In a nutshell
Evidence Action's iron and folic acid (IFA) supplementation program provides technical assistance to increase the coverage of the Indian government’s IFA program, which entails weekly distribution of IFA supplements via schools and community health centers. Evidence Action's technical assistance includes procurement and program monitoring, staff training, and coverage surveys. Since 2018, GiveWell has made ~$14m in grants to support this technical assistance across multiple Indian states. (more)
This page provides a lookback on our 2022 grant. We’re focusing on this grant because our earlier grants were significantly disrupted by the COVID-19 pandemic, as school closures forced IFA programming to transition entirely to community-based delivery. To ease this transition, we agreed to shift the funding we had earmarked for monitoring and evaluation to support programming activities instead. Therefore, we have more information to evaluate our 2022 grant.
We’ve assessed this grant’s performance against initial expectations and identified key lessons learned. This shallow assessment fits into a broader cross-cutting assessment of our technical assistance grantmaking, which you can read about here.
Overall, we think this grant plausibly achieved the impact we expected and we’d probably make it again (more). We think this grant was plausibly around our cost-effectiveness threshold (10x), though with very wide confidence intervals (more). That’s because:
- Coverage surveys commissioned by Evidence Action suggest IFA coverage improved significantly in areas where Evidence Action provided technical assistance (TA). Taken at face-value, this data suggests that coverage increased by 34% between the beginning and two-year mark of the grant. Coverage estimates published by the government suggest a more moderate 17 percentage point improvement, and no noticeable difference between states where we funded TA and India as a whole. (more)
- When we visited this program in September 2024, government officials we spoke to generally spoke positively of Evidence Action’s support, and gave specific examples of bottlenecks they thought they’d helped alleviate. For example, they mentioned Evidence Action’s work in proactively flagging supply constraints in individual schools, which made stockouts less likely. We also commissioned IDinsight to meet separately with government officials without Evidence Action staff present. These meetings also generally yielded positive reviews of Evidence Action’s technical assistance. (more)
- Given the changes in IFA coverage before and after TA, we think these grants were plausibly around our cost-effectiveness threshold (~10x our benchmark). However, these estimates have wide confidence intervals and we haven’t prioritized making other updates to assumptions in our CEA. (more)
Grant activities are still ongoing, and we’re not sure how well coverage increases will be sustained once Evidence Action exits. Evidence Action's view is that program quality will drop after they exit, but still be in a better place than it would otherwise, because of systems they’ve put in place. We plan to monitor this.
Broader takeaways and lessons for our grantmaking: (more)
- Discrepancies between multiple data sources (in this case, data from Evidence Action and the government of India) highlight the importance of triangulating measurement approaches and understanding why they differ. This is something that’s come up in multiple TA grant lookbacks.
- We should have been clearer about milestones we expected this grant to hit before making the grant, to make looking back on these easier.
Published: January 2026
Background
Since 2018, GiveWell has directed ~$14 million to Evidence Action's iron and folic acid supplementation technical assistance program in India.1 The aim of this program is to support the government’s weekly provision of IFA tablets in schools and IFA syrup in community centers for pre-primary aged children.2 Provision of IFA tablets and syrup is mandated by the Indian government via their Anemia Mukt Bharat program. To help boost the coverage of this initiative, Evidence Action staff worked with health and education officials in targeted states to improve supply availability, train teachers and other frontline workers, and strengthen review mechanisms and reporting protocols.3
In 2020, Evidence Action’s work was significantly disrupted by the COVID-19 pandemic, which forced school closures across India.4 IFA supplementation shifted to community-based delivery, and we agreed to let Evidence Action shift the funding we had earmarked for monitoring and evaluation to program activities aimed to support this transition.5 Since we lack data on this period, this lookback focuses on our ~$9 million 2022 grant, when school-based IFA distribution had resumed. This grant sought to increase IFA coverage in five states: Haryana, Jharkhand, Madhya Pradesh, Rajasthan, and Uttarakhand.6
Would we have made these grants again, knowing what we know now?
