What Helped the Voiceless? Historical Case Studies

[Last edit: 10/​15/​20]

[Edit: Shorter ver­sion here.]

In­tro­duc­tion and Sum­mary:

Fu­ture gen­er­a­tions, non-hu­man an­i­mals, and other voice­less groups are harm­fully ne­glected in to­day’s policy mak­ing. What strate­gies for chang­ing this can ad­vo­cates of ne­glected groups learn from times when ex­cluded groups gained poli­ti­cal pro­tec­tions?

Peo­ple in the EA com­mu­nity have called for in­ves­ti­gat­ing moral and poli­ti­cal progress. Still, be­fore this re­port, there were (as far as I’m aware) no EA-mo­ti­vated ac­counts of why coun­tries other than the UK abol­ished the slave trade and slav­ery, of why coun­tries around the globe be­came democ­ra­cies, or of why these in­ter­na­tional shifts hap­pened around the in­dus­trial rev­olu­tion. This re­port aims to provide such an ac­count, and to ar­gue that it has cer­tain im­pli­ca­tions for how we can best help ne­glected groups.

(In ad­di­tion to the most im­me­di­ately rele­vant take­aways from case stud­ies, it seems broadly use­ful for effec­tive al­tru­ists to bet­ter un­der­stand his­tor­i­cal causes of the im­prove­ment and de­te­ri­o­ra­tion of well-be­ing.)

I ex­am­ined his­tor­i­cal case stud­ies of global policy shifts that greatly benefited ex­cluded groups (es­pe­cially groups with limited or no ca­pac­i­ties for self-ad­vo­cacy):

  • The abo­li­tion of the slave trade and slavery

  • Ex­ten­sions of the vot­ing fran­chise over eco­nomic lines

  • Cli­mate change mitigation

  • Ge­netic en­g­ineer­ing governance

Of these, I fo­cused on abo­li­tion and de­moc­ra­ti­za­tion. Sum­maries of these case stud­ies are available here.

In many of the cases stud­ied, poli­cies were cre­ated mainly for the benefit of pow­er­ful groups, and they hap­pened to greatly benefit ex­cluded groups. In other cases, ex­cluded groups benefited be­cause some poli­ti­cal ac­tors tried to benefit them for their own sake. Benefits come more re­li­ably when pow­er­ful ac­tors try to benefit a group, so durable poli­ti­cal power would be ex­tremely valuable for ex­cluded groups, such as fu­ture gen­er­a­tions.

Th­ese case stud­ies mo­ti­vate the ques­tion: How do groups gain or lose rel­a­tively durable forms of poli­ti­cal powerle­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion? I in­tro­duce and ar­gue for a qual­i­ta­tive, ra­tio­nal-choice model that makes pre­dic­tions about poli­ti­cal in­clu­sion, with sup­port from many his­tor­i­cal ex­am­ples.

The model sug­gests that these fac­tors make it more likely that tran­si­tions to poli­ti­cal ex­clu­sion will oc­cur and per­sist:

  • Op­por­tu­ni­ties for prof­itable exploitation

  • Costs of inclusion

  • Ex­clu­sive values

The model also sug­gests that these fac­tors make it more likely that tran­si­tions to poli­ti­cal in­clu­sion will oc­cur and per­sist:

  • A group’s ca­pac­ity for resistance

  • Strate­gic alliances

  • In­ter-so­cietal pressure

  • In­clu­sive values

We can sum­ma­rize the model with this pic­ture:

This model, and the case stud­ies be­hind it, cast doubt on pre­vi­ous the­o­ries of moral cir­cle ex­pan­sion. They sug­gest that in­clu­sive val­ues did mat­ter for in­clu­sive progress, but sev­eral other mo­tives each tended to be similarly in­fluen­tial, or more. In­clu­sive val­ues, then, are not as his­tor­i­cally im­por­tant as is of­ten as­sumed.

If that’s the case, you might ask, why do we have rel­a­tively in­clu­sive val­ues to­day? An ex­pla­na­tion that bet­ter fits his­tor­i­cal ev­i­dence is that so­cial val­ues mainly changed retroac­tively to be­come more in­clu­sive, af­ter other fac­tors mo­ti­vated poli­ti­cal in­clu­sion. You might also won­der, if these other fac­tors mo­ti­vated in­clu­sion, why have in­clu­sive shifts have been so fre­quent in the mod­ern era? This is ar­guably be­cause eco­nomic growth/​in­dus­tri­al­iza­tion strength­ened the re­sis­tance of ex­cluded groups in sev­eral ways, while in other cases, poli­cies mainly aimed at benefit­ing pow­er­ful groups hap­pened to benefit pow­er­less groups.

What im­pli­ca­tions do these case stud­ies and this model have for sup­port­ing groups that can­not ex­ert in­fluence, par­tic­u­larly fu­ture gen­er­a­tions? I make these pro­vi­sional recom­men­da­tions:

  • Given the his­tor­i­cal im­por­tance of the in­fluence of ex­cluded groups and luck, past trends to­ward greater in­clu­sion do not strongly im­ply that poli­ti­cal in­clu­sion will even­tu­ally ex­tend to voice­less groups.

  • On spe­cific is­sues that mat­ter a lot for fu­ture gen­er­a­tions, fo­cus on show­ing pow­er­ful ac­tors that it is in their own in­ter­ests to take ac­tions that hap­pen to greatly benefit fu­ture gen­er­a­tions.

  • Pur­sue sup­port that is con­cen­trated at the na­tional level to have some sup­port on the in­ter­na­tional level, which is ex­tremely valuable.

  • When op­pos­ing a policy, em­pha­size highly salient down­sides: those that are very con­crete (easy to imag­ine), hor­rify­ing, and be­liev­able (highly plau­si­ble/​already pre­sent).

  • Take ad­van­tage of differ­ences in pa­tience—pur­sue poli­cies whose last­ing benefits and po­ten­tially con­cen­trated costs will not come for years or decades.

  • When de­sign­ing any Fu­ture Gen­er­a­tions in­sti­tu­tion, in­clude an ex­plicit man­date for it to con­sider long-term pros­per­ity (in ad­di­tion to ex­is­ten­tial risks aris­ing from tech­nolog­i­cal de­vel­op­ment and en­vi­ron­men­tal sus­tain­abil­ity), in or­der to get more last­ing sup­port from busi­ness.

  • When you have power, do not aim to make the biggest im­prove­ments you can; aim to make im­prove­ments that are ac­cepted widely enough to per­sist, even af­ter you lose power.

Other im­pli­ca­tions in­clude:

  • When con­sid­er­ing long-term risks of ex­treme suffer­ing, we should as­sign more weight to sce­nar­ios in which main­tain­ing some form of large-scale suffer­ing is in the in­ter­ests of pow­er­ful ac­tors, and less weight to sce­nar­ios in which it is not, since suffer­ing is more likely to per­sist in the former cases.

  • If it is not ac­com­panied by other de­vel­op­ments that push in differ­ent di­rec­tions, the au­toma­tion of hu­man la­bor at mas­sive scales will, by de­fault, severely erode democ­ra­cies globally.

More de­tail is available be­low. If you only want to read some of this re­port, you might be es­pe­cially in­ter­ested in the fol­low­ing sec­tions:

Con­text and Ac­knowl­edge­ments:

This is a re­port of the most use­ful-seem­ing parts of the re­search I con­ducted over the sum­mer, through the Stan­ford Ex­is­ten­tial Risks Ini­ti­a­tive (SERI). The views ex­pressed in this re­port do not nec­es­sar­ily re­flect those of SERI or any af­fili­ated or­ga­ni­za­tions.

I am not a pro­fes­sional his­to­rian, so, to an even larger de­gree than usual, all im­por­tant im­pli­ca­tions should be heav­ily scru­ti­nized and con­sid­ered to­gether with the im­pli­ca­tions of other analy­ses.

I am very grate­ful to Jeremy We­in­stein for his men­tor­ship, to SERI for its sup­port, to Felipe Calero for recom­men­da­tions on biose­cu­rity read­ings, and to Linchuan Zhang as well as Zach Fre­itas-Groff for their feed­back.

His­tor­i­cal Case Stud­ies:

Will you not em­brace it? So much good has not been done by one effort as in the Prov­i­dence of God it is now your high priv­ilege to do. May the vast fu­ture not have to lament that you have ne­glected it!

- “The Civil War in Amer­ica,” an 1862 ar­ti­cle from the UK’s Belfast Newslet­ter, call­ing on Amer­i­cans to abol­ish slavery


I de­scribe his­tor­i­cal case stud­ies of times when policy change greatly benefited ex­cluded groups, mainly: mod­ern abo­li­tion, de­moc­ra­ti­za­tion, cli­mate change miti­ga­tion, and ge­netic en­g­ineer­ing gov­er­nance. The case stud­ies are com­par­a­tive or in­ter­na­tional in their fo­cus. After the in­tro­duc­tion, be­low, sum­maries of the case stud­ies are available. Th­ese re­views are not ex­haus­tive; they do not in­clude all im­por­tant nu­ances.


Policy has shifted in ways that greatly benefited largely or com­pletely voice­less groups in the past. Study­ing how these changes hap­pened may yield valuable in­sights for how peo­ple can bring about an­other such change: the es­tab­lish­ment of se­ri­ous poli­ti­cal pro­tec­tions for fu­ture gen­er­a­tions. For this pur­pose, I in­ves­ti­gated sev­eral his­tor­i­cal in­stances when policy shifts greatly benefited groups with limited or no ca­pac­i­ties for self-ad­vo­cacy.

Here, I give in­ter­na­tional ac­counts of four broad policy shifts, re­stricted in scope to the mod­ern era: the abo­li­tion of slav­ery, ex­pan­sions of vot­ing rights, the miti­ga­tion of cli­mate change, and the gov­er­nance of ge­netic en­g­ineer­ing. I dis­cuss other as­pects of the en­vi­ron­men­tal move­ment, an­i­mal ad­vo­cacy, and chil­dren’s ad­vo­cacy more briefly. Dis­cus­sion of im­pli­ca­tions is saved for later. While I fo­cus on fu­ture gen­er­a­tions, I ex­pect these case stud­ies to be highly rele­vant for ad­vo­cates of other voice­less groups, such as non-hu­man an­i­mals.

Case Study Sum­maries:

Case Study Sum­mary—Abo­li­tion­ism:

Since the be­gin­ning of the 19th cen­tury, the most ex­treme forms of slav­ery have se­ri­ously de­clined. The ab­solute num­ber of hu­mans held in chat­tel slav­ery or serf­dom has fallen sig­nifi­cantly, and the pro­por­tion of hu­mans in these con­di­tions has dropped even more dras­ti­cally. The fol­low­ing are ac­counts of the abo­li­tion of the slave trade and slav­ery in the mod­ern era, the only era in his­tory when there has been sig­nifi­cant or­ga­nized op­po­si­tion to these sys­tems.

Mass op­po­si­tion to slav­ery, in differ­ent forms, had rel­a­tively early suc­cess in Haiti and the UK. In Haiti, formerly Saint-Dom­ingue, slaves freed them­selves and abol­ished slav­ery through vi­o­lent rev­olu­tion. In Bri­tain, civili­ans led a mass move­ment against the slave trade and slav­ery. The move­ment’s lead­ers suc­ceeded by mak­ing skil­lful use of unique poli­ti­cal op­por­tu­ni­ties—times when Bri­tons who prof­ited from slav­ery had lit­tle poli­ti­cal abil­ity to stop them. Bri­tish eman­ci­pa­tion ex­tended through­out the Bri­tish Em­pire, in­clud­ing Bri­tish North Amer­ica, Ja­maica, Cape Colony, Aus­tralia, and—af­ter more years of de­lay—In­dia.

After these vic­to­ries, Bri­tish abo­li­tion­ism be­came a ma­jor in­ter­na­tional force for end­ing the slave trade, with sup­port from Bri­tish plan­ta­tion own­ers who wanted to hurt their busi­ness com­peti­tors. The Bri­tish suc­ceeded in pres­sur­ing many other coun­tries into abol­ish­ing their own slave trades.

In France, rev­olu­tion­ar­ies ended slav­ery in an un­suc­cess­ful at­tempt to keep con­trol of their re­bel­ling colony of Saint-Dom­ingue, but Napoleon brought back slav­ery and tried to re-cap­ture Saint-Dom­ingue (it went poorly for his troops). Later rev­olu­tion­ar­ies again banned slav­ery, and the next Napoleon who seized power did not se­ri­ously try to bring it back.

After these early waves of an­ti­s­lav­ery, the un­even de­cline of slav­ery con­tinued. In Cen­tral and Eastern Euro­pean coun­tries—in­clud­ing Rus­sia, Aus­tria, Poland, and Prus­sia—rulers ended serf­dom to avoid its per­ceived con­tri­bu­tions to mil­i­tary weak­ness. In the United States, abo­li­tion­ists ended slav­ery rel­a­tively quickly, but only in colonies where slav­ery was not very prof­itable. Later, an­ti­s­lav­ery ac­tivity in North­ern states made slave-hold­ing South­ern states se­cede, in fear for their in­sti­tu­tion of slav­ery. Civil war fol­lowed, and it ended slav­ery by em­pow­er­ing North­ern abo­li­tion­ists, while cre­at­ing strong mil­i­tary in­cen­tives for them to pur­sue eman­ci­pa­tion.

Other parts of Europe were even slower to abol­ish slav­ery. Spain ended slav­ery in its colonies of Cuba and Puerto Rico be­cause ex­pan­sions in civil liber­ties in­creased the in­fluence of abo­li­tion­ists, while a Cuban war for in­de­pen­dence made the con­tinu­a­tion of slav­ery a mil­i­tary li­a­bil­ity for Spain. The Ot­toman Em­pire ended most of its slave trade be­cause of pres­sure from Western Euro­pean em­pires—es­pe­cially the Bri­tish, and the lead­ers of the Young Turk Revolu­tion then abol­ished slav­ery. By the 1870s, Euro­pean coun­tries had passed eman­ci­pa­tion for their colonies (al­though some colonies did not listen), and the At­lantic slave trade had effec­tively ended.

Colo­nial­ism and world wars then fur­ther global­ized an­ti­s­lav­ery efforts. Euro­peans used the ban­ner of an­ti­s­lav­ery to help jus­tify their con­quests in Africa and Asia, and, fol­low­ing an agree­ment, they grad­u­ally re­duced the slave trade and chat­tel slav­ery in ar­eas they con­quered. Then, post­war in­ter­na­tional in­sti­tu­tions helped co­or­di­nate an­ti­s­lav­ery efforts. Fol­low­ing the Se­cond World War, na­tions de­clared slav­ery a crime against hu­man­ity and a vi­o­la­tion of hu­man rights. Anti-colo­nial states pres­sured re­main­ing Euro­pean coun­tries to agree to end mod­ern slav­ery (slav­ery, more broadly con­ceived) in their colonies. In 1956, na­tions agreed to in­ter­na­tion­ally pro­hibit mod­ern slav­ery. To­day, slav­ery has de­clined, but mil­lions still live in slav­ery.

Case Study Sum­mary—De­moc­ra­ti­za­tion:

Vot­ing rights have ex­panded dra­mat­i­cally over the past few cen­turies, both offi­cially and in prac­tice. Th­ese trends to­ward poli­ti­cal in­clu­sion in­volved two kinds of de­moc­ra­ti­za­tion: the cre­ation of democ­ra­cies, and the re­moval of many re­stric­tions on who could par­ti­ci­pate in them. As a case study, the re­moval of eco­nomic re­stric­tions is es­pe­cially rele­vant to efforts aimed at cre­at­ing last­ing pro­tec­tions for fu­ture gen­er­a­tions.

A wide range of ev­i­dence—the timing of de­moc­ra­ti­za­tion, state­ments by his­tor­i­cal poli­ti­ci­ans, as­sess­ments by his­to­ri­ans, and var­i­ous em­piri­cal stud­ies—sup­ports the claim that, most of the times when democ­racy was cre­ated and ex­tended along eco­nomic lines, this oc­curred be­cause of non-elites cred­ibly threat­en­ing (or some­times ac­tu­ally car­ry­ing out) vi­o­lent rev­olu­tion. Em­piri­cal ev­i­dence is un­clear about whether these con­flicts are driven by con­flict­ing prefer­ences over gov­ern­ment re­dis­tri­bu­tion; other val­ues may be at play.

De­moc­ra­ti­za­tion has mostly oc­curred in in­dus­trial coun­tries be­cause in­dus­tri­al­iza­tion em­pow­ers rev­olu­tion­ary threats. Through sev­eral mechanisms, in­dus­tri­al­iza­tion makes de­moc­ra­ti­za­tion less costly for elites, and it makes con­flict more costly for elites. At the same time, in­dus­tri­al­iza­tion makes it eas­ier for non-elites to co­or­di­nate into forc­ing elites to choose be­tween de­moc­ra­ti­za­tion and con­flict. In the shorter term, na­tional crises of­ten shift power in ways that strengthen rev­olu­tion­ary threats.

While rev­olu­tion­ary threats have ar­guably been the main drivers of de­moc­ra­ti­za­tion, other fac­tors have also been im­por­tant. Changes in val­ues may have been sig­nifi­cant con­trib­u­tors to fran­chise ex­pan­sions, but only if they were driven by the mass ag­i­ta­tion of the dis­en­fran­chised. In ad­di­tion, there are some his­tor­i­cal cases that the rev­olu­tion­ary threat hy­poth­e­sis strug­gles to ex­plain, es­pe­cially early US de­moc­ra­ti­za­tion. In cases like these, elites’ ex­pec­ta­tions that they would re­ceive mil­i­tary, poli­ti­cal, or eco­nomic sup­port from cit­i­zens in ex­change for grant­ing them the vote seem to have been cru­cial con­trib­u­tors to de­moc­ra­ti­za­tion. Other ma­jor waves of de­moc­ra­ti­za­tion—women’s suffrage—fol­lowed the world wars, and there are sev­eral highly plau­si­ble mechanisms by which world wars boosted suffrag­ist move­ments.

Case Study Sum­mary—En­vi­ron­men­tal­ism (Cli­mate Agree­ments Fo­cus):

Modern mass en­vi­ron­men­tal­ism surged in the US in the 1960s, along with sev­eral other move­ments for change, and it achieved early suc­cesses. Sig­nifi­cant vic­to­ries in the early 1970s may have hap­pened be­cause Nixon saw green policy as a way to please re­form­ers with­out agree­ing to more rad­i­cal de­mands for change. Early US reg­u­la­tion in the 1970s fea­tured efforts to limit pol­lu­tants that were di­rectly harm­ful to hu­man health.

Decades of cli­mate diplo­macy have been limited by in­ter­na­tional dis­agree­ments. Coun­tries be­gan to ne­go­ti­ate on cli­mate change in the late 1980s, fol­low­ing a cam­paign by con­cerned sci­en­tists, the Mon­treal Pro­to­col, and an in­tense heat wave. Since then, likely be­cause of differ­ences in economies and elec­toral sys­tems, Europe has pushed for more se­ri­ous cli­mate ac­tion than the US. After es­tab­lish­ing a frame­work for cli­mate agree­ments in the 1992 Earth Sum­mit, na­tions signed the Ky­oto Pro­to­col in 1997. It es­tab­lished spe­cific, bind­ing com­mit­ments for re­duc­ing car­bon emis­sions, but these only ap­plied for de­vel­oped coun­tries, and the US never rat­ified the Pro­to­col. Since then, the US has con­tinued to fail at pass­ing se­ri­ous cli­mate ac­tion leg­is­la­tion. The 2015 Paris Agree­ment was de­signed to by­pass some of the ob­sta­cles that had blocked pre­vi­ous in­ter­na­tional efforts, but its re­sults have been mixed.

De­spite their ma­jor limi­ta­tions, cli­mate ac­tion and en­vi­ron­men­tal­ism more broadly have ac­com­plished much. Cli­mate change miti­ga­tion poli­cies that have already been im­ple­mented are es­ti­mated to re­duce global warm­ing by 2100 by about 1.5℃, rel­a­tive to what would have hap­pened with no cli­mate ac­tion. En­vi­ron­men­tal­ism has also had ma­jor suc­cesses in en­vi­ron­men­tal con­ser­va­tion, and even more suc­cess in the reg­u­la­tion of di­rectly harm­ful pol­lu­tants.

Case Study Sum­mary—Gover­nance of Ge­netic Eng­ineer­ing:

In­ter­na­tional efforts to gov­ern ge­netic en­g­ineer­ing be­gan in the ’70s. In 1972, na­tions signed the Biolog­i­cal Weapons Con­ven­tion, agree­ing to not de­velop biolog­i­cal weapons. A year later, biol­o­gists in­vented ge­netic en­g­ineer­ing tech­niques that could re­sult in the cre­ation of novel, dev­as­tat­ing pathogens. Con­cerned, lead­ing biol­o­gists called to­gether top biol­o­gists wor­ld­wide to a con­fer­ence at Asilo­mar, where they de­vel­oped guidelines for the re­spon­si­ble use of ge­netic en­g­ineer­ing in re­search.

While ge­netic en­g­ineer­ing has not re­sulted in the mass out­break of dis­ease, gov­er­nance has been highly limited in other ways. While ge­netic en­g­ineer­ing tech­nolo­gies have pro­lifer­ated, over­sight con­tinues to cen­ter on the ini­tial guidelines de­vel­oped at Asilo­mar, and reg­u­la­tors mainly en­force limits through fund­ing re­stric­tions. On the in­ter­na­tional stage, diplo­mats have not agreed to equip the Biolog­i­cal Weapons Con­ven­tion with se­ri­ous en­force­ment mechanisms, and the Con­ven­tion did not stop the Soviet Union from de­vel­op­ing a mas­sive pro­gram of ge­net­i­cally en­g­ineered bioweapons.

Case Study Sum­mary—Briefer Case Stud­ies:

An­i­mal ad­vo­cates have had some sig­nifi­cant suc­cesses in ban­ning rel­a­tively small forms of an­i­mal ex­ploita­tion and cru­elty al­to­gether, im­prov­ing the liv­ing con­di­tions of many farmed an­i­mals, and start­ing tech­nolog­i­cal and leg­is­la­tive ini­ti­a­tives that may even­tu­ally re­sult in the end of fac­tory farm­ing. How­ever, an­i­mal ad­vo­cates have not yet suc­ceeded in es­tab­lish­ing se­ri­ous, last­ing poli­ti­cal pro­tec­tions for farmed an­i­mals, whose di­rect ex­ploita­tion is es­pe­cially prof­itable, is prac­ticed on a scale many times larger than hu­mans’ ex­ploita­tion of other hu­mans, and con­tinues to in­crease.

Chil­dren have his­tor­i­cally been sub­jected to many forms of abuse, with lit­tle pro­tec­tion. In the mod­ern era, ac­tion from philan­thropists, gov­ern­ments, and in­ter­na­tional or­ga­ni­za­tions has helped cre­ate some pro­tec­tions for chil­dren. Chil­dren have benefited from ad­vo­cates’ per­cep­tions that pro­tect­ing chil­dren is a promis­ing way to pre­vent so­cial prob­lems, as well as from tech­nolog­i­cal and sci­en­tific ad­vances that have im­proved par­ent-child re­la­tion­ships.


Others af­firm that the rule of a mas­ter over slaves is con­trary to na­ture, and that the dis­tinc­tion be­tween slave and free­man ex­ists by law only, and not by na­ture; and be­ing an in­terfer­ence with na­ture is there­fore un­just.

- Aris­to­tle, Book I of Poli­tics, c. 330 BCE


At the end of the 18th cen­tury, forced la­bor was, as it had been through­out his­tory, the global norm. As one his­to­rian put it, “Free­dom, not slav­ery, was the pe­cu­liar in­sti­tu­tion.” Slav­ery and the slave trade in­volved mass cru­elty and suffer­ing. Two cen­turies later, slav­ery has been out­lawed ev­ery­where (Hochschild, 2005).

Since the be­gin­ning of the 19th cen­tury, the num­ber of hu­mans in con­di­tions of chat­tel slav­ery and serf­dom has de­clined sig­nifi­cantly in ab­solute terms, and even more dras­ti­cally in rel­a­tive terms. The fol­low­ing origi­nal charts, made with es­ti­mates from his­to­rian Sey­mour Drescher (2017) and the In­ter­na­tional Labour Or­ga­ni­za­tion (2017), illus­trate the dra­matic global de­cline, over the past two cen­turies, of the most ex­treme forms of slav­ery.

(Th­ese es­ti­mates have sig­nifi­cant limi­ta­tions: ear­lier data is sparse, the es­ti­mate for 1950 prob­a­bly un­der­states a surge in the use of forced la­bor by au­thor­i­tar­ian gov­ern­ments, and the es­ti­mate for 2016, while it only counts forced la­bor­ers un­der con­di­tions other than debt bondage, still prob­a­bly counts some forms of forced la­bor that were not in­cluded in ear­lier es­ti­mates. Still, the clear broader trends of sig­nifi­cant de­cline are likely ro­bust to these de­grees of er­ror. This spread­sheet, also linked in Ap­pendix A, has more de­tail about the data.)

The demise of slav­ery over the last few cen­turies offers a rele­vant case study for ad­vo­cates of fu­ture gen­er­a­tions. Two groups have faced some similar challenges: both groups have been highly limited in their ca­pac­i­ties for self-ad­vo­cacy (slaves were fre­quently pre­vented from re­ceiv­ing ed­u­ca­tions and form­ing as­so­ci­a­tions, while fu­ture gen­er­a­tions can­not in­fluence their an­ces­tors), and both groups have had wealthy in­ter­ests in­vested in eco­nomic ac­tivi­ties that harm them. The fact that op­po­nents of slav­ery faced these broadly similar ob­sta­cles but had ma­jor suc­cesses any­way sug­gests that ad­vo­cates of fu­ture gen­er­a­tions may have much to learn from the his­tory of abo­li­tion.

The fol­low­ing are com­par­a­tive ac­counts of how the slave trade and slav­ery ended. The Bri­tish Em­pire and the United States are re­viewed most ex­ten­sively. Abo­li­tion and eman­ci­pa­tion in In­dia, France, Haiti, Cen­tral and Eastern Europe, Cuba, Puerto Rico, the Ot­toman Em­pire, and in­ter­na­tional diplo­macy are re­viewed more briefly. The time pe­riod cov­ered is mainly be­tween the 18th cen­tury and the mid-20th cen­tury; be­fore this era, or­ga­nized op­po­si­tion to the slave trade and slav­ery was al­most en­tirely limited to nu­mer­ous slave re­bel­lions that were ul­ti­mately crushed (Drescher, 2017). At least two coun­tries—China and Brazil—are not in­cluded de­spite their his­tor­i­cally large slave pop­u­la­tions, as the au­thor has not had time to find satis­fac­tory his­tor­i­cal in­for­ma­tion on abo­li­tion in these coun­tries.

United King­dom:

(Re­search non­profit Sen­tience In­sti­tute has a very de­tailed nar­ra­tive (An­this, 2017) of the move­ment to abol­ish the slave trade and slav­ery within the Bri­tish Em­pire. The nar­ra­tive be­low may be use­ful to read­ers in­ter­ested in a shorter read, with some­what differ­ent ar­gu­ments about why things hap­pened.)

The emer­gence of or­ga­nized abo­li­tion (Hochschild, 2005):

Large-scale or­ga­ni­za­tion against the slave trade and slav­ery be­gan in Bri­tain. Much of Bri­tain’s wealth (by some es­ti­mates, around 5% of the na­tional in­come) came from its sys­tem of slave trade and slav­ery, through its trade with Bri­tish colonies in the West Indies. Many wealthy Bri­tons were heav­ily in­vested in slav­ery and the slave trade. In the 18th cen­tury, evan­gel­i­cal move­ments called for the re­vival of spiritual life amidst per­cep­tions of im­mense sin (such as gam­bling), in­fluenc­ing some Bri­tons who would later be­come lead­ing abo­li­tion­ists.

Be­fore 1787, there was a lit­tle abo­li­tion­ist ac­tivity in Bri­tain. A 1772 le­gal case with an early abo­li­tion­ist pros­e­cu­tor, the Som­er­set case, was widely in­ter­preted as free­ing all of the few slaves in England, al­though the ac­tual de­ci­sion had been milder. The Quak­ers had been early, staunch op­po­nents of slav­ery, but they were poli­ti­cally and cul­turally marginal­ized as a re­sult of their re­li­gious iden­tity.

