And furthermore, would you be willing to make the change if it meant making changes to the water supply?
From The Big Think
Adding trace amounts of lithium to the drinking water could limit suicides. Two studies, a recent one in Japan and an older one in Texas, have shown that this naturally occurring substance, used as a psychotropic drug to combat bipolar disorder, could have beneficial effects for society: Communities with higher than average amounts of lithium in their drinking water had significantly lower suicide rates than communities with lower levels. Regions of Texas with lower lithium concentrations had an average suicide rate of 14.2 per 100,000 people, whereas those areas with naturally higher lithium levels had a dramatically lower suicide rate of 8.7 per 100,000.
A gated version of the study is available here. There are other benefits beyond a reduction in suicides. From the Texas study:
[T]he homicide rates during the decade studied were consistently lower (by 30-50%) in the high- than the medium- or low-Li counties. The incidences of robbery, burglary, theft and the total crime rates were also lower in the high-Li counties, but the respective differences involving assault were statistically significant.
Drug use was also lower in the area with the highest natural occurring lithium concentrations.
One question I've never seen answered is: How much would it cost a low-lithium city to artificially boost their concentrations to the highest level? We have experience adding trace elements to the water supply (fluoride), so it can certainly be done.
I decided to do a very rough back-of-the-envelope type calculation to see what the cost per person in lithium is.
I am not a chemist, but my understanding is that a city that wanted to increase the elemental lithium in the water supply would use lithium carbonate (if I'm mistaken here, please let me know). The largest bulk price I could find for lithium carbonate is 50kg for $3842.82 (Source), which works out to $76.86/kg or 7.686 cents per gram. This should only be used as a rough estimate – the price may be lower (because a city would be buying in far larger quantities) or higher (because the city would need a grade with less impurities than the one quoted). But for a rough back-of-the envelope calculation, it works.
A city with no-to-little elemental lithium would need to add 70 micrograms/L of elemental lithium to the water supply. Since we're adding lithium carbonate (not pure lithium), we would need roughly 200 micrograms/L. (For reference, there are a million micrograms in a gram).
The average Canadian domestic user uses just over 100,000 L of water a year (Source). At 200 micrograms/L, we would need to add roughly 20 grams per person of lithium carbonate for a total cost of $1.53 per person, or $153,000 per 100,000 people.
The city of Toronto has 3.3 murders/100,000 people (Source). A 30% reduction in this rate would lower it by 1 murder per year per 100,000 people. If our rough back-of-the-envelope calculations are correct and the lithium carbonate method works like the Texas study suggests, $153,000 buys us one less murder. That does not take into account the reductions in rapes, suicides, drug use or thefts.
Will it work? I don't know. It seems like it would be worthy a pilot study or two. Although those levels of elemental lithium are believed to be safe, there may be side-effects we are not considering. There are ethical considerations as well, but it is hard to make a case that adding fluoride to the water supply is ethical but lithium is not – and we've been adding fluoride to drinking water for over half a century.
Edited to add: The Texas study grouped counties into four groups: High lithium (70-160 micrograms), Medium (15-60), Low (0-12) and Low excluding big cities (0-12). If we compare the first group to the fourth, high lithium counties have, per 100,000 population:
- 5 fewer murders
- 5 fewer suicides
- 22 fewer rapes
- 310 fewer burglaries
- 751 fewer thefts
(all of which are statistically significant at the 5% level). There may be a multitude of other factors causing this than lithium levels. There is no guarantee that artificially raising lithium levels to get a county to go from low-lithium to high-lithium would provide these results. But given a (very rough) cost of $153,000 per year, isn't it worth investigating?
Worthwhile Canadian Initiative 
“Given the practice of the cities of Vancouver and Victoria of dumping raw sewage into Georgia Strait, remind me not to drink the water next time I’m in White Rock.”
Vancouver’s water comes from the North Shore mountains. We may be special, but we haven’t figured out how to live off of salt water yet.
“The Precautionary Principle restated:
Keep all economists away from public policy health care issues.”
Well, given that the precautionary principal is irrational, that would be an accurate restatement of the precautionary principal.
Out of curiosity, can you prove that adopting the precautionary principle in all circumstances would not cause any harm? If not, what does that say about the precautionary principle.
AC: “Vancouver’s water comes from the North Shore mountains. We may be special, but we haven’t figured out how to live off of salt water yet.”
Touche. But I suppose swimming at the beach is probably not a great idea.
Nick: “Moreover, why do I find Andrew Coyne’s point so compelling? What Andrew Coyne says ought to be irrelevant. But it doesn’t feel irrelevant.”
I don’t know about you, but that’s not why I think Andrew’s point is helpful. To me, Andrew is pointing out that we can discuss the effects of the policy by interpolation, rather than extrapolation. That significantly decreases the likelihood of getting things wrong.
“Ritalin in the grade school fountains; Viagra in the faculty water bottles.
The Economist Party almost has a full platform.”
If it works, why not?
