2018-12-17

GDP impartially consider'd

We start off in the usual home of bad ideas these days, Twatter. Wherein we find promulgated the thought that Nordhaus may be revered among economists, but ecologists have a very different opinion of his legacy. Many believe that the world's failure to pursue aggressive climate action over the past decades is in large part due to arguments Nordhaus has advanced. This is, of course, drivel. To be fair, as so often with bad ideas on Twatter, the "thought" isn't even novel; it's just a cut-n-paste from some nutter elsewhere. Why is this obvious drivel from nutters that no-one of any degree of sense would take seriously? Because this so massively over-states the degree to which theory, and theory of economics, has affected GW policy. A number of people, some of them very influential, have or had or thought they had excellent self-interested reasons for doing nothing about GW, and this really doesn't require any complex explanations.

Anyway, that's all largely beside the point, because the next stage is mt musing
As I think @jamesannan pointed out some years ago, why should I care whether my descendants are nine times or ten times as rich as I am today? 
Reducing climate impact to GDP is economics answering the wrong question, as well as answering it wrong.
This is the familiar issue2: can we reduce the impacts of GW down to something as crass, crude, distasteful - frankly, my dear, it smacks of trade - as money, when everyone knows that Gentlefolk don't deal with money themselves; they have people to do that for them. This is all familiar stuff, and in Economics and Climatology? from 2012 I was reacting to mt again, that time his:
economics... claims for itself a unique position among the sciences, as the crux, the central weighing mechanism, for all public decision-making.
To clarify (see the comments), mt wasn't objecting to there being a "central weighing mechanism"1, his was objecting to it being economics; I think that he wanted to use the shared morality of civilization as his mechanism instead (and still does); when I tried to probe that (do we have one? How would you know? Do all countries have the same one?) the discussion got bogged down. My objection is that we likely don't have (globally) shared morality, no matter how easy agreement may be around dining tables, and that trying to use such a concept is a recipe for the current mess, not an escape from it. The point about the economics is that it does tell a lot of people that their pet solution is wrong.

Aaanyway, having gone through that and failed to agree yet again, we come to James "tricksy" Annan:
Have also been thinking a little about this recently. One issue is that GDP is a measure of annual turnover, it does not attempt to measure wealth in any way shape or form.
Of course, noting that GDP is not a measure of wealth or income is hardly new. Nonetheless my "Meh; yes, but of course no. Feel free to substitute some other measure; you'd get much the same answers" didn't go down well. If JA is critiquing the problem that GDP is a measure of income rather than wealth, then the answers are (a) yes that's true but uninteresting (if you think otherwise, please provide an interesting consequence); and (b) on the large scale of centuries, GDP growth is exponential, so they're proportional anyway. If he's critiquing the problem that it "only includes monetised transactions, includes government at what it costs rather than the value it adds, doesn't discuss the distribution of income or consumption, only the gross amount and so on and on" (which I think it rather closer to mt's point) then yes, OK, that's a known thing. But like many other known problems that nonetheless survive, it survives because it isn't easy to see a better alternative. Also, although the measure is imperfect, on the large scale there aren't any obvious consequences. Even if you use a rather elastic definition of the word subsidy, the present-day costs of GW aren't large so the stuff we're "missing" from GW related to FF use isn't large, so correcting it would only have a small impact.

Update


Some find this post hard to scute. Naturally, I'm delighted: I didn't spend long years getting an expensive education just so that any old bloke on the wub can understand my deep thoughts. As to what this post is about: read it. Like many of my posts, it doesn't have One True Theme; I'm more reacting the to-me-strange things other people say. It was kicked off, as you can tell by implication, by JA's tweet.

But if it has one theme, it is that nit-picking an idea is cheap; that many nits do not make a nat; that observing that GDP does not describe internal disparity is about as useful as noting the Climate Sensitivity doesn't either; nor does a Mean; yet all remain useful concepts widely used for lack of a better.

I don't think effects on GDP are the best, or at least a good, way of quantifying the effects of human induced climate change is a good summary of this post. I am forced into using GDP for quantifying impacts, but only if this is necessary; for example, for comparing present-day costs of emission reduction versus the future costs of the damage of those emissions aka SCC stuff. At the moment, GDP or some related measure is the only game in town if you want to do that. Throwing up your hands, abandoning the effort, and picking an arbitrary temperature threshold is probably worse; unless the SCC calculations are so uncertain as to be clearly dangerous, and we're forced to say nothing more than "it looks dangerous, we are unable to quantify the risk". I don't think that's true; and if it was true, I think it would be ineffective.