Re­ports and in­ci­dents in­volv­ing the cru­elty of the slave trade trick­led into Bri­tain, mo­ti­vat­ing a few Angli­cans to ally with Quak­ers in pur­su­ing abo­li­tion. Rea­son­ably, op­po­nents of slav­ery be­lieved that end­ing the slave trade was much more fea­si­ble than end­ing slav­ery it­self, and that slav­ery in the West Indies would dwin­dle with­out the slave trade (as plan­ters de­pended on the im­por­ta­tion of slaves to make up for the reg­u­lar deaths of slaves from bru­tal work­ing and en­vi­ron­men­tal con­di­tions). This mo­ti­vated abo­li­tion­ist al­lies, in 1787, to found the So­ciety for Effect­ing the Abo­li­tion of the Slave Trade.

The first mass cam­paigns for abo­li­tion (Hochschild, 2005):

Abo­li­tion­ists be­gan a mass cam­paign against the slave trade. Thomas Clark­son, the So­ciety’s only full-time or­ga­nizer, worked de­ter­minedly to found and con­tact lo­cal abo­li­tion­ist chap­ters (which mostly had mid­dle class lead­ers) and to in­ves­ti­gate the slave trade. The So­ciety pub­lished pam­phlets and books, made pop­u­lar pe­ti­tions, and pi­o­neered the use of newslet­ters, mail fundrais­ing, a logo (a kneel­ing man with the text “Am I not a man and a brother?”), and other iconic images (es­pe­cially di­a­grams of slave ships). Buoyed by new pub­li­ca­tions and Manch­ester ac­tivists’ mass pe­ti­tion or­ga­ni­za­tion, pub­lic in­ter­est in the de­bate over the slave trade soared in 1788. The Com­mit­tee also helped start a Sierra Leone colony of freed slaves, but the colony faced se­vere eco­nomic and lead­er­ship prob­lems.

The first abo­li­tion bill was defeated by sup­port­ers of the slave trade. The West In­dian in­ter­est was widely seen as the strongest lobby in Bri­tain at the time, a time when or­ga­nized in­ter­ests other than re­li­gious sects and trade groups were rare. Lob­by­ists en­gaged in ex­ten­sive pro­pa­ganda and man­aged to ini­ti­ate ex­ten­sive hear­ings. By do­ing this, they de­layed the first abo­li­tion bill pro­posed by Wilberforce, an ally of the So­ciety and an in­de­pen­dent MP. Lob­by­ists ar­gued, among other things, that the slave trade was good for Afri­cans and eco­nom­i­cally cru­cial for Bri­tain. Abo­li­tion­ists wrote hun­dreds of books call­ing for Bri­tish abo­li­tion. How­ever, abo­li­tion was hurt by its as­so­ci­a­tion with the re­cent erup­tion of vi­o­lent rev­olu­tion in France. In 1791, Par­li­a­ment voted down the first abo­li­tion bill.

In re­sponse to their 1791 defeat, abo­li­tion­ists re­dou­bled their efforts. Women led a mass sugar boy­cott (Wol­ls­ten­craft’s first fem­i­nist man­i­festo soon fol­lowed). Abo­li­tion­ists made mass pe­ti­tions. Why did a mass abo­li­tion move­ment emerge in Bri­tain, but not in other slave trad­ing pow­ers? The fol­low­ing fac­tors prob­a­bly con­tributed: good or­ga­niz­ing, per­haps in­dus­tri­al­iza­tion, un­usu­ally dense com­mu­ni­ca­tion net­works, Bri­tish val­u­a­tions of their free­doms, the phys­i­cal dis­tance of slaves (mean­ing that few per­ceived abo­li­tion as a threat to their ways of life), and the de­spised, similar prac­tice of mass naval im­press­ment. Dense com­mu­ni­ca­tion net­works were likely boosted by ur­ban­iza­tion and com­mu­ni­ca­tion tech­nolo­gies that had been less pre­sent in ear­lier eras (Pinker, 2011).

De­spite abo­li­tion­ists’ efforts, sup­port for the slave trade had more in­fluence in 1792. The West In­dian in­ter­est ran a me­dia cam­paign, streaked with de­nial­ism, defend­ing the slave trade. Sup­port for the slave trade was strength­ened by the re­cent be­gin­ning of what would come to be known as the Haitian Revolu­tion, a mas­sive slave re­volt in what was then the French colony of Saint-Dom­ingue; the re­volt pro­vided rhetor­i­cal am­mu­ni­tion against abo­li­tion­ists (e.g. crit­ics blamed abo­li­tion­ists for slaves’ ideas of re­bel­lion), and it boosted the prof­its of Bri­tish slave own­ers by hurt­ing their French com­peti­tors. Slave in­ter­ests were in­fluen­tial enough in the House of Lords to stop that year’s abo­li­tion bill.

The bleak decade (Hochschild, 2005):

In the decade that fol­lowed, abo­li­tion­ist ac­tivity and suc­cess was min­i­mal. This was largely be­cause Bri­tain was en­gaged in war with Revolu­tion­ary France. Fear of rad­i­cal French in­fluence led to an in­tense crack­down on civil so­ciety and dis­sent; abo­li­tion­ists, some of whom had been pub­li­cly hon­ored by the French, pru­dently kept low pro­files, while much of the pub­lic shifted their at­ten­tion to the wars. This added to abo­li­tion­ists’ dis­cour­age­ment from their defeat in 1792.

Mean­while, the French Revolu­tion had helped in­spire the rev­olu­tion in Saint-Dom­ingue, the crown jewel of Euro­pean colonies, and it was far more suc­cess­ful than rul­ing France had ex­pected. Seek­ing to main­tain con­trol over the colony, France freed all its em­pire’s slaves in 1794 (al­though Napoleon would later undo this). Bri­tain, hop­ing to gain spoils and quash ideas of re­bel­lion among slaves, sent scores of troops to Saint-Dom­ingue. Un­pre­pared for trop­i­cal dis­ease, they re­treated af­ter Bri­tish armies ex­pe­rienced dev­as­tat­ing losses, and some men re­turned with bleak views of Caribbean slav­ery. Through an­other re­volt, Saint-Dom­ingue rebels then re­pel­led the in­vad­ing forces of Napoleon, and in 1804 they es­tab­lished the in­de­pen­dent Repub­lic of Haiti, which was already af­flicted by eco­nomic dev­as­ta­tion, au­thor­i­tar­i­anism, and eth­nic con­flict.

Abo­li­tion (Hochschild, 2005):

War with France then cre­ated a ma­jor vuln­er­a­bil­ity for the slave trade, and abo­li­tion­ist lawyer James Stephen no­ticed this. Fol­low­ing Napoleon’s rise to power, Bri­tain was still war­ring with France. By then, how­ever, France had been fight­ing to re­store slav­ery in Saint-Dom­ingue, so Bri­tish war pa­tri­o­tism had shifted to fa­vor abo­li­tion­ists. Stephen con­vinced Wilberforce to pro­pose a ban on the ex­por­ta­tion of slaves to for­eign pow­ers, on the jus­tifi­ca­tion that trade with neu­tral pow­ers—which them­selves were trad­ing with France—em­pow­ered Bri­tain’s mil­i­tary en­e­mies.

This was a brilli­ant move. In­ge­niously, the bill split the West In­dian lobby, be­cause it called for seiz­ing en­emy trade ships—an ap­peal­ing prospect for many mer­chants—and it would weaken the com­pe­ti­tion of plan­ta­tions in Bri­tish colonies. Few no­ticed a ma­jor effect the bill would have, which abo­li­tion­ists were care­ful to avoid em­pha­siz­ing: by cut­ting off much slave trade that was owned by Bri­tons but based in for­eign pow­ers’ colonies, the bill would effec­tively end a large por­tion—by some es­ti­mates, two thirds—of the Bri­tish slave trade. Boosted by war pa­tri­o­tism and Clark­son’s pe­ti­tion-gath­er­ing, the bill passed in 1806.

Sev­eral fac­tors as­sisted the ex­pan­sion of this bill into a com­plete abo­li­tion of the Bri­tish slave trade the fol­low­ing year. First, de­pressed sugar prices re­duced plan­ters’ in­ter­est in im­port­ing new slaves (and hence in pro­tect­ing their abil­ity to do so). Se­cond, the new Prime Minister, Grenville, was more com­mit­ted to abo­li­tion and could gather sup­port for it in the re­luc­tant House of Lords. Third, the Haitian Revolu­tion sunk claims that France would dom­i­nate the sugar mar­ket if Bri­tain ended its slave trade. In ad­di­tion, the pre­vi­ous bill had prob­a­bly slashed in­cen­tives to main­tain the slave trade—per­haps it even in­cen­tivized some plan­ta­tion own­ers to sup­port com­plete abo­li­tion, to hurt their com­peti­tors who still benefit­ted from im­port­ing slaves.

In 1807, Bri­tain—the world’s largest slave-trad­ing power—abol­ished its slave trade, and it then be­gan to en­force abo­li­tion in­ter­na­tion­ally. The UK pro­hi­bi­tion in­cluded a ban on the pur­chase of slaves. This in­cen­tivized the West In­dian in­ter­est to lobby for the in­ter­na­tion­al­iza­tion of the pro­hi­bi­tion and its en­force­ment, in or­der to weaken their com­peti­tors. It did so, and the UK en­forced the ban at sea. By 1815, the UK gov­ern­ment suc­ceeded in get­ting the French king to abol­ish the slave trade, al­though this was not en­forced.


An­tis­lav­ery efforts con­tinued af­ter the abo­li­tion of the slave trade. The end of the slave trade, it turned out, was not enough to end slav­ery in the West Indies. Per­haps wealthy Bri­tons iden­ti­fied with slave own­ers more eas­ily than with slave traders, and eman­ci­pa­tion likely seemed more threat­en­ing to their busi­ness in­ter­ests. Some abo­li­tion­ists founded a so­ciety for grad­ual eman­ci­pa­tion, while a younger gen­er­a­tion pushed for im­me­di­ate eman­ci­pa­tion. The lat­ter efforts were led by women, and these efforts in­cluded an­other mass boy­cott on slave-pro­duced sugar, ex­panded to boy­cott any stores that sold it. Still, the abo­li­tion­ists went more years with­out ma­jor vic­to­ries (Hochschild, 2005).

A surge of ac­tivity and a ma­jor slave re­volt in Ja­maica then con­tributed to eman­ci­pa­tion. In the early 1830s, bills for elec­toral re­form in the UK re­ceived mas­sive pop­u­lar sup­port. En­couraged by the move­ment for re­form, more abo­li­tion­ists pushed for im­me­di­ate eman­ci­pa­tion—the num­ber of lo­cal groups soared, with lower and mid­dle classes offer­ing the high­est rates of sup­port. The surge in abo­li­tion­ist ac­tivity may have in­spired Ja­maicans, in the end of 1831, to ini­ti­ate the largest slave re­volt that ever oc­curred in Bri­tish colonies. The re­volt was quickly defeated, but it offered rhetor­i­cal am­mu­ni­tion to abo­li­tion­ists based on the costs of slav­ery and the venge­ful mis­treat­ment of white mis­sion­ar­ies (Hochschild, 2005).

Elec­toral re­form also weak­ened sup­port­ers of slav­ery. Bills for elec­toral re­form passed the House of Com­mons in 1831 and 1832. When the House of Lords at­tempted to block elec­toral re­form, mass ri­ots threat­ened lords with pop­u­lar up­ris­ing, forc­ing them to pass the Re­form Act 1832 (Ace­moglu and Robin­son, 2000). The re­form re­duced the rep­re­sen­ta­tion of wealthy landown­ers, halv­ing the wealthy pro-slave block in par­li­a­ment (Hochschild, 2005). Pop­u­lar ag­i­ta­tion may have also mo­ti­vated elites to pass eman­ci­pa­tion as a way to satisfy pop­u­lar pres­sure with­out meet­ing more rad­i­cal de­mands for do­mes­tic re­form (An­this, 2017). In ad­di­tion, the re­cent mass dis­tur­bances in re­sponse to lords’ op­po­si­tion to pop­u­lar bills likely gave lords good rea­sons to fear the poli­ti­cal con­se­quences of try­ing to block an­other pop­u­lar bill just a year later.

By 1833, with abo­li­tion­ists strength­ened and their op­po­si­tion weak­ened, the UK passed eman­ci­pa­tion through­out most of its em­pire, with caveats. This in­cluded Ja­maica, Cape Colony, Aus­tralia, and Bri­tish North Amer­ica (which be­came the “North star” for Amer­i­can fugi­tive slaves). How­ever, eman­ci­pa­tion came with an as­terix: an effec­tive 4-year de­lay (which would have been years longer if not for an­ti­s­lav­ery pres­sure), mas­sive com­pen­sa­tion for former slave own­ers (equiv­a­lent to about 40% of the UK’s na­tional bud­get—pri­vate prop­erty rights were treated as sacro­sanct in the UK), and ex­cep­tions in South Asia (Hochschild, 2005).

The suc­cess of Bri­tish abo­li­tion­ists was highly con­se­quen­tial for the in­ter­na­tional his­tory of slav­ery, as Bri­tons went on to pro­mote abo­li­tion in other re­gions by spread­ing abo­li­tion­ist liter­a­ture, pro­vid­ing or­ga­ni­za­tional sup­port to other abo­li­tion­ist or­ga­ni­za­tions, as­so­ci­at­ing free la­bor forces with mil­i­tary power (merely by hav­ing both), ex­tend­ing abo­li­tion to Bri­tish im­pe­rial pos­ses­sions, di­rectly pres­sur­ing other gov­ern­ments to abol­ish their slave trades, and en­forc­ing bans on the slave trade (Drescher, 2017).


By the mid-19th cen­tury, the Bri­tish Em­pire con­trol­led much of In­dia through the Bri­tish East In­dia Com­pany. When the UK abol­ished slav­ery in 1833, it ini­tially ex­empted these ter­ri­to­ries. An­tis­lav­ery ac­tivity con­tinued in the UK, now with fewer op­po­nents. In 1843, the UK dele­gal­ized slav­ery in In­dia. This meant that courts would no longer en­force or oth­er­wise as­sist slav­ery; slaves could leave the own­ers they had been bound to, but there were no agen­cies pub­lish­ing this fact. This amounted to a non-dis­rup­tive and “slow death for slav­ery”—grad­ual abo­li­tion—which Bri­tish tax­pay­ers saw as an ap­peal­ing al­ter­na­tive to an­other ex­pen­sive act of com­pen­sa­tion for slave own­ers (Drescher, 2017).

France and Haiti:

First eman­ci­pa­tion and its par­tial re­ver­sal (Hochschild, 2005):

Soon af­ter the French Revolu­tion, slaves in France’s most pro­duc­tive slave colony or­ga­nized a mas­sive re­volt. By the late 18th cen­tury, France’s colo­nial pos­ses­sions in­cluded Caribbean is­lands that were heav­ily in­vested in sug­ar­cane plan­ta­tions worked by Afri­can slaves. In 1789, the French Revolu­tion over­threw France’s aris­to­cratic monar­chy, but rev­olu­tion­ar­ies were not quick to ex­tend their prin­ci­ples of liberty to their colo­nial slaves. Two years later, a mas­sive slave re­volt erupted in Saint-Dom­ingue (later named Haiti), the ex­tremely pro­duc­tive “crown jewel” of Euro­pean colonies.

The Saint-Dom­ingue slave re­bel­lion was highly suc­cess­ful. Sev­eral fac­tors likely con­tributed to the re­bel­lion’s un­usual strength: ru­mors of the French Revolu­tion helped in­spire and co­or­di­nate rebel ac­tion, peo­ple of mixed de­scent (as a class, wealthy sec­ond-class cit­i­zens who rarely joined slave re­bel­lions) sup­ported the rebel slaves, and many of the slaves had lived and gained mil­i­tary ex­pe­rience in Africa be­fore they were en­slaved.

Re­bel­lion brought about an ini­tial end to slav­ery in Saint-Dom­ingue. In 1794, af­ter sev­eral years of fight­ing be­tween Saint-Dom­ingue rebels and French sol­diers, France eman­ci­pated all its slaves. In do­ing this, France was fol­low­ing the lead of a se­nior offi­cial who be­lieved that eman­ci­pa­tion was France’s best shot at main­tain­ing con­trol over Saint-Dom­ingue. Fol­low­ing France’s eman­ci­pa­tion, mul­ti­ple colo­nial pow­ers and rebel armies con­tinued to fight for con­trol of Saint-Dom­ingue. Rebel leader Tous­saint Lou­ver­ture was vic­to­ri­ous, and he es­tab­lished a mil­i­tary dic­ta­tor­ship in which slav­ery was abol­ished.

Napoleon then tried to bring back slav­ery, but his suc­cess was limited. In 1799, Napoleon seized power in France, and he sent troops to in­vade Saint-Dom­ingue. French forces were tem­porar­ily suc­cess­ful, and Napoleon re­in­stated slav­ery in France’s colonies in 1802. Slaves again re­volted. To­gether with trop­i­cal dis­ease, slave rebels im­posed mas­sive ca­su­alties on French troops, forc­ing them to re­treat from Saint-Dom­ingue. In 1804, rebels es­tab­lished the in­de­pen­dent Repub­lic of Haiti, with slav­ery abol­ished. The Haitian Revolu­tion had been the most suc­cess­ful slave re­volt in his­tory, but Haiti was already be­set by eco­nomic dev­as­ta­tion, au­thor­i­tar­i­anism, and eth­nic con­flict.

Se­cond eman­ci­pa­tion and its per­sis­tence:

Or­ga­nized sup­port for abo­li­tion had never thrived in France, and it re­mained min­i­mal in the decades af­ter Napoleon’s power grab. Do­mes­tic abo­li­tion­ism had dim prospects in France; au­thor­i­tar­ian gov­ern­ment long re­pressed civil so­ciety, and the dev­as­tat­ing Haitian Revolu­tion as well as Bri­tish abo­li­tion­ist pres­sure soured French peo­ple’s at­ti­tudes to­ward abo­li­tion, while as­so­ci­at­ing pa­tri­o­tism with the defense of slav­ery (Drescher, 2015).

A sec­ond eman­ci­pa­tion only came through an­other rev­olu­tion. In 1848, rad­i­cals again seized power in France, and they es­tab­lished the Se­cond French Repub­lic. Lead­ers sym­pa­thetic to the an­ti­s­lav­ery cause soon eman­ci­pated France’s slaves. Fear­ing de­lay, they chose to pass eman­ci­pa­tion be­fore the Na­tional Assem­bly first met, as they feared that leg­is­la­tors would de­lay eman­ci­pa­tion un­til they had se­cured com­pen­sa­tion for slave own­ers (Drescher, 2015).

In 1851, re­pub­li­can France again fell to au­thor­i­tar­i­anism. Un­der the rule of Napoleon III, France im­posed 14-year pe­ri­ods of in­den­tured servi­tude on former slaves. Still, France did not try to last­ingly re­store slav­ery, per­haps be­cause the new em­peror had learned from Napoleon I’s catas­trophic at­tempt to do so (Drescher, 2017).

Cen­tral and Eastern Europe:

Tsar Alexan­der II, who held highly cen­tral­ized power, eman­ci­pated Rus­sia’s serfs in 1861. A ma­jor con­tribut­ing fac­tor to this was the Crimean War of the mid-1850s; Rus­sia’s defeat at the hands of the then-free UK and France con­vinced Rus­sia’s au­to­crat that serf­dom caused un­ac­cept­able weak­ness on the in­ter­na­tional stage (Drescher, 2015).

Events de­vel­oped similarly in other parts of Eastern Europe. Over­all, in Eastern Europe, serf­dom was a much more com­mon form of com­pul­sory la­bor and poli­ti­cal dis­em­pow­er­ment than chat­tel slav­ery. Ac­cord­ing to his­to­rian Sey­mour Drescher, Euro­pean elimi­na­tions of serf­dom fol­lowed a com­mon pat­tern: “Most Euro­pean serf eman­ci­pa­tions,” no­tably Aus­tria, Poland, and Prus­sia, in ad­di­tion to Rus­sia, “were made in an­ti­ci­pa­tion of or re­sponse to mil­i­tary threat or defeat” (2017).

United States:

Early pro­hi­bi­tion in Ge­or­gia:

All of the thir­teen colonies of Bri­tain that would later form the United States al­lowed slav­ery from their found­ing, ex­cept for Ge­or­gia. Ge­or­gia was ini­tially un­der the in­fluence of Bri­tish trustees who pro­hibited slav­ery in the colony. Their mo­ti­va­tions for do­ing this in­cluded want­ing Ge­or­gia to be a colony of small farm­ers that pro­vided op­por­tu­ni­ties to Euro­pean debtors (which would not hap­pen if Ge­or­gia was filled with large-scale slave plan­ta­tions), and be­cause they had se­cu­rity con­cerns (nearby Span­ish Florida offered free­dom to slave defec­tors).

The trustees’ char­ter ex­pired, and Ge­or­gians, whose neigh­bors had made the prof­ita­bil­ity of slav­ery in­creas­ingly ap­par­ent, quickly cre­ated an econ­omy with min­i­mal re­stric­tions for slave own­ers. This came af­ter a brief pe­riod of more limited slav­ery, which might have been a failed at­tempt at dam­age con­trol by the trustees—they knew their power was limited in time and de­gree (Ge­or­gians had started ille­gally im­port­ing slaves, and their ap­peal to se­cu­rity from the Spa­niards had lost rele­vance af­ter the Span­ish were mil­i­tar­ily defeated), so they tried to es­tab­lish a bet­ter al­ter­na­tive to full-fledged slav­ery that colon­ists would ac­cept (Hinks and McKivi­gan, 2007).

Sec­tional di­ver­gence:

Amer­i­can colonies quickly di­verged in the de­gree of their re­li­ance on slav­ery. En­vi­ron­men­tal con­di­tions in South­ern colonies made the pro­duc­tion of cash crops through slave la­bor highly prof­itable, and South­ern economies quickly grew heav­ily in­vested in slave plan­ta­tions. In con­trast, slav­ery was never as prof­itable in New England colonies; en­vi­ron­men­tal con­di­tions did not fa­vor cash crop pro­duc­tion, wheat pro­duc­tion mainly re­quired sea­sonal la­bor, there was a higher sup­ply of cheap non-slave la­bor (less trop­i­cal dis­ease), and us­ing slaves for other eco­nomic ac­tivi­ties (e.g. fur trap­ping, tim­ber pro­duc­tion, and ship build­ing) may have posed higher train­ing, su­per­vi­sion, and se­cu­rity costs (“New England,” 2020; Fenoaltea, 1984).

After the Amer­i­can Revolu­tion­ary War and the wave of liberal ideas it brought, sec­tional differ­ences in at­ti­tudes to­ward slav­ery and its ex­pan­sion emerged. Between 1780 and 1804, North­ern states—be­gin­ning with Quaker-led Penn­syl­va­nia—all abol­ished slav­ery. In the 1780s, ev­ery state that had a Black pop­u­la­tion lower than 6% passed eman­ci­pa­tion laws, while ev­ery state that did not, did not (Be­fore 1790, the US did not col­lect cen­sus data on slav­ery; es­ti­mates of Black pop­u­la­tions are use­ful prox­ies for slave pop­u­la­tions, al­though they are over-es­ti­mates) (“Colo­nial and Pre-Fed­eral Statis­tics,” 1975). Between 1790 and 1804, New York and New Jersey—the two states that, out of slave states, had the low­est rates of slav­ery in those years—passed eman­ci­pa­tion laws (“Re­ca­pitu­la­tion of the Tables,” 1864).

The US abol­ished the im­por­ta­tion of slaves in 1807, but South­ern plan­ta­tions were not highly de­pen­dent on im­port­ing new slaves (slave pop­u­la­tions in­creased nat­u­rally, as South­ern plan­ta­tions were not in the highly fatal busi­ness of pro­duc­ing sug­ar­cane), and there would not be a wide­spread anti-slav­ery move­ment in the US for decades (Hochschild, 2005).

Or­ga­nized sup­port for slav­ery limited abo­li­tion­ist ac­tivity and suc­cess for decades be­fore slave states se­ceded. Abo­li­tion­ist ac­tivity in North­ern states was damp­ened by fears of South­ern se­ces­sion, which were height­ened by the sec­tional na­ture of slav­ery’s di­vi­sion and the US tra­di­tion of fed­er­al­ism. De­spite the North­ern op­po­si­tion to the ex­pan­sion of slav­ery that did ex­ist, South­ern states’ in­fluence main­tained a bal­ance of slave and free states in Congress for decades (in or­der to main­tain fed­eral rep­re­sen­ta­tion of their in­ter­est in slav­ery), while go­ing far to sup­press dis­course (Drescher, 2015). Within the South, up un­til the Amer­i­can Civil War and its af­ter­math im­posed eman­ci­pa­tion on South­ern states, white peo­ple’s at­tempts at or­ga­niz­ing for eman­ci­pa­tion were re­pressed through means that in­cluded vi­o­lence and cen­sor­ship (Fin­nie, 1969).

In the mid-19th cen­tury, west­ward ex­pan­sion re­vived sec­tional dis­agree­ments over the ex­pan­sion of slav­ery, cul­mi­nat­ing in South­ern se­ces­sion. Abo­li­tion­ists, boosted by or­ga­niz­ers who learned from Bri­tish pre­de­ces­sors and were in­spired by Amer­ica’s Se­cond Great Awak­en­ing, gained the sup­port of la­bor­ers who did not want the com­pe­ti­tion of slave plan­ta­tions in new states. Fear­ing abo­li­tion, and cling­ing to the United States’ tra­di­tion of fed­er­al­ism, South­ern states se­ceded from the Union in 1861 (Drescher, 2015).

Un­suc­cess­ful con­ces­sion:

Union states tried mak­ing con­ces­sions to se­ced­ing states, but these were un­suc­cess­ful. Congress, then con­trol­led by the Repub­li­can Party (which had been founded seven years ear­lier to op­pose the ex­pan­sion of slav­ery), voted with a su­per­ma­jor­ity to pro­pose the Cor­win Amend­ment, a con­sti­tu­tional amend­ment that would have per­ma­nently pro­tected South­ern slav­ery from fed­eral in­ter­ven­tion. The Repub­li­can pres­i­dent Lin­coln ex­pressed that he did not op­pose the amend­ment, and he sent it to state leg­is­la­tures for rat­ifi­ca­tion. This was not enough for the South­ern Con­fed­er­acy, which was by then com­mit­ted to in­de­pen­dence, and civil war be­gan a month af­ter Congress pro­posed the amend­ment (be­fore the nec­es­sary num­ber of states rat­ified the amend­ment) (Lup­ton, 2006).


Repub­li­cans’ sup­port for the Cor­win Amend­ment sup­ports the claim that North­ern­ers were not will­ing to go to war end South­ern slav­ery; they wanted war to re­unite the Union, and the war changed North­ern­ers’ in­cen­tives, poli­ti­cal op­por­tu­ni­ties, and per­haps val­ues in ways that fa­vored eman­ci­pa­tion.

First, war em­pow­ered and in­cen­tivized abo­li­tion­ists to pass tem­po­rary mea­sures of eman­ci­pa­tion, re­duc­ing con­ser­va­tives’ abil­ity and in­cen­tives to block per­ma­nent eman­ci­pa­tion. In 1863, Lin­coln passed the Eman­ci­pa­tion Procla­ma­tion, declar­ing the free­dom of all slaves held in re­bel­ling re­gions. In do­ing this, he ex­plained that the source of his power and mo­ti­va­tion was war: his Procla­ma­tion was le­gi­t­i­mate “by virtue of the power” of the pres­i­dent to en­act “a fit and nec­es­sary war mea­sure for sup­press­ing… re­bel­lion,” and it was made for that very pur­pose (Lin­coln, 1863). The Eman­ci­pa­tion Procla­ma­tion was mil­i­tar­ily ad­van­ta­geous to the North be­cause it turned Con­fed­er­ate slave la­bor­ers into Union sol­diers. In ad­di­tion, eman­ci­pa­tion made the war a war over slav­ery, sink­ing the emerg­ing will in cot­ton-hun­gry yet an­ti­s­lav­ery Bri­tain to sup­port the Con­fed­er­acy (“Effect of the Eman­ci­pa­tion”).