The point here isn’t whether cost-benefit analysis works. I don’t think there’s any way we could get away from CBA (a world without CBA would be far more alien than a world with lithium-enriched water). The point is about how to follow the steps of CBA properly. Mike is making a research proposal and indicating how socially relevant it is. Many of the commenters seem to be operating under the assumption that Mike is claiming to know enough to make an actual policy proposal when I think he pretty clearly isn’t doing that.
Mike:
As far as your data is concerned, I was wondering where you got the information on surface water. My source suggests that surface water has negligible lithium and thus my contention that it is quite possible that we are poorly adapted for it as a species. Because of this, there is a real, if small, probability of a disaster that might not reveal itself until long after it’s too late. You’d need to have enormous upside to overcome this possibility.
As far as imposing an excessive standard, I am only proposing to apply it where people don’t have an effective choice about whether to participate. If you can’t prove that something is safe, then I have a right to have doubt and to substitute my judgment for yours. This comes back to the the debate we recently had about the practical limits of scientific analysis in the economics of pollution. When faced with “unknown unknowns” I simply don’t accept that your prior distribution of possible harms is more rational than mine.
And I don’t think this principle constrains very much progress. We managed to eliminate small pox even without forcing everyone to take the vaccine. And yes we are dosed with vitamin D and iodine, but there is plenty of evidence that our ancestors were exposed to plenty more of those substances in the wild, so no cause for concern.
As far as my own exposures go, I am definitely not serene at all. I am quite sure I’m consuming very little lithium. The city I live in has a copious supply of (unfluoridated) surface water. I have a cottage with a well but have the water tested every year. I just spent quite a lot of money replacing the perfectly functional water supply pipe to my house because it was made of lead.
Blik:”Many of the commenters seem to be operating under the assumption that Mike is claiming to know enough to make an actual policy proposal”
Not at all (at least in my case). I just don’t want you substituting your utility and world model for mine. Like I said, feel free to send me a lithium tablet in the mail. But don’t force it down my throat.
Speaking of food fortification, which is the flip side of water fortification, we do it extensively in Canada. Iodization of salt has been mandatory since 1949. Along with cretinism, iodine deficiency also leads to goitres, or swollen thyroid gland. If you haven’t seen this condition you can be forgiven. Iodization has completely eliminated this problem in Canada.
Fortification or enrichment of white flour was also legislated at the same time. White flour has to have thiamin, riboflavin, niacin, folic acid and iron added to it to replace levels lost during milling.
Milk is fortified with Vitamin D. This has tossed rickets into the same memory bin as goitres.
I’m quite aware of mental illness. The answer to physch therapy or drugs is “yes”. The body is very complicated. If a causal link can be established with certain lithium levels and reduced mental illness, I say go for it.
The comment on naturally occurring lithium has an interesting parallel with fluoridation. The wiki article on the topic states that ground water in Arizona and most interestingly Southern Ontario west of Toronto has a naturally-occurring fluoride level above the recommended limit. De-fluoridation has to be employed in this case. I cross-checked and verified this fact with Ontario sources. Note that this is a concern specific to groundwater. Surface water, i.e. rivers or the Great Lakes are unaffected. This eliminates many urban areas from the problem.
Determinant: “If a causal link can be established with certain lithium levels and reduced mental illness, I say go for it.”
That and strong scientific evidence that as a species we have somehow recently become deficient in lithium is what it would take to convince me.
K: “Not at all (at least in my case). I just don’t want you substituting your utility and world model for mine. Like I said, feel free to send me a lithium tablet in the mail. But don’t force it down my throat.”
The government has to implicitly choose one utility function over all others. I don’t see any sensible way around that.
Of course, I’m the kind of person who questions whether the world described in Brave New World is actually any worse than the one we have now.
K: “That and strong scientific evidence that as a species we have somehow recently become deficient in lithium is what it would take to convince me.”
What if “natural” behavior isn’t socially optimal? Maybe our behavior is evolutionarily optimal, but not too violent from society’s point of view? Maybe overdosing on lithium is the least intrusive way to make people more moral beasts.
Another issue is the question of whether the optimal diet is really the one that most closely matches what our ancestors ate. I mean, our ancestors didn’t lead our lives. They had different patterns of activity, different sensory experiences, and different lifespans.
“A city with no-to-little elemental lithium would need to add 70 micrograms/L of elemental lithium to the water supply.”
You can’t add “elemental” lithium to the water supply. Here’s what happen’s if you try – http://www.youtube.com/watch?v=8ypUVpwgcAA
You want to add lithium ION to the water supply I assume.
And while ingesting certain quantities of the element lithium may be safe, ingesting “elemental” lithium is not recommended
Thanks for your comment, MOB – I was worried I’d err in the chemistry somewhere.
“Many of the commenters seem to be operating under the assumption that Mike is claiming to know enough to make an actual policy proposal when I think he pretty clearly isn’t doing that.”
Agreed – I don’t think you’d want to even consider doing something like this without a lot more study. That and I can’t see it ever happening – it creeps too many people out (myself included).