Yes, though with important caveats about our confidence level. Our assessment suggests these grants likely achieved cost-effectiveness around our current funding threshold (~10x our benchmark) (more), based on reviewing updated coverage estimates (more).
If we were making similar grants today, we would probably commission our own independent coverage surveys (rather than outsourcing this to our grantee) and make clearer predictions about key milestones we expected the grant to achieve and when (more).
How did implementation go?
Coverage and prevalence estimates
Coverage
For our 2022 grant, we earmarked funding for Evidence Action to commission independent coverage surveys.7 They did this by running an open tender and selecting a data collection firm in each of the states where their IFA program operates.8 Aggregating across states and age groups, this data implies a 34 percentage point increase in IFA coverage two years after Evidence Action’s assistance began.
The government of India also publishes its own IFA coverage estimates as part of the performance indicators of its Anemia Mukt Bharat program. This data suggests a more moderate 17 percentage point coverage improvement across comparable time periods.9
Notes: Government data was pulled from the Anemia Mukt Bharat dashboard on August 19th 2025. In our experience, historic data on this platform sometimes changes, so this chart may not be replicable using data pulled at a later date. We don’t have permission to publish state-level estimates.
To dig into potential reasons for this discrepancy, we spoke to several people with familiarity with the government estimates, including people involved with IFA programming in India and several government officials. All of these people generally expressed negative views on the quality of the government data. We also noticed that the data changed when we pulled it in 2024 vs. 2025, which seems like an additional flag on data quality.10
Unfortunately, we can’t evaluate how these coverage improvements compared to what we predicted at the time of making the grant, as we didn’t structure our cost-effectiveness model in a way that made these predictions explicit.11 We think this was a mistake in hindsight, and plan to always include predictions about coverage changes in our cost-effectiveness models for TA programs going forward. (more)
Generally speaking, we view these coverage results as a positive update, but we’re cautious about putting too much weight on them. Importantly, Evidence Action’s TA was not randomized, as they deliberately targeted states with high rates of anemia to begin with.12 Coverage in these states may have improved faster in the absence of TA as the government may have been incentivized to devote more resources to them.
To probe this, we looked at how coverage changed in the five states Evidence Action supports compared to coverage across India generally, using data on IFA coverage that the government collects. Overall, these data suggest that coverage in states without TA increased at a similar rate as coverage in states with TA.13 We interpret this as a negative update on this grant having counterfactual impact, though we don’t put much weight on this due to concerns about data quality and non-randomized treatment.
Prevalence
Unlike for Evidence Action’s deworming program, we don’t have measures of anemia (the downstream health outcome IFA targets) before and after Evidence Action’s TA.14 The government of India plans to publish findings from a diet and biomarkers survey in 2025 or 2026, which includes estimates of anemia prevalence rates across states.15 We plan to use the outcome of this survey to triangulate our assumptions about coverage increases.
Discussions with local stakeholders
From September 2 to September 6, 2024, we visited with Evidence Action staff and government officials in India as part of a site visit to inform this lookback. We spent three days in Delhi speaking to Evidence Action staff involved with the running of the program, and two days in Uttarakhand (one of the states where we support TA), where we spoke with officials at various levels of the Indian government (block, district, state).16 In the Uttarakhand Ministry of Health, we spoke to the civil servant directly responsible for the implementation of IFA in schools, with Evidence Action staff members present.
These officials generally spoke positively of Evidence Action’s work, and cited similar reasons to those given for their deworming work.17 In particular, they mentioned Evidence Action’s help in keeping tabs on the expiry of procurement contracts, their help with developing standard operating procedures, training teachers, and holding block officials accountable for submitting coverage measures into the Health Ministries electronic commodity tracking system (eLMIS).18
They also mentioned other ways that Evidence Action provided support specific to IFA:19
- Procurement: Evidence Action reviewed procurement contracts and suggested protective clauses for the government. Though these haven’t been activated yet, they may protect the government in the future if a supplier doesn’t fulfil their obligations.
