Larks

• Larks 16 Jul 2019 3:46 UTC

  54 points

  on: Age-Weighted Voting

  I performed a very cur­sory liter­a­ture re­view on the sub­ject. Over­all it seems the psy­chol­ogy re­search sug­gests that older peo­ple dis­count the fu­ture less than younger peo­ple, which might sug­gest giv­ing their votes more weight.

  Usu­ally such a brief pe­rusal of the liter­a­ture would not give me a huge amount of con­fi­dence in the core claims; how­ever in this case the con­clu­sion should seem prima fa­cie very plau­si­ble to any­one who has ever met a young boy.

  In no par­tic­u­lar or­der:

  Age Differ­ences in Tem­po­ral Dis­count­ing: The Role of Dis­po­si­tional Affect and An­ti­ci­pated Emo­tions:

  Ad­vanced age was as­so­ci­ated with a lower ten­dency to dis­count the fu­ture, but this effect reached statis­ti­cal sig­nifi­cance only for the dis­count­ing of de­layed gains.

  Age-Re­lated Changes in De­ci­sion Mak­ing:

  Older adults are not always risk-averse, and their abil­ity to post­pone grat­ifi­ca­tion tends to ex­ceed that of younger adults.

  Aging and al­tru­ism in in­tertem­po­ral choice:

  Re­search on life span changes in mo­ti­va­tion sug­gests that al­tru­is­tic mo­tives be­come stronger with age, but no prior re­search has ex­am­ined how al­tru­ism af­fects tol­er­ance for tem­po­ral de­lays. Ex­per­i­ment 1 used a re­al­is­tic fi­nan­cial de­ci­sion mak­ing task in­volv­ing choices for gains, losses, and dona­tions. Each de­ci­sion re­quired an in­tertem­po­ral choice be­tween a smaller-im­me­di­ate and a larger-later op­tion. Par­ti­ci­pants more of­ten chose the larger-later op­tion in the con­text of dona­tions than in the con­text of losses; thus, part­ing with more of their over­all cap­i­tal when the act of do­ing so benefited a char­ity. As pre­dicted, the mag­ni­tude of this “al­tru­ism effect” was am­plified in older rel­a­tive to younger adults. -

  Fol­low­ing Ad­vice Be­cause it’s Been Paid For: Age, the Sunk-Cost Fal­lacy, and Loss Aver­sion:

  [Y]ounger adults com­mit the sunk-cost fal­lacy more fre­quently, and make nor­ma­tively cor­rect de­ci­sions less fre­quently,than older adults when hy­po­thet­i­cal sunk costs are at stake … Younger adults made fewer ini­tial in­vest­ments of money on the Calorie Es­ti­ma­tion Task than older adults. Younger adults demon­strated the sunk-cost fal­lacy more fre­quently, over­in­vested more af­ter demon­strat­ing the fal­lacy, and made the nor­ma­tively cor­rect de­ci­sion less fre­quently than older adults on the hy­po­thet­i­cal self-re­port mea­sure of the sunk-cost fal­lacy. Younger adults in­di­cated that they were more averse to hy­po­thet­i­cal mon­e­tary losses on a Trade­off Loss Aver­sion Task and more averse to hy­po­thet­i­cal mon­e­tary losses on a De­lay Dis­count­ing Loss Aver­sion Task. … Older adults did not demon­strate the sunk-cost fal­lacy on the Calorie Es­ti­ma­tion Task by fol­low­ing more ex­pen­sive ad­vice more closely than less ex­pen­sive ad­vice. -

  De­ci­sion Mak­ing in Older Adults: A Com­par­i­son of De­lay and Prob­a­bil­ity Dis­count­ing Across Ages:

  [O]lder adults dis­counted de­layed re­wards less steeply than young adult

  Acute stress and al­tru­ism in younger and older adults:

  Even un­der acute psy­choso­cial stress, older adults make more al­tru­is­tic choices than younger adults.

  Age-re­lated differ­ences in dis­count­ing fu­ture gains and losses:

  Re­sults in­di­cated that im­paired older adults dis­counted the fu­ture more than unim­paired older adults. In­ter­est­ingly, mid­dle-aged adults dis­counted fu­ture gains at a similar rate as im­paired older adults, but dis­counted fu­ture losses less than im­paired older adults (and similarly to unim­paired older adults).

  Age‐re­lated differ­ences in de­lay dis­count­ing: Im­me­di­ate re­ward, re­ward mag­ni­tude, and so­cial in­fluence:

  Younger adults ex­hibited higher lev­els of de­lay dis­count­ing than older adults.

  The only ev­i­dence I’ve found that older peo­ple might dis­count more was a Chi­nese study. I’m not sure why Chi­nese peo­ple would be differ­ent in this re­gard, though ob­vi­ously their cul­ture is differ­ent in many ways—or, the pa­per might just be wrong.

  Age differ­ences in de­lay dis­count­ing in Chi­nese adults:

  The cur­rent study aimed to clar­ify this re­la­tion­ship us­ing a rel­a­tively large sam­ple of Chi­nese adults with a wide age range (viz., 18 to 86 years old). A to­tal of1288 in­di­vi­d­u­als com­pleted the Mone­tary Choice Ques­tion­naire. Re­sults showed that the rate of de­lay dis­count­ing in­creased with age across adult­hood, with younger par­ti­ci­pants (18–30 years) dis­count­ing lessthan both mid­dle-aged par­ti­ci­pants (31–60 years) and older par­ti­ci­pants (over 60 years); and mid­dle-aged par­ti­ci­pants dis­count­ing less than older par­ti­ci­pants. Fur­ther­more, when the re­ward mag­ni­tude was large, par­ti­ci­pants were more likely to wait for de­layed re­wards.

