Paul Campos points to this post where political analyst Nate Silver writes:
If I’d told you 10 years ago a president would seek re-election at 81 despite a supermajority of Americans having concerns about his age, and then we’d hit 8% inflation for 2 years, you wouldn’t be surprised he was an underdog for reelection. You’d be surprised it was even close! . . .
Trump should drop out! . . . Biden would lose by 7 points, but I agree, the Republican Party and the country would be better served by a different nominee.
Campos points out that the claim that we “hit 8% inflation for 2 years” is untrue—actually, “Inflation on a year over year basis hit 8% or higher for exactly seven months of the Biden presidency, from March through September of 2022, not “two years.” It did not hit 8% in any calendar year”—and I guess that’s part of the issue here. The fact that Silver, who is so statistically aware, made this mistake is an interesting example of something that a lot of people have been talking about lately, the disjunction between economic performance and economic perception. I don’t know how Nate will respond to the “8% inflation for 2 years” thing, but I guess he might say that it feels like 8% to people, and that’s what matters.
But then you’d want to rephrase Nate’s statement slightly, so say someting like:
If I’d told you 10 years ago a president would seek re-election at 81 (running against an opponent who is 77) despite a supermajority of Americans having concerns about his age, and with inflation hitting 9% in the president’s second year and then rapidly declining to 3.5% but still a concern in the polls . . .
If Nate had told me that ten years ago, I’m not sure what I’d have thought. I guess if he’d given me that scenario, I would’ve asked about the rate of growth in real per-capita income . . . ummm, here’s something . . . It seems that real per-capita disposable personal income increased by 1.1% during 2023. These sorts of numbers depend on what you count (for example, real per-capita GDP increased by 2.3% during that period) and what is your time window (real per-capita disposable personal income dropped a lot in 2002 and then has gradually increased since then, while the increase in GDP per capita has been more steady).
In any case, economic growth of 1 or 2% is, from the perspective of recent history, neither terrible nor great. Given past data on economic performance and election outcome, I would not be at all surprised to find the election to be close, as can be seen in this graph from Active Statistics:
The other thing is a candidate being 81 years old . . . it’s hard to know what to say about this one. Elections have often featured candidates who have some unprecedented issue that could be a concern to many voters, for example Obama being African-American, Mitt Romney being Mormon, Hillary Clinton being female, Bill Clinton openly having had affairs, George W. Bush being a cheerleader . . . The age issue came up with Reagan; see or example this news article by Lou Cannon from October, 1984, which had this line:
Dr. Richard Greulich, scientific director of the National Institute on Aging, said Reagan is in “extraordinarily good physical shape” for his age.
Looking back, this is kind of amazing quote, partly because it’s hard to imagine an official NIH scientist issuing this sort of statement—nowadays, we’d hear something from the president’s private doctor and there’d be no reason for an outsider to take it seriously—and partly because of how careful Greulich was to talk about “physical shape” and not mental shape, which is relevant given Reagan’s well-known mental deterioration during his second term.
The 2020 and 2024 elections are a new thing in that both candidates are elderly, and, at least as judged by some of their statements and actions, appear to have diminished mental capacity. When considering the age issue last year (in reaction to earlier posts by Campos and Silver), I ended up with this equivocal conclusion:
Comparing Biden and Trump, it’s not clear what to do with the masses of anecdotal data; on the other hand, it doesn’t seem quite right to toss all that out and just go with the relatively weak information from the base rates. I guess this happens a lot in decision problems. You have some highly relevant information that is hard to quantify, along with some weaker, but quantifiable statistics. . . . I find it very difficult to think about this sort of question where the available data are clearly relevant yet have such huge problems with selection.
Both Biden and Trump were subject to primary challenges this year, and the age criticisms didn’t get much traction for either of them. I’m guessing this is because, fairly or not, there was some perception that the age issue had already been litigated in earlier primary election campaigns where Biden and Trump defeated multiple younger alternatives.
Putting this all together, and in response to Nate’s implicit question, if you had told me 10 years ago that the president would seek re-election at 81 (running against an opponent who is 77) despite a supermajority of Americans having concerns about his age, and with inflation hitting 9% in the president’s second year and then rapidly declining to 3.5% but still a concern in the polls, then I’d probably first ask about recent changes in GDP and per-capita income per capita and then say that I would not be surprised if the election were close, nor for that matter would I be surprised if one of the candidates were leading by a few points in the polls.
What about Nate’s other statement: “Trump should drop out! . . . Biden would lose by 7 points, but I agree, the Republican Party and the country would be better served by a different nominee”?
Would replacing Trump by an alternative candidate increase the Republican party’s share of the two-party vote by 3.5 percentage points?
We can’t ever know this one, but there are some ways to think about the question:
– There’s some political science research on the topic. Steven Rosenstone in his classic 1983 book, Forecasting Presidential Elections, estimates that politically moderate nominees do better than those with more extreme views, but with a small effect of around 1 percentage point. When it comes to policy, Trump is pretty much in the center of his party right now, and it seems doubtful that an alternative Republican candidate would be much closer to the center of national politics. A similar analysis goes for Biden. In theory, either Trump or Biden could be replaced by a more centrist candidate who could do better in the election, but that doesn’t seem to be where either party is going right now.
– Trump has some unique negatives. He lost a previous election as an incumbent, he’s just been convicted of a felony, and he’s elderly and speaks incoherently, which is a minus in its own right and also makes it harder for the Republicans to use the age issue against Biden. Would replacing Trump by a younger candidate with less political baggage be gain the party 3.5 percentage points of the vote? I’m inclined to think no, again by analogy to other candidate attributes which, on their own, seemed like potential huge negatives but didn’t seem to have such large impacts on the election outcome. Mitt Romney and Hillary Clinton both performed disappointingly, but I don’t think anyone is saying that Romney’s religion and Clinton’s gender cost them 3.5 percentage points of the vote. Once the candidates are set, voters seem to set aside their concerns about the individual candidate.
– Political polarization just keeps increasing, which leads us to expect less cross-party voting and less short-term impact of the candidate on the party’s vote share. If the effect of changing the nominee was on the order of 1 or 2 percentage points a few decades ago, it’s hard to picture the effect being 3.5 percentage points now.
The other thing is that Trump in 2016 and 2020 performed roughly about as well as might have been expected given the economic and political conditions at the time. I see no reason to think that a Republican alternative would’ve performed 3.5 percentage points in either of these elections. It’s just hard to say. Trump is arguably a much weaker candidate in 2024 than he was in 2016 and 2020, given his support for insurrection, felony conviction, and increasing incoherence as a speaker. If you want to say that a different Republican candidate would do 3.5 percentage points better in the two-party vote, I think you’d have to make your argument on those grounds.
P.S. You might ask why, as a political scientist, I’d be responding to arguments from a law professor and a nonacademic pundit/analyst. The short answer is that these arguments are out there, and social media is a big part of the conversation; the analogy twenty or more years ago would’ve been responding to a news article, magazine story, or TV feature. The longer answer is that academia moves more slowly. There must be a lot of relevant political science literature here that I’m not aware of . . . obviously, given that the last time we carefully looked at these issues was in 1993! I can read Campos and Silver and social media, process my thoughts, and post them here, which is approximately a zillion times faster and less effortful than writing an article on the topic for the APSR or whatever. Back in the day I would’ve posted this on the Monkey Cage, and then maybe another political scientist would’ve followed it up with a more informed perspective.
