Control

In a new blog post, Adam Curtis discusses the problems with economics and the difficulty of treating the socio-political sphere as if it was a machine. This repeats arguments advanced in his series The Trap and Pandora's Box - that the Thatcherite/laisser-faire (and even Blairite) project of marketizing all aspects of public life (education, the health service, social services, etc) is not truly a freeing, but the construction of an incredibly complex network of measurements, penalties, and rewards, all of which are susceptible to manipulation by the people involved (politicians, managers, staff) who gain rewards or punishments depending on the data fed into the system; inspired by cybernetics, feedback engineering, and systems theory.

One good example is British railways. Before privatisation, almost the entire railway system was run by a single entity, British Rail, which combined passenger and freight operations, rolling stock, track, stations, safety, planning, staffing, and all other aspects of running a railway. The 1993 railways act broke British Rail into over 100 companies: the track is controlled by one company, which contracts out maintenance, and train services are put on by a number of different companies, who lease rolling stock from other companies. A hugely complex bureaucracy is required to adjudicate between all these organisations: train operators are penalised for unpunctual trains, but have found tricks like cancelling trains or adjusting timetables that can be used to manipulate the metrics. Subsidies doubled in real terms from 1994 (just before privatisation) to 2005, with no increase in punctuality or quality of service. Wikipedia has a good summary.

As a result you do not have a free-market system where people are free to innovate and provide new services; you have a huge costly bureaucracy which constrains people into maximising semi-arbitrary metrics rather than competing directly for customers. Of course, with a railway system, it is hard to allow two operators to run rival services and compete against each other, but this reduces the choices to state ownership or unregulated monopoly (with each company building their own lines, as happened in the 19th century). The present system is neither free enterprise nor state control, but the worst of both worlds. The toughest question is why such a system exists: when railway privatisation was planned, some people wanted BR privatised as 1 or 4 companies, but the Adam Smith Institute, a group of right-wing semi-libertarian ideologues pushed for the bizarre system now in place. Their reasoning remains opaque: it cannot be motivated by naive laisser-faire economics, but can only be a product, as Curtis suggests, of strange counter-intuitive ideas lifted mistakenly from cybernetics about how a complex system of data flows (and money flows) can produce a self-regulatory system, little considering the effects on either the cost of the system or the services it provides.