 Minusta on keskusteltu, että sieltä on tärkeää sääntä, johon on tärkeää assoissani, jossa sääntä on tärkeää assoissani. Joten järjestöä, että se on tosiaan järjestö, joka on tärkeää y-varmia, joten on tärkeää assoissani. Minä katsomaan esi-elmää 500-lipuksiä. Silloin varoittamme mennä koko ajan, ja mennään esim. Let's assume now that we want to make the claim that naming a woman as a CEO causes profitability to increase. So we can attribute this profitability difference to the women CEOs. Now why would this kind of causal claims be important? There are two reasons. First of all, causal claims allow us to make policy recommendations. For example, if we can make a valid causal claim, then we can make claims that we should increase the women CEOs. So with the more women CEOs we can make that kind of recommendations. Another important reason for making causal claims is that if we don't make a causal claim, then someone else will interpret our results causally. So when this difference was published originally in 2005, there were many discussions online on various newspapers and whether we should, based on this result, nominate more women as CEOs. Even let's take another example, there's a report that women-led companies are more profitable, it's not a unique observation to this particular study. Here's a report by Mackinsey. They showed that there's a difference between men and women-led companies. So women-led companies are more profitable and then they say that while this profitability difference doesn't allow us to make a causal claim, they nevertheless think that there should be a policy recommendation, that we have more women on boards or as CEOs. So what do you make of that? When someone reads that kind of claim that we can't make a causal claim, but nevertheless we think that there could be a good thing about more women-led companies, of course people will interpret that in order to improve your financial performance, nominate women to a CEO position. So people will make causal interpretations of your data. So you have to either make the interpretation yourself to guarantee that it's valid or you have to explicitly cause that it's not a causal relationship and you should refrain from making any policy implications like the Mackinsey people did. So how do we make a causal claim then? We have identified that there is a difference of 4.7% point and let's say that we have some way identified that that can be by chance only. So there is a consistent association that women-led companies are more profitable than men-led companies. How do we know that it's a causal effect? We have to ask the question of why is there a difference? We need different explanations to rule out, different theories to rule out alternative explanations. There is a reason for the correlation in the data, we just have to discover what is the reason. So is it the reason because women-led companies are more profitable than because of the CEO gender? Or is there some other reason that certain companies tend to be led by women and certain companies tend to be more profitable? To do that we need a theory. So the theory was a set of connected propositions or claims that explain what happens, how, when and why. So the important part in this example is that why are the return on assets between the men and women-led companies different? We need to have that white question. And a big part of doing quantitative research is to think what kind of rival or alternative explanations we have for our data. We have, for example, these explanations. We could say that women as a CEO causes firm performance, the first explanation, but it's not a direct effect. Rather it's that women facilitate top management team work and better top management team work lead to firm performance. Or it could be that smaller companies are more profitable and smaller companies are more likely to hire women. That would be an example of a spurious relationship. Or certain industries are more profitable, certain industries are more likely to hire women. For example, if we look at return on assets, mining industry, they have large assets. So the return on assets in that industry is pretty low compared to the mean of all industries. Then mining companies are more likely to be run by men than women. So that would be a reason to suspect that there's a spurious correlation. Or it could be a reverse causation. So we could say that because a company is profitable, they can afford to hire women. Or it could be that the women are better CEOs and that influences company performance. Why this kind of argument makes sense is because women are still discriminated against in CEO decisions. Only 22 out of 500 companies were led by a woman. That means that the last woman or the worst woman who gets to be a CEO in that sample is likely to be a lot better CEO than the last man. Because there are so many more men in the sample. So that would be that it's not actually that the women are better. Or there's something about being a woman that causes the company to be better. But it's a selection effect. So that's also a plausible alternative explanation. Then we need to consider which ones of these are the most relevant. Because we need to collect additional data. We need to have the variable for the CEO gender. We need to have the variable for the profitability. And then what else? We need to collect data to rule out these alternative explanations. And we would at least need the industries because that's easy to get. We would need the company sizes that's easy to get as well. These are top management team performance. That's more difficult to get. Skills more difficult to get. So it's a trade off of what is easily available and what we actually need. Then we start ruling out these alternative explanations. So we have to consider now three conditions for causality. And we can make a causal claim by showing that there is a statistical association between the cause X and the effect Y. That's the first step. The association may not be a correlation. It could be some other kind of association. But there must be an association. If cause and effect don't depend on one another, we can't make a causal claim. Then we would have to show that there is a direction of influence so that the X, the cause always comes before the Y, the effect and not the other way around. And then we have elimination of rival explanations. So how do we rule out the possibility of this correlation being an industry effect that influences the CEO selection decisions and also influences profitability? How do we know that it's not a firm size effect? There's a very simple strategy for ruling out the direction. And that is we just measure the cause before the effect. If we measure the CEO gender now and profitability the next year, it's implausible to say that profitability in the future caused the company to choose a women's CEO now. Of course there could be some profitability expectations that influence that, but that's a different thing. So we measure the cause before the effect. Elimination of the rival explanation is the hard part. We have two empirical strategies for that. And one is randomized assignment and do an experiment. In this case, we would take the 500 companies randomly assigned a randomly chosen finished man to half of the companies randomly assigned randomly chosen finished women to another half of the companies see which half is more profitable two years from now. That's of course impractical to do, but that's the experimental way. We manipulate the independent variable and then we observe the dependent variable after a delay. In practice, in business research, we do static system modeling or controlling for alternative explanations. We say that the company profitability is a function of a CEO gender plus some other things that could correlate with CEO gender. And then we test which one of those is the strongest predictor of performance taking the other plausible explanations into consideration.