This video has been doing the rounds recently, and deservedly so:
If you aren’t aware of the Antikythera mechanism, here’s Wikipedia:
The Antikythera mechanism (/ˌæntɪkɪˈθɪərə/ AN-tih-kih-THEER-ə) is an ancient Greek hand-powered orrery, described as the oldest example of an analogue computer,[1] used to predict astronomical positions and eclipses decades in advance.[2][3][4] It could also be used to track the four-year cycle of athletic games which was similar to an Olympiad, the cycle of the ancient Olympic Games.[5][6][7]
https://en.wikipedia.org/wiki/Antikythera_mechanism
This artefact was among wreckage retrieved from a shipwreck off the coast of the Greek island Antikythera in 1901.
The video is a fascinating walk through very modern attempts at recreating the Antikythera mechanism – although the recreation does raise a fascinating question, as the video mentions. How the hell could the ancient Greeks have possibly got this thing going back then? Watch the video to understand how complex this thing is!
And while watching the video was endlessly fascinating, I couldn’t help but thing of one little thing: the model is wrong.
Please, don’t misunderstand me. If, like me, you are a fan of learning about new things, and are fond of spending half your day staring out the window thinking about how they could have possibly done this back then, then you will know that I’m not trying to run the mechanism down. All I’m saying is that the model is wrong in the sense that we now know that the Earth is not at the centre of the solar system. The sun is.
That, in fact, is what makes the mechanism even more awesome. In spite of a wrong assumption, they had a working model of the solar system in the sense that they could predict a variety of astronomical events with near perfect accuracy.
The thing is, I think you could say the same thing about a lot of economic models. In spite of wrong assumptions, we economists too have working models of the economy in the sense that we can predict a variety of economic events with near perfect accuracy, until one day, suddenly, we can’t.
And when models don’t work, we take a look at the specifications of the model, we collect better data and we wonder if there’s something wrong with the world. All of which helps, but every now and then, perhaps it makes sense to ask if our core assumptions could be wrong.
A useful thing (for me, at any rate) to think about the next time I teach, say, the IS-LM model.
EconForEverybody 