• A $1.88 million increase in costs for the four provinces in which VillageReach is currently working over the period of the project plan (2010-2014). This includes:16
  • An additional $666,000 in indirect costs of the program (35% of the increase). These are VillageReach headquarters costs that are officially allocated to the Mozambique project. VillageReach explains that in its previous budget estimate it underestimated the level of indirect costs that would be necessary to support the program, We do not feel that we have a complete understanding of what caused this underestimate.
  • $440,000 to cover per diems (25% of the increase; discussed above).
  • An additional $228,500 for vehicle purchases and maintenance (12% of the increase).
  • $275,000 to extend VillageReach's planned period of engagement each of the provinces (15% of the increase).
  • $273,000 in various other costs including a greater number of in-country staff and higher than expected costs for hotels, flights, and surveys.

  VillageReach notes that some of the increase in budgeted costs is due to unexpected inflation in Mozambique and an increase in the value of the Mozambican currency relative to the U.S. dollar.17

• VillageReach has removed four provinces (Inhambane, Sofala, Tete, Manica) from the budget, reducing the budget by $2.13 million.18 VillageReach writes:19
  
  We will focus our resources for the coming year on the four provinces in which we are currently engaged. No new provincial deployments will begin until the following are achieved:

  • marked improvement in reported distributions and routine data collection;
  • identified minimum funding level of 70% of the estimated costs for a complete engagement (3-, 4-years) in each additional province; and
  • hire of VillageReach staff for each new province in advance of any new engagements.

• Beyond the costs allocated to the Mozambique project, we now estimate that the VillageReach will need $302,000 less in unrestricted funds over the period of the project than previously thought. Non-Mozambique costs are funds spent on "general program," "program development," management, and fundraising (some of these costs are allocated to the Mozambique project as "indirect" costs). There were two changes in the budget for these categories: (a) higher actual and projected costs for FY 2011-2012 (by $108,000), and (b) one less year included in the budget due to the reduction in the number of provinces covered, causing a decrease in the total budget of $410,000.20

In sum, the budget changes reduce expected costs of the project by $557,000.21 Note that the revised budget is one year shorter than the earlier budget. The implications of this change for VillageReach's room for more funding are discussed below.

As discussed above, VillageReach's cost estimates for the 4 provinces it is working in have increased and it is currently unsure whether it will expand to the remaining 4 provinces in the original plan. The facts that (a) VillageReach's overall budget has stayed roughly the same as it was; (b) it now intends to cover half as many provinces as it previously did; and (c) the four provinces it no longer intends to enter in the near term are the ones that had significantly lower baseline vaccination rates than the provinces it has already entered22 leads us to the conclusion that VillageReach's cost-per-child-vaccinated has increased. We have yet to fully nail down the details of this change, but we will post an update when we do.

Room for more funding as of February 2012

VillageReach shared FY 2011 and first quarter FY 2012 revenue data with us,23 as well as revenue projections for the remainder of FY2012 as of December 31, 2011.24

Our current room for more funding estimates are based on annual expected revenue of $980,000 in FY201225 and $850,000 in FY2013 and FY2014 (excluding contracts).26 These are estimates VillageReach has provided. VillageReach told us that only revenues that it believes with high confidence it will receive are included in "anticipated" revenue for FY2012.27 The revenue projections for FY2013 and FY2014 are based on rough estimates of individual and foundation giving.28

VillageReach's expected total costs for the project have been reduced slightly, but the number of provinces has been been cut in half and as a result VillageReach is now only budgeting through FY2014 instead of FY2015. Since VillageReach has reduced the budgeted timeframe for the project by a year, expected revenues for the project have been reduced by $850,000.29

VillageReach received $130,000 in GiveWell-influenced donations in the first quarter of FY 2012 (October to December 2011).30

We estimate that VillageReach has a funding gap of $1.44 million for its work in its first four provinces.31 This does not include room for more funding in the other four provinces where it has put further work on hold.32 Without further budget changes or additional revenues from sources that are not currently anticipated by VillageReach, we estimate that project will be out of money by mid-2013.33

Re-analysis of the Mozambique pilot project (2002-2007)

Through following VillageReach's current scale-up project in Mozambique, we have learned more about the challenges to high quality data collection in the country, and have sought to improve our understanding of the quality of the pilot project evaluation. In a separate report on the evaluation of the Cabo Delgado pilot project, we examine two pieces of evidence from the pilot project evaluation: changes in vaccine stockouts in clinics served by VillageReach and province-level changes in vaccination rates in Cabo Delgado and neighboring provinces.

