28/08/2012 § 6 Comments
Due to the initiative and effort of Eric Crampton, there is now an economics blog on Sciblogs. It’s called The Dismal Science and will pull posts from several excellent New Zealand economics blogs. Thanks, Eric, and yay us.
It has interesting and Darwinian implications for climate change. For example, the Kapiti Coast has just informed residents of low-lying areas that their properties will be under water if the sea level rises.
I’ve seen similar maps for several areas in the lower North Island — Kapiti, Wellington, the Hutt Valley. They are informative, but they are also just a question of physics and maths. If your house is 1 metre above sea level and the sea rises 2 metres, given that 1 < 2, it should be pretty easy to work out the consequences. It gets complicated, of course, with more precise measurements and predictions, and accounting, too, for storm surges and so on. But the essence is that water seeks its own level and some numbers are smaller than bigger numbers.
For that reason, the reaction of one affected resident was curious:
The sea-risk report is ‘just another thing to bamboozle residents’, say Elizabeth and Terence O’Brien, from Raumati.
These individuals have decided that the report isn’t about preparing and planning and informing, it is about ‘bamboozling’. The nice thing about methodological individualism is that we can work with that. There is no need to argue, cajole, or convince. I can look at low-lying beachfront property and wonder how long it will last; they can see it as a great investment opportunity. We can both be pleased in the present with our accurate foresight. Then, in the future, we will individually bear the consequences of our beliefs and choices.
Obviously, the whole topic of climate change is more complex. But, in the end, individuals need to make decisions about their own circumstances, using the information and theories they think are most germane. And that’s pretty interesting to study.
13/08/2012 § 3 Comments
Last week, I was reading a report on the economic impacts of climate change on New Zealand’s agricultural sector. I have written one or two such reports myself and try to keep up with other people’s research. It struck me that there’s a huge difference between these academic or policy documents and the public debate in the popular media.
The academic and policy documents make it very clear that climate change is neither good nor bad for New Zealand. The main variables are temperature, rainfall, and distribution of extreme events. That last variable is difficult to analyse, so it generally comes down to the first two. The impacts are quite different by region. In some areas, a little more warmth and rain make everything grow better. In other areas, reduced rainfall makes farming more difficult.
Climate change researchers try to grapple with this variation with a series of models built on models. The main IPCC climate change scenarios are analysed with global climate models. The modellers feed in a set of assumptions and see what happens to different regions, but these regions are defined fairly broadly. Then, those modellers feed in a new set of assumptions to see what happens.
New Zealand researchers then have to take the international results and downscale them to account for local variations. That’s another set of modelling. The downscale modelling has to choose which internal climate change model and which scenario from that model to use as a basis. For example, a lot of New Zealand work relies on the HadCM3 (Hadley Centre Coupled Model v3) international model, and tends to start with the ‘central’ scenarios as opposed to the extreme ones.
There is also the very important question of how people will adapt. New Zealand is generally temperate and well supplied with water. Other countries are very successful with higher temperatures and less rainfall, so it’s clear that we could make a few changes and still live fairly well.
The economic research is essentially a large number of chained assumptions from scenario to international model to downscaling model to economic model. At each step along the way, we could have informed and thoughtful debates: how we select likely scenarios, which international model is best for New Zealand, how downscaling captures local conditions, how weather affects production, the extent and speed of economic adaptation, etc.
Instead, the popular media seems stuck in a loop. Is climate change happening? Are humans to blame? When will we run out of time?
It’s climate change as soap opera. Which suggests we’ll be seeing the same storylines in another 50 years, just with a new cast of characters.
03/07/2012 § 9 Comments
The Government announced that the ETS will be modified. The changes hold off on implementing some parts of the ETS, and most notably keep agriculture out for a few years.