CIP also raises the "catastrophe" point. In answer I don't really have an answer. Of their nature, they are hard to predict or factor in; they add, in a very hard to quantify way, to risk, or if you prefer to "expected" cost.

Notes


1. As I said then, the only alternative to a central mechanism is a decentralised mechanism, where everyone uses their own prejudices. Which is the bad bits of the current political mess.

2. Or indeed two issues, because the "nine times or ten times as rich as I am today" leads off into yet another discussion, which I don't think we need to go into for the moment.

Refs


Ship of fools.
* SHOCKING PRESIDENTIAL PORTRAIT  MARRS OTHERWISE RESPECTABLE PAPER ON MONKEY FECES
* The 2008 Financial Crisis - Arnold Kling

24 comments:

CapitalistImperialistPig said...

Economists and the World Bank have thought about the GDP vs. wealth problem. Wealth is harder to measure, but the best estimates show a strong correlation, with, to be sure, some outliers. It's not true that GDP measures turnover - it measures production, like the name says, and it's hardly surprising that it correlates strongly with wealth, since one meaning of wealth is the ability to command or produce goods and services. Of course GDP is a good proxy for income, since income is just goods and services received - it's hard to recieve them if they don't get produced.

Even if you think of wealth as a brick of gold under the bed, it's ultimate value is not intrinsic but in its ability to command goods and services.

Probably the most logical way to think economically about environmental consequences is look at both wealth and production. If your production is destroying your wealth, as selling all your oil to buy automobiles, it behooves you to think about what happens when the oil is all gone.

On another point, if the only question were whether our descendents were going to be nine times as wealthy as us or ten, we are worrying about nothing important. If the far more reasonable question is whether they will be wealthier than us or much poorer, then it behooves us to think about it. Every exponential law hits its limit.

Phil Hays said...

Chessboard. Regardless of how much rice you have, sooner or later you will run out of rice.

Andy Mitchell said...

I would note that the same UK politicians who said GDP was so important we could not possibly cut carbon emissions, are now saying a drop in GDP is well worth paying for a no deal Brexit.

Michael Tobis said...

I'm not sure I understand what question you are posing.

Here are some that I pose:

Is economics as currently constituted a proper tool for optimising policy that has impacts over centuries?

To the extent it is useful, is maximizing economic activity the objective we should be seeking?

Is well being measured by wealth? Is it impossible to imagine two scenarios, one with high growth and one with low or negative growth, where well-being is unambiguously better in the latter case?

Are the IAMs used by economists to project future climate impacts suited for purpose?

Are economic metrics the right driver for human decision-making?

I am happy to argue in the negative for any of those. But I'm genuinely not sure what question you are posing here.

rconnor said...

> “although the measure is imperfect, on the large scale there aren't any obvious consequences.”

Ummm, how about the consequence where you get to the point where GDP rises (“Hooray!” yells the economists) but where 1% of the people own 82% of the wealth[1] (“Hmmm, not’s not ideal” says 99% of the people).

As much as you’d like to avoid it, the metric you use to determine if a climate policy is good or not is a moral choice. GDP is great for utilitarians. But utilitarianism ain’t the only game in town. And neither is GDP.

Other possible metrics (as examples):
- Social mobility index
- Inequality index
- Rawlsian-style Difference Principal

[1] https://www.bbc.com/news/business-42745853

William Connolley said...

> what question you are posing

I'm not posing a question.

> Is economics as currently constituted

Does this mean anything? Back in 2012 you attempted to distinguish "economics is the science of collective decision-making and the other is: economics is the stuff economists talk about" but when pushed, you couldn't really take this anywhere. I think all you have is a vague uncertainty about economics; I've not seen any evidence you understand it well enough to question it.

> but where 1% of the people own 82% of the wealth

That isn't a consequence of the measure.

Also, you've misread the article you "quote". It doesn't say what you think it says. Note that the figure caption re 1% vs 99% appears to be simply made-up bollox.

rconnor said...