• Larks 15 Jul 2019 3:28 UTC

  21 points

  on: Age-Weighted Voting

  In the UK’s Euro­pean Union mem­ber­ship refer­en­dum, the vot­ing pat­tern was heav­ily cor­re­lated with age: older peo­ple were much more likely to vote to leave the EU than younger peo­ple. My cur­rent (poorly in­formed) un­der­stand­ing is that, in terms of the cor­re­la­tion be­tween age and con­ser­vatism, both ag­ing it­self and co­hort effects play a role. If the lat­ter is sig­nifi­cant in this case, this sug­gests that, in twenty years’ time, most of Bri­tain’s elec­torate will be in favour of be­ing part of the EU. If so, then a huge amount of time and effort will have been wasted in the tran­si­tion costs of leav­ing and re­join­ing.

  First of all I think us­ing live poli­ti­cal ex­am­ples like this is not a great idea.

  It also seems pos­si­ble to me (though I am less cer­tain) that the ‘older peo­ple are right­fully more scep­ti­cal of offi­cial cost es­ti­mates’ al­ter­na­tive the­ory I de­scribed in this com­ment might ap­ply here. One of the key ar­gu­ments against Brexit was that it would lead to a high de­gree of short-term dis­rup­tion, due to un­cer­tainty, dis­rup­tions to sup­ply chains, and a loss of ex­port mar­kets on the con­ti­nent. In con­trast, many of the benefits, like pro­tec­tion from fur­ther in­sti­tu­tional de­cline in the EU, were longer-term.

  Yet af­ter the fact it is worth not­ing that many of the anti-Brexit fore­casts for im­me­di­ate large nega­tive con­se­quences have been proved wrong, at least so far. For ex­am­ple the Trea­sury fore­cast that un­em­ploy­ment would rise by 520,000 af­ter a Leave vote, whereas in ac­tual fact the UK labour mar­ket has im­proved sig­nifi­cantly since then.

  There’s also a big differ­ence be­tween want­ing to Re­main and want­ing to Re­join later on:

  • Once the UK has left, the tran­si­tion is­sues will be sunk costs—in­deed tran­si­tion costs will be­come a rea­son to not re­join!

  • Sta­tus quo bias will switch to sup­port­ing in­de­pen­dence.

  • The out­side view also sug­gests the UK would want to re­main in­de­pen­dent, as the other in­de­pen­dent west­ern eu­ro­pean coun­tries like Nor­way, Ice­land and Switzer­land have be­come if any­thing less likely to join now than they were in the past (see for ex­am­ple here and here).

  • This also seems to be the case in the UK—for ex­am­ple, the idea of join­ing the Euro, de­spite be­ing sup­ported by the Con­ser­va­tive Party in 1990 and Tony Blair later, has been to­tally dis­cred­ited even among Re­main sup­port­ers.

  So over­all I don’t think Brexit-re­lated tem­po­ral in­con­sis­tency is a great rea­son to sup­port in­creas­ing the weight on younger vot­ers.

• Larks 15 Jul 2019 3:02 UTC

  8 points

  on: Age-Weighted Voting

  How­ever, em­piri­cally there’s ev­i­dence that gen­er­a­tions tend to vote in their self-in­ter­est when it comes to is­sues that have differ­ent costs and benefits across time. Here’s Gabriel Ah­lfeldt sum­ming up some re­sults from a re­cent pa­per:

  “[O]lder vot­ers are less likely to sup­port mea­sures that pro­tect the en­vi­ron­ment, pro­mote sus­tain­able use of en­ergy or im­prove trans­port. Older vot­ers are also less likely to sup­port ex­pen­di­tures on ed­u­ca­tion or welfare poli­cies, such as un­em­ploy­ment benefits, but they are more likely to sup­port ex­pen­di­tures on health sys­tems. The rea­sons for these ten­den­cies can be differ­ent in ev­ery cat­e­gory. But it is difficult to find a sin­gu­lar ex­pla­na­tion other than gen­er­a­tional self-in­ter­est, which would ex­plain why older vot­ers tend to be gen­er­ally less sup­port­ive of ex­pen­di­tures that benefit other gen­er­a­tions and pro­jects that have pos­i­tive ex­pected effects in the long run, but costs in the short run. It fits the bill that where it is harder to think of gen­er­a­tional-spe­cific in­ter­ests such as on ques­tions re­lated to an­i­mal pro­tec­tion, women’s rights or ur­ban de­vel­op­ment, there is also no ev­i­dence of a gen­er­a­tion gap.”

  I don’t think this is a very fair sum­mary. Tak­ing the en­vi­ron­men­tal ex­am­ple, for ex­am­ple, older peo­ple are more sup­port­ive of nu­clear power:

  From the whole model in Table2, we know that five so­ciode­mo­graphic vari­ables show a sig­nifi­cant im­pact on the ac­cep­tance of nu­clear power. Age and ed­u­ca­tion have a pos­i­tive im­pact on ac­cep­tance, whereas gen­der, so­cial class, and res­i­dence have a nega­tive im­pact on it. As peo­ple be­come older, they re­veal more pos­i­tive at­ti­tude to­ward nu­clear power. Such an older-age effect confirmed the re­search find­ings by Slovic et al. [8]. En­light­en­ment from ed­u­ca­tion in­creases sup­port for nu­clear power. Ac­cord­ing to Kim et al. [16], higher ed­u­ca­tion brings out more ac­cep­tance. - Com­par­a­tive Anal­y­sis of Public At­ti­tudes to­ward Nu­clear Power En­ergy across 27 Euro­pean Coun­tries by Ap­ply­ing the Mul­tilevel Model

  Yet be­ing pro-nu­clear power seems like the more long-ter­mist solu­tion. Nu­clear power sta­tions are very ex­pen­sive, and cost billion of dol­lars to cre­ate, but once built they can gen­er­ate elec­tric­ity very cheaply and re­li­ably for a very long time—and with es­sen­tially no car­bon emis­sions. The rel­a­tive op­po­si­tion of young peo­ple sug­gests that ei­ther older peo­ple are not vot­ing in a more short-sighted fash­ion, or that the in­ex­pe­rience of young peo­ple out­weighs this effect.

