Nutrition International — Rice Fortification in India (September 2024)

Note: This page summarizes the rationale behind a GiveWell grant to Nutrition International. Nutrition International staff reviewed this page prior to publication.

Published: July 2025

Summary

Background

Iron deficiency is the most common cause of anemia, a condition in which the production of hemoglobin, a protein in red blood cells that helps transport oxygen, is diminished.1 There is some evidence that fortifying rice with iron can reduce the prevalence of iron deficiency and anemia.2 Rice can be fortified with iron by creating fortified rice kernels (FRK) made of rice flour and fortified premix, and then mixing them with raw rice kernels.

Nutrition International is an international non-governmental organization that provides technical assistance (TA) to governments in delivering nutrition-related interventions, including rice fortification. GiveWell has previously made grants to Nutrition International to support its Vitamin A supplementation campaigns in Chad. This is the first grant we’ve made to support its fortification work.

Nutrition International has worked on rice fortification in Madhya Pradesh with support from Global Affairs Canada3 and the Bill and Melinda Gates Foundation.4 This is the first time it will work on rice fortification in West Bengal.

What we think this grant will do

With this grant, Nutrition International will:

1. Provide technical assistance (TA) to improve the coverage of quality fortified rice that meets the minimum fortification standard,5 thereby increasing the amount of iron, as well as folic acid and vitamin B12, in fortified rice that is distributed through public distribution systems (PDS) in two Indian states, West Bengal and Madhya Pradesh.
2. Commission IDinsight to carry out an impact evaluation of its program in both West Bengal and Madhya Pradesh. IDinsight will collect data on (i) coverage of adequately fortified rice and (ii) average iron content of the rice before program implementation. It will then collect this data two years later to estimate how much both metrics have improved.

Why we made this grant

Our best guess is that this grant is 11x as cost-effective as unconditional cash transfers (GiveWell’s benchmark for comparing programs).

We detail why we think this grant is cost-effective below. In simple terms, we think this grant is cost-effective because:

• Iron deficiency burden is high in India. We expect the program to increase NaFeEDTA-equivalent iron intake by 0.3mg per week per targeted person.6 Because iron deficiency burden is high, we expect even small improvements to iron intake will reduce the population’s anemia-related morbidity (~45% of program benefits), improve long-term income of children benefiting from the program (~45% of benefits), decrease neonatal mortality (~5% of benefits) and improve income of anemic population working in physically demanding jobs (~5% of benefits).
• The program reaches many people at a low cost per person. The program involves TA to the government to improve rice fortification programs targeting a large number of people (roughly 300 million people over three years) at a relatively low cost. This translates to an additional cost of about $0.01 per person reached each year by the program.

A sketch of our cost-effectiveness analysis for the grant is below, using West Bengal as an example:

What we are estimating	Best guess (rounded)	Confidence intervals (25th - 75th percentile)	Implied cost-effectiveness
Donation to Nutrition International	$2,430,969
Anemia morbidity averted
Annual cost per person	$0.01
Number of people reached over three years	170,500,000	85,000,000-260,000,000	5x - 17.1x
Increase in coverage of rice fortified above minimum standard over the counterfactual, average over 3 years	8%	2%-14%	2.3x - 20.9x
Total additional iron intake per individual per week, NaFeEDTA-equivalent (mg)	0.30
Annual anemia years lived with disability (YLDs) in this population in absence of an iron fortification program	2,100,000
Reduction in anemia prevalence from iron fortification trials	29%
Internal and external validity adjustments	-98%
Adjusted reduction in anemia from iron fortification	0.6%	0.2%-1%	6.5x - 15x
Cost-effectiveness estimate from morbidity averted
Annual anemia YLDs (all ages) averted by iron fortification	13,000
Moral weight for each YLD averted	2.3
Cost-effectiveness estimate	3.7x
Development effects
Historical estimate from the United States of long-run annual income increase among iron deficient population, one year of fortification	0.32%
Adjustment for differences in iron increase between historical US study and current program settings	-96%
Replicability adjustment	-60%	-90% - -30%	7x - 14.8x
Estimate of income increase per year of iron fortification among iron-deficient population, current program settings	0.005%
Present value of total lifetime benefits per person with dietary iron deficiency among people under age 15 (units of increase in ln(income)	0.131%
Proportion of children in current programs who would benefit (with baseline dietary iron deficiency)	35%
Cost-effectiveness estimate from development effects
Total present value of units of increase in units of annual ln(income) among people under age 15	21,000
Units of value assigned to increasing ln(income) for one person for one year	1.44
Cost-effectiveness estimate	3.7x
Summary of primary benefits (% of modeled benefits)
Anemia morbidity averted	44%
Development effects	44%
Neonatal deaths averted	6%
Increased physical work capacity	5%
Additional adjustments
Adjustment for additional program benefits and downsides	10%
Adjustment for improved behavioral activities and long term benefits	32%
Adjustment for diverting other actors’ spending into SMC (“leverage”)	-7%
Adjustment for diverting other actors’ spending away from SMC (“funging”)	-5%
Overall cost-effectiveness(multiples of cash transfers)	11x

You can see our full cost-effectiveness analysis for this grant here and a simple version here.

