IAMs – Open Thread

There’s been an interesting debate about IAMs. IAMs are Integrated Assessment Models that are used to develop mitigation pathways. In this article, Kevin Anderson argues that IAMs are simply the wrong tools for the job, while Jessica Jewel clarifies the role of IAMs and suggests that IAMs play a central part in the climate debate. Given that things have been somewhat quiet here, I thought it would be interesting to see if we could get a discussion going about this issue.

I wanted to just highlight two comments that caught my eye. Kevin Anderson says

The algorithms embedded in these models assume marginal changes near economic equilibrium, and are heavily reliant on small variations in demand that result from marginal changes in prices.

As I understand it, models of socio-economic systems are largely empirical; the system is too complicated to describe via fundamental equations, and so the models are based on what has happened in the past. I think it’s possible to run these models under scenarios that differ greatly from what we’ve experienced before, but I suspect they’re mostly valid when considering how we, and our economies, would respond to small perturbations. Given that what would be required to achieve some of our goals, it’s not really clear that such models really are all that useful.

Jessica Jewel, however, argues that IAMs to answer what if questions about the future consequences of decisions or developments. I think this is a perfectly valid argument. All models are essentially wrong, but can still be very useful for understanding how a system might evolve.

However, Jessica Jewel also says

The challenge of evaluating feasibility is that the pathways are constrained not only by economic costs and technical complexity, but also by socio-political acceptability.

This is something I’ve always struggled with. Of course, we do need to be aware of socio-political reality, but physical reality doesn’t really care about what we regard as socio-politically acceptable. To achieve some of our climate targets would require doing things that are essentially unprecedented (for example, changing our entire energy infrastructure in a matter of decades). How do we do so if we limit ourselves to doing things that we’re comfortable doing? Of course, I’m not suggesting that we consider doing things that would be regarded as morally unacceptable, but it seems likely that we may have to consider pathways that will change our lifestyles in ways that we may not be entirely happy about.

Of course, maybe socio-politically acceptable includes pathways that we would regard as socio-politically uncomfortable, but this never seems clear to me. It’s also quite possible that I simply don’t really understand IAMs and that they are a perfectly suitable tool for helping us determine mitigation pathways. However, there just seems to be a disconnect between economists suggesting optimal pathways that would lead to around 3.5oC of warming, and scientists who are suggesting that it’s imperative that we limit warming to 1.5oC.

As usual, I’ve written too much; the idea was to try and provoke a discussion. I’ve also tried to be slightly provocative, so feel free to challenge this. It would certainly be interesting to hear from some who thinks IAMs really are useful tools for developing mitigation pathways.

Update:
I should probably have made clearer that there are essentially two types of IAMs (which I have found rather confusing, and may still not fully understand). The ones being discussed in the debate between Kevin Anderson and Jessica Jewel are – I think – ones that try to actually model the evolution of the energy system and other GHG-emitting systems and are (I think) coupled to a simple climate model. The other type has simplified relationships between the various factors (economic growth, costs of mitigation, damages due to climate change, etc) and are used for cost-benefit analyses. The one that suggested an optimal pathway that would lead to warming of 3.5oC was one of the latter, and was the DICE model developed by William Nordhaus. Thanks to Steve Forden for pointing this out.

Links:
Debating the bedrock of climate-change mitigation scenarios – Nature article with the debate between Kevin Anderson and Jessica Jewel.
The human imperative of stabilizing global climate change at 1.5°C – article by some of the authors of the IPCC SR1.5 report.
How Two Wrongs Make a Half-Right – 2013 post by Michael Tobis highlighting some potential issues with IAMs.
A Review of Criticisms of Integrated Assessment Models and Proposed Approaches to Address These, through the Lens of BECCS – a paper reviewing the criticisms of IAMs.

Advertisements
This entry was posted in Carbon tax, Climate change, economics, GRRRROWTH, Philosophy for Bloggers, Uncategorized and tagged , , , , , , , . Bookmark the permalink.

102 Responses to IAMs – Open Thread

  1. I’ve added some links at the end of the post, a couple of which are mentioned in the post (the article with the debate between Kevin Anderson and Jessica Jewel and the article suggesting that it is imperative that we limit warming to 1.5C). However, I’ve also included a link to an older post of Michael Tobis’s that discusses IAMs and a link to a recent paper that review the criticism of IAMs. The latter seems to suggest that it’s complicated; some of ther criticism is valid but that these models are still useful.

  2. Kevin Anderson argues that IAMs are simply the wrong tools for the job, while Jessica Jewel clarifies the role of IAMs and suggests that IAMs play a central part in the climate debate.

    I have not read the debate (yet), but generally it is hard to make the case to stop using tools without providing a better alternative.

  3. Victor,
    I think that is one of the arguments made in the article I linked to at the end. I think there is a recognition that they could be improved (as could all models) but we still probably need some way of trying to determine how we could follow various possible future pathways.

  4. Steven Mosher says:

    so. rogers iron law.

  5. Steven Mosher says:

    when building a model it helps to have experiments.. how do we do difficult things, how do people react to sacrifice?
    sweden should experiment. ban air travel.
    uk can experiment, ban meat.
    get some data. does the public have the stomach for doing what is right.?

  6. Steven,
    The so-called “iron law” clearly applies in many circumstances, but it is a social construct, so isn’t really a law (and, as many have pointed, Brexit seems to violate it). If achieving some of our goals would require breaking it, it might be useful if there were some high-profile political scientists who understood it who could then help to find ways to do so (rather than implying that it’s simply impossible).

  7. Joshua says:

    > This is something I’ve always struggled with. Of course, we do need to be aware of socio-political reality, but physical reality doesn’t really care about what we regard as socio-politically acceptable.

    Bingo.

    On top of the problem of never addressing the “cost” of negative externalities, alarmist doomsayers predict catastrophe and third world children starving – because of aCO2 mitigation and a certain type of economic cause and effect – based on past economic behaviors, which as we all know have been based on questionable rationality and all manor of biases such as recency bias, and that are not like physical laws of reality (which, given quantum mechanics, are dubious enough unto themselves).

    Behavioral/economic iron laws can be reversed at the drop of a hat if there is the will to do so, which there may well inevitably be in a few decades out because of increased climate change. As hard as it is for some to believe, there are no laws that prevent people in richer countries from helping children in poorer counties even if we decide to use less fossil fuel.

    Imagine that!

    What makes this even more ironic is that the mitigation = economic disaster doomsayer alarmists criticize climate scientists for projecting into the future based on! modeling past phenomena.

  8. David B. Benson says:

    Can’t help myself.

    Where are the YOUARE? models?

  9. Joshua,

    Behavioral/economic iron laws can be reversed at the drop of a hat if there is the will to do so, which there may well inevitably be in a few decades out because of increased climate change.

    Indeed. I wonder if any economist has done a study where they develop socio-economic models based only on what we knew up until something like 1930, then tried to project socio-economic development forward till today and then compared with what actually happened. It seems a pretty obvious thing to do, so it may well have been done.

  10. Steven Mosher says:

    im suggesting iron law research and experiments because every law has exceptions. how far can you push people?

    who knows. try. experiment. uk folks outlaw meat. collect data.

    yes yes, who knows when the law will not apply. test. find out.

  11. Steven,
    Of course experiments can be useful. I can imagine, though, that it can be very difficult if it’s clear that it’s simply an experiment (I was not in the UK at the time, but I believe that Margaret Thatcher’s attempt to test a Poll Tax on Scotland went down very badly). Another point might be that we (we being our elected representatives) could decide that we simply need to break the iron law. Then the goal of IAMs, or socio-economic models, would be to determine how to do so, rather than determining if it’s possible, or not.

  12. I suspect the iron law only really applies to the US, if it applies anywhere, there is no reason to suppose that something acceptable in the UK would be acceptable in the US, it wont be a single factor experiments. For example I rather doubt petrol prices in the UK would be tolerated in the us. We have done that experiment already, the public here will tolerate much higher prices, so what next?

    The only iron law related to Roger is that every discussion about climate will end up about how he has been attacked by those who disagree with him. It is just hyperbolic rhetoric that hopes to become true if repeated often enough.

  13. I suspect the iron law only really applies to the US, if it applies anywhere, there is no reason to suppose that something acceptable in the UK would be acceptable in the US, it wont be a single factor experiments.