P.P.S. In a followup post, Campos introduces a concept I’d not heard before, the “backup quarterback syndrome”:
Nate Silver has fallen for the backup quarterback syndrome, which is the well-known fact that, on any team that isn’t completely dominating its competition, the backup quarterback tends to be the most popular player, because fans can so easily project their fantasies onto that player, since the starting quarterback’s flaws are viewed in real time, while the backup quarterback can bask in the future glory attributed to him by optimism bias.
I disagree with Campos regarding Nate here: it’s my impression that when Nate expresses strong confidence that a replacement Republican would do much better than Trump, and speculates that a replacement Democrat would do much better than Biden, Nate is not making a positive statement about Ron DeSantis or Gretchen Whitmer or whomever, so much as comparing Trump and Biden to major-party nominees from the past. Nate’s argument in support of the backup quarterback is based on his assessment of the flaws of the current QB’s.
That said, I like the phrase, “backup quarterback syndrome.” It does seem like a fallacy. It’s probably been studied (maybe not specifically in the football-fan context) in the heuristic-and-biases literature.
Even using official CPI real median personal income trended down since 2019
https://fred.stlouisfed.org/series/MEPAINUSA672N
Real median household income down even stronger
https://fred.stlouisfed.org/series/MEHOINUSA672N
Daniel:
Based on the literature, the usual comparison focuses on the change during the year leading up to the election. Things being bad since 2019 would not reflect badly on the current incumbent. That said, your comment illustrates an important general point, that different measures of the economy will give different results, and there are no strong theoretical or empirical reasons to prefer one measure over others.
We just don’t have the 2023 numbers yet to put on that graph, but I’d be surprised if it weren’t down from 2022 in terms of median household. Maybe flat.
Ideally we’d see for example each of the deciles of income, and adjusted by the basket of goods each decile actually buys.
It’s not like we don’t have about 500 people employed by the Bureau of Economic Analysis with a budget of ~$100M, plus the BLS with 2100 people and a budget of $655M and the consumer expenditures survey, average price series, and the ACS,
Does anyone else think that’s the most hilarious thing you’ve ever heard? almost $800M and we don’t have deciles of real income?
Daniel
You can find some information closer to what you are asking for:
https://www.bea.gov/news/2024/personal-income-and-outlays-april-2024
https://apps.bea.gov/iTable/?reqid=19&step=2&isuri=1&categories=survey&_gl=1*114ah75*_ga*MTY0NTg5MzE2NS4xNjc2OTg3Mzgw*_ga_J4698JNNFT*MTcwMjQ3MjU2Ni4xMDQ4LjEuMTcwMjQ4MDAyOC4wLjAuMA..#eyJhcHBpZCI6MTksInN0ZXBzIjpbMSwyLDNdLCJkYXRhIjpbWyJjYXRlZ29yaWVzIiwiU3VydmV5Il0sWyJOSVBBX1RhYmxlX0xpc3QiLCI0MjkiXV19
I realize these are not exactly what you are asking for, but it isn’t like there is a plot by the BEA to not show recent data or not provide distributional data. I suspect some of the limitations concern their reluctance to release preliminary data if it is subject to change when they have experience that change is likely.
What I haven’t looked for is access to a full database that would permit construction of your own data series. As it is, however, there are so many series available that many desired time series can be accessed. I get your point about the budgeted resources for collecting this data, but in many ways they do a good job of providing data of reasonable quality and consistency. If we spend too much money on that, it is probably due to the inefficiency of the public sector. But your dismissive attitude I find off-putting.
As we’ve discussed before, I don’t know if you are suggesting a conspiracy against providing valuable information or incompetence or both. And I’m beginning to worry that your evident frustration mirrors the way that much of the population feels. The reason I worry about that is because it is what leads many to irresponsible choices such as voting for Trump, or not believing that there is such a thing as truth. I don’t think those are true for you, but the frustration sound quite similar to me.
Dale, I wouldn’t say a “conspiracy” it’s just … the dominant paradigm is entrenched and deeply flawed and one reason it’s entrenched is it protects the powerful (by allowing policy to be built off measurements that preference what the rich and powerful want).
It’s obvious and been obvious for years, and I don’t mean a couple years I mean my entire lifetime, that Economics as practiced and official economic measures are a lie. The world doesn’t work like textbook economics models, and not just a little bit but sometimes entirely *opposite* (see my links elsewhere on 1970’s surveys showing unit costs usually decline with increased output)
Shockingly the BEA income tables you show aren’t even labeled for units… I assume it’s millions of dollars because all the numbers are things like 0.0524 (52.4k dollars?) and they only publish them since 2015.
I guess 9 years is better than nothing, but we’ve been collecting the Consumer Expenditure survey since 1984!
My impression is the whole thing is just self-serving self-dealing. The Democrats exist as a make-work welfare system for college educated political science, economics, and law students, either as career politicians or as bureaucrats while Republicans exist as a rent extraction mechanism for big businesses, the ultra wealthy, and military and police.
Just no-one is in it to make the kids my children go to school with healthy, happy, and financially comfortable people.
so, why not vote for Trump (at least he’s entertaining)?
I pose that because I think that is part of the current thinking. I’m not in disagreement with your characterization of the 2 parties, but I am fearful of throwing it all out. I’d like to think we’ll get something better, but the odds are greater that we will get things we’ve seen before – autocracy, dictatorship, abuse of power. We have elements (too many) of these already, but it can easily get worse. Trump is the epitome of self-serving self-dealing. A ruler for our times.
Dale, I don’t find him entertaining and I believe under him we are likely to get if not civil war at least massive increases of violence. So I prefer Biden, because I think he will keep the outright street violence to a minimum. It’s a low bar.
I agree with nearly everything you (Daniel Lakeland) say in this thread. I have a slight disagreement with
“The Democrats exist as a make-work welfare system for college educated political science, economics, and law students, either as career politicians or as bureaucrats while Republicans exist as a rent extraction mechanism for big businesses, the ultra wealthy, and military and police.”
Both the Democrats and the Republicans are rent extraction mechanisms for big business, the ultra wealthy, and military and police. The difference is which businesses they favor: the Republicans shill for energy and manufacturing corporations whereas the Democrats shill for finance and tech.
Clyde:
The industries favored by the two parties have changed over the years; see the book Golden Rule by the political scientist Thomas Ferguson. Here’s what we wrote in Red State Blue State:
Clyde, I don’t disagree with you at all, consider it an adopted modification to my rant. Also, jeez, what’s the chances someone would agree so enthusiastically with me? I appreciate the support.
Daniel –
I started a long response to you the last time we discussed this issue – then got busy and distracted and kind of lost momentum to finish it. I hope to get back to it.
In the meantime you might find this discussion of some interest:
https://open.spotify.com/episode/6JKs9d43xMYdDh9RXVdtse?si=4m87Lv2OTg-MQBf4cWBIZw
Joshua, I had never read that The Atlantic article mentioned in the summary of the podcast, and yet, there it is, the same basic quantities I keep talking about, housing, food, healthcare, childcare/education, transportation. They don’t mention them all, but they mention housing, healthcare, and education/childcare.
Interestingly I’m working through the ACS 5 year data from 2018-2022 right now, looking at trying to estimate a “subsistence” level of housing + utilities so I can make a dimensionless ratio of income to subsistence income. I’m not doing the “right” thing (with a massive Bayesian model) I’m just doing the easier things (like filtering and taking averages). It’s not really very clean, because the data is a bit weird. There are retired people who own their home and live in a home that’d cost $6000/mo to rent, and then there are families of 4 who live in a similar house across the street and pay the $6000/mo rent. There are people in rural Appalachia who have no rent but there’s 12 people in a shack…
Different people were affected differently. But maybe what I should do is take my own advice and focus on families of two parents and 1-6 children because that experience is what will be remembered by the children in 15 years.