Update on contract engagements

In our review of VillageReach, we focus primarily on its Mozambique program because we believe individuals' donations primarily support this program.34 VillageReach also implements programs in other countries as projects that are contracted with specific funders.

Our previous understanding of these projects was that spending on a project followed the signing of a contract in which a funder agreed to pay all costs of the project. VillageReach has told us that past contract engagements have been fully paid for by funding restricted for these engagements.35

After further discussion with VillageReach, it is now our understanding that the timeline for contract engagements is often not as straightforward as we had assumed. VillageReach staff, primarily those based in the US, may work on a variety of projects and programs, including the Mozambique program. VillageReach’s forecast of expenditure and related revenues (from foundations, individuals and contract sources) is based on its visibility into pending opportunities. The organization may have excess capacity of staff from time to time, as it gains and completes its work. As a result, unrestricted funds may need to be used to cover this excess capacity until new programs or contracts begin.36 Therefore, it is not clear that contract revenues and spending will always match. If VillageReach fails to win enough contract funding to cover its contract engagement budget, the effective cost of the Mozambique project may rise because unrestricted funds may be reallocated to cover the contract budget. VillageReach told us that it will reduce its contract budget if revenues are lower than expected, in order to limit using unrestricted funding.37 As of February 2012, VillageReach projected a $215,000 funding gap for contracts and a $125,000 funding gap for restricted foundation grants (these grants may be restricted to either Mozambique or non-Mozambique projects).38 VillageReach appears to have surplus from FY2011 for non-Mozambique projects.39

Correction to previous update

In our March 2011 update on VillageReach, we wrote:

We misinterpreted the study. The sample for the study was 12 to 23 month olds; under 12 month olds were not surveyed.40 The vaccination rate for "children below 12 months of age" referred to the percentage of children who were vaccinated before they turned a year old, not the percentage of current children under the age of 12 months who were vaccinated (since this group was not included in the sample).

Furthermore, because the study was carried out in April 2010, none of the children would have been vaccinated during VillageReach's pilot project.41

Therefore, our previous statement was incorrect.

Global vaccine supply

We noted previously that there was a global shortage of the pentavalent vaccine, due to the World Health Organization decertifying a supplier. In August 2011, a second supplier was decertified.42 In January 2012, VillageReach told us that the global supply problems had been resolved, but that there had been difficulties in shipping vaccines within Mozambique (from the central level to the provincial level) due to insufficient space on domestic airlines.43

March 2012 sources

• Beale, John. VillageReach Director of Strategic Development. Email to GiveWell, March 4, 2011.
• Beale, John. VillageReach Director of Strategic Development. Email to GiveWell, February 10, 2012.
• Beale, John. VillageReach Director of Strategic Development. Phone conversation with GiveWell, February 16, 2012.
• Beale, John, and Leah Barrett. VillageReach Director of Strategic Development and Program Manager. Phone conversation with GiveWell, January 12, 2012.
• Dos Santos, A. P. 2010. Relatório do inquérito provincial de cobertura vacinal: Maputo. VillageReach asked that we keep this document confidential.
• GiveWell. Changes in VillageReach budget (February 2012) (XLS).
• GiveWell. Incremental children vaccinated (February 2012) (XLS).
• Mozambique National Institute of Statistics. Multiple indicator cluster survey (2008) (PDF).
• VillageReach. Budget variance report (FY2011).
• VillageReach. Contract engagements (2011). VillageReach asked that we keep this document confidential.
• VillageReach. Health system strengthening in Mozambique (July 2010) (PDF).
• VillageReach. Health system strengthening in Mozambique: Update (November 2011) (PDF).
• VillageReach. Mozambique budget change (February 2012). We have not yet received permission to post this document.
• VillageReach. Mozambique budget with GiveWell additions (February 2011, updated April 2011) (XLS).
• VillageReach. Mozambique budget with GiveWell additions (November 2011, updated February 2012) (XLS).
• VillageReach. Status of VillageReach work following 1-­year evaluations (January 2012). VillageReach has not yet received permission from the government to publish this document.
• VillageReach. Tentative schedule for VillageReach evaluation activities (August 2011) (DOC).
• VillageReach. Tentative schedule for VillageReach evaluation activities (January 2012) (PDF).
• VillageReach. Vaccination coverage baseline survey for Cabo Delgado province (PDF).
• World Health Organization. Panacea Biotec DTP-based combination and monovalent hepatitis B Vaccines delisted from WHO list of prequalified vaccines. http://www.who.int/immunization_standards/vaccine_quality/DTP_mono_hepb_... (accessed February 22, 2012). Archived by WebCite® at http://www.webcitation.org/61ta7u09C.