- Operational support: Unlike National Deworming Day, which only takes place once or twice a year, IFA supplementation happens weekly, which demands more constant attention on school stockouts, reporting, and adherence to operating procedure. The head of IFA programming for Uttarakhand stressed the usefulness of gap-filling; even a relatively straightforward program like providing IFA in schools can entail a lot of moving parts between multiple layers and silos of government, and having a trusted partner on-hand to catch delays or gaps can be helpful. The official responsible for IFA is also responsible for deworming and ~20 other health programs, and they said that Evidence Action’s TA had freed up her time to focus on other programs in her portfolio.
It’s possible that the officials we spoke to felt pressured to report positively on the program as Evidence Action staff members were present for these meetings. We also commissioned IDinsight, a research organization not involved in the delivery of the program, to speak separately with government officials without Evidence Action staff being present.20 These meetings also generally yielded positive reviews, including:
- Supply chain management: One government official described the importance of Evidence Action’s monthly reports flagging shortages so that supplies could be reallocated effectively.21
- Gap-filling: Another government official noted that Evidence Action’s assistance led to a greater focus on the IFA program, allowing it to maintain its coverage.
We also view these reviews as a positive update, though acknowledge that there may be incentives to exaggerate the impact of programs to surveys done on behalf of donors.
How cost-effective do we think these grants were?
Based on a retrospective cost-effectiveness model, we think these grants were plausibly around our current cost-effectiveness threshold (10x), though with wide uncertainty intervals.
| Cost-effectiveness we modeled at the time | Retrospective cost-effectiveness estimates | |||
|---|---|---|---|---|
|
25th percentile 22 |
Best-guess | 75th percentile | ||
| Evidence Action IFA |
12x 23 |
2x | 8x | 16x |
To estimate this, we put our 2022 cost-effectiveness model for IFA into a new template we’ve built for modeling TA grants. Key assumptions in this model include:
- Effect of program on coverage: We put some weight on data reported by Evidence Action and some weight on data reported by the government
- How coverage would have changed: We make rough forecasts about how IFA coverage would have changed in the absence of TA. To estimate counterfactual impact, we take the ‘wedge’ between: i) our estimated effect of the program on coverage and ii) how we expect coverage would have changed otherwise. These assumptions are illustrated below.
- Program costs: We based cost estimates on our 2022 ~$9m grant. We decided to confine this lookback to our 2022 grant for reasons discussed here.
We did not revisit the downstream assumptions in our IFA model – e.g., anemia burden in India, or the effect of IFA supplementation on anemia.
We separately estimated retrospective cost-effectiveness of this grant as part of our renewal investigation. That retrospective analysis estimated this grant as 19x.24 This difference is partly explained by updated ‘downstream’ assumptions – e.g., in our 2025 CEA, we’re estimating a smaller reduction in the burden of anemia year-on-year (which inflates cost-effectiveness),25 and a smaller effect of IFA on the anemia burden (which deflates cost-effectiveness).26 We haven’t used our updated downstream assumptions in this model to keep things consistent with the other TA lookbacks. This difference is also partly driven by different investigators on our team making different assumptions about the effect of this grant on counterfactual coverage. This highlights the degree of uncertainty in our estimates and the possibility of interpreting the M&E data in different ways.
Did we set ourselves up to learn?
We should have been more explicit about milestones we expected this grant to hit, and made explicit assumptions about how we expected coverage to change in our cost-effectiveness model. We plan to do both more consistently in future.
How calibrated were our forecasts?
We made 5 forecasts related to this grant, all of which were realized as ‘yes’. Some of these forecasts related to project milestones being hit; others related to subsequent GiveWell grant decisions.
| Forecast | Confidence (%) | Realization date | What happened? |
|---|---|---|---|
| EvAc will have completed a baseline survey. | 80% | 12/31/2022 | Yes |
| EvAc will have completed an endline survey. | 80% | 12/31/2024 | Yes |
| The difference in coverage between baseline and endline survey will be >12% (for reference, EvAc’s estimate of their effect on coverage is 14% after 2 years). | 60% | 3/31/2025 | Yes |
| Our best guess of EvAc IFA's cost-effectiveness will be above 8x our benchmark in all five states in India that it is currently operating in. | 60% | 6/30/2025 | Yes |
| We will have recommended funding for geographic expansion of the program in at least one location (such as additional states in India, Nigeria, or another country). | 60% | 6/30/2025 | Yes |
Sources
- 1
Our list of previous grants to this program can be found here.