Anyway, I enjoy this kind of stuff, but I do find it frustrating because of the difficulty of doing the “right” thing (ie. the large Bayesian model with say 5-10 parameters per county and all 3000+ counties with hierarchical priors etc)
Daniel
You might find this proceeding of interest: https://www.cpuc.ca.gov/affordability. It’s your state and they have put a considerable effort into this (though I am far from agreeing with their efforts). Much of their data comes from the ACS, although I don’t think they do anything about the relatively large sampling errors in the data. But you may find it interesting to get involved.
https://postimg.cc/v4NW40wM
Shows a plot from some preliminary calculations. This is dimensionless ratio of income/(housing related costs) across the entire country. with housing related costs being an estimate of “minimum” costs for each state and size of household.
Minimum in this conception being the median cost (rent/mortgage, property tax, electricity, fuel, water, etc) for people with High School Diploma or less living in housing with exactly 4 rooms (regardless of household size).
You can see that actually the modal ratio actually shifted upwards. But still, modally, those people are making about 2 to 2.5 times this minimum housing level. This doesn’t yet include an estimate of food costs, transportation costs, child or healthcare costs… so add those together, it seems like you’re going to have the modal household making something closer to 1 on this kind of measure, meaning, “barely scraping by”.
That’s just sad to me. Of course some people are making 5-10x this kind of minimum, so not everyone is doing poorly, it just seems bad if we have “most” people close to the edge at all times.
Daniel –
Here’s the transcript for the podcast.
https://www.nytimes.com/2024/06/07/opinion/ezra-klein-podcast-annie-lowrey.html?showTranscript=1
Joshua, that discussion is exactly 100% my thesis. For example the CPI doesn’t look inflated, basically because people have been adapting, they’ve been cutting back, moving out a little farther from their preferred location, not going to the doctor for their injured shoulder, getting their parents to watch the kids some days instead of sending them to a daycare near their work… etc etc… so on paper everything looks fine, but in reality we’re just tracking people’s actual expenditures which are contracting in real terms… so it doesn’t show up in the numbers somehow. People are moving into their parents house and taking care of their 85 year old parents instead of working… on paper those people have no housing expenses, but also no income, and no choices in their life.
that’s exactly her thesis, and it’s mine too. And I say it’s been going on as a trend since 9/11 or a little before, like since 2000. And to a lesser extent it’s been going on since the 1970’s in some ways. it’s not like a recent phenomenon but it’s become enough of a problem of accumulation that people are actively having it in their face now: “I have few choices, I can barely hold together rent, food, transport, healthcare, kids education, childcare, etc and that’s with two incomes and a better than median household income”… and people hate it. Can’t blame em.
The only problem as I see it, later in the transcript, is that she doesn’t take the final leap… So CPI is shit, wages “adjusted for inflation” haven’t gone up, they’ve likely gone down since 2019 continuously (in terms of ability to buy basic needs). What has happened is people who owned houses refinanced, and that cut their housing costs, and gave them some breathing room, and that’s sticky, but if you rent I think you probably got screwed.
I’m working on the data analysis, need to calculate in food and healthcare and taxes… but by the time I’ve done food, healthcare, taxes, transport and housing… (Income / That Index) is going to look bleak I think. The mode of that statistic might be quite near 1. It might be 40% of the US that is on the boundary of poverty/insolvency as defined by that. I don’t know.
I’m gonna do my best and maybe we can discuss it here on the blog or something.
If you want to follow along I’m releasing pieces of the analysis a little bit at a time on Mastodon @[email protected]
I found much of this post a distraction from its central question: would replacing either or both of these candidates move the needle enough to make a difference? If it is a close election, then the answer would almost certainly be “yes.” But to get any sensible answer to the question, we have to know something about who the replacement would be. I find very few politicians that would justify me voting for them on the basis of their character and/or their policies. Very, very few (close to zero). Perhaps personality would persuade a large enough share of voters to switch or decide to turn out – by personality I mean fame, glitter, stage presence, appearance, etc. The truth is that my voting has become almost entirely an opportunity to vote against someone. I think that is true for many people, although there may be a Trump exception to that (though it baffles me as to why).
Dale:
Because the election could be very close, just about any change could be decisive. That’s a point we’ve discussed in the past, so I didn’t bring it up here.
To me, the central questions are: (a) Nate’s claim that, given Biden’s age and inflation two years ago, we should be surprised that the election “is even close,” and (b) Nate’s claim that a different Republican candidate would get 3.5 percentage points more of the popular vote.
Given the current level of political polarization, I’m inclined to be skeptical about both Nate’s claims. But there are a lot of moving parts here, hence all the distractions in my above post.
In the past I’ve slammed political analysts for getting the facts wrong (for example here); in this case, Nate’s statements are conditional and can’t be proved to be true or false, so my discussion has an inherent fuzziness.
My own view is that voting is a terrible idea, and it was a mistake to move away from the original form of democracy… random assignment.
https://en.wikipedia.org/wiki/Sortition
Of course modern random assignment should be randomly selected uniform from all people, not just a few rich people. But beyond that, pick a big enough sample and you’re just about guaranteed to be doing better than any system relying on career politicians.
Let’s say 400 people and a 4 year term, interleaved 1/4 turnover each year…
I’ve always like that idea. I even extended it to include a random sample of everyone, including infants, prisoners, etc. There are some logistical issues I can’t solve, but I far prefer random selection to the process we now have.
I’m not sure infants is worthwhile, but you might argue that a parent of the infant could serve as a proxy, which makes parents more likely to be chosen, which isn’t a bad thing as they directly represent both themselves and their offspring.
It’d be a hard sell to put convicted murderers, rapists, domestic assaulters, or major frauds in congress, I think it’d be fine to exclude active prisoners.
But you could easily have a rule that allowed the congressional body itself to reject anyone with a sufficient vote. Say 75-80% vote to remove someone.
Logistically I don’t think this is a major difficulty. You’d publish open source code for a cryptographic random number generator and a database of all SSNs in sorted order, then you’d seed the random number generator by shuffling a deck of cards with a mechanical device and fanning them out, and have 20 people pick randomly from the fan.. put the string describing the chosen cards in order into the RNG and then have it run… everyone could verify what happened.
You would pick enough that when people didn’t want to serve you’d just move down the list.
What depresses me, as a natural scientist, is what is not being discussed here. Scientists have known that atmospheric CO2 was increasing, and the effect of that on global temperatures, since at least the 60s; I learned about this in 1971. In 1978, Mercer’s prescient article in Nature laid out one of the scary consequences: a 5 m increase in sea level (look up Mercer West Antarctic Ice Sheet on Google Scholar to find the article – it gives a good description of the basic physics). Now, with recent news about the Thwaites Glacier, we have confirmation that the disaster Mercer warned about seems to well underway. For all his faults, Biden has pushed significant efforts to deal with climate change, and Trump is all “Drill, baby, drill! Jeez, don’t you guy have children or grandchildren to worry about?
Economists have known that the assumption they made in the first chapter of Econ 101 that marginal cost of production should rise was wrong for like 100 years with good solid survey data from the early 70’s, and purely practical reasons to believe that it’s a stupid model (ie. engineers design factories to operate at high efficiency when near their capacity, and businesses plan for higher capacity than what they currently need) and yet it’s still a fundamental idea in economics
https://profstevekeen.substack.com/p/profit-maximization-in-the-real-world?utm_source=publication-search
Is it any wonder that public trust in govt is at an all time low? https://www.pewresearch.org/politics/2023/09/19/public-trust-in-government-1958-2023/
it should be! government is in the business of lying to the public and selling itself off to the highest bidder, which is usually Exxon or Google or Microsoft or Disney or Halliburton or Lockheed
Wow that blog from Steve Keen is great! I didn’t realize that the firmament of Neoclassical Economics relied so extensively on the assumption that marginal costs rise with output. I’ve long done profit maximization to help clarify issues in agriculture and natural resource economics and have literally never made that assumption, in fact usually the opposite. In agriculture, we see quite clearly that marginal costs per unit of production almost invariably decline with higher yields (output), with a few notable exceptions.