What can we make of this? It seems like a game theory approach might help. Imagine a 2×2 game. The players are NZ and the Rest of the World, and they can choose to implement an ETS or not. Here’s a possible pay-off matrix:
|NZ||Yes ETS||No ETS|
|No ETS||-10, 0||0,-1000|
What do these values mean? If everyone implements an ETS, then we just tick along as we are and there are no positive or negative consequences: 0,0. This is, apparently, a rose-tinted view of the future, but it will work for the moment. If no one implements an ETS, then there are serious consequences. New Zealand, however, doesn’t suffer too badly (there’s some science behind that). There are some unders and overs, but we’ll call it even. In that case (bottom right-hand cell): 0, -1000.
Then there are the two cells in which NZ and ROW take different approaches. If NZ implements and no one else does, then we pay a small cost but the world still suffers the same consequences: -5, -1000. In the reverse situation, NZ pays a reputational cost but the world turns out okay: -10,0.
Now, let’s add some probabilities. What’s the probability that the ROW will implement an ETS? Ten per cent? If so, the value to NZ of adopting an ETS is (-5 x 0.9) = -4.5, while the value of stalling is (-10 x 0.1) = -1. So, yes, stalling is worthwhile.
Given this payoff matrix, stalling is worthwhile until the probability of the ROW implementing an ETS rise above 33%. That is, it doesn’t have to be likely to happen — we just need a better than 1-in-3 chance.
The Greens are predictably exercised about the Government’s decision. In their statement, they quote the co-leader:
“New Zealand is missing the chance to protect and enhance its $20 billion clean, green brand,” said Mrs Turei.
This may be true — I don’t know. The two things that occur to me are:
- given the lack of action in other countries, I’m not convinced that greenhouse gas emission activism by New Zealand will be noticed or valued. If the voters overseas don’t care enough, why will the consumers overseas care?
- we can add some payoff in the matrix to estimate the potential impact. What if it is worth an extra 10 per cent in export revenue ($20b/GDP)? Let’s add ‘1’ to both NZ-Yes ETS payoffs. The value of implementing the ETS is then ((1-4) x 0.9) = -2.7. It still isn’t worthwhile.
The cold calculus of climate change is that, for the moment, ETS is a branding exercise. And it doesn’t look like the brand can pay for itself.
02/04/2012 Comments Off on Black Death – once more with feeling
Seamus Hogan brought in some mathematical notation, so I’ve been trying to figure out the notation for what I was thinking. Here goes:
VoSLs can be estimated by finding examples of people paying to reduce their risks of death. For example, if consumers pay $5,000 more for a car with a safety record that gives them a 1% lower probability of dying in an accident, that suggests a VoSL of $500,000. The equation is:
VoSL = Value / Probability.
We can use this to think about how much to spend now to reduce future impacts of climate change. We multiply the number of lives saved by the VoSL, and then use discounting to compare the future value to present spending:
(VoSL * lives) / (1 + discount rate)^years = spending.
But here’s the problem: VoSL in the future is affected by the quality of life in the future, which is affected by spending on climate change. That is:
Value = f(population),
because when lots of people die, life gets cheaper — this was the quote about the Black Death in the earlier post.
Population = g(spending), so
Value = h(spending).
Putting this back into the estimate of the right level of spending, we get:
((h(spending)/probability)*lives / (1 + discount rate)^ years = spending.
Instead of being able to look at future potential losses and calculating the ‘correct’ level of spending now, we find that spending is on both sides of the equation. Rearrangement gives us:
lives / (1 + discount rate)^ years = spending/(h(spending)/probability).
You can think of the LHS as a constant — the per-life discounting to be applied to future spending. Once you decide the timeframe (100 years?) and the discount rate (good luck), this is just a number. The RHS is a ratio — the ratio of current spending to future VoSL as a function of current spending. Since VoSL increases with spending, we aren’t guaranteed a unique solution. It all depends on h(spending), on that function. This was Seamus’s point — with concave functions for population and quality of life, there are potentially multiple equilibria.
Basically, the reasoning above says that two positions are as economically rational as each other:
- We put a high value on life, and we should preserve future lives by spending now to address climate change.
- It’s all going to descend into brutal chaos, anyway, so we don’t need to bother (insert gratuitous reference to the frightening cultural phenomenon that is The Hunger Games).