    Yes. It certainly seems that things that are perfectly normal in one country (people using public transport) is quite unusual in others (or in parts of others). Similarly, how we behave can also change over time (things regarded as normal a few decades ago, are now unacceptable). So, there certainly doesn’t seem to be some simple “iron law” that applies across all time and across all of society.

  14. This the Iron Law you are talking about?

    A weird law, as fighting climate change is an investment in future prosperity, which is the reason for most investments in this modern age.

    The US iron law is that special interests with loads of money will win over the general interest as long as bribing politicians (with campaign contributions and jobs) and media (with ads) is legal.

  15. Victor,
    Indeed. If you accept mainstream economics, then including a carbon tax would properly internalising externalities (i.e., it would include all the costs) and should then lead to the most efficient market-based pathway. The iron law seems to suggest that you can’t implement something that is too disruptive, or that people won’t accept at the time (at least, I think this is what it means). However, this suggests that – in the medium to long term – the iron law could essentially prevent us from implementing policies that would be economically optimal. Utimately, people will suffer in some respecsts because they won’t accept something that might seem inconvenient in the short-term even though it would be beneficial in the medium to long-term.

  16. The people who claim to be the biggest supporters of capitalism seem to be its biggest opponents who reject its consequences when they do not hurt the poor.

  17. It’s not like we have a lot of time for experiments at this point. Sure, we can run parallel strategies and see which ones work best (indeed, Canada is effectively doing so now, for instance), But we are still faced with the fierce urgency of now. Jurisdictions need to act and eggs will be inevitably broken.

  18. Willard says:

  19. mt says:

    I’m happy to stir the pot.

    I do not think IAMs add value to the alternative of nothing at all. I think they might deserve marginal credit as an undergraduate modeling exercise, but they are nothing more than that, not deserving of attention, never mind a “Nobel” prize in economics.

    How do we think about our economic future? It’s very hard.

    There are two fundamental issues here, and my complaints need to be divided into two categories.

    The first is, what is our objective? Is measuring GDP (*) really a good optimization target? Optimization mathematics is immensely powerful, as many other aspects of life show us. But if you optimize for the wrong thing, you can get monstrous results. It’s as if we decided that height were the best metric of health for children and we started feeding them growth hormone just before puberty to get them to be eight feet tall. I think we want to be happy. Perhaps we even want to be wealthy, in some sense. But we do not, as individuals, want to maximize economic throughput. The amount we buy and sell isn’t even tightly coupled to our personal profit. GDP measures churn, not benefit! (Historically it has kept people employed without government intervention, but even that connection is being blurred now.)

    The second is whether these models have any skill at all! Even ifd it were GDP that we ought to care about, what can these models tell us about it into the future?

    In scientific modeling, we speak of the two tests of models: verification and validation. The verification stage is to determine that the model as implemented is actually a reasonable implementation of the model as designed. In fact, IAMs are frequently (and specifically the prototype DICE model is) implemented in a spreadsheet. Spreadsheets are nightmares in terms of testing. So I suspect that even though these models are astonishingly simple (fancy buolding a GCM in Excel) they may be buggy.

    But it’s at the validation stage that quantitative economics really fails. There is an astonishing lack of an actual testing of models against reality. If you look through the economic literature, you simply don’t see the sorts of testing that are performed exhaustively and formally with GCMs. In fact, you’d be hard pressed to find any testing at all. Economic models of all sorts are simply elaborated guesses.

    When it comes to IAMs, it’s worse still. Those of us with any interest in physical modeling are steeped in differential equations, particularly differential equations of evolution. Economics begins with an obsession with “marginal rates”. This essentially means small-signal linearization. You can dig deep into the literature and find someone who is aware of time derivatives (Tol, I think, sent me a reference once) but it is introduced with great fanfare as an especially deep insight. Nonlinear dynamical systems are well outside their purview. Indeed, the simpler idea of regimes of applicability of specific models, elementary in fluid dynamics, is as far as I know utterly absent. (What would the dimensionless numbers that set the boundaries between regimes even look like?)

    IAMs lack anything resembling an explicit time derivative, other than marginal rates which are mostly guesswork. They are a series of quasi-equilibrium calculations taken on a decadal basis. Is that time step sufficient to resolve the system dynamics? Do you think any economist on earth has even heard of the CFL criterion?

    What would be the right time step? Well, perhaps there isn’t one, When you really think about it, one of the most common critiques that is applied to climate modeling (if you can’t predict the weather out three weeks, how can you tell us anything about the climate in a century) has good answers in climatology, but not in economics. The system is woefully nonlinear.

    Our current circumstances are profoundly impacted by social media. How could economics have predicted such a thing? The situation we face is delicately balanced on decisions made by a few individuals – Mr. Gates, Mr. Jobs, Mr. Zuckerberg, Mr. Bezos come to mind immediately. We can’t model any of these individuals effectively. Can we model the net of their decisions and its impact on how those decisions impact the emergent future? Perhaps this isn’t important to the tasks at hand for IAMs. But before you put any confidence in IAMs, you should at least make a case that these come out in the wash. I haven’t seen such a case made.

    Much as we would like to have quantitative ways of thinking about our collective future, I don’t think the problem is tractable in that way. The pretense that it is, strikes me as economic turf-protection, unbased in reality.

    Can economics tell us nothing? I wouldn’t go that far. The small signal analysis that economists more habitually perform does seem to have validity for topics like what interest rates should be, how much money should be made available, how currencies should be exchanged, etc.

    Can it tell us anything about climate, though? I can think of a couple of things. The concept of externalities is valid. The idea that we need to create an artificial shortage of fossil fuels follows, and from that, the fact that we need the price of fossil fuels to rise sharply follows as well. This doesn’t appear to be clear to a lot of people who should know it, and if anything, I think it should be more explicitly explained by economists more regularly.

    I expect there’s more. But nothing particularly useful comes from IAMs. It’s cargo cult science. Producing numbers with a computer is not a difficult thing. Making the numbers meaningful is harder. I don’t see any evidence that IAMs add any value, and I suggest we ignore them.

    ===

    (*) Actually it’s really gross global product that I think we are after here, not gross “domestic” or “national” product at all. I don’t see what domesticity has to do with it.(Is there some reason they started talking about GDP rather than GNP? Is there a distinction?)

  20. MT,

    I’m happy to stir the pot.

    Thanks 🙂

    I sometimes think that some IAMs, in particular those developed in Excel, should really be called calculations, rather than models. I do find the use of Excel somewhat bizarre, given that there have been quite straightforward programming languages that have been around for a while now. However, I also realise that it can take some time to learn a new one (I went from IDL to python a couple of years ago; took a month or so over summer and I’m really pleased that I did it).

  21. Willard says:

    > Spreadsheets are nightmares in terms of testing.

    No idea why you’d say that, mt:

    This week, economists have been astonished to find that a famous academic paper often used to make the case for austerity cuts contains major errors. Another surprise is that the mistakes, by two eminent Harvard professors, were spotted by a student doing his homework.

    https://www.bbc.com/news/magazine-22223190

    Where are auditors when we need them?

  22. verytallguy says:

    Climate Change Policy: What Do the Models Tell Us?

    Very little. A plethora of integrated assessment models (IAMs) have been constructed and used to estimate the social cost of carbon (SCC) and evaluate alternative abatement policies. These models have crucial flaws that make them close to useless as tools for policy analysis: certain inputs (e.g. the discount rate) are arbitrary, but have huge effects on the SCC estimates the models produce; the models’ descriptions of the impact of climate change are completely ad hoc, with no theoretical or empirical foundation; and the models can tell us nothing about the most important driver of the SCC, the possibility of a catastrophic climate outcome. IAM-based analyses of climate policy create a perception of knowledge and precision, but that perception is illusory and misleading.

    https://www.nber.org/papers/w19244

  23. vtg,
    Thanks. Pindyck’s criticisms are discussed quite extensively in this paper.

  24. Joshua says:

    > I wonder if any economist has done a study where they develop socio-economic models based only on what we knew up until something like 1930, then tried to project socio-economic development forward till today and then compared with what actually happened.

    That would be interesting.