- 2
For more on Evidence Action’s IFA program, see this section of our 2022 grant page.
- 3
Source: GiveWell’s conversations with Indian government and Evidence Action staff, September 2024 (unpublished)
- 4
Source: GiveWell’s conversations with Indian government and Evidence Action staff, September 2024 (unpublished)
- 5
Source: GiveWell’s conversations Evidence Action staff (unpublished)
- 6
See this section of our 2022 grant page for more on the activities that this grant funded.
- 7
See this section of our 2022 grant page for more on our monitoring and evaluation plan.
- 8
Source: GiveWell’s conversations Evidence Action staff (unpublished)
- 9
Source: GiveWell’s analysis of Indian government IFA data (unpublished)
- 10
Source: GiveWell’s analysis of Indian government IFA data (unpublished)
- 11
The cost-effectiveness analysis from our 2022 grant page can be found here.
- 12
Source: GiveWell’s conversations Evidence Action staff (unpublished)
- 13
Based on data from the government of India (Anemia Mukt Bharat), we estimate that IFA coverage in states where Evidence Action provided TA increased by 19 percentage points from 57% in 2022 to 74% in 2024 while in other states coverage increased by 19 percentage points from 43% to 62%.
Source: GiveWell’s analysis of Indian government IFA data (unpublished) - 14
See this section of our lookback at a grant to Evidence Action’s Deworm the World program in India.
- 15
Source: GiveWell’s conversations with Indian government and Evidence Action staff, September 2024 (unpublished)
- 16
India is organized into states/union territories which are broken into districts then sub-districts then blocks then villages/urban wards. A state (e.g., Uttarakhand) is the primary sub-national government unit. A district is the next tier, overseen by a District Magistrate/Collector. Below that, states use various sub-district terms (e.g., tehsil/taluk), and—especially for health, education, and rural development delivery—community development “blocks” (administrative clusters of villages/Gram Panchayats) led by a Block Development Officer. Programs often plan and supervise implementation at the block level and aggregate reporting at the district and state levels. Source: The Indian government’s census website
- 17
See this section of our lookback at a grant to Evidence Action’s Deworm the World program in India.
- 18
Source: GiveWell’s conversations with Indian government and Evidence Action staff, September 2024 (unpublished)
- 19
Source: GiveWell’s conversations with Indian government and Evidence Action staff, September 2024 (unpublished)
- 20
For more on this grant, see our page here.
- 21
Source: IDinsight conversations with Indian government officials (unpublished).
- 22In our CEA, we’ve modeled three scenarios with different assumptions about the impact of Evidence Action’s program on coverage:
Our best guess of the effect
A 25th percentile guess, representing what we consider a conservative estimate of program impact
A 75th percentile guess, representing what we consider an optimistic estimate of program impact
- 23See our 2022 IFA CEA, which represents our thinking at the time of making the grant, here.
- 24
At the time of publishing this lookback, we have not yet published a page on our renewal of Evidence Action’s IFA program in India. This cost-effectiveness estimate comes from the CEA associated with that investigation and will be published when the grant page is made public.
- 25
In our 2025 model, we’re assuming that the anemia burden in these locations declines by 1% a year (as measured in years lost to disability (YLDs), in-line with past trends. In the 2022 model, we were assuming a decline of 4% a year.
- 26
In our 2025 model (unpublished), we’re assuming IFA reduces the prevalence of anemia by 28%. In our 2022 model, we’re assuming IFA reduces the prevalence of anemia by 48% (this estimate was not made explicit when we initially modeled this program’s cost-effectiveness, but we’ve calculated an implied figure as part of this lookback).