The Dangers of Low Expectations

Daniel Finkelstein wrote in The Times yesterday here that it is important that politicians level with the public on the difficult trade offs involved in a lockdown, and what easing it off might in practice mean. You can read his piece here. There are some strong arguments made, but I do have one big problem with the article. It does seem to underplay the potential a mass testing and tracing regime might have to keep transmission rates (R) below, or roughly at 1. In particular, he does not explicitly mention the fact that South Korea seems to have done precisely this, to great effect. Note that this does not imply that a lockdown can be relaxed any time soon, as this strategy does not work if the number of cases greatly exceeds testing capacity, as Julia Gog explains here. In practice this means both bringing down case numbers in the mean time through more quite drastic social distancing measures (a lockdown) while developing testing capacity.

But I think there is a broader point here: while some degree of understanding of the complexities and difficulties of the situation is important, and it isn't useful for the public to just blame politicians for a natural catastrophe (particularly if some level of goodwill is required for mass compliance to lockdown measures), it's important that expectations of the government aren't set too low either. If it is indeed possible, or there is a good chance it is possible to ease the lockdown at some point in the future without another major outbreak if testing capacity is significantly increased, it is vital that the government does this. And if the government is failing to do this even though this is possible (albeit logistically difficult), it is important this fact is in the public domain and proper scrutiny is applied (which is why I argued here, that Starmer was right to ask questions about an exit strategy).

Now it might well turn out that such a strategy, while workable from as a purely epidemiological perspective, is not logistically workable in the UK at this late stage. But it's important that this isn't just left for the government to decide in secret. As with the initial delay of social distancing measures, it might well turn out that choices are made on fairly flimsy reasoning. And it would be a tragedy if here too the road not taken led to much better place.

This is just one example of the dangers of setting expectations too low. But there are others too, most notably in the provision of PPE to care and hospital workers. While the logistical challenges are undoubtedly immense, it cannot suffice to simply let any failures go unchallenged when other countries seem to manage much better. And if, as may well be the case, it turns out that mass adoption of mask wearing also turns out to be a crucial tool for allowing the lockdown to ease, it's vital that a strategy is in place to ensure the mass production and distribution of masks. A situation in which the government is allowed to brush aside problems as resulting purely from the natural features of the virus would be a dangerous one. So of course, we must be reasonable with our expectations and not try underestimate the logistical challenges of this pandemic. But expectations let's not expect any less of our government than what we expect of other wealthy countries.

Some thoughts on price and value

The relationship, or lack thereof, between wages and the value of work has seemed a particularly pertinent question lately. Many of the workers whose salaries are below the threshold for tier 2 visas are the very same workers deemed 'essential' during lockdown.  There are some quite familiar arguments as to why wages often do not seem to reflect 'value'; these include different levels bargaining power, positive and negative social externalities of some work, and the impact of established norms and expectations about what kind of wage or salary a certain profession should receive.But I'd be tempted to go further than this: even if we were to imagine a world with equality of bargaining power and without significant economic externalities, we should not expect wages to approximate some kind of absolute value distinct from price, simply because value, as a quantifiable or quantified entity is itself an emergent phenomenon of systems of exchange and distribution.

Imagine a simple exchange relationship between two individuals. Individual A gathers food, and individual B plays music to entertain A. A easily can gather enough food for both individuals to survive, so there is really no need for them both to do so. Perhaps we imagine this pair to be mercantile enough to come to some kind of arrangement: B plays music for A for a certain amount of time a day in exchange for food. You could imagine a ratio between the two: 1 meal for 1 hour of music. Perhaps some days A doesn't feel like listening to music, some days B is less hungry than others, and instead sticking to some bartering ratio, they instead denominate both things in terms of some third item both have scarce supplies of it takes some effort to replenish.

We could then start to think about something called 'price', this being the ratio at which each item is exchanged for this third, common denominator. But what would this 'price' have to do with 'value' in an absolute sense? Life might be boring without music, and life couldn't continue without food. These are different kinds of value, for which there is no quantifiable means of comparing, except in terms of exchange ratios for which they are exchanged, or price. And it is fine to refer to those quantities associated with each thing, so long as we remember that all we are really describing by that quantification is volumes of exchange. This does not mean that this ratio would not be informed by value: if A didn't like B's music, A wouldn't be happy with this relationship. And the ratio of exchanges would have to be such that the volume exchanged did not exceed what A or B was willing to part with in exchange for what the other offers. But the ratio doesn't quantify how much better off A and B are for carrying out these exchanges, just that they are better off, and that they occur in a ratio that both are willing to agree to.