    Kind of related… I heard recently about models that track the trend in the number of hours people working in random with technological development. The two tracked as models predicted up until a certain point – people working fewer hours in parallel with new technologies…up until relatively recently when people stopped working less depite new technology.

  25. Joshua says:

    … in tandem…

  26. mt says:

    The Pindyck piece doesn’t go far enough IMO. It simply accepts the “social cost of carbon” approach and puts an informal upper bound of “as much as $200/tonne” on it.

    There is no strong reason to assume that 1) emissions costs, particularly costs of CO2 emissions,
    are independent of prior emissions nor 2) that they are bounded.

    I think to the contrary that CO2 emissions, being essentially cumulative, do more marginal damage the more total damage has already been done. Consider that if we believe that there is some “final tipping point”, a forcing beyond which civilization would collapse to the point where money is meaningless, the total social cost of that tipping can only be regarded as unbounded, essentially infinite, and the social cost per unit emission therefore similarly unbounded.

    A fixed number therefore seems wrong to me, and bracketing it with the highest number any economist has dared to publish is not helpful.

    I believe these models remain terribly underconstrained, and I expect one can get any SCC out of them one wants. I am not an expert and I could be wrong, but this is my current impression.

    Interestingly, some people think this of physical climate model sensitivity as well. I am confident they are wrong about climate physics, so I could be wrong as well about climate economics. Again, I admit this as a real possibility and stand ready to eat my words.

    That said, I can’t resist noting that many of those people who think climate models are arbitrary tend to be especially under the sway of economic thinking. Economists like to think of themselves as mathematical scientists, but unlike other disciplines, they severely lack an empirical tradition to constrain their models. It’s not surprising if they expect other disciplines to operate in similar ways. But this idea that models exist independent of empirical constraints may tell us more about them than it it tells them about us.

  27. MT,

    Consider that if we believe that there is some “final tipping point”, a forcing beyond which civilization would collapse to the point where money is meaningless, the total social cost of that tipping can only be regarded as unbounded, essentially infinite, and the social cost per unit emission therefore similarly unbounded.

    This is indeed an issue I’ve had and which I tried to get across in the post. There are serious climate scientists implying (strongly in some cases) that 4C of warming would be incompatible with organised global community and economic assessments suggesting that the impact of such a level of warming would be quite modest (and – as we all know – Nordhaus’s optimal pathway has a mean of 3.5C).

    Would be interesting to know if economists have tried to reconcile these apparently contradictory positions.

  28. mt says:

    I find myself missing the “like” and “share” features on this conversation!

  29. That reminds me that Steve Forden pointed out a recent paper that suggests that

    Here we show that the global warming limit that minimizes this century’s total economic costs of climate change lies between 1.9 and 2 °C if temperature changes continue to impact national economic growth rates as observed in the past.

    However, if you look at the figures, 4C of warming only leads to Global welfare losses of 12%.

  30. John Hartz says:

    Hot off the press…

    The world’s leading climate science organizations have joined forces to produce a landmark new report for the United Nations Climate Action Summit, underlining the glaring – and growing – gap between agreed targets to tackle global warming and the actual reality.

    The report, United in Science, includes details on the state of the climate and presents trends in the emissions and atmospheric concentrations of the main greenhouse gases. It highlights the urgency of fundamental socio-economic transformation in key sectors such as land use and energy in order to avert dangerous global temperature increase with potentially irreversible impacts. It also examines tools to support both mitigation and adaptation.

    “The Report provides a unified assessment of the state of our Earth system under the increasing influence of anthropogenic climate change, of humanity’s response thus far and of the far-reaching changes that science projects for our global climate in the future. The scientific data and findings presented in the report represent the very latest authoritative information on these topics,” said the Science Advisory Group to the UN Secretary-General’s Climate Action Summit.

    “It highlights the urgent need for the development of concrete actions that halt the worst effects of climate change.”

    Landmark United in Science report informs Climate Action Summit, WMO Press Release, Sep 22, 2019

    The url for the report: itself:

    https://public.wmo.int/en/resources/united_in_science

  31. mt says:

    “Losses of 12%” could mean many things; one needs more precise language to analyze carefully.

    Suppose, then, that this refers to the size of the aggregate of all earthly economies in 100 years.

    An economy that has grown by 12 % in a century has had an annual growth rate of less than one eighth of a percent higher than an economy that has not grown at all:

    >>> 1.12 ** 0.01
    1.001133929265272

    The precision by which we compare values over a century apart yields far less accuracy than this 12 per cent.

    What the claim means at best is that the economic impact of climate change is undetectable by their methods.

    It would be nice if they were to give us a confidence window around their results! I think it is all pretty much just “we think it’s pretty near zero impact but we don’t really know how near”, i.e., ¯\_(ツ)_/¯

  32. MT,

    It would be nice if they were to give us a confidence window around their results!

    In Nordhaus’s recent paper, he does provide uncertainty ranges. The optimal pathway has a 5% to 95% range for temperature by 2100 of 2.5C to 4.6C. So, a 5% chance of the warming exceeding 4.6C in a pathway that his analysis seems to suggest would be optimal (in a cost-benefit analysis sense, I think).

  33. izen says:

    @-ATTP
    “Indeed. I wonder if any economist has done a study where they develop socio-economic models based only on what we knew up until something like 1930, then tried to project socio-economic development forward till today and then compared with what actually happened.”

    This.

    As far as I understand it, the ‘skill’ of GCMs is often related to how well they hindcast. Many are run up with initial conditions from the past and only if they arrive at realistic results comparable with present conditions are they considered good enough to use for projection.
    I have no idea if IAMs follow a similar method of assessing the ‘skill’ of the model and would be interested to hear from anyone with a deeper knowledge of HOW IAMs are graded on their accuracy.
    I have read articles and at least one academic paper that claimed that the adherence to certain rules of mathematical formalism had become more important in the academic realms of economics, (especially among economeretricians) than any coherence with ‘real world’ processes.

  34. izen says:

    GCMs are just starting to have some success with projecting regional and local climate change.
    I wonder if IAMs can predict inflation rates in different Nations?
    At present most Nations fall between zero and 10% inflation, with a few outliers that have small negative rates and a few with rates well into double figures.
    Those are often the Nations with serious internal political problems that are also the source of most migration. I would suspect the cost of that is not modelled in IAMs.
    Or the cost of a wall….

  35. mt says:

    Ken – “The optimal pathway has a 5% to 95% range for temperature by 2100 of 2.5C to 4.6C. ”

    That’s not the question I was asking. I was for confidence bounds in his evaluation of any given scenario *in terms of his own metric*, not in terms of the physical outcome of the optimum emission trajectory that his model spits out.

    Given the nature of the optimization, this would be crucial in dollar-denominated tradeoffs.

  36. mt says:

    @izen The main climate models are over a million lines of code now. (Not saying this is necessarily a good thing to optimize for, but there it is.)

    Human society is vastly more complicated than the climate system, for which basic physical principles have long been known. One would anticipate considerable complexity in its modelling. But what we get from our Nobel-worthy study is a few hundred Excel cells, a couple of dozen formulas, a 10 year time step, and a presumed “underlying” growth rate, a problem which you haven’t even gotten me started on yet.

    I think some of the IAMs do in fact have a notion of “continent” in them. You can break this down by country if you like. Would suggest you start by coding it in a testable language. Seems like a couple of weeks’ work, and it would be a substantial improvement over the state of the art.

    But I still wouldn’t believe it. As we know, all models are wrong, but only a few of the possible wrong models are useful, and most aren’t. You have quite a job to do to convince me that yours is useful.

    Not that I matter all that much either as an individual or as representing a demographic. Unfortunately, the powers that be are in the habit of listening to economists, and not to physical scientists, and that matters much more than my opinion.

    So I wish they would be more astute in which directions they direct their skepticism.

  37. mt,

    That’s not the question I was asking. I was for confidence bounds in his evaluation of any given scenario *in terms of his own metric*, not in terms of the physical outcome of the optimum emission trajectory that his model spits out.

    I’m probably just being dense, but how would you determine this? Seems like might require quite a bit of expert judgement.