  Similarly, lets look at the core case study that Ah­lfeldt et al. dis­cuss: Stuttgart’s new train sta­tion. They find that older peo­ple tended to vote against the tran­sit in­vest­ment, and sug­gest this is be­cause old peo­ple wouldn’t be al­ive for as long to reap the benefits.

  How­ever, they omit to men­tion that, like many large gov­ern­ment pro­jects, the pro­ject ended up run­ning dra­mat­i­cally over-bud­get and be­hind sched­ule:

  Ac­cord­ing to newly re­leased offi­cial calcu­la­tions, the to­tal costs for the Stuttgart 21 railway sta­tion could reach EUR 8.2 billion, twice as high as the es­ti­mate of EUR 4.5 billion pre­sented at the start of the pro­ject. The in­crease was driven by higher build­ing costs due to a boom­ing Ger­man con­struc­tion in­dus­try and tight do­mes­tic la­bor mar­ket, prob­lems con­cern­ing the build­ing site, and a de­sire by the man­age­ment of Deutsche Bahn to cre­ate a fi­nan­cial buffer against the risk of fu­ture price in­creases. The new figures re­leased were based on an as­sess­ment by the ac­coun­tancy firm Price Water­house Coop­ers (PwC) and the en­g­ineer­ing com­pany Emch+Berger.

  Ad­di­tion­ally, the com­ple­tion date of the pro­ject was pushed back for a fifth time to 2025, four years later than origi­nally an­ti­ci­pated. (source)

  It seems rea­son­able to think that older vot­ers would have de­vel­oped a sense of sus­pi­cion over time about up­front es­ti­mates—an in­tu­itive sense of plan­ning fal­lacy, and prob­lems with the poli­ti­cal pro­cess. With the benefit of these more ac­cu­rate views, it would be ra­tio­nal for them to be more likely to re­ject the pro­ject; the prob­lem lies in the naïvety of the young vot­ers.

  The au­thors seem to have con­sid­ered a similar idea… but strangely as­sume that ex­pe­rience could only make one more in favour of large pro­jects:

  Ex­pe­rience with similar pro­jects in the past com­bined with a sense of morale could the­o­ret­i­cally im­ply that the like­li­hood of sup­port could in­crease with age,if the pro­ject is per­ceived as so­cially de­sir­able.

  What’s more, that age brings wis­dom should have been an ob­vi­ous hy­poth­e­sis, as if you look at the con­trols in their ta­bles, you’ll see that higher lev­els of ed­u­ca­tion also lead to op­po­si­tion to the pro­ject (as­sum­ing I read the table cor­rectly; the au­thors con­ve­niently did not de­scribe this re­sult in the text). So ba­si­cally they found that both more ed­u­cated and more ex­pe­rienced peo­ple op­posed the pro­ject, a pro­ject that did in­deed fall short of ex­pec­ta­tions, and yet at­tributed this to self­ish­ness in­stead!

• Larks 15 Jul 2019 0:59 UTC

  44 points

  on: De­brief: “cash prizes for the best ar­gu­ments against psychedelics”

  First of all, thanks for run­ning this—I think prizes are a great idea—and con­grat­u­la­tions to the win­ners!

  I some­what dis­agree with your take here how­ever:

  I was sur­prised by the ra­pidity of sub­mit­ted ar­gu­ments – four hours af­ter an­nounc­ing the prize, there were six sub­mit­ted ar­gu­ments (in­clud­ing the win­ning ar­gu­ment).

  None of these en­gaged with the pro-psychedelic ar­gu­ments I made in the main post, in­stead they ap­peared to be ar­gu­ing gen­er­ally against psychedelics be­ing a promis­ing area.

  I’m chalk­ing this up to a “cached ar­gu­ments” dy­namic – peo­ple seem to have quick takes about top­ics stored some­where in mem­ory, and when asked about a topic, the im­me­di­ate im­pulse is to de­liver the pre-stored quick take (rather than en­gag­ing with the new ma­te­rial be­ing pre­sented).

  If this is true, it’s a hur­dle on the road to good dis­course. To have a pro­duc­tive dis­agree­ment, both par­ties need to “clear out their cache” be­fore they can ac­tu­ally hear what the coun­ter­party is say­ing.

  You asked for the best ar­gu­ments against psychedelics, not for counter-ar­gu­ments to your spe­cific ar­gu­ments in favour, so this doesn’t seem that sur­pris­ing. Prob­a­bly if you had speci­fi­cally asked for counter-ar­gu­ments, as op­posed to merely say­ing they were there to ‘seed dis­cus­sion’, peo­ple would have in­ter­acted with them more.

• Larks 14 Jul 2019 20:03 UTC

  8 points

  on: Age-Weighted Voting

  Thanks for writ­ing this! I found it very in­ter­est­ing, and cer­tainly far above most EA posts about poli­tics.