If we extend this to real world transactions, the fact that one person is willing to pay £10 for something might show that they prefer having that thing than having £10, and by extension, other things they could have spent that £10 on. But it doesn't tell us much about the value of that preference being satisfied in relation to some other person's preference being satisfied of an equal monetary value (£10 extra spending money is obviously worth more to someone who is broke and might need the money to pay a landlord than someone who is very wealthy, but can we quantify how much more?).

But does this mean prices tell us nothing about value? No, they do tell us something. They you what transactions people are willing to enter into. They allow you to make extraordinary predictions about what you can expect of a vast array of economic interactions with agents you know little, or nothing, about. This is in and of itself of great value- it makes societies 'legible' to observers, and allows for the coordination of economic activities and the division of labour without central planning. But this is not the same as reflecting 'value' in an every day sense. Prices, and by extension wages, do not allow for the worth different individuals attach to things to quantified by a universal common denominator. Rather, they allow us to predict what agents are willing to do for them, and the ratio in which these actions will occur. This is true before we even get into thinking about inequalities of bargaining power in the labour market.

Coronavirus: three deadly misconceptions

Most crises provoke their fair share of bad ideas, and the Coronavirus pandemic is no exception. Three misconceptions seem to stand out in particular, and I thought I'd try and spell out why these ideas didn't work.

1. Lockdowns need to be 'timed' correctly to have the maximum impact.

This was the go to justification for the delay of implementing social distancing measures in the UK. Superficially, there appeared to be some logic to it. If lockdowns are time limited, it might make sense to ensure they are in operation during the period where the infection curve peaks. The logic is alluring until you realise that when that peak occurs is itself determined by the time a lockdown is started. So far as we can tell from Italy, Spain and China, this is roughly 3 weeks later. This time-lag is independent of the level of spread before the start of a lockdown (assuming, that is, some existing community transmission). The only situation in a delay would reduce the burden on a health system is the one in which it was merely a one-off act designed to mitigate Covid's 'natural' spread to a large majority of the population, in other words where we had given up on actually preventing people from getting the disease.

2. There is a trade-off between suppressing Covid and the economy

This one also sounds pretty intuitive, as social distancing measures obviously reduce economic activity. Indeed, if they didn't, they probably wouldn't be working. People can't go to restaurants, bars, pubs, cinemas, etc. Most can't go to work, and only some of this can effectively be shifted to work done at home. Clearly this is going to cause substantial economic disruption. What's less clear is that the alternative would prevent much of this disruption from occurring anyway. High rates of infection would inevitably mean large numbers of people dropping out (hopefully temporarily) of the workforce, and a large proportion of discretionary spending would likely fall regardless, as people worried about catching Covid going out and feared a loss of income. What lockdowns hope to achieve is limit this period of disruption to a shorter time frame, after which restrictions can hopefully be eased off when accompanied by greater levels of testing. And if you are going to do a lockdown, the earlier the better in economic terms, as this means it can be relaxed at an earlier stage. Moreover, in the long term, covid suppression will likely lead to a swifter recovery, as a larger proportion of the workforce will have been protected and less adjusting will be necessary.

3. Test and trace can only work if we manage to find and isolate every single person infected with coronavirus.

If this were the case, test and trace programs would have a hopeless task. It is simply inconceivable that health care workers could manage to trace and track every single person infected with coronavirus, and a strategy based on that aim would be futile. But this is not what test and trace programs aim to achieve. Rather, they aim to find a large proportion of people who have contracted coronavirus at an early stage and get those people to isolate themselves, thereby reducing the average rate of transmission. If this is reduced to a number below 1, the numbers will exponentially decline regardless of the fact that many people slip through the net. And if average rates infections are kept just a little above 1, increase can be kept sufficiently low to allow for significant periods of relaxation of lockdown rules. What effectively test and trace aims to achieve is to render Covid a less infectious disease.