On top of this, the Civil War made it clear that, with­out the per­ma­nent eman­ci­pa­tion of the 13th Amend­ment, con­flict was likely to reemerge in the fu­ture. One Rep­re­sen­ta­tive who changed his vote to sup­port the 13th Amend­ment de­clared (“Record of the US House, 1865):

noth­ing short of the recog­ni­tion of their in­de­pen­dence will satisfy the south­ern con­fed­er­acy. It must there­fore be de­stroyed; and in vot­ing for the pre­sent mea­sure I cast my vote against the cor­ner-stone of the south­ern con­fed­er­acy, and de­clare eter­nal war against the en­e­mies of my country

While, pre­vi­ously, it had seemed that pur­su­ing eman­ci­pa­tion would have the cost of risk­ing civil war, by 1865 that cost had been paid, and per­ma­nent eman­ci­pa­tion seemed nec­es­sary for pre­vent­ing fu­ture war.

The Civil War did not only strengthen North­ern in­cen­tives to end slav­ery; it also con­cen­trated power in the hands of North­ern abo­li­tion­ists. Lin­coln’s war pow­ers, as noted above, em­pow­ered him to make the Eman­ci­pa­tion Procla­ma­tion, and Union vic­tory in 1865 al­lowed Union lead­ers to force South­ern states to rat­ify the 13th Amend­ment (“Land­mark Leg­is­la­tion”). In ad­di­tion, mil­i­tary suc­cesses and war pa­tri­o­tism (the pub­lic’s de­sire to ex­press unity, and per­haps de­sire for vengeance) con­tributed to Repub­li­cans hav­ing the elec­toral suc­cess in 1864 to pass per­ma­nent eman­ci­pa­tion. Repub­li­cans had pre­vi­ously failed to pass the amend­ment in Congress, and they cam­paigned on it in 1864, declar­ing in their party plat­form that, “as Slav­ery was the cause, and now con­sti­tutes the strength, of this re­bel­lion… jus­tice and the na­tional safety de­mand… a death-blow at this gi­gan­tic evil” (“Rat­ify­ing the 13th Amend­ment, 2012; “The Plat­forms,” 1864).

Per­ma­nent eman­ci­pa­tion may have also benefited from the ex­plicit con­sid­er­a­tion of fu­ture gen­er­a­tions. When Lin­coln learned that the 13th Amend­ment needed two ad­di­tional votes to pass, he re­port­edly af­firmed (Kaller):

The abo­li­tion of slav­ery by Con­sti­tu­tional pro­vi­sions set­tles the fate, for all … time, not only of the mil­lions now in bondage, but of un­born mil­lions to come – a mea­sure of such im­por­tance that those two votes must be procured

An 1862 Bri­tish news­pa­per also called on Amer­i­cans to end slav­ery, in terms that (other than their re­li­gious em­pha­sis) sound like they be­long in a Nick Bostrom book (“Effect of the Eman­ci­pa­tion”):

Will you not em­brace it? So much good has not been done by one effort as in the Prov­i­dence of God it is now your high priv­ilege to do. May the vast fu­ture not have to lament that you have ne­glected it!

The United States passed a per­ma­nent pro­hi­bi­tion on slav­ery in 1865, just four years af­ter leg­is­la­tors had al­most per­ma­nently pro­tected slav­ery.

Cuba and Puerto Rico:

By the mid-19th cen­tury, slav­ery re­mained com­mon in Span­ish Cuba and Puerto Rico, and slave own­ers were es­pe­cially in­fluen­tial in Cuba. Cuba’s elite was highly de­pen­dent on sugar plan­ta­tions worked by slaves. While the UK had man­aged to pres­sure Spain into agree­ing to end the trade of slaves to its colonies, these bans were not effec­tively en­forced. The Span­ish re­pressed Cuba’s emerg­ing abo­li­tion­ism, partly be­cause of its as­so­ci­a­tions with “liber­al­ism, anti-colo­nial­ism, and the ever-med­dling Bri­tish” (Ber­gad, 2017).

In con­trast, abo­li­tion­ist ac­tivity emerged in Puerto Rico. This was largely pos­si­ble be­cause, un­like Cuba, Puerto Rico had a large free la­bor force (a cheap sup­ply of la­bor), and it was shift­ing to coffee pro­duc­tion by non-slaves. Abo­li­tion­ist con­cerns that emerged fo­cused on how slav­ery cre­ated risks of slave rev­olu­tion, and how the im­por­ta­tion of slaves cre­ated de­mo­graphic shifts that racists dis­liked (Ber­gad, 2017).

Mili­tary de­vel­op­ments in Cuba then con­tributed to eman­ci­pa­tion. In 1868, Cubans be­gan their first re­volt for in­de­pen­dence from Spain. Rebels seized a part of Cuba that had fewer slaves, and Cubans from re­gions that re­lied more heav­ily on slave la­bor did not sup­port the rebels. In 1869, rebels granted slav­ery to any slaves who fled to them, and the next year they abol­ished slav­ery in rebel-held ter­ri­tory. This meant that the con­tinu­a­tion of slav­ery would weaken re­sis­tance to Cuban in­de­pen­dence (pro-Span­ish landown­ers would pre­sum­ably be bet­ter off if they had to pay their la­bor­ers a lit­tle than if their la­bor­ers fled) (Ber­gad, 2017).

Then, mil­i­tary de­vel­op­ments in Spain added to the pres­sure from Cuban rebels, fi­nally bring­ing about eman­ci­pa­tion. In 1868, Spain un­der­went a liberal rev­olu­tion. This rev­olu­tion ended Span­ish re­pres­sion of the new and im­pact­ful Span­ish Abo­li­tion­ist So­ciety, which Puerto Ri­cans had helped found. Un­der abo­li­tion­ist pres­sure from within and mil­i­tary pres­sure from Cuban rebels, Spain passed grad­ual eman­ci­pa­tion for Cuba and Puerto Rico in 1870. Pres­sure from mul­ti­ple sources (in­clud­ing in­ter­na­tional pres­sure) then es­ca­lated, lead­ing the Span­ish gov­ern­ment to make eman­ci­pa­tion come sooner, in 1886 (Ber­gad, 2017).

The Ot­toman Em­pire:

The en­slave­ment of peo­ple from Africa and Cir­cas­sia was en­trenched in the 19th cen­tury Ot­toman Em­pire, with neg­ligible do­mes­tic op­po­si­tion. Ot­toman prac­tices sur­round­ing slav­ery differed sig­nifi­cantly from ear­lier Euro­pean prac­tices: it was cus­tom­ary for slave own­ers to free their slaves af­ter seven to ten years of la­bor, and slaves were more of­ten used for house­hold la­bor. Do­mes­tic op­po­si­tion to slav­ery was al­most non-ex­is­tent in the Ot­toman Em­pire, as well as in other Is­lamic so­cieties, per­haps be­cause au­thor­i­tar­ian rule re­pressed civil so­ciety (Fer­gu­son and Toledano, 2016).

In­ter­na­tional pres­sure and do­mes­tic rev­olu­tion then brought about the de­cline of Ot­toman slav­ery. By the mid-19th cen­tury, the Ot­tomans were highly eco­nom­i­cally de­pen­dent on Euro­pean pow­ers, and Western Euro­pean em­pires—es­pe­cially the Bri­tish—pres­sured the Ot­tomans to end their slave trade. In 1857, as a re­sult of this pres­sure, the Ot­toman sul­tan banned the slave trade and im­ple­mented par­tial en­force­ment. Pres­sure in­creased in the years lead­ing up to the Bri­tish-ini­ti­ated 1889 Brus­sels Anti-Slav­ery Con­fer­ence, com­pel­ling the sul­tan to strengthen en­force­ment of the ban. With in­creased en­force­ment, very lit­tle of the slave trade con­tinued. In 1908, the liberal Young Turk Revolu­tion over­threw the Ot­toman Em­pire, and the Young Turks abol­ished slav­ery (Fer­gu­son and Toledano, 2016).

In­ter­na­tional Abo­li­tion­ism:

Bri­tish pres­sure and the end of the transat­lantic slave trade:

When the UK banned its own slave trade, it be­gan a long his­tory of pres­sur­ing other coun­tries to ban their own slave trades, likely be­cause of strong do­mes­tic sup­port for these diplo­matic efforts. After Bri­tish plan­ta­tion own­ers lost their own abil­ity to im­port slaves, they sup­ported the in­ter­na­tion­al­iza­tion of the ban on im­port­ing slaves, so that their busi­ness com­peti­tors would not have that ad­van­tage. Ad­ding to busi­ness pres­sure, Bri­tish abo­li­tion­ists had built a large and in­fluen­tial net­work of or­ga­niz­ers who could and did mo­bi­lize the Bri­tish pub­lic against the slave trade and slav­ery. This left abo­li­tion­ists with as­sets in ex­pe­rience, or­ga­ni­za­tion, net­works, and pub­lic sup­port, which they then turned to­ward in­ter­na­tional abo­li­tion. In ad­di­tion to ex­ert­ing in­ter­na­tional pres­sure through UK poli­tics, Bri­tish abo­li­tion­ists helped fund and oth­er­wise sup­port other coun­tries’ an­ti­s­lav­ery so­cieties (Drescher, 2017; Hochschild, 2005).

Bri­tish pres­sure for in­ter­na­tional abo­li­tion achieved sig­nifi­cant suc­cesses in the ne­go­ti­a­tions fol­low­ing the Napoleonic Wars. Around 1815, the UK co­erced sev­eral other Euro­pean coun­tries to end their colo­nial slave trades, and they made the in­ter­na­tional Congress of Vienna de­clare that

the slave trade has been con­sid­ered by just and en­light­ened men in all ages, as re­pug­nant to the prin­ci­ples of hu­man­ity and uni­ver­sal moral­ity.

The UK then drew on this sup­posed con­sen­sus to pres­sure more coun­tries into bilat­eral agree­ments for ban­ning their slave trades, and for en­forc­ing those bans with pa­trol ships. Var­i­ous other North­ern Euro­pean coun­tries soon eman­ci­pated their own slaves.

Abo­li­tion­ist suc­cesses in other Euro­pean states came later. Spain and Por­tu­gal con­tinued to trade slaves for longer, and the US, be­fore its own eman­ci­pa­tion of slaves, pre­vented block­ades against Cuba’s slave trade. Mean­while, the French Em­pire had grown more liberal, al­low­ing op­po­nents of slav­ery to hold a large in­ter­na­tional an­ti­s­lav­ery meet­ing in Paris in 1867. Through con­tinued efforts, by the 1860s and ’70s, Euro­pean coun­tries had passed eman­ci­pa­tion for their colonies (al­though some colonies, e.g. Cuba, did not im­ple­ment eman­ci­pa­tion), and the At­lantic slave trade had effec­tively ended (Drescher, 2017).

The de­cline of the slave trade and slav­ery in Euro­pean em­pires:

An­tis­lav­ery ac­tivity then shifted to the Eastern Hemi­sphere, largely through colo­nial­ism. Euro­peans used the ban­ner of an­ti­s­lav­ery to help jus­tify their colo­nial in­cur­sions. Some coun­tries, such as Ja­pan, sup­ported the abo­li­tion of the slave trade so they would be more per­ceived by Euro­peans as civ­i­lized. Abo­li­tion­ist ac­tivity re­surged af­ter the 1885 Ber­lin Con­fer­ence. The Bri­tish ini­ti­ated the or­ga­ni­za­tion of the Brus­sels Anti-Slav­ery Con­fer­ence, which in 1890 signed an agree­ment to end the slave trade (while leav­ing slav­ery it­self). Euro­peans, per­haps be­cause they did not want oth­ers to kid­nap la­bor­ers whom they could ex­ploit, effec­tively re­duced the slave trade as their colo­nial con­trol ex­panded (Drescher, 2017).

End­ing chat­tel slav­ery it­self in parts of Africa and Asia con­trol­led by Euro­peans was of­ten harder for op­po­nents of slav­ery. After all, Euro­pean colo­nial­ists of­ten de­pended on the co­op­er­a­tion of lo­cal slave-own­ing elites, and com­pen­sa­tion would again be ex­pen­sive. In ad­di­tion, ral­ly­ing pub­lic op­po­si­tion was harder, as many Old World slaves had less bru­tal roles than Caribbean plan­ta­tion slaves, and Euro­peans tended to have a lower sense of re­spon­si­bil­ity for slav­ery they had not started.

Still, chat­tel slav­ery de­clined in new Euro­pean colonies. As a par­tial solu­tion to the difficul­ties of eman­ci­pat­ing slaves in Africa and Asia, Euro­peans who con­quered much of Africa in the late 19th cen­tury tended to imi­tate the way the UK abol­ished slav­ery in its In­dian pos­ses­sions: grad­ual eman­ci­pa­tion through the dele­gal­iza­tion of slav­ery. The Bri­tish Anti-Slav­ery So­ciety also suc­cess­fully de­manded ac­tion against par­tic­u­larly ex­treme forms of ex­ploita­tion by weaker colo­nial pow­ers, such as Por­tu­gal’s An­gola and Leopold’s Congo (Drescher, 2017).

World wars and the global pro­hi­bi­tion of mod­ern slav­ery:

The First World War cre­ated new op­por­tu­ni­ties for in­ter­na­tional an­ti­s­lav­ery. Fol­low­ing the war, states cre­ated the League of Na­tions. The League, led by Euro­peans, fa­cil­i­tated the in­ter­na­tional co­or­di­na­tion of abo­li­tion. Work­ing within the League, the UK led a com­mis­sion that re­sulted in the Slav­ery Con­ven­tion of 1926. This con­ven­tion defined slav­ery as own­er­ship of a per­son, and sign­ing na­tions agreed to sup­press the slave trade and end slav­ery. Euro­peans tended to see slav­ery as an in­sti­tu­tion in demise.

Slav­ery then de­clined fur­ther through in­ter­na­tional efforts af­ter the Se­cond World War. The war in­volved a tem­po­rary surge of forced la­bor, es­pe­cially in Nazi con­cen­tra­tion camps. After the Allied vic­tory, Nazi lead­ers were charged, among other things, with slav­ery, as a crime against hu­man­ity. The 1948 Univer­sal Dec­la­ra­tion of Hu­man Rights de­clared var­i­ous forms of mod­ern slav­ery (a broader no­tion of slav­ery, in­clud­ing not only chat­tel slav­ery, but also serf­dom, debt bondage, and forced mar­riage) as vi­o­la­tions of hu­man rights. By 1956, most of the states in the United Na­tions had banned slav­ery. Anti-colo­nial pow­ers—no­tably, the Soviet Union—de­manded an end to mod­ern slav­ery in Euro­pean colonies, per­haps to re­duce their Euro­pean ri­vals’ rev­enue streams. The Sup­ple­men­tary Con­ven­tion on the Abo­li­tion of Slav­ery of that year ex­panded the ear­lier Con­ven­tion, in­ter­na­tion­ally pro­hibit­ing mod­ern slav­ery (Drescher, 2017).



Vot­ing rights have ex­panded dra­mat­i­cally over the past few cen­turies. The fol­low­ing chart shows changes in offi­cial vot­ing rights over time in a set of coun­tries made up of the ten coun­tries that to­day have the largest pop­u­la­tions, and the ten coun­tries that to­day have the largest economies (to­gether, to­day these coun­tries make up more than half of the world’s pop­u­la­tion) (Roser, 2019).

Fur­ther data sug­gests that this trend to­ward ex­panded vot­ing rights has been ac­com­panied by a gen­uine ex­pan­sion of poli­ti­cal power, al­though the lat­ter trend is weaker. The fol­low­ing chart shows re­gional changes over time in the Van­hanen’s In­dex of Democ­racy, which serves as a proxy for equal­ity in de facto poli­ti­cal power by be­ing pro­por­tional to both voter par­ti­ci­pa­tion and the strength of op­po­si­tion par­ties (Roser, 2019).

Th­ese trends to­ward poli­ti­cal in­clu­sion in­volved both the cre­ation of democ­ra­cies and the re­moval of var­i­ous re­stric­tions on who could par­ti­ci­pate in them. In many coun­tries, the re­moval of eco­nomic and sex-based suffrage re­stric­tions en­fran­chised the ma­jor­ity of adults, while the re­moval of re­li­gious, racial, and eth­nic re­stric­tions on vot­ing rights en­fran­chised sig­nifi­cant minori­ties.

Un­der­stand­ing elites’ re­movals of eco­nomic re­stric­tions on the fran­chise seems es­pe­cially rele­vant for figur­ing out how fu­ture gen­er­a­tions can gain poli­ti­cal pro­tec­tion. In both cases, pow­er­ful ac­tors have had strong fi­nan­cial in­cen­tives to main­tain ex­clu­sive power (the wealthy tend to not want vot­ing rights to be ex­tended to peo­ple who would vote for gov­ern­ment re­dis­tri­bu­tion, and pre­sent gen­er­a­tions have in­cen­tives to not share in­ter­gen­er­a­tional goods with fu­ture gen­er­a­tions). Be­cause of this similar­ity, I fo­cus on the re­moval of eco­nomic re­stric­tions on vot­ing rights. Why did these changes hap­pen?

Three broad ex­pla­na­tions for de­moc­ra­ti­za­tion are best sup­ported by his­tor­i­cal ev­i­dence. Most com­monly, elites ex­tended the fran­chise to miti­gate threats of rev­olu­tion. In some cases, they may have been gen­uinely per­suaded by the mass ag­i­ta­tion of the dis­en­fran­chised. In the early United States, and per­haps in other cases, elites de­moc­ra­tized to gain eco­nomic, mil­i­tary, or poli­ti­cal sup­port from the newly en­fran­chised. One highly plau­si­ble way to think about how these many mo­ti­va­tions for de­moc­ra­ti­za­tion in­ter­acted is that there were many in­sti­tu­tional “hur­dles” that stood in the way of de­moc­ra­ti­za­tion, and var­i­ous mo­ti­va­tions were sig­nifi­cant to differ­ent de­grees for over­com­ing differ­ent hur­dles (Aidt and Franck, 2019).

Threats of Revolu­tion and Global De­moc­ra­ti­za­tion:

A wide range of ev­i­dence sup­ports the claim that, most of the in­stances when democ­racy was cre­ated and ex­tended along eco­nomic lines, this was caused by non-elites cred­ibly threat­en­ing or ac­tu­ally car­ry­ing out vi­o­lent rev­olu­tion.

In the UK, Swe­den, France, and Ger­many, de­moc­ra­ti­za­tion usu­ally fol­lowed mass dis­tur­bances, sug­gest­ing it was driven by fear of rev­olu­tion. All three ma­jor cases of eco­nomic fran­chise ex­pan­sion in the UK (in 1832, 1867, and 1884), as well as all three ma­jor cases of fran­chise ex­pan­sion in Swe­den (1866, 1909, and 1918), were pre­ceded by mass ag­i­ta­tion. In France, eco­nomic fran­chise ex­pan­sion fol­lowed rev­olu­tion in 1789 and 1848 (al­though in both cases the fran­chise was soon con­tracted, and gen­uine re-ex­pan­sion only came af­ter France’s mil­i­tary defeat in 1870, for rea­sons less clearly linked to the threat of rev­olu­tion). In Ger­many, fran­chise ex­pan­sion fol­lowed rev­olu­tion in 1848, and (af­ter ex­pan­sion in 1870 that did not go far in ex­tend­ing de facto vot­ing pow­ers) the fran­chise was fur­ther ex­tended in 1919, af­ter mass ag­i­ta­tion that bor­dered on rev­olu­tion fol­lowed Ger­man mil­i­tary defeats (Ace­moglu and Robin­son, 2000).

The above ex­am­ples are highly typ­i­cal of de­moc­ra­ti­za­tion through­out mod­ern global his­tory. Re­view­ing his­tor­i­cal re­search, poli­ti­cal economists Ace­moglu and Robin­son (2005) con­clude—with sig­nifi­cant sup­port from his­to­ri­ans—that, in na­tion af­ter na­tion, de­moc­ra­ti­za­tion has tended to fol­low mass dis­tur­bances:

As illus­trated by the Bri­tish, Ar­gen­ti­nian, and South Afri­can poli­ti­cal his­to­ries… most tran­si­tions to democ­racy, both in nine­teenth- and twen­tieth-cen­tury Europe and twen­tieth-cen­tury Latin Amer­ica, took place amid sig­nifi­cant so­cial tur­moil and rev­olu­tion­ary threats. In ad­di­tion, the cre­ation of demo­cratic so­cieties in most former Euro­pean colonies in the 1950s and 1960s was the re­sult of pres­sure by the dis­en­fran­chised and rel­a­tively poor colo­nials against the coloniz­ing power. Such threats of tur­moil and so­cial di­s­or­der similarly ac­com­panied the re­cent spate of de­moc­ra­ti­za­tion in Africa (Brat­ton and van der Walle 1997) and Eastern Europe (Bunce 2003).

The timing of many cases of de­moc­ra­ti­za­tion, then, sug­gest that they were driven by non-elites cre­at­ing threats of rev­olu­tion.

Ad­di­tional analy­ses provide fur­ther ev­i­dence for the rev­olu­tion­ary threat hy­poth­e­sis—the view that elites per­ceiv­ing threats of rev­olu­tion was a ma­jor cause of de­moc­ra­ti­za­tion. In some cases, as quoted by Ace­moglu and Robin­son (2000), elites ex­plic­itly stated that this was their mo­ti­va­tion:

When in­tro­duc­ing the elec­toral re­form to the Bri­tish par­li­a­ment in 1831, the prime minister Earl Grey said, ‘There is no-one more de­cided against an­nual par­li­a­ments, uni­ver­sal suffrage and the bal­lot, than am I . . . The Prin­ci­pal of my re­form is to pre­vent the ne­ces­sity of rev­olu­tion. . . . I am re­form­ing to pre­serve, not to over­throw’ (quoted in [Evans 1983]).

On top of such state­ments, more sup­port comes from so­cial sci­en­tists’ use of var­i­ous met­rics as prox­ies for elites’ per­cep­tions of rev­olu­tion­ary threat. Th­ese have in­cluded bond yields in the UK (these are higher when peo­ple are less cer­tain that the gov­ern­ment will ex­ist in the fu­ture) as well as records of ri­ots near UK elec­toral dis­tricts, droughts in Sub-Sa­haran Africa (which tend to pre­cede ri­ots), and the oc­cur­rence of rev­olu­tions in neigh­bor­ing coun­tries. As re­viewed by Aidt, et al. (2015), as well as Aidt and Franck (2019), these stud­ies and oth­ers gen­er­ally provide sig­nifi­cant sup­port for the rev­olu­tion­ary threat hy­poth­e­sis.

Mo­ti­va­tions for Con­flict Over the Vot­ing Fran­chise:

An ap­peal­ing ex­pla­na­tion for why non-elites threaten elites with rev­olu­tion is that they want last­ing, ma­te­rial benefits. Ace­moglu and Robin­son (2005) ar­gue that non-elites at­tempt to change poli­ti­cal in­sti­tu­tions in or­der to turn short-lived power, such as a tem­po­rary boost in co­or­di­nat­ing ca­pac­ity from an eco­nomic crisis, into more last­ing power: poli­ti­cal rep­re­sen­ta­tion. Non-elites want durable power, the ar­gu­ment goes, to make gov­ern­ments pass re­dis­tribu­tive policy that ma­te­ri­ally benefits them, both in the near and long term. At the same time, elites’ ex­pec­ta­tion that de­moc­ra­ti­za­tion will bring re­dis­tri­bu­tion in­cen­tivizes them to op­pose de­moc­ra­ti­za­tion, es­pe­cially in highly un­equal so­cieties. This view is sup­ported by a con­sis­tent trend, al­though some other hy­pothe­ses also pre­dict this trend: through­out the Western Hemi­sphere, greater in­equal­ity did in fact cor­re­spond with slower de­moc­ra­ti­za­tion (Enger­man and Sokoloff, 2005).

How­ever, em­piri­cal ev­i­dence is mixed about whether de­moc­ra­ti­za­tion ac­tu­ally causes in­creased re­dis­tri­bu­tion (Hor­pedahl, 2011). Why, then, do non-elites sup­port these rev­olu­tion­ary threats, and why do elites put up re­sis­tance?

Sev­eral ex­pla­na­tions are highly plau­si­ble. One pos­si­bil­ity is that elites and non-elites have had wrong ex­pec­ta­tions about the ma­te­rial con­se­quences of poli­ti­cal change. This would not be unique in poli­ti­cal his­tory; peas­ants who sup­ported rev­olu­tions that were hi­jacked by dic­ta­tors and brought about famines pre­sum­ably had in­cor­rect ex­pec­ta­tions. A sec­ond pos­si­bil­ity is that peo­ple value vot­ing rights, not for ma­te­rial benefits, but for rea­sons such as so­cial sta­tus and civil liber­ties (while the elites need to be in­cen­tivized be­cause they face a real or illu­sory risk of re­dis­tri­bu­tion, or be­cause they hold ex­clu­sive val­ues) (Hor­pedahl, 2011). A third pos­si­bil­ity is that de­moc­ra­ti­za­tion does cause re­dis­tri­bu­tion, but mea­sur­ing this is hard, so em­piri­cal stud­ies have not yet yielded con­clu­sive re­sults. The am­bi­guity of the ev­i­dence might re­sult from a mix­ture of these con­trib­u­tors. This lack of clar­ity sug­gests that the rev­olu­tion­ary threats which con­tribute to de­moc­ra­ti­za­tion are still far from well un­der­stood.

De­moc­ra­ti­za­tion, In­dus­tri­al­iza­tion, and Na­tional Crises:

Another link is em­piri­cally clearer: de­moc­ra­ti­za­tion has mostly oc­curred in the mod­ern era, es­pe­cially in in­dus­tri­al­ized coun­tries. Why?

One rea­son why in­dus­tri­al­iza­tion would fa­vor democ­racy is that in­dus­trial elites have less to fear from democ­racy than agri­cul­tural elites. In­dus­trial elites can more eas­ily avoid taxes (per­haps be­cause their cap­i­tal is eas­ier to hide), so they have less to fear from de­moc­ra­ti­za­tion. This effect is even stronger with ex­ten­sive global­iza­tion, which is to a large ex­tent driven by in­dus­tri­al­iza­tion’s con­tri­bu­tions to com­mu­ni­ca­tion and trans­porta­tion tech­nolo­gies. In ad­di­tion, in­dus­trial elites tend to be less in­vested in eco­nomic in­sti­tu­tions that la­bor­ers strongly dis­fa­vor (e.g. slav­ery, which is more prof­itable for agri­cul­ture), so again they have less to fear from de­moc­ra­ti­za­tion (Ace­moglu and Robin­son, 2005).

In­dus­tri­al­iza­tion also fa­vors de­moc­ra­ti­za­tion by cre­at­ing elites who are more vuln­er­a­ble to con­flict. Rel­a­tive to landown­ers, in­dus­trial elites have more vuln­er­a­ble cap­i­tal (hu­man and phys­i­cal cap­i­tal is more eas­ily de­stroyed than land), so it is more costly for them to re­spond to rev­olu­tion­ary threats with vi­o­lent re­pres­sion (Ace­moglu and Robin­son, 2005). In­dus­tri­al­iza­tion also in­creases eco­nomic spe­cial­iza­tion and in­ter­de­pen­dence, mak­ing it eas­ier for one eco­nomic in­ter­est group to threaten oth­ers by with­hold­ing its ser­vices. This makes in­dus­trial elites even more vuln­er­a­ble to an­gry masses. In­dus­tri­al­iza­tion, then, changes elites’ in­cen­tives in ways that fa­vor de­moc­ra­ti­za­tion (Con­gle­ton, 2004).

On top of all this, in­dus­tri­al­iza­tion con­tributes to de­moc­ra­ti­za­tion by im­prov­ing non-elites’ abil­ity to or­ga­nize (Con­gle­ton, 2004). In­dus­tri­al­iza­tion in­creases pop­u­la­tion den­si­ties while im­prov­ing com­mu­ni­ca­tion and trans­porta­tion tech­nolo­gies, mak­ing it less costly for non-elites to or­ga­nize. In ad­di­tion, in­dus­tri­al­iza­tion raises in­comes, in­creas­ing the pub­lic’s will­ing­ness to spend to pres­sure the gov­ern­ment into pro­vid­ing pub­lic goods. Per­haps higher in­comes also tend to em­power cit­i­zens to fight for civil liber­ties, which later fa­cil­i­tate re­formist or rev­olu­tion­ary or­ga­ni­za­tion.