  18-27yr olds: 6x vot­ing weight
  28-37yr olds: 5x vot­ing weight
  38-47yr olds: 4x vot­ing weight
  48-57yr olds: 3x vot­ing weight
  58-67yr olds: 2x vot­ing weight
  68+yr olds: 1x vot­ing weight

  One effect of this is to cre­ate a big dis­con­ti­nu­ity at 18 years, out of all pro­por­tion to the marginal in­crease in knowl­edge and wis­dom. One op­tion would be to have the peak later, as a rough sum of the knowl­edge curve and the in­cen­tive curve, e.g.:

  18-27yr olds: 1x vot­ing weight
  28-37yr olds: 2x vot­ing weight
  38-47yr olds: 3x vot­ing weight
  48-57yr olds: 3x vot­ing weight
  58-67yr olds: 2x vot­ing weight
  68+yr olds: 1x vot­ing weight

  An al­ter­na­tive fix would be to give par­ents of minor chil­dren ex­tra votes. This would aim to ex­ploit the fact that par­ents typ­i­cally ex­hibit re­ally high lev­els of al­tru­ism to­wards their chil­dren, so ar­guably ac­tu­ally have longer-term hori­zons than child­less 18yr olds de­spite lower life ex­pec­tancy. We see this be­havi­our in vot­ing through par­ents ap­par­ent ob­ses­sion with good schools and safety for their chil­dren. So we might have some­thing like:

  0-7yr olds: par­ents +4 votes each
  8-17yr olds: par­ents +3 votes each
  18-27yr olds: 6x vot­ing weight
  28-37yr olds: 5x vot­ing weight
  38-47yr olds: 4x vot­ing weight
  48-57yr olds: 3x vot­ing weight
  58-67yr olds: 2x vot­ing weight
  68+yr olds: 1x vot­ing weight

• Larks 12 Jul 2019 16:42 UTC

  14 points

  in reply to: Paul_Christiano's comment on: Age-Weighted Voting

  Yes—hence the stan­dard pair of ar­gu­ments:

  The young peo­ple who will have to live with the con­se­quences of Brexit voted to Re­main.

  vs

  The older peo­ple with ac­tual ex­pe­rience of liv­ing out­side the EU voted to Leave.

• Larks 12 Jul 2019 11:59 UTC

  2 points

  on: Seek­ing EAs to In­ter­view on Ca­reer Change Resources

  Could you ex­plain how this re­lates to 80k’s con­tent? My im­pres­sion was that a lot of ma­te­rial was aimed at col­lege stu­dents, which seems like ‘the ear­lier stages of a ca­reer change’.

• Larks 8 Jul 2019 2:41 UTC

  19 points

  on: Cor­po­rate Global Catas­trophic Risks (C-GCRs)

  Over­all I thought this was a weak ar­ti­cle. In many cases the rea­son­ing seems to be ig­no­rant of sim­ple eco­nomics. Below I point out sev­eral is­sues; oth­ers I omit­ted in a vain at­tempt at brevity.

  Cor­po­rate Growth

  In­cen­tives cre­ate strong se­lec­tive pres­sure for cor­po­ra­tions to grow (e.g. be­cause they benefit from economies of scale).

  It is definitely true that there are in­cen­tives for firms to grow up the their effi­cient size. How­ever, equally there are in­cen­tives not to grow fur­ther, and many com­pa­nies are around these lev­els—for a lit­tle back­ground on the the­ory see for ex­am­ple wikipe­dia. Cer­tainly, when I was an equity an­a­lyst, when news one of my com­pa­nies had just an­nounced an ac­qui­si­tion crossed the tape, I’d ex­pect the share price to be down on the news—whereas if they an­nounced a spin-off, I’d ex­pect the stock to be up!

  Rev­enues of the For­tune 500 in­creased from 58% to 73% of US GDP

  This is very mis­lead­ing.

  We can see this by not­ing that, rather per­versely, one of your pro­posed poli­cies would ac­tu­ally in­crease cor­po­rate rev­enue as a % of GDP. Why? By break­ing up ver­ti­cally in­te­grated com­pa­nies, we would add a new layer of trans­ac­tions that would be counted as rev­enue de­spite hav­ing no ad­di­tional eco­nomic sub­stance. If I have a so­lar panel fac­tory and an in­stal­la­tion busi­ness, no rev­enue is recorded when the pan­els move from the fac­tory to the in­stal­la­tion trucks, only when they go to the end cus­tomers. If you split my com­pany into two, we now record rev­enue twice, de­spite no fun­da­men­tal change—illus­trat­ing why this is a silly met­ric. It is more ap­pro­pri­ate to com­pare prof­its to GDP, or rev­enue to gross trans­ac­tions.

  Peo­ple more com­monly look at S&P prof­its as a % of GDP, which is bet­ter, but still not great. You men­tion as an aside that this is partly due to in­ter­na­tional rev­enues, but with­out not­ing how this un­der­mines the point. Over the last few decades we have seen con­sid­er­able global­i­sa­tion, as US com­pa­nies have ex­panded into over­seas mar­kets, so this ra­tio would have in­creased even if the com­posi­ti­tion of the do­mes­tic US econ­omy re­mained to­tally un­changed.

  One of the first les­sons one learns in ac­count­ing anal­y­sis is that ra­tios have to make eco­nomic sense—the nu­mer­a­tor and de­nom­i­na­tor have to re­fer to the same thing—which is not the case here. In­stead we should be com­par­ing do­mes­tic prof­its to GDP.

  Why com­pare cor­po­rate rev­enue to state tax rev­enue?

  One of the key parts of the ar­gu­ment is that we can mean­ingfully com­pare cor­po­rate rev­enue to state tax rev­enue as a mea­sure of power. How­ever the jus­tifi­ca­tion for this in the ar­ti­cle is very weak:

  Re­cently, re­searchers have pro­posed that states’ “bud­gets” (tax rev­enues) and cor­po­ra­tions’ “bud­gets” (rev­enues minus their prof­its) are com­pa­rable in that they are crude, but use­ful prox­ies for their power (de­spite cor­po­ra­tions and states spend­ing on differ­ent things).

  How do this ran­dom PhD stu­dent re­searchers jus­tify this ap­proach?

