Women’s Empowerment: Founders Pledge report and recommendations

See be­low for the ex­ec­u­tive sum­mary; the full re­port can be found on our web­site.

Com­ments are wel­come, and feel free to share the re­port wher­ever you think it could be helpful!

Up­date 14-02-2019 12.12 pm BST:

This is to thank you all once more for all your com­ments here, and to let you know they have been use­ful and we have in­cor­po­rated some changes to ac­count for them in a new ver­sion of the re­port, which will be pub­lished in March or April. They were also use­ful in our in­ter­nal dis­cus­sion on how to frame our re­search, and we plan to keep im­prov­ing our com­mu­ni­ca­tion around this through­out the rest of the year, e.g. by pub­lish­ing a blog post /​ brief on cause pri­ori­ti­sa­tion for our mem­bers.

One thing I still want to stress here to avoid mis­con­cep­tions: FP gen­er­ally chooses the ar­eas we re­search through cause pri­ori­ti­sa­tion /​ in a cause neu­tral way, and we do try to fully an­swer the ques­tion ‘how can we achieve the most good’ in the ar­eas we in­ves­ti­gate, not (even) shy­ing away from harder-to-mea­sure im­pact. In fact, we are mov­ing more and more in the lat­ter di­rec­tion, and are de­vel­op­ing re­search method­ol­ogy to do so (see e.g. our re­cently pub­lished method­ol­ogy brief on policy in­ter­ven­tions). Some of our re­ports so far, in­clud­ing this one, have been an ex­cep­tion to these rules for prag­matic (though im­pact-mo­ti­vated) rea­sons, mainly:

  • We quickly needed to build a large enough ‘ba­sic’ port­fo­lio of rel­a­tively high-im­pact char­i­ties, so that we could make good recom­men­da­tions to our mem­bers.

  • There are some causes our mem­bers ask lots of ques­tions about /​ are ex­tra in­ter­ested in, and we want to be able to say some­thing about those ar­eas, even if we in the end recom­mend them to fo­cus on other ar­eas in­stead, when we find bet­ter op­por­tu­ni­ties there.

But there’s definitely ways in which we can im­prove the fram­ing of these ex­cep­tions, and the com­ments pro­vided here have already been helpful in that way.

Up­date 21-12-2018 1.18 pm BST:

To provide some con­text (thanks Sie­beRozen­dal for the com­ment):

  • We se­lected to work on this par­tic­u­lar cause at least par­tially be­cause of a large in­ter­est in our mem­ber com­mu­nity.

    • This does not mean that the choice of writ­ing this re­port was a cause-par­tial choice: for Founders Pledge to do the most good we ob­vi­ously need to take our com­mu­nity’s prefer­ences into ac­count.

    • Nei­ther does it mean that one couldn’t ar­rive at women’s em­pow­er­ment as a high-po­ten­tial cause area through cause pri­ori­ti­sa­tion, given cer­tain val­ues.

  • The re­port isn’t writ­ten speci­fi­cally for an EA tar­get au­di­ence. How­ever, it’s find­ing should be of in­ter­est to a cause-neu­tral reader. In par­tic­u­lar, our re­search shows:

    • Band­han and Village En­ter­prise are both likely more cost-effec­tive than GiveDirectly, but not than GiveWell’s other recom­mended char­i­ties, tak­ing GiveWell’s value in­puts;

    • StrongMinds is likely more cost-effec­tive than GiveDirectly if looked at in terms of DALYs averted and with GiveWell’s value in­puts, though again not than GiveWell’s other recom­mended char­i­ties. (If looked at through the lense of im­prov­ing sub­jec­tive well-be­ing, it could ar­guably be at least as cost-effec­tive as many, if not all, GiveWell’s recom­men­da­tions.);

    • No Means No Wor­ld­wide is a harder case, as it re­quires value and em­piri­cal judge­ments with very large un­cer­tain­ties, plus sex­ual vi­o­lence is ob­vi­ously a very sen­si­tive area to dis­cuss. My cur­rent per­sonal best guess is that they are in the same bal­l­park as GiveDirectly, but I could be wrong by a large mar­gin in ei­ther di­rec­tion.

EXECUTIVE SUMMARY

The cause area

One hun­dred four coun­tries still have laws pre­vent­ing women from work­ing in spe­cific jobs; only 56% of women giv­ing birth in Africa de­liver in a health fa­cil­ity; and at least 35% of women wor­ld­wide have ex­pe­rienced some form of phys­i­cal or sex­ual vi­o­lence. Th­ese are just some of the challenges that women and girls around the globe face to­day.

