GiveDirectly Comprehensive Summary

Last updated: December, 2024

Problem

Poverty is strongly associated with lower subjective wellbeing (Sacks et al., 2010; Tan et al. 2020; OWID, 2022).

Intervention

Cash transfers provide money directly to individuals in need. They have been shown to improve various outcomes, including: material poverty (Kabeer & Waddington, 2015), human capital (Baird et al., 2013b; Millán et al., 2019), social capital (Owusu-Addo et al., 2018), health ​(Lagarde et al., 2007; Behrman et al., 2011; Crea et al., 2015), intimate partner violence (Baranov et al., 2021; Buller et al., 2018), child labour (Kabeer & Waddington, 2015), the spread of HIV (Pettifor et al., 2016), spending on tobacco and alcohol (Evans & Popova, 2014; Handa et al., 2018), and labour supply (Baird et al., 2018; Banerjee et al., 2017).

Organisation

GiveDirecty’s core programme provides $1,000 lump-sum payments to people living in poverty, many of whom live in sub-Saharan Africa.

Evaluation

Methods

To estimate the impact of GiveDirectly, we aggregated all randomised controlled trial (RCT) evidence from lump-sum cash transfers we found in our previous systematic review and meta-analysis¹There we searched for all RCTs and quasi-experimental studies – including peer-reviewed publications and grey literature (e.g., reports, pre-prints, and working papers) – conducted from 2000-2020 examining the impact of cash transfers on self-reported subjective wellbeing and mental health outcomes. (RCTs = 10, natural experiments = 2, participants = 24,027; McGuire et al., 2022a). The studies were mostly of a very high quality.

The measures of wellbeing included in our meta-analysis contained self-reported measures of life satisfaction, happiness, and affective mental health (MHa).

We estimated the total effect of GiveDirectly’s lump-sum cash transfers over time using a meta-regression model that assumes the benefits dissipate linearly through time. 

Because improving one person’s wealth and wellbeing can benefit others who are close to that person, we also estimated the spillover effects on household members. Thus, the total effect was the impact on recipients and household members over time.

Impact

Our model suggests that $1,000 cash transfers have a total effect over time of 2.10 WELLBYs on the recipient. We estimate that recipients live with four other household members on average, and that each receive 86% of this benefit, resulting in 7.12 WELLBYs for the other household members combined (9.22 WELLBYs for the whole household).

Cost

The estimated costs for GiveDirectly are based on an average of their reported expenses from 2012 to 2019, weighing recent years much more heavily.

We estimate that the total cost to GiveDirectly to provide each $1,000 transfer is $1,221. The additional $221 comes from the overhead costs of running the organisation (salary for staff, office lease, etc.).

Cost-effectiveness

The cost-effectiveness of GiveDirectly is $132 per WELLBY. This means for every $1,000 donated, the organisation creates 7.55 WELLBYs.

To quantify the uncertainty in our results, we simulated our data thousands of times varying key parameters to see how the results might differ across a range of reasonable inputs (e.g., Monte Carlo simulations)²Monte Carlo simulations allow us to treat inputs in a cost-effectiveness analysis (CEA) – often merely stated as point estimates – as distribution. Thereby, this allows us to communicate a range of probable values (i.e., uncertainty around the point estimates). See our cost-effectiveness methodology for more detail.. This is illustrated in Figure 1 below.

Figure 1. Density plot of the quantified uncertainty around GiveDirectly’s cost-effectiveness

Note. The diamond represents the central estimate of cost-effectiveness. The shaded area is a probability density distribution and the whiskers represent the 95% confidence interval.

Quality of evidence

Our quality of evidence assessment is stringent. We assess quality of evidence according to an adapted version of the ‘GRADE’ criteria, a widely-used and rigorous tool for assessing evidence quality across healthcare and research fields. The GRADE criteria for evidence quality are very stringent, so we expect very few interventions that we evaluate for wellbeing in LMICs (which tend to be less well-studied) will score more than ‘moderate’ on the quality of their evidence. Considering most decisions about charities are made with little-to-no evidence, this is a substantial improvement.