  For this pur­pose, we com­pare the rev­enues of states (mainly taxes col­lected) with the rev­enues of cor­po­ra­tions, as sug­gested by Jeffrey Har­rod, who ar­gued that we should see rev­enues (minus prof­its) as a “bud­get” of firms in anal­ogy to gov­ern­ments.6767 Har­rod, “The Cen­tury of the Cor­po­ra­tion”.View all notes Ad­mit­tedly, this is a crude proxy for power or in­fluence, nonethe­less it is in­struc­tive to jux­ta­pose the fi­nan­cial scale of states and cor­po­ra­tions di­rectly. (link)

  That’s right—they ba­si­cally just posit it. There is no ac­tual jus­tifi­ca­tion for as­sum­ing that a com­pany with rev­enues of $100bn is more pow­er­ful than a state with taxes of $90bn.

  It makes sense that there is no jus­tifi­ca­tion—be­cause the state­ment is clearly ab­surd. Con­sider the ex­am­ple from the ar­ti­cle: com­par­ing the Aus­tralian State to Wal­mart, the lat­ter of which is sup­pos­edly more pow­er­ful.

  • • Aus­tralia has a pow­er­ful mil­i­tary, in­clud­ing ten fri­gates, two de­stroy­ers and six sub­marines. Their air force has over 100 com­bat planes, and they have 30,000 ac­tive sol­diers in the mil­i­tary. Us­ing this they have par­ti­ci­pated in a num­ber of wars. In con­trast, to my knowl­edge Wal­mart has zero fri­gates, zero de­stroy­ers, zero sub­marines, and zero com­bat planes.

    • In the event of a war, fur­ther­more, the Aus­tralian gov­ern­ment might be able to in­tro­duce con­scrip­tion, dra­mat­i­cally in­creas­ing their man­power. Wal­mart is un­likely to be able to do so as they have no such le­gal au­thor­ity.

    • With the benefit of these weapons, the Aus­tralian gov­ern­ment is able to achieve a great deal. For ex­am­ple, they are able to force al­most ev­ery sin­gle Aus­tralian to pay them taxes. In con­trast, Wal­mart is re­li­ant on peo­ple vol­un­tar­ily giv­ing them money, has to offer mer­chan­dise in re­turn, and is un­able to im­prison peo­ple who re­fuse. Speak­ing of which, the Aus­tralian gov­ern­ment has around 40,000 peo­ple in prison—Wal­mart con­spicu­ously lacks its own prison sys­tem. Even in cases where some­one di­rectly steals from them, Wal­mart is forced to rely on the as­sis­tance of the US gov­ern­ment to deal with the situ­a­tion.

    • Fi­nally, the Aus­tralian state is able to effec­tively bar large groups of peo­ple from en­ter­ing its ter­ri­tory on the ba­sis of their na­tion­al­ity, po­ten­tially im­pris­on­ing them for long pe­ri­ods on re­mote is­lands to achieve this. In con­trast, Wal­mart’s abil­ity to pre­vent peace­ful peo­ple from en­ter­ing its stores is ex­tremely limited.

  This ar­ti­cle, how­ever, rather than in­ter­ro­gat­ing the claim that cor­po­rate rev­enues can be di­rectly com­pared to state taxes, in­stead ac­cepts it com­pletely. In the space of a few short para­graphs we move from propos­ing it as an in­ter­est­ing idea to ac­cept­ing it as the ba­sis for claims like:

  71 out of the 100 most pow­er­ful ac­tors are corporations

  And some of the claims made in this sec­tion are sim­ply bizarre. For ex­am­ple,

  “Shell’s rev­enue/​bud­get to sell fos­sil fuels is $272bn”

  Where does this num­ber come from? I checked their 2018 num­bers on bloomberg and got rev­enue of $388bn. More im­por­tantly, rev­enue is not at all the same thing as ‘bud­get to sell fos­sil fuels’. That would be part of SG&A, which is ob­vi­ously much smaller—around $11bn. In­stead, most of their rev­enue is spent buy­ing phys­i­cal in­puts for e.g. their re­fin­ing busi­ness.

  Or take this one:

  For in­stance, Ap­ple spends >$1bn/​year on mar­ket­ing. Lob­by­ing bud­gets could in­crease to a similar level.

  Ac­cord­ing to OpenSe­crets, the largest US cor­po­rate spender on lob­by­ing in 2018 was Alpha­bet (Google), spend­ing just un­der $23m. It is tech­ni­cally true that their bud­get could in­crease to over a billion; how­ever, this would re­quire a >33,000% in­crease.

  Or this claim:

  In con­trast, state power is cur­tailed by manda­tory and en­ti­tle­ment spend­ing tak­ing up an in­creas­ing share of their bud­gets (e.g. >70% in the US).

  This spend­ing is not re­ally manda­tory at all—if states did not want to give out benefits there is no re­quire­ment for them to do so. We know this is true be­cause there have been many pow­er­ful states that spent very lit­tle on benefits. If states are weak­ened by en­ti­tle­ment spend­ing, it is be­cause they choose to do so.

  In con­trast, cor­po­ra­tions have to spend the vast ma­jor­ity of their rev­enue on raw re­sources, salaries etc. in or­der to pro­duce the goods and ser­vices they sell. Buy­ing pota­toes is not op­tional for McDon­alds!

  Do cor­po­ra­tions cause ex­po­nen­tially larger ex­ter­nal­ities over time?

  This sec­tions seems like a ran­dom grab-bag of com­plaints with a lack of model-based think­ing. In par­tic­u­lar, there doesn’t seem to be any rea­son to think that ex­ter­nal­ities grow ex­po­nen­tially, ex­cept that cor­po­rates (and the econ­omy as a whole) grow ex­po­nen­tially. The differ­ence be­tween ex­ter­nal­ities grow­ing ex­po­nen­tially rel­a­tive to cor­po­rate size, and ex­ter­nal­ities re­main­ing fixed rel­a­tive to cor­po­rate size, might seem like a triv­ial point, but one of the pro­posed ‘solu­tions’ - to break up large cor­po­ra­tions—ac­tu­ally ends up rest­ing on this con­fu­sion.