  38. verytallguy says:

    Nordhaus on IAMS

    A couple things to note:

    1) The social cost of carbon varies *by a factor of five* according to the discount rate chosen from 2.5-5%
    2) Damage functions are as low as just 10% of global output for a six degree temperature rise. This is similar to the effect of Brexit on the UK economy, for comparison. Yet a six degree temperature rise would have a cataclysmic effect on ecosystems worldwide, render parts of the globe literally uninhabitable and guarantee future sea level rises to flood most major cities in the world. Are these really comparable?

    Some more general points:

    IAMS fundamentally not fit for purpose: They are essentially considering small perturbations around constant, positive economic growth unconstrained by resource limitations. Climate change on the other hand, will unchecked transform the entire natural world.

    Even within their own narrow applicability, the uncertainties are so large as to render results totally meaningless. Discount rates are arbitrary but small changes make huge impacts on the outputs of IAMS

    There is no reason looking at historic economic analysis to believe we have any skill whatever in predicting out a century ahead the economics of any decision taken today, be it climate, technology, societal or other.

    Summary: We should be talking about values, not value. Not because value doesn’t matter, but because at the timescales under discussion, it is unknowable.

  39. Dave_Geologist says:

    Our current circumstances are profoundly impacted by social media. How could economics have predicted such a thing? The situation we face is delicately balanced on decisions made by a few individuals – Mr. Gates, Mr. Jobs, Mr. Zuckerberg, Mr. Bezos come to mind immediately. We can’t model any of these individuals effectively. Can we model the net of their decisions and its impact on how those decisions impact the emergent future?

    SPOILER ALERT!

    mt, I’m reminded of Asimov’s Foundation Series. Psychohistory purported to model the actions of human populations hundreds or thousands of years into the future. But it fell over when a telepathic mutant warlord, The Mule came along. Fortunately he was sterile (hence the self-deprecating name) and had no telepathic offspring, so things got back on track. But later it turns out that even before The Mule, there had actually been a secret organisation behind the scenes manipulating people and populations to keep the Empire and its successors on the desired track. Even without an unforeseen mutant, the predictions broke down due to what we’d nowadays call chaos. Then, in a further regression. it turns out that the real actors behind the scenes were robots (an AI, in modern terms), manipulating us For Our Own Good. Once the robots got smart enough, they were compelled to do so under the Zeroth Law: “A robot may not harm humanity, or, by inaction, allow humanity to come to harm”.

    Perhaps IAMs are today’s psychohistory. Now all we need is a benign AI overlord (in parallel Asimov stories, Multivac and its successors, although the boundaries become blurred when the human race uploads itself into the AI and becomes non-corporeal). If you want to know how the story ends, read The Last Question.

  40. verytallguy says:

    Posting links to anywhere without so much as a word of explanation is a crime; if the link is to the Mail it’s a crime against humanity and SM should therefore be banned.

  41. Joshua says:

    > and SM should therefore be banned.

    Much too light a sentence for a crime against humanity. Especially for a repeat offender.

  42. Dave_Geologist says:

    You went from IDL to python only a couple of years ago ATTP?

    Now I don’t feel so old! IDL originated from early VAX/VMS/Fortran, and its syntax still shows its heritage. Actually, DCL/Fortran was more my thing, in the PDP days, so maybe I feel old again. I’m glad I steered clear of MCR, because DCL to Vax was a smooth transition with most things staying the same. MCR struck me as more unix-y, and I hated Unix. I even bought the book when I could no longer put it off. And yet here I am, typing away on Linux… suppose you can get used to anything after a few decades 😉 .

  43. Should probably make it a formal part of the comments/moderation policy, but – yes – we do require some context, rather than simply dumping a link.

  44. You went from IDL to python only a couple of years ago ATTP? Now I don’t feel so old!

    I felt old hearing how quickly ATTP picked up Python. I switched from Matlab to R some years ago and have the feeling I will never master R like I did Matlab.

  45. Willard says:

    > rather than simply dumping a link.

    More so when it’s to throw red meat, in this case both figuratively and literaly at least in the cool kids’ sense.

    Now that I have to explain myself, here’s a newsie showing that intra-G10 exodus has already started:

    Let’s wait a few more years and I predict cities will return to moats.

  46. wmconnolley says:

    > Damage functions are as low as just 10% of global output for a six degree temperature rise. This is similar to the effect of Brexit on the UK economy, for comparison. Yet a six degree temperature rise would have a cataclysmic effect on ecosystems worldwide, render parts of the globe literally uninhabitable and guarantee future sea level rises to flood most major cities in the world.

    I think this goes towards the core of people’s problems with the IAMs. But, you don’t have to just sit back and accept it. That bit is not a black box. You can look at where it comes from and see how reasonable it is. Mind you, having said that, you provide no sources.

  47. wmc,

    Mind you, having said that, you provide no sources.

    If you look at Table A-1 in Nordhaus’s recent paper, you see that for a temperature change of 6.1C, the damages (as a % of output) is 12.7%. You can also download the DICE model from here, which I have done but haven’t really looked at it in any detail.

  48. To follow up a bit, I have wondered why someone hasn’t modified the IAM damage function so that it’s essentially quadratic at low levels of warming but becomes very large at some high level of warming (maybe 8C, or 10C).

  49. mt says:

    Well, then you’d get a different result, of course.

    But this is really the point, isn’t it? You want a model that can tell you something other than what you assumed going in.

    I don’t think the people working on the economic projections of climate impact know how to do that, and I’m not at all convinced it’s even possible.

  50. wmconnolley says:

    > Table A-1 in Nordhaus’s recent paper, you see that for a temperature change of 6.1C, the damages (as a % of output) is 12.7%

    I just looked at https://www.nber.org/papers/w22933.pdf. Table A-1 is on page 41. I can’t see anything corresponding to 6.1 oC or 12.7% there. Am I missing it? Did you point me at the wrong thing?

  51. WMC, The PDF I had downloaded must have been a preprint. It’s now A-4 on page 44 (baseline scenario). Numbers are slightly different, but similar enough.

  52. verytallguy says:

    Wot ATTP sed.

    I accept my 10% was a low eyeball, but I’m sure you get the drift.

    IAMS merely codify assumptions and values, and produce wildly different outcomes with mildly different inputs.

    Specifically:

    Assumption that climate change does nothing more than cause mild perturbation to human existence, even to 6 degrees and above

    The value of every ecosystem on earth is low to negligible.

    Our views on these assumptions and values are where the debate should be, not whether an IAM which had made these implicitly gives 5% more economic output in 100 years time.

  53. wmconnolley says:

    Ah, so that shows damages (95 percentile) of 11.51% at temperature (ditto) of 6.33 at 2100. As against 3.22% (central) at 4.08. So the damages are heavily non-linear, at least. Is 3% damage at +4 oC obviously wrong? It seems consistent with their inputs:

    “The current version of the DICE model continues to rely on existing damage studies. The
    method for estimating the damage function was the following: The new estimates started with
    a survey of damage estimates by Andrew Moffat and Nordhaus (2017). The underlying damage
    estimates were identified and collected. The survey turned up 27 studies which contained 36
    useable damage estimates. For example, a recent version of the PAGE model estimated that the
    impact of a 4 °C warming would be to lower global output by 2.9%”.

    So I think there’s no point railing at the IAMs, if what you want to complain about is their damage function.

  54. izen says:

    @-wc
    “For example, a recent version of the PAGE model estimated that the impact of a 4 °C warming would be to lower global output by 2.9%”

    The page model referred to IS an IAM.

    I find it difficult to take seriously analysis that estimates just a 2.9% reduction in GDP growth by 4C when the impact of the existing ~1C rise has already reduced maize and wheat yields by 3.8 and 5.5%, respectively. Whatever assumptions are made about GDP Grooowth, a reduction in the calories available to the global population from wheat and maize of over 18% seems… problematic.
    A meta-study indicates that –

    https://www.pnas.org/content/114/35/9326
    …each degree-Celsius increase in global mean temperature would, on average, reduce global yields of wheat by 6.0%, rice by 3.2%, maize by 7.4%, and soybean by 3.1%.