We think the evidence supporting the effect of GiveDirectly is of high quality. The primary reasons for this are:

• There is a sizeable number of high-quality, GiveDirectly-specific RCTs studying the programme as it operates at scale.
• There are a large number of additional studies on cash transfer programmes, making for a large evidence base.
• The general evidence for cash transfers shows no clear signs of publication bias.
• We also found five cash transfer studies that captured household spillover effects, which decreases our uncertainty in this influential parameter for cash transfers.
• Many studies contain both subjective wellbeing and affective mental health outcomes, so we have good data on each and are confident that the effects do not vary substantially between outcome types.

Depth of our analysis

In-depth. We believe that we have reviewed all of the available evidence. We would be surprised if there were important studies that we weren’t aware of. We have spent considerable time on our analysis. Our meta-analysis of the effect of cash transfers on subjective wellbeing has undergone academic peer review and was published in Nature Human Behaviour.

Funding need

GiveWell estimated that GiveDirectly could absorb $456 million between 2021 and 2023. We think that this could be a lower bound for future funding needs over the next three years³From the GiveDirectly website: “With the successes in our remote targeting and payment technologies in a variety of settings, emerging relationships with key telecom partners in Africa, and current operational capabilities, we have the technical capacity to reach more than 5M people per year. This translates to the potential ability to deliver in excess of $ 5B per year to those living in extreme poverty today.”. Because of GiveDirectly’s low overhead costs, and primarily digital deployment mechanism, it appears unusually capable of scaling. It does not appear to have a clear limit on the amount of funding it could reasonably absorb.

Conclusion

GiveDirectly has strong evidence to support its effectiveness. However, we estimate that its programme is several times less cost-effective at increasing wellbeing than other charities we have evaluated, so it is not one of our recommended charities to donate to. However, it could be a good fit for donors who value the certainty of the evidence, the simplicity of the intervention, or the autonomy it provides to recipients.

FAQs

• Do you include community spillovers?
  • We are aware of new evidence about community spillovers which has updated the cost-effectiveness evaluation of GiveDirectly by other evaluators. We are yet to consider this evidence and whether it will lead to an update of our numbers. Note that our analysis of GiveDirectly is different from other evaluators in that it uses a wellbeing framework, and it already included many qualities that others might not yet have included such as a meta-analysis of many studies, longterm effects, and empirical household spillovers (McGuire et al., 2022b; McGuire et al., 2022e).
• You mention results in WELLBYs for GiveDirectly, but your reports do not mention results in WELLBYs?
  • Our reports on cash transfers and GiveDirectly are older than when we started systematically using WELLBYs as our outcome of interest. To use WELLBYs we convert our SD-year results by multiplying them by the typical standard deviation on a wellbeing scale, which we estimate to be 2 SDs. Before 2024, this number was 2.17, but we updated it after careful consideration, which has reduced the effect of every charity, including GiveDirectly (from 8 WELLBYs per $1,000 to 7.55 WELLBYs per $1,000). See our cost-effectiveness methodology for more detail.
• What’s the relationship between HLI and GiveDirectly?
  • HLI and GiveDirectly are completely independent organisations. We are not paid by GiveDirectly, and would never accept money to recommend any organisation.

Is your question missing from this list? Contact us at hello@happierlivesinstitute.org.

Our reports so far

We published a meta-analysis of the effect of cash transfers on subjective wellbeing in Nature Human Behaviour; as far as we know, this was the first meta-analysis of any intervention measured in terms of subjective wellbeing. We assessed the cost-effectiveness of GiveDirectly in our cash transfers cost-effectiveness analysis. We compared the cost-effectiveness of psychotherapy/StrongMinds to cash transfers/GiveDirectly in our psychotherapy vs cash transfer comparison. Most recently, we updated our analysis to include the impact of household spillovers in our household spillover report.