The other factors informing our decision to make this grant are:

• We believe Nutrition International is well-positioned to carry out this program. Nutrition International has a long history of working on nutrition programs and of working in India.7 We believe they have a clear sense of relevant bottlenecks in rice fortification programs there and are well positioned to address them.
• We believe this grant has high learning value. We think IDinsight's evaluation will enable us to learn more about the cost-effectiveness of rice fortification programs in India and other geographies. This is important because we expect that iron fortification may make up a large portion of our nutrition-related grantmaking in the future. (more)

Our main reservations

We detail our main uncertainties and open questions about this grant below. At a high level:

• We’re unsure how accurate anemia burden estimates are. Our CEA relies on estimates of the prevalence of anemia from the Institute of Health Metrics and Evaluation (IHME)'s Global Burden of Disease (GBD) project, but we don’t know how accurate these estimates are. We haven’t deeply reviewed IHME's methods or other sources of data on anemia prevalence. (more)
• We don’t know how effectively rice fortification programs in West Bengal and Madhya Pradesh will reduce anemia burden. We think the effectiveness of fortification programs is likely to depend on a number of context specific factors. We include a number of adjustments in our CEA to account for differences between the trial conditions of the studies providing our evidence for the effectiveness of rice fortification and the likely program implementation settings in Madhya Pradesh and West Bengal, as well as for risks of bias in those trials, but we remain unsure about how accurate these are. (more)
• We’re unsure about existing coverage of rice fortified above minimum standards, the iron content of fortified rice, or how much we should expect Nutrition International’s program to improve these metrics: Our CEA models program benefits based on two factors: increased coverage of rice fortified above minimum standards and improved iron content in fortified rice. However, we are uncertain about current levels of adequately fortified rice, its iron content, and the program's expected impact on these metrics. (more)
• We’re uncertain about how much we’ll learn from IDinsight’s evaluation. IDinsight's endline evaluation will measure the quantity and iron content of rice fortified above minimum standards following two years of program implementation. However, we're unsure whether this analysis will be able to demonstrate the program's direct impact on changes to these metrics. (more)

The intervention

Anemia is highly prevalent in India.8 People with anemia may feel tired, weak, and short of breath. It has also been associated with other negative physical, psychological, and cognitive effects.9

There is some evidence that iron fortification reduces the prevalence of iron deficiency and anemia.10 Rice, which is a staple food in India, can be fortified with iron by creating fortified rice kernels made of rice flour and micronutrient premix, and then mixing them with raw rice kernels. This is a complex process, with a number of potential challenges, including:11

• Sometimes not enough fortified rice kernels are added to the rice, which means the rice will contain less iron than recommended.
• The kernels containing micronutrient premix can be differentiated from standard rice, and some consumers may remove the fortified kernels before cooking.
• Fortified rice kernels have a different density from standard rice, which increases the likelihood that the premix will be unevenly mixed with the regular rice kernels.
• Myths and misconceptions among beneficiaries lead to fortified rice not being consumed.

The government of India has mandated that the rice distributed through government-run public distribution systems (PDS) must be fortified.12 However, states may face barriers in complying with this directive, including:

• A lack of state-level government capacity to develop and implement standardized guidelines for rice fortification,13 monitor the program’s success by testing fortified rice to ensure it meets quality standards,14 and implement successful education campaigns.
• Lack of millers and premix producers’ technical capacity to meet quality standards.15

As part of this investigation, GiveWell spoke with government officials from Madhya Pradesh and West Bengal who confirmed that both states are facing such implementation bottlenecks and are looking for support.16

Does Rice Fortification work?

Ahead of this investigation, we conducted a meta-analysis of six randomized and nonrandomized controlled trials that measured the impact of iron-fortified rice on the prevalence of anemia. This analysis found that fortified rice reduces the prevalence of anemia by 29%, though we have some open questions about these findings that we detail in our report.

In our cost-effectiveness analysis, we model the primary benefits of rice fortification as coming from its impact on (1) morbidity associated with iron deficiency and anemia (~45% of benefits) and (2) improving the long-term income of children benefiting from the program (~45% of benefits). We also expect the program to decrease neonatal mortality (~5% of benefits) and improve the income of the anemic population working in physically demanding jobs (~5% of benefits).17 However, we have some uncertainties about the effectiveness of rice fortification in India that we outline below.

The grant

With this grant, Nutrition International will provide TA to improve the coverage of fortified rice that meets the minimum fortification standard, thereby increasing the amount of iron in fortified rice that is distributed through PDS in two Indian states, West Bengal and Madhya Pradesh. We think this TA will improve the quality of rice distributed by:

• Support for governments in setting up sustainable supply chain mapping, to avoid deterioration of the fortified rice.18
• Assisting with the development of standard operating procedures, monitoring and evaluation plans, and quality control guidelines.19
• Supporting the development and implementation of educational campaigns.20
• Training millers and premix suppliers on best production practices and quality assurance.21

Nutrition International will also commission IDinsight to carry out an impact evaluation of its program in both West Bengal and Madhya Pradesh.22 IDinsight will collect data on (i) coverage of adequately fortified rice and (ii) average iron content of the rice before program implementation. It will then collect this data two years later to estimate how much both metrics have improved.23 IDinsight’s evaluation will also incorporate process tracing, providing a qualitative assessment of the extent to which Nutrition International’s program is responsible for those improvements.24

We are uncertain about the current quality and iron content of fortified rice in both states and how much we should expect Nutritional International’s program to improve these (more below). As such, we are recommending a small disbursement of funding initially, with the remaining funding conditional on baseline results that indicate that the current quality of fortification is low enough for the program to remain cost-effective.

Budget for grant activities

This is a three year, $5.5 million grant, split into two phases:

• Phase 1 is an initial disbursement of $764,000 to enable IDinsight to design and implement a baseline evaluation. The budget breakdown for Phase 1 is:25

Cost Categories	Total (USD)	%
Program Activities
Evaluation design and IRB approval - IDinsight	117,733	15%
Baseline Evaluation (pre/post test with process tracing) - IDinsight	546,440	72%
Total Direct Costs	664,173	87%
Indirect 15%	99,626	13%
Grand Total	763,799	100%

• Phase 2 is contingent on baseline results showing:
  • The proportion of rice samples from warehouses meeting the minimum fortification standard is no higher than 60% and that the average iron content of rice fortified below the minimum standard is no higher than 23mg iron/kg; or
  • The proportion of rice samples from warehouses meeting the minimum fortification standard is no higher than 80% and that the average iron content of rice fortified below the minimum standard is no higher than 14mg iron/kg; or
  • Combinations of the proportion of rice samples from warehouses meeting the minimum fortification standard and average iron content of rice fortified below the minimum standard between these two options, that we estimate imply a similar opportunity for impact.26

Those conditions were chosen to ensure the current quality of fortification is low enough for the program to be above our funding bar. It’s possible that the quality of fortified rice is already high enough in both states that we will find that Phase 2 funding will not be cost-effective.