August 4, 2011 – progress in Mozambique and update on contract engagements

August 2011 summary

• By province:
  • Cabo Delgado: We discussed the baseline vaccination coverage survey and 6-month progress report from Cabo Delgado in the previous update. There have been no additional updates due from the province since then.
  • Niassa: In this update, we discuss the baseline vaccination coverage survey and 6-month progress report from Niassa. The quality of data in these reports is below what we consider necessary to accurately track VillageReach's progress over time or to evaluate its impact at the end of the project. In addition, we are concerned about the progress to date in Niassa where there are several data points indicating problems there (more).
  • Maputo: In this update, we discuss the baseline vaccination coverage survey from Maputo (more). The baseline survey from Maputo found extremely high vaccination rates (significantly above the levels VillageReach used when it estimated the impact of the program). No progress reports from the province have yet been due.
• VillageReach has provided a schedule of its planned monitoring and evaluation activities for June 2011-June 2012. By August 2011, it expects to have completed a health center baseline survey in Maputo (June), one-year progress evaluations in Cabo Delgado (July) and Niassa (August), and a health center baseline survey and cost study in a fourth province, Gaza (August), though it states that "evaluation reports or results of the field work would not be anticipated until at least one month following data collection activities to ensure adequate time for data entry and analysis." It expects to produce progress reports from each province every 6 months.44
• VillageReach reports that it has seen improvements in the supply of the pentavalent vaccine (more), and we will expect to see this reflected in future progress reports. We discussed in the previous update that there were shortages of this vaccine caused by a global supply disruption.
• VillageReach continues to contract with specific parties to carry out projects in different parts of the world. VillageReach has told us that all of these projects are fully funded, but we have not seen an expense breakdown by project to confirm this.
• VillageReach's room for more funding and future plans have not changed significantly since our last update (more).
• We are in the process of asking questions about of the quality of the data VillageReach collected during its pilot project. The poor quality of early data from Niassa has made us question whether we have sufficient detail on how well VillageReach was able to collect data in the past (more).

Cabo Delgado

No additional progress reports were expected since the last update. VillageReach told us that it is collecting monthly data in the province but that it is only able to share progress reports with us every 6 months, due to government restrictions on data sharing.45 We expect the next progress report from Cabo Delgado in, or soon after, August 2011.46

VillageReach told us that the issue discussed in the previous update regarding delays in the flow of funds to the provincial government has been resolved and distributions are going ahead as planned. It told us that the delays were caused by the flow of funding to the government.47 VillageReach has not received reimbursement for the funds it had lent the government, but told us that it still expects to.48

Niassa vaccination coverage baseline and progress to 6 months

Baseline vaccination coverage survey

The vaccination coverage baseline survey in Niassa covered 6 districts selected by the government "to represent 3 well-performing and 3 poorly-performing districts"49 (and is representative within these districts).50 We are concerned that these districts may not be representative of the province as a whole, and that the future evaluation of VillageReach's impact in the province could be inaccurate as a result. VillageReach told us that it will compare the baseline results with 2011 Demographic and Health Survey (DHS) data (expected to be available in late 2012 or early 2013) as a check of their accuracy. It also told us, "We consulted with Mark Kane [an epidemiologist and vaccination expert who originally advised us on our initial program evaluation] who determined that it meets our provincial level baseline needs."51