I think Nick’s ‘height tax’ analogy is a good one.
BoB Smith: Out of curiosity, can you prove that adopting the precautionary principle in all circumstances would not cause any harm? If not, what does that say about the precautionary principle.
Nope. Not much either way.
Why I believe it is relevant in this case is because of the importance of water to the ecosystem. Notwithstanding that one doesn’t know what happens to perfectly normal individuals when given increased levels of lithium (the evidence so far suggests positive effects for those with some forms of mental illness – what about the side effects for others?) what happens to the treated water after it goes down your home drain?
It ends up in the water table – and eventually into habitats for other animals/plants etc. I believe there are studies demonstrating how expired drugs flushed down the toilet, or the increased use of birth control pills – does end up affecting reproduction/sex of amphibians fish etc.
In addition, while concentrations of some elements (mercury for example) in water may not be toxic to humans, they do concentrate lower down in the food chain – hence why some fish is banned for eating in some waterways for this very reason. Even ocean born albacore tuna is risky to consume in high quantities.
Ask anyone familiar with the Green Party policies, or some professions. The Precautionary Principle is well established – maybe not suitable in all cases. This one certainly seems to qualify, well before one gets to the economic cost/benefit analysis.
Bob Smith: Aha! I get it now. You are saying that the Precautionary Principle is not just false, but self-contradictory. We don’t in fact follow the PP, so following the PP would be a dangerous experiment. I was a bit slow on the uptake. Interesting point.
taste: agreed that it’s better to be safe than sorry, but keep in mind we’re not talking a drug like viagra or birth control pills – things you wouldn’t find in the environment sans humans introducing it. Lithium is already present in the environment (to varying degrees), and humans seem to need small amounts of it in their diet:
http://www.jacn.org/cgi/content/full/21/1/14
But perhaps instead of adding it to the water supply (much of which isn’t consumed by humans anyway), it would be better to add it to some convenient food (milk? break?). Or maybe table salt?
Also, if it was added to food, then people so minded could choose to buy “no lithium added” varieties.
Mike: I don’t actually have a problem with the height tax. I’m tall but sympathetic to those who are short. Unlike forced consumption of a psychoactive drug, the effects of a height tax are totally predictable and induce no stress in me whatsoever. Unless you can provide strong evidence that our natural condition (i.e. not just some municipality in Texas) involved higher levels of lithium, I will consider this equivalent to a proposal to add Prozac (or speed 🙂 to the water supply.
Blik: “The government has to implicitly choose one utility function over all others.I don’t see any sensible way around that.”
Governments, in the short to medium term, are bound by many things including constitutions. So, at best, they are performing constrained optimizations. I’m proposing what I consider to be a reasonable constraint on government action. I suspect that a lot of the “heeby-jeebies” expressed by commenters (including Mike) relate to the bounds of government action.
“Of course, I’m the kind of person who questions whether the world described in Brave New World is actually any worse than the one we have now.”
I suspect that exactly sums up most of our disagreement. When science does find the drug that will make you happy all the time, the economists will be the ones who will be first in line for it, like AAA subprime CDO investors in March 2007. Some kind of professional blindness to black swans.
“Unless you can provide strong evidence that our natural condition (i.e. not just some municipality in Texas) involved higher levels of lithium”
And Japan. And river deltas such as the Nile. But you do have me curious how wide spread this is – I’m going to place a call to the MOE.
Patrick’s link is quite interesting, for those who haven’t seen it: http://www.jacn.org/cgi/content/full/21/1/14
K: “Unlike forced consumption of a psychoactive drug,”
Oh come now. Go easy with the hyperbole!
This is exactly what I meant by people getting the heebie jeebies because of the association of lithium with crazy.
But as the link I provided shows, lithium is not a psychoactive drug. It’s a required nutrient. One consequence of deficiency is malfunction of the brain.
If you want to draw analogies, it seems to me iodine is probably more appropriate that prozac.
Patrick’s link is interesting. This whole thread (post and comments) is interesting.
I’m still totally hung up on my own hang-up. When I read Patrick’s link, I come away thinking “Oh, that’s OK then, because we are not talking about introducing some new additive drug to the water; we are merely remedying some deficiency in our natural diet.”
Why should the status quo, of what is considered “normal and natural”, have such a big effect on my thinking? What’s the difference between adding something, and replacing something that’s missing? Just like Philippa Foot’s trolley problem http://en.wikipedia.org/wiki/Trolley_problem , it doesn’t make sense from the generalised utilitarian/consequentialist calculus.
We don’t in fact follow the PP, so following the PP would be a dangerous experiment.
-Also heard on the BP Deepwater Horizon drilling rig, pre April 20th
Why should the status quo, of what is considered “normal and natural”, have such a big effect on my thinking? What’s the difference between adding something, and replacing something that’s missing?
