theory in the first place.
Psychological make-up
The most solid example of a paradox of equality is to be found in the research on the psychological make-up of the mind based on the Five Factor Model, also known as The Big Five, which measures the five personality traits
- Extraversion
- Neuroticism
- Openness to experience
- Conscientiousness
- Agreeableness
In a large study, the women reported higher levels of extraversion, neuroticism, conscientiousness, and agreeableness, whereas the men scored higher than the women on openness (Schmitt et al., 2008). One would expect that differences between men and women would be larger in countries with larger gender inequality, but this turns out not to be true. Costa et al. (2001) were the first to report this surprising finding in the paper ”Gender differences in personality traits across cultures: Robust and surprising findings”. This study compared data about The Big Five from 22 countries, and correlated the reports with measures of gender equality on the country level. There was a strong correlation between the level of individualism in the country and the magnitude of the difference between genders on The Big Five. In more individualistic (Western) cultures, differences between men and women in the psychological traits measured by the Big Five were larger than in less individualistic cultures.
Schmitt et al. (2008) performed a more comprehensive study of The Big Five in 55 countries with 17,637 respondents. In this study the classification of the 55 countries included several measures of formal gender equality. Again the finding was that participants from more egalitarian countries reported larger gender differences on The Big Five than those from less egalitarian countries. This applies also when other methods of assessing personality traits are used (Schmitt et al., 2017). Two additional studies (Kaiser, 2019; Mac Giolla and Kajonius, 2018) have replicated these findings.
Thus, there is compelling evidence that gender differences between men’s and women’s psychological traits do not diminish in more egalitarian societies, rather on the contrary.
What this means is that there might after all be a case for explaining differences of outcome by individual choice. Simply put: if women are, on average, kinder and more conscientious than men, then that could lead to systematic differences in choice of for example vocation, and what is more, these systematic differences in choice do not diminish in more egalitarian countries. This should, however, not be exaggerated: the differences are rather small compared to the massive difference we see in for example the choice of jobs. What is important about these findings is that they show that individual choices may be systematic over groups.
Women in STEM
Looking more directly at choice of vocation in egalitarian and less egalitarian societies, a similar pattern emerges. The STEM (science, technology, engineering, mathematics) occupations have a low “representation” of women, and efforts have been made for decades to attract more women, albeit with limited success. The relative lack of women in STEM is seen as a loss of talent which these areas can ill afford.
Stoet & Geary (2018) studied the relation between gender equality of countries – as expressed by the WHO Global Gender Gap Index – and women’s choice of STEM educations. A clear inverse association was found: the more equal the country, the less women in STEM. Charles & Bradley (2009) found that Colombia, Bulgaria, and Tunisia, countries that we do not regard as gender-egalitarian, had the lowest gender segregation in educations, whereas Finland, Hong Kong, South Africa, and Switzerland were the four countries that had the most pronounced segregation of educations. They concluded that the genders are better able to express their “gendered selves” in more developed countries.
These examples should suffice to show that the prediction of gender role theory that gender differences would disappear in more egalitarian societies is unsupported by evidence. Simply put, women choose health educations and men choose STEM more often when their choices are less constrained. There is much to be said for the theory that lack of constraints leads to a more diverse labour market.
Unless one would challenge that Western societies have made substantial progress in equality of opportunity during the last 50 years, the implication is that more equality of opportunity is not necessarily associated with more equality of outcome. This forces us to reject the theory that inequality of outcome is invariably a result of inequality of opportunity.
Conclusion
The idea that inequalities of outcome are due to constraints on what women can do, is contradicted by the observation that countries with less constraints exhibit greater differences between genders in for example choice of vocation. With the example of the gender balance among professors at Danish universities, I have shown that even when there is inequality of outcome there can be equality of opportunity. More examples of co-existing inequality of outcome and equality of opportunity could be presented, but extensive discussion of the complex issues pertaining to each is not within the scope of this paper. Such discussions are conspicuously absent from the public discourse which to a large extent implements Phillips’ principle of equating inequality of outcome with inequality of opportunity. This is conceptually untenable. We cannot assume off-hand that an observed inequality of outcome is an inequality of opportunity. If we do, we are likely to misdirect affirmative action.