> “That isn't a consequence of the measure.” "you've misread the article"

I should have said “new wealth” not just “wealth”; my mistake for slopping wording. Regardless, the point remains: when GDP is the sole (or main) measure of “good” it puts all focus on the net growth and deems the distribution of that growth to be irrelevant. We then get to situations where 82% of new wealth goes to the richest 1% and the poorest 50% see no increase in wealth and we considered that a “good” economy. That’s a problem.

https://www.oxfam.org/en/research/reward-work-not-wealth


To bring this back to climate change specifically, see the hypothetical situation below:

We put in a carbon tax at $X, which turns out to be the perfect social cost of carbon. The revenue is used to off-set local taxes and allows for GDP to grow Y%. Economically and using GDP as the measure, that’s a win. The folks at the Adam Smith Institute put up a “Mission Accomplished” banner at the office.

The problem is that we can, using GDP as our metric for success, call that carbon tax “a win” without even considering the CO2 emissions or the climate change damages reduced (or lack thereof). Indonesia could be underwater in 2050 in that scenario and it would still be considered a “win” using GDP as the measure. Related question: what is the impact to GDP of mass forced migration caused by climate?

Tom said...

Sigh. Economics is the science of allocating scarce resources. Its goal is not to point the way to ever-higher economic activity. Its goal is to help maximize utility for the highest number of stakeholders.

It is the perfect tool to use when analyzing issues such as those discussed here. It obviously is not the only tool. It equally obviously is not a perfect tool.

I still remember the shocked horror over at Collide-a-Scape when I almost casually mentioned that price is not the only signal in the marketplace. People don't really seem to understand the dismal science. Maybe that's why economists seem so... dismal.

William Connolley said...

> the point... GDP... deems the distribution of that growth to be irrelevant

It's a good point, which is why I made it in my article. Why are you re-making it?

> https://www.oxfam.org/en/research/reward-work-not-wealth

Oxfam's ideas - Last year saw the biggest increase in billionaires in history, one more every two days. This huge increase could have ended global extreme poverty seven times over. 82% of all wealth created in the last year went to the top 1%, and nothing went to the bottom 50% - are stupid and unthinking.

The 50% in extreme poverty are not living in the West. No-one in the West is in the global bottom 50%. The problem that the worst off 50% have is shit govt, lack of rule-of-law; and lack of capitalism. Focussing on disparities in wealth will not help anyone; trying to solve the "problem" by trying to redistribute a fixed amount of wealth will not help.

> Indonesia could be underwater in 2050

That is not possible.

rconnor said...

I’ll simplify the intuition pump for you:

- A carbon tax is set that (as accurately as humanly possible) captures the social cost of carbon.
- Revenue from the carbon tax is used to offset local taxes.
- The carbon tax increased local GDP.
- The carbon tax did not significantly reduce emissions.
- Climate change led to significant damages in the future, particularly in other areas of the planet where now climate refugees are an issue.
- The carbon tax left no reserve to pay for those damages because the revenue was already used to offset local taxes.
- Global GDP doesn’t actually go down that much because those areas were damages were the worst had low GDP to begin with.
- Inequality between countries rises.
- Was the carbon tax successful?
- Did the carbon tax successfully address climate change?

William Connolley said...

Your last question is weird, because you've already answered it in your assumptions. Did you not read what you wrote? I also don't understand your third point; and don't believe your fourth. So, your (IMO invalid) assumptions makes the combination uninteresting. The problem of inequality between countries is real; if you were interested in solving that, you'd turn your attention to the shit govt in the poorer places, not climate.

rconnor said...

An intuition pump is a not a future prediction of reality; it is a hypothetical situation.

Example: Parfit’s teletransporter. The plausibility of a teletransporter ever being constructed in the real world is irrelevant to the discussion. The reader assumes one exists in the hypothetical and then deals with the resulting question of personhood.

In my situation, it does not matter if you think it’s (im)plausible that a carbon tax (on its own) would not significantly reduce emissions*. You just have to consider a hypothetical where a carbon tax was enacted that (1) accurately captured the social cost of carbon, (2) corresponded to an increase in GDP (whether directly or inspite of, it doesn’t matter – the point is it didn’t stifle economic growth) but (3) failed to (on its own) prevent significant climate change damages and ask yourself if the carbon tax was successful.