    The math is above my pay grade, but it seems the IAMs have been examined and found wanting.

    https://scholar.harvard.edu/weitzman/files/damagesfunctionglobalwarming.pdf
    With this kind of fundamental non-robustness, the outcomes of CBAs or IAMs are held hostage by core structural uncertainties about how high temperature change and high productive capacity should be combined to yield utility. Such a dismal message is not intended to bring despair to the economics of climate change, nor to negate the need for further study and numerical simulations to guide policy. Instead, this message is just another warning, in a growing series of cautionary tales, that the particular application of CBAs or IAMs to climate change seems more inherently prone to being dependent on subjective judgments about structural uncertainties than most other, more ordinary, applications of CBAs or IAMs.

  55. tw2017 says:

    What is the source for the idea that “the impact of the existing ~1C rise has already reduced maize and wheat yields by 3.8 and 5.5%, respectively”

    Here are real yields per hectare from World Bank data:
    https://data.worldbank.org/indicator/ag.yld.crel.kg
    It’s simply true that total crop production, and total productivity per hectare, have been growing steadily for decades. If you look for any major crop, you see the actual figures look like the above reference.

  56. Steven Mosher says:

    “I felt old hearing how quickly ATTP picked up Python. I switched from Matlab to R some years ago and have the feeling I will never master R like I did Matlab.”

    easiest way is to read some master code and ask questions everyone on the R help list is a master.

  57. Steven Mosher says:

    “More so when it’s to throw red meat, in this case both figuratively and literaly at least in the cool kids’ sense.”

    my sense was people would get the obvious connection between me suggesting, tongue in cheek, that the Brits should give up meat, and then finding someone who suggests, for reals, the same thing.

    Here is the thing. given some folks arguments above.. “urgency urgency break some eggs”
    given the distain for models, at least these types of models.. what takes their place?

    In short, everyone uses the crisis to promote their pet value, or dont eat your pets values.
    ( psst dog is yummy )

    Think of all of the opportunity to virtue signal while the house burns down.
    anything to distract people from the actual problem which has nothing to do with emissions in the west.

    better put some brain cells on adapting to 3C, cause that is what we will see.

  58. izen says:

    @-tw2017
    “What is the source for the idea that “the impact of the existing ~1C rise has already reduced maize and wheat yields by 3.8 and 5.5%, respectively” ”

    https://science.sciencemag.org/content/333/6042/616?ijkey=d81edbf05ed0824c93d433ccc93cdf1c9eb1d04d&keytype2=tf_ipsecsha
    Given the hill-shaped yield-temperature function, predicted decreases are larger the warmer a country is to begin with. A 1°C rise tends to lower yields by up to 10% ….

  59. izen says:

    @-tw2017
    “It’s simply true that total crop production, and total productivity per hectare, have been growing steadily for decades. ”

    True but not simple.
    That improvement in crop yields since 1965 has NOT been with the same plant with the same cultivation pattern.
    Better irrigation infrastructure, vastly increased fertiliser, herbicide, and pesticide use, (derived from hydrocarbons) and mechanised/industrial agricultural methods have been employed.
    But the 4 main food plants have well established optimum temperatures for cultivation. breeding and perhaps GM can shift those slightly, but much maize and most wheat is already grown in areas that are warmer than that optimum.

    https://www.sciencedirect.com/science/article/pii/S2212094715300116

  60. izen says:

    @-tw2017

    Perhaps the US Midwest can switch from maize to rice….
    https://izenmeme.wordpress.com/iowafloodz1/

  61. WMC,
    That quote is a little odd, since PAGE is another IAM. So, it’s not clear that damage estimates are not also coming from IAMs (I may be confused). Should also clarify again what I pointed out in the update. I think there are two different types of IAMs (why they couldn’t have different names is not clear to me). The ones we’re discussing are the ones that do cost-benefit analysis and – as I understand it – are quite simple (Nordhaus’ model is in an Excel workbook). There is another type that actually tries to model the evolution of energy systems and also includes a simple climate model. I think you can also do cost-benefit analyses with these (one of which I link to in the links section at the base of my post). These – I think – are somewhat more sophisticated than the Nordhaus-type IAMs.

  62. wmconnolley says:

    > when the impact of the existing ~1C rise has already reduced maize and wheat yields by 3.8 and 5.5%, respectively

    I don’t follow you. Agriculture is globally ~3.5% of GDP (https://data.worldbank.org/indicator/NV.AGR.TOTL.ZS) and is, one presumes, one of the more sensitive parts of it. So your numbers are fully consistent with IAM results. And of course total yields are rising.

    > since PAGE is another IAM

    Quite possibly, but that’s not quite the point, the point is to find something actually *wrong* with the damage function.

  63. verytallguy says:

    “the point is to find something actually *wrong* with the damage function.”

    That it implies very low (<10%) damages for a temperature rise (6 degrees) which we expect to be cataclysmic.

  64. verytallguy says:

    < should read ~

  65. wmconnolley says:

    Probably the damage estimates are less reliable at the high end, so focussing on the 95% percentile is probably a mistake. Is ~3% for 4 oC “obviously” wrong? If so, why?

  66. WMC,
    I wasn’t intended to focus on the high end. I was simply providing a link that shows that warming of ~6C would be estimated – by IAMs – to produce damages of ~10% (you commented that noone had provided a source, so I did).

    Is 3% for 4C wrong? I don’t know. However, there are certainly a number of quite credible people (Kevin Anderson, Shellnhuber, for example) who seem to imply that it would be incredibly damaging. My hand wavy argument would be that it would seem surprising that we could produce a change comparible to that of the change between a glacial and an inter-glacial, and on a timescale about 10 times faster than the glacial/inter-glacial transition, and discover that the impact was pretty minor. I guess it’s possible that the impact is major and yet the economic damages were minor, but that might suggest that quantifying it in this way is not a particularly good way of describing the impacts.

  67. wmconnolley says:

    > I wasn’t intended to focus on the high end

    My particular comment re focus was addressed to vtg.

    > who seem to imply that it would be incredibly damaging

    Yes. But “incredibly” is just a word. I think the people who dislike the “low” damage functions need to do rather better. KA for example is a pro – can’t he do some research and provide something better?

    > it’s possible that the impact is major and yet the economic damages were minor, but that might suggest that quantifying it in this way is not a particularly good way of describing the impacts

    This is possible. In which case, all the angst about the damage function would turn out to be misplaced, and you should have been focussing your energy on “yes the economy will be fine, but the natural world will be ruined”.

  68. WMC,
    There is an alternative perspective. If someone came to me and said “I’ve designed a rocket that can get to alpha Centauri and back in two years” I’d simply say “you’ve made a mistake”. I wouldn’t need to work out where, it would clearly be wrong. So, in some sense, it might be reasonable to expect those who develop IAMs and damage functions to sanity check their results, especially if there are credible people who seem to be pointing out that some of their estimates appear to not make sense.

    In which case, all the angst about the damage function would turn out to be misplaced, and you should have been focussing your energy on “yes the economy will be fine, but the natural world will be ruined”.

    Not sure it’s angst (it’s just a discussion). It is possibe that the economy will be fine, but the natural world will be ruined, but I’m not sure these are entirely consistent positions. What would an economy look like if the natural world was ruined? A small fraction of the world’s population living in extreme wealth, while the rest starved? I really don’t know, but my point was more that if this were the case, what would “the economy will be fine” actually mean to the average person? Alternatively, what would “the damages are 3% of output” mean to the average person?

  69. verytallguy says:

    Well, it looks like we agree at least that the damage function is clearly wrong for large temperature changes.

  70. verytallguy says:

    [contd]… and as it’s at the high end where the impacts are focussed, that seems like an obvious fatal flaw to me.

    The other fatal flaw is the economic modelling. Apologies, I don’t have time to expand on that right now.

  71. Ben McMillan says:

    The issue is that a world where natural systems break down is that it seems plausible that this will involve large regions of the world which become practicably uninhabitable due to drought, famine, and extreme weather. The resulting famine, displacement, and warfare are likely to have very large effects on the people involved, and worldwide.

    These are not in IAMs. Instead, trivial things like the cost of heating and cooling houses dominate the damage estimates (at least for the one I was looking at, think it was DICE). There is an implicit assumption that everyone in 2100 will be rich enough that they can just buy food and water from somewhere else in the world, live in air-conditioned houses, and move above the new sea level.

    Displacement, famine and warfare are pretty typical results of natural systems breaking down. Even something seemingly minor like potato harvests failing has caused significant migration, starvation and misery.