Let’s take a single factor optimization for simplicity, something like N fertilizer application. We see quite clearly a saturating *yield* response to additional N (with tons of under-modeled variability), but the cost per unit N is, *at worst*, constant. In reality, you tend to get a better rate buying in larger quantities and can reduce application costs per unit N with up-scaled equipment and/or heavier application rates (each trip of the tractor with a fertilizer spreader attached costs about the same in terms of labor and diesel, so may as well get more done per trip, etc.) all of which tends to a declining marginal cost with application rate.
It is alarming to me that the august establishment of high brow economic influencers seem to get this wrong in quite fundamental ways for both industry and agriculture.
Side bar, I’ve loosely followed Steve Keen’s work for a couple years now and mostly like what he’s up to quite a lot!
I disagree that this is inconsistent with the theory. Those predicted marginal cost curves decrease, *then* increase.
Hungry Helen’s cookie factory sounds like academia, which is essentially the government.
The issue is that a normal business won’t operate in the increasing marginal cost regime. Only governments and those heavily supported by govs (eg, education, healthcare, military contractors) would, or even could, do such a thing.
The theory has room for both.
Anon, yes it’s true if you go high enough in production it could rise again, but that’s not the regime business operates in. Therefore it isn’t relevant to questions about business behavior. But neoclassical econ assumes that this is how prices and production get set. If prices are high enough that you make money on the marginal item then make more items, reduce your price, and sell them for more total profit… Stop when the marginal price equals the market price.
This is a pretty fundamental argument in neoclassical econ, you’ll hear in like third year macroecon classes (at least when I took it in the 90s). It’s fundamentally flawed! The market price could never be set that way because in real world production as you make more it gets cheaper per unit. Eventually if you get near capacity you open a new factory and are back at the same regime.
Also fixed costs are usually an order of magnitude higher than the economy assumptions. (More like 30-40% than 3-4%)
Stupid phone keyboard…
*Third year microeconomics
Also, stop when the marginal *cost* equals market price.
Daniel
Clearly the idea that “you open a new factory and are back at the same regime” is similarly flawed. Factories take time to build, so in the meaningful short-run, you are stuck with the factory size you currently have. Now, I’m willing to accept the empirical reality that most firms are operating in the range where their marginal costs are either constant or declining, but that does not necessarily invalidate the market supply curve. Provided that there are numerous firms (which will not be the case when marginal costs are significantly declining – conditions well recognized in neoclassical economics) with different cost structures, then the relative ability of each firm to increase output without incurring rising marginal cost might be swamped by the variation in production across firms. It can remain true that each firm will try to sell “as much as it can” despite the fact that it would have trouble if, for example, it could conceivably double sales (need that new factory and need it now), but that significant increases in sales would require multiple firms to increase production. If those firms have higher marginal costs than other firms, then the industry supply curve will effectively be upward sloping.
I haven’t completely digested Keen’s argument – and I am quite open to wholesale criticism of neoclassical theory – but I am not convinced at this point. I was trained in neoclassical theory and find much of it unsatisfying, but Keen’s critique seems overdone to me. And without a clear substitute, it leaves me wondering, now what? For example, if we decide there is no such thing as a supply curve, and if we don’t accept downward sloping market demand curves, and if we don’t believe all markets are monopolies (though I think concentration is a bigger problem than most economics do, I think we can all agree that the entire economy is not monopolistic), then how does he propose that prices get determined?
In fact, my biggest issues with neoclassical theory are with welfare economics – the policy implications derived from the theory. These can be largely divorced from much of the theory itself. I think that much of neoclassical theory dealing with the operation of markets (including what happens when many important assumptions are relaxed, such as perfect and symmetric information) yields valid insights. I am less inclined to believe that the theory is sufficient for reaching policy conclusions as a result.
Dale, thanks for your comments. I’m far from an expert in neoclassical stuff. I consider myself as more or less like a guy with an undergrad minor in econ, having read outside of school and taken some basic micro and macro classes in undergrad. Maybe I’ve got the equivalent economic specific background of 4-5 undergrad level classes. So I appreciate your greater background. Of course, on the other hand I’ve studied a lot of dynamical systems and differential equations and data analysis in other contexts, so I think I have more complimentary background.
My impression is Keen says there’s no such thing as equilibrium, and desires an economics built on dynamics and differential equations. I more or less agree. Of course such things could have temporary equilibria, but generally outside forces will be constantly moving things around. What he wants is a mechanistic rule for actually how factories typically respond to price and sales data.
Yes, of course building factories take time, but his point is also that people design and start building them ahead of time, and have generally a lot of capacity available for all but the hottest new products.
So, let’s imagine a world in which there’s a commodity item and 1000 factories for it. Let’s say bolts since I use that same commodity in an ODE dynamic model I built for a book I’m writing, in that model it’s counterfeiters invading a market to seek rent selling low quality crap to people who absorb the cost by implementing their own testing procedure, but for now let’s just assume people selling normal bolts.
A factory is operating at cost below the current market price, but they’re not going to ramp up production immediately today, because they know they can’t sell all those bolts they’ll have to warehouse them. So they hire some more sales team, and those guys try to find buyers. As they find buyers they start to get more orders, and they ramp up production. But now advertising/marketing is an additional cost. I don’t think it’s well accounted for by calling it a cost of production though, because it’s independent of how many you produce. You might be at the beginning of selling your new product and choose to utilize very high marketing while only making a few products per day, in hopes that your sales ramp up, and you can start to use your factory capacity to finally make money, etc. Or you might be well established and withdraw marketing just leaving enough to keep your established customers reminded that you’re a good source of what they need.
In the example we’re discussing, unless people start using bolts they didn’t need before, what’s actually happening is that your factory is stealing customers away from some other marginal factory. One way to compete is to offer kind of better “service”. Like, you’ll ship bolts ahead of time on a predictive basis, and you’ll extend credit to companies that have variable sales, and you do a little bit better quality testing, etc. But, one thing the sales team might do is in fact offer discounts to good customers who buy in large quantity, so yes, as time goes on prices may fall. Let’s suppose you’ve been producing a lot and competing with another firm and prices have been falling, but your factory is still competing because marginal prices are declining for you. Let’s say the other factory is selling and therefore producing fewer and fewer, so through time their cost per bolt is rising and their profit is falling.
Your competitor sees they’re headed for a problem, and decides to invest in ways to make their factory more efficient, and/or buys some of their own marketing talent. If it doesn’t work, they may go out of business, and then your factory gets a rush of new orders, which may make you near capacity and then you buy out their older factory and upgrade it to use your better processes, or build a new factory of your own.
on the other hand, perhaps the competitor figures out a new market for bolts, so they start selling to people you don’t know about, and their cost drops.