The budget breakdown for Phase 2 is:27

Cost Categories	Total (USD)	%
Program Activities	2,638,911	56%
A. Program implementation to scale and sustain rice fortification	1,890,513	40%
National-level Technical Assistance	138,785
Program Implementation in West Bengal	1,046,770
Funding gaps in program implementation in Madhya Pradesh	704,959
B. Documentation and learning	748,398	16%
Endline Evaluation (pre/post test with process tracing) - IDinsight	555,306
Benefit Cost Analysis & Cost Analysis	72,409
Studies to produce quality fortified rice and fortified rice kernels	120,682
Program Management	1,453,902	31%
Country Program and Technical Support	762,961
Regional and Global Program and Technical Support	394,555
Other Direct Costs and Shared Utilities	204,823
Travel to Support Program Implementation	91,563
Total Direct Costs	4,092,813	87%
Indirect 15%	613,922
Grand Total	4,706,735

Our understanding is that the total program budget is roughly 3 times higher than Nutrition International’s previous programs to support the fortification of PDS rice in India.28 This is due to the following programmatic changes:29

• Increasing the number of spot tests by a factor of 10.30
• Roughly doubling the number of meetings with state and district officials.
• Roughly doubling the number of supervision visits by Nutrition International field staff for millers and premix suppliers.
• Establishing and supporting state-led Program Management Units (PMUs) in West Bengal and Technical Support Units (TSUs) in Madhya Pradesh to provide ongoing technical assistance to ensure rice fortification programs are sustained and implemented.

Nutrition International has told us it needs five years of continuous implementation to ensure that benefits persist once the program is completed.31 However, we are recommending a three-year grant disbursed in two phases to start, since we have a number of open questions about the effectiveness of rice fortification in India and about the program (more below).

If we make Phase 2 of the initial grant, we might decide to extend our support for an additional two years, for a total of five years of continuous program implementation. This depends on what we learn about Nutritional International’s program and about the effectiveness of rice fortification more broadly over the course of this initial grant.

The case for the grant

We are recommending this grant because:

• We believe that Nutrition International’s technical assistance program to support iron fortification is cost-effective. We currently estimate it to be about 11 times as cost-effective as unconditional cash transfers, our benchmark for comparing programs. We typically recommend grants that are 10x cash or above, for non-top charity programs. (more)
• We believe Nutrition International is well-positioned to carry out this program. Nutrition International has a long history of working on nutrition programs and of working in India. Experts we spoke with confirmed its experience providing technical support for program implementation, M&E, and educational campaigns. We believe they have a clear sense of relevant bottlenecks in rice fortification programs and the ability to address them.
• We believe this grant has high learning value. IDinsight's evaluation will enable us to learn about the quality of rice fortification in India, the extent to which Nutrition International's program increases coverage of adequately fortified rice and their ability to predict this effect. This information could help us identify additional funding opportunities to support rice fortification in India, as well as opportunities to support Nutrition International’s fortification work in other geographies. This is especially important because we expect that iron fortification may be a large portion of our nutrition-related grantmaking in the future. (more)

High cost-effectiveness

Our best guess is that this grant is 11 times as cost-effective as unconditional cash transfers (GiveWell’s benchmark for comparing different programs). The main reasons we expect this grant to be cost-effective are discussed below.

