{"items": [{"author": "Alexei", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901208428962", "anchor": "fb-901208428962", "service": "fb", "text": "Thanks for writing this up. I've heard this argument a lot recently as well, so it's good to get some more concrete numbers behind it.", "timestamp": "1505978052"}, {"author": "Marcus", "source_link": "https://plus.google.com/111675838261170541573", "anchor": "gp-1505999434190", "service": "gp", "text": "A couple of comments:\n
\n1) This is not meant to be a support of the original paper, which I think has some framing problems (namely, showing only the absolute impact of an action on emissions, rather than the relative impact per unit effort or cost).\n
\n2) I think your range of \"value of carbon\" is very skewed. \n
\na) I'd argue that the former EPA SCC value is actually on the conservative end (it doesn't take into account a lot of climate change effects, such as changes to air quality or labor supply due to warming, both of which have been identified as large impacts, and the use of a constant discount rate rather than a Ramsey framework approach likely undervalues impacts in the distant future). \n
\nb) The use of marginal cost of abatement (based on Cool Earth) as a way to estimate the impact of emissions is, in my opinion, very misleading. This would be the equivalent to saying that we can measure the value of a human life by looking at the cheapest option available to save lives. So since Givewell estimates that the cost per life saved for Deworm the World is $900, this reasoning would lead us to find a value of a human life of only $900. \n
\n3) If we really want to determine the \"exernality\" impact of having a kid, it would be a complicated analysis involving all sorts of impacts. I would imagine that in a hypothetical labor poor, resource rich world, having an average kid would have positive externalities. In a labor rich, resource poor world, having an average kid would have negative externalities. My instinct is that we are a little closer to the latter than the former (where \"resources\" includes clean air, biodiversity, access to open spaces, and so forth, in addition to more traditional kinds of capital and mineral resources), but I acknowledge having not thought this through thoroughly. And of course, any specific child-having decision is a very personal decision for a given family, and shouldn't just be driven by externalities calculations. \n
", "timestamp": 1505999434}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901252141362", "anchor": "fb-901252141362", "service": "fb", "text": "Note: the discussion in https://www.facebook.com/.../permalink/1530657430323942/... is continuing and brings up some additional ideas. A key one is that if we put a lot of effort into reducing greenhouse gas emissions then the cost of offsets would climb dramatically, so using Cool Earth numbers for something as long-term as children probably doesn't make sense.", "timestamp": "1506005023"}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1506006427857", "service": "gp", "text": "@Marcus\n I'm not trying to answer this question, which would be an enormous amount of work (hence my \"whether it would be better for people to have more children or fewer children in general is not at all a settled question\") but instead argue that their conclusion doesn't follow even within their incredibly limited framework.", "timestamp": 1506006427}, {"author": "Kathryn", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901262435732", "anchor": "fb-901262435732", "service": "fb", "text": "One problem with having kids is that being a \"good\" parent seems to mean that you hope for as good or better than you had as a child. So in addition to requiring more resources just by existing, children become the catalyst for consuming more etc. I have seen 2 studies (long ago so I cannot cite the sources) one showing that poor children are not such a drain on resources because they do not consume as much...they walk to school, have one pair of shoes (if any) etc. whereas middle income or rich kids are much worse for the environment and the resource consumption. The other study showed that for poor families especially in rural areas around the world, where children are expected to work in the fields etc, a child that reaches about 9 years of age becomes an asset rather than a drain on family as they begin producing more than they are consuming. So this is a strong economic incentive to have as many children as possible.", "timestamp": "1506009458"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792", "anchor": "fb-901266397792", "service": "fb", "text": "\"... to avoid somewhere between $80 and $6,000 in yearly emissions!\"

This statement hides some complexity in what we're actually trying to avoid, right? We're talking about some amount of yearly emissions, per kid, for the kid's lifetime, right?

Then of course there's the kids' kids.

\"people have many other effects other than emitting greenhouse gases\" is a good point, but much of the rest of the post seems to suggest the stakes are small in a way that doesn't seem right.", "timestamp": "1506011216"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792&reply_comment_id=901274631292", "anchor": "fb-901266397792_901274631292", "service": "fb", "text": "→ \"Then of course there's the kids' kids.\"

The paper tries to account for that: \"For the action 'have one fewer child,' we relied on a study which quantified future emissions of descendants based on historical rates, based on heredity (Murtaugh and Schlax 2009). In this approach, half of a child's emissions are assigned to each parent, as well as one quarter of that child's offspring (the grandchildren) and so forth. This is consistent with our use of research employing the fullest possible life cycle approach in order to capture the magnitude of emissions decisions.\"