So, anyway there’s several factors involved:
1) Information distribution in the market (affected by marketing)
2) Price
3) cost of production (both fixed and average per-unit)
4) Marketing cost (a separate category of cost with independent controls)
5) Other firms demand
Note that as bolt prices drop, companies may begin to design new products so that they use say 2 bolts instead of say 4 screws. So through time again even at fixed price, the demand for the product can change in time.
to me, neoclassical economics thinks entirely about processes that occur on the order of milliseconds to a week (high frequency trading or up to stores restocking shelves weekly). But there are lots of processes that take months to years to play out (designing new products, opening new factories, reworking the efficiency of existing factories, marketing campaigns, related product development, competing on quality (a smaller phone charger, a smaller battery pack, a higher capacity battery pack, a paint that dries faster, a paint that coats better).
The idea that “at equilibrium, the price is equal to the marginal cost of production” is a complete fiction. absolutely rotten science.
My knowledge is far from complete, but I am aware of attempts to replace neoclassical theory with
The Energy Basis of Value: Georgescu-Roegen
The Labor Theory of Value: Karl Marx (this one is different than the others)
Anti-equilibrium: Janos Kornai
An Evolutionary Theory of Economic Change: Nelson and Winter
And a theory I can’t recall the name of by a professor I had as an undergraduate (whose name I can’t recall – my problem, he was brilliant and I liked his theory a lot)
All of these got serious attention but none got much traction. Of course, we can chalk that up to all the usual reasons why fields resist change. But they also shared a failure to convincingly demonstrate that neoclassical theory needed to be abandoned rather than incrementally modified. I have used some of these ideas (and others) in my own critiques/departures from neoclassical theory. But I am not convinced that the whole throwing the whole thing out is a step forward. I do personally think that economic theory is fundamentally at odds with ecological systems, but I don’t know what to do with that. As we have often said, all theories are wrong, some are useful. I do think neoclassical theory (at least micro) is useful. But I’ll propose a new saying: all useful theories can and will be misused. And neoclassical micro is a case in point.
I don’t think the issue for me or Keen is theory of value, it’s lack of dynamics and insistence on equilibrium.
Suppose we’ve got a ball on a spring, and it’s idealized to have no air resistance etc. So I start it oscillating… where’s the equilibrium?
It just doesn’t exist! The ball literally oscillates forever.
In economics, how does the price get set? It’s a process that goes on in time. The marketing people call their contacts and say “we’re having a sale on bolts if you buy 1000 we give you a free box of 100” and sometimes people take that deal, and sometimes they don’t. Or, consider the quantity to be made, the factory runs for a few days and warehouses the results on site, and someone is in charge of packing and shipping boxes out… Every week or so they do a count of how many boxes are in the warehouse, and if the count is too low they run an extra shift that week, and if it’s too high they run shorter shifts. Their goal is “keep an amount on hand that’s between 500 and 1000 boxes at all times” and they operate like a thermostat… turn on an extra shift when they hit 500 and turn off a shift if they hit 1000.
Occasionally maybe they get a really big order, and they have say 30 days to ship all the boxes, and so they track that order’s outstanding boxes each day and see if they’re on track to ship all of them by 30 days, and if not they add shifts until the linear extrapolation line comes in at day 29 or 28 because no-one likes to be behind they’d rather be ahead of the game.
The factors that determine price and quantity are *not* the intersection of a supply and demand curve, they’re factors like the rate at which information about the availability of bolts from different suppliers diffuses through the network of people who want bolts and the efficiency of the middlemen at finding new takers, and the outcomes of tests like offering a discount for a week to see how many more bolts will be ordered, and whether a new product is being made by a new customer starting this week.
A more reasonable model might be something like
dB/dt = k (AvgStock – Bstock)
dBstock/dt = B – Sales
Sales = N * F(Price,t)
dN/dt = G(AdvSpend) + H(K)
dK/dt = Q(AdvSpend) * (Pop – K)
where B is the rate of bolt production, AvgStock is the desired average quantity in the warehouse and Bstock is the actual quantity in the warehouse. F is a nonlinear function of Price, G is a nonlinear function of instantaneous advertising spend rate, H is a function of the number of people who know about your firm, K, and Q is a nonlinear learning rate in the population as a function of advert spend.
https://fred.stlouisfed.org/graph/?g=1oYav
Here’s a graph of the global price of copper as a fraction of GDP (times some constant) per metric tonne (1000 kg or 1 Mg). I’m pretty sure you once told me you worked on the question of predicting copper prices… so tell me why should we decide this was ever in equilibrium rather than continuously in flux like a wheel bouncing over some rough terrain according to constantly shifting pushes and pulls?
Certainly some of those rapid dips and rises were due to specific shocks, like the 2008 financial crash, but even outside of that, there was a long downward trend from 2011 to 2016 and then oscillations from 2016 to now, that looks like something well outside of equilibrium to me.
The point is, the price is not set in any sense by “the marginal cost of producing a tonne of copper”. I just refuse to believe that the marginal cost of producing a tonne of copper continuously declined from 2011 to 2016 and then oscillated back and forth. The technology didn’t change, the wages didn’t change much (in ratio to GDP so, some kind of inflation adjusted). I’d guess that after the 2008 crash, investment in tech stuff drove a need for more copper for wiring and such, prices went high, for a decade mines worked extra time extracting large profits, while prices fell, then oscillations as demand changed a bit, the COVID pandemic drove prices up a bit because mines cut back production for exogenous reasons (ie. to prevent covid or to comply with laws or because shipping was affected) and then as things got ironed out the price declined further.
at no point was “marginal cost of production” a factor in any of this.
Yes, I did win a copper price forecasting contest. I made no assumption regarding whether price was in equilibrium or not – I based it on time series analysis of past prices + additional variables that should affect it. You are mistaken when you focus on the marginal cost as THE price. Copper is not a perfectly competitive industry and we would not expect price changes to always be reflecting marginal cost changes. Demand still matters, as does industry organization and strategy. The more you think about those things, the more relevant dynamic models become – but I would stop short of rejecting all models that lack a dynamic framework. Comparative statics can sometimes suffice – it depends on how rapidly things adjust (which differs across markets greatly).
My point is not about this particular market. You seem to want to reject supply/demand/equilibrium as a model because you think it fails to describe the copper market. Do you really want to say it doesn’t adequately describe any market whatsoever? I believe your answer is yes, as I think you have indicated something like this in the past. That is not my position. Rejecting all neoclassical micro theory because the textbook version is static would be a poor idea. Rejecting it because you believe costs are irrelevant to market prices would be a reason to throw out the whole theory, but I don’t believe that to be true.
Dale, I think equilibrium concepts are useful as approximations to an underlying dynamic theory.
Let me give an example, we go back to the weight on the spring. I hold the spring in my hand, with the weight on the other end. Then I move my hand up and down slowly, say it takes a minute to move 3 inches… Now the ball moves up together with my hand, as if they were connected by a stiff bar, because I move so slowly that the spring doesn’t stretch. But if I move my hand quickly (0.1 seconds to move 3 inches), then the spring stretches or compacts and then the dynamics are much more complicated and the ball will oscillate.
If Econ wants to describe dynamical models and then say “in these particular situations, we expect the static approximation to be sufficient” then fine. But if they never make this dynamic model and then just assert facts about equilibria implying some stuff, then this is not a scientific procedure, it’s more like a religion. There’s an asserted “demand curve” and a “supply curve” and the supply curve is determined in part by increasing marginal cost of production, a fact that is patently false, and literally no-one can measure the supply curve, it’s not even in principle measurable, and supposedly it’s moving around in time, but the price is always in equilibrium or close to it by basically faith…. it’s angels on pins, it’s holy trinity, it has nothing to do with science.
on the other hand, with a dynamic model we could potentially say that static approximations are ok. For example 1999 to 2003 the copper price changed slowly but was mostly constant. That might be a time period where we could describe it as mostly near to an equilibrium.