References:
A. H. Eagly, Wood W and Johannesen-Schmidt MC (2004) Social role theory of sex differences and similarities: Implications for the partner preferences of women and men. In: Eagly AH, Beall EA, and Sternberg R (eds) The Psychology of Gender. 2nd ed. New York: Guilford.
Alper J (1993) The Pipeline Is Leaking Women All the Way Along. Science 260(5106): 409–411.
Charles M and Bradley K (2009) Indulging Our Gendered Selves? Sex Segregation by Field of Study in 44 Countries. American Journal of Sociology 114(4). The University of Chicago Press: 924–976. DOI: 10.1086/595942.
Costa Jr. PT, Terracciano A and McCrae RR (2001) Gender differences in personality traits across cultures: Robust and surprising findings. Journal of Personality and Social Psychology 81(2): 322–331. DOI: 10.1037/0022-3514.81.2.322.
Dahlerup D (2007) Electoral gender quotas: Between equality of opportunity and equality of result. Representation 43(2). Routledge: 73–92. DOI: 10.1080/00344890701363227.
European Commission (2019) She Figures 2018. Brussels: European Commission, Directorate-General for Research and Innovation. doi: 10.2777/936.
Gvozdanovic J and Maes K (2018) Implicit bias in academia: A challenge to the meritocratic principle and to women’s careers – And what to do about it. League of European Research Universities (LERU).
Kaiser T (2019) Nature and evoked culture: Sex differences in personality are uniquely correlated with ecological stress. Personality and Individual Differences 148: 67–72. DOI: 10.1016/j.paid.2019.05.011.
Mac Giolla E and Kajonius PJ (2018) Sex differences in personality are larger in gender equal countries: Replicating and extending a surprising finding. International Journal of Psychology 0(0). DOI: 10.1002/ijop.12529.
Phillips A (2004) Defending Equality of Outcome. Journal of Political Philosophy 12(1): 1–19. DOI: 10.1111/j.1467-9760.2004.00188.x.
Schmitt DP, Realo A, Voracek M, et al. (2008) Why can’t a man be more like a woman? Sex differences in Big Five personality traits across 55 cultures. Journal of Personality and Social Psychology 94(1): 168–182. DOI: 10.1037/0022-3514.94.1.168.
Schmitt DP, Long AE, McPhearson A, et al. (2017) Personality and gender differences in global perspective. International Journal of Psychology 52(S1): 45–56. DOI: 10.1002/ijop.12265.
Ståhle B (1999) Alder, Køn Og Rekruttering i Dansk Universistetsforskning. København: Uni-C.
Ståhle B (2007) Fornyelse i Forskerstaben. Forskerpersonal Og Forskerrekruttering På Danske Universiteter 2004-2006. København: Uni-C. Available at: https://ufm.dk/forskning-og-innovation/statistik-og-analyser/forskere-ved-universiteterne/fornyelse-i-forskerstaben-forskerpersonale-og-forskerrekruttering-pa-danske-universiteter-2004-2006-1.pdf.
Ståhle B (2011) En Forskerstab i Vækst. Forskerpersonale Og Forskerrekruttering På Danske Universiteter 2007-2009. Uni-C Statistik & Analyse.
Stoet G and Geary DC (2018) The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education. Psychological Science 29(4): 581–593. DOI: 10.1177/0956797617741719.
Uddannelses- og Forskningsministeriet (2017) Mænd og kvinder på de danske universiteter – Danmarks talentbarometer 2017. Uddannelses- og Forskningsministeriet. Available at: https://ufm.dk/publikationer/2017/filer/maend-og-kvinder-pa-de-danske-universiteter-danmarks-talentbarometer-2017.pdf.
Uddannelses- og Forskningsministeriet (2018) Forskerrekruttering på Universiteterne 2015-2017. Uddannelses- og Forskningsministeriet. Available at: https://ufm.dk/forskning-og-innovation/statistik-og-analyser/forskere-ved-universiteterne/forskerrekruttering-pa-universiteterne-2015_2017.pdf.
Uddannelses- og Forskningsministeriet (n.d.) Publikationer — Uddannelses- og Forskningsministeriet (Publications from the Ministry and the Agencies). Available at: https://ufm.dk/publikationer (accessed 30 January 2020).
Watson D and Hjorth J (2015) Women’s grants lost in inequality ocean. Nature 519(7542): 158–158. DOI: 10.1038/519158d.