• High anemia burden in India. IHME data suggests that anemia burden in India is high,32 and that burden appears even higher in West Bengal and Madhya Pradesh.33 However, we are uncertain about the accuracy of burden data and our estimate of burden reduction, which is based on several rough adjustments (see below).
• The program reaches many people at a low cost per person. The program involves technical assistance to the government and rice industry to improve rice fortification programs targeting a large number of people (roughly 292 million person-years of support34 ) at a relatively low cost. We estimate that the total cost per person reached annually by the program is around $0.01, which we calculate as the total program costs divided by the number of beneficiaries of the program over three years.35
• Nutrition International’s effect on iron intake. We think that Nutrition International will improve the coverage of rice fortified above minimum standards by about 8% over the three years of the program.36 We expect this increased coverage to translate to increases of NaFeEDTA-equivalent iron intake by about 0.3mg per week for beneficiaries of the program.37
• Reduced anemia morbidity (~45% of program benefits). We estimate that Nutrition International’s program will reduce anemia morbidity among beneficiaries by 0.6%. We base this on the findings of an internal meta-analysis, which shows that fortified rice reduces the prevalence of anemia by 29%. To this estimate, we apply several adjustments, including:
  • A 6% upward adjustment for the higher proportion of anemia in these states attributed to dietary iron deficiency.38
  • A 0.2% downward adjustment for the impact of infectious disease and inflammation on iron absorption.39
  • A 98% downward dose adjustment to account for difference in iron intake between trial settings and program implementation settings.40
  • A 150% upward fortification duration adjustment to account for our estimate that the short duration of the trials in our meta-analysis allow them to capture less than half of rice fortification’s full impact.41
  • A 16% downward adjustment adjustment for cooking methods, as some cooking methods result in a higher risk of discarding kernels made with fortification premix.42
  • A 25% downward adjustment for tea and phytic acid consumption, as both inhibit the absorption of iron and are consumed regularly in India.43
  • A 20% downward adjustment for lower effectiveness of fortification at scale, as our best guess is that fortification at scale would be less effective than in trial contexts.44
  • A 25% downward validity adjustment to account for risk of bias in the studies included in our meta-analysis.45
• Increased future income (~45% of program benefits). We think that the program will improve children's long-run cognitive and/or physical development, leading to small increases in income later in life. We estimate that this will lead to long-term annual income increases of .005%. This estimate relies on Niemesh 2015, a historical quasi-experiment based on the introduction of iron fortification of wheat flour and bread in the United States in the 1940s. From this study's findings, we calculate an increase of 0.32% in income per year of consuming iron-fortified wheat flour and bread. To this estimate, we apply two adjustments:
  • An adjustment for differences in iron increases between study and program settings. The program evaluated in Niemesh 2015 increased iron intake by 8mg/week,46 while this program would increase NaFeEDTA-equivalent iron intake by 0.3mg/week. This leads to a 96% downward adjustment.
  • A 60% downward replicability adjustment to account for the risk that the findings in Niemesh 2015 may be overstated.47
• Reduced neonatal mortality (~5% of program benefits). We think the program will reduce neonatal mortality by 0.1% among pregnant women who benefit from the program. This estimate of the effect of iron fortification and supplementation on reduced neonatal mortality comes from an unpublished internal meta-analysis of three RCTs of iron supplementation during pregnancy, which finds an overall reduction in risk of neonatal mortality of 26% (95% CI 1% to 46%). After internal and external validity adjustments, we then apply those benefits to the estimated number of pregnant women in the cohort who do not already receive iron through supplementation during pregnancy (66%) to arrive at our estimate.
• Increased physical work capacity (~5% of program benefits). We estimate that iron fortification increases productivity in working-age people with anemia with physically demanding jobs by an average of 0.1% and income by an average of 0.01%.
• Benefits from educational campaigns and long term program benefits. We include a rough 15% upward adjustment to account for unmodeled benefits from educational campaigns to address misconceptions about fortified rice among beneficiaries, which we think will lead to increases in the amount of fortified rice consumed. We include a second 15% upward adjustment to account for increases in iron concentration and improved coverage of adequately fortified rice that would continue accruing following completion of the program.

As mentioned above, Nutrition International has told us it needs five years of continuous implementation to ensure that benefits persist once the program is completed. If we made an additional $3 million grant to support the program for an additional two years, our best guess is that this grant would be 13 times as cost-effective as unconditional cash transfers. This estimate is based on several uncertain assumptions.48 We have not yet investigated these assumptions in depth, but plan to do so should we decide to consider an extension to this grant.

High learning value

We expect this grant and IDinsight's evaluation will provide information that will inform our grantmaking on fortification. In particular, we think:

• It will help us understand whether there might be cost-effective funding opportunities to improve the quality of fortified rice in India, by providing:
  • Estimates of adequately fortified rice coverage and average iron content for rice distributed through PDS in West Bengal and Madhya Pradesh at baseline.
  • Estimates of the effect Nutrition International has on coverage of adequately fortified rice and average iron content, through a before-and-after comparison and qualitative information on Nutrition International’s contribution to any observed increases.
• It will help us better understand Nutritional International’s work by providing:
  • Estimates of the effect Nutrition International has on coverage of adequately fortified rice and average iron content.
  • A qualitative assessment of whether and how Nutrition International's additional activities and increased budget affected their impact.
  • Feedback on Nutrition International's ability to predict baseline coverage and its own impact.

We conducted a value of information analysis to determine whether the benefits of conducting an impact evaluation justify the cost. For simplicity, we only focused on the effect the information would have on our funding to Nutrition International (and set aside the effect it might have on funding opportunities to improve the quality of fortified rice in India beyond Nutrition International). Over the next 5 years, we think it’s plausible that we will allocate approximately $17 million in funding to Nutrition International (including work in India and other countries)

We found that, for the evaluation to be 10 times as cost-effective as unconditional cash transfers (our grantmaking bar), it must increase the probability of increasing the cost-effectiveness of this $17 million by 3x by at least 25%. This seems plausible to us because:

• Fortification appears to be an inexpensive and high-impact intervention.
• Nutrition International is a major player in the fortification space.
• We currently have little information on the effect Nutrition International’s activities have on the coverage of adequately fortified rice, and their ability to predict this impact, so the impact evaluation should provide insights that could inform future estimates.

However, we have some concerns about how much we can learn from IDinsight's evaluation (see below)

Risks and reservations

We have open questions about rice fortification, about Nutrition International’s TA program, and about the budget and structure of this grant, as follows:

Uncertainties about rice fortification

• How accurate are anemia burden estimates in India? Our CEA relies on estimates of the prevalence of anemia from IHME's GBD project. IHME's estimates rely on complex mathematical models that we do not fully understand.49 We have not yet investigated IHME's modeling choices in depth, considered alternative sources for estimating anemia prevalence, or discussed anemia prevalence estimates with experts in depth.
• How effective will rice fortification programs be in West Bengal and Madhya Pradesh at reducing anemia burden? The effectiveness of fortification programs is likely to depend on a number of context-specific factors. We include a number of adjustments in our CEA to account for differences between the trial conditions of the studies providing our evidence for the effectiveness of rice fortification and the likely program implementation settings in Madhya Pradesh and West Bengal, as well as for risks of bias in those trials.50 Many of these are based, in part, on subjective best guesses, and we remain unsure about the magnitude of those adjustments. We are also uncertain whether these adjustments are comprehensive as we have not examined the fortification process in detail. We are planning to do additional research on these adjustments before making further grants for iron fortification in India.