Ed­u­ca­tion might seem like an­other his­tor­i­cally cru­cial con­trib­u­tor to pro-democ­racy or­ga­niz­ers, but his­tor­i­cal timing sug­gests oth­er­wise. Some so­cial sci­en­tists (Drazen, et al., 2007) sug­gest that in­dus­tri­al­iza­tion’s con­tri­bu­tions to ed­u­ca­tion (e.g. through ris­ing in­comes, lower trans­porta­tion costs, and in­dus­tri­al­ists’ de­mands for bet­ter trained work­ers) are cru­cial for non-elites’ abil­ities to or­ga­nize effec­tively. How­ever, ex­pan­sions in pub­lic ed­u­ca­tion fre­quently came af­ter de­moc­ra­ti­za­tion, so it is not clear how they could have been its cause (Ace­moglu and Robin­son, 2000; Enger­man and Sokoloff, 2005). Still, broader trends of ris­ing liter­acy, per­haps re­sult­ing from the cre­ation of tech­nolo­gies like the print­ing press, likely con­tributed to some de­gree to non-elites’ abil­ities to co­or­di­nate their re­sis­tance (Pinker, 2011).

In brief, in­dus­tri­al­iza­tion fa­vors de­moc­ra­ti­za­tion be­cause it makes de­moc­ra­ti­za­tion less costly for elites, and it makes con­flict more costly for them, while mak­ing it eas­ier for non-elites to co­or­di­nate into forc­ing elites to choose be­tween de­moc­ra­ti­za­tion and con­flict.

While in­dus­tri­al­iza­tion’s effects help ex­plain long-term trends in the in­fluence of rev­olu­tion­ary threats, na­tional crises go far in ex­plain­ing their shorter-term vari­a­tions. His­tor­i­cally, de­moc­ra­ti­za­tion fre­quently fol­lowed war (Ace­moglu and Robin­son, 2000). As sig­nifi­cant ex­am­ples, mil­i­tary defeats and eco­nomic crises in­crease dis­satis­fac­tion with ex­ist­ing in­sti­tu­tions, pro­vid­ing fo­cal points of co­or­di­na­tion for sup­port­ers of de­moc­ra­ti­za­tion, while re­duc­ing elites’ will­ing­ness and abil­ity to deal with rev­olu­tion­ary threats through vi­o­lent re­pres­sion (eco­nomic dev­as­ta­tion leaves elites with less to lose from re­dis­tribu­tive poli­cies, and it leaves gov­ern­ments with fewer re­sources for re­pres­sion).

The Limited Role of Ide­olog­i­cal Change:

Hav­ing con­sid­ered the rev­olu­tion­ary threat hy­poth­e­sis, we turn to con­sid­er­ing other hy­pothe­ses for why de­moc­ra­ti­za­tion has oc­curred.

One va­ri­ety of al­ter­na­tives cen­ters on ide­olog­i­cal change, and claims (du­bi­ously) that states of­ten de­moc­ra­tized be­cause elites’ val­ues changed in fa­vor of in­clu­sion. The timing of de­moc­ra­ti­za­tion, as dis­cussed pre­vi­ously, is strong ev­i­dence against hy­pothe­ses that do not at­tribute a ma­jor role to the dis­en­fran­chised in bring­ing about ex­pan­sions of democ­racy. If elites had be­come en­light­ened on their own, one would not ex­pect the timing of de­moc­ra­ti­za­tion to so of­ten co­in­cide with non-elites pos­ing threats (Ace­moglu and Robin­son, 2000).

Another view re­mains plau­si­ble: non-vi­o­lent forms of mass ag­i­ta­tion help per­suade elites about the need for re­form. This view has some sup­port from the ap­par­ent in­fluence of non-vi­o­lent demon­stra­tions on sup­port for de­moc­ra­ti­za­tion in the UK in 1831, as well as later events of de­coloniza­tion and civil rights in the US (Aidt and Franck, 2019). Ide­olog­i­cal change driven by mo­bi­liza­tion of the dis­en­fran­chised may have been par­tic­u­larly im­por­tant in the US civil rights move­ment. Since en­force­ment of ma­jor leg­is­la­tion to pro­tect Afri­can Amer­i­can vot­ing rights fo­cused on the South, poli­ti­cal ac­tors of other states may have had weaker in­cen­tives to main­tain poli­ti­cal ex­clu­sion, so they would have been more re­cep­tive to ide­olog­i­cal change. Over­all, the timing of de­moc­ra­ti­za­tion sug­gests that changes in val­ues may have been sig­nifi­cant con­trib­u­tors to fran­chise ex­pan­sions, but only if they were driven by the mass ag­i­ta­tion of the dis­en­fran­chised.

Early US De­moc­ra­ti­za­tion and its Strate­gic Alli­ances:

The above hy­pothe­ses, which both in­volve mass mo­bi­liza­tion among the dis­en­fran­chised, do poorly at ex­plain­ing early de­moc­ra­ti­za­tion in the US. Over the early 19th cen­tury, many US states re­moved eco­nomic re­stric­tions on who could vote, but, as his­to­rian Don­ald Rat­fliffe (2013) writes, such changes usu­ally “did not re­sult from wide­spread pop­u­lar de­mands for a wider suffrage.” Among these, US fron­tier states de­moc­ra­tized es­pe­cially quickly: no state that joined the Union af­ter the thir­teen colonies had prop­erty re­quire­ments for vot­ing, and the few eco­nomic re­quire­ments some did have (e.g. tax-based re­quire­ments) were minor and short-lived. In con­trast, the origi­nal thir­teen colonies were gen­er­ally much slower to re­move ex­plic­itly eco­nomic fran­chise re­stric­tions (Enger­man and Sokoloff, 2005). The ab­sence of pop­u­lar de­mands for early US de­moc­ra­ti­za­tion sug­gests that it had mo­ti­va­tions other than rev­olu­tion­ary threats, mo­ti­va­tions that were es­pe­cially strong in fron­tier states.

A third va­ri­ety of hy­pothe­ses about elites’ mo­ti­va­tions help ex­plain early US de­moc­ra­ti­za­tion, es­pe­cially in fron­tier states. Th­ese hy­pothe­ses as­sert that elites give some peo­ple the right to vote be­cause elites ex­pect these peo­ple to re­spond to new vot­ing rights in ways that poli­ti­cally, mil­i­tar­ily, or eco­nom­i­cally benefit elites. One hy­poth­e­sis of this kind offers a well-sup­ported ex­pla­na­tion for the rapid de­moc­ra­ti­za­tion of US fron­tier states: elites in fron­tier states faced la­bor short­ages, and they had to com­pete with other fron­tier states for set­tlers from older states. As a re­sult, elites in fron­tier states passed many poli­cies to at­tract set­tlers, in­clud­ing the ap­peal­ing offer of vot­ing rights for poorer set­tlers (as long as they were white men). In other words, elites in fron­tier states ex­tended the fran­chise be­cause they ex­pected to benefit from the re­ac­tion of the newly en­fran­chised: im­mi­gra­tion and eco­nomic ser­vices (Enger­man and Sokoloff, 2005).

Similar mo­ti­va­tions help ex­plain de­moc­ra­ti­za­tion in the origi­nal thir­teen colonies that formed the US, which were gen­er­ally much slower to re­move ex­plic­itly eco­nomic fran­chise re­stric­tions. The timing of fran­chise ex­ten­sions (e.g. soon af­ter the War of 1812) and the sur­round­ing de­bate sug­gest that these elites ex­pected to gain a differ­ent benefit in ex­change for de­moc­ra­ti­za­tion: more sup­port­ive mil­i­tary ser­vice, and per­haps more will­ing tax-pay­ment, from the newly en­fran­chised (Hor­pedahl, 2011; Keyssar, 2000). Per­haps poli­ti­cal op­por­tunism—elites ex­tend­ing the vote be­cause they ex­pect the newly en­fran­chised to vote for them—had some role in US and UK de­moc­ra­ti­za­tion (al­though var­i­ous as­pects of UK de­moc­ra­ti­za­tion sug­gest that this mo­ti­va­tion was minor in the UK, at most) (Ace­moglu and Robin­son, 2000; Aidt and Franck, 2019; Keyssar, 2000).

Per­haps these varied mo­ti­va­tions were in­fluen­tial for early US de­moc­ra­ti­za­tion, but less in­fluen­tial for other cases of de­moc­ra­ti­za­tion, be­cause in the early US there were rel­a­tively high lev­els of eco­nomic equal­ity among white men (Enger­man and Sokoloff, 2005). This may have caused US elites to not per­ceive as much eco­nomic risk in ex­tend­ing the fran­chise to other white men.

In short, elites’ ex­pec­ta­tions that they would re­ceive mil­i­tary, poli­ti­cal, or eco­nomic sup­port from cit­i­zens in ex­change for grant­ing them the vote help ex­plain cases that the rev­olu­tion­ary threat hy­poth­e­sis strug­gles to ex­plain, es­pe­cially early US de­moc­ra­ti­za­tion.

Women’s Suffrage:

Mov­ing on from ex­plic­itly eco­nomic suffrage, the charts at the be­gin­ning of this sec­tion show that ma­jor waves of de­moc­ra­ti­za­tion—women’s suffrage—fol­lowed the two world wars. Why? Efforts for suffrage definitely did not be­gin af­ter the wars; the US, for ex­am­ple, had had an ac­tive women’s suffrage move­ment for gen­er­a­tions, and it had achieved sig­nifi­cant suc­cess in Western states be­fore WWI (Hig­gins, 2019). As with wealth-based re­stric­tions, sex-based re­stric­tions on vot­ing may have been more vuln­er­a­ble in fron­tier states be­cause of these states’ greater eco­nomic equal­ity. The differ­ences in eco­nomic con­sump­tion be­tween men and women may have been smaller than the differ­ences in con­sump­tion be­tween wealthy and poor men; this may have al­lowed for shifts to­ward in­clu­sive so­cial val­ues to be ma­jor con­trib­u­tors to women’s suffrage, in­stead of be­ing over­rid­den by eco­nomic self-in­ter­est.

There are sev­eral mechanisms, with sup­port from ar­gu­ments that suffrag­ists used, by which it is highly plau­si­ble that the world wars bolstered women’s suffrage (Hig­gins, 2019; “Women’s Suffrage,” 2018). First, world wars ex­panded the roles of women, un­der­min­ing sex­ist be­liefs that sup­ported fran­chise re­stric­tions. Se­cond, world wars mo­ti­vated Allied coun­tries to por­tray them­selves as bea­cons of democ­racy, fur­ther un­der­min­ing the idea of ban­ning half of the pop­u­la­tion from vot­ing. Third, the dev­as­ta­tion of the world wars in­creased men’s per­cep­tions of the value of hav­ing an elec­torate re­luc­tant to go to war, and men ex­pected the elec­torate to be more paci­fis­tic if it in­cluded women. Fourth, war bolstered a fair­ness ar­gu­ment, and per­haps the real and per­ceived risk of los­ing women’s sup­port for fu­ture war efforts, as women had borne mas­sive costs from wars they did not vote for. The suc­cess of women’s suffrage in some coun­tries also prob­a­bly ac­cel­er­ated suc­cess in other coun­tries, as con­tem­po­ra­ne­ous vic­to­ries es­tab­lished the time in which women’s ad­vo­cates were liv­ing as a fo­cal point of co­or­di­na­tion for in­ten­sified efforts.

En­vi­ron­men­tal­ism (Cli­mate Agree­ments Fo­cus):


En­vi­ron­men­tal­ism and es­pe­cially cli­mate diplo­macy pose highly rele­vant case stud­ies for peo­ple in­ter­ested in bring­ing about vic­to­ries for fu­ture gen­er­a­tions. After all, en­vi­ron­men­tal ad­vo­cates have achieved sig­nifi­cant suc­cesses for fu­ture gen­er­a­tions, as well as other en­tities that have no di­rect poli­ti­cal power: ecosys­tems. Th­ese suc­cesses have in­cluded ma­jor cli­mate change miti­ga­tion efforts, as well as en­vi­ron­men­tal con­ser­va­tion. With this mo­ti­va­tion, the fol­low­ing is an ac­count of in­ter­na­tional cli­mate change poli­tics. It pays ad­di­tional at­ten­tion to de­vel­op­ments in the US as well as other en­vi­ron­men­tal out­comes.

Emer­gence of Modern En­vi­ron­men­tal­ism in the US:

The mod­ern en­vi­ron­men­tal move­ment arose most promi­nently in the United States in the 1960s. At least four fac­tors were likely ma­jor con­trib­u­tors to the surge in con­cern and ad­vo­cacy for the en­vi­ron­ment. First, it had been over a decade since the be­gin­ning of the US’ post­war eco­nomic boom. This pro­vided Amer­i­cans the eco­nomic com­fort for con­sid­er­ing seem­ingly dis­tant prob­lems, with­out newfound grat­i­tude for the sys­tems that pro­vided this com­fort (Dryzek, et al., 2002). Se­cond, in­ter­na­tional de­vel­op­ments spurred con­tem­po­ra­ne­ous US move­ments for civil rights, fem­i­nism, and peace, in­creas­ing ac­tivists’ ex­pe­rience in mo­bi­liza­tion and open­ness to­ward crit­i­cisms of cur­rent sys­tems. Third, Rachel Car­son pub­lished Silent Spring in 1962, rais­ing pub­lic aware­ness of the harms of hu­mans’ un­reg­u­lated treat­ment of the en­vi­ron­ment (Dunn, 2012). And fourth, in the US poli­ti­cal sys­tem, in­ter­est groups had (as they still have) much ac­cess to in­fluen­tial poli­ti­ci­ans and vot­ers, heav­ily in­cen­tiviz­ing so­cial move­ments to or­ga­nize as in­ter­est groups (Dryzek, et al., 2002).

Around 1970, Nixon passed leg­is­la­tion that con­sti­tuted ma­jor vic­to­ries for the US en­vi­ron­men­tal move­ment, in­clud­ing the cre­ation of the En­vi­ron­men­tal Pro­tec­tion Agency and the re­quire­ment that gov­ern­ment agen­cies eval­u­ate poli­cies’ en­vi­ron­men­tal im­pacts—poli­cies that were in­ter­na­tion­ally imi­tated (“Home”). Mass mo­bi­liza­tion likely con­tributed to these suc­cesses; ear­lier in 1970, 20 mil­lion Amer­i­cans par­ti­ci­pated in demon­stra­tions for the first Earth Day (Combs, et al., 2020). A plau­si­ble ex­pla­na­tion for Nixon’s en­thu­si­asm for en­vi­ron­men­tal leg­is­la­tion is that he saw it as a way to weaken move­ments call­ing for sys­temic change, as act­ing for the en­vi­ron­ment would be ac­quiesc­ing to rel­a­tively mod­er­ate de­mands and mak­ing en­vi­ron­men­tal­ists more op­ti­mistic about ex­ist­ing sys­tems (Dryzek, et al., 2002). Fur­ther US reg­u­la­tion in the 1970s fea­tured efforts to limit pol­lu­tants that were di­rectly harm­ful to hu­man health (Combs, et al., 2020).

Emer­gence of Cli­mate Change as a Global Poli­ti­cal Is­sue:

Over the ’80s, sci­en­tists brought in­ter­na­tional at­ten­tion to cli­mate change. Con­tinued re­search on ear­lier dis­cov­er­ies and im­proved com­put­ing ca­pac­i­ties cre­ated a strength­en­ing sci­en­tific con­sen­sus on the ex­is­tence and dan­gers of hu­man-caused cli­mate change. A small group of these con­cerned sci­en­tists, boosted by spon­sor­ship from sym­pa­thetic gov­ern­men­tal and in­ter­na­tional or­ga­ni­za­tions, used work­shops, con­fer­ences, ar­ti­cle pub­li­ca­tions, and their con­tacts among poli­cy­mak­ers to raise the promi­nence of cli­mate change as a risk that needed to be ad­dressed, and to gen­er­ate po­ten­tial poli­cies (Bo­dan­sky, 2001).

Two de­vel­op­ments in the late ’80s ac­cel­er­ated cli­mate ac­tion. First, the dis­cov­ery of the dan­gers of ozone de­ple­tion, and the sub­se­quent ne­go­ti­a­tions cul­mi­nat­ing in the promis­ing Mon­treal Pro­to­col of 1987, brought in­creased at­ten­tion to the pos­si­bil­ity and im­por­tance of harm­ful emis­sions to the at­mo­sphere. Se­cond, North Amer­ica ex­pe­rienced a mas­sive heat wave and drought in 1988, bring­ing in­creased at­ten­tion to the costs of global warm­ing. Fol­low­ing these events, gov­ern­ments be­gan ne­go­ti­a­tions for in­ter­na­tional ac­tion on cli­mate change (Bo­dan­sky, 2001).

Early Cli­mate Ne­go­ti­a­tions and Differ­ences:

Early ne­go­ti­a­tions re­vealed differ­ences in coun­tries’ de­sires that would prove con­tentious for decades to come. Europe pushed for strict limi­ta­tions, while the US in­sisted on flex­i­bil­ity (Bo­dan­sky, 2001). Why these differ­ences? One po­ten­tial ex­pla­na­tion is that fos­sil fuel pro­duc­tion or con­sump­tion made up a much larger por­tion of the US econ­omy than Euro­pean economies, strength­en­ing US eco­nomic and poli­ti­cal in­cen­tives for lax cli­mate ac­tion. The ev­i­dence for this ex­pla­na­tion is limited. Around 1992, when na­tions met at the Earth Sum­mit, the US did pro­duce sig­nifi­cantly more oil than the EU, but lev­els of coal pro­duc­tion were similar. The US and the EU also had similar to­tal lev­els of con­sump­tion for these two fos­sil fuels, as well as similar GDPs (Ritchie and Roser). Th­ese similar­i­ties sug­gest that differ­ences in cli­mate poli­tics had other ma­jor causes.

Although US and EU lev­els of coal pro­duc­tion were similar, the in­dus­try con­text of this was not. The US had large coal re­serves that promised fu­ture prof­its, while coun­tries such as Ger­many were pay­ing the costs of coal sub­sidies (Bo­dan­sky, 2001). In ad­di­tion, the abil­ity of in­dus­try to turn wealth into poli­ti­cal strength likely differed be­tween the re­gions. US leg­is­la­tors still pay much at­ten­tion to busi­ness’ po­ten­tial to sway vot­ers through at­tack ads, while Euro­pean coun­tries tend to limit the in­fluence of wealthy in­ter­ests through strict re­stric­tions on cam­paign ad­ver­tise­ment (Lizza, 2010; Atwill, 2009).

Another likely con­trib­u­tor to Europe’s stronger en­vi­ron­men­tal­ism is elec­toral sys­tems. The pro­por­tional rep­re­sen­ta­tion of much of Europe, in con­trast to the plu­ral­ity vot­ing sys­tem of the US, means that en­vi­ron­men­tal­ists dis­satis­fied with ma­jor par­ties are in­cen­tivized to vote for al­ter­na­tives (“Elec­toral Sys­tems”). This in­cen­tivizes ma­jor par­ties to in­cor­po­rate en­vi­ron­men­tal­ism into their plat­forms in or­der to avoid los­ing ma­jor con­stituen­cies, as hap­pened in Ger­many (Dryzek, et al., 2002). Likely re­flect­ing and re­in­forc­ing these differ­ences, the US shifted con­trol of cli­mate poli­cies from for­eign and en­vi­ron­men­tal ministries to do­mes­tic ones—where in­dus­try in­ter­ests are bet­ter rep­re­sented—much ear­lier than other de­vel­oped coun­tries did (Bo­dan­sky, 2001).

Another ma­jor di­vi­sion that showed it­self early was that be­tween de­vel­oped coun­tries and de­vel­op­ing coun­tries. While de­vel­oped coun­tries had cre­ated im­mense wealth in large part through un­sus­tain­able de­vel­op­ment, the emis­sions of de­vel­op­ing coun­tries were ris­ing rapidly. Each group of coun­tries tended to use one of these ar­gu­ments to ar­gue that the other group should bear the costs of cli­mate change miti­ga­tion (Bo­dan­sky, 2001).

The FCCC and the Ky­oto Pro­to­col:

Fol­low­ing sev­eral years of ne­go­ti­a­tions, na­tions met at the Earth Sum­mit in 1992 and signed the Frame­work Con­ven­tion on Cli­mate Change. As na­tions had wanted con­sen­sus, re­luc­tant na­tions had much in­fluence. The Sum­mit mainly cre­ated a frame­work for fu­ture ne­go­ti­a­tions; it did not cre­ate legally bind­ing com­mit­ments (Bo­dan­sky, 2001).

Then, coun­tries be­gan ne­go­ti­a­tions to­ward more con­crete agree­ments. A 1996 dec­la­ra­tion in­di­cated that coun­tries were will­ing to act with­out con­sen­sus, re­duc­ing the in­fluence of ma­jor oil pro­duc­ers (Bo­dan­sky, 2001).

Cul­mi­nat­ing con­tinued ne­go­ti­a­tions, the 1997 Ky­oto Pro­to­col es­tab­lished legally bind­ing, spe­cific com­mit­ments to emis­sion re­duc­tion. Th­ese had ma­jor limi­ta­tions: de­vel­op­ing coun­tries were not bound to emis­sions, and the US never rat­ified the agree­ment.

Months be­fore Ky­oto, the US Se­nate had passed a unan­i­mous re­s­olu­tion, declar­ing they would not rat­ify any treaty that se­ri­ously harmed the US econ­omy or did not man­date com­mit­ments for de­vel­op­ing coun­tries (Hovi, et al., 2012). Why, then, did na­tions sign an agree­ment strictly man­dat­ing emis­sions re­duc­tions from de­vel­oped coun­tries only?

In­ter­views with peo­ple who were del­e­gates at the time sug­gest sev­eral rea­sons—cen­tered on do­mes­tic poli­ti­cal pres­sures—for why del­e­gates passed a treaty the US would not rat­ify. It was not for lack of want­ing US par­ti­ci­pa­tion—Euro­pean del­e­gates made sig­nifi­cant con­ces­sions in­tended to win US in­volve­ment. A fac­tor that prob­a­bly did con­tribute is that do­mes­tic poli­ti­cal pres­sure pushed both Euro­pean and Clin­ton-Gore del­e­gates to pur­sue an am­bi­tious agree­ment, even if this meant lower chances that it would be passed. Per­haps this was be­cause it is eas­ier to blame oth­ers for a treaty that oth­ers re­fused to rat­ify, than to blame oth­ers for one’s own cre­ation of an un­am­bi­tious treaty. The pur­suit of an un­re­al­is­ti­cally am­bi­tious treaty by the Clin­ton-Gore ad­minis­tra­tion may have fur­ther con­tributed to the Euro­peans (who like other del­e­ga­tions fo­cused mainly on treaty tech­ni­cal­ities) over­es­ti­mat­ing the chances of US rat­ifi­ca­tion (Hovi, et al., 2012).

Ad­di­tion­ally, the dis­mal re­sult of the treaty may have been less ob­vi­ous than it seems for US del­e­gates. They seem to have seen an­other path to rat­ifi­ca­tion: win­ning over com­mit­ments from de­vel­op­ing coun­tries as side agree­ments. How­ever, US ne­go­tia­tors did not suc­ceed in cre­at­ing side agree­ments; ei­ther they over­es­ti­mated the ease of do­ing so, or they took a well-calcu­lated risk that went poorly (Hovi, et al., 2012).

In the end, Clin­ton did not even sub­mit the treaty to the Se­nate for rat­ifi­ca­tion, leav­ing Euro­pean coun­tries as the main par­ties to the treaty (Hovi, et al., 2012).

US Do­mes­tic Cli­mate Failure, 2010:

In 2010, US leg­is­la­tors raised promis­ing cli­mate change leg­is­la­tion, but it failed to pass. This was af­ter Obama was elected pres­i­dent of the United States with con­gres­sional su­per­ma­jori­ties, fol­low­ing a cam­paign in which he em­pha­sized pri­ori­tiz­ing cli­mate. Pow­er­ful op­po­si­tion likely con­tributed to the bill’s failure in sev­eral ways. First, US leg­is­la­tive pro­ce­dures (es­pe­cially the Se­nate fili­buster) cre­ated many veto op­por­tu­ni­ties. Se­cond, Repub­li­can par­ti­san­ship was on the rise, and it in­volved in­tense ob­struc­tion­ism. Third, US cam­paign laws and poli­ti­cal norms gave (as they still give) im­mense in­fluence to wealthy in­ter­ests—or­ga­nized op­po­si­tion and the threat of it diminished poli­ti­cal sup­port, while in­cen­tiviz­ing bill spon­sors to make ma­jor con­ces­sions.

Re­cent events and poor co­or­di­na­tion also helped the bill’s op­po­nents defeat it. The US had only re­cently be­gun re­cov­ery from the Great Re­ces­sion, so the eco­nomic risks of cli­mate ac­tion were es­pe­cially salient. In ad­di­tion, an un­for­tu­nately timed event—the Deep­wa­ter Hori­zon oil spill—made the bill’s con­ces­sions to offshore drilling have highly salient costs. On top of this, Obama and other lead­ing Democrats failed to co­or­di­nate well with spon­sors of the bill (per­haps be­cause they had de­pri­ori­tized cli­mate ac­tion); their un­strate­gic ac­tions took away lev­er­age from spon­sors of the bill, while their lack of sup­port con­tributed to driv­ing out Gra­ham, a key Repub­li­can col­lab­o­ra­tor who feared for his own elec­toral risk (Lizza, 2010; Weiss, 2010).

In short, the 2010 Se­nate bill for cli­mate ac­tion failed for rea­sons that likely in­cluded these: the US poli­ti­cal sys­tem makes ac­tion against busi­ness in­ter­ests difficult, and Repub­li­can par­ti­san­ship, re­cent events, and co­or­di­na­tion failures made ac­tion even harder.

The Paris Agree­ment:

After fur­ther failures, in­ter­na­tional cli­mate ne­go­tia­tors low­ered their am­bi­tions. The 2015 Paris Agree­ment was cre­ated to by­pass the fac­tors that had doomed pre­vi­ous ne­go­ti­a­tions. Un­like pre­vi­ous agree­ments, which had sought top-down re­quire­ments to solve the col­lec­tive ac­tion prob­lem, the Paris Agree­ment fo­cused on hold­ing states to pledges they would de­ter­mine for them­selves. This differ­ence was a ma­jor con­trib­u­tor to par­ti­ci­pa­tion in the treaty by de­vel­op­ing coun­tries, which had ar­gued that similar ex­pec­ta­tions for coun­tries with differ­ent lev­els of de­vel­op­ment would be un­fair. In ad­di­tion, the agree­ment did not cre­ate bind­ings on leg­is­la­tures, so Obama could pass it as an ex­ec­u­tive agree­ment rather than seek­ing su­per­ma­jor­ity ap­proval from the Se­nate (Aschwan­den, 2015).

Some as­pects of the Paris Agree­ment were am­bi­tious; oth­ers were not. Sign­ing na­tions de­clared a global aim of limit­ing global warm­ing to “well be­low 2℃,” and to try to keep it be­low 1.5℃. Largely be­cause of US efforts, legally bind­ing miti­ga­tion was weak. So were com­pli­ance mechanisms. On the other hand, the treaty in­volved strong pro­ce­dures for es­ca­lat­ing com­mit­ments and en­sur­ing na­tional trans­parency. Na­tions widely praised the agree­ment, as­sert­ing, among other praises, that it was fair. While its con­crete obli­ga­tions were vague, the agree­ment was sym­bol­i­cally sig­nifi­cant: af­ter decades of dis­agree­ment, na­tions had agreed on the need for mul­ti­lat­eral ac­tion to se­ri­ously miti­gate cli­mate change.

Sev­eral fac­tors con­tributed to this suc­cess. First, the US and China had signed a bilat­eral agree­ment the pre­vi­ous year. Se­cond, the del­e­gates’ French hosts man­aged lo­gis­tics mas­ter­fully (they re­duced the num­ber of play­ers by bring­ing to­gether key play­ers in ne­go­ti­a­tions se­cret from other del­e­ga­tions, and they pre­sented a fi­nal agree­ment—in which all play­ers had made sig­nifi­cant com­pro­mises and re­ceived sig­nifi­cant con­ces­sions—as “take it or leave it”). Third, the vast ma­jor­ity of coun­tries had already cre­ated do­mes­tic cli­mate plans. Do­mes­tic en­vi­ron­men­tal efforts likely con­tributed to this; it may have also helped that del­e­gates’ pre­vi­ous ne­go­ti­a­tions had en­gaged them with other coun­tries’ ar­gu­ments and ex­am­ples, which per­suaded them of the eco­nomic fea­si­bil­ity and im­por­tance of cli­mate ac­tion (Dimitrov, 2016).