Because you understand that a new equilibrium will be established. That will be different than the current state – similar to economics. Just because lithium occurs naturally in higher quantities in other environments doesn’t mean that the ecosystem where this occurs has not adjusted/evolved accordingly – beyond just some measured human traits in a small portion of the population.
Nick: “We don’t in fact follow the PP, so following the PP would be a dangerous experiment.”
That is a lovely piece of abstract logic. It’s also disabused of any relationship to the practical application of the PP. The PP just says that “unknown unknowns” are likely to be bad. It seems to be true. For example, random emissions into the environment seem, historically, to have been more likely to cause harm than good. And the distributions of the harm are unquantifiably fat tailed, and some doubt that they even have finite moments.
There’s a pretty good recent Taleb paper in which he characterizes the problem as model errors being concave. It also discusses some of the other common themes on this site like the fragility of debt. So despite being unquantifiable, the PP is not meaningless. It just implies that we need worry about the inherent concavity of unknowns, and that robust systems need to be engineered for the tails rather than the means.
“What’s the difference between adding something, and replacing something that’s missing?”
It’s because evolution has adapted you to it! You are well suited for your environment because you descend from those who were not killed by it. Change it in some random way, and it might kill you. Change it a lot, and it will kill you a lot!
Mike: “And river deltas such as the Nile. But you do have me curious how wide spread this is – I’m going to place a call to the MOE”
Cool! And river delta data might be pretty compelling as our ancestors are likely to have spent a significant amounts of time there.
Have to say that I’m enjoying immensely this whole discussion. Among other things, the suggestion that an idea is in “poor taste” now has a prominent place in my personal Pantheon of Incredibly Lame Arguments.
Among other things, the suggestion that an idea is in “poor taste” now has a prominent place in my personal Pantheon of Incredibly Lame Arguments.
Can I submit your name down the road to the folks at École Polytechnique de Montréal or their supporters who, through an act of Parliament, are concerned about the timing of meetings in other parts of the country, seemingly unrelated?
She also called on mayor-elect Gordon McKay to make a commitment that
Midland Council would never again hold a meeting that conflicts with the Dec. 6 National Day of Remembrance, created in 1991 by an Act of Parliament.
http://www.midlandfreepress.com/ArticleDisplay.aspx?e=2890455
“For example, random emissions into the environment seem, historically, to have been more likely to cause harm than good.”
Do they? So, if you could, you’d reverse the industrial revolution? Seems to me that that involved (amongst other things) people randomly pumping out tons of carbon dioxide without giving too much thought about the “unknown unknowns”. Good thing too, I wouldn’t like an 18th century standard of living and I bet neither would you.
Yet, had the precautionary principle been in place in 1710, there no doubt would have been some government official saying “Awfully sorry Mr. Newcomen, but I don’t think you’ve canvassed all the unknown adverse effects of this so-called “steam-engine” and the emmisions it will produce, so I’m afraid we can’t let you use it.” (I know, I know, Newcomen didn’t invent the modern steam engine, but he significantly improved the original design).
Nick: “We don’t in fact follow the PP, so following the PP would be a dangerous experiment.”
Exactly.
Moreover, the precautionary principle isn’t really a health or science based decision making rule (even though it’s often framed a such). Rather it’s just an inherently conservative decision making rule intended to make change difficult and to favour the status quo (which is why it’s often promoted by interest groups – European farmers, for example – with an interest in preserving the status quo). But there’s no obvious reason for thinking that maintaining the status quo is neccesarily a better way of protecting human health or well-being (or whatever else it is you’re looking to protect). Yes, it means that you avoid unknown harm, but at cost of avoiding the benefits of that policy. A priori it isn’t clear that we should proceed on the assumption that the former outweight the latter.
Moreover, the unkown unkowns work both ways. There may be (and often are) unknown, but beneficial, side effects of certain policies (inventions, treatments, etc.) which are not (and can’t be) contemplated at the time those policies (inventions, treatments, etc.) are introduced. Newcomen invented his steam engine to pump out mines. I doubt that he contemplated that steam engines might realistically be adapted to power boats, trains (since they wouldn’t be invented for another century), factories (apparently it was tried, unsuccesfully with Newcomen steam engines), etc, and yet his engine was an important step in that direction.
With a surname like Smith, I surmise that many distant relatives died from unknown consequences of “modern life” including the plague, and the Great Smog.
Funny how science and knowledge progresses from such otherwise unforeseen catastrophes. For some. Others wish to remain in the era of Adam Smith.
“With a surname like Smith, I surmise that many distant relatives died from unknown consequences of “modern life” including the plague, and the Great Smog”.
True, are you suggesting ( that they would have lived long and healthy lives in the absence of “modern life”?
As an aside, since when is the “plague” a consequence of modern life? Given that the worst (recorded) outbreaks of plague generally occurred before “modern” times (the Plague of Justinian and the Black Death) and the Third Pandemic of the 19th and 20th century, while definitely modern, was most deadly (not coincidently) in parts of the world not generally considered (at the time) to have been inflicted with “modern life” (China and India).
“Do they? So, if you could, you’d reverse the industrial revolution?”