Uncertainties about the program

• What is the baseline coverage of rice fortified above minimum standards and the iron content of fortified rice? The current coverage of adequately fortified rice and iron content of fortified rice in Madhya Pradesh and West Bengal are key inputs in our CEA, but we are relatively uncertain about both. Our coverage estimate relies on Nutrition International’s best guess that coverage of adequately fortified rice in Madhya Pradesh and West Bengal is between 50-60%, but other stakeholders we spoke with gave us significantly higher estimates.51 We are also uncertain about the average iron content of fortified rice. The estimates we use in our CEA rely on Nutrition International data from a district in Madhya Pradesh, but we don’t know how representative this data is of other districts. We expect to learn more about these uncertainties through IDinsight’s baseline evaluations ahead of program implementation.
• How much will Nutrition International’s program improve iron content and coverage of adequately fortified rice, and how much would it improve counterfactually? We estimate that Nutrition International’s program will increase the iron content of fortified rice by 1.5 mg per kg of rice over the course of the program relative to the counterfactual. This is based on an uncertain estimate that Nutrition International's program will increase adequately fortified rice coverage by 10 percentage points over the counterfactual, and that this will result in an increase of iron content by 1.86 mg per kg of rice.52 To this, we add a skeptical adjustment of 20% to account for our uncertainty about how much Nutrition International’s program will increase coverage of adequately fortified rice and what the iron content of that rice will be.
• We’re uncertain about how much we’ll learn from IDinsight’s endline evaluation. IDinsight's endline evaluation will measure the quantity and iron content of rice fortified above minimum standards following two years of program implementation. To determine the program's impact, they'll use a method called process tracing.53 This approach will examine the TA activities in relation to the changes to fortified rice coverage and iron content in an attempt to establish whether and how the program caused the observed changes. However, these results will be necessarily qualitative, as we won’t have a quantitative estimate of the quality and content of fortified rice without Nutrition International to compare against. We're unsure whether this analysis will prove the program's direct impact or how valuable the resulting qualitative data will be.
• The amount of fortified rice consumed in Madhya Pradesh and West Bengal. Our cost-effectiveness estimate relies on the total amount of rice distributed through social protection programs that is consumed in each state. These estimates are important inputs into our calculations for the cost per person reached. We start by the amount of rice held in state warehouses, based on estimates shared by Nutrition International.54 To estimate the amount of fortified rice that is consumed, we apply two discounts to the estimate of total rice fortified: a 25% discount to account for wastage at the warehouse level and a 20% discount to account for wastage at the household level. We are uncertain about the estimates of the amount of fortified rice produced in each state, as well as how accurate our subjective adjustments are. We plan to investigate these questions through conversations with Nutrition International and other stakeholders in greater depth should we decide to consider an extension to this grant.
• Budget uncertainties. As above, our understanding is that this budget is roughly 3 times higher than Nutrition International’s previous expenditures under their institutional Grant from Canada to support the fortification of PDS rice in two states in India. While we have a general understanding of the additional activities this funding will support, we have open questions about how the additional funds will be used and how it will ultimately impact the program’s effectiveness.

Plans for follow up

• We plan to analyze the data from IDinsight’s baseline evaluation to ensure that the program looks cost-effective before making the second phase of this grant by October 2025.
• We plan to investigate questions around how much rice is fortified and how much is consumed in each state through conversations with Nutrition International and other stakeholders in greater depth before we decide to expand rice fortification in new states or geographies (likely September 2025) or renew this grant (September 2026).
• We plan to analyze the data from IDinsight’s endline evaluation to determine the difference in average iron concentration of fortified rice and coverage of adequately fortified rice relative to baseline data, and what effect the program had on affecting these outcomes.
• We will continue discussions with Nutrition International to get aligned on organizational priorities and program budgets, and to better understand its process for rice fortification by the end of 2025.
• We plan to look more deeply into iron deficiency anemia burden in India before we commit further funding to iron fortification there. We expect to begin work on this at the end of 2024.
• We have a number of open questions about external validity adjustments in our model that we plan to investigate more deeply, including:
  • The effect of inflammation on iron absorption
  • Whether the effect of iron intake and fortificant concentration on anemia burden are linear
  • The effect on inhibitors (e.g. tea and phytic acid) on program effectiveness

Internal forecasts

For this grant, we are recording the following forecasts:

Confidence	Prediction	By time	Resolution
60%	We receive baseline results by end of September 2025	End of September 2025	-
65%	Conditional on receiving results, we disburse the rest of the grant by end of October 2025	End of October 2025	-
60%	Nutrition International’s effect on coverage of rice fortified above minimum standards is at least 16pp increase (20pp*0.8 adjustment) from baseline after 2 years of program implementation	End of September 2027	-
60%	The average iron content for fortified rice with Nutrition International support, for rice fortified below maximum standards, is at least 1.5mg/kg higher than it was at baseline after 2 years of program implementation	End of September 2027	-
60%	We decide to make more than $17m in grant to Nutrition International in the next 5 years	September 2029	-

Our process

• We began conversations with Nutrition International about iron fortification in South Asia in late 2022, and this grant at the beginning of 2023. The investigation was briefly put on hold while Nutrition International facilitated discussions between West Bengal and central government officials.
• Our initial investigation focused on program implementation in additional states, and included an impact evaluation run by Nutrition International estimating the effect of rice fortification on anemia.
  • We shifted to focus to only West Bengal and Madhya Pradesh because the other states fell below the 10x bar in our cost-effectiveness model.
  • We also set aside the impact evaluation component. When we began our investigation, the PDS rice fortification program was in pilot stages, and we were exploring the possibility of running a difference-in-difference evaluation; the program has now already been rolled out, so this would no longer be possible.
• We spoke with stakeholders from BMGF, GAIN, and various government officials from West Bengal and Madhya Pradesh.
• A team of senior researchers and external experts reviewed key assumptions behind our rice fortification CEA.

