Taking Systemic Change Seriously

This is meant to be a rough re­sponse to the at­ti­tude that sys­temic change is too difficult/​in­tractable as well as a re­sponse by perfor­mance that EAs don’t think about sys­temic change. Note: by sys­temic change I’m refer­ring to many pos­si­ble changes in the fun­da­men­tal struc­ture of eco­nomic, poli­ti­cal and in­ter­na­tional sys­tems, not nec­es­sar­ily to what lots of peo­ple naively as­sume to be the one true proper method of sys­temic changeTM.

EAs have seemed to con­gre­gate at the ex­tremes of di­rect ro­bust aid (poverty, veg ads) and mas­sive tech­nolog­i­cal risks and trans­for­ma­tion (x risk, abo­li­tion­ism) with­out many peo­ple in the mid­dle. This is cu­ri­ous and cries out for an ex­pla­na­tion. There are a few peo­ple work­ing in policy spaces to im­prove how gov­ern­ments deal with the afore­men­tioned is­sues, but none of that re­ally counts as mid­dle ground or sys­temic change in my opinion (and many peo­ple out­side EA would agree). It’s just ap­plied ac­tivism and poli­tics. Sys­temic change means im­prov­ing hu­man so­ciety’s abil­ity to solve many prob­lems and be more eth­i­cal in a gen­eral long term sense. Some things that would be sys­temic change in­clude chang­ing the way our poli­ti­cal sys­tems op­er­ate, al­ter­ing the struc­ture of the in­ter­na­tional or­der and re­mov­ing the in­fluence of cap­i­tal on so­ciety.

Since so few EAs have se­ri­ously ap­proached sys­temic change, it’s likely that there are more un­der­de­vel­oped ideas in this in­ter­ven­tion space than in other in­ter­ven­tion spaces, which in­di­cates that it might be a bet­ter cause area than we would naively ex­pect. Also, if we are un­cer­tain about cause ar­eas then sys­temic change makes sense as a way to at­tack a va­ri­ety of prob­lems (but if you think that just a few par­tic­u­lar causes are by far the most im­por­tant then spend­ing your time on sys­temic change seems in­effi­cient). I think sys­temic change makes the most sense if you ex­pect new im­por­tant is­sues to arise in the fu­ture. Th­ese con­sid­er­a­tions in­di­cate that the value of sys­temic change is co­var­i­ant with the value of move­ment build­ing.

I want to sketch a bet­ter pic­ture of what sys­temic change should ‘look like.’ I can give sev­eral desider­a­tum for a sys­temic change effort:

  1. It should be great in ex­pec­ta­tion. In other words, look­ing at the po­ten­tial and likely re­sults of ac­tivism should re­veal large im­prove­ments for the fu­ture of sen­tient life.

  2. It should be ro­bust. It should not rely upon any one poli­ti­cal ide­ol­ogy, any one em­piri­cal ex­pec­ta­tion, or any one frame­work of moral­ity or de­ci­sion the­ory. Given the opac­ity of sys­temic change and the fact that we will prob­a­bly never get much re­li­able feed­back about the out­comes of our ac­tions, we should de­mand high prior con­fi­dence.

  3. It should be scal­able. At least, it should be the sort of thing where a tiny frac­tion of the EA com­mu­nity—maybe fewer than 10 peo­ple—could ac­com­plish some­thing non-neg­ligible OR have a small prob­a­bil­ity of achiev­ing some­thing sig­nifi­cant. Other­wise we will prob­a­bly be wast­ing our efforts for the time be­ing.

  4. It should be ide­olog­i­cally safe de­pend­ing on how wide­spread and pub­lic we want the cam­paign to be. Ideally, it should be ide­olog­i­cally pos­i­tive by get­ting new peo­ple on board with EA in gen­eral. But if we are per­ceived as hav­ing views which are re­pug­nant or offen­sive then we may lose in­fluence with many peo­ple. This is a real pos­si­bil­ity: peo­ple have sneered at EA be­cause of at­ten­tion given to ex­is­ten­tial risk and peo­ple have re­jected it be­cause of the re­fusal to broadly con­demn cap­i­tal­ism. You can­not please ev­ery­one but we should think about these fac­tors, es­pe­cially when we want the move­ment to have in­sti­tu­tional clout with cur­rent elites. It sucks, but it’s the in­tel­li­gent road to take.

Be­fore we get into par­tic­u­lar in­ter­ven­tion ideas, the first ques­tions we should an­swer are: how good of a cause area do we think sys­temic change might be, a pri­ori? And are the above desider­ata suit­able?

Now I can think of sev­eral ex­am­ples of sys­temic change which would fit some or all of the above crite­ria.
  1. Ev­i­dence/​sci­ence/​im­pact based gov­er­nance: Gen­er­ally in­still­ing a cul­ture of more ra­tio­nal de­ci­sion­mak­ing in the gov­ern­ment would im­prove its abil­ity to im­ple­ment a wide range of pro­grams effec­tively, as would changes in the struc­ture of poli­ti­cal pro­cesses that are de­signed to take bet­ter guidelines into ac­count. There are more spe­cific pro­pos­als that can be in­ves­ti­gated in this area, such as futarchy.

  2. World gov­ern­ment: For very ba­sic game-the­o­retic rea­sons, more cred­ible power in the hands of in­ter­na­tional or­ga­ni­za­tions and the U.N. in par­tic­u­lar could go a long way to­wards solv­ing global co­or­di­na­tion prob­lems (like ex­is­ten­tial risks) and re­duc­ing war. It would set a prece­dent in poli­ti­cal re­la­tions that might con­tinue in­definitely. Re­mov­ing the veto from the U.N. se­cu­rity coun­cil is a pos­si­ble step in this di­rec­tion. On the other hand, shoring up the E.U. could be crit­i­cal to pre­vent­ing a re­verse trend in the com­ing decades. I think this cause is po­ten­tially the best, de­pend­ing on how well it can be meshed with the very defen­si­ble re­al­ist un­der­stand­ing of in­ter­na­tional re­la­tions.

  3. Public own­er­ship of the means of pro­duc­tion: plac­ing more ques­tions about pro­duc­tion in demo­cratic hands could go a long way to­wards re­duc­ing poverty and in­ter­na­tional con­flict, ac­cord­ing to var­i­ous the­o­rists. How­ever, these claims are con­tentious and di­vi­sive. It is pos­si­ble that im­ple­ment­ing this change would re­duce ex­is­ten­tial risks as well, due to alle­vi­at­ing co­or­di­na­tion prob­lems and ram­pant con­sump­tion drives.

Now the ques­tions to be an­swered are: how good might the above causes be, and what other types of sys­temic change should we in­ves­ti­gate?