We have equilibria in physics for thermodynamics. For example steam in a boiler. If we add heat energy we convert mass from liquid to gas in a closed container and the pressure changes are related to the heat added. But if we open the blowoff valve all of a sudden pressure is changing dynamically, temperature is changing dynamically, steam condenses to water, some work is done on a piston, etc. It’s a dynamic process.
Neoclassical econ fails because it doesn’t have an underlying dynamic theory that clearly connects to reality it just has some kind of approximation that is unfalsifiable (you can’t observe the demand/supply curves).
The Washington Post has a somewhat related story today: https://www.washingtonpost.com/climate-environment/2024/06/12/trump-federal-scientists-climate-environment/
Both Biden and Trump were already president though. Under one the warming flatlined, under the other it skyrocketed:
https://www.columbia.edu/~mhs119/Temperature/Ts_variousMeans.png
In fact this correlation between warming vs party of US president goes back many decades. There is probably nothing more correlated with the tempetature record.
Anon:
Yes, this is related to the well-known correlation (see for example this from Larry Bartels) between political parties and economic growth. The economy has on average grown faster when Democrats have been president and, at least in the short term, economic contraction is good for the environment. It’s hard for this issue to get much political traction: Republican politicians aren’t out there asking for votes on the grounds that they will contract the economy and thus reduce global warning.
What happened to temperature in a given four years is meaningless, given the times scales involved with warming and the natural variability in temperature. Remember the famous ‘pause’ in warming? It resulted from adding a variable stationary signal to a steadily increasing signal. For reliable information on climate change, go to RealClimate.Org.
I agree with Dale (in his comment below) that the time scales make climate change hard to deal with politically, but that is not a good reason not to try.
Sounds like there is no evidence voting democrat has any effect on climate change. In fact such evidence could not exist even in principle (according to your theory of climate).
Yet, it is seemingly very important to vote democrat due to this topic.
I’d guess its more like Andrew proposed, there is some connection between presidential party, economic activity, and the temperature record.
Then we see all the same pattern all the time of “war on drugs” -> more drugs; “war on terror” -> more terrorists -> “war on poverty” -> more poverty -> “war on cancer” -> more cancer. I could go on.
So I would expect “war on climate change” -> more climate change.
Anon:
I disagree with you, not necessarily on the question of which party will be better at remediating climate change, but on the idea that all we can do is reason based on general cases (in your framing, “‘war on climate change’ -> more climate change). There’s some physics here! Also, even more generally, there was a war on crime, now there’s less crime; there was a war on cancer, now many cancers can be treated; there was a war on the Axis, the Axis lost; etc. Some policies work, some don’t, it’s complicated, that doesn’t mean that it’s a bad idea to try.
Anyone who thinks that we will do anything meaningful about climate change until it’s way too late is deluded. It’s already way too late. We could have done something in 1971 after The Limits to Growth, but we didn’t.
CO2 output from coal per capita globally has been more or less constant since 1910 at 2 tons/yr/person +- 0.25t
https://ourworldindata.org/explorers/co2?facet=none&country=CHN~USA~IND~OWID_WRL&Gas+or+Warming=CO%E2%82%82&Accounting=Territorial&Fuel+or+Land+Use+Change=Coal&Count=Per+capita
US dramatically turned down its coal output starting in 2000 ish but only because China dramatically built more capacity.
Ocean Surface Temps have been at world-record levels by a considerable margin **every day for over a year** (since March 2023)
https://climatereanalyzer.org/clim/sst_daily/
Air surface temperatures are similarly way out of the normal range:
https://climatereanalyzer.org/clim/t2_daily/?dm_id=world
So there’s every evidence that as of today it’s already way too late. So my prediction has come true already, we won’t have done anything until after it’s way too late.
When it comes to all-fossil emissions, they were constant between 1920-1950 but ramped up to a new more or less constant rate by mid 1970’s and have been 4.4-4.7 tons / person / yr ever since.
https://ourworldindata.org/explorers/co2?facet=none&Gas+or+Warming=CO%E2%82%82&Accounting=Territorial&Fuel+or+Land+Use+Change=All+fossil+emissions&Count=Per+capita&country=CHN~USA~IND~OWID_WRL
While of course population has exponentially increased.
The Limits To Growth predictions basically suggest the human population will collapse to about half its peak population over the next 70 years. This is more or less a qualitative or semi-quantitative prediction.
Here’s the archive copy of the book. See page 124 for the “standard run” their kind of “best guess of what happens if nothing changes”
https://web.archive.org/web/20240104230738if_/https://collections.dartmouth.edu/content/deliver/inline/meadows/pdf/meadows_ltg-001.pdf
It’s entirely likely that timeframe or quantity is off by some considerable amount, but it seems entirely plausible we hit something like 9-10B by 2050 and then decline to 4B by say 2100 or 2200 after major reductions in fertility as well as the effect of floods, farming failures, and war.
Most of the death will be concentrated in places like India, Indonesia, African countries, and soforth. Fertility decreases are already making china population decline. Europe and US will need to get their shit together to avoid war, they’ve already got one going on in Ukraine, and a second one in the middle east.
The key fact to remember is **both Democrats and Republicans in the US will fight any meaningful change tooth and nail**. Not that there’s no difference between them, but neither one is going to do what is needed to change this, because what’s needed is changes in fossil fuels which would themselves cause unspeakable death. No-one wants to say “sorry Indians, you all have to die because we’re cutting your access to energy by a factor of 10”. And the only way to enforce that on the Indians would be nuclear war. So it’s not going to happen. The US and Europe already have enough solar and wind capacity they could cut fossil fuels by 90% and avoid massive death, but India, it’d just be a bloodbath, China similarly.
Nothing of consequence will happen because we’re on a major flat plateau of policy where no matter what we try we produce death. Stay on course? Death. Major output changes? Death. The future is Death. Those who come out the other side may see a better world, but only after 2-3 billion people die and the rest stop having children because conditions are too awful.
Thanks for coming to my TED talk. I’ll go enjoy some sort of lunch, while I still can.
Daniel Lakeland: “Nothing of consequence will happen because we’re on a major flat plateau of policy where no matter what we try we produce death. Stay on course? Death. Major output changes? Death. The future is Death. Those who come out the other side may see a better world, but only after 2-3 billion people die and the rest stop having children because conditions are too awful.”
What is this bold prediction based on? Just the (very stylized) models in the 1972 book?
Climate change will indeed likely result in deaths, but also technology and global wealth are improving, so human well being can still be improving on average over time (https://www.theatlantic.com/science/archive/2021/11/how-bad-will-climate-change-get/620605/), even as things are worse than they would be with earlier, better investments and policy changes. And existing fertility declines haven’t really been because things are awful, but precisely because they are good and people are making different choices.
Dean, the prediction is based on the general behavior of models of the type from the 1970’s book, yes. But also the observed behavior of the world in the last say 20 years. Some examples:
1) Here’s 500 years of Cod catches in Eastern Canada https://ourworldindata.org/grapher/long-term-cod-catch
2) Here’s fish catch in the UK https://ourworldindata.org/grapher/fish-catch-uk
3) Insect populations are decreasing drastically https://en.wikipedia.org/wiki/Decline_in_insect_populations
4) The world has been losing 5-7M hectares of forest a year for decades. https://ourworldindata.org/grapher/annual-change-forest-area?tab=chart&country=~OWID_WRL
And of course all the fossil fuel burning evidence from previous post… not only is per capita output more or less constant rather than declining, but human population is exponentially growing still.
The capacity to feed and house and protect from the elements is dependent on energy, and arable land.
