Introduction

The UK government has consistently failed to consider gender in its response to Covid-19, despite the fact men and women have been affected in distinct ways. Men have been disproportionately infected with the virus [1], while women have been disproportionately affected by the polices introduced to inhibit disease transmission [2]. As has been the case in previous outbreaks [3, 4, 5], women have suffered from the consequent effects of epidemics.

During COVID-19, women have disproportionately performed paid work on the frontlines of the pandemic. Of the NHS workforce, 77% are women [6], a number that increases to 90% if including the social care workforce. Working on the Covid-19 frontlines has required long, gruelling work conditions in full PPE (assuming it is even available) and has imposed a mental toll through working with infected and dying patients. The challenge has been exacerbated further by the fact that many have changed their domestic arrangements to reduce the risk of exposing their families to infection [7, 8]. Women have also absorbed the unpaid work associated with COVID-19 social changes: as schools and offices have closed, there has been an increased domestic workload within households. This includes, but is not limited to, home-schooling, additional cooking, and additional cleaning, alongside the mental load of it all. Numerous data sources have demonstrated that while men’s unpaid labour has increased compared to their baseline during the pandemic, women’s unpaid labour has grown significantly more [9, 10]. What’s more, women also appear to have absorbed much of the non-developmental work in the household (what needs to happen to keep everyone alive—cooking, cleaning, washing), compared to the developmental care (such as playing games or learning with children), compounding household bargaining disparities [11]. Women tend to assume these social roles for three main reasons: firstly, social gender norms that continue to prevail in UK society mean women typically assume more domestic labour within dual parent households; secondly, the gender pay-gap dictates that when parents are making decisions as to which parent remains working in paid employment and which parent reduces paid employment to assist with their children’s care, it tends to be the higher earner—most often the man—who continues to work [12]; and finally, women have disproportionately been furloughed or made redundant during the pandemic [9], meaning that they are available to take on the additional unpaid labour. This latter instance is due to the interaction between the sectors of the economy that have been shut down by government order and highly feminised employment patterns [13].

Beyond changes to labour, women have been affected by increasing rates of domestic violence, with proxy indicators suggesting soaring rates of violence and femicide [15]; women have also suffered from the need to redirect health service resources towards COVID-19. As NHS services have been rationed, moved online, or been cancelled, maternity and sexual reproductive health services have been acutely affected [16]. This has left women feeling less confident throughout the course of their pregnancies and has seen concerning trends of poorer birth outcomes start to emerge [17].

Given that government policy reflects what information ministers are given, we sought to understand whether and how the government had taken gender into account. The Scientific Advisory Group for Emergencies (SAGE) is the group of experts that advise the government on scientific and technical aspects of emergency response and recovery [18]. Selected by the SAGE Secretariat, led by the chief medical officer (CMO) and chief scientific officer (CSO), participants are chosen based on their expertise in areas the Secretariat deems to be necessary for mounting a robust response to a crisis, such as the COVID-19 pandemic: what is necessary typically depends on how the government defines the crisis. The government’s definition of the COVID-19 crisis focused on its epidemiological nature and the need to move towards ‘herd immunity’ through vaccinations while also finding ‘some way to shield the most vulnerable’ [19]. This narrow epidemiological approach, by definition, excluded broader social considerations, including gender, from SAGE’s ambit, despite recognition of the need for social science to be engaged in any epidemic response, as shown during the Ebola outbreak in West Africa [20].

This speaks to a broader issue within the politics of evidence, which is whose voices matter and what is considered to be necessary evidence [21, 22], as well as the politicisation of the expert selection process, with experts selected because their views accord with the government’s desires, rather than because of the rigour of their analysis [23]. Compounding this was the fact that those invited into the room were circumscribed in what advice they could provide. It was assumed by members of the SAGE that issues beyond their remit concerning ‘operational matters such as testing, protective equipment, and respirators, or the need to check on care homes… were being addressed’ elsewhere by government [24]. Thus, in analysing the content of SAGE advice and its interrelation with the gendered effects being experienced, we also need to consider the limits of SAGE’s mandate and what should be occurring elsewhere in the government response, such as through the Minister for Women and Equalities or the Government Equalities Office. However, given the prominent role that SAGE has played in producing evidence on which the government bases its policy decisions, and the repeated claims that the government is ‘following the science’, we sought to understand whether gender had ever been a feature of this advice and if not, whether this contributes to No 10’s failure to consider gender within its policy response.

