True or false: female leaders are doing better in the Covid-19 crisis

I was originally posed this question following the Forbes story “What Do Countries With The Best Coronavirus Responses Have In Common? Women Leaders“, published on 13th April. It called out seven countries with women leaders – Germany, Taiwan, New Zealand, Iceland, Finland, Norway and Denmark – who all appeared to have done better in the crisis than most countries.

I took a look at the data with a healthy skepticism that leader gender would be a factor in coronavirus response performance. Even though the gender differences might have a positive influence in such a crisis situation, I was also aware that each country’s response has been informed by a large team of experts (scientific, health, political, economic etc.) who are – presumably – a mix of genders.

The result surprised me, and so I have updated the analysis with the latest data and reproduced this below.

The analysis uses the Council for Women Leaders as the basis to determine which countries have a current woman leader, and then assumed all the others are men.

The data used for comparison is the average daily deaths per million of population since the death rate in each country reached 10 per million. Therefore, countries that have never reached 10 deaths per million are also excluded (which excludes, therefore, New Zealand).

For similar reasons, countries with a population of less than 1 million were also excluded on the basis that the sample size was too small, and especially given the relatively small number of countries with a woman leader, could have led to significant distortions.

This left just six countries with women leaders in the data set: Denmark, Estonia, Germany, Norway, Romania and Serbia. This data is presented below:

This data clearly supports the hypothesis that countries with female leaders have experienced fewer deaths per million and a lower growth rate in deaths.

So we took a closer look at the two most commonly cited explanations for differences in country-to-country comparison: how big the population is, and how dense the population is (per km, not IQ!).

The population explanation hypothesis

This might go something like this:

Women leaders tend to lead smaller countries, and so that is why they seem to do better – because it is somehow easier.

Here is the same graph as above but now with the average population of the countries included shown in dashed lines based on 2018 population data.

So yes, overall, men leaders tend to lead larger populations.

Does the total population really make a difference though? I’ve not seen or done analysis on that questions, but it is fair to say that some of the largest countries – China and US for example – actually show a lower death per million rate. So perhaps this actually suggests the male leaders should be doing better?

The population density explanation hypothesis

This goes something like this:

The virus spreads by person-to-person infection, so higher population density would make it more difficult to social distance and isolate, and therefore we might expect to see a correlation between higher population density and higher rates of infection and death per million of population.

Same graph again but this time showing average population density in the dashed lines.Data here is sourced from the Wikipedia page which in turn sources the most recent population density data from a large number of sources. See https://en.wikipedia.org/…/List_of_countries_and… for details.

The countries meeting the criteria to be included show that male leaders do indeed preside over more densely populated countries – approximately 30% higher density than the average for female leaders.

The reason it changes over time here is because fewer countries have such a long history since reaching the 10 deaths per million threshold.

It is perhaps no surprise, then, that those that do have this longer history (i.e. saw outbreaks earlier) are the ones that have higher population densities, although this correlation is significant only for male leaders, despite there being a smaller differential between population density of countries with male or female leaders in general.

Conclusion

It does indeed seem pretty clear that there is a correlation between female leadership and lower deaths from Covid-19, although this is based on only six qualifying countries with female leaders.

The question, though, is this correlation causative? In other words, do women leaders really cause fewer deaths, or is there some other underlying cause?

A recent meme showed the strong correlation between forest fires and ice cream sales. Just because there are more fires at the same time more ice creams are sold, does not mean that the sale of ice cream causes forest fires (or vice versa!). Both in fact are driven by the temperature – an independent causal factor.

It might be that the female leadership has directly caused fewer deaths. Perhaps because women tend to be more compassionate, more level-headed in a crisis, and less likely to care about what people think about them.

Or maybe there is an independent causal factor for our correlation between female leaders and fewer deaths – perhaps societies that are elect female leaders are more likely to listen to their leader’s advice and follow the rules?

Either way, at this time, be grateful if you have a female leader – statistically it’s likely you’re faring this storm better than those who do not.