In this re­port, we fo­cus on women’s em­pow­er­ment, by which we mean im­prov­ing the lives of women and girls. We re­searched char­ity pro­grammes to find those that most cost-effec­tively im­prove the lives of women and girls. As a heuris­tic for find­ing the most cost-effec­tive in­ter­ven­tions, we chose to fo­cus on pro­grammes aimed at low- and mid­dle-in­come coun­tries.

Our process

We used a top-down ap­proach to se­lect char­i­ties. First, we cat­e­gorised women’s em­pow­er­ment in low- and mid­dle-in­come coun­tries into twelve sub­fields. We then re­viewed liter­a­ture and in­ter­viewed twenty ex­perts in these sub­fields. This yielded a short­list of eleven promis­ing in­ter­ven­tions across sub­fields, in­clud­ing the grad­u­a­tion ap­proach to com­bat ex­treme poverty, em­pow­er­ment-self-defence courses to pre­vent sex­ual vi­o­lence, and in­ter­per­sonal group ther­apy to treat de­pres­sion.

With this short­list, we be­gan eval­u­at­ing char­i­ties. We started with a longlist of 163 women’s-em­pow­er­ment char­i­ties, and nar­rowed it down to a short­list of 15 char­i­ties based on our in­ter­ven­tion re­search and a quick scan of or­gani­sa­tional strength. We then com­pared the short­listed or­gani­sa­tions us­ing more de­tailed in­for­ma­tion on both cost-effec­tive­ness and strength of ev­i­dence. By our crite­ria, four char­i­ties es­pe­cially stood out. For each of those, we in­ves­ti­gated or­gani­sa­tional strength and plans, which led us to recom­mend three and pro­vi­sion­ally recom­mend the fourth.

We also recom­mend char­i­ties that are highly cost-effec­tive in im­prov­ing women’s lives but do not fo­cus ex­clu­sively on women’s em­pow­er­ment. We dis­cuss these or­gani­sa­tions, in­clud­ing those recom­mended by our re­search part­ner GiveWell, in other re­search re­ports on our web­site.

Char­ity recommendations

StrongMinds

What do they do? StrongMinds im­ple­ment In­ter­per­sonal Group Psy­chother­apy (IPT-G), train­ing laypeo­ple to treat women suffer­ing from de­pres­sion in Uganda.

Does the in­ter­ven­tion work? Ev­i­dence for the effi­cacy of IPT-G in low-re­source set­tings comes from two ran­domised con­trol­led tri­als (RCTs) and StrongMinds’s own quasi-ex­per­i­men­tal im­pact as­sess­ment.

Is the in­ter­ven­tion cost-effec­tive? We es­ti­mate that StrongMinds pre­vent the equiv­a­lent of one year of se­vere ma­jor de­pres­sive di­s­or­der for a woman at a cost of $200–$299, with a best guess es­ti­mate of $248.

What are the wider benefits? There are in­di­ca­tions of im­prove­ments in em­ploy­ment, nu­tri­tion, phys­i­cal health, hous­ing, and chil­dren’s ed­u­ca­tion.

Are they a strong or­gani­sa­tion? They have a good track record and a strong fo­cus on gen­er­at­ing ev­i­dence. They are trans­par­ent about their mis­takes and les­sons, and are com­mit­ted to con­tin­u­ous im­prove­ment.

Is there room for fund­ing? StrongMinds could pro­duc­tively use an ex­tra $5.1 mil­lion in fund­ing through 2020.

Band­han’s Tar­get­ing the Hard­core Poor programme

What do they do? As part of their Tar­get­ing the Hard­core Poor (THP) pro­gramme, Band­han provide women liv­ing in ex­treme poverty in In­dia with a pro­duc­tive as­set, a sav­ings ac­count, busi­ness train­ing, men­tor­ing, con­sump­tion sup­port, and in­for­ma­tion on ed­u­ca­tion and health. They also work with the In­dian gov­ern­ment and other NGOs to scale up their model.

Does the in­ter­ven­tion work? A high-qual­ity long-term RCT sup­ports the effec­tive­ness of Band­han’s THP pro­gramme. Ad­di­tional ev­i­dence gath­ered in differ­ent con­texts sug­gests that the ‘grad­u­a­tion ap­proach’ adopted by Band­han can effec­tively ad­dress ex­treme poverty.

Is the in­ter­ven­tion cost-effec­tive? We es­ti­mate that Band­han’s THP pro­gramme dou­bles a par­ti­ci­pant’s con­sump­tion for one year at a cost of $41–$134, with a best guess es­ti­mate of $62. This sug­gests that Band­han’s pro­gramme can bring about nom­i­nal gains in con­sump­tion of about $1.77 for each $1.00 donated. Ad­just­ing for pur­chas­ing power, this is equiv­a­lent to gains of $7.27 for each $1.00 donated.

What are the wider benefits? There is some ev­i­dence that the pro­gramme im­proves food se­cu­rity, phys­i­cal health, and sub­jec­tive well-be­ing.