    I guess you could model this by assuming countries are essentially wiped out past some threshold. I don’t know why IAMs don’t do this kind of thing but I expect it is because (as pointed out above) most economics focuses on marginal changes and single-equilibria systems with simple dynamics. These kind of spreadsheet models also don’t do things like stock-market crashes well.

  72. Chubbs says:

    The biggest financial risk from climate change is all the bad investments that are being made due to climate denial.

    “All told, one-third of equity and fixed income assets issued in global financial markets can be classified as belonging to the natural resource and extraction sectors, as well as carbon-intensive power utilities, chemicals, construction, and industrial goods firms. Decarbonization would essentially strand those assets, resulting in losses in asset values for the energy sector of $1 trillion to $4 trillion. In the broader industrial sector, the stranded asset risks could rise to $20 trillion.”

    https://foreignpolicy.com/2019/07/20/why-central-banks-need-to-step-up-on-global-warming/

  73. VTG said:
    “The other fatal flaw is the economic modelling. Apologies, I don’t have time to expand on that right now.”

    Guessing what the fatal flaw is.

    Except for resource economics (which is mostly geological constraints) economic modeling is largely based on game theory. This says that any model is subverted by gaming the system as people become aware of the model. See also Lucas_critique, Goodhart’s law, Campbell’s law, Jevon’s paradox, etc

    So with game theory, the question to ask is “how much will be kept in the ground?”
    Everyone will agree to keep some in the ground, but then someone will subvert the system and gain advantage by extracting the resources. Or someone will suggest that we must extract the fuel to power the machinery that will help us find alternatives. So assume it’s extracted until it’s all gone. That’s why resource economics has at least some validity as its governed by constraints, ie arguing over RCP8.5

    This is all arguable but how I rationalize why I don’t fall into the “fatal flaw” black-hole of economic modeling.

  74. wmconnolley: “I don’t follow you. Agriculture is globally ~3.5% of GDP

    When there is no longer enough food, this percentage quickly rises. It is kinda a necessity of life.

    ATTP: “What would an economy look like if the natural world was ruined? A small fraction of the world’s population living in extreme wealth, while the rest starved?

    For am IAM that would be a small impact as long as those wealthy people are doing fine and investing in bigger fortresses.

  75. Willard says:

    > KA for example is a pro – can’t he do some research and provide something better?

    “KA” is just a name. I think people who put faith in IAMs should do better than mention a name, e.g.:

    The paper analyzes the effects of varying climate impacts on the social cost of carbon and economic growth. We use polynomial damage functions in a model of an endogenously growing two-sector economy. The framework includes nonrenewable natural resources which cause greenhouse gas emissions; pollution stock harms capital and reduces economic growth. We find a big effect of the selected damage function on the social cost of carbon and a significant impact on the growth rate. In our calibration a quartic damage function raises the social cost of carbon by more than a factor of ten compared to the linear function. In the social optimum the growth rate remains positive even when the damage function is highly convex. We test the robustness of the results by adding pollution decay, lowering the elasticity of intertemporal substitution, and addressing uncertainty, which does not alter our results. We find that high marginal climate damages require stringent climate policies but do not preclude positive economic growth despite convexity, provided that policies are designed in an efficient manner.

    https://doi.org/10.1007/s10640-018-0219-y

    Can’t they at least pay due diligence to the lichurchur they stan?

  76. mt says:

    “When there is no longer enough food, this percentage quickly rises. It is kinda a necessity of life.”

    This is essentially the argument of https://dothemath.ucsd.edu/2012/04/economist-meets-physicist/ which we really need to bring into the conversation at this point. (Read it first before you read this comment, please.)

    The counterargument that someone presented to me once (not sure if it was William) is that salt is necessary, but salt is very inexpensive.

    The question then boils down (this is not obvious, but I think it’s true) to whether food (and other necessities) will ever be so cheap to produce that people will continue to produce it even if it is a vanishingly small part of the economy, so much so that people will willingly trade it for massages, movie tickets, blog posts, and other very light impact products. Like salt.

    I think this is unlikely. I don’t think *real* things will ever be a vanishingly small part of the economy: we require a whole series of physical inputs to make those necessary processes happen. This argument seems a bit circular, I admit.

    But look at it this way, we can turn food (feeding creative professionals) into fantasies (superhero movies e.g.), but we can’t turn superhero movies back into food. The fundamental idea of money-based economics is that guns and butter are interchangeable. And superhero movies. But you can’t eat superhero movies.

    There is a real economy, call it Smil-world (after Vaclav Smil), with measurable fluxes of energy and materials. Insofar as we are concerned with a single planet, I think we can agree that Smil-world is bounded.

    The question then boils down to whether we can imagine an economy in which Smil-world continues to supply our needs and desires, but forms an ever-diminishing part of our economic activity and motivations. It seems unlikely to me, but I find it frustratingly difficult to justify my belief.

    But growth proponents ask us to believe even more than this. They expect this dematerialization of our interests to take place very rapidly, and in parallel with a non-increase in the scope of Smil-world as a whole and a rapid decline in the parts of it that are presently problematic. Here is where for me it seems little more than wishful thinking.

  77. Willard says:

    Speaking of Smil-worlds (bold text is a question):

    You debunk overly rosy projections by techno-optimists, who say we can solve all our problems with smarter computers, and economists, who promise endless capitalist growth. In many countries, the downside of material growth now seems greater than the upside, which leads to what you call “anthropogenic insults to ecosystems”. Is that a fair summary?
    Yes, I think so. Without a biosphere in a good shape, there is no life on the planet. It’s very simple. That’s all you need to know. The economists will tell you we can decouple growth from material consumption, but that is total nonsense. The options are quite clear from the historical evidence. If you don’t manage decline, then you succumb to it and you are gone. The best hope is that you find some way to manage it. We are in a better position to do that now than we were 50 or 100 years ago, because our knowledge is much vaster. If we sit down, we can come up with something. It won’t be painless, but we can come up with ways to minimise that pain.

    https://www.theguardian.com/books/2019/sep/21/vaclav-smil-interview-growth-must-end-economists

    My main beef with Smil is that he’s writing books. They are not cheap. They are not peer-reviewed. The epistemic moat becomes an economic moat. I dislike economic moats. A taxonomy:

    http://reactionwheel.net/2019/09/a-taxonomy-of-moats.html

  78. Salt is currently cheap, but it used to be expensive. Solders were paid in salt, the word salary stems from salt.

    WMC would sign a song of praise to capitalism about it. I will sing a song a praise to the governments that created the infrastructure, the rule of law and the peace that made this possible. I hope someone will join in singing about the activists that forced the powerful to give us some rule of law and peace. A wonderful choir of praise.

    Food is currently also cheap for the most rich billion of people, while it is a large part of the budget of poor people and for us in the past.

    I did not want to claim that it is inevitable that food becomes scares and expensive, but the argument that it is currently cheap is not strong. The price depends on supply and demand. If demand becomes larger than supply the prices will quickly rise, first killing the poor, then eating into our household budgets.

    We have some buffer in all the meat we are eating. If growing grains becomes harder we could eat more of it ourselves and less of it to cattle. But under the economic system of WMC a billionaire eating a steak is prioritised over the life of a poor person.

  79. wmconnolley says:

    > https://dothemath.ucsd.edu/2012/04/economist-meets-physicist/

    https://wmconnolley.wordpress.com/2015/05/13/agricultural-land-value-as-a-percentage-of-gdp/
    http://noahpinionblog.blogspot.com/2012/11/murphys-law.html

    >> wmconnolley: “I don’t follow you. Agriculture is globally ~3.5% of GDP”
    > When there is no longer enough food, this percentage quickly rises.

    I think you’ve missed the point. Which is that while it’spossible to assert that GW has impacted yields relative to what they would have been sans GW, the yields are still rising, so overall production is still rising. So shortage of food is not currently relevant, and there’s no clear reason to think it will be (see also the falling value of ag land as % of GDP).

    > the argument that it is currently cheap is not strong

    Again, you’ve missed the point. The argument is that it is “cheap” and becoming everywhere a lower proportion of people’s total spend, over time.

    > ATTP: “What would an economy look like if the natural world was ruined?

    FWIW, I don’t think that’s a plausible scenario. A ruined natural world would impact the economy strongly.