Methods

We analysed SAGE minutes related to the COVID-19 crisis from 22 January to 22 December 2020. We conducted two forms of policy analysis to understand the gender considerations within these SAGE documents. The first was to search for explicit reference to sex and/or gender. This comprised searching for the following express mentions: women; woman; man; men; male; female; gender; pregnan*, and sex. The second step was to look for gendered issues and where the implications for gender had not been acknowledged or considered as such, arising through a gender-blind advisory process. This involved a careful reading of all minutes, looking for issues that we know to be inherently gendered within UK society and those which have been already identified as gendered by the growing literature on gender and COVID-19. This includes, but is not limited to, issues relating to care, healthcare workers, schools, childcare, education, sex-disaggregated data, domestic violence, care homes, social care, and other sectors of the economy that are heavily feminised (or indeed masculinised). To ensure that we didn’t miss other contributions to SAGE analysis, for those SAGE meetings in which we found explicit and/or unacknowledged gender considerations, we also included in our analysis all the meeting documents provided for that session. This was to allow us to understand whether gender issues had been raised but not carried through into the consolidated advice given to government. Finally, we categorised each of these appearances into gendered themes: education, sexual and reproductive health, labour, healthcare workforce, childcare, domestic violence, sex disaggregated data, and feminised sectors of the economy.

Given the increasing noise surrounding the role of women’s participation in leadership and COVID-19 advisory groups [25], we also analysed the composition of SAGE meetings, where publicly available, to understand the gender parity of these meetings. We did this by calculating the proportion of women from the total number of scientific expert attendees reported for each SAGE meeting. A minority of experts chose to have their names redacted, although this was limited to 0, 1, or 2 redactions per meeting.

Results

The summary findings from the analysis are in Table 1. Below, square brackets relate to the SAGE meeting where this content was considered. Where similar findings were found in supporting papers to minutes, only the minutes are recorded.

Discussion

Ongoing research continues to highlight the many gendered effects and consequences of the COVID-19 pandemic, and whilst topics that have significant effects on women are discussed at length during SAGE meetings, acknowledgement of the relationship that issues such as schools, care, healthcare workers, and particular employment sectors have with gender is frequently absent. When women are expressly considered, it is predominantly linked to sex (biological) rather than gender (social), such as considering the risks COVID-19 may pose to pregnancy and analysis of sex-disaggregated data (case data that is broken down by male and female). Where social behaviours are included and gender is considered, the evidence from the meetings suggests that this concern only extends to men, most particularly young men who are thought to be the least likely to adhere to lockdown restrictions [43, 44]. It should be acknowledged that there has been some shift, however, and that as the pandemic has progressed, there has been explicit consideration of the downstream effects of government policy on women. Nonetheless, this has remained in supplementary documentation, suggesting that gendered concerns never featured within discussions of SAGE, while much of the content in the supplementary papers verges on reproducing gendered stereotypes and roles, rather than engaging substantively with the gender issues.

SAGE do appear to be acutely aware of the equity aspects of their recommendations, as well as the broader consequences of policy advice [43] stating ‘it is vital to measure the impacts of these interventions (beyond disease surveillance) where possible’ [12]. Yet this has predominantly referred to ethnic or socio-economic inequalities, for example: ‘there is some evidence that poorer people are engaging in more social contacts outside the home’ [28] and ‘there is likely to be an interplay of social, economic, biological, and pre-pandemic health risk that vary across [ethnic] groups’ [59]. It therefore appears that gender considerations have been missing. Given that there is growing evidence attesting to the fact that COVID-19 has a disproportionate effect on women, this is a significant concern [12, 45, 46].

This reflects a broader concern, which is what disciplines and types of evidence are deemed worthy of SAGE’s concern. Indeed, SAGE states ‘once policies are formulated, SAGE should review them through the lenses of epidemiological modelling and behavioural science’ [14], potentially excluding a raft of other academic perspectives. For example, even when considering school closures, a heavily gendered issue, the meeting minutes suggest that the focus was on the behavioural science and whether parents would keep their children out of school regardless or whether children might congregate in other areas, proving a risk for disease transmission [17]. There appears to have been limited consideration of these policies from a broader social science perspective, or any gender perspective, with such considerations only found in three background documents out of several hundred. We cannot say whether this is a chicken or an egg problem: those in the room were experts in epidemiology and behavioural science, and therefore it is unsurprising that these areas were the focal points of discussion. Yet it is also important to recognise the expertise missing from SAGE. This reflects the selection process, which in turn reflects the knowledge or information that the government deems valuable to responding to the crisis and what the CMO and CSO believe SAGE’s mandate is. On this occasion, as the above shows, the focus remained on epidemiological and behavioural considerations, to the detriment of the broader consequences on society. This reflects the UK government’s position that an effective response to COVID is reliant on individual responsibility, rather than social and structural barriers. What is so disappointing is that this narrow frame has already been shown to neglect the socio-economic consequences of outbreaks, particularly those affecting women [47, 48]. Indeed, a parliamentary review of SAGE convened for the 2014–16 West Africa Ebola outbreak reported that ‘many of our witnesses emphasised that … ensuring that the membership of SAGE included social scientists, were [sic] extremely important in controlling [the] outbreak’ but did not mention their importance in understanding and mitigating the harmful downstream impacts of public health interventions [49].