First 6 months of operation

• The program is currently operating in only part of the province (6 out of 16 districts).52 VillageReach told us that this is a result of lack of government funding for the project, and that it plans to expand to the full province in the future.53
• VillageReach reports that monitoring data, including on the key indicator vaccine stock outs, has been very incomplete and of poor quality.54 Lacking this data for the early stages of VillageReach's involvement in the region will make it difficult to assess the effect of the program in the future.
• Distributions in the first 6 months of the program were often late because government per diems for field staff were delayed.55 VillageReach told us that since the progress report was written, "There is improvement in the timing of getting the per diems, but there is still room for improvement."56
• VillageReach told us that progress has been slower than in Cabo Delgado because (a) Niassa lacks external funding to support the government's costs for the project (Cabo Delgado's government receives funding from a private foundation to support the project); (b) Niassa does not have experience with the program, unlike Cabo Delgado which was the setting of the pilot project; and (c) Niassa is less densely populated than Cabo Delgado. We have asked VillageReach for more information on its current expectations for progress in each province they are working in and plan to work in, and VillageReach told us that it will provide this in the future.57 We expect an outlook on Gaza, VillageReach's next province in early August.58

Note on quality of the Niassa 6-month report

• Significant portions of the Niassa 6-month progress report we first received appear to be copied and pasted from the Cabo Delgado 6-month report.59 When we asked VillageReach about this, it told us (a) these reports were prepared for the provincial governments and the message VillageReach was communicating to each was similar;60 (b) the Niassa report was written in Portugese and translated into English and Cabo Delgado report was used as a template for this.61 We have since received a re-translated report, as well as the original Portuguese report; translation error seems to be responsible for one of the examples we found, but not the others.

  We agree with VillageReach that these progress reports may not, on their own, be the best way for VillageReach to communicate its progress to us, and VillageReach has offered to "provide a summary written view of our opinions on the progress and challenges" with each future progress report, though we do not yet know if VillageReach will clear these summaries for public posting because of issues related to its relationship with the Mozambique government and what information it can share publicly.

• As was true with the Cabo Delgado 6-month report discussed in the previous update, the Niassa 6-month report does not present a chart of each of VillageReach's core indicators; for example, rather than showing the percentage of health centers with a stock out of at least one vaccine (which is VillageReach's official definition of its "stock outs" indicator), it separates stock outs by individual vaccine and does not include data on non-vaccine supplies. We previously requested that VillageReach chart all standard indictors and we have not seen progress on this; VillageReach told us that it intends to update this in future reports.62

We expect the next progress report from Niassa in, or soon after, September 2011.63

Maputo baseline vaccination coverage survey

The government of Mozambique conducted a survey of vaccination rates in the province of Maputo in 2009,64 which VillageReach plans to use, together with the 2011 Demographic and Health Survey results, as the baseline vaccination coverage survey for evaluating its work in the province.65

Based on the baseline study, current vaccination rates in Maputo seem extremely high66 - significantly above the levels VillageReach used when it estimated the impact of its program (VillageReach has not been granted permission by the government to share details publicly).67 Given the high starting rate of vaccination in the province, we expect the "cost per additional child vaccinated" in Maputo to be substantially higher than previously expected.

VillageReach told us that it chose to work in Maputo because the province expressed a desire to work with it and that it believes there is room for improvement in areas other than vaccination rates.68 We have not seen evidence that VillageReach has the ability to observe and estimate humanitarian effects of improvements other than increases in vaccination rates.

We expect a progress report from Maputo in, or soon after, January 2012.69

Update on the pentavalent vaccine shortage

In our previous VillageReach update, we noted that there was a global shortage of the pentavalent vaccine, which may have partially caused stock outs of the vaccine in the provinces in which VillageReach is active. When we asked VillageReach whether this situation had changed, it told us that it had seen improved supplies in Niassa and Cabo Delgado.70 VillageReach told us that the government has not given it permission to share data on stock levels.71 We will expect to see this improvement reflected in future progress reports.