They're citing http://www.sciencedirect.com/.../pii/S0959378008001003 which I haven't read", "timestamp": "1506014747"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792&reply_comment_id=901275709132", "anchor": "fb-901266397792_901275709132", "service": "fb", "text": "→ Ok, then am I understanding right that the paper accounts for this as part of its argument, but your numbers ignore it? That could be part of the reason for the different takeaways, right?", "timestamp": "1506015072"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792&reply_comment_id=901276522502", "anchor": "fb-901266397792_901276522502", "service": "fb", "text": "→ Here's the paper, looking at it now: http://www.jefftk.com/murtaugh2009.pdf", "timestamp": "1506015216"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792&reply_comment_id=901278313912", "anchor": "fb-901266397792_901278313912", "service": "fb", "text": "→ Ok, I've skimmed the paper, and it's kind of silly. The emissions/year cost of having children looks like it's the emissions of all of your descendants over all time, in proportional to how related you are (1/2 weight for children, 1/4 weight for grandchildren, ...), divided by your life expectancy. For example, the emissions of one child in the US plus their share of their children's emissions and so on comes to 9441 tCO2e, which divided in two (two parents) and then by life expectancy at reproductive age gives you ~120 tCOTe/y.

The dividing by life expectancy is a distraction: let's look at that 9441 tCOTe. The only reason they get a bounded number for this is that they model humans as going extinct by reproducing below replacement levels. If reproduction levels were 2.01 instead of under 2, would they say the tCO2e cost of children was infinite? This is just really bad futurism, where you draw curves way out and don't consider anything else that might change (technology, evolution, etc)", "timestamp": "1506015947"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901266397792&reply_comment_id=901278423692", "anchor": "fb-901266397792_901278423692", "service": "fb", "text": "→ :( yeah", "timestamp": "1506016017"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268174232", "anchor": "fb-901268174232", "service": "fb", "text": "I'm also a bit confused about some aspects of the \"social cost of carbon\" calculations:

Does the \"3%\" in \"\"3% Average\" refer to a 3% discount rate? If I'm understanding that right, a 3% discount rate weights next 24 years roughly equal to the whole rest of the future.

I'm also confused about why the estimates increase. I'm surprised 2015 emissions are half as bad as 2050 emissions. They write:

\"estimates of the social cost of these greenhouse gases increase over time because future emissions are expected to produce larger incremental damages as physical and economic systems become more stressed in response to greater climatic change, and because GDP is growing over time and many damage categories are modeled as proportional to gross GDP\"

It seems that either (a) carbon must come out of the atmosphere much quicker than I thought, or (b) the EPA is thinking about marginal effects very different from how I would.", "timestamp": "1506011991"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268174232&reply_comment_id=901269815942", "anchor": "fb-901268174232_901269815942", "service": "fb", "text": "→ Oh, maybe the discount rate is just for converting to current dollars, not for actually caring about the future less than the present? Don't think it's that but I'm not sure.", "timestamp": "1506012913"}, {"author": "Nix", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268174232&reply_comment_id=901312625152", "anchor": "fb-901268174232_901312625152", "service": "fb", "text": "→ > Does the \"3%\" in \"\"3% Average\" refer to a 3% discount rate? If I'm understanding that right, a 3% discount rate weights next 24 years roughly equal to the whole rest of the future.