Whilst the gender blindness across SAGE minutes is notable, it is not the fault of SAGE members per se, but rather the Secretariat that is responsible for selecting participants. Instead of looking at the approach taken by SAGE, we need to consider why a gender advisor did not have a seat at the table, perhaps in the same way we might question why an advisor in black, Asian and minority ethnic integration was only introduced to SAGE in June [40] despite decades of NHS work and policy on ethnic health inequalities [50]. We know that gender advisors participate in decision-making in other areas of emergency response, such as climate change and humanitarianism, and so the idea of a gender advisor is not entirely foreign to such environments [5].

We cannot expect experts in epidemiology to consider all potential impacts of the outbreak and response. For example, despite debate between public health and the economy during COVID-19, there are no economists on SAGE. This is left to government economists (and any other economic advisors that this government consults). It maybe that gender considerations are also left to governmental gender advisors and assessors. The UK government has the Government Equalities Office (GEO), tasked with understanding gendered (and indeed any) inequalities created or compounded by UK government policy, and they may be equipped to address this in the context of COVID-19. However, evidence would suggest they have not done so, with only three COVID-19 related publications published by the GEO over the course of 2020, a mere fraction of a percentage of the more than 2700 published documents from across Westminster departments. Further, the GEO would seem to not be a government priority, judging from our ongoing interview work, with Conservative governments stripping funds (and power) from the GEO (forthcoming).

The UK government is required by the Equality Act 2010 to undertake assessments of the potential impacts of its policies on people and groups with protected characteristics [51]. However, it is not known at what point the UK government undertook such an assessment of their COVID-19 policies. To date, only the Department of Health and Social Care has published one, on 28 July 2020, four months after the first lockdown was introduced, and the assessment focuses on the areas of the COVID-19 response related to the department itself, rather than on the wider social impacts that needed to be considered in order to assess the impacts of the government responses on gender, and women in particular. Our interview data from the aforementioned project anecdotally revealed that, when such assessment was proposed in Cabinet, the staff member who suggested it was mocked for suggesting it at a time of crisis, demonstrating a wider narrative within the government that gender doesn’t matter during the ‘tyranny of the urgent’ in health emergency management [52]. This is particularly concerning given the wider approach of the Johnson Government to issues of inequality: in her December 2020 speech, Minister for Women and Equalities, Liz Truss, claimed that issues of race, sexuality, and gender were ‘fashionable’ rather than factual and that they detracted from the more significant issues of poverty and geographical inequalities [53]. The government’s commitment to extend its equalities work beyond the nine protected characteristics of the Equality Act is commendable, particularly as sex and gender reassignment are included as protected characteristics whereas gender is not. However, as COVID-19 has shown, there is still plenty of work to be done across the issues of gender, sexuality, and race and the intersectional manner in which inequalities of geography, income, gender, race, sexuality, disability interact.

The UK government has recognised gender within its COVID-19 response, actively highlighting in the COVID-19 recovery plan that pandemics can have gendered effects [54]. However, they only acknowledge this within their discussion of the international arena, as though gender inequality only occurs overseas. Meanwhile, the UK government closed schools in March without considering the gendered effects of the domestic labour that it would transfer to the home. The government either ignored, or failed to recognise, the impact this was having on women for some time, with businesses only permitted to furlough parents on childcare grounds in July 2020, importantly, however, at the employer’s discretion. Data from the TUC on the third lockdown has shown that 73% of employers have refused such requests [55]. The government’s response with regard to pregnancy was even more delayed, with NHS guidance not permitting women to have a birth partner during labour until 14 December, nine months into the pandemic [56]. As well as failing to introduce policies to assist women, the government decided to suspend gender pay gap reporting in March 2020 [57], suggesting that companies didn’t need to do this whilst they were managing the changes as a consequence of COVID-19. We now know, however, that gender pay gaps were a key determinant of women absorbing the domestic costs associated with lockdown and school closures [12]. We also don’t know whether, were SAGE to include gender informed advice, the government would listen and engage with it. SAGE are advisors, while policy is made by the coronavirus strategy and operations committee, government ministers, and ultimately, the prime minister [58]. These individuals may decide to ignore gender sensitive advice. Indeed, the government’s disregard for gender issues is evident from the fact that the role of minister for women, rather than being held by a single individual with a clear mandate, has been amalgamated into Liz Truss’s Ministry of International Trade. Not only does she not sit on the coronavirus strategy and operations committee within Cabinet, but she holds a role particularly crucial and time-consuming when Britain is attempting to renegotiate global trade deals independent from the EU.