    > WMC a billionaire eating a steak

    I don’t particularly like steak, thanks. Humous is nicer.

    > Vaclav Smil
    http://mustelid.blogspot.com/2019/09/words-for-word-god.html (part two).

    > a quartic damage function raises the social cost of carbon by more than a factor of ten compared to the linear function
    The comparison to linear isn’t interesting, I think; DICE uses quadratic. Is the quartic well-fitted? The higher the degree the more likely for it to do wacky things outside its calibration regime.

  80. mt says:

    “But under the economic system of WMC a billionaire eating a steak is prioritised over the life of a poor person.”

    Yes. The way I often put it is that the beef cow outbids the Haitian, who starves not for want of food but for want of money. A situation which I find bizarre and unacceptable.

    But the food system has a *lot* of slack. The production of grain for cattle feed is very efficient in dollars (it has to be, it is only feeding a beef cow, which in turn has a profit margin of about $50). But it is very inefficient in land and inputs. Intensive, sustainable agriculture can produce an enormous amount of food with much smaller inputs of anything besides labour.

    I am worried about many things in our future, but a resource-driven food shortage is not one of them. We don’t have a resource-driven food shortage now, and I am confident that even the youngest among us will not live to see one. What we have now is rank injustice, and we may augment it with rank stupidity. But if people starve, I really have a hard time believing it will be about a physical incapacity to produce food for 10 billion or so.

    But that clearly doesn’t make us immune to an institutional incapacity. And as wealth and power concentrate, we clearly see that the present-day motivation to feed everybody is not economic. Why?

    Allow me to speculate. (I won’t bet my life on this idea but I wonder if there’s something to it.)

    Could this be *because* the food economy is such a small part of the food-and-fantasy economy? Is cheap food as a fraction of the total economy going to feed more people, or fewer? It seems a paradox to suggest the latter, but look around you.

    If it’s impossible for Smil-world to become too small a fraction of money-world, we should consider what the symptoms of tending that way would be.

    Essentially, they could amount to us feeding cows and not people; fueling airplanes and not bellies. Why bother feeding poor people with cheap food when the profit margins are so small?

    If the solution to feeding the world sustainably is labour-intensive, and yet the proportion of the economy devoted to food shrinks, the person working on the farm is increasingly the one least likely to eat! Admittedly, it seems utterly implausible, except for the fact that it seems to be happening.

  81. WMC,

    FWIW, I don’t think that’s a plausible scenario. A ruined natural world would impact the economy strongly.

    Yes, I would agree. I was responding to your comment about the economy being fine but the natural world being ruined. I’m mostly not convinced that the economic models correctly price the impact of severe impacts to the natural world. I’m not even sure that this is something that would be all that easy to do. It is a very complex system. Hence, maybe it’s fine to use these type of models to understand how we can influence society and our economies on short timescales, but not really okay to use them to determine how things will probably be in many decades time?

  82. My understanding of the argument against “the physicist” is that there is no reason why continued economic growth has to involve continually increasing energy use, and the continued exploitation of the natural world. I think I can understand that this is theoretically possible. However, I am confused by how the system would respond if key things (like agriculture) became a vanishingly small part of the global economy. Also, as I understand it, we haven’t really demonstrated this level of decoupling in reality. So, yes, maybe we can have continued economic growth that manages to avoid having increasing environmental impact, but it does seem to be something that has yet to be demonstrated (happy to be proven wrong).

  83. izen says:

    @-wc
    “The argument is that it is “cheap” and becoming everywhere a lower proportion of people’s total spend, over time.”

    Often food is ‘cheap’ because governments heavily subsidise basic foods as a means of ensuring social stability. Its the modern equivalent of the Roman ‘bread and circuses’. This because while food production is increasing and supply exceeds demand, the market as a way of meating(?) the need costs more than the means of a large % of the population and therefore the generation of profit for private enterprise.

    There are indications that the increasing production and cost of food is driven not by a need to feed people but as a subsidised, and gamed market system. (corn ethanol?)
    Food may be becoming a lower proportion of people’s total spend because there is a government and market interest in any additional wealth accrued by consumers getting spent on more profitable consumer goods and services.

    The problems arise when there IS an actual shortage of food and governments lack the means to subsidise the increased cost. Man (and woman) may not live by bread alone, but if there is 20% less of it, the suggestion that they eat cake is historically ineffective. Starvation of soylent green become the alternatives.

    A real world example.
    It is widely accepted that the troubles in Syria started with crop failure due to droughts and insufficient oil production to pay for imports and subsidies the ensure the general population could be fed at affordable prices.
    The impact on GDP is well established, as is the impact on social stability.

    https://www.ceicdata.com/en/indicator/syria/real-gdp-growth

    To echo mt I think a shortage of super-hero movies and online games has a smaller effect on social stability that a shortage of food. Syria become an example in microcosm of the effects of shortage of real material assets which have a far greater impact on GDP than I think IAM’s can capture.

    Here is the description of the PAGE09 IAM.
    https://www.jbs.cam.ac.uk/fileadmin/user_upload/research/workingpapers/wp1104.pdf
    I struggle to see this any anything other than an arid mathematical formalism applied to how physical processes can impact National and Global economies. It certainly seems to have little relevance to the situation as it evolved in Syria, or the developing problems in South America in Venezuela, Argentina, Honduras, El Salvador, Guatemala, Nicaragua…

  84. wmconnolley says:

    > your comment about the economy being fine but the natural world being ruined

    Actually that’s your scenario not mine; it is my rephasing of your “it’s possible that the impact is major and yet the economic damages were minor”.

    > I am confused by how the system would respond if key things (like agriculture) became a vanishingly small part of the global economy

    This is mt’s puzzlement. I’ve already pointed you at part of the answer: https://wmconnolley.wordpress.com/2015/05/13/agricultural-land-value-as-a-percentage-of-gdp/ – it has already happened in part.

  85. WMC,

    Actually that’s your scenario not mine; it is my rephasing of your “it’s possible that the impact is major and yet the economic damages were minor”.

    Yes, but it doesn’t really capture what I was getting at. I was suggesting that if we end up in a scenario where the natural world is ruined and yet economists could still claim that the damage was only a few percent of GDP, then maybe the latter has lost any real meaning.

    This is mt’s puzzlement. I’ve already pointed you at part of the answer:

    Yes, I get that and – having re-read Noahpinion’s post – another answer also seems to be that no economist is actually considering time periods beyond about 100 years (possibly even that 100 years is somewhat stretching things). If so, then agriculture will clearly not become a vanishingly small part of the economy on this timescale. Still not quite sure what happens on longer timescales, but maybe the answer is simply “we don’t really know”.

  86. Ben McMillan says:

    Although much of the world has a large surplus of water, food, and places to live that are suitable for habitation, these resources are shared very unevenly. So ‘agriculture is a small part of world gdp’ is true but glosses over the large fraction of the population living near subsistence (about 25%).

    So it happens frequently that lots of people to starve while other people use grains to make fuel for their cars. Shortages of food are very relevant in the current world. You just can’t treat the world like a homogeneous lump and get the right answer.

    As VV said, for an IAM this looks like a small issue. Even in countries/regions where there are a large proportion of subsistence farmers, the economy is dominated by the activities of rich people, so you just see a small drop in GDP when all the crops fail. In reality something more dramatic happens.

  87. wmconnolley says:

    Err yes that’s true in the current world, so has nothing to do with GW. And of course the fraction and absolute number of those living in absolute poverty is declining.

  88. Ben McMillan says:

    The issue is the combination of people living in food/water/land insecurity and effects of global warming.

    You can’t just wave away poor people or assume they won’t exist in some shiny future world. One might hope the proportion living in poverty in 2050 is very small but that is at best a risky bet to make. And the people who will lose out aren’t getting a say in making the bet.

    Note that historically speaking the current period is an aberration, growth-wise, so 100 years of extrapolation seems unwise.