In the five years since, the re­sults of the Paris Agree­ment have been mixed. On the one hand, many coun­tries set in­suffi­ciently am­bi­tious tar­gets, many coun­tries have failed to reach their tar­gets, and the Trump ad­minis­tra­tion has de­clared its in­ten­tion to with­draw from the agree­ment as soon as it is legally el­i­gible to do so, which will be in Novem­ber (Roberts, 2019). On the other hand, ma­jor poli­ti­cal en­tities, in­clud­ing In­dia, the EU, and the US Cli­mate Alli­ance, have taken ma­jor cli­mate ac­tion since the agree­ment, such as in­vest­ing heav­ily in re­new­able en­ergy, while ex­plic­itly refer­enc­ing the com­mit­ments made in Paris (“Na­tional Elec­tric­ity,” 2018; “Clean En­ergy,” 2020; “Cli­mate Lead­er­ship,” 2019).

En­vi­ron­men­tal Out­comes:

Ozone layer: The ozone layer’s de­ple­tion stopped and has re­cently started re­cov­ery. The 1987 Mon­treal Pro­to­col might not have been as di­rectly im­pact­ful as some later agree­ments; alone, it was far from enough to stop ozone de­ple­tion (Ritchie and Roser, 2018).

For­est cover: Con­tem­po­rary for­est lev­els of for­est cover are much lower than pre-in­dus­trial lev­els. Still, global for­est cover has re­mained sta­ble since at least 2000. This has pre­sum­ably been be­cause “[s]ince 1990 Europe has seen an in­crease in forests while Africa and the Amer­i­cas saw forests de­clin­ing” (Roser, 2013). Europe’s in­crease in for­est cover is largely due to af­foresta­tion pro­grams (pro­grams that cre­ate forests where there had not pre­vi­ously been forests) at EU-wide as well as state-wide lev­els (Noack, 2014)

Air pol­lu­tion: Deaths from pol­lu­tion have de­clined sig­nifi­cantly, largely be­cause out­door pol­lu­tion has re­mained sta­ble, while in­door pol­lu­tion has de­clined steadily, at least since the ’90s (Ritchie and Roser, 2019). In the US, sev­eral forms of air pol­lu­tion have de­clined sig­nifi­cantly over the last sev­eral decades: sulfur diox­ide and ni­tro­gen ox­ide (ma­jor con­trib­u­tors to acid rain), lead (which is toxic), and ground-level ozone as well as par­tic­u­late mat­ter (which are harm­ful to in­hale) (“Air Qual­ity,” 2020).

Plas­tic: Plas­tic pro­duc­tion has ex­ploded, al­though man­age­ment has sig­nifi­cantly im­proved in de­vel­oped coun­tries, re­duc­ing ocean pol­lu­tion (Ritchie and Roser, 2018).

Pes­ti­cides: Pes­ti­cide use in­creased fairly steadily in the 90s and 2000s, and it has largely lev­el­led off over the past decade (Roser, 2019).

Cli­mate change: Car­bon diox­ide emis­sions have mas­sively in­creased, es­pe­cially since the be­gin­ning of the 20th cen­tury, and even more so af­ter the cen­tury’s first half. Re­cently, emis­sions in China have briefly stag­nated, while emis­sions in other de­vel­op­ing coun­tries of Asia con­tinue to climb. Anal­y­sis by Cli­mate Ac­tion Tracker sug­gests that, with cur­rent poli­cies, Earth will un­dergo about 3℃ of warm­ing (rel­a­tive to pre-in­dus­trial tem­per­a­tures) by 2100, with catas­trophic re­sults. Even if na­tions cre­ated new poli­cies suffi­cient to meet all their de­clared tar­gets, about 2.7℃ of warm­ing would still oc­cur (Ritchie and Roser, 2019).

Still, cli­mate ac­tion has achieved ma­jor suc­cesses. Anal­y­sis by Cli­mate Ac­tion Tracker sug­gests that cli­mate change miti­ga­tion poli­cies that have already been im­ple­mented will re­duce global warm­ing by 2100 by about 1.5℃, rel­a­tive to what they would have been if gov­ern­ments had made no policy changes. Con­tribut­ing to this suc­cess, US, EU, and UK car­bon emis­sions have dropped sig­nifi­cantly, even af­ter ac­count­ing for emis­sions that have been out­sourced (Ritchie and Roser, 2019). The timing of these de­clines in car­bon emis­sions sug­gests that they were caused by nat­u­ral gas re­plac­ing dirt­ier fos­sil fuels in the US, and by the 2005 launch of cap-and-trade reg­u­la­tions in the EU (which at the time in­cluded the UK).

Gover­nance of Ge­netic Eng­ineer­ing:


The gov­er­nance of ge­netic en­g­ineer­ing has re­duced a sig­nifi­cant threat to fu­ture gen­er­a­tions: cer­tain en­g­ineered pathogens could bring about hu­man ex­tinc­tion, keep­ing fu­ture gen­er­a­tions from ex­ist­ing. In the early 1970s, biol­o­gists and diplo­mats took sig­nifi­cant steps to­ward en­sur­ing the re­spon­si­ble use of ge­netic en­g­ineer­ing tech­nolo­gies. Their ser­vice to fu­ture gen­er­a­tions poses a rele­vant case study for those in­ter­ested in fu­ture-ori­ented gov­er­nance, es­pe­cially for iden­ti­fy­ing tac­tics that have been suc­cess­ful for the re­spon­si­ble man­age­ment of emerg­ing risks, and for iden­ti­fy­ing ar­eas where ex­ist­ing in­sti­tu­tions that gov­ern such risks could be im­proved.

Early Gover­nance Ini­ti­a­tives:

Since the ’70s, the Biolog­i­cal Weapons Con­ven­tion has been the main agree­ment gov­ern­ing state use of ge­netic en­g­ineer­ing. Biolog­i­cal weapons had been an in­ter­na­tional is­sue since well be­fore hu­mans dis­cov­ered ways to cre­ate novel pathogens. In the Se­cond World War, var­i­ous states had bioweapons pro­grams. As a part of in­ter­na­tional efforts to pur­sue non-pro­lifer­a­tion in the decades fol­low­ing WWII, states cre­ated the Biolog­i­cal Weapons Con­ven­tion in 1972. Nixon had ended the US bioweapons pro­gram three years prior, and this had bolstered in­ter­na­tional sup­port for such a con­ven­tion. The con­ven­tion for­bade the de­vel­op­ment, pro­duc­tion, and stock­piling of biolog­i­cal weapons and re­lated tech­nolo­gies. (Later diplo­macy clar­ified that the Con­ven­tion also for­bade the use of bioweapons). The Con­ven­tion did not cre­ate mechanisms for in­spec­tion or en­force­ment (“Biolog­i­cal Weapons Con­ven­tion,” 2003).

The fol­low­ing year, in 1973, biol­o­gists dis­cov­ered how to cre­ate novel pathogens. Early ge­netic en­g­ineer­ing in­volved re­mov­ing parts of plas­mids, loops of DNA in bac­te­ria, and re­plac­ing them with DNA from an­other or­ganism. While still rudi­men­tary, ge­netic en­g­ineer­ing promised to en­able new ad­vances in medicine and agri­cul­ture, while also cre­at­ing the po­ten­tial for sci­en­tists to cre­ate novel pathogens that could pose un­prece­dented risks.

In the same year, Soviet leader Brezh­nev both signed the Biolog­i­cal Weapons Con­ven­tion, and vi­o­lated it by se­cretly ini­ti­at­ing a pro­gram to ge­net­i­cally en­g­ineer pathogens for use in war­fare. The Soviet pro­gram for ge­net­i­cally en­g­ineered bioweapons would soon be­come the largest such pro­gram in the world (F. Calero Forero, per­sonal com­mu­ni­ca­tion, Aug. 10, 2020).

In 1974, eleven con­cerned and es­teemed biol­o­gists—in­clud­ing James D. Wat­son, co-dis­cov­erer of DNA—pub­lished a joint let­ter, ini­ti­at­ing biol­o­gists’ self-gov­er­nance on ge­netic en­g­ineer­ing re­search. Not­ing the en­dorse­ment of two sci­en­tific as­so­ci­a­tions, the au­thors called for a mora­to­rium on par­tic­u­larly risky ex­per­i­ments in ge­netic en­g­ineer­ing un­til guidelines were de­vel­oped, as well as an in­ter­na­tional meet­ing to de­velop ap­pro­pri­ate prac­tices. The au­thors ex­plic­itly ex­pressed com­mit­ment to pre­cau­tion­ary ac­tion amidst high un­cer­tainty that might not be eas­ily re­solv­able.

Eight months later, and with fi­nan­cial sup­port from pres­ti­gious sci­en­tific or­ga­ni­za­tions, 140 of the world’s best molec­u­lar biol­o­gists gath­ered at Asilo­mar to de­velop pro­pos­als by which the sci­en­tific com­mu­nity could self-reg­u­late re­search in ge­netic en­g­ineer­ing. See­ing the use of ge­netic en­g­ineer­ing in war­fare as more dis­tant, sci­en­tists ex­plic­itly chose to not raise that is­sue and in­stead fo­cus on miti­gat­ing risks from in­ad­e­quate cau­tion. Fol­low­ing the con­fer­ence, rep­re­sen­ta­tives from the Na­tional In­sti­tutes of Health used the meet­ing’s ten­ta­tive con­clu­sions to de­velop guidelines for ge­netic en­g­ineer­ing re­search (Rogers, 1975).

Gover­nance Since the ’70s:

Since those early gov­er­nance ini­ti­a­tives, deaths from the ir­re­spon­si­ble or mil­i­tary de­vel­op­ment of ge­net­i­cally en­g­ineered weapons have been min­i­mal—far from the mil­lions of deaths, or more, that seem like plau­si­ble out­comes of en­g­ineered pathogens. Still, this track record might not last, as gov­er­nance has not im­proved sub­stan­tively on early ini­ti­a­tives, while dan­gers have in­creased.

Palmer et al. (2015) write that, since Asilo­mar:

Our strate­gies and in­sti­tu­tions for man­ag­ing biolog­i­cal risk in emerg­ing tech­nolo­gies have not ma­tured much… con­clu­sions [at Asilo­mar] led to the re­com­bi­nant DNA guidelines still used to­day.

Over­sight in the US, they ex­plain, comes from an over­lap­ping patch­work of com­mit­tees and agen­cies that mainly re­strict ex­per­i­men­ta­tion by re­strict­ing fund­ing. Over­sight of aca­demic biotech­nol­ogy re­search in other states has similar limi­ta­tions.

In­ter­na­tional gov­er­nance has also seen highly limited change. The Biolog­i­cal Weapons Con­ven­tion re­mains “the foun­da­tion of the in­ter­na­tional biolog­i­cal arms con­trol regime,” and it con­tinues to lack mechanisms for in­spec­tion or en­force­ment. The ab­sence of these mechanisms has been partly caused by the great difficulty of ver­ify­ing bioweapons-re­lated com­pli­ance; pathogens can be quickly pro­duced or de­stroyed in large quan­tities, by fa­cil­ities that can claim to be us­ing their tech­nolo­gies for peace­ful phar­ma­ceu­ti­cal pur­poses. Na­tions did agree on a mea­sure for ini­ti­at­ing in­ves­ti­ga­tions—through the UN Se­cu­rity Coun­cil—but this could not have done much to de­ter the Soviet Union from cre­at­ing its mas­sive bioweapons pro­gram, as it held veto power in the Coun­cil (“Biolog­i­cal Weapons Con­ven­tion,” 2003).

Efforts to strengthen the Con­ven­tion have not suc­ceeded. Decades of diplo­macy have man­aged to ex­pand Con­ven­tion mem­ber­ship and to cre­ate mea­sures for build­ing trust be­tween par­ti­ci­pants (which may help make para­noia-driven defec­tion less likely), but more sub­stan­tive ad­vances have been limited by del­e­gates’ ex­ten­sive post­pone­ment of de­ci­sion mak­ing and failures to reach agree­ment (“Biolog­i­cal Weapons”). An Ad Hoc Group’s 2001 pro­posal of sub­stan­tive mea­sures for in­creas­ing com­pli­ance, in­clud­ing ran­dom in­spec­tions, may have been the near­est the Con­ven­tion has come to se­ri­ous en­force­ment. The US sank the pro­posal, ar­gu­ing that it would mainly hurt le­gi­t­i­mate ac­tors such as phar­ma­ceu­ti­cal com­pa­nies (“Biolog­i­cal Weapons Con­ven­tion,” 2003).

Per­haps the most sig­nifi­cant gov­er­nance change that has oc­curred has not been a change to the Con­ven­tion, but the de­cline of the Soviet bioweapons pro­gram. This oc­curred af­ter the col­lapse of the Soviet Union and the defec­tion of a lead­ing sci­en­tist (F. Calero Forero, per­sonal com­mu­ni­ca­tion, Aug. 10, 2020).

While gov­er­nance in­sti­tu­tions have not changed sub­stan­tively, risk has greatly ex­panded. The 21st cen­tury has seen the ac­cel­er­a­tion of tech­nolo­gies for edit­ing, syn­the­siz­ing, and giv­ing re­pro­duc­tive ad­van­tages to genes. Ex­am­ples of these tech­nolo­gies in­clude CRISPR and gene drives. As these tech­nolo­gies have im­proved, ac­cess to ge­netic en­g­ineer­ing has pro­lifer­ated be­tween coun­tries and within coun­tries (S. Luby, per­sonal com­mu­ni­ca­tion, Apr. 15, 2020). In 2015, the Nu­clear Threat Ini­ti­a­tive es­ti­mated that 16 coun­tries have had, or are cur­rently sus­pected of hav­ing, biolog­i­cal weapons pro­grams. At the same time, the tech­nolo­gies and knowl­edge for cre­at­ing high-risk pathogens are much fur­ther within the reach of other states, ter­ror­ist groups, aca­demics, and even am­a­teurs than they were in the 1970s (“The Biolog­i­cal Threat,” 2015; S. Luby, per­sonal com­mu­ni­ca­tion, Apr. 15, 2020).

The in­creas­ing ac­cessibil­ity of ge­netic en­g­ineer­ing con­sti­tutes a ma­jor risk. It would only take one par­tic­u­larly un­for­tu­nate ac­ci­dent or malev­olent act to cre­ate catas­tro­phe. As the COVID-19 pan­demic has shown, global health sys­tems’ ca­pac­i­ties for han­dling global out­breaks of dis­ease are highly limited.

Briefer Case Stud­ies:

An­i­mal Ad­vo­cacy—Suc­cesses and Limi­ta­tions:

This is a brief overview of suc­cesses and limi­ta­tions of the mod­ern an­i­mal ad­vo­cacy move­ment. Be­cause ma­jor shifts to­ward poli­ti­cal in­clu­sion have not oc­cured in these cases, I fo­cus on the smaller changes that have hap­pened. I also fo­cus on an­i­mals that are un­der hu­man con­trol, rather than wild an­i­mals, be­cause we know much less about the well-be­ing of wild an­i­mals (al­though we do have good rea­sons to drop the idea that their lives are idyl­lic).

Rise of an­i­mal slaugh­ter: As coun­tries have grown wealthier and cre­ated cheaper ways to raise an­i­mals for slaugh­ter and con­sump­tion, an­i­mal pro­duc­tion and con­sump­tion has ex­ploded.

Over the last 80 years, the num­ber of land an­i­mals slaugh­tered globally has in­creased more than eight-fold, to more than 70 billion land an­i­mals a year. This num­ber is the re­sult of steady in­crease since around 1985. The vast ma­jor­ity of these are chick­ens. Meat con­sump­tion per cap­ita has risen along with the to­tal pop­u­la­tion (Ritchie and Roser).

Over the last 80 years, the pro­duc­tion of aquatic an­i­mals has more than tripled. Per cap­ita con­sump­tion has in­creased sub­stan­tially. As pro­duc­tion from cap­ture (wild catch) fish­ery has slowed its in­crease, aqua­cul­ture (seafood farm­ing) has in­creased mas­sively, es­pe­cially in East Asia and the Pa­cific. Aqua­cul­ture now pro­duces a lit­tle more than half of all seafood (Ritchie and Roser, 2019).

The vast ma­jor­ity of an­i­mals used and kil­led by hu­mans are chick­ens and fish, used and slaugh­tered for con­sump­tion. The use of an­i­mals for pur­poses such as races, product test­ing, fur, and cir­cuses is com­par­a­tively tiny (“Why Farmed,” 2016).

Caged egg-lay­ing hens: The EU has banned the use of bat­tery cages for hens. In ad­di­tion, six US states (Cal­ifor­nia, Mas­sachusetts, Ore­gon, Wash­ing­ton, Michi­gan, and Colorado) have passed pro­hi­bi­tions on cages for egg-lay­ing hens. Sev­eral did so through pub­lic refer­enda (Colorado did so through the cred­ible threat of a fur­ther-reach­ing pub­lic refer­en­dum) (“An­i­mal Welfare,” 2008; Brown, 2020).

De­beaking: Seven Euro­pean coun­tries, in­clud­ing Ger­many, have banned de­beaking. In ad­di­tion, “Aus­tria en­gaged in a pro­cess whereby farm­ers who wished to con­tinue to beak trim paid a penalty that was re­dis­tributed to farm­ers keep­ing in­tact-beak flocks, and by 2005 fewer than 5% of flocks were beak-trimmed” (Ni­col, 2018).

Ag gag laws: As of June 2020, six US states have ac­tive laws crim­i­nal­iz­ing farm whistle­blow­ing. Five other states had pre­vi­ously passed similar laws, but these have been struck down as un­con­sti­tu­tional. In 17 other states, at­tempts at pass­ing such laws were defeated (“What is Ag-Gag,” 2020).

“Right to harm” amend­ments: As of 2019, 20 US states have passed con­sti­tu­tional amend­ments pro­hibit­ing hunt­ing reg­u­la­tions, and 2 states (Mis­souri and North Dakota) have passed con­sti­tu­tional amend­ments “for­ever” pro­hibit­ing re­stric­tions on mod­ern farm­ing prac­tices. Some at­tempts at pass­ing similar amend­ments in other states have been defeated. (“Bal­lot Mea­sure,” 2019; “Right to Farm”).

Fur farm­ing: Ja­pan, Cal­ifor­nia, and 14 Euro­pean coun­tries, in­clud­ing Ger­many and the UK, have passed laws ban­ning or effec­tively phas­ing out all fur farm­ing, in some cases by rais­ing welfare stan­dards enough to make it not prof­itable (“Fur Bans”). In Ja­pan, the effec­tive ban was mo­ti­vated by con­cerns over in­va­sive species es­cap­ing from fur farms (“Ja­pan,” 2016).

Cos­met­ics test­ing: As of 2018, the UK, the EU, Is­rael, Nor­way, In­dia, New Zealand, Taiwan, Switzer­land, Gu­atemala, Cal­ifor­nia, and Colom­bia have banned or phased out all an­i­mal test­ing for cos­met­ics (“Timeline: Cos­met­ics”; McClain, 2002).

An­i­mals in cir­cuses: Many small-scale ju­ris­dic­tions, as well as Cal­ifor­nia and Mex­ico, have banned the use of an­i­mals in cir­cuses (“Cir­cus Bans”).

Recog­ni­tions of an­i­mal sta­tus: Numer­ous coun­tries, es­pe­cially in Europe have legally rec­og­nized non-hu­man an­i­mals as sen­tient or at least not “things” (Hud­son, 2019).

An­i­mal product al­ter­na­tives: In­ter­est in plant-based and cul­tured al­ter­na­tives to an­i­mal prod­ucts has surged over the past few years. Alter­na­tive pro­tein com­pa­nies have re­cently raised hun­dreds of mil­lions of dol­lars in fund­ing, ma­jor meat com­pa­nies have launched meat al­ter­na­tive product lines, and re­tail sales of plant-based prod­ucts topped $5 billion (Crosser, 2020).

Ini­ti­a­tives to ban fac­tory farm­ing: Over the past few years, leg­is­la­tive efforts to ban fac­tory farm­ing have been launched in Switzer­land, France, and the US (“Swiss to Vote,” 2019; Klein, 2020; “France Launches,” 2020).

Ad­vo­cacy for Chil­dren:

For much of his­tory, chil­dren have been severely mis­treated. His­tor­i­cally wide­spread forms of abuse have in­cluded ne­glect, fear-based ma­nipu­la­tion, beat­ings, sex­ual abuse, in­fan­ti­cide, and en­slave­ment. Lloyd deMause’s The His­tory of Child­hood (1974) pro­vides a bleak sum­mary of his­tor­i­cal child­hood: “Cen­tury af­ter cen­tury of bat­tered chil­dren grew up and in turn bat­tered their own chil­dren.”

Op­po­si­tion to these abuses was rare, but not en­tirely ab­sent. In 374, for ex­am­ple, Rome crim­i­nal­ized in­fan­ti­cide, due to in­fluences in­clud­ing the Chris­tian Church and per­haps pop­u­la­tion con­cerns. In the West, Chris­ti­an­ity strongly as­so­ci­ated sex and sex­ual de­sire with sin, con­tribut­ing to the idea that chil­dren were (at least in those re­spects) in­no­cent. The later writ­ings of Rousseau also con­tributed to the as­so­ci­a­tion be­tween chil­dren and in­no­cence. Or­ga­nized chil­dren’s ad­vo­cacy origi­nated from the efforts of philan­thropists, es­pe­cially in states with laws that per­mit­ted cit­i­zen as­so­ci­a­tions aimed at so­cial re­form. In 1741, Bri­tish philan­thropist Thomas Co­ram opened a hos­pi­tal for aban­doned chil­dren, be­cause “he couldn’t bear to see the dy­ing ba­bies ly­ing in the gut­ters and rot­ting on the dung-heaps of Lon­don” (deMause, 1974).

In the 19th cen­tury United States, in­dus­tri­al­iza­tion, ur­ban­iza­tion, and im­mi­gra­tion made many chil­dren highly vuln­er­a­ble to eco­nomic ex­ploita­tion or other harms, while mak­ing chil­dren’s hard­ships and their later so­cial costs (e.g. in­creased crime) more visi­ble. In 1874, philan­thropists who had helped found the Amer­i­can So­ciety for the Preven­tion of Cru­elty to An­i­mals founded the world’s first child pro­tec­tive agency, The New York So­ciety for the Preven­tion of Cru­elty to Chil­dren. Within months, the So­ciety be­gan col­lab­o­ra­tion with lo­cal po­lice to in­ves­ti­gate and pros­e­cute cases of child maltreat­ment. Over the decades that fol­lowed, US ad­vo­cates of chil­dren’s pro­tec­tion at­tained ma­jor leg­is­la­tive vic­to­ries, in­clud­ing some just a few years af­ter the So­ciety’s found­ing. Ad­vo­cates formed more agen­cies to en­force these le­gal pro­tec­tions for chil­dren, which had pre­vi­ously been non-ex­is­tent (“His­tory”).

While ex­ploita­tive child la­bor and other forms of child abuse re­main wide­spread, chil­dren’s ad­vo­cates have had sig­nifi­cant suc­cesses. As noted above, many coun­tries have passed le­gal pro­tec­tions for chil­dren, and they en­force these. In the US, chil­dren’s ad­vo­cates took a few years to es­tab­lish some le­gal pro­tec­tions for chil­dren, but it was not un­til sev­eral gen­er­a­tions later—in the 1930s—that they man­aged to es­tab­lish se­ri­ous fed­eral re­stric­tions on child la­bor. Many coun­tries have now passed chil­dren’s la­bor pro­tec­tions, which have been boosted by com­pul­sory ed­u­ca­tion laws (Bis­sell, 2015; “His­tory”; Takan­ishi, 1978). Build­ing on ear­lier in­ter­na­tional agree­ments and dec­la­ra­tions, na­tions signed the Con­ven­tion of the Rights of the Child in 1989, declar­ing their com­mit­ments to pro­tect­ing nu­mer­ous chil­dren’s rights. All UN mem­ber states ex­cept for the US have rat­ified the Con­ven­tion (“Con­ven­tion on the Rights,” 1989).

One likely con­trib­u­tor to the suc­cesses of chil­dren’s ad­vo­cates was how, when in­dus­tri­al­iza­tion made many so­cial prob­lems more visi­ble in coun­tries where pub­lic as­so­ci­a­tion for re­form was fea­si­ble, re­form­ers of­ten saw child pro­tec­tion as a promis­ing way to pre­vent so­cial prob­lems from aris­ing. In other words, a ma­jor mo­ti­va­tion of chil­dren’s ad­vo­cates was not chil­dren’s own well-be­ing, but chil­dren’s dan­ger­ous po­ten­tial to grow up into adults who caused so­cial prob­lems.

In ad­di­tion, ad­vances in health, sci­ence, and tech­nol­ogy may have also boosted chil­dren’s pro­tec­tion, by im­prov­ing par­ent-child re­la­tions in sev­eral ways. First, the rise of birth con­trol in­creased the pro­por­tion of chil­dren who were wanted. Se­cond, re­duc­tions in child mor­tal­ity made it less risky for par­ents to in­vest in emo­tional bonds with their chil­dren. And third, im­prove­ments in de­vel­op­men­tal psy­chol­ogy have helped par­ents bet­ter un­der­stand chil­dren’s needs (Hart, 1991; Takan­ishi, 1978).

Model and Im­pli­ca­tions:


In this part of this re­port, I draw on ear­lier parts to cre­ate and ar­gue for a qual­i­ta­tive, ra­tio­nal-choice model that makes pre­dic­tions about when shifts to poli­ti­cal in­clu­sion—which is uniquely benefi­cial—oc­cur and per­sist. I then dis­cuss the model and case stud­ies’ im­pli­ca­tions for the­ory of moral cir­cle ex­pan­sion, poli­ti­cal strate­gies, and in­sti­tu­tional de­signs, with a fo­cus on strate­gic im­pli­ca­tions for sup­port­ing fu­ture gen­er­a­tions.


In many of the stud­ied cases, poli­cies were cre­ated mainly for the benefit of pow­er­ful groups, and they hap­pened to greatly benefit ex­cluded groups. In other cases, ex­cluded groups benefited be­cause some poli­ti­cal ac­tors tried to benefit them for their own sake. Benefits come more re­li­ably when pow­er­ful ac­tors try to benefit a group, so durable poli­ti­cal power would be ex­tremely valuable for ex­cluded groups, such as fu­ture gen­er­a­tions.

How do groups gain or lose rel­a­tively durable forms of poli­ti­cal powerle­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion? I in­tro­duce and ar­gue for a qual­i­ta­tive, ra­tio­nal-choice model that makes pre­dic­tions about poli­ti­cal in­clu­sion, with sup­port from many his­tor­i­cal ex­am­ples. The model sug­gests that these fac­tors make it more likely that tran­si­tions to poli­ti­cal ex­clu­sion will oc­cur and per­sist: op­por­tu­ni­ties for prof­itable ex­ploita­tion, costs of in­clu­sion, and ex­clu­sive val­ues. The model also sug­gests that these fac­tors make it more likely that tran­si­tions to poli­ti­cal in­clu­sion will oc­cur and per­sist: a group’s ca­pac­ity for re­sis­tance, strate­gic al­li­ances, in­ter-so­cietal pres­sure, and in­clu­sive val­ues.

This model and these case stud­ies sug­gest that ear­lier the­o­ries of moral cir­cle ex­pan­sion have over­stated the his­tor­i­cal im­por­tance of in­clu­sive val­ues. Sev­eral other fac­tors, re­sult­ing from ex­cluded groups’ ca­pac­i­ties to ex­ert in­fluence, have each been as in­fluen­tial as in­clu­sive val­ues, or more. In­clu­sive val­ues, then, are not so his­tor­i­cally im­por­tant; per­haps so­cial val­ues have changed retroac­tively to be­come more in­clu­sive, af­ter other fac­tors mo­ti­vated poli­ti­cal in­clu­sion. This sug­gests that we should not ex­pect past trends to­ward greater in­clu­sion to con­tinue un­til groups that can­not ex­ert in­fluence, such as fu­ture gen­er­a­tions, are also in­cluded. In­clu­sive shifts have been fre­quent in the mod­ern era, not so much be­cause of the ex­pan­sion of com­pas­sion, but more be­cause eco­nomic growth/​in­dus­tri­al­iza­tion strength­ened the re­sis­tance of ex­cluded groups, and in other cases, be­cause poli­cies mainly aimed at benefit­ing pow­er­ful groups hap­pened to benefit pow­er­less groups.

