On Experts (Long Post)

There’s been quite a bit made lately of the fact that Michael Gove has said that that “British people are sick of experts”. Of course, the reaction has been swift and ruthless, and as always, extended to the point of absurdity. (I’ve always found it interesting that people who attempt to point out what they consider absurd often go so far as to themselves become absurd.)

Reactions pointed out that experts are experts because they spend a lot of time doing what they do and studying what they do, and generally living what they do. I consider myself an expert at computer security, and on computers in general, because, well, I’ve spent a lot of time doing what I do and studying what I do, and generally living what I do. This is a rational basis for proving expertise, and mostly when experts speak they should be listened to because they know their science.

The keyword there is science.

Economics is not a science.
Yes, it may be portrayed as a science, and many people think they can prove that it is, and while there are generally accepted principles of economics, but there are no falsifiable experiments in economics, and certainly no way to even conduct experiments on the scales of economies, so anyone who tells you that they are an “expert on the economy” is merely an expert at reducing something to what has happened before and may or many not happen again. Real scientists can not only tell you what happened before but can tell you what will happen again because their work literally proves that sort of thing.

Let’s contrast computer science and economics. If I want to study how a computer works, I can, as I’ve done in the past, take it apart and research the parts. I can go to the Intel plant and watch chips fabricated, and having learned the concepts behind transistors, logic gates and the like, apply actual experimentation to how things work with computers and get results based on hypotheses. My experiments will not only explain how things occur, they can be used to reproduce events occurring after – to regularly predict what will happen. Given certain inputs, I know that I will get the same outputs consistently. Furthermore, other people can take those same questions, use different equipment, and with the same inputs, obtain the same outputs. They will get the same results by repeating my experiment in their own lab.

No such thing is possible with economics.

Can you take apart an economy? You might be able to study sections of it, and conduct polls and/or behavioural analyses of certain actors in the economy, but there is no way to conduct experiments. Furthermore, the concept of an economy is based on the nebulous idea of a “rational actor” and economists have been finding to their dismay for some time that no such rational actor exists. (Some of the more interesting work to come out of economics research has to do with what things look like when you presume actors are partially or not-at-all rational.) Personally, as an agorist, I think that leaving the market to work as it should will bring us to the glorious agorist future and things will be as they should without any of this “elitist” nonsense or people who think they can plan for us and get it horribly wrong.

Wait, can’t you create computer models of economies?

Yes, you can do that, but the economic models will only be as good as the data from which they are created, which as I’ve already mentioned, depends on a whole host of assumptions and other things that can’t readily be tested. Again, the “units” in the economic models need to adequately reflect the “units” (people) in the economy and that is a difficult task at best.

What about science’s flip-flopping nature? I thought eggs were good, but then they were bad, but now they’re good again….
This is another issue that gets brought into the mix. Legitimate experts, scientists, are constantly learning new things. So with respect to the old things, they may be wrong. The history of many things (eggs, fat, cholesterol, smoking) has been at one time good for you (cigarettes were originally marketed for their stress-reducing capabilities) and at other times bad for you. Ignoring a presumption that the non-expert dismisses experts for this seeming back-and-forth (though it obviously happens, part of what Gove was talking about) there is a distinct difference here. This is the progress of science correcting previous errors. This is falsifiable experimentation brining to light new information where potential new technology or something else has given us a better insight into occurrences.

Wait, but economic experts exist though…
Taking aside for a moment the difficulty of economics as a science, let’s look at the track record of so-called “economic experts”. Even without a solid basis for “sciencing” the economy, surely some people must be better at analysing patterns than others?

That’s precisely the case. Only none of them were the commonly-referred-to-as-experts.

Who predicted the 2008 financial crisis? Not one of the financial experts. People like Peter Schiff saw it coming, and people like Michael Burry figured it out, but the experts, including the economics teams of nation states and “experts” at central banks had no clue until it literally collapsed on them. Ben Bernanke said in 2007 that the US economy would strengthen in 2008, and this in July of 2007 – only six months before 2008 and only nine months before Lehman and Bear Stearns.

Furthermore, the people who regularly predict things well are often derided, while the “experts” (Bernanke and his ilk like Paul Krugman) are considered over and over again to be the paragons of truth while consistently getting things wrong. They are rewarded for it! Bernanke gets hundreds of thousands of dollars per speech, and it can be conclusively shown that he literally had no idea what was coming until if landed in his lap.

Or how about people, such as George Osborne, the British Chancellor of the Exchequer, who literally predicted a year of “DIY Recession” if people voted for Brexit? This man, tomorrow, will completely reverse course and tell the world that there is nothing to worry about in an attempt to calm the markets. Was he lying then or is he lying now? The answer is clearly that he was lying then, because while there will be some short-term pain, the reality is more likely to be that things will be fine after a month or two once the news is no longer such a shock, and after other countries start down the same path. Clearly, this “expert” who is qualified enough to be the Chancellor of the Exchequer in the British government, is essentially proven a liar either then or now. Yet we’re to take his expertise at face value when he was willing to sell it down the river for political expedience?

What about outside economics?
One of the tweets I posted on the evening of the vote (the 23rd into the 24th) is that the real losers of the evening were the British pollsters. The British pollsters, who presumably would be considered “experts” had gone and gotten it completely wrong about the Remain side winning, and were predicting a close but foregone conclusion. They were wrong. Yes, experts can sometimes get it wrong, but they were wrong consistently. They were wrong in the previous general election, and from what I understand don’t have a great track record for anything. Nate Silver, made famous for his new methods where statistical methods and polling are concerned, was completely wrong on the rise of Donald Trump, giving him only a 2% chance at the outset of the political campaign start in the US.

I’m not saying that experts aren’t experts because they’re sometimes wrong. You can be an expert and still get things wrong. The problem becomes when you get them consistently wrong – over a period of time. This is the very definition of the opposite of expert. Yet people who do this, whether it’s because there’s no alternative or the media need to feed the beast that is the 24-hour news cycle, continually hype these people and their predictions. (I’d point out that the media mostly wouldn’t know science if it punched them, so there’s that problem too, but that’s another post.)

The end result of this is that it comes down to science. If something is a science, it should eventually bring us to a result that is consistent and correct. This relies on experimentation and the reproducibility of results. There are many areas that may seem like sciences, but do not actually include the underlying mechanisms for testing and consistent result reporting based on those tests. Economics is one of these, and the idea that you can be sick of these “experts” is legitimate.