  89. angech says:

    “the people who will lose out aren’t getting a say in making the bet.”
    “You can’t just wave away poor people”
    ” historically speaking the current period is an aberration, growth-wise”
    One is tempted to make the observation that the last 100 years is when the population has expanded or exploded by possibly 4 times. The increasing warmth of near 1C did not seem to cause any problems in food production, water availability or procreativity.
    The opposite in fact.
    If you do not get a say in the bet chances are that you will lose.
    “You just can’t treat the world like a homogeneous lump and get the right answer.”
    If the world was a homogeneous lump the right answer would actually be obvious, I quite agree with you knowing that you meant you should not ……

  90. Dave_Geologist says:

    Re the physicist and economist mt: they should have read Larry Niven’s Ringworld series.

    Easy solution: move your planet into interstellar space like the Puppeteers did, 1/10th of a light year from its star, and just have a small fusion source to provide light and very little heat. Except, IIRC, they ultimately had to impose birth control when the planetary population got to a trillion. Still, moving out came in handy when their Sol-like star went red giant. And the technology was re-purposed to let them flee the galaxy when they learned of the Core Explosion (which Niven explained by a cascade of supernovas, but would nowadays probably have re-written as an Active Galactic Nucleus – the Milky Way has gone quasar, we just don’t know it yet).

    Or build a Ringworld like the Pak did. You just have to invent scrith , a material with tensile strength comparable to the strong nuclear force and near-infinite thermal conductivity, and stick giant radiator fins on the outside. Mind you, they has to cannibalise all the planets in the system as feedstock for the scrith-making machine.

    I’m sure our economist could insert something like that into an IAM at an appropriate point in the future.

  91. JCH says:

    Growth is families. It’s not business. Growth is essential for a society to have families. When people stop having them, growth will disappear. See Japan. So, for those who have children of child bearing age, the trends are pretty evident. Great grandparents had huge families. Grandparents had large families. Aunts and uncles and parents averaged two kids, or less. Children, nieces and nephews, great jobs, big houses, really nice cars, no kids at all. At best, they are discussed.

  92. Willard says:

    > And of course the fraction and absolute number of those living in absolute poverty is declining.

    Absolute poverty has been fixed at around 2 bucks per day. While this measure features in common talking points, explaining how it’s related to growth is unclear. After all, nobody can really live a life we can understand with two bucks per day. Certainly not the homeless from the industrialized world.

    Thy wiki tells me that most of the 1990-2010 decline happened in China, India, Pakistan, and Vietnam. Could be linked to sweat shops. Intriguingly:

    [T]he number of people living in extreme poverty in Sub-Saharan Africa rose from 290 million to 414 million over the same period.

    https://en.wikipedia.org/wiki/Extreme_poverty

    All in all, getting to zero extreme poverty is as symbolic as staying under 1.5C.

  93. Willard says:

    Good news everyone. Elizabeth Anderson has been named a 2019 MacArthur “genius grant” fellow:

    In the summer of 1979, Elizabeth Anderson, then a rising junior at Swarthmore College, got a job as a bookkeeper at a bank in Harvard Square. Every morning, she and the other bookkeepers would process a large stack of bounced checks. Businesses usually had two accounts, one for payroll and the other for costs and supplies. When companies were short of funds, Anderson noticed, they would always bounce their payroll checks. It made a cynical kind of sense: a worker who was owed money wouldn’t go anywhere, or could be replaced, while an unpaid supplier would stop supplying. Still, Anderson found it disturbing that businesses would write employees phony checks, burdening them with bounce fees. It appeared to happen all the time.

    […]

    The bank experience showed how you could be oppressed by hierarchy, working in an environment where you were neither free nor equal. But this implied that freedom and equality were bound together in some way beyond the basic state of being unenslaved, which was an unorthodox notion. Much social thought is rooted in the idea of a conflict between the two. If individuals exercise freedoms, conservatives like to say, some inequalities will naturally result. Those on the left basically agree—and thus allow constraints on personal freedom in order to reduce inequality. The philosopher Isaiah Berlin called the opposition between equality and freedom an “intrinsic, irremovable element in human life.” It is our fate as a society, he believed, to haggle toward a balance between them.

    In this respect, it might seem odd that, through history, equality and freedom have arrived together as ideals. What if they weren’t opposed, Anderson wondered, but, like the sugar-phosphate chains in DNA, interlaced in a structure that we might not yet understand? What if the way most of us think about the relation between equality and freedom—the very basis for the polarized, intractable political division of this moment—is wrong?

    https://www.newyorker.com/magazine/2019/01/07/the-philosopher-redefining-equality

    No wonder I never liked Berlin.

  94. mt says:

    I see nobody has taken the bait on my speculation above. I guess it is because it doesn’t make sense – I find it hard to credit my own argument that people would be better fed if food were more expensive. Oh well.

    So those of us who find the infinite growth on a finite planet prospect implausible ought to come up with a failure mechanism – how would the system fail if real actual things formed a vanishingly small proportion of the economy?

    My answer presumes something that isn’t automatic – that this world where most of what we trade is essentially made of fantasy would result in a concentration of “wealth” (whatever that means in this fantastic economy).

    Certainly if stuff is vanishingly important, unskilled labour also is. I think my speculation unconsciously assumed as much. But perhaps we could patch this together with a universal basic income. Then poor people could allocate necessities and rich people could go on trading fantasies to their hearts’ content?

    That said, there is this: https://en.wikipedia.org/wiki/Garrett_relation#/media/File:Garrett_Relation.tif

    So far there’s little evidence of this dematerialization.

  95. Michael 2 says:

    “equality and freedom have arrived together as ideals.”
    Freedom (1) is at the starting gate and equality (2) waits at the finish line. Only some that start will reach the finish line.
    1. For purposes of this metaphor I define freedom as similar to choice, acting and not being acted upon. As nearly everyone is acted upon it is more of a theoretical thing.
    2. I don’t have a definition of equality and neither do you.

  96. mt says:

  97. izen says:

    @-W
    “Much social thought is rooted in the idea of a conflict between the two. If individuals exercise freedoms, conservatives like to say, some inequalities will naturally result.”

    The conflict between freedom and equality is largely down to defining freedom as a matter of consumption of goods and services. This is not an exclusive definition, but is certainly the dominant one in modern capitalist systems and is particularly strong in the US. In any society where there is a scarcity of goods and services the exercise of the ‘freedom’ to have a wide choice of food, housing and occupation is rationed by government validated tokens of access to these assets.

    Until relatively recently in many societies women did not have the freedom to own material goods or access to services, they had to depend on a male ‘guardian’ to mediate. Even the personal access and use of the government ration tokens was not socially acceptable. The dowry system makes this explicit.

    The Golden Rule, or tit-for-tat suggests an ethical constraint on the ‘freedom’ to impose harm on others, and by extension a moral limit on communal governance limiting individual freedom unless it causes harm to others.The freedom to own a gun is justified in some societies as a means of defending the ownership of goods and services, in others the freedom is curtailed because the rate of suicide and murder is seen as exceeding the incidence of effectively defending the individual freedom to access scares resources.
    But historically many societies dealt with the problem of scarcity by restricting the ability to use government ration cards, otherwise known as money, to a small proportion of the population. Many were forced to generate the assets enjoyed by Free-men as slaves or serfs.

    In a post-scarcity society (see Ian M Banks ‘Culture’) there can be equality of freedom.
    But when scarcity of assets is present, freedom becomes measured by how much money you can obtain either by work, or by owning at least a share in the means of production.

    Arbeit macht frei

  98. mt says:

    Case in point: https://news.mongabay.com/2019/09/global-fisheries-deprive-local-communities-of-key-nutrients-study-finds/

    I think part of the confusion may result from the fact that “impoverished” peoples were until recently pre-monetary. They acquired resources through networks of mutual obligation, not through money. See, pretty much, all of cultural anthropology, and also David Graeber’s book Debt.

    So when food is cheap for everybody in the economy, a price is imposed for people who used to live outside it. They used to live, perhaps not comfortably, but fairly reliably, on zero dollars. Now they have to buy food with two dollars and compete with the damn beef cows for it.

  99. angech says:

    “Absolute poverty has been fixed at around 2 bucks per day“.

  100. Ben McMillan says:

    The issue is whether IAMs provide a useful prediction of the future: but they don’t even do a very good job of explaining the 20th century.

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4978232/
    ‘the economic growth model used by a prominent BC-IAM had little predictive power over the 20th century’

    This is from 2016: the economic components of IAMs are described as ‘largely untested’!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.