I draw on this study’s anal­y­sis of his­tor­i­cal case stud­ies to make pro­vi­sional recom­men­da­tions for ad­vo­cates of fu­ture gen­er­a­tions.


In the pre­vi­ous part of this re­port, I pre­sented case stud­ies of times when policy change greatly benefited ex­cluded groups that had lit­tle or no abil­ity to ad­vo­cate for them­selves. Here, I seek to gen­er­al­ize from those case stud­ies to les­sons for to­day’s efforts to benefit ex­cluded groups, with a fo­cus on les­sons for pro­tect­ing the in­ter­ests of fu­ture gen­er­a­tions. By “ex­cluded group,” I mean a group with lit­tle or no offi­cial power to in­fluence policy mak­ing. In this part of this study, to avoid re­dun­dancy, I of­ten make claims about spe­cific his­tor­i­cal de­vel­op­ments with­out offer­ing sup­port­ing ar­gu­ments; these ar­gu­ments and more cita­tions are in the pre­vi­ous part of this re­port.

Two Ways That Ex­cluded Groups Benefit—Luck and Power:

The case stud­ies from the pre­vi­ous part of this re­port sug­gest that it is use­ful to dis­t­in­guish be­tween two broad ways in which poli­ti­cal ac­tors benefit ex­cluded groups: they might hap­pen to benefit ex­cluded groups while at­tempt­ing to benefit a poli­ti­cally in­cluded group, or they might benefit ex­cluded groups for their own sake.

In many cases, poli­cies were cre­ated mainly for the benefit of pow­er­ful groups, and they hap­pened to greatly benefit ex­cluded groups. Th­ese benefits were un­in­ten­tional, in the sense that poli­ti­cal ac­tors’ in­ten­tions to benefit ex­cluded groups were not ma­jor causes of the policy that brought about these benefits. In­stead, benefit­ing an ex­cluded group was a means to, or a side effect of, benefit­ing a pow­er­ful group. The case stud­ies from ear­lier offer sev­eral ex­am­ples of times when ex­cluded groups re­ceived ma­jor benefits, but largely for the sake of some­one else:

  • Early efforts to gov­ern ge­netic en­g­ineer­ing have gone far in pro­tect­ing fu­ture gen­er­a­tions: they have made it harder for sci­en­tists to cre­ate novel pathogens that could keep fu­ture gen­er­a­tions from ex­ist­ing (by wiping out hu­man­ity) or at least harm their well-be­ing. How­ever, these early efforts seem to have been driven by con­cerns over risks to pre­sent gen­er­a­tions, who have much poli­ti­cal power.

    • Nei­ther the let­ter that called for sci­en­tists to con­vene at Asilo­mar nor the con­fer­ence’s sum­mary state­ment so much as men­tioned risks to fu­ture gen­er­a­tions (Berg, et al., 1974, 1975).

  • Ma­jor suc­cesses in en­vi­ron­men­tal con­ser­va­tion have state­ments of in­ten­tion that are al­most en­tirely con­cerned with benefits to hu­mans, not the en­vi­ron­ment. While these mis­sion sys­tems do sug­gest that their benefits to fu­ture gen­er­a­tions were in­ten­tional, they also sug­gest that ecosys­tems and species have been mainly pre­served, not for their sake, but be­cause they hap­pen to be con­ve­nient for hu­mans.

    • The US Na­tional Parks Ser­vice de­scribes its mis­sion as pre­serv­ing na­tional parks “for the en­joy­ment, ed­u­ca­tion, and in­spira­tion of this and fu­ture gen­er­a­tions” (“About Us”).

    • The main treaty pro­tect­ing en­dan­gered species from in­ter­na­tional trade be­gins by declar­ing that en­dan­gered species “must be pro­tected for this and the gen­er­a­tions to come” (“Con­ven­tion on In­ter­na­tional Trade,” 1973).

  • Ad­vo­cacy for chil­dren was sig­nifi­cantly mo­ti­vated by its po­ten­tial for re­duc­ing fu­ture crime and other so­cial prob­lems; chil­dren were of­ten pro­tected be­cause they would be the “re­deemers of so­ciety,” rather than be­cause they already mat­tered (Hart, 1991).

Some (e.g. Willi­ams, 1944) have ar­gued that slav­ery de­clined be­cause it ceased to be in the eco­nomic in­ter­ests of pow­er­ful ac­tors. In an ex­treme ver­sion of this hy­poth­e­sis, the UK abol­ished slav­ery be­cause slave plan­ta­tion own­ers thought they would make more money if they re­ceived com­pen­sa­tion for their slaves than if they con­tinued us­ing their slaves. If this were the case, it would mean that the abo­li­tion of slav­ery, too, was an ex­am­ple of ex­cluded groups benefit­ting for the sake of some­one else. Is this view about the de­cline of slav­ery cor­rect?

This cyn­i­cal anal­y­sis of abo­li­tion is prob­a­bly too cyn­i­cal. On top of many ar­gu­ments that eco­nomic his­to­ri­ans have made, the UK passed eman­ci­pa­tion a year af­ter elec­toral re­form slashed the rep­re­sen­ta­tion of West In­dian plan­ta­tion own­ers (Hochschild, 2005). This sug­gests that plan­ta­tion own­ers had been block­ing, not sup­port­ing, com­pen­sated eman­ci­pa­tion. Also, in the US, the use of slaves in­creased un­til the Civil War, sug­gest­ing that US slav­ery was not in eco­nomic de­cline be­fore eman­ci­pa­tion (“Re­ca­pitu­la­tion of the Tables,” 1864). The abo­li­tion of slav­ery, then, was prob­a­bly not a case of slaves get­ting lucky.

The ex­am­ples of “benefits for some­one else’s sake” above are a use­ful but un­re­li­able form of pro­tec­tion; they will not go fur­ther than ar­eas where the in­ter­ests of in­cluded and ex­cluded groups hap­pen to over­lap. For­tu­nately, there have been more re­li­able sources of benefits.

In other cases, ex­cluded groups were benefited be­cause some poli­ti­cal ac­tors tried to benefit ex­cluded groups for their own sake. In these cases, some strate­gic ac­tors aim for poli­cies that are very benefi­cial for an ex­cluded group, so they no longer need luck. Poli­cies that greatly benefit the ex­cluded group, while im­pos­ing small costs or no costs on an in­cluded group, will more of­ten be passed. This means benefits come more re­li­ably when some pow­er­ful ac­tor is pur­su­ing them for their own sake. Ex­am­ples of these benefits in­clude many cases of abo­li­tion­ism, ex­pan­sions of vot­ing rights, and an­i­mal welfare pro­tec­tions. (Law­mak­ers who were co­erced into de­moc­ra­ti­za­tion may not have been try­ing to benefit ex­cluded groups for their own sake, but the pow­er­ful dis­en­fran­chised groups who did the co­erc­ing were try­ing to do just that.)

Be­cause poli­cies will more re­li­ably benefit a group if some pow­er­ful poli­ti­cal ac­tor is try­ing to benefit that group for their own sake, a policy shift that is par­tic­u­larly valuable for any group is poli­ti­cal in­clu­sion—a rel­a­tively durable in­crease in a group’s poli­ti­cal power, es­pe­cially through le­gal pro­tec­tions or poli­ti­cal rep­re­sen­ta­tion. When an ex­cluded group gains poli­ti­cal power, its mem­bers will tend to use that power to gain a wide range of benefits for them­selves; the ex­cluded group be­comes the pow­er­ful poli­ti­cal ac­tor seek­ing benefits for their own sake. For ex­am­ple, when former slaves gained some le­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion in the US, they used this new poli­ti­cal power to se­cure ac­cess to pub­lic ed­u­ca­tion (“Afri­can Amer­i­cans and Ed­u­ca­tion,” 2019).

In sum, ex­cluded groups some­times hap­pen to benefit from poli­cies; other times, in­fluen­tial ac­tors try to benefit them. Benefits come more re­li­ably when pow­er­ful ac­tors try to benefit a group, so durable poli­ti­cal power would be ex­tremely valuable for ex­cluded groups, such as fu­ture gen­er­a­tions. (Of course, they can­not be di­rectly em­pow­ered, but they can be em­pow­ered through prox­ies that have power to act for their in­ter­ests.) Given the im­por­tance of poli­ti­cal in­clu­sion—its ten­dency to last­ingly bring about a wide range of benefits for pre­vi­ously ex­cluded groups—it would be use­ful to know what makes shifts to poli­ti­cal in­clu­sion hap­pen and per­sist.

Qual­i­ta­tive Model—Con­trib­u­tors to Poli­ti­cal In­clu­sion and Ex­clu­sion:

Back­ground and As­sump­tions:

This model aims to make pre­dic­tions about when tran­si­tions to poli­ti­cal in­clu­sion (e.g. le­gal pro­tec­tions or poli­ti­cal rep­re­sen­ta­tion) are more likely to oc­cur and per­sist. It is largely in­spired by Ace­moglu and Robin­son’s (2005) model of de­moc­ra­ti­za­tion. While Ace­moglu and Robin­son’s model fo­cuses on vot­ing rights, this model aims to be broader in ap­pli­ca­tion, mak­ing pre­dic­tions about le­gal pro­tec­tions (e.g. pro­tec­tions from slav­ery) as well as vot­ing rights.

As noted be­fore, this model is about par­tic­u­larly durable forms of poli­ti­cal in­clu­sion: le­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion. Th­ese may be called de jure, or offi­cial, forms of poli­ti­cal power. Un­like some other ways in which a group might gain poli­ti­cal power (e.g. a surge of sup­port in re­sponse to a na­tional crisis), le­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion tend to be rel­a­tively long-last­ing.

Why are le­gal pro­tec­tions and poli­ti­cal rep­re­sen­ta­tion es­pe­cially durable? Be­cause, if a coun­try op­er­ates un­der the rule of law, with­draw­ing le­gal pro­tec­tions or poli­ti­cal rep­re­sen­ta­tion re­quires go­ing through a difficult pro­cess for chang­ing policy, and the new poli­ti­cal power of a formerly ex­cluded group makes it more difficult for peo­ple to suc­cess­fully go through that pro­cess. Sig­nifi­cant le­gal pro­tec­tions or poli­ti­cal rep­re­sen­ta­tion, once given, tend to be difficult to with­draw. This makes these pow­ers es­pe­cially benefi­cial for ex­cluded groups, and the fo­cus of this model.

The model makes the fol­low­ing sim­plify­ing as­sump­tions:

  • There are two pri­mary poli­ti­cal ac­tors: group A and group B.

    • The two groups may differ in offi­cial poli­ti­cal power and wealth.

    • There are also poli­ti­cal ac­tors from other so­cieties (e.g. the pres­i­dent of an­other na­tion). They may in­cen­tivize group A to be in­clu­sive or ex­clu­sive.

  • Each group has co­her­ent prefer­ences over poli­ti­cal in­sti­tu­tions (offi­cial rules about who has power and how they can use it).

    • Groups’ prefer­ences re­sult from their per­cep­tions of costs and benefits; groups pre­fer in­sti­tu­tions that they per­ceive as more benefi­cial and less costly.

    • Groups mostly (but not en­tirely) care about their own ma­te­rial well-be­ing.

    • (I make the above as­sump­tions, not be­cause I think they are ob­vi­ous, but be­cause they seem to make ac­cu­rate pre­dic­tions.)

    • Group A may hold some in­clu­sive or ex­clu­sive val­ues.

  • Group A can at­tempt to poli­ti­cally ex­clude (re­duce the de jure power of) group B, or group A can poli­ti­cally in­clude group B.

    • If group B is in­cluded, they will have more in­fluence over what gov­ern­ment poli­cies are passed.

    • If group B is ex­cluded, group A will be more able to ex­ploit group B for eco­nomic gain.

    • Group B might be able to re­sist at­tempts at ex­clud­ing it or keep­ing it in ex­clu­sion, e.g. through vi­o­lent re­volt (not ap­pli­ca­ble for fu­ture peo­ple).


This sec­tion uses the above frame­work to rea­son about the effects that sev­eral fac­tors have on the like­li­hood of tran­si­tions to poli­ti­cal in­clu­sion, or away from it. The iden­ti­fi­ca­tion of rele­vant fac­tors draws heav­ily from the his­tor­i­cal case stud­ies from the pre­vi­ous part of this re­port. Con­clu­sions are sum­ma­rized in the be­gin­ning of the next sec­tion.

Given the above as­sump­tions, what makes it more likely that group A will choose to poli­ti­cally in­clude group B, which has been ex­cluded? In other words, what con­tributes to tran­si­tions to­ward poli­ti­cal in­clu­sion? His­tor­i­cal ex­am­ples of such tran­si­tions in­clude the eman­ci­pa­tion of slaves and serfs, and ex­pan­sions of the vot­ing fran­chise.

  • In some cases of in­clu­sion, group A will­ingly in­cludes group B (this will­ing­ness may re­sult from co­er­cion), be­cause group A is will­ing to pay the costs of in­clud­ing group B. What makes this likely?

    • Tran­si­tions to poli­ti­cal in­clu­sion are more likely in­so­far as group A’s per­ceived benefits from in­clud­ing group B are high. What fac­tors, to the de­gree that they are pre­sent, con­tribute to this?

      • Effec­tive re­sis­tance from group B, or the ex­pec­ta­tion of fu­ture effec­tive re­sis­tance, means that in­clu­sion al­lows group A to benefit from avoid­ing the costs of fu­ture re­sis­tance; this is es­pe­cially in­fluen­tial if a fo­cus­ing event/​wors­ened con­di­tions help group B co­or­di­nate, if re­pres­sion is an un­ap­peal­ing al­ter­na­tive be­cause group A has vuln­er­a­ble hu­man cap­i­tal or few re­sources, or if group A has high-re­turn in­vest­ment op­por­tu­ni­ties other than over­com­ing re­sis­tance (high op­por­tu­nity costs)

      • Strate­gic al­li­ance: group A ex­pects that, if in­cluded, group B will use its new power in a way benefi­cial for group A, such as by sup­port­ing group A poli­ti­cally or mil­i­tar­ily.

      • In­clu­sive val­ues: group A val­ues in­creas­ing group B’s durable poli­ti­cal power for non-self-in­ter­ested rea­sons (this is more likely if mem­bers of group A have the or­ga­ni­za­tional pro­tec­tions of civil liber­ties).

      • In­ter-so­cietal pres­sure for in­clu­sion is strong, which is more likely if an in­fluen­tial in­ter­na­tional ac­tor is not ex­clu­sive in the way that is be­ing con­sid­ered, and has in­clu­sive val­ues (per­haps in­sincerely, to jus­tify/​dis­tract from its own harm­ful poli­cies) or stands to gain from strate­gic al­li­ances (e.g. weak­en­ing the ex­ploit­ing group) or oth­er­wise re­duc­ing ri­vals’ prof­its.

    • Tran­si­tions to poli­ti­cal in­clu­sion are more likely in­so­far as group A’s per­ceived costs from in­clud­ing group B are low. What fac­tors, to the de­gree they are pre­sent, con­tribute to this?

      • Group A does not prof­itably ex­ploit group B.

      • Un­der the in­clu­sion be­ing con­sid­ered, in­clu­sive in­sti­tu­tions are not costly for group A, e.g. group B is too small or wealthy to pass high taxes un­fa­vor­able to group A.

  • In other cases of in­clu­sion, in­clu­sion hap­pens with­out group A choos­ing it; group B uses its (per­haps un­offi­cial) power to vi­o­lently re­place the ex­ist­ing gov­ern­ment with one that in­cludes group B (fu­ture gen­er­a­tions lack this op­tion). What makes this likely?

    • Group B’s per­ceived benefits from vi­o­lent rev­olu­tion are high. What con­tributes to this?

      • Cur­rent in­sti­tu­tions are highly costly for Group B rel­a­tive to those group B would cre­ate if in­cluded, e.g. end­ing very costly ex­ploita­tion or re­dis­tribut­ing wealth from a wealthy elite.

    • Group B’s per­ceived costs from vi­o­lent rev­olu­tion are low. What con­tributes to this?

      • Risk of failure or very costly vic­tory is low, be­cause group A has a low ca­pac­ity to re­press re­bel­lion, or group B can effec­tively co­or­di­nate (this is es­pe­cially likely fol­low­ing fo­cus­ing events/​wors­ened con­di­tions, and if ex­clu­sion does not make co­or­di­na­tion difficult/​im­pos­si­ble through mea­sures such as pro­hi­bi­tions on ed­u­ca­tion).

What about tran­si­tions to­ward poli­ti­cal ex­clu­sion? What makes it more likely that group A will choose to poli­ti­cally ex­clude group B, which has been ex­cluded? His­tor­i­cal ex­am­ples of such tran­si­tions in­clude the en­slave­ment of Afri­cans, the es­ca­lat­ing marginal­iza­tion of Jewish peo­ple un­der the Nazis, con­trac­tions of the vot­ing fran­chise, and au­thor­i­tar­ian coups.

  • In these cases, group A is will­ing to pay the costs of ex­clud­ing group B, which had been in­cluded. What makes this likely?

    • Tran­si­tions to poli­ti­cal ex­clu­sion are more likely in­so­far as group A’s per­ceived gains from ex­clud­ing group B are large. What fac­tors, to the de­gree they are pre­sent, con­tribute to this?

      • Group A can prof­itably ex­ploit group B, e.g. be­cause group A has many (already ex­ist­ing or po­ten­tial) re­sources that group A could seize, such as nat­u­ral re­sources, wealth, or la­bor power (if the econ­omy makes it prof­itable to use forced la­bor).

      • Group B has the mo­ti­va­tion and for­mal power to make the cur­rent in­sti­tu­tions costly for group A, e.g. by pass­ing re­dis­tribu­tive poli­cies if group B is much less wealthy; ex­clud­ing group B elimi­nates these costs.

      • Ex­clu­sive val­ues: Group A val­ues ex­clud­ing group B for non-self-in­ter­ested rea­sons (e.g. dur­ing or af­ter war).

    • Tran­si­tions to poli­ti­cal ex­clu­sion are more likely in­so­far as group A’s per­ceived costs from ex­clud­ing group B are low. What fac­tors, to the de­gree they are pre­sent, con­tribute to this?

      • Power is so asym­met­ric that group B has lit­tle power to re­sist ex­clu­sion im­posed by group A (re­sis­tance cre­ates costs di­rectly, es­pe­cially if group A is highly in­vested in hu­man cap­i­tal, and it cre­ates risk of no/​re­duced benefit from at­tempt­ing to ex­clude a group).

      • Lack of strate­gic al­li­ances: group A does not see it­self as benefit­ing poli­ti­cally/​mil­i­tar­ily from group B’s poli­ti­cal in­clu­sion.

      • In­ter-so­cietal pres­sure for in­clu­sion is weak.

The Model, Con­densed:

We can con­dense the above rea­son­ing into the fol­low­ing sum­mary. (The main sim­plifi­ca­tion is only list­ing fac­tors once if they both make tran­si­tions to in­clu­sion more likely and tran­si­tions to ex­clu­sion less likely, or vice versa. In ad­di­tion, vi­o­lent rev­olu­tion is grouped to­gether with the threat of it, un­der “ca­pac­ity for re­sis­tance.”)

  • The fol­low­ing fac­tors, when high, make it more likely that tran­si­tions to greater poli­ti­cal ex­clu­sion will oc­cur and per­sist:

    • The de­gree of ex­ist­ing/​po­ten­tial prof­itable ex­ploita­tion of the group un­der consideration

    • The costs of ex­ist­ing/​po­ten­tial in­clu­sion (e.g. higher taxes)

    • Ex­clu­sive val­ues (e.g. dur­ing/​af­ter war)

  • The fol­low­ing fac­tors, when high, make it more likely that tran­si­tions to greater poli­ti­cal in­clu­sion will oc­cur and per­sist:

    • The ca­pac­ity of a vuln­er­a­ble group to effec­tively re­sist tran­si­tions to/​per­pet­u­a­tion of ex­clu­sion (strength­ened by high co­or­di­nat­ing ca­pac­ity and vuln­er­a­ble au­thor­i­ties)

    • Ex­ist­ing/​po­ten­tial strate­gic al­li­ances be­tween a vuln­er­a­ble group and an in­fluen­tial group (e.g. the vuln­er­a­ble group, if in­cluded, would elec­torally/​mil­i­tar­ily/​eco­nom­i­cally benefit those who are already in­cluded)

    • In­clu­sive val­ues (for these to emerge when rul­ing in­ter­ests profit from ex­clu­sion, civil liber­ties are very helpful)

    • In­ter-so­cietal pres­sure for in­clu­sion (mo­ti­vated by in­ter-so­cietal in­clu­sive val­ues, strate­gic al­li­ances, or de­sires to oth­er­wise re­duce ri­vals’ prof­its)

A scale, vi­su­ally rep­re­sent­ing con­trib­u­tors to poli­ti­cal in­clu­sion and ex­clu­sion. When fac­tors on the left side—prof­ita­bil­ity of ex­ploita­tion, costs of in­clu­sion, and ex­clu­sive val­ues—cre­ate stronger in­cen­tives for pow­er­ful ac­tors, this model pre­dicts that tran­si­tions to­ward ex­clu­sion are more likely to oc­cur and per­sist. When the other fac­tors—dis­cussed above—cre­ate stronger in­cen­tives, this model pre­dicts that in­clu­sion is more likely to oc­cur and per­sist.

His­tor­i­cal Ex­am­ples:

This sec­tion offers ex­am­ples of his­tor­i­cal cases when the pres­ence of each of the fac­tors iden­ti­fied above, or their ab­sence, seem to have had the pre­dicted effect on tran­si­tions to in­clu­sion or ex­clu­sion. While this is not too sur­pris­ing—many of these ex­am­ples in­formed the cre­ation of the model—the model gains sup­port from the wide range of cases that its rel­a­tively sim­ple frame­work suc­cess­fully pre­dicts. Th­ese ex­am­ples fo­cus on abo­li­tion­ism and de­moc­ra­ti­za­tion, be­cause these are es­pe­cially clear ex­am­ples of durable poli­ti­cal in­clu­sion, and that is what the model makes pre­dic­tions about.

Ex­am­ples of op­por­tu­ni­ties for prof­itable ex­ploita­tion fa­vor­ing poli­ti­cal ex­clu­sion:

  • The mass en­slave­ment of Afri­cans for the transat­lantic slave trade was uniquely prof­itable.

    • Euro­peans’ trade in guns made it very prof­itable for some Afri­cans to en­slave other Afri­cans, be­cause guns were very use­ful. Euro­peans cre­ated this very large de­mand for slaves be­cause send­ing Afri­can slaves to the Amer­i­cans would be highly prof­itable, es­pe­cially be­fore the 19th cen­tury, for rea­sons that in­cluded the fol­low­ing:

      • Euro­peans had a large la­bor short­age in trop­i­cal re­gions of the Amer­i­cas, which had great en­vi­ron­men­tal con­di­tions for grow­ing cash crops. Espe­cially be­fore in­dus­tri­al­iza­tion, this agri­cul­tural work was very la­bor-in­ten­sive, and it was rel­a­tively difficult for re­sis­tant la­bor­ers to sab­o­tage it.

      • Afri­cans tended to be more re­sis­tant than Euro­peans to trop­i­cal dis­ease.

      • There were not yet coun­tries tak­ing ma­jor steps to block the slave trade.

      • Many Afri­can so­cieties did not yet have many guns, so whole so­cieties were highly vuln­er­a­ble to slave cap­tors armed with guns, mak­ing it rel­a­tively cheap for Euro­peans to ex­change guns for slaves (Whatley, 2008).

  • In Haiti, the US South, the French colo­nial em­pire, and the Ot­toman em­pire, slav­ery was only abol­ished once the gov­ern­ment—which greatly prof­ited from slav­ery—was vi­o­lently re­placed by an­other gov­ern­ment that did not profit so much from slav­ery (Drescher, 2015; Fer­gu­son and Toledano, 2017). In con­trast, slav­ery was abol­ished rel­a­tively early in ju­ris­dic­tions where slav­ery was not very prof­itable (e.g. New England, part of Western Europe) (Fenoaltea, 1984).

  • The use of in­ter­gen­er­a­tional re­sources and the mis­treat­ment of farmed an­i­mals has also been highly prof­itable, likely driv­ing and sus­tain­ing the ex­clu­sion of fu­ture gen­er­a­tions and non-hu­man an­i­mals.

  • Schol­ars of geno­cide—study­ing geno­cides in­clud­ing those car­ried out by Amer­i­can coloniz­ers and set­tlers, the Nazis, the Turks, Stalin’s regime, and Mao’s regime—have con­cluded that “eco­nomic mo­ti­va­tions are ex­tremely im­por­tant to geno­cide” (N. Naimark, per­sonal com­mu­ni­ca­tion, May 19, 2020).

    • In all of these cases, op­pres­sors per­ceived (ac­cu­rately, or un­der the in­fluence of mis­lead­ing stereo­types) that the groups they sought to elimi­nate held great wealth, which the op­pres­sors could take if they mur­dered these peo­ple.

    • Geno­cide is, in im­por­tant ways, a case of poli­ti­cal ex­clu­sion; elimi­nat­ing a group last­ingly re­duces the group’s poli­ti­cal power, e.g. their abil­ity to keep oth­ers from steal­ing their wealth.

Ex­am­ples of high costs of in­clu­sion fa­vor­ing poli­ti­cal ex­clu­sion:

  • Through­out the Western hemi­sphere, states with more eco­nomic in­equal­ity—states where elites had more to lose from pop­ulist democ­racy—were slow­est to de­moc­ra­tize, and they fre­quently re­verted to rule by the few (Ace­moglu and Robin­son, 2005; Enger­man and Sokoloff, 2005).

    • Where these high per­ceived costs were ab­sent (es­pe­cially in the early US, and even more so in US fron­tier states, where dis­en­fran­chised white men were a rel­a­tively small and wealthy group—less threat­en­ing to elites), de­moc­ra­ti­za­tion hap­pened rel­a­tively quickly (Enger­man and Sokoloff, 2005).

  • In an­other case where high costs of in­clu­sion were ab­sent—the pas­sage of laws pro­tect­ing chil­dren from cru­elty (not in­clud­ing child la­bor)—poli­ti­cal pro­tec­tions came very quickly af­ter or­ga­ni­za­tions be­gan ad­vo­cat­ing for them (“His­tory”).

Ex­am­ples of ex­clu­sive val­ues con­tribut­ing to poli­ti­cal ex­clu­sion:

  • De­hu­man­iza­tion, es­pe­cially dur­ing war, has been a fre­quent con­trib­u­tor to geno­cide (N. Naimark, per­sonal com­mu­ni­ca­tion, May 19, 2020). War­time ex­clu­sive val­ues have also fre­quently con­tributed to other forms of poli­ti­cal ex­clu­sion, such as the US in­tern­ment of Ja­panese Amer­i­cans.

  • Racism has helped sus­tain race-based sys­tems of op­pres­sion.

Ex­am­ples of effec­tive re­sis­tance con­tribut­ing to poli­ti­cal in­clu­sion:

  • In Haiti, slaves used vi­o­lent re­volt to re­place the ex­ist­ing gov­ern­ment with one that abol­ished slav­ery (Hochschild, 2005).

  • Slave re­sis­tance was a sig­nifi­cant mo­ti­va­tor for the pas­sage of abo­li­tion and eman­ci­pa­tion by the Bri­tish, whose an­ti­s­lav­ery efforts later be­came ex­tremely in­fluen­tial in­ter­na­tion­ally (Hochschild, 2005).

    • The UK’s 18067 abo­li­tion of the slave trade came a few years af­ter Haiti, fol­low­ing the most suc­cess­ful slave re­bel­lion in hu­man his­tory (which was dev­as­tat­ing for French own­ers of slave plan­ta­tions), es­tab­lished in­de­pen­dence in 1804.