This is my understanding, and I think is pretty standard in governmental CBAs (whether it is right or not). For some discussion about how to deal with long term discount rates, see:
http://science.sciencemag.org/content/341/6144/349.full", "timestamp": "1506027474"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268174232&reply_comment_id=901315369652", "anchor": "fb-901268174232_901315369652", "service": "fb", "text": "→ It's been pointed out to me privately that most of the discount rate applied comes not from valuing the future less, but from the assumption of continuing growth, meaning we'll be richer in the future, making marginal consumption less valuable.", "timestamp": "1506028318"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268289002", "anchor": "fb-901268289002", "service": "fb", "text": "Regardless of why the estimates increase, it looks like you're using the 2015 estimates when really you should use estimates across the potential child's lifetime, right?", "timestamp": "1506012081"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268289002&reply_comment_id=901275344862", "anchor": "fb-901268289002_901275344862", "service": "fb", "text": "→ I think you're right. Their 2050 numbers are ~2x their 2015 numbers, but after 2050 you'd expect it to keep going up I guess? Though at some point we'll have figured out how to emit less, deal with high GHG levels, or gone extinct.", "timestamp": "1506015024"}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268289002&reply_comment_id=901275704142", "anchor": "fb-901268289002_901275704142", "service": "fb", "text": "→ Well, really that's too sharp a division. We might be dealing with it, say, in a way where that there's a significant cost per tCO2e.", "timestamp": "1506015071"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901268289002&reply_comment_id=901277914712", "anchor": "fb-901268289002_901277914712", "service": "fb", "text": "→ This is one of the things that makes me worry that either we're not interpreting the EPA's \"social cost\" numbers right or they're not calculating what we're interested in.", "timestamp": "1506015782"}, {"author": "Marcus", "source_link": "https://plus.google.com/111675838261170541573", "anchor": "gp-1506022811366", "service": "gp", "text": "I also retract 2a, because on re-reading, I realize you were using the 95% bound of the SCC, which is a decent high example. (2b still stands, though). \n
", "timestamp": 1506022811}, {"author": "Jeff Kaufman", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901392495092", "anchor": "fb-901392495092", "service": "fb", "text": "After in person discussion I'm more skeptical of the EPA numbers, and I don't know if they're supposed to represent cost to the whole world or just cost to the US. The right numbers would be similar to \"what would be the right price for a carbon tax, all things considered\" with there's probably some economics work on? Though tail risks from climate change might not be weighted highly enough?", "timestamp": "1506050897"}, {"author": "Nix", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901392495092&reply_comment_id=901394211652", "anchor": "fb-901392495092_901394211652", "service": "fb", "text": "→ They are global estimates. Some discussion on page 17 from the technical support document: https://19january2017snapshot.epa.gov/.../sc_co2_tsd...", "timestamp": "1506051763"}, {"author": "David Chudzicki", "source_link": "https://www.facebook.com/jefftk/posts/901187770362?comment_id=901392495092&reply_comment_id=902665808362", "anchor": "fb-901392495092_902665808362", "service": "fb", "text": "→ Haven't looked into very closely, but I'm still pretty confused by how this seems to imply global warming isn't that big of a deal.", "timestamp": "1506689297"}, {"author": "Eric", "source_link": "https://plus.google.com/118361373248657996863", "anchor": "gp-1506078450421", "service": "gp", "text": "@Marcus\n I don't understand your point in 2b. Aren't we trying to measure the \nmarginal\n cost (to the environment) of having an extra kid?\n
\n
\nIf so, the marginal cost of abatement seems like a very appropriate comparison!", "timestamp": 1506078450}, {"author": "Marcus", "source_link": "https://plus.google.com/111675838261170541573", "anchor": "gp-1506081536268", "service": "gp", "text": "If you are trying to decide where to spend your money to do the most good, then finding the cheapest marginal cost of abatement is useful. If you are trying to estimate how much damage you are causing, then the marginal cost of abatement gives you zero information. Let me try another example:\n
\n
\nI'm trying to decide whether to burn some coal in my stove. I know that burning that coal will create 1 ton of particulate matter. I also know that I could install a scrubber on my neighbor's smokestack for $10 that would eliminate 1 ton of particulate matter. Does this mean that the cost to the environment of my creating 1 ton of PM is $10? No! The cost is a function of the number of deaths and other impacts caused by my PM emissions. \n
\n
\nIf we already know the marginal cost to the environment, then it is useful to compare the marginal cost to the environment to the marginal cost of abatement. If the former is larger than the latter, then it means that we aren't doing enough abatement. But there is no a priori reason to assume that the two are equal, or that the latter tells us anything about the former. \n
", "timestamp": 1506081536}, {"author": "Eric", "source_link": "https://plus.google.com/118361373248657996863", "anchor": "gp-1506082252518", "service": "gp", "text": "@Marcus\n Gotcha, I see what you're saying.\n
\n
\nI now agree with you that the cost in terms of harm to the world may be radically different from the cost of abatement.\n
\n
\nDo you also agree with me that the cost of abatement is super decision relevant?\n
\n
\nIf I'm thinking about whether to have another kid, I could be deciding between:\n
\n
\n1) Have zero additional children.\n
\n2) Have one more kid.\n
\n3) Have one more kid and pay the cost of abatement.\n
\n
\n#1 and #3 have the exact same impact on the environment. If impact on the environment would have been my reason for choosing #1 over #2, I should just switch to comparing #1 vs #3. If the additional cost of abatement (on top of all the other costs and benefits of having children) is low enough, then I can do #3 and have an additional kid with clear conscience.\n
\n
\nDoes that sound right to you?", "timestamp": 1506082252}, {"author": "Marcus", "source_link": "https://plus.google.com/111675838261170541573", "anchor": "gp-1506085803682", "service": "gp", "text": "#3 is the general theory behind \"carbon offsets\" which are quite popular in some circles. In principle, it seems reasonable, conditional on a number of factors (one standard list for offsets is \"real, permanent, quantifiable, verifiable, enforceable, and additional\"). ", "timestamp": 1506085803}]}