  Firstly we should note that the cita­tions men­tioned do not ac­tu­ally sup­port the ar­ti­cle’s point:

  There is no con­sen­sus amongst economists on whether gov­ern­ments should reg­u­late the size of big cor­po­ra­tions more gen­er­ally (e.g. through pre­vent­ing M&A or break­ing them up).[62],[63]

  The linked polls dis­cuss spe­cific gov­ern­ment in­ter­ven­tions in spe­cific cases, as is stan­dard in the an­titrust liter­a­ture, not gen­eral reg­u­la­tion of cor­po­rate size. I am not aware of any cred­ible economists who be­lieve in a carte blanche re­stric­tion on the size of cor­po­ra­tions. I have liter­ally never heard any economist sug­gest a policy which would re­strict com­pa­nies that have grown or­gan­i­cally due to economies of scale and good ex­e­cu­tion, like Net­flix or Ap­ple.

  But more im­por­tantly, break­ing large cor­po­rates up into smaller pieces would only re­duce ex­ter­nal­ities if those ex­ter­nal­ities were su­per-lin­ear in cor­po­rate size. This doesn’t seem to be the case how­ever—if you split up cor­po­ra­tion with two fac­to­ries, you still have two com­pa­nies with one fac­tory each, pro­duc­ing just as much pol­lu­tion in to­tal. (They might even pro­duce slightly more as smaller firms are of­ten slower to adopt new tech­nolo­gies).

  The ar­ti­cle might also have benefited from dis­cus­sion of the track records of non-cap­i­tal­ist sys­tems, like the pointless slaugh­ter caused by the so­viet whale quota sys­tem. The ar­ti­cle men­tions Fukushima… but not Ch­er­nobyl, de­spite the lat­ter be­ing an ex­am­ple of much worse be­havi­our. Similarly, it blames cor­po­rates for obe­sity, but does not men­tion the difficul­ties com­mu­nist states had in avoid­ing mass star­va­tion. Nor at these prob­lems only in so­cial­ist coun­tries—for ex­am­ple, even in the US pri­vately-run power plants are cleaner than those run by the gov­ern­ment:

  Our em­piri­cal sub­jects are pub­lic and pri­vate en­tities’ com­pli­ance with the U.S. Clean Air Act and Safe Drink­ing Water Act. We find that, com­pared with pri­vate firms, gov­ern­ments vi­o­late these laws sig­nifi­cantly more fre­quently and are less likely to be pe­nal­ized for vi­o­la­tions.

  It is not pos­si­ble to re­duce cor­po­rate power in a vac­uum—the pro­posed solu­tions largely in­stead trans­fer it to gov­ern­ments. But given that, even ac­cord­ing to the bizarre rev­enue=tax method­ol­ogy adopted, states were still more pow­er­ful than their cor­po­ra­tions, this would mean con­cen­trat­ing more power in the most pow­er­ful en­tities. En­tities which in the past have re­peat­edly brought mankind to the brink of nu­clear war—an ex­ter­nal­ity far greater than that of any mere cor­po­ra­tion.

• Larks 2 Jul 2019 3:22 UTC

  10 points

  on: Op­ti­miz­ing Ac­tivi­ties Fairs

  Great ar­ti­cle! Glad to hear you’re do­ing well. To think I thought we were be­ing ag­gres­sive get­ting 3 ta­bles back in 2012!

  Two more tips I found helpful:

  • If you can, have a term card printed out to give peo­ple. Ideal place for a cou­ple of para­graphs about EA and a list of all your events for the term.

  • Pre-fill the signup sheet with your own names and emails, be­cause an empty list makes you look un­pop­u­lar, and no-one wants to join an un­pop­u­lar club.

  • Prac­tice speak­ing the pitch out loud a cou­ple of times be­fore­hand. It’s likely that you’ll grav­i­tate to­wards a slightly differ­ent word­ing, which should be­come very nat­u­ral.

  • I spent a lot of the time in front (and slightly to the side) of the table, which was much bet­ter for in­ter­cept­ing peo­ple. This did an­noy the fair staff, but not quite enough to in­cur any ac­tual sanc­tion. How­ever, this might mat­ter less if you can to­tally dom­i­nate with 6 ta­bles!

  • Usu­ally I think di­ver­sity stuff is non­sense, but I did try to en­sure there was always one girl and one guy on the stall, which I agree is very helpful in this case.

  • Slightly older guys might want to shave.

  • Smile! While ex­haust­ing, this can be a very fun ex­pe­rience.

• Larks 27 Jun 2019 20:53 UTC

  19 points

  on: Needed EA-re­lated Ar­ti­cles on the English Wikipedia

  In­ter­est­ing list. I know some peo­ple and or­gani­sa­tions (which I am ob­vi­ously not go­ing to name) pre­fer not to have Wikipe­dia pages—is it pos­si­ble that some of the above groups might fall into this cat­e­gory?

• Larks 27 Jun 2019 20:53 UTC

  3 points

  on: [Link] A shift in ar­gu­ments for AI risk

  Meta: It might be helpful to have a fea­ture to au­to­mat­i­cally link linkposts to the rele­vant page on LW, if the ar­ti­cle had pre­vi­ously been linkposted there.

  In this case there wasn’t much dis­cus­sion any­way, though this post and com­ments also seem very rele­vant.

• Larks 27 Jun 2019 3:27 UTC

  9 points

  on: “Mo­ral Bias and Cor­rec­tive Prac­tices” and the pos­si­bil­ity of an on­go­ing moral catastrophe

  Thanks very much for shar­ing this. I ac­tu­ally ex­pected to like it a lot—it’s an is­sue I think about a lot—but ended up be­ing quite un­satis­fied.

