3. Summary of included studies.

Intervention category	Intervention type	Studies and study designs
Improve buying power	Unconditional cash transfers	18 RCTs: Ahmed 2019a; Ahmed 2019b; Asfaw 2014; Baird 2013a; Brugh 2018; Daidone 2014; Fenn 2015; Fernald 2011; Gangopadhyay 2015; Haushofer 2013; Hjelm 2017; Hoddinott 2013; Merttens 2013; Miller 2011; Pellerano 2014; Schwab 2013; Skoufias 2013; Tonguet Papucci 2015 3 prospective controlled studies: Aguero 2006; Breisinger 2018; Renzaho 2017
	Conditional cash transfers	9 RCTs: Baird 2013a; Evans 2014; Gertler 2000 (PROGRESA); Hidrobo 2014c; Kandpal 2016; Kurdi 2019; Kusuma 2017a; Macours 2012; Maluccio 2005 5 prospective controlled studies: Andersen 2015; Ferre 2014; Huerta 2006 (PROGRESA); Leroy 2008 (PROGRESA); Lopez Arana 2016
	Income generationd	6 RCTs: Beegle 2017; Darrouzet Nardi 2016; Marquis 2018; Olney 2016; Osei 2017; Verbowski 2018 11 prospective controlled studies: Alaofe 2016; Alaofe 2019; Asadullah 2015; Doocy 2017; Jodlowski 2016; Kangmennaang 2017; Katz 2001; Kennedy 1989; Murshed E Jahan 2011; Porter 2016e; Weinhardt 2017
Food prices	Food vouchers	4 RCTs: Fenn 2015b; Hidrobo 2014c; Jensen 2011; Ponce 2017 0 prospective controlled studies
	Food rebates/subsidies	1 RCT: Chen 2019 3 prospective controlled study: Andaleeb 2016; Chakrabarti 2018; Sturm 2013
Infrastructure changes	—	0 identified
Social environment	Village savings and loans	1 RCT: Kusuma 2017b 1 prospective controlled study: Brunie 2014

^aBaird 2013 assesses both conditional and unconditional cash transfers.
^bFenn 2015 assesses both unconditional cash transfers and food vouchers.
^cHidrobo 2014 assesses both conditional cash transfers and food vouchers.
^dThis includes different interventions that aimed to generate income of participants (e.g. integrated agricultural programmes, community development programmes).
^ePorter 2016 assessed a public works (80%) (cash/food‐for‐work) or unconditional cash transfer government programme (20%). Results were reported for the entire population, not disaggregated according to intervention received.

RCT: randomised controlled trial.