The point is that if GDP is the go-to metric, then the answer is yes, it was successful. However, this seems counterintuitive because there were still significant damages and the carbon tax isn’t designed to increase GDP, it’s designed to minimize damages from climate change (which the second question hints at). This highlights the faults of using GDP as the go-to metric for evaluating climate change policy.

(*And I’ve already discussed why a carbon tax, on its own, won’t likely be enough to prevent significant damages (ex. most recently by discussing the low price elasticity of fuel) but you’ve dodged those discussions.)

William Connolley said...

Parfit?

Michael Tobis said...

Mr Nordhaus just won the "Nobel" for a model that comprises, exclusive of initialization, 22 lines of Excel code which iterate on 5 year time steps. There is a built-in "underlying growth rate" which begs the most important question to which the models are applied, a question which itself is of dubious merit..

The IAMs are noise. I don't need a deep understanding to understand that.

The economy is far more complex and less constrained than the climate system. But you'd be hard pressed to build even a 1-dimensional RCM in as little code as won Nordhaus his prize.

The more theoretical versions of economics are almost exclusively quasi-equilibrium calculations, but as far as I know the field wouldn't know a justification of quasi-equilibrium if it bit them in the rump.

I haven't seen much, but I've seen enough.

William Connolley said...

> won the "Nobel" for a model

But of course he didn't; he cite is much wider than that, for the ideas of integrating climate into economics. But this is far too detailed; the ignorance / uncertainty of which I'm accusing you (sorry) is far more basic than IAMs; see the "Is economics as currently constituted..." stuff. So I think what I'm saying is that you can't see past the surface. If you think you can write a critique of economics, then it would be good if you tried to do so.

The idea that the economy is oh-so-complex (and so present day economic theory tells us nothing useful, which I take to be what you're saying) had exact analogues amongst the GW denialists, but who say that climate is oh-so-complex (and so present day climate theory tells us nothing useful).

rconnor said...

> "Parfit?”

Squirrel?

I’ll repeat:

In my situation, it does not matter if you think it’s (im)plausible that a carbon tax (on its own) would not significantly reduce emissions*. You just have to consider a hypothetical where a carbon tax was enacted that (1) accurately captured the social cost of carbon, (2) corresponded to an increase in GDP (whether directly or in spite of, it doesn’t matter – the point is it didn’t stifle economic growth) but (3) failed to (on its own) prevent significant climate change damages and ask yourself if the carbon tax was successful.

The point is that if GDP is the go-to metric, then the answer is yes, it was successful. However, this seems counterintuitive because there were still significant damages and the carbon tax isn’t designed to increase GDP, it’s designed to minimize damages from climate change (which the second question hints at). This highlights the faults of using GDP as the go-to metric for evaluating climate change policy.

(*And I’ve already discussed why a carbon tax, on its own, won’t likely be enough to prevent significant damages (ex. most recently by discussing the low price elasticity of fuel) but you’ve dodged those discussions.)

I’d appreciate if you could answer (preferably in a good-faith way) whether you think the hypothetical carbon tax was successful. Or if you really can’t even possibly consider the hypothetical, then perhaps you want to defend your position that it is so unthinkable that a carbon tax *on its own* could possibly fail at mitigating significant future climate change related damages. Especially in light of criticisms that it could fail (ex. due to price elasticity, the difference between capital and operational budget constraints amongst industrial companies, that it neither guarantees emission reductions nor does it create a pot to pay for/to prevent damages, etc.)

William Connolley said...

> ask yourself if the carbon tax was successful... designed to minimize damages from climate change

Your question appears to me to be neither interesting nor sensible. The question "is thing X successful" can only be answered if you know the answer to "what ws thing X intended to do?" If you are a strict Pigouvian, and your aim is that every externality should be internalised by an appropriate tax, then the carbon tax you describe was a success. However, you have (suddenly, now) declared that the success criteria is minimising GW damage. Since the premise of your question is the failure of this aim, your question answers itself. Your assertion that if-GDP-metric-then-success appears to me nonsensical.

You've now used up your quota of stupid questions for a while.

Michael Tobis said...

I do not claim that economic theory tells us nothing useful.

