Be assertive, but don’t come off as bossy. Be warm and modest, but not so much that people think you’re incompetent. It may be 2026, but gender stereotypes still affect how women in leadership positions are perceived – and, as a result, how they choose to communicate. 

Many women leaders navigate a frustrating double bind rooted in how they frame their messaging. Promotion-focused language, emphasising bold action and opportunity, can clash with gender norms; prevention-focused language, highlighting caution and risk-avoidance, may reinforce perceptions of low competence.

This conundrum is amplified in crisis situations, such as during the Covid-19 global pandemic, or when a firm is facing an existential threat. In these scenarios, leaders must frame decisions in ways that motivate followers to change their behaviours while still retaining their support. 

Although all leaders must strike this balance, it’s tricker for women because the framing that tends to be most effective in a crisis – using action-oriented, promotion-focused language and avoiding risk-based, prevention-focused language – is also the framing most likely to clash with gender norms that expect women be cautious and conscientious. To lead effectively in such critical times, women may have to defy these stereotypes, risking backlash.

Through a gendered lens

My co-authors* and I conducted research to test the impact of using gender-stereotypical language on both the approval ratings of state governors in the United States during Covid-19 and their ability to promote effective social-distancing behaviours. 

Specifically, we looked at the regulatory focus with which governors framed their policies. Regulatory focus distinguishes between two strategies: prevention, which focuses on safety, security and avoiding a situation getting worse; and promotion, which focuses on growth, achievement and advancing to a better situation. Whether a leader adopts more of a prevention or promotion focus can play a key role in their ability to exert authority effectively.

Regulatory focus is also relevant to gender. Research suggests that women are expected to behave in prevention-oriented ways and are rewarded for doing so, while men are rewarded for promotion-oriented behaviours. Gender-stereotypical language for women is therefore characterised by more prevention-focused words (e.g. “safe”, “protect” and “lose”) and fewer promotion-focused words (e.g. “take”, “gain” and “win”).

Since women who conform to gender stereotypes are evaluated more favourably in terms of likeability, one might also expect them to be more effective when using prevention-based language. But for women in leadership roles, things are more complicated: the same qualities that make them well-liked can also undermine perceptions of their competence. In other words, when women adhere to gender stereotypes, they’re viewed as more likeable but less effective.

Communicating in a crisis

We analysed over 3,700 documents capturing the Covid-19 communications of 41 male and 9 female governors to determine whether they used more prevention- or promotion-focused language. This data was drawn from press conferences and press releases.

We also assessed compliance with governors’ policies by measuring changes in people’s movement to public spaces relative to pre-pandemic levels, using data from Google’s Community Mobility Reports. We then harnessed survey data capturing governors’ approval ratings to determine how positively they were perceived. Finally, we analysed over 24,000 posts on X (formerly Twitter) by US governors during the early days of the pandemic. We interpreted reposts as a proxy for intended behavioural compliance and likes as a proxy for general approval.

The double bind

We found that female governors who used gender-stereotypical language received higher approval ratings, but this didn’t translate into leadership effectiveness. In fact, it was associated with lower compliance with their social-distancing policies. On the flip side, female governors who used more counter-stereotypical, low-prevention-focused language relative to their pre-pandemic baseline were more effective leaders, but were perceived more negatively.

This highlights the double bind for female leaders – one that goes beyond perceptions of competence to affecting their ability to influence constituents’ behaviour. This is especially striking given that in many high-profile leadership roles, being both liked and effective is essential. Yet, in the context of our study, these goals required diametrically opposed strategies for women.

We also found that when male governors used prevention-focused language, this was negatively associated with both behavioural compliance and approval. In other words, men were similarly rewarded for following gender norms and punished for deviating from them. 

Moreover, we consistently found that constituents’ behaviours and perceptions were more sensitive to the language choices of female leaders. This is likely because of the greater scrutiny women face, and is in line with research suggesting that women’s actions are more likely to be put under a microscope. Although this amplifies the personal and societal stakes of the trade-offs women leaders must contend with, it could also represent an opportunity: women may be able to lead more effectively precisely because their actions elicit stronger responses.

Practical implications 

Our research uncovers a consistent pattern that shows just how stark – and unfair – the systemic trade-offs for women leaders can be. Although the responsibility for change ultimately lies with organisations and broader society, our findings point to some practical takeaways that could help women in leadership roles navigate this double bind. 

Given the importance of language choices and the way leadership communication is framed, promotion- and prevention-focused messaging could be deployed strategically to navigate these trade-offs – not because women should have to, but because understanding these dynamics can offer some agency in an unequal system. In practice, women leaders may need to strike a balance of using enough stereotypical language to maintain approval while employing sufficient counter-stereotypical language to attain their goals.

Our findings also suggest that public perception may not be a reliable basis for evaluating women in leadership roles. Perceptions undeniably matter for achieving and maintaining positions of power. However, when these are driven by gender bias rather than actual performance, they risk becoming misaligned with organisational and societal goals. In such cases, institutions should resist letting perception drive evaluation and reduce its weight in how female leaders are assessed.

In real-world data from one of the greatest crises in modern times, we found that women leaders faced a stark choice between being effective and being liked. In the context of Covid-19, for female governors, the trade-off might have been between saving constituents’ lives and saving their own careers. 

We may wish for leaders to always prioritise being effective – but our results suggest that the more leader-like a woman is, the less secure her position may be. By exposing a key dimension of this double bind, we hope to help employers and societies design and implement evaluation systems that overcome such biases. Without this, gender equality in positions of power may remain elusive.

*Sandra Matz, Columbia University; Friedrich Götz, University of British Columbia; and Ashley Martin, Stanford University.