Sources

Document	Source
GBD 2021 Anemia Collaborators, 2023	Source (archive)
GiveWell, CEA of Nutrition International's Iron Fortification of Rice in India Program	Source
GiveWell, Literature review of evidence on tea and infection in children vs. adults	Source
GiveWell, Meta-analysis of fortified rice impact on anemia in RCTs, 2023	Source
GiveWell, Nutrition International — Vitamin A Supplementation in Chad, 2022-2024 (January 2022)	Source
GiveWell, Translating fortification effects from efficacy trials to the programs we fund	Unpublished
Government of Maharashtra, Distribution of the fortified rice through PDS and Milling	Source (archive)
Institute for Health Metrics and Evaluation Global Burden of Disease project, 2021	Source
Niemesh 2015	Source (archive)
Nutrition International responses to GiveWell questions	Unpublished
Nutrition International, India RF Program Scale-up with GW Funding	Unpublished
Nutrition International, Our Programs - India	Source (archive)
Nutrition International, Proposal to support rice fortification project in India, 2024	Unpublished
Nutrition International, Questions and answers on TA for fortification of rice in India	Unpublished
Nutrition International, Rice Fortification in India - GiveWell Budget Phase 1, 2024 (unpublished).	Unpublished
Nutrition International, Rice Fortification in India - GiveWell Budget Phase 2, 2024 (unpublished).	Unpublished
Nutritional International, Responses to GiveWell Follow-up Questions, 2023	Unpublished
Peña-Rosas et al. 2015	Source (archive)
Peña‐Rosas et al. 2019	Source (archive)
Petry et al. 2016	Source (archive)
World Health Organization, Anaemia fact sheet, 2023	Source (archive)

• 1

  “Chronic iron deficiency frequently turns into iron-deficiency anaemia. While iron deficiency is the most common cause of anaemia, other causes such as acute and chronic infections that cause inflammation; deficiencies of folate and of vitamins B2 , B12 , A, and C; and genetically inherited traits such as thalassaemia and drepanocytosis (sickle-cell anaemia) may be independent or superimposed causal factors (WHO 2001; WHO 2015a)." Peña-Rosas et al. 2015, p. 7.

• 2

  Our internal meta-analysis finds that fortified rice reduces the prevalence of anemia by 29%, though we have some open questions about these findings that we detail in our report.

• 3

  Nutrition International, comment on a draft of this page, December 2024 (unpublished).

• 4

  Leonie Vierck, Deputy Director, Nutrition, Bill & Melinda Gates Foundation, conversation with GiveWell, April 2023 (unpublished).

• 5
  • "As per the Government of India’s food fortification standards, rice is to be fortified with iron, folic acid (vitamin B9) and vitamin B12 (Cyanacobalamine or Hydroxycobalamine). In addition, there are a few micronutrients that could also be added to rice optionally."
  • The minimum fortification standard for rice distributed through PDS is 28mg of iron per kg of rice. “The GoI recommends that rice be fortified with between 28 and 42.5 mg of iron per kg of rice with ferric pyrophosphate as source of iron. At this consumption level and with the existing range of iron level in the GoI’s rice fortification standards, both women of reproductive age (WRA) (15 to 49 years old) and adult men will receive between 4.7 to 7.1 mg of iron per day."

  Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), pg. 15.

• 6

  The rice distributed through PDS will be fortified with ferric pyrophosphate. However, the fortificants used in the trials we base our analysis on differ, and each compound may have different bioavailability. In order to compare the effects of these trials with the anticipated effects of Nutrition International’s program, we adjust the total additional iron intake by an absorption factor relative to sodium iron ethylenediaminetetraacetate (NaFeEDTA), a highly bioavailable iron fortificant. See our calculations here.

• 7

  "For more than 20 years, Nutrition International has been working in India alongside government ministries to strengthen and address the country’s nutrition challenges." Nutrition International, Saving lives and improving the health of women and children in India.

• 8

  Roughly 42% prevalence in 2021 among those 5-14 years old according to IHME.

• 9

  “Anaemia causes symptoms such as fatigue, reduced physical work capacity, and shortness of breath. Anaemia is an indicator of poor nutrition and other health problems.” World Health Organization, Anaemia fact sheet, 2023.

• 10

  Our internal meta-analysis finds that fortified rice reduces the prevalence of anemia by 29%, though we have some open questions about these findings that we detail in our report.

• 11

  “The rice fortification through SSNPs is being scaled up quickly. Some of the quality challenges that we expect are:
  • varying levels of micronutrients;
  • easily distinguishable kernels;
  • changes to fortified rice after being cooked;
  • loss of micronutrients during storage and transport;
  • segregation of the FRK during transport and
  • myths and misconception among beneficiaries.”
  Nutrition International, Responses to GiveWell questions (unpublished), March 1, 2023

• 12

  “For the prevention of the high volume of the Anaemia in the country and the state, the central government has decided to distribute the fortified rice, prepared by mixing the FRK (Fortified Rice Kernel – FRK) (made from rice flour and which contains the high volume of the micronutrients such as Iron, Folic Acid and Vitamin B12) in the CMR in the proportion of 1 (FRK):100 (CMR), through the public distribution system.” Government of Maharashtra, Distribution of the fortified rice through PDS and Milling, April 2022

• 13

  Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished)

• 14

  Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished)

• 15

  Rajan Sankar, Advisor, Bill and Melinda Gates Foundation India, conversation with GiveWell, March 30, 2023 (unpublished).

• 16

  Unpublished conversations with Mr. Parvez Siddiqui, Principal Secretary, Food & Supplies Department, West Bengal Government and Faiz Kidwai, ex-Principal Secretary, Madhya Pradesh State Food and Civil Supplies & Consumer Protection Department (Government Official), March 27, 2023

• 17

  See this section of our CEA.