It doesn't follow that it tells us anything useful about climate policy.

Just because we want an answer to a question doesn't mean that some given methodology can be stretched enough to provide a useful answer.

Nordhaus' model along with its its ilk is a violation of EInstein's quote, (possibly apocryphal as most of them are) "things should be made as simple as possible and no simpler". I think it's obvious that it's a casual hack with little basis in reality. But suppose I am wrong. How would you convince me?

I guess you'd have to obtain some testable insight from the model. Any candidates?


William Connolley said...

> obtain some testable insight

The ones you fight against:

* that it is very hard to get a high cost for GW damage into the future, when measured against anything numerical;
* that the discount rate is very important, and that we don't understand it properly;
* that our descendants will be richer than us; which poses the question of why we should "reduce our wealth" (aka, increase it less) in order to make them richer.

Your answer to all of these, at least as far as I can tell, is to retreat from the quantitative to the qualitative, in a way that's reminiscent to me of the retreat of religion in the face of science. It is possible that you're right and that's the correct answer. But in that case I'm doubtful it's a powerful answer, and I'm dubious it would lead to good policy. But it's also possible you're wrong, and we should stick to the numerical grounds.

Phil Hays said...

* that it is very hard to get a high cost for GW damage into the future, when measured against anything numerical;

If you stop at 2100, sure. No everyone agrees that the world ends in 2100, or that we can just stop caring about what happens after 2100.

Look to the PT for a historical analog... and that was only 8 C to 14 C.


* that the discount rate is very important, and that we don't understand it properly;

It is clear that some understand the discount less well than others.


* that our descendants will be richer than us; which poses the question of why we should "reduce our wealth" (aka, increase it less) in order to make them richer.

Don't follow a trend out a window. While our descendants might be richer, at some point wealth must plateau. Sooner or later, you run out of rice.

Tom said...

Much of this discussion should really begin with something that is not well-quantified: Value at Risk. Some have tried to aggregate different estimates of the hit to GDP caused by global warming. As our host intimates, it leads to a somewhat lower amount of growth than would otherwise be the case. IPCC central estimates of sea-level rise will eat up about 0.25% of coastal areas (and storm surge will make more land iffy) according to Tol & Yohe. Stern posits the hit at 4-5% of GDP, but uses a discount rate that isn't well, sane. (His impacts prediction should ideally be an HTML form where you can enter different values for discount and other metrics he uses, such as population rise.

The IPCC does not see climate change as an existential threat to humanity or its various civilizations. Of course they might be wrong, but as their answer agrees with my ignorant prejudice, I advocate looking to their estimate of impacts and base mitigation and adaptation on those estimates. Their impacts estimates are freely available on their website and look eminently addressable.

William Connolley said...

> stop at 2100, sure

My canned reply is Should we care about the world after 2100?, for those lurking (are you lurking? Write a comment just to say "hello!"). In this context, I'd extend my original to ask "do you really need to go past 2100 to get high damages? That seems a lot like conceding the game."

> historical analog

Not "historical". And, consider rates of change.

> some understand the discount less well than others

I agree entirely. Now do you see how pointless your comment was?

> at some point wealth must plateau

...as our forefathers centuries ago would doubtless have said. Anyway, you're wrong.

Incidentally, since Nordhaus comes into all this, this recent thread by him may be of interest.

> Tom

Hello. I'm sorry not to address your comments; you'll have to try saying something I disagree with :-)

Tom said...

Gee, was it more fun when we fought like cats and dogs? I don't really know... BTW, I never, ever intentionally misspelled your name. :)

Phil Hays said...

Geologic historical, perhaps?

So you can put twice the rice on each square? 1 on the first, 2 on the second and so on? Who would have thunk it? 2^64 is only 1.8*10^19

So there is there enough rice? 5×10^14 kg of rice... is a lot of rice. Mass of biosphere is comparable at 4x10^12. Quite a bit smaller, as well.

Maybe you can do it. Right.

Then do 64 more squares. Up to about 10^34 kg of rice.

Four chess boards is about 3*10^72 kg. Mass of universe is only about 1.5 × 10^53

Sooner or later, you run of of rice.

Thread on Nordhaus is interesting, eh?

"But basic principle, that more direct investments in tech would bring faster adoption and better politics has held up well. "