Tech Mahindra Foundation — Rapid Surveys on Tea Consumption and Iron Supplementation Timing in India (March 2025)

Note: This page summarizes the rationale behind a GiveWell grant to Tech Mahindra Foundation. Tech Mahindra Foundation staff reviewed this page prior to publication.

Published: September 2025

1. The organization

Tech Mahindra Foundation (TMF) is a survey firm based in India focused on education, employability, and disability programs.1 This is GiveWell’s first grant to TMF.

2. The grant

This $22,000 grant will fund rapid surveys to collect data on several key uncertainties in our CEA for IFA supplementation programs in schools:2

1. The time of day IFA supplements are delivered in schools relative to meal times.
2. Whether students consume tea close to the time of IFA supplementation.
3. The extent to which students take IFA supplements outside of school settings.

TMF informed us that schools would be closed in Karnataka during the survey period and that students in Karnataka do not consume tea at school.3 Due to these scheduling and contextual differences, the survey methodology will differ by state:

• In Rajasthan: TMF will conduct school-based surveys focused on IFA and tea consumption habits in schools across three districts. They will target approximately 900 students between the ages of 5-19 years.
• In Karnataka: TMF will conduct household surveys focused on IFA consumption outside of schools. These will target 600-900 households with an equal split between rural and urban areas. They will survey an estimated 1.5 children per household (including one child in each household and a second child only if of a different gender).4

At the time of publishing this page, we have received the survey results, but not yet incorporated them into our CEA. Our preliminary analysis suggests that the results will not significantly change our estimate of the cost-effectiveness of IFA programs.

3. The case for the grant

We are recommending this grant because:

• The data collected may help resolve uncertainties in our CEA, including:
  • When school children consume IFA supplements relative to meal times, which can affect iron absorption. Our cost-effectiveness analysis currently assumes that 100% of students consume IFA supplements after a meal, which we incorporate as a -40% adjustment to bioavailability-adjusted iron intake relative to trial settings.5 If we learn that a large proportion of students actually consume IFA on an empty stomach, we think our cost-effectiveness estimates could increase by approximately 20%.6
  • Whether older students drink tea around the time they take supplements, reducing iron absorption. Tea consumption can reduce iron absorption by 80-85% when consumed near the time of supplementation.7 Our current model doesn't include a downward adjustment for tea consumption. Our understanding is that tea consumption in children under 12 is negligible in India, but we lack data for older students.8 If tea consumption is common in this age group, it could reduce our cost-effectiveness estimates for them by approximately 20%.9
  • The extent to which students take IFA supplements outside of school settings. Our cost-effectiveness analysis assumes that IFA supplementation occurs primarily through school-based programs. However, we lack data on whether students receive IFA supplements from other sources, such as healthcare facilities, pharmacies, or government health programs outside of schools. Through the household surveys in Karnataka, we hope to understand the baseline level of IFA supplementation from non-school sources. We may opt to factor IFA from additional sources as an additional adjustment in our CEA.
• The cost of the surveys is low relative to the potential value of information. At $22,000, this grant represents a small fraction of the several million U.S. dollars in potential funding we may direct to Evidence Action's IFA technical assistance program in India in 2025. Given that updates from these surveys could shift our cost-effectiveness estimates by over 20%, we consider this a worthwhile investment to improve our decision-making.

4. Risks and reservations

Our main reservations about this grant are:

• We may not receive results in time to inform our upcoming renewal grant. We expect to make a decision about a potential several million U.S. dollar grant to renew funding for Evidence Action's IFA technical assistance program in 5 states in India and extend the program to Karnataka by June, 2025. However, we don’t expect to receive results from these surveys until May 21, 2025.10 This timing could make it difficult to incorporate the findings into our investigation for this grant.
• Uncertainties about government permissions and data privacy issues in India. TMF will need to receive permission from district authorities to conduct the school-based surveys.11 We have some uncertainties about timelines to obtain government permissions for surveys in both states, including potential data privacy concerns related to collecting student information in schools. It’s possible these issues could lead to delays in survey implementation, which could limit our ability to use this data for the upcoming grant investigation.
• Limited experience with the survey firms. We have not previously worked with TMF, so we have limited experience with the quality and reliability of their work. However, it was recommended as a good candidate to carry out this work by Lalit Kumar of the GiveWell Research Council.

