Winner of the New Statesman SPERI Prize in Political Economy 2016


Monday, 19 October 2020

Why do some find the economics/health trade-off so hard to get? Because it’s like the Phillips curve.


The distinction between the short run and long run traditional Phillips curve (not the New Keynesian variety) is so ingrained in economists that it seems obvious to us. We sometimes forget that it took many years for policymakers to understand it. I think something similar is happening with the economics/health trade-off in this pandemic.


There is undoubtedly a short term trade-off between imposing greater restrictions (in extremis, lockdowns) to improve health outcomes (a good) and the negative effect that has on the economy (a bad). In much the same way as there is generally a short term trade off between expanding the economy to reduce unemployment (a good) and getting higher inflation (a bad). But in both cases that is not the end of the story.


The problem highlighted by the traditional Phillips curve model is that this extra inflation just grows over time. What may be an optimal trade-off between lower unemployment and higher inflation today will tomorrow become sub-optimal, because inflation feeds on itself and will rise again. The theory of the vertical long run Phillips curve is that there is only one level of unemployment that keeps inflation steady. Anything less than that will lead to steadily rising inflation. (Please don’t bombard me with comments about the validity of the Phillips curve, because I’m only interested in the theory here as I think it tells us something about how we are failing to deal with coronavirus.)


The trade-off between health and the economy works the same way in a pandemic. Boris Johnson got advice from SAGE about what was required to bring the virus under control, but he also knew that this advice would damage the economy. The way he put it at a news conference is that he was trying to balance the good (less infections) with the bad (hit to the economy) by doing less than SAGE asked.


Unfortunately, even if that trade-off made sense at the time, if R>1 (as it now is) the number of infections will grow over time. So a balancing act that makes sense today makes no sense tomorrow. The analogy with the Phillips curve is that there is only one level of economic restrictions which will get R=1, and that is much more restrictive than what might seem optimal in the short run if you ignore the dynamics.


The analogy can be pushed further. If you keep on trying to get unemployment below its natural rate (the rate at which inflation is steady), inflation will keep rising. After a time it is no good just stabilising inflation, because its level is much too high. Equally we now need a period of R<1 to get the number of cases down, and that means a period when economic restrictions are greater than the level that would keep R=1. The longer government puts this off, the longer and deeper any lockdown will be, and therefore the bigger the hit to the economy will eventually be.


This is why the government’s attempts to balance what SAGE suggests with the impact on the economy are not only misguided, but positively bad for the economy. Even when policymakers are convinced to impose sufficient restrictions to get R=1, because it took too long getting there infections and therefore deaths are much too high, so we will need some kind of strict lockdown to get back to something tolerable. In short, delays in getting R=1 mean a lockdown becomes inevitable, and in a lockdown with high infection numbers the economy really suffers


If analogies with the Phillips curve are not meaningful for you, just think of what the government is doing as ignoring the dynamics. Balancing the economy with health in the short run is futile, because it ignores the dynamic problem that infections will increase. Having modest economic restrictions now keep R>1, so the problem gets worse, hospitals become overwhelmed and governments are forced to do much more harm to the economy later on. The only exception is if they follow the anti-lockdown crusade, which means letting hundreds of thousands of people die. It is the equivalent, using our Phillips curve analogy, to generating hyperinflation. 


Policymakers in many countries, not just the UK, have failed to grasp the nature of this dynamic problem. In the first wave they typically imposed the harshest of lockdowns to be sure of getting R well below one. They gradually relaxed lockdown, achieving in the summer a position where R was close to one. After that three kinds of mistakes occurred.


The first was the one described above. In the UK, for example, even with cases beginning to rise in August, policymakers tried to encourage people to mix in restaurants and stop working at home. (Of course they came under intense pressure from sectors that had been closed down to relax prematurely.) The second mistake is more speculative. As time goes on, people relax and forget what they are supposed to do and what they are not supposed to do. To use the Phillips curve analogy, in both cases the natural rate begins to rise.


But by far the biggest mistake was the third. Everyone wanted schools to restart in September, and the government thought universities should restart in October. Those are two large increases in social mixing, which mean that to keep R at one some further restrictions were required. In terms of the Phillips curve there was an upward shift in the natural rate. Policymakers failed to understand that, and so the second wave began. Looking at it this way, the second wave was completely preventable if policymakers had recognised that shift and taken appropriate action, but instead they just crossed their fingers.


If we want schools and maybe universities to stay open, does that mean we have to keep pubs and restaurants permanently closed to avoid a repeat of what happened this Autumn? The evidence from countries like South Korea is not necessarily. A really good test, trace and isolate (TTI) infrastructure can exert strong downward pressure on R, and can handle the occasional spike before we need to resort to lockdown. Using the Phillips curve analogy, TTI puts downward pressure on the natural rate.


The biggest failure of this government and others over the summer was to be content with TTI operations that were unable to handle spikes. The failures of the UK’s centralised and privatised operation are well known, and have been described repeatedly in earlier posts. A good TTI system requires being able to get tests locally on demand and having the results back within 24 hours for around 80% or more of people. But it also requires people to isolate after being told they might have the virus, and the UK has put little effort into ensuring people are willing and able to do that. Indeed the Cummings episode did the opposite. 


The message of this post is that anyone who talks of balancing health against the economy in this pandemic is being very short-sighted. Unfortunately many politicians are naturally short-sighted, and are used to balancing pressures from different directions, which is why we have seen cases rise in many countries around the world. The irony is that taking firm action to suppress the virus to very low levels is popular, and as we have seen very recently will reward any politician with the foresight to think about health rather than the economy.  




No comments:

Post a comment

Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.