      • It was widely be­lieved that slaves who had been kid­napped and en­slaved were more likely to rebel than peo­ple who were born slaves, so fears of slave re­bel­lions would have been strong mo­ti­va­tors for abol­ish­ing the slave trade.

      • France’s loss of its wealthiest slave colony also meant the UK would have less to lose against French busi­ness com­peti­tors if it abol­ished the slave trade.

    • The UK’s 1833 pas­sage of eman­ci­pa­tion for most slaves in the Bri­tish Em­pire closely fol­lowed the 1831-2 Bap­tist War—the biggest slave re­bel­lion that ever oc­cured in Bri­tish colonies, and it de­stroyed the prop­erty of many slave plan­ta­tions.

    • Th­ese in­fluen­tial cases of slave re­sis­tance tended to oc­cur af­ter ru­mors of an­ti­s­lav­ery poli­ti­cal events in­creased slaves’ ca­pac­ity to re­sist, by helping them co­or­di­nate their efforts.

  • Slave re­sis­tance more gen­er­ally in­creased se­cu­rity and su­per­vi­sion costs, es­pe­cially for cer­tain types of la­bor (e.g. for la­bor tasks that were prof­itable in New England and Canada, as well as in in­dus­trial so­cieties). This con­tributed to the re­duced eco­nomic vi­a­bil­ity of slave la­bor in these so­cieties, re­duc­ing the use of slav­ery and the sup­port for keep­ing it (Fenoaltea, 1984).

  • Dur­ing the transat­lantic slave trade, many Afri­can so­cieties did not have many guns, so their mem­bers had low abil­ities to re­sist en­slave­ment by slavers armed with guns.

  • Most cases of de­moc­ra­ti­za­tion through­out his­tory came through the dis­en­fran­chised or­ga­niz­ing to di­rectly threaten the elite’s in­ter­ests (Ace­moglu and Robin­son, 2000, 2005; Aidt, et al., 2015; Das­gupta and Ziblatt, 2015).

    • When and where the ca­pac­ity of the dis­en­fran­chised to re­sist has in­creased, de­moc­ra­ti­za­tion has been more fre­quent than in other cases. Ex­am­ples of this in­clude the fol­low­ing (Ace­moglu and Robin­son, 2000, 2005):

      • When Euro­pean mer­chants liv­ing un­der monar­chs gained wealth (em­pow­er­ing the At­lantic “bour­geois rev­olu­tions”)

      • When na­tional crises (e.g. mil­i­tary defeats, de­pres­sions) or im­proved com­mu­ni­ca­tion tech­nolo­gies helped the dis­en­fran­chised co­or­di­nate their resistance

      • When in­dus­tri­al­iza­tion made elites’ wealth more vuln­er­a­ble to violence

      • When world wars drained Euro­pean rulers’ abil­ities to re­press do­mes­tic as well as colo­nial resistance

  • In ma­jor cases when the vot­ing fran­chise was con­tracted, a com­mon fea­ture was that the wider pub­lic had re­cently lost its abil­ity to re­sist ex­clu­sion through or­di­nary law­mak­ing pro­cesses, be­cause rad­i­cal pro­gres­sives who rep­re­sented the wider pub­lic gov­erned in ways that made them lose pop­u­lar sup­port.

    • Un­pop­u­lar forms of gov­er­nance by rad­i­cals have in­cluded mass ex­e­cu­tions, anti-Church poli­cies, failing to end civil con­flict, be­ing as­so­ci­ated with mil­i­tary defeat, and hav­ing gained power through vi­o­lence.

    • (Peo­ple might vote against poli­ti­ci­ans who rep­re­sent them be­cause poli­ti­cal rep­re­sen­ta­tion is not the only thing that vot­ers care a lot about.)

    • This seems to have oc­curred in the First and Se­cond French Repub­lics, in the Re­con­struc­tion Era US South, in 1880s Colom­bia, and in 1920s Hun­gary.

Ex­am­ples of strate­gic al­li­ances con­tribut­ing to in­clu­sion:

  • In sev­eral cases, slave liber­a­tors ex­pected that eman­ci­pa­tion would di­rectly or in­di­rectly boost their mil­i­tary power:

    • In Cen­tral and Eastern Euro­pean states—in­clud­ing Rus­sia, Aus­tria, Prus­sia, and Poland—rulers ended serf­dom to avoid its per­ceived con­tri­bu­tions to mil­i­tary weak­ness, pre­sum­ably through the eco­nomic or mil­i­tary in­fluence of peas­ants’ choices. This hap­pened partly be­cause rulers no­ticed the mil­i­tary strength of the UK and France (e.g. in the Crimean War), which by then only used offi­cially free la­bor (Drescher, 2015, 2017).

    • Both Union­ists in the Amer­i­can Civil War and in­de­pen­dence fighters in Cuba’s Ten Years’ War offered free­dom to slaves who ran away from en­emy plan­ta­tions (Ber­gad, 2017; Lin­coln, 1863). Avoid­ing this mil­i­tary li­a­bil­ity was one mo­ti­va­tor for Spain’s eman­ci­pa­tion of Cuban slaves. In ad­di­tion, France’s first eman­ci­pa­tion was largely an at­tempt to weaken Haiti­ans’ sup­port for in­de­pen­dence (Hochschild, 2005).

  • De­moc­ra­ti­za­tion was some­times driven by elites’ de­sires to gain eco­nom­i­cally or mil­i­tar­ily use­ful-to-the-elite con­tri­bu­tions.

    • Fron­tier states in the US, com­pet­ing for set­tlers who would fill their la­bor short­ages, de­moc­ra­tized quickly (Enger­man and Sokoloff, 2005).

    • The origi­nal thir­teen colonies in the US were not as in­cen­tivized to ap­peal to set­tlers, and they were slower to de­moc­ra­tize (Enger­man and Sokoloff, 2005). Still, when they did, a ma­jor mo­ti­va­tion was want­ing to gain more loyal sup­port for mil­i­tias, and per­haps for poli­ti­cal par­ties, from the formerly dis­en­fran­chised (Hor­pedahl, 2011; Keyssar, 2000).

Ex­am­ples of in­clu­sive val­ues con­tribut­ing to in­clu­sion:

  • The UK, New England, and Puerto Rico had in­fluen­tial abo­li­tion­ist as­so­ci­a­tions—which in­cluded many mem­bers who were hor­rified by slav­ery—that pushed for eman­ci­pa­tion (Hochschild, 2005; Drescher, 2015, 2017; Ber­gad, 2017). In con­trast, the in­fluence of such or­ga­ni­za­tions was largely miss­ing in states that lacked strong pro­tec­tions for civil liber­ties (e.g. au­thor­i­tar­ian France, the US South, and many so­cieties in Africa and Asia); these tended to be much slower to abol­ish slav­ery, and they more of­ten did so mainly in re­sponse to ex­ter­nal pres­sures (Fer­gu­son and Toledano, 2017; Fin­nie, 1969).

  • Early ad­vo­cates of chil­dren’s pro­tec­tion seem to have had highly hu­man­i­tar­ian con­cerns (deMause, 1974).

Ex­am­ples of in­ter-so­cietal pres­sure con­tribut­ing to in­clu­sion:

  • In­ter-so­cietal pres­sure mo­ti­vated by in­clu­sive val­ues (in­clud­ing for PR rea­sons): Europe and es­pe­cially the UK in­ter­na­tion­ally pro­moted an­ti­s­lav­ery (Drescher, 2017; Fer­gu­son and Toledano, 2017). In­ter­na­tional sanc­tions on South Africa’s apartheid regime may also have been an ex­am­ple of this.

  • In­ter-so­cietal pres­sure mo­ti­vated by strate­gic al­li­ances: states such as Span­ish Florida offered free­dom to slaves who ran away from ri­val colonies (Hinks and McKivi­gan, 2007).

  • In­ter-so­cietal pres­sure mo­ti­vated by de­sires to oth­er­wise re­duce ri­vals’ prof­its: Bri­tish plan­ta­tion own­ers sup­ported Bri­tish diplo­matic efforts for ban­ning the im­por­ta­tion of slaves to their com­peti­tors’ plan­ta­tions, and anti-colo­nial states later pro­moted an­ti­s­lav­ery in Euro­pean em­pires (Hochschild, 2005; Drescher, 2017).


This sec­tion elab­o­rates on and clar­ifies sev­eral points about the model in­tro­duced in the above sec­tions.

  • The model’s claims about tran­si­tions are equiv­a­lent to (differ­ent) claims about per­sis­tence; any state of af­fairs is more likely to per­sist when tran­si­tions away from it are less likely. In more ab­solute terms, per­sis­tence hap­pens when the con­di­tions nec­es­sary for a tran­si­tion away from the pre­sent state are not met.

    • Poli­cies are per­pet­u­ated be­cause those in power, on bal­ance, want to per­pet­u­ate them. As a re­sult, changes hap­pen when the col­lec­tive prefer­ences of those in power change (which may be be­cause of a change of which peo­ple are in power).

  • The way in which this model is re­lated to Parts 2’s qual­i­ta­tive model of in­sti­tu­tional per­sis­tence and change is that this model de­scribes spe­cific cases of the other model’s much broader re­la­tion­ships.

  • Ja­cobs’ (2011) work can be use­fully thought of as adding nu­ance to how the above fac­tors work and in­ter­act:

    • The de­gree to which a cost/​benefit be­comes a per­ceived cost/​benefit de­pends largely on how much at­ten­tion one pays to it; some events (e.g. sud­den catas­tro­phes) in­crease how much at­ten­tion peo­ple pay to some costs or benefits, while poli­cies that ob­scure costs/​benefits (e.g. by spread­ing them out) de­crease how much at­ten­tion these get.

    • Elec­toral risk, when it ex­ists (e.g. when a poli­ti­cal party in a democ­racy has a strong op­po­si­tion), cre­ates in­cen­tives that bring poli­ti­ci­ans’ per­cep­tions of policy costs/​benefits more in line with the per­cep­tions of vot­ers. This likely fa­cil­i­tates the ex­clu­sion of fu­ture gen­er­a­tions; its im­pact on other of­ten-ne­glected groups is less clear.

    • How easy/​hard a poli­ti­cal sys­tem makes it for ac­tors to veto a pro­posed policy change largely de­ter­mines the de­gree to which some in­cluded ac­tors can block a policy, if they per­ceive that a policy change will be net costly for them, while other in­cluded ac­tors con­sider it net benefi­cial.

  • The char­ac­ter­is­tics of in­cluded groups are some­times changed by in­ter­nal poli­ti­cal shifts, which may be caused by in­ter­nal changes (e.g. liberal rev­olu­tions, au­thor­i­tar­ian coups, civil wars, shift in vot­ers’ sup­port for var­i­ous poli­ti­cal par­ties) or for­eign in­ter­ven­tions.

  • In­ter­na­tional pres­sure takes two main forms: ne­go­ti­a­tion with in­ter­na­tional in­cen­tives, and bols­ter­ing the ex­cluded group’s ca­pac­ity/​will­ing­ness to re­sist (e.g. fund­ing re­sis­tance, or offer­ing free­dom to run­away slaves).

  • Em­pow­ered ac­tors can and of­ten do use their power to al­ter the above fac­tors to make the per­sis­tence of their preferred state more likely (e.g. abo­li­tion­ists’ ban on the slave trade made re-tran­si­tion­ing to slav­ery, af­ter eman­ci­pa­tion, much more costly; wealthy peo­ple’s pro­mo­tion of civil liber­ties in the early US seems to have helped them pro­tect prop­erty rights).

  • In­vest­ment can fa­vor the per­sis­tence of ex­clu­sion through eco­nomic in­vest­ment into prof­itable ex­ploita­tion, or through ide­olog­i­cal in­vest­ment into ex­clu­sive val­ues (e.g. us­ing ex­clu­sive val­ues as jus­tifi­ca­tions when challenged by other so­cieties). In­vest­ment can fa­vor the per­sis­tence of in­clu­sion through in­vest­ment (e.g. poli­ti­cal/​eco­nomic pro­jects) re­li­ant on strate­gic al­li­ances, or through ide­olog­i­cal in­vest­ment into in­clu­sive val­ues (e.g. in­ter­na­tional pro­mo­tion of in­clu­sion).

  • Cases of poli­ti­cal in­clu­sion could be (over-sim­plis­ti­cally) clas­sified as: in­clu­sion (will­ing or un­will­ing) due to the in­fluence of the ex­cluded, or in­clu­sion due to the benev­olent in­fluence of the in­cluded.

  • Ex­clu­sion may come mainly from busi­nesses that de­pended on ex­clu­sion. In these cases, in­clu­sion de­stroys or weak­ens or­ga­nized sup­port for ex­clu­sion, mak­ing it harder for ex­clu­sion to re-emerge. Similarly, a de­cline in civil liber­ties may de­stroy or­ga­nized hu­man­i­tar­ian in­ter­ests.

  • This model has im­por­tant limi­ta­tions, in­clud­ing these:

    • The fac­tors iden­ti­fied are not straight­for­wardly dis­junc­tive. This means that, when us­ing the model, one should be care­ful to avoid dou­ble-count­ing a sin­gle shift in in­cen­tives. For ex­am­ple, if the econ­omy changes so that slaves can more eas­ily de­stroy cap­i­tal, one might want to count this as in­creased ca­pac­ity for re­sis­tance, or as de­creased prof­ita­bil­ity of ex­ploita­tion, but not as both.

    • The model is qual­i­ta­tive; it does not make very pre­cise pre­dic­tions.

    • There is some ev­i­dence (e.g. as dis­cussed in the pre­vi­ous part’s sec­tion “Mo­ti­va­tions for Con­flict Over the Vot­ing Fran­chise”) against the as­sump­tion that eco­nomic mo­tives are much more in­fluen­tial than other ide­olog­i­cal mo­tives.

      • As fur­ther ev­i­dence, the in­fluence in US poli­tics (and per­haps in other coun­tries) of the Chris­tian right—which in­cludes efforts to re­strict ac­cess to birth con­trol and abor­tions—seems to be an ex­am­ple of cul­tural/​ide­olog­i­cal mo­tives be­ing very in­fluen­tial, even when peo­ple have eco­nomic (and other) rea­sons to fa­vor differ­ent poli­cies.

    • The model is not ex­haus­tive. There could eas­ily be more fac­tors that fa­vor in­clu­sion or ex­clu­sion, and differ­ent case stud­ies might make these more clear.

Strate­gic Im­pli­ca­tions:

And if all I can do is speak, it is for you I shall speak… the calamities that have no mouths

- Aimé Cé­saire, Note­book of a Re­turn to the Na­tive Land, 30

Im­pli­ca­tions for Mo­ral Cir­cle Ex­pan­sion:

In this sec­tion, I in­tro­duce ex­ist­ing the­o­ries about the ex­pan­sion of peo­ple’s cir­cles of con­cern—in­clu­sive moral progress—which have em­pha­sized the benev­olence of pow­er­ful ac­tors. Then, I ar­gue that these views over­state the im­por­tance of in­clu­sive val­ues, as benev­olence has only been one con­trib­u­tor among many to ma­jor his­tor­i­cal shifts to­ward greater in­clu­sion. I pro­pose a differ­ent the­ory of moral cir­cle ex­pan­sion, one which does not have op­ti­mistic im­pli­ca­tions for the fu­ture in­clu­sion of fu­ture gen­er­a­tions. I also re­spond to two po­ten­tial ob­jec­tions.

Philoso­phers and so­cial sci­en­tists have no­ticed and the­o­rized about his­tor­i­cal trends to­ward greater in­clu­sion, typ­i­cally em­pha­siz­ing benev­olent val­ues. Singer’s (1981) clas­sic work on this sub­ject, The Ex­pand­ing Cir­cle, notes: “The cir­cle of al­tru­ism has broad­ened from the fam­ily and tribe to the na­tion and race, and we are be­gin­ning to rec­og­nize that our obli­ga­tions ex­tend to all hu­man be­ings.” This, Singer ar­gues, is the nec­es­sary re­sult of hu­mans’ col­lec­tive rea­son­ing on ethics: hu­mans’ evolved de­sires for “a dis­in­ter­ested defense of one’s con­duct” mean that “[e]th­i­cal rea­son­ing, once be­gun, pushes against our ini­tially limited eth­i­cal hori­zons, lead­ing us always to­ward a more uni­ver­sal point of view.” This view sug­gests that in­clu­sive val­ues have been very im­por­tant, and that they will con­tinue ex­pand­ing their scope.

Psy­chol­o­gist Steven Pinker and philoso­pher Allen Buchanan offer ad­di­tional ex­pla­na­tions of progress to­ward in­clu­sion. In his book The Bet­ter An­gels of Our Na­ture, which draws ex­plic­itly from Singer’s ear­lier work, Pinker (2011) em­pha­sizes the im­por­tance of “moral dis­cov­ery” and En­light­en­ment hu­man­ism in ex­plain­ing the de­cline of slav­ery and despo­tism. (Pinker does give some credit to denser com­mu­ni­ca­tion net­works for fa­cil­i­tat­ing re­sis­tance against despots, but he mostly high­lights chang­ing val­ues and ideas.) Ap­proach­ing the is­sue with less sym­pa­thy to util­i­tar­i­anism, Buchanan’s book The Evolu­tion of Mo­ral Progress em­pha­sizes the im­por­tance of “safe and sta­ble ecolog­i­cal and so­cial cir­cum­stances,” which he ar­gues “are hos­pitable to cos­mopoli­tanism and en­courag­ing of in­clu­sive val­ues” (Brown­stein and Kelly, 2020). De­spite their differ­ences, Buchanan’s the­ory of ex­pand­ing cir­cles of con­cern makes the same sug­ges­tion as Singer and Pinker’s ex­pla­na­tions: poli­ti­cal changes to­ward greater in­clu­sion have been mainly caused by the chang­ing val­ues of pow­er­ful ac­tors.

With their em­pha­sis on evolv­ing so­cial val­ues, ex­ist­ing the­o­ries of moral cir­cle ex­pan­sion also sug­gest that, since in­clu­sion has ex­panded in the past, it is highly plau­si­ble—per­haps in­evitable—that in­clu­sion will con­tinue ex­pand­ing, un­til the in­ter­ests of cur­rently ne­glected groups such as fu­ture gen­er­a­tions have en­tered into de­ci­sion mak­ers’ con­sid­er­a­tion. Ap­ply­ing this idea to analo­gous con­cerns over an­i­mal welfare, re­search non­profit 80,000 Hours (Todd, 2017) writes:

Will the fu­ture be bet­ter? [...] moral con­cern for other be­ings seems to have in­creased over time — the ‘ex­pand­ing moral cir­cle’ — so we ex­pect that peo­ple in the fu­ture will have more con­cern for an­i­mal welfare

Some might take this to mean that peo­ple should be highly op­ti­mistic about civ­i­liza­tion’s “de­fault” tra­jec­tory, or that chang­ing so­cial val­ues to fa­vor in­clu­sion is a highly promis­ing way of helping ex­cluded groups such as fu­ture gen­er­a­tions.

How­ever, his­tor­i­cal case stud­ies of in­clu­sion cast doubt on these ar­gu­ments. As I ar­gue in the ear­lier case stud­ies and anal­y­sis, the ex­ist­ing or po­ten­tial in­fluence of the ex­cluded (what they did or would do with power) was usu­ally a ma­jor force driv­ing in­clu­sion. At least, this was the case for two ma­jor in­stances of in­clu­sive progress: the abo­li­tion of slav­ery, and ex­ten­sions of vot­ing rights. In­clu­sion so fre­quently re­quired pow­er­ful ac­tors to be co­erced, vi­o­lently re­placed, or offered benefits in ex­change for in­clu­sion that the­o­ries which fo­cus on benev­olence com­pletely miss most of the his­tor­i­cal mo­ti­va­tors of in­clu­sion.

The his­tor­i­cal im­por­tance of pres­sure from ex­cluded groups means ex­ist­ing the­o­ries of moral cir­cle ex­pan­sion are, at best, highly in­com­plete. We might ask Singer and Buchanan: if trends to­ward greater in­clu­sive val­ues are as his­tor­i­cally im­por­tant as they sug­gest, why did it so of­ten take re­bel­lions, ri­ots, and strate­gic al­li­ances to bring about in­clu­sion? Even if their the­o­ries about the causes of in­clu­sive val­ues are cor­rect, the philoso­phers over­state the his­tor­i­cal im­por­tance of in­clu­sive val­ues; benev­olence has only been one of many his­tor­i­cal mo­ti­va­tors of in­clu­sion.

The his­tor­i­cal case stud­ies re­viewed ear­lier more strongly sug­gest a new the­ory of moral cir­cle ex­pan­sion:

  • Var­i­ous fac­tors—es­pe­cially the re­sis­tance of the ex­cluded, the po­ten­tial for elites to ally with the ex­cluded, and per­haps some in­clu­sive val­ues—mo­ti­vate tran­si­tions to­ward greater poli­ti­cal in­clu­sion.

    • Th­ese tran­si­tions to­ward greater poli­ti­cal in­clu­sion tend to per­sist, largely be­cause poli­ti­cal power, once given, is uniquely difficult to take away.

  • After poli­ti­cal in­clu­sion oc­curs, so­cial val­ues change retroac­tively to be more in­clu­sive. There are sev­eral ways in which this might hap­pen:

    • The formerly ex­cluded group might use its new power to spread val­ues that are in­clu­sive to it­self, to avoid be­ing re-ex­cluded.

    • Peo­ple tend to in­ter­nal­ize their so­ciety’s norms into their own val­ues (Hen­rich, et al., 2005, 2010). After in­clu­sion oc­curs, peo­ple will be in­ter­nal­iz­ing more in­clu­sive norms.

    • In­fluen­tial or­ga­ni­za­tions might in­vest ide­olog­i­cally in the new in­clu­sion, mak­ing it difficult (psy­cholog­i­cally and for pub­lic re­la­tions) for them to op­pose it (e.g. a la­bor union might ad­vo­cate for bet­ter work­ing con­di­tions by ar­gu­ing that cur­rent con­di­tions re­sem­ble the evils of slav­ery, or a coun­try might jus­tify du­bi­ous in­ter­na­tional policy by ap­peal­ing to the su­pe­ri­or­ity of democ­racy).

    • The in­clu­sion of a group of­ten de­stroys busi­nesses that de­pended on ex­clu­sion, so, af­ter in­clu­sion, there are fewer or­ga­ni­za­tions mo­ti­vated to spread ex­clu­sive val­ues.

The first part of this hy­poth­e­sis—on the causes of in­clu­sion—is sup­ported by the his­tor­i­cal case stud­ies and model re­viewed ear­lier. The sec­ond part of this hy­poth­e­sis re­mains spec­u­la­tive; I have not con­sid­ered it in de­tail, but there are highly plau­si­ble mechanisms by which it could be the case. It also fits ac­tual his­tor­i­cal cases of in­clu­sion much bet­ter than hy­pothe­ses which (wrongly) claim that in­clu­sive val­ues gen­er­ally came be­fore poli­ti­cal in­clu­sion, and were its main cause.

The claim that changes in eth­i­cal ar­gu­ments are largely the effects—not the causes—of ma­jor poli­ti­cal shifts also has sig­nifi­cant sup­port from find­ings in moral psy­chol­ogy. Stud­ies sug­gest that, when we en­gage in moral rea­son­ing, it is usu­ally not to make up our minds, but to make up jus­tifi­ca­tions for what we have already de­cided (Haidt, 2007).

If this view is cor­rect, prospects for the in­clu­sion of fu­ture gen­er­a­tions and other to­tally pow­er­less groups are rel­a­tively dim. Most of the his­tor­i­cal mo­ti­va­tors of de­liber­ate in­clu­sion re­quired ex­cluded groups to be ca­pa­ble of ex­ert­ing in­fluence, so these past trends do not give us much rea­son to ex­pect that de­ci­sion mak­ers’ cir­cles of con­cern will con­tinue ex­pand­ing un­til they in­clude groups that lack this ca­pac­ity, such as fu­ture gen­er­a­tions:

  • Groups with no ca­pac­ity to ex­ert in­fluence can­not effec­tively re­sist ex­clu­sion.

  • Groups with no ca­pac­ity to ex­ert in­fluence can­not poli­ti­cally, mil­i­tar­ily, or eco­nom­i­cally benefit those who in­clude them; they are not ap­peal­ing al­lies.

  • In­ter-so­cietal pres­sure is limited; it can­not be mo­ti­vated by the prospect of a strate­gic al­li­ance that cuts across so­cieties, and it can­not work by boost­ing the re­sis­tance of the ex­cluded.

In these cases, poli­ti­cal ac­tors still have some po­ten­tial mo­ti­va­tors for in­clu­sion, but they have fewer: in­clu­sive val­ues, as well as in­ter-so­cietal pres­sure mo­ti­vated by in­clu­sive val­ues that ex­tend across so­cieties, or by de­sires to re­duce ri­vals’ prof­its. At the same time, po­ten­tial mo­ti­va­tors for ex­clu­sion are no weaker for groups that can­not ex­ert in­fluence.

The fol­low­ing pic­ture illus­trates this:

When a group has no ca­pac­ity to ex­ert in­fluence, many his­tor­i­cally im­por­tant mo­ti­va­tors of in­clu­sion can­not ex­ist. As a re­sult, the bal­ance of costs and benefits, as per­ceived by pow­er­ful ac­tors, is tilted—per­haps heav­ily—in fa­vor of ex­clu­sion.

One might ob­ject: if progress to­ward greater in­clu­sion has been caused mainly by the in­fluence of the ex­cluded, why has most in­clu­sive progress hap­pened over the last few cen­turies? This, very plau­si­bly, has been mostly be­cause eco­nomic growth—es­pe­cially in­dus­tri­al­iza­tion—has greatly in­creased the abil­ity of ex­cluded hu­mans to gain and hold on to poli­ti­cal in­clu­sion.

One way that in­dus­tri­al­iza­tion prob­a­bly fa­vors in­clu­sion is by strength­en­ing ex­cluded groups. In­dus­tri­al­iza­tion im­proves com­mu­ni­ca­tion tech­nolo­gies, and it drives in­creases in pop­u­la­tion den­sity. Th­ese make it eas­ier for ex­cluded groups to or­ga­nize re­sis­tance. Eco­nomic growth also seems to em­power some newly wealthy ac­tors to de­mand civil liber­ties, and these also make it eas­ier for other ex­cluded groups to or­ga­nize re­sis­tance (Ace­moglu, et al., 2005).

In­dus­tri­al­iza­tion may have also fa­vored in­clu­sion by caus­ing elites to be more vuln­er­a­ble to re­sis­tance, mak­ing ex­clu­sion less ap­peal­ing. Slaves can more eas­ily mess up in­dus­trial tasks than agri­cul­tural ones (Fenoaltea, 1984). Similarly, ri­ots of the dis­en­fran­chised do not threaten landown­ers’ wealth as much as they threaten the ma­chin­ery and hu­man cap­i­tal that makes up the wealth of in­dus­trial elites (Ace­moglu and Robin­son, 2005). In brief, we can ex­plain the timing of these mod­ern era in­creases in in­clu­sion through the im­pacts of in­dus­tri­al­iza­tion: mak­ing ex­cluded groups more able to re­sist ex­clu­sion, and mak­ing con­flict more painful for elites.

One might also ob­ject: if in­clu­sive val­ues are not very strong con­trib­u­tors to in­clu­sion, then why have there been many policy shifts that greatly benefited pow­er­less groups, in­clud­ing fu­ture gen­er­a­tions? Ac­tu­ally, other his­tor­i­cal events may ap­pear to have been suc­cesses for the in­clu­sion of voice­less groups, but they were mainly suc­cesses for groups that were already pow­er­ful. As ar­gued ear­lier, en­vi­ron­men­tal con­ser­va­tion and the early gov­er­nance of ge­netic en­g­ineer­ing (as well as chil­dren’s pro­tec­tion, to a lesser ex­tent) seem to have been largely done for the benefit of ex­ist­ing, pow­er­ful ac­tors.

