Economics by definition requires maths and econometrics

At the end of his interview with the FT, Jean Tirole makes an unsurprisingly but frustrating statement.  He not only argues that economics is about mathematical modeling and statistical testing, but he then (implicitly) equates this to being a science.  There is little doubt that the type of modelling most economists forces clarity and helps to establish causal chains.  Yet the idea that this is the only way we can discuss economic problems continues to be really questionable for two reasons. First, you can be clear about your causal model with words.  Tirole recognizes that economic ideas have to be translated into words to get to the general public but, following his reasoning, you will lose clarity in that process.  Second, there are many economic issues (e.g. the role of the elite) where there is not enough data available and there are "qualitative" processes to consider... surely we need to consider those processes and surely they are economic in nature.