• 18

  “To support the governments in ensuring the quality of fortified rice, Nutrition International will map out supply chain challenges related to shelf-life of the FRKs and fortified rice and with local academic
  institutions undertake analysis of the shelf-life of the product being distributed through the SSNPs.” Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), p. 8

• 19

  “At the state-level, Nutrition International proposes to introduce and scale inclusion of fortified rice in the SSNPs from pilot districts to the entire state. This support will include supply chain mapping, capacity strengthening through training and supervision, and monitoring and evaluation of the program in line with national standards, operating procedures, and quality assurance/quality control guidelines. In addition to strengthening the capacity of government officials to manage, plan for and monitor the fortification system in line with established best practices, Nutrition International proposes to support the establishment of Technical Support Units in targeted states to oversee, monitor and lead the program.” Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), p. 3

• 20

  “Communication materials on the benefits of consuming fortified rice provided to stakeholders and address myths and misconceptions on consumption of fortified rice ... Nutrition International will roll out the national community engagement plans to the targeted states to address consumer awareness and misconceptions regarding consumption of fortified rice.” Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), p. 3

• 21

  “Nutrition International will build the capacity of the millers and premix suppliers for adoption of best production practices and quality assurance norms for production of adequately and uniformly fortified rice, including establishing upfront capacity and continuing supply with quality. Also, in collaboration with BIS/FCI departments, Nutrition International will build the capacity of manufacturers of blenders/extruders regarding recommended standards of the equipment and monitor designs available to ensure that more fortified rice kernel (FRK) producers are able to meet quality standards.” Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), p. 3

• 22

  "In addition, Nutrition International will engage the services of IDinsight to conduct a process and impact evaluation…To measure the degree to which this grant increased the availability of adequately fortified rice under SSNPs in Madhya Pradesh and West Bengal, an impact evaluation will be conducted." Nutrition International, Proposal to support rice fortification project in India, 2024 (unpublished).

• 23

  "The impact evaluation will focus on the quality of rice after the program has been in place in each of the two states for a specific period, comparing the metrics between baseline and endline to assess improvement in quality of fortified rice." Nutrition International, Proposal to support rice fortification project in India, 2024 (unpublished).

• 24

  “The qualitative component of process tracing looks at the causal mechanism in detail, thus, we will describe whether and how well program implementation takes place over time and how assumptions we made in the Theory of Change hold up in reality (Was the State-led Program Management Units established and built?)... As we do not have a quantitative causal estimate with a counterfactual, we will not be able to attribute the change over time to the intervention. However, using process tracing with both its quantitative and qualitative component, we can say whether the intervention was the likely main contributor to the change over time. What we can also do is to look at the partial correlation of activities with the program outcomes to get a sense of the strength of this associational relationship.” Unpublished email with Nutrition International, July 29, 2024.

• 25

  Nutrition International, Rice Fortification in India - GiveWell Budget Phase 1, 2024 (unpublished).

• 26

  The minimum threshold criteria for phase 2 approval were determined through quantitative analysis to maintain program cost-effectiveness. Our benchmarking indicated that when either (a) no more than 60% of rice samples meet the minimum fortification standard with substandard samples averaging ≤23mg iron/kg, or (b) no more than 80% of rice samples meet the minimum standard with substandard samples averaging ≤14mg iron/kg, the intervention still meets our 10x cash bar for grantmaking. These thresholds represent the minimum values for both proportion and iron content that would allow the program to remain cost-effective. This is based on unpublished versions of our cost-effectiveness analysis with parameters altered to evaluate minimum criteria.

• 27

  Nutrition International, Rice Fortification in India - GiveWell Budget Phase 2, 2024 (unpublished).

• 28

  "NI’s Rice Fortification program budget in India for FY 24/25 is CAD $783,092 (USD $560k) annually, covering National TA and 3 States." Nutrition International, India RF Program Scale-up with GW Funding, 2024 (unpublished).
  Excluding the funding for the phase 1 IDinsight evaluation, this program budget would be $4.7m over three years, or roughly $1.6m per year ($4.7m / 3 = $1.56m). This is approximately three times the 2024-2025 annual budget of $560k ($1.6m / $560k = 2.85).

• 29

  Unpublished email with Nutrition International, July 29, 2024.

• 30

  Unpublished email with Nutrition International, July 29, 2024.

• 31

  NI is proposing an ambitious approach to ensure the scaling of rice fortification in at least two target states within 3-5 years. However, to validate the work, ensure its sustainability and generate evidence through a rigorous effectiveness-implementation hybrid design study, NI would prefer consideration of a 5-year grant period." Nutrition International, Questions and answers on TA for fortification of rice in India, 2023 (unpublished), p. 1

• 32

  IHME's estimate of the rate of all-cause anemia among the population of India in 2021 was 43709.484 per 100,000. In comparison, IHME's estimate of the global rate of anemia in 2021 was 24328.192 per 100,000.

• 33

  We estimate the burden of anemia in India on the basis of IHME data. We then adjust this estimate upward by 13% in Madhya Pradesh and 23% in West Bengal, based on state-level IHME data.

• 34

  We calculate that this program will cover approximately 97.4 million people for three years. 97.4 million people*3 years = 292.2 million person-years.

  Note that Nutrition International calculates coverage differently; they instead estimate unique beneficiaries. "NI estimates the numbers of SSNP beneficiaries with access to fortified rice in areas supported by the project to be 130 million. This figure is based on the number of Public Distribution System (PDS) cardholders in project areas as of September 2024, adjusted by projected coverage rates, and reflects the estimated number of unique individuals reached over the life of the grant." Nutrition International, comments on a draft of this page, June 26 2025 (unpublished). This differs from GiveWell’s approach, and as a result, total beneficiary figures vary depending on the methodology used.

• 35
  • See these rows in our CEA for these calculations.
  • To estimate program beneficiaries, we multiply the quantity of fortified rice in each state by a subjective wastage adjustment of 25%, and divide this by the amount of rice allocated per person per day. See these rows in our CEA for more details.