No, but way to take down a straw man, Bob!
I didn’t say innovation was bad on average. I said the unknown
consequences are more likely to be bad than good. Part of the reason is that randomness is bad for systems that require order. A second reason is that the positive consequences are likely to be seeked out as property and internalized, while the negative ones that can be externalized merit no attention from private research capital, and are therefore more likely to be unexplored. I didn’t think it was that subtle, but I suppose you can misconstrue anything if you try hard enough. In the context of your industrial revolution example it means that dumping waste into the environment is probably bad. Like radiation hitting a gene, it might be good. But probably it’s bad. That’s all.
I made this case extensively a short while ago here . I don’t want to rehash it all as it was quite long. However, I was debating someone who, it turned out, didn’t believe in any concept of utility at all, so it’s quite possible there were unexplored perspectives. The Taleb paper above describes it as concavity of the tails of model errors, which is a useful financial markets perspective.
K: “I said the unknown consequences are more likely to be bad than good.”
How can you say that? It’s the very nature of “unknown” consequences that that’s a analysis that can’t be made.
K: Like radiation hitting a gene, it might be good. But probably it’s bad.
Speaking of unkown unknowns, let me introduce you to the evolving theory of radiation hormesis (http://www.dose-response.org/about/mission.htm), to say nothing of the old (but apparently forgotten) treatment of gas gangrene using xrays (http://www.21stcenturysciencetech.com/Articles%202007/20_1-2_Gangrene.pdf)
Bob: “It’s the very nature of “unknown” consequences that that’s a analysis that can’t be made.”
No it isn’t. It is the nature of systems that are dependent on being highly organized that if you do random things to them they will be less organized. The living earth is not a random collection of atoms and can only function when it exists in a highly specialized subset of available states. But as you do random things to it, you make it more and more like a random collection of atoms and it will be less able to perform the processes from which we gain utility. This is a truth that derives from physics rather than economics.
“let me introduce you to the evolving theory of radiation hormesis”
It took me 2 minutes uncover the fact that as far as “dose-response hormesis” is concerned, the scientific consensus is that there is no there there : “The scientific research base shows that there is no threshold of exposure below which low levels of ionizing radiation can be demonstrated to be harmless or beneficial.” I.e. it only “works” when the level of radiation is so low that the effects are indistinguishable from noise. Kind of like homeopathy. Same underlying “theory” too.
As far as your second example goes, yes, bacteria don’t like x-rays either. If the bacteria are eating your leg and are threating to eat the rest of you then irradiating them may hurt them more than you. Same goes for malignant tumors, but you need really good aim, or you will cause more cancer than you are curing. It is a testament to the danger of radiation that it has long since been abandoned as an antibiotic.
K: “In the context of your industrial revolution example it means that dumping waste into the environment is probably bad.”
OK, but let’s go back to that example. Newcomen believes a steam engine is beneficial, and has some concrete beneficial uses in mind (pumping out mines). On the other hand, my hypothetical government official, rightly, says “dumping waste into the environment is probably bad”, although neither he, nor Newcomen can quantify the degree of “badness”. “Badness” might mean (i) “making tin mines smell worse” (who cares?)(ii) funcking up the air of British cities (definitely not good – although recent empirical research suggests that the higher wages in factory cities likely offset the adverse environmental conditions), (iii) global warming (not good – though probably wouldn’t have bothered the people skating on the Thames in Newcomen’s day) or (iv) poison gas clouds spreading accross Europe killing all life (not plausible in restrospect, but in 1710?)
In that context, since Newcomen is unlikely to be able to establish that his steam engine will not (or is unlikely to) cause harm (because, pollution likely will cause harm), and the extent of that harm can’t be known, doesn’t the precautionary principle mitigate in favour of telling Newcomen to go back to the drawing board and come back once he’s got a hydrogen powered engine? Couldn’t my hypothetical government official in 1710 plausibly say that the “potential adverse effects are not fully understood [they weren’t], [and] the activit[y] should not proceed” (to quote the understanding of the precautionary principle set out in the UN World Charter for Nature)?
Moreover, even on the weak understanding of the precautionary principal, in that example it would seem to mitigate against allowing Newcomen to operate his machine given that, initially at least, the social benefit of the Newcomen steam engine (while no doubt significant for the mining industry and Royal Navy dockyards) was still limited. Of course, without developing the Newcomen steam engine, the unknown (at the time) benefits resulting from the further development of steam engines might never have been discovered. You have to wonder what the principled basis is for favouring a decision rule which gives weight to possible negative unknown unknowns (however you measure that), but not the possibility of positive ones.
K: “It took me 2 minutes uncover the fact that as far as “dose-response hormesis” is concerned, the scientific consensus is that there is no there there”
It took you two minutes to find a wikipedia page for radiation hormesis, what, is your server down?