Whilst some point to the need for greater women’s participation to ensure a gendered perspective is taken, as our data have shown, and as has been evidenced in myriad other contexts, having more women at the table does not automatically equate to more gender consideration in policy [59, 60]. Increasing the number of women participants in SAGE did not lead to greater awareness of gender issues. Indeed, there was no increase of explicit considerations of gender or women in meetings where more women were participating. Thus, whilst increasing the participation of women is important for the normative goal of gender parity in public life and leadership, this should not be seen as a synonym for gender advice. Being a woman doesn’t make you an expert in gender, no more than being French would make you an expert in French politics.

The lack of gender considerations will undoubtedly have long term effects. Whilst we have already seen more women furloughed than men, and women have lost their jobs 5% points more than men during 2020 [9], these numbers will likely only increase when the furlough scheme ends and widespread unemployment follows. Women who have been furloughed, including those on childcare grounds, have potentially demonstrated to their employers that they are non-essential for business continuity, particularly at a time of economic precarity. Moreover, if schools are not made safe to facilitate teaching, and if the precarious childcare sector fails and more facilities close, this will have long term effects on women’s labour force participation as women continue to absorb this unpaid care work. In the wake of the Ebola outbreak, 13 months after the start of the crisis, 63% of men had returned to work in Sierra Leone, compared to only 17% of women [14]. This is not only important for gender equality but also for the control of the pandemic in the UK, and health and livelihoods thereafter: people working on insecure or informal contracts, including 12.1% of Black and minority ethnic women, 6.4% of white women, and 5.5.% of white men, have been less likely to stay home if required to self-isolate due to a lack of sick pay and the risk of job loss [61]. With an additional 2.7 million people applying for Universal Credit between 12 March and 8 October 2020 [62], food banks have reported record increases in need throughout the pandemic period while government ministers refused to give free school meals to 1.3 million vulnerable children during school holidays, until public outcry forced a reversal in their decision [63, 64]. A key determinant of healthy child development is family income, suggesting that the impacts of COVID-19 and government response and recovery decision-making will be multi-generational [65, 66].

There can be many ‘unknown unknowns’ when it comes to pandemic planning and response, and those trained as modellers and biomedical scientists cannot be expected to universally consider the many social facets of an epidemiological response to an outbreak. However, the gendered impacts of government decision-making and the gendered impacts of infectious disease outbreaks are now well evidenced, and UK policymakers have a statutory requirement to consider the inequality impacts of all policy. In order to effectively consider the likely epidemiological, social, and economic impacts of NPIs, pandemic planning and response firstly needs to be framed beyond the lens of the biomedical sciences and secondly needs to include the full breadth of expertise that the UK has to offer, including those social scientists, gender specialists, and representatives of civil society organisations that work at the coalface and see the reality of government policy and/or inaction every day. As the government’s chief emergency scientific advisory group, SAGE currently presents its consolidated advice to government in a gender-blind form. Paired with an apparent lack of gender specialists feeding into government policy otherwise, such policies will inevitably remain gender-blind, leading to the reinforcement and exacerbation of gender inequality and making the lives of women that much harder.

Whilst this paper has focused on the role of gender in the UK COVID-19 context, there are meaningful implications for other countries that might have similar emergency advisory groups and/or are considering more robust permanent institutions to mitigate against future crises. Gendered impacts of COVID-19 are global trends [5], despite the context specific gender norms within borders. Our ongoing global research project on gender and COVID-19 has demonstrated the extent of the gendered implications and downstream effects of the pandemic, which transcend languages, social groups, and cultural settings. More alarmingly, despite this universality of gendered effects, very few countries have taken a gender-sensitive approach to pandemic response, and where they have this has tended to have a narrow focus on gender-based violence. Thus, it is crucial that gender advisory is a key part of emergency preparedness and response expertise presented to governments, and global policy and norms should seek to encourage and perpetuate this. This is particularly important in settings where gender inequality is entrenched and where a gender-blind policy response may significantly reinforce inequalities across society. Simply, we want governments to ask ‘where are the women?’