• 36

  To calculate this, we multiply Nutrition International's estimate of the increase in fortified rice resulting from their support by an adjustment for the benefits of partially-fortified rice and a best-guess downwards skeptical adjustment. See here in our CEA.

• 37

  We apply a discount factor for absorption of ferric pyrophosphate relative to NaFeEDTA.

• 38

  We expect that iron fortification programs only alleviate anemia cases that are caused by inadequate iron intake. We adjust for the difference in proportion of anemia cases caused by dietary iron deficiency in the trials in our meta-analysis compared to program implementation settings. This results in a 6% upward adjustment.

• 39

  Our understanding is that iron absorption may be inhibited by the body's inflammatory response to infections, and that this may reduce the effectiveness of iron fortification and supplementation programs in areas with high infection prevalence. We adjust for this concern using an "inflammation-exposure index" developed by Petry et al. 2016. We estimate that program and trial setting are similar in this respect, resulting in a very small 0.2% downward adjustment.

• 40

  We estimate that the program will increase NaFeEDTA-equivalent iron intake by 0.3mg/week, which is roughly 2% of the increase seen in trial programs (14mg/week), resulting in a 98% downward adjustment.

• 41

  The trials in our meta-analysis last a weighted average of 6 months, but we think only 40% of the full effect of fortification is realized over this timeframe. As a result, we estimate trials capture less than half of the program’s impact, so we apply a 150% upward adjustment.

• 42

  Findings suggest that rinsing, soaking, or frying hot-extruded FPP-fortified rice and cooking it with the absorption method causes little iron loss, resulting in ~80-100% iron retention at the point of consumption. The excess water cooking method tends to cause more iron loss, with figures for iron retention ranging from 55-100%. We are not aware of evidence on the type of cooking method used in program geographies. Nutrition International has told us that “absorption method and cooking in excess water that is drained off later are the most common cooking methods used across different states in India.” (Nutrition International, Responses to GiveWell Follow-up Questions, 2023 (unpublished)). We assume those are equally common, and therefore take an average of 90% and 77% (the mid-points of the ranges for the two cooking methods) to estimate iron retention. 90% retention being a 10% downward adjustment, and 77% being a 23% downwards adjustment, the average is (10% + 23%)/2 = 16.5%.

• 43

  We would expect tea consumption to be lower and phytic acid consumption to be higher in trial settings compared to program settings, because trials mostly included children who are less likely to drink tea than adults. We apply a 25% downward adjustment to account for this. Calculations for the adjustment for tea consumption can be found on this section of the CEA. We then double that adjustment to get a rough estimate of the effects of both tea and phytic acid.

• 44

  Our prior is that fortification at scale would be less effective than in trial contexts. While we have tried to model the main differences, we are concerned we might be missing some (more below). We roughly account for this with our 20% downward adjustment.

• 45
  • This is in line with Peña‐Rosas et al. 2019’s assessment, which flags the level of certainty as “low”, due to risk of bias (the baseline characteristics were not similar in all groups and the method of randomisation was unclear in half of the studies) and indirectness (studies used different cut‐off levels of hemoglobin to define anemia)
  • “Downgraded 2 levels: one for serious limitations in study design or execution (risk of bias) and one for indirectness. The baseline characteristics were not similar in all groups and the method of randomisation was unclear in half of the studies. Also studies used different cut‐off levels of haemoglobin to define anaemia. Hardinsyah 2016; Parker 2015 (C); Perignon 2016 (C); Radhika 2011 used WHO cut‐off levels, Hotz 2008 used CDC criteria and Angeles‐Agdeppa 2008 and Thankachan 2012 did not name the criteria they used.” Peña‐Rosas et al. 2019

• 46

  The linked sheet (here) shows 1.2mg/day of NaFeEDTA-equivalent increase in iron consumption. Per week, this is 1.2 mg/day x 7 days/week = 8.4mg/week.

• 47

  This is a subjective validity adjustment to account for the evidence being non-randomized and from a long time ago, and to account for gender disparities in income effects (Niemesh 2015 only finds statistically significant effects (at the 1% level) when the analysis is restricted to men). Our reasoning (and the size of our adjustment) is very similar to our discussion of a replicability adjustment for natural experiments on the long-run income impacts of malaria programs. See here for more information.

• 48

  These assumptions are:

  • The average annual costs of the program in years 4 and 5 would be the same as the average annual costs over years 1-3.
  • That benefits extend for 10 years (5 years after the end of the program)
  • That (i) coverage of adequately fortified rice without Nutrition International support increases by 5 percentage points per year and caps at 75% coverage; and (ii) coverage of adequately fortified rice with Nutrition International support increases by 10 percentage points per year and caps at 85%.

• 49

  See "GBD 2021 anemia collaborators 2023" for a discussion on how anemia is estimated in GBD 2021.

• 50

  See sections above

• 51

  For example, a stakeholder from World Food Program India suggested a coverage rate of 85%-88%. We are uncertain about the basis for these differing estimates. World Food Program, conversation with GiveWell, October 9, 2024 (unpublished).

• 52

  Nutrition International estimates that its program will increase the proportion of adequately fortified rice by about 30 percentage points over three years. We assume that coverage would increase by 15 percentage points without Nutrition International support. Unpublished email with Nutrition International, August 21, 2024.

• 53

  "In addition, IDinsight will pursue a process tracing evaluation, which is similar to a pre-post evaluation paired with a comprehensive mixed methods procedure. This is a within-case evaluation approach that uses descriptive and diagnostic evidence collected over time to study the underlying causes of known outcomes in a specific case of interest." Nutrition International, Proposal to support rice fortification project in India, 2024 (unpublished).

• 54

  See these rows in our CEA.