In any event, you might want to read it (and the documents it links to) a bit more closely. What it suggests is that, as of yet, there is not sufficient evidence to depart from the conventional linear no-threshold model. Fair enough, given that radiation hormesis is, as I said, an envolving theory. However, notwithstanding what wikipedia sees as the consensus view (a conclusion that is subject to some debate given that many of the sources it cites predates the sharp increase in research on the issue over the past decade) there is a body of literature, from respectable sources (peer reviewed journals, the French Academy of Science) which provides some support for the hypothesis.
In any event, appeals to scientific consensus have never cut much weight with me, as it’s really a lazy mental shortcut for people who don’t want to think about an issue. What was the scientific consensus on a heliocentric solar system circa 1542? Science isn’t a popularity contest.
K: “It is a testament to the danger of radiation that it has long since been abandoned as an antibiotic.”
You really should read the Cutler article I linked to, as he explains why the use of X-rays to treat gas gangrene fell into disuse. It wasn’t because of the dangers of radiation.
Bob: the scientific consensus before Copernicus wasn’t irrational. It was just short on data and analytical tools. I don’t have any reason to question the scientific consensus here. They don’t even have a underlying mechanism to explain how radiation could be beneficial. On top of the consensus being against them it sounds far to much like homeopathic reasoning which is unequivocally nonsense. “Not sufficient evidence to depart from conventional…” coupled with no known underlying mechanism just sounds like “probably wrong”. You could say the same thing about pretty well any alternative theory.
Assuming that we are not adapted to higher levels of lithium (and I’ll allow, Mike, that maybe we are) then the unknown consequences of lithium are much like the unknown effects of radiation. Unknown, but probably bad. The principal consequence of the PP is that any model of harm forms a lower bound of the expected harm. The reality of any particular situation is that we must consider who are the recipients of both the benefits and the possible harm and if not the same, how we can prevent the consequences for the possible victims or alternatively consult them on what they consider to be reasonable remedies, constraints or compensation. I.e. We need to take into account when possible the potential victims own assessment of their own risk rather than relying entirely on scientific consensus or government experts. The models are simply incapable of assessing the uncertainties, and those uncertainties should be assessed by principals (who will always charge additional risk premium for tail risk) rather than agents. I think that’s about as close as I can get.
K: “When science does find the drug that will make you happy all the time, the economists will be the ones who will be first in line for it, like AAA subprime CDO investors in March 2007. Some kind of professional blindness to black swans.”
There will always be failures of prediction and the PP diminishes the odds of major mistakes leading to declines relative to the status quo. The point that I believe the rest of us are trying to make is that there’s no reason to think that following the PP will lead to better overall results than not following it, if one assumes that policy makers possess any useful information whatsoever. I think it’s entirely plausible that a world that avoided financial disasters by strictly adhering to the PP would be worse than the world in which we live.
Blik: “there’s no reason to think that following the PP will lead to better overall results than not following it, if one assumes that policy makers possess any useful information whatsoever.”
First, let me take back the swipe at economists. It’s not economists. It’s agents of other peoples interests. Agents (government scientists, fund managers, rating agencies, etc) use models to evaluate risk, and they tend to be unable to (prevented from, incentivized not to) make allowances for uncertainties outside their models. Markets, which are dominated by agents, may underestimate rare events (Nassim Taleb thinks so), but expert cost benefit analysis is intrinsically worse.
The Gaussian copula is a good examples of a model that takes into account things we know, i.e. the historical rates and correlations of default. It served as a method of interpolating, but also extrapolating to events that never occurred before. It was used by rating agencies to evaluate CDO risk. Then stuff happened. Stuff that wasn’t inside those models, and as it happened that stuff was worse than the stuff inside the models. And the stuff that happened is extremely unlikely to happen again in a very long time. But that doesn’t mean that nothing is going to happen again. It just means that we don’t know what that thing is, but any sane principal in the market will still allow for the possibility of such an event in their risk assessment.
And we see that senior tranche CDOs and S&P options display a very large pricing “skew”. So markets do assume that unknown unknowns are predominantly bad, even though they have no idea what those events will be. When agents of other peoples interests fail to take the same thing into account then we have a loss of optimal allocation. The solutions are to either 1) somehow get the agents to think more like principals or 2) find a way to consult the principals directly about their preferences. But the PP tells us to err in the direction of assuming that things will be worse than predictions lead us to expect; exactly like rational agents in the markets.
To be really clear, that last sentence should read “…predictions based on stuff we know…”
Here’s an idea: Every year we take a vote:
1) What should the rate of carbon tax be?
2) How much lithium would you like in your water?
…
Then take the median. And don’t use the carbon tax revenues to balance the budget. Just redistribute them equally per citizen. Answering these questions is surely a lot less taxing than trying to figure out how to allocate your RRSP. Lets make like we’re adults.
The principal consequence of the PP is that any model of harm forms a lower bound of the expected harm.
Except every action (or inaction) has the potential to cause harm.
