I read through Deidre McCloskey’s “The Secret Sins of Economics” last night. The two sins she places on economists are 1. Our theories and their results, because they are based on various assumptions, don't really show concrete results. When one changes the assumptions, the results change... so the theories aren't that useful, per-se. 2. Economists focus too much on statistical significance when doing their empirical analyses. These two sins, she argues, causes us not to focus enough on answering "how much" questions.
For example, suppose I take a sample of 100 students and examine their attendance and their performance on the final exam in my macro class. Suppose I find that a 1 person increase in student attendance over a semester in my macro class is correlated with a 0.001 increase in the average score on the final exam (test scores are out of 100). Further suppose that number is significantly different from 0. That's all well and good, but that doesn't "prove" anything.
First, the true value of my population parameter - the correlation between attendance and average test scores is still unknown. I don't know what the true population value is. It may actually be 0, but I got misleading results - I was unlucky. Merely reporting statistical significance and making conclusions solely on it misses the point - we have evidence that suggests something is likely in the population, but we don’t *know* for sure.
Second, 0.001 is not that big of a number. How much effort should a teacher take to improve attendance?
Third, we haven't even said anything about cause and effect. Statistical significance, by itself, does not tell us this. Theory, by itself, does not tell us this.
Good scientific inquiry needs to have theory and empirics - it's not necessary for one person to do everything, she argues, but it is necessary for the body of literature to have both components.
She's right - in general, but I don't think she's right to claim that these are the sins of economics, but these are things that *economists* do. Her point, from what I gather, is that economists need to focus on the whole package of scientific inquiry in order to root out what is. But we do not.