Take any policy and relevant outcomes over time. It could be gun control vs crime rate, or education expenditures vs test scores.
What if they both behave randomly. What will the data show?
So 10 years of data over 50 states - The max and min correlations wind up being around +/- 0.7. So even if policies have no impact on society, there will always be some data supporting any hypothesis.
The implicit test of hypotheses is whether they sound plausible. Of course no one believes that reduced numbers of pirates cause global warming, but they will believe that an increase in gun control might lead to less crime.
Doing a “rationale check” on ourselves is fine from the Austrian position. AE is a priori and can therefore be validated rationally. However if the proposition is not a priori, as in the gun control case, reasoning cannot help us. Indeed, from the fact that we can argue the effects of gun control both ways illustrates this.
tl;dr - I think Austrians need more demonstrations of how statistics proves nothing.
I think they need more examples of how statistics can illustrate but not form theory. Because, unfortunately enough, people are wired to look for evidence and patterns and such. It’s what we do more or less. Trying to avoid it means missing a crucial means of getting a message across to people.
In which case though, who gives a shit, about theory or history? Posit magic elves are responsible for everything or anything else for that matter, because then Milton Friedman is right and predictive power and reliability are all that count. This part of Austrianism always annoys me. Yes, there are epistimological limits on what we can know. Yes, we can’t run controlled experiments like the hard sciences. That doesn’t mean there is a complete disconnect between theory and history though. A correct theory should demonstrate some superiority on some level to incorrect theories. There should be some predictive power. There should be more congruance with history as facts are accumulated about that history, not less. It should more consistently generate explanations of historical data. It should recommend areas of historical investigation and what we might find.
The problem with a strict separation of theory and history is that it leaves you with the ivory tower or quackery and nothing in between. Theory is only as good as its applicability to the world. A theory that has no predictive power for the future and no explanatory power for the past is useless and redundant. I think Austrians need to do what so many main stream economists have done, cut the bullshit and start inserting their value judgements. There’s no point in ‘debating’ a smug prick like Krugman who will stab you in the back at the first opportunity if you don’t arm yourself with at least as big a knife and the means to use it. And then, get dirty. Argue that the theory presents a clearer, more accurate picture of the Great Depression, or this or that other time in history. Argue the strengths of the Austrian explanation vs others.
They can and do. Theory can’t come from nothing. The problem is cherry picking, which means looking at only the things you want to correlate within the scope that helps find the desired pattern.
“They can and do. Theory can’t come from nothing. The problem is cherry picking, which means looking at only the things you want to correlate within the scope that helps find the desired pattern.” - Caley McKibbon
According to you. Not according to me. All data collection is ‘cherry picking’ because no matter what the intentions of skills of the data collector or statistician some subjective judgement about what to include vs exclude and the relative significance or weight of what is included will be present. The economy viewed as a system has varying degrees of being coupled at any given time and has no constant relationships between its factors. trying to draw a theory based on such data is pointless.