Arable land use per person has declined, but slowly and stabilized over the last decade across the world:
https://ourworldindata.org/grapher/arable-land-use-per-person
Half of the worlds usable land is already used in agriculture: https://ourworldindata.org/global-land-for-agriculture and technology is not reducing the per person usage rapidly as shown above.
So, at similar growth rates, within a couple decades we will be at the point where we need 100% of land as agriculture to feed the population. Being within 50% of that point is already a dangerous situation (today).
Meanwhile there’s deaths from climate issues, such as heat… https://apnews.com/article/record-heat-deadly-climate-change-humidity-south-11de21a526e1cbe7e306c47c2f12438d
If heat deaths are growing exponentially in AZ, then how is India handling them? We know they’re way way underreported.
Then there’s water availability. Mexico city is subsiding 10-20 inches per year https://www.washingtonpost.com/climate-environment/2024/06/04/mexico-city-sinking-water-crisis/ and will run out of water. Similar loss of water supply is well documented in places like South Africa https://www.nationalgeographic.com/science/article/partner-content-south-africa-danger-of-running-out-of-water. CA miraculously got a lot of rain the last 2 years, but was headed for empty reservoirs before.
What’s not appreciated is that a lot of these situations have wildly nonlinear responses. Imagine CA water supply continued on its crash course for another 5 years? As long as there’s some water people survive fine. But when there’s NO water, shit hits the fan in a big way. The US might compensate by transporting water, but Africa and India and such? not so much.
If we cut fossil fuels by a massive factor, we can probably produce stable climate, but doing so would obviously crash out the world agricultural output, and we need ~ 2000 calories / person / day and there is **nothing** technology can do about that. At best it can reduce the land and energy use per calorie, but we see that’s not continuing to happen, and a lot of those improvements have already occurred
Yes, people have reduced their fertility rate as their lives have gotten better and their children are healthier and not dying of childhood diseases. But also in a period of time where agriculture and transportation were flourishing, not being stressed by the necessity of severe restrictions on fossil fuel emissions.
In the absence of *severe* fossil fuel emissions reduction, un-developed areas face periods of severe heatwaves, floods, storms, etc. In the presence of severe fossil fuel emissions reductions they face starvation and massive decline in quality of life, while for a while the severe weather etc is still occurring (probably decades of high heatwaves built into the climate even if we cut CO2 emissions by a big factor today).
The most likely response is migration, from areas that are arid and hot or constantly flooded to areas where those problems are less, which means encroaching onto the existing 50% of the land area that’s in use for agriculture, causing problems for food production as well.
There’s no good solution, because all of the issues take decades to play out, are hard to predict, and the value of any intervention is extremely uncertain so as to have a hard time knowing which ones actually will benefit.
In the end, if we agree that to avoid catastrophic heatwaves and crop loss in floods and lack of fresh water we need to be closer to the CO2 output of say 1930, that was around 1.5 Billion Tonnes / yr, and we’re currently at 30 Billion Tonnes / yr, so we need a factor of 20 reduction. That’s not a small amount. (got those numbers from estimating derivatives off this curve)
https://ourworldindata.org/explorers/co2?facet=none&Gas+or+Warming=CO%E2%82%82&Accounting=Territorial&Fuel+or+Land+Use+Change=All+fossil+emissions&Count=Cumulative&country=CHN~USA~IND~OWID_WRL
A factor of 10 reduction might be possible in developed countries through solar and wind and storage (water pumping, stacking blocks, batteries, electric trains full of dirt chugging up mountains… whatever). A factor of 10 reduction in India seems unlikely to result in anything but massive hardship. The production of infrastructure to enable that 10x reduction in India would require massive building projects and solar panels and nuclear plants and soforth, projects that require *temporarily* *increasing CO2 output* and by “temporarily” I mean, for 30 years.
Economists often talk about improvements to technologies, but we simply don’t have clear improvements to technologies that can support food production and clean water production at global scale with current populations, in the presence of a 90% reduction in carbon output, nor is there a fast timescale to transition to energy technology that could create the 90% reduction in carbon output. Shit is baked into the system right now as far as I can tell.
Daniel
You are even more of a pessimist than I am. I don’t share many of the concerns you raise. I was impressed by The Limits to Growth when it came out, but for it pathbreaking work, not its accuracy. Time and again, projections of doom have not played out. I have more confidence in the ability of technologies to handle a number of the problems you refer to (better ways to grow food, replacing meat with non-meat, etc.). Where I share your pessimism is in the distribution of the impacts. Global warming will cost many lives – they will be lives in poor places. How those people will respond to these threats I do not know. Even if the damage is contained to them, they must weigh on whatever moral conscience we have left.
Then there are the less obvious, nonlinear, and greater uncertainty effects: species loss, pandemics, global war. These worry me greatly. But lumping them all together runs the risk of outright rejection of the position as a “doomsday prophet.” Hey, that might be another synonym for “torment executioner.”
Dale. Limits to Growth isn’t about accuracy of numerical predictions, it’s about qualitative behavior. For example, you have a weight on the end of a spring, with a damper. For some amount of damping, the system will decay slowly to equilibrium, for smaller amounts of damping it will oscillate while decaying, for no damping it will oscillate forever. I don’t need to match the behavior of the system to 7 decimal places to know that if you overly damp something it won’t oscillate at all… if it’s gonna oscillate, I don’t even need to know the frequency exactly, I can be satisfied with “this system will oscillate and decay on a timescale of minutes”.
Similarly, LTG tells us that *of course* there’s a carrying capacity of the earth, and *of course* that carrying capacity is dependent on technology, and *of course* the factors that matter for capacity are things like food, water, and pollution, and of course our ability to overcome problems is a function of the energy we can expend per person and the technology used for it all and of course the timescales of interest for change are ~100 years or so.
We can grow a lot of food right now because climate isn’t too weird, and we have fossil fuels. But we can’t have both forever… either we reduce fossil fuel output by some factor of 70,80,90% or something, or we will continue to have trending-worse climate and food production will suffer, as will fresh water, and destruction of housing viability through desertification or flooding or windstorms.
Then there’s timescale issues. How long does it take to build new dams? How many locations do we have we can do it in? How long does it take to convert farming to sheltered farming that is impervious to flooding? How long does it take to build solar farms and wind farms? How long does it take to approve and construct nuclear plants for overnight power?
If a king waved a wand and said “go forth and maximize the resilience of human populations” what timescale would that take? Building dams is at least 5 years at accelerated paces. Building solar power plants is decades, as is wind farms. Building a nuke plant is likely 10 years and ignores the waste handling and security implications.
If you imagine a curve of water per capita for a region that looks like 10/(t+1) where t=0 is now, and 1 is let’s say the water needed for a person in a year, then as you go through time, water related deaths are zero all the way out to time 10, and then somewhere between 10 and 15 years they go way up in a vertical boundary layer.
Sure you could build desalination plants, but, on what timescale? How big? And how much fossil fuels to do the construction? and what distribution system?
The point isn’t that 10/(t+1) is reality, it’s just that there’s a boundary layer and for most regions of the world have been away from that boundary layer for all of history. Some regions of the world will hit that boundary layer in the next 20 years and cause massive problems. We don’t know where or when.
Just this week here’s what the weather looks like: https://www.windy.com/-Temperature-temp?temp,36.984,-107.688,5 Las Vegas is at 107F. What happens when Las Vegas goes to 120F every day for 60 days? Now what happens when that happens in Pakistan, India, or rural China? That kind of thing is *happening today*. New Delhi will be 107F at its local noon or so. That will be true or even worse every day for as far as we can predict (out to Friday the 21st on windy.com) and has been true for a while now (a week or two?).
https://en.climate-data.org/asia/india/delhi/new-delhi-30/
says that “climate” suggests highs of 85 to 95F in June. So we’re talking about 20F higher than normal every day for weeks.