Are they a strong or­gani­sa­tion? Band­han is a spe­cial­ised or­gani­sa­tion with a good track record. They are care­ful to main­tain high-qual­ity de­liv­ery of their pro­gramme; they are com­mit­ted to ev­i­dence; and they have been trans­par­ent through­out our anal­y­sis of their pro­gramme. One point for im­prove­ment, how­ever, is that their web­site lacks up-to-date in­for­ma­tion.

Is there room for fund­ing? The key im­ped­i­ment pre­vent­ing Band­han from scal­ing up is fund­ing, as they have all the re­quired in­fras­truc­ture and ca­pac­ity in place. Another $24 mil­lion would al­low them to reach an ad­di­tional 60,000 house­holds over the com­ing six years. For effi­ciency, we recom­mend a min­i­mum dona­tion to Band­han’s THP pro­gramme of $320,000.

Village Enterprise

What do they do? Village En­ter­prise provide busi­ness and fi­nan­cial-liter­acy train­ing, seed fund­ing, men­tor­ing, and ac­cess to busi­ness sav­ings groups to peo­ple liv­ing in ex­treme poverty in Sub-Sa­haran Africa.

Does the in­ter­ven­tion work? A re­cent high-qual­ity RCT pro­vides ev­i­dence that sup­ports Village En­ter­prise’s pro­gramme. There is also some ex­ter­nal ev­i­dence that the ‘grad­u­a­tion ap­proach’ on which Village En­ter­prise’s model is based effec­tively ad­dresses ex­treme poverty.

Is the in­ter­ven­tion cost-effec­tive? We es­ti­mate that Village En­ter­prise dou­ble a par­ti­ci­pant’s con­sump­tion for one year at a cost of $157–$367, with a best guess es­ti­mate of $250. This sug­gests that Village En­ter­prise’s pro­gramme can bring about nom­i­nal gains in con­sump­tion of about $0.99 for each $1.00 donated. Ad­just­ing for pur­chas­ing power, this is equiv­a­lent to gains of $2.18 for each $1.00 donated.

What are the wider benefits? There is some ev­i­dence that the pro­gramme im­proves sub­jec­tive well-be­ing.

Are they a strong or­gani­sa­tion? Village En­ter­prise are a strong or­ga­ni­za­tion, and rou­tinely ac­count for ev­i­dence and cost-effec­tive­ness in de­ci­sion-mak­ing. They have strong mon­i­tor­ing and learn­ing pro­cesses and are out­stand­ingly trans­par­ent and ac­countable.

Is there room for fund­ing? They could pro­duc­tively use an ex­tra $28 mil­lion in fund­ing through 2021.

No Means No Wor­ld­wide [pro­vi­sional]

What do they do? No Means No Wor­ld­wide (NMNW) train in­struc­tors to teach their ‘IMpower’ courses to both boys and girls, to help pre­vent sex­ual as­sault. They also work with large NGOs and gov­ern­ments to scale these courses up.

Does the in­ter­ven­tion work? Ev­i­dence sug­gests that NMNW’s IMpower in­ter­ven­tion re­duces the in­ci­dence of sex­ual vi­o­lence in sev­eral set­tings and for girls at differ­ent ages. This ev­i­dence comes mostly from two RCTs and two quasi-RCTs.

Is the in­ter­ven­tion cost-effec­tive? We es­ti­mate that NMNW pre­vent a sex­ual as­sault for $9–$757, with a best guess es­ti­mate of $62 per case averted.

What are the wider benefits? There is ev­i­dence that NMNW’s pro­gramme de­creases nega­tive gen­der at­ti­tudes among boys and re­duces rates of preg­nancy-re­lated school dropouts.

Are they a strong or­gani­sa­tion? NMNW are ex­cep­tion­ally com­mit­ted to gen­er­at­ing ev­i­dence; are trans­par­ent about their perfor­mance and mo­ti­va­tions; and have a good track record sup­port­ing IMpower im­ple­men­ta­tion.

Is there room for fund­ing? NMNW could pro­duc­tively use an ad­di­tional $7 mil­lion in fund­ing through 2021.

Why is our recom­men­da­tion pro­vi­sional? Based on the cur­rent ev­i­dence, we feel con­fi­dent recom­mend­ing NMNW to donors with a spe­cific in­ter­est in avert­ing sex­ual as­sault. Depend­ing on the re­sults of an in­de­pen­dent eval­u­a­tion of NMNW’s IMpower pro­gramme, which are due at the end of 2018, we may ei­ther recom­mend NMNW more gen­er­ally to donors in­ter­ested in women’s em­pow­er­ment; keep recom­mend­ing them only to donors in­ter­ested in avert­ing sex­ual as­sault; or de­cide not to recom­mend them