  His­to­ri­ans are of­ten crit­i­cised for draw­ing quite gen­eral con­clu­sions from rel­a­tively sparse data. For ex­am­ple, I re­call read­ing See­ing Like a State and be­ing con­cerned that he was draw­ing con­clu­sions from what seemed like a mere hand­ful of data points, even if new and in­ter­est­ing ones. Or re­cently I read the re­cent BIS re­port on the im­pact of bank cap­i­tal re­quire­ments on fi­nan­cial crises and growth, and was con­cerned about the rel­a­tively small sam­ple set these con­clu­sions were be­ing based on—even though this data spanned many coun­tries and over a hun­dred years.

  But in this case, An­der­son is draw­ing ridicu­lously strong con­clu­sions from one sin­gle data point—a some­what bi­ased ac­count of the end of slav­ery in a sin­gle coun­try. There is no recog­ni­tion that other is­sues in other ge­ogra­phies and other points in time played out rather differ­ently—even ex­am­ples as prox­i­mate as the his­tory of slav­ery in England and the Em­pire. De­spite this, she pre­sents her con­clu­sions as be­ing a nec­es­sary law of his­tory:

  Stronger meth­ods are needed to coun­ter­act the bi­ases in­duced by so­cial power. My case study of a so­ciety-wide change in moral be­lief, from proslav­ery to abo­li­tion­ist, fo­cused on two such meth­ods. First, con­tentious poli­tics—ac­tive, prac­ti­cal, mass re­sis­tance to the moral claims em­bod­ied in so­cial in­sti­tu­tions en­forced by and cater­ing to the pow­er­ful—is needed to ac­ti­vate gen­uine prac­ti­cal rea­son­ing across all lev­els of so­ciety. The pow­er­ful won’t re­ally listen to rea­son—that is, to claims from be­low—un­til they no longer have the power to rou­tinely en­force their de­sires. Se­cond, the sub­or­di­nated and op­pressed must ac­tively par­ti­ci­pate in that con­tention. [em­pha­sis added]

  It’s also the case that, if her views are true, this is very bad news. A key part of EA is pro­mot­ing the welfare of groups with no voice, like an­i­mals or the as-yet-un­born. Groups that we have con­cluded de­serve con­sid­er­a­tion, not be­cause of ac­tivism by these groups, but by con­sid­er­ing our moral in­tu­itions. If the only way of im­prov­ing the treat­ment of a group is for that group to prac­tice ‘mass re­sis­tance’ then it is liter­ally im­pos­si­ble for an­i­mal rights to ever im­prove, or the rights of chil­dren, or fu­ture gen­er­a­tions. For­tu­nately, the fact that we do have an­i­mal welfare laws, and child welfare laws, and en­vi­ron­men­tal laws, and so on, sug­gests that she is in­deed wrong.

  She seems to miss that moral de­liber­a­tion and re­flec­tive equil­ibrium have a vi­tal ad­van­tage: they are asym­met­ric weapons, that cut more read­ily against false­hood, while tend­ing to give suc­cor to the truth. In­deed, they are at the core of the EA pro­ject.

  In con­trast, she pre­sents a very dystopian world, where there is lit­tle true moral de­bate—or if there is, it is of lit­tle con­se­quence. All there seems to be is a strug­gle be­tween pow­er­ful groups and al­most-as-pow­er­ful groups. There is no great ten­dency here for the arc of his­tory to bend to­wards jus­tice—just lead­ers tel­ling the masses that they have been op­pressed by a cruel minor­ity and need to rise, which his­tory has shown to be a harbinger of tyranny at least as of­ten as of liber­a­tion.

• Larks 22 Jun 2019 18:40 UTC

  11 points

  on: EA Men­tal Health Sur­vey: Re­sults and Anal­y­sis.

  One per­son re­ported miss­ing 150 hours of work due to men­tal health prob­lems, which if true would put them as the third most pro­duc­tive per­son in the en­tire sam­ple, at 75 hours /​ week. It seems plau­si­ble to me that this per­son in­stead mis­es­ti­mated their la­tent pro­duc­tivity, and I won­der if this might be true for other peo­ple. Per­haps in the ab­sence of men­tal ill­ness they would have come up with some other rea­son/​ex­cuse to at­tribute their lack of pro­duc­tivity.

• Larks 14 Jun 2019 20:57 UTC

  4 points

  on: Melbourne Univer­sity Effec­tive Altru­ism Study

  Some of the ques­tions seem some­what ge­og­ra­phy spe­cific—I sus­pect differ­ent moral in­tu­itions may be at play de­pend­ing on the na­tion­al­ity of the re­spon­dent for in­dex­i­cal rea­sons. You might want to con­sider cus­tomiz­ing the prompts based on IP ad­dress or similar. (De­liber­ately vague to avoid spoilers).

• Larks 8 Jun 2019 19:45 UTC

  8 points

  on: Could the crowd­fun­der to pros­e­cute Boris John­son be a high im­pact dona­tion op­por­tu­nity?

  Just a few days af­ter this post was made, the law­suit was thrown out by the high court as poli­ti­cally mo­ti­vated and vex­a­tious. Ap­par­ently the ac­tivist who brought the law­suit spent more than he raised, so it seems that the marginal dona­tion is ba­si­cally just a cash trans­fer to him.