Other poli­cies that may seem to have helped fu­ture gen­er­a­tions prob­a­bly did not do so for their sake. Ac­tion on cli­mate change has been highly limited, and the limited ac­tion that has been taken may have re­sulted from the great im­por­tance of cli­mate change for pre­sent (em­pow­ered) gen­er­a­tions. Other ma­jor suc­cesses of en­vi­ron­men­tal­ism have also offered (and em­pha­sized) sig­nifi­cant benefits for already pow­er­ful ac­tors: the reg­u­la­tion of di­rectly harm­ful pol­lu­tants, and ecolog­i­cal mod­ern­iza­tion (Combs, et al., 2020; Dryzek, 2002). In these cases, fu­ture gen­er­a­tions mainly got lucky, so these events are not strong ev­i­dence for a re­li­able trend of ever-broad­en­ing benev­olence fa­vor­ing the in­clu­sion of fu­ture gen­er­a­tions.

In brief, while pre­vi­ous the­o­ries of in­clu­sive progress have em­pha­sized in­clu­sive val­ues, case stud­ies of abo­li­tion­ism, de­moc­ra­ti­za­tion, and other in­stances of in­clu­sion sug­gest that in­clu­sive val­ues have only been one cause of in­clu­sion many, and that the in­fluence of ex­cluded groups has been cru­cial for most his­tor­i­cal mo­ti­va­tors of in­clu­sion. This sug­gests that in­clu­sive val­ues are not so his­tor­i­cally in­fluen­tial (per­haps so­cial val­ues changed to be­come more in­clu­sive, af­ter other fac­tors mo­ti­vated poli­ti­cal in­clu­sion). This also sug­gests that we should not ex­pect past trends to­ward greater in­clu­sion to con­tinue un­til pow­er­less groups like fu­ture gen­er­a­tions are also in­cluded. Un­der this un­der­stand­ing, in­clu­sive shifts have been fre­quent in the mod­ern era be­cause eco­nomic growth/​in­dus­tri­al­iza­tion strength­ened the re­sis­tance of ex­cluded groups, and other cases of poli­cies that benefited pow­er­less groups were mainly aimed at benefit­ing pow­er­ful groups.

Im­pli­ca­tions for Poli­ti­cal Strat­egy for Fu­ture Gen­er­a­tions:

The case stud­ies and anal­y­sis above sug­gest that cer­tain poli­ti­cal strate­gies are par­tic­u­larly promis­ing for benefit­ing fu­ture gen­er­a­tions. This is just one anal­y­sis, so its im­pli­ca­tions should be con­sid­ered to­gether with the strate­gic im­pli­ca­tions of other analy­ses, not adopted with­out scrutiny. While these sug­ges­tions fo­cus on fu­ture gen­er­a­tions, many are also highly ap­pli­ca­ble to ad­vo­cat­ing for other pow­er­less groups, such as non-hu­man an­i­mals.

  • On spe­cific is­sues that mat­ter a lot for fu­ture gen­er­a­tions, fo­cus on show­ing pow­er­ful ac­tors that it is in their own in­ter­ests to take ac­tions that hap­pen to greatly benefit fu­ture gen­er­a­tions.

As ar­gued above, his­tor­i­cal case stud­ies and my model based on them sug­gest that fu­ture gen­er­a­tions’ lack of in­fluence elimi­nates most of the usual mo­ti­va­tions for in­clu­sion. As a re­sult, com­pared to groups that have his­tor­i­cally gained poli­ti­cal in­clu­sion, the poli­ti­cal in­clu­sion of fu­ture gen­er­a­tions is sig­nifi­cantly less likely to oc­cur, and if it oc­curs, it is sig­nifi­cantly more likely to be re­versed. This means that last­ing poli­ti­cal in­clu­sion—ar­guably the most ro­bust and durable way to benefit a group for its own sake, at least within so­cieties that are already wealthy—will be very difficult to achieve for fu­ture gen­er­a­tions. While poli­cies that benefit fu­ture gen­er­a­tions are less likely to be passed for their sake, an­other strat­egy re­mains as promis­ing as ever: get­ting pow­er­ful ac­tors to pur­sue their own in­ter­ests in ways that help fu­ture gen­er­a­tions. This strat­egy seems to have been be­hind ma­jor suc­cesses in en­vi­ron­men­tal­ism and the gov­er­nance of ge­netic en­g­ineer­ing.

For fu­ture gen­er­a­tions, this ap­proach has im­por­tant limits: with­out in­sti­tu­tions that value fu­ture gen­er­a­tions for their own sake, the pro­tec­tion of fu­ture gen­er­a­tions will be much less ro­bust than it could be. Still, this broad ap­proach seems highly promis­ing for spe­cific is­sues that threaten fu­ture gen­er­a­tions, such as the ir­re­spon­si­ble gov­er­nance of de­ci­sion-mak­ing com­puter sys­tems. For ideas on how this ap­proach could be used for cre­at­ing in­sti­tu­tions that do value fu­ture gen­er­a­tions for their own sake, see the sug­ges­tion be­low about gain­ing the sup­port of busi­ness.

  • Pur­sue sup­port that is con­cen­trated at the na­tional level to have some sup­port on the in­ter­na­tional level, which is ex­tremely valuable.

If there is a lit­tle sup­port for a policy, and sig­nifi­cant op­po­si­tion against it, in many coun­tries, then the policy will prob­a­bly have vir­tu­ally no na­tions ad­vo­cat­ing for it on the in­ter­na­tional stage. If, in­stead, sup­port­ers of the policy are con­cen­trated enough to be able to shape the in­ter­na­tional policy of a few coun­tries, then the policy can have na­tions sup­port­ing it in­ter­na­tion­ally. In other words, the same to­tal amount of na­tional-level sup­port can mean differ­ent amounts of in­ter­na­tional sup­port, de­pend­ing on how it is dis­tributed. Op­ti­mally, sup­port is also not too con­cen­trated, since the benefits of in­creased sup­port fall sharply af­ter a policy gets enough sup­port to win poli­ti­cal bat­tles.

A similar dy­namic should work at smaller scales, e.g. the US Se­nate will have more sup­port­ers of a policy if these sup­port­ers are not too diffused across US states. His­tor­i­cally, af­ter the Bri­tish Em­pire banned its own slave trade, its strong op­po­si­tion to the slave trade on the in­ter­na­tional stage seems to have been mas­sively in­fluen­tial for the in­ter­na­tional abo­li­tion of the slave trade.

This con­sid­er­a­tion sug­gests that ad­vo­cates of the rep­re­sen­ta­tion of fu­ture gen­er­a­tions should con­cen­trate their re­sources on a few na­tions—fo­cus­ing on gain­ing in­ter­na­tional cham­pi­ons for fu­ture gen­er­a­tions—rather than spread­ing out their efforts. On the other hand, differ­ences in ad­vo­cates’ com­par­a­tive ad­van­tages, as well as diminish­ing marginal re­turns to effort, are rea­sons to not con­cen­trate re­sources to an ex­treme de­gree. This op­ti­mal bal­ance might look like build­ing on ex­ist­ing efforts in the UK and other coun­tries where there has been ac­tive sup­port for the rep­re­sen­ta­tion of fu­ture gen­er­a­tions. If the tac­tics of these fu­ture gen­er­a­tions ad­vo­cates and their level of philan­thropic fund­ing have not been in­formed by the very high ex­pected value of gain­ing an in­ter­na­tional cham­pion for fu­ture gen­er­a­tions, they should be.

  • When op­pos­ing a policy, em­pha­size down­sides that are very con­crete (easy to imag­ine), hor­rify­ing, and be­liev­able (highly plau­si­ble/​already pre­sent).

Highly suc­cess­ful move­ments have fre­quently em­pha­sized very con­crete, hor­rify­ing, and be­liev­able down­sides of the poli­cies they op­posed. Abo­li­tion­ists put up di­a­grams of slaves crammed into slave ships. En­vi­ron­men­tal­ists called ozone de­ple­tion a “hole in the ozone layer.” Early sup­port­ers of safety guidelines for ge­netic en­g­ineer­ing re­search em­pha­sized po­ten­tial catas­tro­phes that were already very plau­si­ble with ex­ist­ing tech­nolo­gies. Given how limited and im­por­tant peo­ple’s at­ten­tion is, it is highly plau­si­ble that these tac­tics are very helpful for helping peo­ple rec­og­nize the sig­nifi­cance of policy down­sides.

For ad­vo­cates of fu­ture gen­er­a­tions: In pub­lic ad­vo­cacy, peo­ple con­cerned with re­duc­ing ex­is­ten­tial risks should put more em­pha­sis on the large down­sides posed by ex­ist­ing or near fu­ture tech­nolo­gies (e.g. syn­thetic biol­ogy, AI), with de­tails and vi­su­als that are difficult to for­get, and less em­pha­sis on rel­a­tively ab­stract ar­gu­ments about risks to the very long-term fu­ture.

Ad­di­tional sug­ges­tions:

  • Or­ga­nize, and think of or­ga­nized in­ter­ests as the most im­por­tant ac­tors in poli­tics.

    • Just about ev­ery his­tor­i­cal case study re­viewed sup­ports this view.

  • Draw heav­ily on pres­ti­gious in­fluencers to shift norms within fields.

    • This seems to have worked very well for sci­en­tists who suc­cess­fully nor­mal­ized con­cern over risks from ge­netic en­g­ineer­ing re­search.

  • Treat at­ten­tion as a scarce and valuable re­source—it is. Direct it to­ward the is­sues you raise, the benefits of the poli­cies you pro­pose, and the costs of poli­cies you op­pose.

    • This view has sig­nifi­cant sup­port from psy­cholog­i­cal stud­ies (Ja­cobs, 2011), as well as the ap­par­ent his­tor­i­cal im­por­tance of events that fo­cused the pub­lic’s at­ten­tion on cer­tain costs (the Haitian Revolu­tion, the Bap­tist War, Love Canal, the Deep­wa­ter Hori­zon oil spill, var­i­ous crises that cat­alyzed de­moc­ra­ti­za­tion, etc).

  • Highly promis­ing ways of pro­tect­ing fu­ture gen­er­a­tions from au­thor­i­tar­i­anism in­clude pro­mot­ing in­ter­na­tional peace, pro­mot­ing eco­nomic sta­bil­ity, hav­ing rep­re­sen­ta­tives of the wider pub­lic gov­ern ac­cept­ably, and boost­ing the pub­lic’s abil­ity to re­sist emerg­ing forms of re­pres­sion.

    • The wider pub­lic los­ing sup­port for its rep­re­sen­ta­tives has been a fre­quent con­trib­u­tor to democ­racy shrink­ing or be­ing de­stroyed. The pub­lic’s abil­ity to or­ga­nize re­sis­tance has ar­guably been the main driver of de­moc­ra­ti­za­tion. Another fre­quent cause of re­ver­sions to au­thor­i­tar­i­anism has been coups or­ches­trated by elites who are dis­satis­fied with democ­racy, but promis­ing ways of pre­vent­ing this are less clear.

  • When op­por­tu­ni­ties for change are limited, bide your time—lay the ground­work for tak­ing fu­ture op­por­tu­ni­ties when they arise.

    • Many move­ments seem to have suc­ceeded with this tac­tic, in­clud­ing move­ments for Bri­tish abo­li­tion­ism, US civil rights, women’s suffrage, and de­coloniza­tion.

Im­pli­ca­tions for In­sti­tu­tional De­sign for Fu­ture Gen­er­a­tions:

The case stud­ies and anal­y­sis above sug­gest that cer­tain in­sti­tu­tional de­signs are par­tic­u­larly promis­ing for benefit­ing fu­ture gen­er­a­tions. As be­fore, these sug­ges­tions should be con­sid­ered along with other analy­ses, and many of these sug­ges­tions are also highly ap­pli­ca­ble to ad­vo­cat­ing for other pow­er­less groups, such as non-hu­man an­i­mals.

  • Take ad­van­tage of differ­ences in pa­tience—pur­sue poli­cies whose last­ing benefits and po­ten­tially con­cen­trated costs will not come for years or decades.

Many of the biggest suc­cesses by op­po­nents of slav­ery and cli­mate ac­tion ad­vo­cates came through the pas­sage of poli­cies whose last­ing benefits and po­ten­tially con­cen­trated costs would not kick in for years or longer. When it comes to poli­ti­cal agree­ments and trades, peo­ple con­cerned with fu­ture gen­er­a­tions have an ad­van­tage that should fa­vor suc­cess over the long-run: greater pa­tience. Poli­ti­cal short-ter­mism usu­ally works against fu­ture gen­er­a­tions, but it can work for fu­ture gen­er­a­tions if poli­ti­ci­ans’ and lob­by­ists’ con­cern with the short term keeps them from strongly op­pos­ing com­mit­ments to one day care about fu­ture gen­er­a­tions. This tac­tic should be es­pe­cially use­ful in poli­ti­cal trades: com­mit­ments to even­tu­ally pri­ori­tize the long term are cheap for short-ter­mists and very valuable for longter­mists.

For fu­ture gen­er­a­tions, this might look like ad­vo­cat­ing for poli­cies, such as com­mit­tees or funds for fu­ture gen­er­a­tions, that will not be im­ple­mented for a decade or more. The difficult part might then be to keep gov­ern­ments from re­vers­ing these com­mit­ments when the once-dis­tant im­ple­men­ta­tion date grows near. To pre­vent that, in­vest­ing many ac­tors in the policy seems very use­ful: offer­ing pres­ti­gious po­si­tions to in­fluen­tial ac­tors in ad­vance; get­ting poli­ti­ci­ans to pub­li­cly de­clare their ide­olog­i­cal com­mit­ments to the policy in ad­vance (as a way to con­grat­u­late them­selves on their then-re­cent ac­com­plish­ment); and cre­at­ing benefits (e.g. checks) for the pub­lic that start out small and grad­u­ally in­crease, while fea­tur­ing re­minders to re­cip­i­ents about how much the policy will even­tu­ally benefit them. More broadly, these com­mit­ments can be seen as a col­lec­tive ver­sion of com­mit­ments that im­pa­tient in­di­vi­d­u­als pre­fer to make, to deal with prefer­ences that are in­con­sis­tent over time.

  • Any Fu­ture Gen­er­a­tions in­sti­tu­tion should be ex­plic­itly man­dated to con­sider long-term pros­per­ity, in ad­di­tion to ex­is­ten­tial risks aris­ing from tech­nolog­i­cal de­vel­op­ment and en­vi­ron­men­tal sus­tain­abil­ity.

As ar­gued ear­lier, the eco­nomic mo­tives of busi­ness are typ­i­cally a ma­jor force work­ing against the in­clu­sion of fu­ture gen­er­a­tions. So far, poli­ti­cal in­sti­tu­tions rep­re­sent­ing fu­ture gen­er­a­tions have mostly fo­cused on en­vi­ron­men­tal sus­tain­abil­ity, and Jones et al. (2018) ar­gued com­pel­lingly that these in­sti­tu­tions should also be tasked with the miti­ga­tion of ex­is­ten­tial risks. Th­ese past and pre­sent pro­pos­als seem to be miss­ing a ma­jor op­por­tu­nity for helping the poli­ti­cal in­clu­sion of fu­ture gen­er­a­tions oc­cur and per­sist: ad­vo­cates of fu­ture gen­er­a­tions can last­ingly diminish the op­po­si­tion of busi­ness in­ter­ests—or turn it into sup­port—by de­sign­ing pro-fu­ture in­sti­tu­tions so that they visi­bly con­tribute to ar­eas where fu­ture gen­er­a­tions and far-sighted busi­nesses have com­mon in­ter­ests, such as long-term trends in in­fras­truc­ture, re­search and de­vel­op­ment, ed­u­ca­tion, and poli­ti­cal/​eco­nomic sta­bil­ity.

Stud­ies of sev­eral in­vest­ment trends sug­gest that busi­nesses do tend to value long-term benefits—they are not that im­pa­tient—so these long-term benefits do have the po­ten­tial to ap­peal to busi­ness in­ter­ests (Ja­cobs, 2011). Per­haps sur­pris­ingly, there are good rea­sons to think this strat­egy is es­pe­cially promis­ing in the US. The US poli­ti­cal sys­tem makes it rel­a­tively easy for at­tempts at mak­ing poli­cies to get blocked by op­po­si­tion, so when busi­ness in­ter­ests have the op­tion of benefit­ing through long-term in­vest­ment, they rarely have a more ap­peal­ing al­ter­na­tive of benefit­ing through re­dis­tribu­tive poli­cies. This ar­gu­ment is sup­ported by past trends in US policy in­vest­ment into pen­sions (Ja­cobs, 2011). So, es­pe­cially in the US and also in other na­tions, ad­vo­cates for fu­ture gen­er­a­tions will have much bet­ter prospects for last­ing suc­cess if they ally with busi­ness.

His­tor­i­cally, Bri­tish abo­li­tion­ism benefited greatly from de­sign­ing the 1806 par­tial ban on the slave trade so that it gained the sup­port of many slave own­ers and slave traders. Later, in­ter­na­tional abo­li­tion­ism benefited im­mensely from the self-in­ter­ested sup­port of Bri­tish plan­ta­tion own­ers (who wanted to keep their com­pe­ti­tion from get­ting slaves). Similarly, the 2010 US Se­nate at­tempt for cli­mate ac­tion, and the 2020 cage-free egg law in Colorado gained much sup­port—per­haps cru­cial sup­port—through this his­tor­i­cally suc­cess­ful tac­tic: mak­ing con­ces­sions (or even ugly trades) to win the sup­port of some of the usual op­po­si­tion. Ad­vo­cates of fu­ture gen­er­a­tions, I sug­gest, should try some­thing similar (af­ter more care­fully con­sid­er­ing down­side risks).

  • When you have power, do not aim to make the biggest im­prove­ments you can; aim to make im­prove­ments that are ac­cepted widely enough to per­sist, even af­ter you lose power.

Surges of poli­ti­cal sup­port rarely last, so any policy whose main­te­nance de­pends on one’s cur­rent level of power is un­likely to per­sist. Be­cause of this, dras­tic, sud­den policy changes are fre­quently re­versed in the span of a few years, es­pe­cially if they do not quickly cre­ate large con­stituen­cies in­vested in main­tain­ing the policy (e.g. as eman­ci­pa­tion does). For ex­am­ple, many vic­tors of rev­olu­tions wrote rad­i­cal con­sti­tu­tions that were quickly shred­ded by au­thor­i­tar­ian takeovers. Similarly, the in­sti­tu­tions that have most strongly rep­re­sented fu­ture gen­er­a­tions have been re­versed (Jones et al., 2018). When ad­vo­cates of fu­ture gen­er­a­tions have the power to shape policy, they should aim for the best poli­cies that will per­sist past the next ad­minis­tra­tion.

Ad­di­tional sug­ges­tions:

  • Iden­tify and pur­sue poli­cies that in­cen­tivize those cov­ered by them to sup­port their fur­ther ex­pan­sion.

    • Abo­li­tion­ism benefited greatly from the built-in ex­pan­sive­ness of slave trade bans: once plan­ta­tion own­ers could not im­port slaves, they sup­ported diplo­macy to keep their com­peti­tors from do­ing just that.

    • This seems most fea­si­ble when busi­nesses can­not eas­ily re­lo­cate and are com­pet­ing with busi­nesses in an­other na­tion, as well as when a larger net­work cov­ered by a policy in­creases the policy’s benefits for each per­son cov­ered.

  • De­sign and pur­sue poli­cies that by­pass typ­i­cal sources of ve­toes.

    • The Paris Agree­ment was rel­a­tively suc­cess­ful, in part be­cause it was de­signed to by­pass usual sources of dis­agree­ment, as well as the US Se­nate.

Other Im­pli­ca­tions:

In ad­di­tion to the im­pli­ca­tions dis­cussed above, which fo­cus on rep­re­sent­ing fu­ture gen­er­a­tions, this re­port’s case stud­ies have sig­nifi­cant im­pli­ca­tions for other ar­eas that many in the effec­tive al­tru­ism com­mu­nity are in­ter­ested in. As be­fore, this is just one anal­y­sis, so its im­pli­ca­tions should be con­sid­ered to­gether with crit­i­cisms and other analy­ses.

  • When con­sid­er­ing long-term risks of ex­treme suffer­ing, we should as­sign more weight to sce­nar­ios in which main­tain­ing some form of large-scale suffer­ing is in the in­ter­ests of pow­er­ful ac­tors, and less weight to sce­nar­ios in which it is not, since suffer­ing is more likely to per­sist in the former cases.

In sev­eral cases, such as the abo­li­tion of slav­ery in New England and much of Western Europe, the cre­ation of the first men­tal health hos­pi­tals, and the cre­ation of the first laws in the US against non-prof­itable cru­elty to chil­dren, re­form came very quickly (of­ten within a few months or years) af­ter or­ga­nized sup­port for re­form be­gan. In other cases, such as many cases of de­moc­ra­ti­za­tion or eman­ci­pa­tion—when re­form threat­ened wealthy in­ter­ests—in­clu­sive progress was much slower, of­ten tak­ing gen­er­a­tions. This pat­tern (and its straight­for­ward ex­pla­na­tion, that policy out­comes are largely de­ter­mined by the rel­a­tive in­fluence of or­ga­nized in­ter­ests sup­port­ing and op­pos­ing poli­cies) sug­gests that ex­treme suffer­ing is much more likely to per­sist in the long-term fu­ture if it ad­vances the in­ter­ests of pow­er­ful ac­tors.

This is a rea­son to put less weight on risks from fringe ac­tors be­ing sadis­tic or from hu­mans spread­ing wild an­i­mal suffer­ing (since in both of these cases, un­der the ques­tion­able as­sump­tion that poli­tics would be work­ing roughly as it works to­day, there would be stronger poli­ti­cal coal­i­tions op­pos­ing than sup­port­ing this suffer­ing), and more weight on risks from very pow­er­ful sadis­tic ac­tors, black­mail, and re­search stud­ies or cog­ni­tive la­bor in­volv­ing large-scale suffer­ing.

  • If it is not ac­com­panied by other de­vel­op­ments that push in differ­ent di­rec­tions, the au­toma­tion of hu­man la­bor at mas­sive scales will, by de­fault, severely erode democ­ra­cies globally.

While con­cerns about the im­pacts that ar­tifi­cial in­tel­li­gence will have on democ­racy have of­ten fo­cused on me­dia and surveillance tech­nolo­gies, these are not the only ma­jor dan­gers: the di­rect eco­nomic im­pacts of mas­sive-scale au­toma­tion might be enough to kill democ­racy. Stud­ies of de­moc­ra­ti­za­tion sug­gest that tran­si­tions to democ­racy are of­ten driven by the eco­nomic power of non-elites, and that these tran­si­tions tend to be much slower and more fre­quently re­versed when there are very high lev­els of wealth in­equal­ity (pre­sum­ably be­cause, then, elites have more to fear from gov­ern­ment re­dis­tri­bu­tion of wealth). On its own, then, the au­toma­tion of billions of jobs would mas­sively boost the few own­ers of cap­i­tal while sap­ping the eco­nomic power of scores of la­bor­ers, in­cen­tiviz­ing elites to de­stroy democ­ra­cies and mak­ing it more difficult for non-elites to bring democ­racy back.

For­tu­nately, sev­eral fac­tors may pre­vent such a de­cline of democ­racy. If au­toma­tion-based economies are highly vuln­er­a­ble to hack­ing or other forms of con­flict, elites may not be so will­ing to re­press calls for democ­racy. Alter­na­tively, if most elites do not care much for the differ­ence be­tween be­ing very wealthy and be­ing even more wealthy, they may be quite risk averse and there­fore want to avoid con­flict with non-elites who de­mand democ­racy. On the other hand, boosted surveillance and mil­i­tary tech­nolo­gies might make fears of rev­olu­tion have lit­tle rele­vance.

Some Po­ten­tial Ques­tions for Fur­ther Study:

  • What do other case stud­ies sug­gest about the hy­pothe­ses in this re­port, and what other hy­pothe­ses do they sug­gest?

    • It is not clear, for ex­am­ple, how well this model does at ex­plain­ing de­coloniza­tion, the US civil rights move­ment, or women’s rights move­ments.

  • In some cases, such as with some of the ac­tivi­ties of re­li­gious groups in US poli­tics, cul­tural con­cerns seem to over­ride eco­nomic con­cerns. Do they? On a similar vein, how im­por­tant has con­cern for fu­ture gen­er­a­tions been in driv­ing en­vi­ron­men­tal­ism?

    • If these are cases in which non-eco­nomic mo­tives were very in­fluen­tial, how can al­tru­ists cre­ate similarly im­pact­ful cul­tural con­cerns for fu­ture gen­er­a­tions or other ne­glected groups?

    • What does this im­ply for the as­sump­tion that poli­ti­cal ac­tors are mostly mo­ti­vated by eco­nomic self-in­ter­est?

  • In more depth than what was cov­ered here, what does ex­ist­ing re­search sug­gest about whether or not de­moc­ra­ti­za­tion causes wealth re­dis­tri­bu­tion?

    • If it does not tend to, how would it be best to ad­just this re­port’s model, which as­sumes that poli­ti­cal de­ci­sions are mostly mo­ti­vated by eco­nomic in­ter­ests?

  • What are the flaws in and ar­gu­ments against the recom­men­da­tions made above? If these are less strong than the ar­gu­ments in fa­vor of some recom­men­da­tion, what are con­crete ways in which peo­ple can im­ple­ment it?

    • What op­por­tu­ni­ties are there for poli­ti­cally ac­tive al­tru­ists (e.g. those work­ing to end the large-scale mis­treat­ment of non-hu­man an­i­mals) to make progress by push­ing for a policy that is in the in­ter­ests of some of the usual op­po­si­tion?


Fu­ture gen­er­a­tions mat­ter greatly, yet their well-be­ing is harm­fully ne­glected in to­day’s poli­ti­cal de­ci­sion mak­ing. What strate­gies for chang­ing this can ad­vo­cates of fu­ture gen­er­a­tions learn from his­tory? To gain in­sights into this ques­tion, this study ex­am­ined his­tor­i­cal case stud­ies of in­stances when policy change did much to benefit ex­cluded groups, mainly: abo­li­tion­ism, de­moc­ra­ti­za­tion, cli­mate ac­tion, and ge­netic en­g­ineer­ing gov­er­nance. As pre­dicted by a ra­tio­nal-choice frame­work, these case stud­ies sug­gest that sev­eral fac­tors make poli­ti­cal ex­clu­sion more likely to oc­cur and per­sist: op­por­tu­ni­ties for prof­itable ex­ploita­tion, costs of in­clu­sion, and ex­clu­sive val­ues. They also sug­gest that sev­eral fac­tors make poli­ti­cal in­clu­sion more likely to oc­cur and per­sist: a group’s ca­pac­ity for re­sis­tance, strate­gic al­li­ances, in­ter-so­cietal pres­sure, and in­clu­sive val­ues.

This has im­por­tant strate­gic im­pli­ca­tions for sup­port­ing fu­ture gen­er­a­tions and other voice­less groups. First, the drivers of in­clu­sion mostly de­pend on the in­fluence of the ex­cluded, or on ex­cluded groups get­ting lucky, so past trends to­ward greater in­clu­sion do not strongly im­ply that poli­ti­cal in­clu­sion will even­tu­ally ex­tend to voice­less groups. Per­haps in­clu­sive so­cial val­ues have mainly formed retroac­tively.

Ad­di­tional sug­ges­tions for ad­vo­cates of fu­ture gen­er­a­tions in­clude these: on spe­cific is­sues, fo­cus on show­ing pow­er­ful ac­tors how helping fu­ture gen­er­a­tions is in their own in­ter­est; con­cen­trate re­sources in a few na­tions; em­pha­size salient down­sides of poli­cies you op­pose; take ad­van­tage of differ­ences in pa­tience; man­date Fu­ture Gen­er­a­tions in­sti­tu­tions to con­sider long-term pros­per­ity; and pur­sue poli­cies with wider sup­port than the strongest poli­cies you could pass. Other im­pli­ca­tions are that risks of ex­treme suffer­ing are much higher if they in­volve suffer­ing that is in the in­ter­ests of pow­er­ful ac­tors, and that the di­rect im­pacts of au­toma­tion would, if not alle­vi­ated by other fac­tors, severely un­der­mine democ­racy.


This spread­sheet con­tains graphs, data, and elab­o­ra­tion (with sources) on de­mo­graphic trends in slav­ery over the past two cen­turies, globally and in the US.

Ap­pendix B—Abo­li­tion­ist Quotes that Sound Like Effec­tive Altru­ist Quotes:

Here is a list of these quotes.


Here is this re­port’s bibliog­ra­phy, or­ga­nized by top­ics.