Why shouldn't we decide every year where the rail tracks should go based on passenger demand? And dismantle/build trains and stations every year?

Because rail travel, just like universities, depends on infrastructures - material and social. Let's do the "thought experiment". 1/ https://twitter.com/Phil_Baty/status/1265896483565469697

The basic fallacy committed here is the assumption that education follows short term supply-demand fluctuations. It doesn't, because it is mostly capital and infrastructure dependent and requires long-term planning. You cannot build a university program overnight. 2/

And when we say "infrastructure" we don't only mean buildings or equipment. I research infrastructures in the past and the best way to explain their durability is through a network model. Those networks are slow to emerge, but when they coalesce they tend to be robust. 3/

In 2007 the average length of a PhD was 8.2 years. In some fields it can take up to 13 years. And this does not include postdocs and previous qualifications. But even then, to become a lecturer you need to publish and network and... be lucky. 4/

Those cycles of social infrastructures are way longer than short supply and demand swings. Hate to break it to you, but education requires long-term central planning to be successful. If you switch to short-term you will run out of social resources in a few years. 5/

Because the most important infrastructures of universities are not (surprise) the shiny buildings and glass-covered business schools, but the people and the connections they build. And by resetting those networks every year you end up with no way to replenish them. 6/

This is why university education is more like rail travel than supermarket shopping (while being, of course, special in its own way): you need to model long-time trends and shape them, because you are first and foremost building a network that needs time to coalesce. 7/

Those cheap, shiny, disposable adjuncts are not carrots, they don't grow in one season. If you can't take the central planning argument think this: you will be throwing away the money you invested in training a lecturer for 8 years who then will not be needed. 8/

And after the cycle will finish you will... run out of the disposable adjuncts and will have to spend a lot of money to restart it, because you did not pay attention to the fact that you are building a network and not a supermarket. 9/

This also (partly) explains the influence of the Roman Empire on early medieval Britain but you will have to take my word on it until my book gets published. 10/

The 2007 data was taken from this article:

https://www.nytimes.com/2007/10/03/education/03education.html

11/

Just wanted to add that there is an even deeper problem with this THE article: it advocates a "thought experiment" that entertains the idea of condemning hundreds of thousands to perpetual instability and economic anxiety as a *solution*. Let that sink in. 12/

Btw. one can argue that the best option for higher education is the Lange-Lerner model in which the state owns the means of production (and thus investment) and the market allocates goods but it would have to be broadened to include network value. Might blog about it. 13/