More basically, I don’t think economics gives us the answer (provides a unique solution). Instead, it is an ethical problem. Once we sort out the ethical goals, economics can help with the means to achieve them efficiently.
30/03/2012 § 3 Comments
I have been reading about the Black Death, the pandemic that swept through Europe from 1347 to 1350. This has nothing to do with my day job, in case you’re wondering. Wikipedia has this to say (from the Bubonic plague entry):
It is commonly believed that society subsequently became more violent as the mass mortality rate cheapened life and thus increased warfare, crime, popular revolt, waves of flagellants, and persecution.
That seems a pretty uncontroversial statement. As a matter of historical record, there was an increase in manifestations of violence. The violence was directed towards oneself (flagellation) or towards others, and was both petty (crime) and structural (peasant revolts). Historians have connected the dots: life became even nastier, more brutish, and shorter; its value was reduced; the costs of producing violence decreased; therefore, people produced more violence. It’s a simple question of the cost of a factor of production.
That’s a backward look. What if we look forward?
Let’s say that climate change leads to plague and famine. Wouldn’t we expect the same change in the value of a life? Couldn’t we change the sentence above to read, ‘It is commonly believed that society will become more violent as mass mortality cheapens life…?’ It seems, again, relatively uncontroversial to say this.
Now, let’s look at it from the perspective of today. Today, when we are contemplating what to do about climate change, we have in mind or in models some value of a statistical life (VoSL). That is, we put a value on the lives of our descendants. What value should we put on them? If we expect that climate change will lead to a cheapening of VoSL, should we use our current value, or the value we expect to arise in the future?
It’s even worse. Our current actions can affect climate change. We can make it more or less severe. The more severe we make it, the more those future lives may be cheapened, and the less we need to be concerned with the impacts. By taking action to reduce climate change, we are increasing the necessity of taking action. This is, mind you, in a purely in a statistical, intergenerational utility optimising framework.
So whose VoSL should we use, ours or theirs?
07/12/2011 § 1 Comment
It turns out that Prof Weitzman is a Lacanian and doesn’t even know it. At yesterday’s public lecture in Wellington, he made two observations about the symbolic order and climate change:
- the problem of climate change is unprecedented in human history, for its complexity, uncertainty, public good aspects, and potentially catastrophic outcomes; and
- it would be helpful for mobilising actions to deal with climate change if something ‘symbolic’ (his word) happened, but that didn’t harm human welfare too much.
He had me at ‘symbolic’. What could he mean? And what does it mean for climate change?
The first point means that ‘climate change’ has a particular position in the symbolic order that probably doesn’t correspond to the physical damage it could cause human, um, construction, civilisation, society. That is, climate change exists in our thinking like all other signifiers. It exists in relation to other signifiers, in a particular place in the symbolic order. That place is dictated by how our thinking has developed historically. It is rooted in what has happened. If the situation is unique, without precedent, then we are likely to have difficulty thinking about it.
The second point completes the thought. If we can’t ‘think’ climate change properly given our current symbolic order, then that order needs to change. To change that order, we need to insert a new signifier into the symbolic order or change the relationships amongst the existing signifiers. In other words, our definition of ‘climate change’ has to change. Our conception of climate change as it relates to other signifiers has to change.
That’s what Weitzman meant by needing something symbolic to happen. Something would have to happen that shifts the symbolic order enough to provoke people — individuals, communities, nations — into action.
Interestingly, one of the other signifiers that may have to shift is ‘discount rate’. We have specific uses and definitions of discount rates. They have a place in economic analysis. But, given the long time periods involved with climate change, discounting means that future damages are essentially inconsequential. So, reacting to climate change — deciding that it is consequential — means shifting the relationship between ‘discount rate’ and ‘climate change’.
Finally, this will all happen in the Symbolic. We will not start doing anything until we think it is necessary. Weitzman wasn’t hopeful about humanity’s ability to do anything about climate change until it has already taken hold and done serious, obvious damage. Even then, it will be about recognising the events as signifying ‘consequential climate change’, as opposed to just plain old climate change.