Putting lithium in the water can cause potential harm (unknown adverse affects that are presumed to be harmful), not putting lithium can cause potential harm (i.e., possibly 30% more murders, rapes, robberies, etc. than would be the case is lithium was in put in the water). Indeed, that only emphasizes the inherent irrationality of the precautionary principal, in that it is a decision making rule whose outcome can depend on how a question is framed, when a truly science based rule should reach the same outcome regardless of how the question is framed.
Consider the following example. We have two otherwise identical towns, except one (Town A) has low lithium levels in the water and the other (Town B) has high lithium levels. Now, imagine a scenario where the citizens of town A are proposing the adoption of Mike’s propsal, to add lithium to the water to have a high lithium level. Meanwhile, the citizens of Town B, responding to your (perfectly legitimate) concerns about the potential dangers of high lithium levies, are proposal to filter the water to remove lithium so that they will have low lithium levels. For simplicity sake, let’s assume that changing lithium levels is costless (or at least is not material).
Now, there are potential (anduncertain) benefits and potential (and uncertain) harms of each proposal (and, not concidentally, the harms of one proposal are the benefits to the other, and vice-versa). Now, consider how your formulation of the precautionary principal, “that any model of harm forms a lower bound of the expected harm” would play-out. In Town A, the precautionary principal would say, yes, there may be benefits to adding lithium, but they’re uncertain, and the potential harms are unknown and potentially significant (your argument), so we shouldn’t add lithium to the water. In Town B, the precautionary principal would say, yes, there may be benfits to taking lithium out of the water, but they’re unkown, and the potential harms, while uncertain, are potentially significant, so we shouldn’t take lithium out of the water.
The end result is that the precautionary principle suggests that Town A should keep its lithium level low, while Town B should keep it high. That’s an odd result, since, one would have thought that high levels of lithium are either good or bad (and vice-versa with low levels of lithium) for both towns (since they’re otherwise identical). And yet the use of the precautionary principle tells us, in this example, that high lithium is preferrable for Town B, but not for Town B. One would have thought that a truly scientific decision-making rule would reach the same result as to what is the best course of action for the citizens of Town A and Town B, given that they are otherwise identical (i.e. high or low lithium is either good and bad for these people regardless of what the status quo is).
And that’s the point, the precautionary principle isn’t a science based decision making rule, but rather one that favours the status quo (which is why its conclusions depend on how a question is framed and what the treatment option is). That’s a good rule if you believe that the status quo is generally preferrable to change, but a priori it generally isn’t clear what the rational basis for that belief is.
Correct me if I’m wrong, but my understanding is that lithium does not have any psychotropic effects on healthy individuals. So if more lithium in the water supply reduces homicides/suicides, this is probably due to the effect it’s having on the small percentage of people who have mental problems, rather than by mellowing everyone out.
If that’s right, then one needn’t be concerned about governments using lithium in the water as a means to stupefy the populace, since for the vast majority putting lithium in the water would have zero effect.
Bob, that’s a straw man again. You can’t possibly think I’m advocating for any random status quo. We are adapted for the environment of our ancestors. Evolution doesn’t happen cause you move from Calgary to Edmonton. It happens over thousands of years. That’s why we spent so much time debating the mineral content of surface water which was the likely source for the vast majority of our ancestors.
Blackadder: yup. If you know that then no worries.
K: “Bob, that’s a straw man again. You can’t possibly think I’m advocating for any random status quo. We are adapted for the environment of our ancestors. Evolution doesn’t happen cause you move from Calgary to Edmonton. It happens over thousands of years. That’s why we spent so much time debating the mineral content of surface water which was the likely source for the vast majority of our ancestors.”
I’m not entirely sure what you’re advocating. It would be helpful if you would explain why you think my hypothetical example is a straw man. I take it from your suggestion that you’re not advocating for any random status quo, that you’re saying that we shouldn’t add lithium in low lithium towns, but that we should remove it from high lithium towns. That’s a perfectly reasonable position to take (regardless of whether its right or wrong), but as I point out below, it isn’t supported by reference to the precautionary principal.
My first observation is that I’m not sure what any of that has to do with evolution. Is there any basis for believing that humans evolved to live in an environment with naturally low lithium levels rather than high one? If there is significant regional variability in naturally occuring lithium levels (which the original article seems to suggest) one would expect that humans would have evolved to live in both low and high lithium environments (since the ability to consume water with both high and low lithium levels safely would have been an undoubted evolutionary advantage that would have allowed our hunter-gatherer ancestors to travel more freely). If so, that would seem to undermine your point. In any event,I don’t think your evolutionary argument for the the precautionary principal, in this context, is particularly well thought out.
My second observation is that your statement that you’re not arguing for “any random status quo” I think highlights the point that the precautionary principle doesn’t rise to the level of a “principle” if its application depends on the whether or not one prefers the status quo. In my hypothetical example, the status quo is the “state of nature” in both towns, i.e. naturally occuring lithium levels. It just so happens that the state of nature in both towns differ (making for an easy thought experiment). A principled application of the precautionary principle, therefore, would put the onus, and the particularly heavy evidentiary burden, on the person proposing to intervene with that natural state to justify that intervention. And, in my hypothetical example, that would lead to the odd result that the precautionary principal calls for polar opposite policies for otherwise identical people depending on their starting point. But, you seem to be suggesting that the application of the principal depends on whether or not your want to preserve any particular state of nature. As such, it is not a principled decision making rule, but only a moderately useful debating tactic (gee, who can be opposed to taking precautions and avoiding harm) for those who wish to preserve any particular status quo.
