Sunday, April 26, 2020

Surge pricing, anyone?

One social contribution that I tentatively attribute to Uber is popularisation of the concept of surge pricing. That is, we try to call an Uber and all-too-frequently get told that we have to pay a premium at this particular point in time, due to high demand. On the other hand, recent shortages of toilet paper, paracetamol, and certain foods were not accompanied by any kind of surge pricing, and the limits imposed on how much stuff you can buy, were not so effective in keeping these goods available. At this point, things have improved, although I have not been able to buy flour recently: the shortage of flour in the supermarkets I visit seems to be chronic.

Now I appreciate the objection to charging to charging a premium to allcomers, rich and poor alike, in the context of vital food and medicine. (Although, limits on purchases can also be criticised as being unfair to a single purchaser buying for a large family, or a key worker who is short of time and doesn't want to search excessively for a desired item.) In trying to advocate surge pricing, let me turn instead to possible examples less controversial, such as hairdressers and garden centres. When these establishments are allowed to reopen, it seems reasonable that they should charge a premium (temporarily). Not only do they need the money, but it would help to control a flood of customers all causing long queues and infecting each other at close quarters. To be honest, I’m not optimistic that this will happen, since they will still worry about accusations of price-gouging, plus there’s the question of how big a premium is appropriate.

An article in the Economist highlights a related problem, which is the difficulty of measuring the rate of inflation, at a time when various goods and services (whose prices get used to measure inflation) are unavailable. Coming back to flour, it may be felt that some of it (not all!) should be sold at market price, meaning one that some people will pay, but where it stays on the shelves for a few days, at least. There is a moral case against selling goods too cheaply, which is that it becomes an attempt to hide a problem — a successful attempt, if inflation cannot be measured.

Finally, the problem discussed here touches on a defect at the heart of traditional economic theory, which is the celebrated existence of “correct” prices, unaccompanied by a means of arriving at those prices. The Algorithmic Game Theory community has quite rightly worried about price discovery and its computational obstacles. But the obstacles are also social, and status quo bias plays a big part.

Tuesday, April 21, 2020

Short-term prospects for UK universities

A round-up of gloomy reading material I have been taking in.

UK universities fear huge budget holes as Chinese students stay home has worrying figures about the dependence of UK universities on students from China in particular. Note the discussion about Aberdeen University towards the end of the article. had hoped to earn £50m from overseas students this year, 20% of its overall income, but that “is now likely to fall substantially, because of the restrictions and uncertainty created by coronavirus”

How can universities climb out of the coming financial abyss? considers prospects of universities closing or merging, or obtaining bridging loans from banks or government, possibly with strings attached (such as closure of certain courses seen as not very valuable). It is suggested that some prospective UK students may decide to defer going to university, although for the ones that do, “those students may have an edge in the graduate jobs in three or four years’ time, though it’s not clear how many students, if any, will be making that wholly rational calculation.”

Universities UK: state bailout required to save institutions. Some comments added by readers made the point that senior administrative salaries ought to be cut, in the context of requesting a government bailout.

Coronavirus: universities face a harsh lesson discusses universities in the UK, USA, and Australia, with emphasis on the likely loss of income from overseas students, especially from China. The ‘overexposure to the Chinese market’ story seems to be widespread.

What will higher education look like after coronavirus?
The Office for Students will need to design and put in place a multi-billion pound stabilisation fund to prevent the collapse of scores of vulnerable English universities. Access to this fund should be subject to strict non-negotiable conditions, including the phased closure of poor-quality and low-value courses under teach-out arrangements to ensure that students can complete their studies. (This seems easy to say; harder to say what’s meant by a poor-quality and low-value course. I don’t think some smirking reference to Gender Studies does the job.) More optimistic about the long-term; argues that demand for higher education will increase in developing countries. The article has an unfortunate digression into an argument that universities that attract plenty of students, should continue to expand at the expense of ones that don’t.

Monday, January 06, 2020

On leaving the EU

For someone based in the UK who works in any kind of international market, it looks reasonable to consider how his business strategy should be affected by leaving the EU. In the case of CS theory research, this perhaps runs counter to an idealised view in which all research is global, and should not be affected by squalid political considerations. The interest inherent in any specific problem or result ought to be independent of where it was studied. On a related note, it may be felt that it’s the research topic that chooses the researcher, not the researcher who chooses the topic. On the other hand, even in CS theory, ones political environment and associated social networks may have a stronger effect than we would like to acknowledge.

When I was a graduate student, Algorithms and Computational Complexity was relatively under-represented in the UK, compared with today. The UK theory community was dominated by so-called “Euro-theory”, which at the time did not seem to exhibit obvious points of contact with algorithms research. People like me had to look west for assurance that our research was of wider interest than what was apparent in our own backyard. To further justify that west-looking approach, it was clear that the USA was, in terms of research, the undisputed world leader. Then as now, it collected the lion’s share of Nobel prizes. It had industrial research labs producing leading CS theory research, such as IBM, AT&T, and NEC, while Europe had nothing similar. For me, this sowed the seeds of a defiantly Atlanticist attitude to CS theory research — appropriate for Brexit Britain, perhaps? — that the best way to pursue high-quality research was via links with colleagues in the USA.

Fast-forward to about five years ago, and my attitude had softened. The UK’s algorithms-and-complexity community steadily grew, and is much larger than it was in the early 90’s. Travel within Europe is relatively quick and cheap, with no visa issues. The European research community became a bit of a comfort-zone, while the USA’s research ethic seemed comparatively intense and high-pressure. The perennial question of “Where’s my next STOC/FOCS paper going to come from?” has always seemed less urgent in Europe.

The EU has attempted to unify Europe’s academic research activity, which is supported by diverse governments, and gives rise to diverse complaints among European colleagues. I am reminded of the “unhappy families” quote from Anna Karenina. In an attempt to give it a bit of unity, there’s some EU funding for research, concerning which we have this criticism from Andre Geim’s Nobel prize speech: “I can offer no nice words for the EU Framework programmes which, except for the European Research Council, can be praised only by Europhobes for discrediting the whole idea of an effectively working Europe.” For my part, I recently tried to get a grant from the European Research Council but they turned me down. If I were a rational agent, I should at this point be one of Geim’s Europhobes; of course in reality things are not quite so simple. But at this point I reckon the US research funding system looks like the least worst. A recent article at Athene Donald’s blog discusses the post-EU era and the idea of a DARPA-like research agency for the UK.

So, leaving the EU looks like an opportune point to dust off the above-mentioned “Atlanticist attitude”. These days China is also becoming more important. But I hope that Europe will not give up on us, but will compete strenuously with them for our attention.