I can easily see the most developed parts of the world adapting by migration, technology, and soforth. I don’t see that being viable for Brazil, Argentina, India, Indonesia, Pakistan, rural China, Sub Saharan Africa, etc… and we’re talking about a population of maybe 4B in those regions.
We aren’t going to see human extinction, but it’d be surprising if the eventual carrying capacity is much about 5B people. Over the next hundred years, population will likely peak and then decline, and then eventually technology investment over 50 or 100 years together with population decline will create a very reasonable place to live. Unfortunately that’ll be my great-grandkids… and only if we avoid a conventional world war on the scale of WWII.
This is a tangent from the OP, but…
Climate activists concluded from the experience of failed carbon tax legislation at the federal level that direct curtailment of fossil fuels was politically unattainable. This left investment and regulation. Aside from the scientific side (which in my view clearly illuminates the need for economy-wide restriction of carbon emissions), the political assessment is at best mixed. Investment in green energy is not opposed by business (in general), which gives it political traction, and Biden did manage to move this agenda forward, but it has been less of a PR plus than expected. Perhaps this is because of the lead time in investment planning. Meanwhile, detailed sectoral regulation (e.g. prohibiting new gas stoves) has shown itself to be vulnerable to “populist” attack. This has been made clear in Europe, and I doubt the US will prove to be all that different. We need a serious rethinking of the politics of climate mitigation.
Peter,
You are making the argument for Biden’s Build Back Better proposal, which would have been a big deal, and, but for Sinema and Manchin, would have passed.
On the scientific side (the arithmetic of the remaining carbon budget), the investment-centered strategy is incomplete. But now we’re looking at the politics. In the leadup to the IRA, there was a widespread belief that a green investment initiative would be a political winner. Well, not yet. Had BBB gone through, do you think that would have made a noticeable *political* difference?
Would any of it make any meaningful difference at all?
The US is 360M out of 8B people or ~5%
If you wiped the entire US population off the map with one finger-snap the global CO2 consumption would drop by on the order of .7 tons per person, which would be about the level of 1992 or so, and still dramatically more than we need to be at (clearly the climate crisis was already a problem in 1992).
My guess is emissions rates would have to drop to say the level of 1900 or so to eventually equilibriate at something long-term viable.
That’s about 1.2T/person/year or about 25% of current output rate. This could be accomplished by cutting the population in half, and cutting the output rate of the remaining population in half.
Over the next 100 years I fully expect population to drop to 4-5B, both through death and through fertility reduction. Wind and solar energy production together with overnight storage and maybe Nuclear could then create a sustainable future.
That’s assuming the loss of 90% of all fish in the ocean and 75% of all flying insects (which has already occurred) doesn’t lead to even worse agricultural issues than the climate directly (ie. storms, heat, desertification)
Andrew: you have mischaracterize inflation.
1) YoY Inflation was over 6% for 18 consecutive months
2) YoY inflation was over 8% for 6 consecutive months and peaked over 9%
2) annual inflation was over 6% for two consecuitive years
ttps://www.usinflationcalculator.com/inflation/current-inflation-rates/
Inflation did not “decline rapidly to 3%” – but even if it had, inflation at 3.4% (today’s number) is still twice the historical average. Long-run inflation in the US has averaged about 1.2% annually since 1800, so 6% inflation is 5x the long-run average
https://macrohive.com/wp-content/uploads/2021/11/inf2.png
Real wage growth was negative for 2021-2022.
https://www.stlouisfed.org/on-the-economy/2023/feb/nominal-wage-growth-individual-level-2022
This chart shows that in Feb-2023, the two lines crossed, but are now converging again. But the combined area under the curves since the start of Biden’s presidency is clearly negative, where it had been positive for nearly the entire previous decade:
https://www.statista.com/statistics/1351276/wage-growth-vs-inflation-us/
Beyond that, and beyond the tremendous noise in the inflation and wage data, “the economy” isn’t just wages and inflation. What about housing costs and availability? What about the still-exploding homeless problem? What about crime? Why are you so sure problems in one area (e.g., crime, homelessness) don’t smudge into people’s views about other areas, like the economy?
On net, while both you and Nate Silver misrepresented inflation, nate was quite a bit closer than you are.
Chipmunk:
I wrote, “inflation hitting 9% in the president’s second term and then rapidly declining to 3.5%,” I mean to write “second year,” not “second term,” so I went into the above post and fixed that. In the BLS time series cited by Campos, year-on-year inflation peaked at 9.1% in June 2022, declined basically linearly to 3% in June 2023, and has been between 3.1% and 3.7% since then. So I think that my statement, “inflation hitting 9% in the president’s second year and then rapidly declining to 3.5%,” is accurate. And I so disagree with you that in my statement I misrepresented inflation (aside from my silly error of typing “second term” instead of “second year”). Your three statements at the top of your comment are completely consistent with my summary!
I agree with you that there are many different ways of measuring economic performance. As I wrote in my above post, “These sorts of numbers depend on what you count (for example, real per-capita GDP increased by 2.3% during that period) and what is your time window (real per-capita disposable personal income dropped a lot in 2002 and then has gradually increased since then, while the increase in GDP per capita has been more steady).” I focused on recent changes in real per-capita GDP and disposable personal income because these are measures that I’ve seen in election forecasts.
You ask, “Why are you so sure problems in one area (e.g., crime, homelessness) don’t smudge into people’s views about other areas, like the economy?” I have no such certainty! The economic statistics that researchers and analysts, including myself, use in forecasts are only part of the story. People often talk about the unemployment rate too, but I didn’t get into that because the purpose of this post was not to consider all the possibly economic and social statistics out there. Different voters care about different things, and these predictions have noise, which is why, contra Nate, I don’t think we should be surprised that the polls are currently close. I’m not saying the polls had to be close, just that giving the mediocre (not-terrible but not-great) recent economic performance, and our usual view that the election is influenced by economic performance in the year or so preceding the election, a close election is not surprising.
In my post I am not trying to make a case that the economy is going very well, or very poorly. The situation is complicated, and, as is nearly almost always the case, one can make a fundamentals-based argument in either direction. Even if all the economic statistics were better than they currently are, one could argue that there are fundamental economic problems from deficit financing, along with big bills coming up to pay for pensions, health care, and adaptation to climate change. You can come at this from the left, right, or center and have big concerns about economic policy and performance. Some of this has come up in the above comment thread. From the other direction, there are analysts coming from left, right, and center who argue that the economy is performing reasonably well and that the above concerns are overblown. In my post I’m not trying to argue either of these, nor am I trying to split the difference; I’m just saying that, if you want to talk about economic performance as a predictor of the vote, you should not be, in Nate’s words, surprised that the election “was even close.” I did point Nate to this post so maybe he’ll share his perspective on all this.
Biden doesn’t seem to have “diminished mental capacity” to me. I suspect he does take stimulants before a big event such as the State Of The Union. But when he has that energy boost, he sounds completely on top of things. The problem is that on his bad days – where he’s obviously tired with little energy to spare – well, his age can be heard in his voice. It’s disturbingly frail and feeble at times, and it just doesn’t look good to the public.
Trump, now, it’s hard to say. To be scrupulously fair to the guy, he’s under an enormous amount of stress, putting it mildly. That sort of situation would likely cause anyone to sound cranky and ranty. But it doesn’t seem that he’s all that different in cognitive quality from how he’s always been (low bar, I know).