• Larks 6 Jun 2019 21:49 UTC

  10 points

  on: [Link] Book Re­view: The Se­cret Of Our Suc­cess | Slate Star Codex

  See also this fol­low-up for ex­tended quotes:

  https://​​slat­estar­codex.com/​​2019/​​06/​​05/​​list-of-pas­sages-i-high­lighted-in-my-copy-of-the-se­cret-of-our-suc­cess/​​

  In 1971, the an­thro­pol­o­gist David Boyd was liv­ing in the New Guinea village of Irakia, and ob­served in­ter­group com­pe­ti­tion via pres­tige-bi­ased group trans­mis­sion. Con­cerned about their low pres­tige and weak pig pro­duc­tion, the se­nior men of Irakia con­vened a se­ries of meet­ings to de­ter­mine how to im­prove their situ­a­tion. Numer­ous sug­ges­tions were pro­posed for rais­ing their pig pro­duc­tion but af­ter a long pro­cess of con­sen­sus build­ing the se­nior men of the village de­cided to fol­low a sug­ges­tion made by a pres­ti­gious clan-leader who pro­posed that they “must fol­low the Fore’” and adopt their pig-re­lated hus­bandry prac­tices, rit­u­als, and other in­sti­tu­tions. The Fore’ were a large and suc­cess­ful eth­nic group in the re­gion, who were renowned for their pig pro­duc­tion. The fol­low­ing prac­tices, be­liefs, rules, and goals were copied from the Fore’, and an­nounced at the next gen­eral meet­ing of the com­mu­nity:

  1) All villagers must sing, dance and play flutes for their pigs. This rit­ual causes the pigs to grow faster and big­ger. At feasts, the pigs should be fed first from the oven. Peo­ple are fed sec­ond.

  2) Pigs should not be kil­led for break­ing into an­other’s gar­den. The pig’s owner must as­sist the owner of the gar­den in re­pairing the fence. Dis­putes will be re­solved fol­low­ing the dis­pute re­s­olu­tion pro­ce­dure used among the Fore’.

  3) Send­ing pigs to other villages is tabooed, ex­cept for the offi­cial fes­ti­val feast.

  4) Women should take bet­ter care of the pigs, and feed them more food. To find ex­tra time for this, women should spend less time gos­siping.

  5) Men must plant more sweet pota­toes for the women to feed to the pigs, and should not de­part for wage la­bor in dis­tant towns un­til the pigs have grown to a cer­tain size.

  The first two items were im­ple­mented im­me­di­ately at a rit­ual feast. David stayed in the village long enough to ver­ify that the villagers did adopt the other prac­tices, and that their pig pro­duc­tion did in­crease in the short term, though un­for­tu­nately we don’t know what hap­pened in the long-run.

• Larks 6 Jun 2019 1:27 UTC

  25 points

  on: Could the crowd­fun­der to pros­e­cute Boris John­son be a high im­pact dona­tion op­por­tu­nity?

  It’s not re­ally about Brexit, it’s about hon­esty in poli­tics.

  I think this is a lit­tle naive. There are many ex­am­ples of poli­ti­ci­ans ly­ing—pick­ing just one is a clear ex­am­ple of poli­ti­cal bias. If they want this to be about hon­esty, they should si­mul­ta­neously pros­e­cute a leader Con­ser­va­tive and Labour poli­ti­cian, or Leave and Re­main- even if they think the suit against Boris is the most likely to suc­ceed. I think the fact that they’re called ‘Brexit Jus­tice’ makes their mo­ti­va­tions pretty un­am­bigu­ous.

  But even with an ideal­ised case that was non-par­ti­san, the is­sue at stake isn’t truth vs lies but who gets to de­ter­mine what is ac­cept­able to say. If they win, they will es­tab­lish a prece­dent that it is ille­gal for op­po­nents of the gov­ern­ment to say things that the state holds to be lies. Even if the cur­rent gov­ern­ment only used this for good pur­poses, there is no guaran­tee that a fu­ture one would. His­tory is full of ex­am­ples of gov­ern­ments sup­press­ing ‘lies’ that turned out to be the truth.

• Larks 30 May 2019 21:22 UTC

  2 points

  in reply to: Linch's comment on: 2018 AI Align­ment Liter­a­ture Re­view and Char­ity Comparison

  You’re wel­come! And thanks, fixed.

• Larks 30 May 2019 1:16 UTC

  15 points

  in reply to: Jeff_Kaufman's comment on: There’s Lots More To Do

  Do you mean the num­ber of op­por­tu­ni­ties in the fu­ture, or the abil­ity to donate larger amounts of money right now? We could do:

  What is the most dol­lars you would be will­ing to donate in or­der to save the life of a ran­domly cho­sen hu­man? As­sume this is the only op­por­tu­nity you’ll ever get to save a life by donat­ing—all other money you have must be spent on your­self and your fam­ily.

  and also the en­dow­ment effect re­ver­sal:

  If offered a choice be­tween sav­ing a ran­dom stranger’s life and an amount of money, what is the small­est num­ber of dol­lars you would have to be offered to choose that op­tion? As­sume you will only be offered this once, and do not have any op­por­tu­ni­ties to spend or donate money to save other peo­ple.

  Your ver­sion seems good too, though I would worry that in­tro­duc­ing the tem­po­ral el­e­ment and back­ground of progress might bias things in some way.

• Larks 29 May 2019 23:46 UTC

  18 points

  on: There’s Lots More To Do

  He’s set­ting a thresh­old of $5k for how much we’d be will­ing to pay to avert a death, which is much too low. I do agree there is some thresh­old at which you’d be very rea­son­able to stop try­ing to help oth­ers and just do what makes you happy. Where this thresh­old is de­pends on many things, es­pe­cially how well-off you are, but I would ex­pect it to be more in the $100k range than the $5k range for rich-coun­try effec­tive al­tru­ists.

  
  

  It would be in­ter­est­ing to see some data on this. Maybe the EA sur­vey could ask some­thing about it? Some­thing like:

  What is the most you would be will­ing to spend in or­der to save the life of a ran­domly cho­sen hu­man? As­sume that AMF and other char­i­ties do not ex­ist—the al­ter­na­tive is spend­ing the money on your­self.

  I can imag­ine the dis­tri­bu­tion be­ing quite wide.