Bob: Ground water is likely to be high mineral. Surface water isn’t. People in Texas probably drink ground water as there is not much surface water. Our ancestors drank surface water as they didn’t have wells. I thought this was clear. Sorry. I could be wrong but I don’t see why it’s “poorly thought out.”
It’s 1920. You’ve invented a machine for X-raying feet for shoe fitting. Nobody has any idea what the effects of X-rays on the body are. You market it to shoe stores as a way to make it more fun for kids to go shoe shopping. And anyways, any unknown impacts are equally likely to be bad or good. Right?
It’s 1957. You run a pharmaceutical company that has developed a drug that relieves nausea. You have no idea whether the drug could have any impact on a developing fetus. You decide to market it for the relief of morning sickness because on average, any unknown effects on the fetus are equally likely to be bad or good. Right?
How about mercury for making nice hats? Hexachlorophene for babies’ bottoms? Tobacco? Just stuff i can think of off the top of my head. Do you think there was ever any chance that tobacco could have been as big a benefit as it has caused harm? Why do we even bother with phase 1 drug trials if we literally cannot think of any mechanism by which a drug can do harm?
There are sound theoretical reasons why the unexpected effects of drugs and pollutants are likely to be bad. And experience tells us it’s true. That, and the fact that the associated upsides are likely to be privatized, and it makes a lot of sense for the public to act in a precautionary manner.
“I lived for a while during elementary school in a village that didn’t fluoridate its water. No matter. A nurse from the Public Health Unit came around to each class every week and gave everyone a paper cup with fluoride rinse in it. A two minute swish for everyone. It was the ritual when we came back from gym class as that was when the nurse put the cups at everyone’s desk.”
That would be astoundingly expensive and a massive waste of everyone’s time, wouldn’t it? Compared to fluoridating municipal water?
Also, someone mentioned lithium orotate above. There are no scientific studies demonstrating lithium orotate efficacy, but it may at some future date be shown to work.
K: “You decide to market it for the relief of morning sickness because on average, any unknown effects on the fetus are equally likely to be bad or good. Right?”
I don’t think most of us are advocating that sort of argument. My view is that proper cost benefit analysis consists of two parts.
One is to formulate a prediction of how likely each potential outcome is. This prediction is a guess and in practice it won’t specify a likelihood for every possible outcome because there are too many possibilities to list. The prediction will take into account all information that one has. Ones lack of perfect information will be accounted for in the relatively high probabilities assigned to relatively extreme events.
The other part is the assignment of a utility value to each potential outcomes. We very often don’t do this explicitly, but any analysis implicitly includes utility values. The assignment of these values is at its core philosophical and personal. There’s no way to establish an objectively correct set of utility values. There’s no getting around that.
These two components are clearly imperfect. How accurate should they be? I would say that the only solid rule is that one should give a best case and a worst case scenario, with reality falling in between them. I think that’s how decision making should occur. If that method is followed, some terrible events will surely occur, but those accidents should be balanced (in true welfare terms) by the opportunities that are taken.
K: “Here’s an idea: Every year we take a vote:
1) What should the rate of carbon tax be?
2) How much lithium would you like in your water?
…
Then take the median. And don’t use the carbon tax revenues to balance the budget. Just redistribute them equally per citizen. Answering these questions is surely a lot less taxing than trying to figure out how to allocate your RRSP. Lets make like we’re adults.”
First of all, I don’t see that as very far from what I would advocate. My one (major) concern is the problem of rational ignorance among voters. Voters often choose to be much less informed about policies than would be necessary for them to vote for policies approaching the optimum.
I agree with Bob Smith’s characterization of the precautionary principle as non-scientific decision-making criteria that tends to support the status quo.
It conforms to what I observe in public fisheries management–commercial (sic) and recreational. It suggests low-cost regulatory solutions to what are typically deep structural problems such as first-come, first-served rules of access. It suggests sustainable fisheries management is possible despite regulations that allow unrestricted access and/or unlimited effort, and encourages wasteful public resource freeriding.
Furthermore, I would argue that the precautionary principle has contributed to an increase in fraudulent biological science in recent years on the part of state managers, academics and activists.
Blik: That’s how you model the things you know. But, you’ve neglected the question of whether the expected value of the things you never imagined, and therefore never considered, is negative. But if we agree on the the correct recourse, maybe we can agree to disagree about the reasons.
Westslope: on the contrary. It suggests the things will turn out worse than what is predicted by models of what you know, and so regulatory measures will fail to anticipate some harmful impacts. Kind of like the fisheries.