Glass cliff is one of the words on our Word of the Year 2016 shortlist; in this article, the inventors of the term, S. Alexander Haslam and Michelle Ryan, explore its history and explain why they coined it. Find out more about our Word of the Year 2016, post-truth.
On 11 November 2003, Elizabeth Judge wrote an article that appeared on the font page of the Business Section of The Times in which she sought to explore the impact of women having broken through the glass ceiling by securing senior leadership positions on the boards of UK companies. The key question in which she was interested was whether the women’s leadership had been a help or hindrance to company performance.
To answer this question Judge compared the performance of the ten companies with most women on their boards (e.g., Marks & Spencer, J Sainsbury, Shell) to the performance of five companies with no women on their boards (e.g., Schroders, Anglo American, Rio Tinto). What she observed was that since the beginning of the year, the share price of all the latter companies had all increased (by an average of over 20%) whereas the share price of the six of the former companies had decreased (by an average of nearly 10%). On this basis, she drew an alarming conclusion (subsequently altered in the online text): that “The triumphant march of women into the country’s boardrooms has … wreaked havoc on company performance”.
Questioning The Times account
It so happened that on the day that this article was published one of us (Alex) had been invited to go up to London to give a talk about women in leadership to an interest group organised by Dianah Worman from the Chartered Institute for Personnel Development. Upon arriving at the meeting, though, the group was in a state of some distress after having read The Times article. For clearly Judge’s conclusions were a major setback to the business case that the group was seeking to mount for promoting more women into senior leadership positions.
However, as heated discussion ensued, Alex pointed out that Judge’s conclusion was entirely unwarranted — since the evidence on which it was based was correlational. It was entirely possible, then, that these patterns were a reflection not of the negative impact that women in leadership have on company performance but of the tendency for women to secure leadership positions only when companies are doing poorly.
As things stood, though, there was no way to establish which (if either) of these interpretations were correct. This was clearly an interesting question for some more robust research to resolve. And so, upon returning to Exeter, Alex went straight to Michelle to discuss the possibility of doing just this. This was the obvious thing to do because, while his main interests were in general processes of leadership, she had recently completed her PhD on the ways in which gender differences are shaped by identity and context. At the time they were both working together on a Leverhulme-funded project looking at the impact of group identity on decision making — an issue to which the Judge article was broadly relevant.
Our discussions about how to proceed centred on what would be the best way to provide definite insight into the causal question that Judge was addressing. We recognized that the ideal way to do this would be to conduct controlled experiments, but that these would take a long time to conduct and that their findings would lack the rhetorical force of Judge’s findings. Accordingly, we decided to focus our initial efforts on doing forensic archival research to unpick the sequence of events that contributed to the findings reported in The Times. To do this, Michelle went through the painstaking process of forensically examining the records of all nineteen company boards which had appointed women in the previous year, as well as those of a matched sample which had appointed men.
As the results of this process emerged, two things became clear. First, Judge’s core observation was correct: women’s appointment did coincide with poor company performance. However, second, her interpretation of this pattern was clearly wrong. For rather than company performance declining after women were appointed, it was actually the case that their appointment tended to be preceded by a period of consistently poor performance. Moreover, there was no evidence of a similar pattern for men.
These results were exciting for a range of reasons – primarily because it was apparent that our findings gave us new insight into the conditions that women face when they are given leadership positions. For while previous ‘glass ceiling’ research had been concerned largely with issues of quantity of leadership positions occupied by women, these new findings opened up our eyes to the fact that their quality was also an issue. That is, the findings suggested that there were reasons to be concerned not just about the number of leadership positions given to women, but also about the nature of those positions.
Deciding on terminology
As we discussed these issues more, one key question concerned how best we should characterize the phenomenon we had uncovered. Clearly this related to previous discussions of the glass ceiling (defined by Oxford Dictionaries as ‘an unacknowledged barrier to advancement in a profession, especially affecting women and members of minorities’) — where the reference to ‘glass’ refers to a problem that is real but invisible. This was true in the present case too, but here we were dealing with a situation in which women were being promoted into leadership positions that were risky and precarious as a consequence of the poor performance of the companies they were leading. This made it likely that their actions would be under the spotlight and that they would be blamed for problems that were not their fault (as witnessed in Judge’s original article).
It was in the context of these discussions — of which there were a huge number — that we came up with, and settled upon, the epithet by which the tendency to appoint women to precarious (or simply suboptimal) leadership positions has come to be known: the glass cliff. From the moment we started to tag our research in this way it was clear that people understood what we were talking about and that the nomenclature resonated with the experiences of a great many women — not only in business, but in politics, in education, in policing, in law; indeed, pretty much everywhere that leadership positions are on offer.
Certainly we encountered a degree of resistance along the way too. Principally this came from men (and anonymous journal reviewers) who objected to the identification of yet another form of gender-based discrimination. Yet as our research progressed, it became increasingly clear that the glass cliff was a real and widespread phenomenon.
Moreover, the terminology we had settled upon quickly gained traction. On the one hand, it found influence with other researchers. As a result, Google Scholar today identifies over 1,500 research publications that refer to the glass cliff. On the other hand, the term has proved to be of interest (and use) to agenda setters, policy makers, and social commentators more broadly. The New York Times identified the glass cliff as a one of the big ideas that shaped 2008, and Google now identifies over 60,000 references to the term on the Internet. It would be rare for a week to go by without a newspaper or magazine article being written somewhere in the world to address the phenomenon in some shape or form.
One obvious reason for this resonance is that in the decade since we did our original research a large number of women have been promoted to prominent but precarious positions of leadership — women such as Angela Merkel and Julia Gillard. This was particularly true in the wake of the crises triggered by the 2007-8 financial meltdown when women such as Mary Barra, Marissa Mayer, Johanna Sigurdardottir, Ellen Pao, Jill Abramson, Janet Yellen, Theresa May, and Hillary Clinton came to the fore. Indeed, as Emily Peck recently observed in the Huffington Post, these are the women of the glass cliff generation.
In one sense, of course, this is a cause for concern. But in another, it is something to celebrate. Moreover, because we have a term to describe their predicament we are well placed to understand and learn from it, and to work to towards a world in which their (and our) futures are less perilous.
S. Alexander Haslam is Professor of Social and Organizational Psychology and Australian Laureate Fellow at the University of Queensland and the University of Exeter. He is the author, with Steve Reicher and Michael Platow, of The New Psychology of Leadership: Identity, Influence and Power (Psychology Press, 2011).
Michelle Ryan is Professor of Social and Organisational Psychology at the University of Exeter and Professor of Diversity at the University of Groningen. She is the editor, with Manuela Barreto and Michael Schmitt, of The Glass Cliff in the 21st Century: Understanding Barriers to Gender Equality (American Psychological Association, 2009).