5. Plans for follow up

• Lalit Kumar of the GiveWell Research Council will meet with TMF regularly during the project period.
• We expect to receive preliminary survey findings by May 21, 2025. We plan to update our IFA cost-effectiveness analysis once the survey data becomes available.

6. Internal forecasts

We did not record any internal forecasts at the time of making this grant.

7. Our process

Our process for this grant recommendation included:

• Identifying the need for additional information on tea consumption and IFA supplement timing as part of our investigation into Evidence Action's IFA technical assistance program
• Consulting with Lalit Kumar of the GiveWell Research Council, who recommended Tech Mahindra Foundation as an appropriate survey implementation partner
• Adapting our approach when we learned that schools in Karnataka would be closed during the survey period, shifting to household-based surveys in that state
• Reviewing the survey instruments developed by Tech Mahindra Foundation for both the household-based surveys in Karnataka and school-based surveys in Rajasthan

8. Relationship disclosures

We do not have any relationship disclosures to make regarding this grant.

Sources

Document	Source
Disler et al. 1975	Source
GiveWell, Evidence Action — Iron and Folic Acid (IFA) Supplementation in India (August 2022)	Source
GiveWell, Internal Forecasts	Source
GiveWell, What We Learned From Red Teaming Our Iron Grantmaking (March 2025)	Source
Hurrell et al. 2000	Source
Hurrell, Reddy, and Cook 1999	Source
Kabil Professional Services (KPS), Homepage	Source
Lazrak et al. 2021	Source
MacPhail et al. 1980	Source
Tea Board of India, Executive Summary of Study on Domestic Consumption of Tea in India (2007)	Source
Tech Mahindra Foundation (TMF), Homepage	Source
Tech Mahindra Foundation (TMF), Project Proposal and Budget	Unpublished

• 1

  More information on TMF is available on their website here.

• 2
  • The $22,000 budget is based on unpublished conversations with TMF
  • Rapid surveys are focused data collection efforts designed to gather specific information quickly and efficiently to inform time-sensitive decisions. Unlike comprehensive surveys that may take months to plan, implement, and analyze, rapid surveys typically have shorter timelines, targeted scope, and streamlined methodologies.
  • These uncertainties were previously mentioned in our page on ‘What We Learned From Red Teaming Our Iron Grantmaking’.

• 3

  Source: Unpublished conversation with TMF

• 4

  Source: Unpublished conversation with TMF

• 5

  These figures are based on an internal version of our IFA CEA that, as of the time of publishing this page, has not been published.

• 6

  This estimate of 20% is a subjective guess, we have not modeled out these changes in advance of the grant.

• 7

  At least five trials consistently report that tea consumption reduces the absorption of dietary iron from a variety of fortificants (including NaFeEDTA) by about 80-85% (high confidence). Therefore, tea is potentially a very serious concern for the effectiveness of iron fortification programs. These studies include:

  • Lazrak et al. 2021 reports that Moroccan green tea reduces the absorption of iron from NaFeEDTA from a flour-based meal by >85% in both anemic and non-anemic women (relative to water).
  • Hurrell et al. 2000 reports that black (Assam) tea reduces the absorption of iron from white flour bread fortified with NaFeEDTA or ferrous sulfate by 84% and 82%, respectively.
  • Hurrell, Reddy, and Cook 1999 reports that black (Assam) tea reduces the absorption of iron from white flour bread fortified with ferrous sulfate in men and women without anemia by 79-94% (relative to water).
  • MacPhail et al. 1980 reports that tea reduces the absorption of iron from sugar water containing NaFeEDTA by 85%.
  • Disler et al. 1975 reports that drinking tea (I think black)# reduces iron absorption from bread fortified with ferric chloride or a meal of potato and onion soup with rice fortified with ferrous sulfate by about 75-80% (relative to water).#

• 8

  The penetration of tea in the kids segment below 12 years of age which on an average comprise of ~25% of the overall population is almost negligible." Executive Summary of Study on Domestic Consumption of Tea in India, Tea Board of India (2007)

• 9

  This is a rough estimate of what we believe is a plausible effect, we have not performed sensitivity analyses to estimate the potential impact of the results of this grant on our CEA.

• 10

  Source: Unpublished email from TMF

• 11

  Source: Unpublished email from Lalit Kumar, a member of GiveWell’s Research Council with extensive experience in program implementation