Room for more funding and future plans

We estimate that, based on conservative revenue projections, VillageReach has a funding gap for its Mozambique project of about $965,000. We estimate that VillageReach has sufficient funding for the Mozambique program through 2013, but that, with current projections of future funding and costs, it would run out of money in mid-2014.72

In April 2011, we estimated VillageReach's funding gap at $1.03 million, but we later discovered errors in our calculation of secured and expected revenues (details in the footnote).73 We have tracked about $163,000 to VillageReach since mid-April (as of July 19, 2011); we underestimated VillageReach's funding gap by about $98,000 in April.

We have not requested an update on expected costs for the Mozambique project. Our current estimate of VillageReach's revenues is based (a) VillageReach's secured and anticipated funding for FY2011 as of March 31, 2011; (b) VillageReach's estimate of expected funding from individuals and foundations in future years (excluding funding resulting from a GiveWell recommendation); and (c) funds we tracked to VillageReach between April 1 and July 15, 2011.

Additional provinces:

• VillageReach plans to begin work in Gaza province in September 2011,74 following a costing study and health center survey in August.75 VillageReach plans to do a vaccination coverage survey in 2012, and expects also to have data from 2010 and 2011 surveys in the province (conducted by the government).76 We have not yet seen the data from 2010 or 2011 (2011 data is not expected until late 2012 or early 2013).77 On July 29, 2011, VillageReach told us that it would have an update on Gaza to share within a week.78 We expect this update to include progress toward baseline health center and costing surveys and program preparations, as well as an outlook on how similar the program in Gaza is likely to be to the pilot program in Cabo Delgado.
• We previously calculated that VillageReach needed to raise $200,000 in unrestricted revenue in 2011 in order to start working in a fifth province, Inhambane, by early 2012, and that it had achieved this as of April 2011. In July 2011, VillageReach told us that a baseline study was not yet scheduled for the province, but that discussions with the province, which were expected in summer 2011,79 were on schedule.80

Update on contract engagements

In our review of VillageReach, we focus primarily on its Mozambique program because we believe individuals' donations primarily support this program.81 VillageReach also implements programs in other countries as projects that are contracted with specific funders. VillageReach told us that all current contract engagements are fully funded by restricted funding.82

In our main review of VillageReach, we discuss a project in South Africa that VillageReach expected to support, in part, with non-contract funding. VillageReach told us that this project has been cancelled.83

VillageReach also told us that it is no longer financially supporting the VidaGas project.84

We requested a full organization budget broken down by project as we have seen from VillageReach in the past, but it told us that it could not provide us with that breakdown because it had recently revised its budgeting system.85 Therefore, we have not been able to verify how unrestricted funding is spent.

Questions about pilot project data

In the course of discussing the problems with data collection in the first few months of the program in Niassa, some questions were raised about the quality of the data collected in VillageReach's pilot project. We have not yet explored these questions and plan to do so in the next few months. We will post updates on this page.

August 2011 sources

• Beale, John. VillageReach Director of Strategic Development. Email to GiveWell, May 27, 2011.
• Beale, John. VillageReach Director of Strategic Development. Email to GiveWell, June 29, 2011.
• Beale, John. VillageReach Director of Strategic Development. Phone conversation with GiveWell, July 14, 2011.
• Beale, John, and Leah Barrett. VillageReach Director of Strategic Development and Program Manager. Phone conversation with GiveWell, May 12, 2011.
• Beale, John, and Leah Barrett. VillageReach Director of Strategic Development and Program Manager. Phone conversation with GiveWell, July 15, 2011.
• Dos Santos, A. P. 2010. Relatório do inquérito provincial de cobertura vacinal: Maputo. VillageReach asked that we keep this document confidential.
• VillageReach. Budget (FY 2011 to March 31, 2011). VillageReach asked that we keep this document confidential.
• VillageReach. Cabo Delgado Dedicated Logistics System six-month process evaluation (February 2011). VillageReach asked that we keep this document confidential.
• VillageReach. Comments on GiveWell's VillageReach Update (July 2011).
• VillageReach. Contract engagements (2011). We have requested permission to post this document.
• VillageReach. Health system strengthening in Mozambique (July 2010) (PDF).
• VillageReach. Mozambique budget with GiveWell additions (February 2011, updated July 2011) (XLS).
• VillageReach. Niassa Dedicated Logistics System six-month process evaluation (February 2011). VillageReach asked that we keep this document confidential.
• VillageReach. Tentative schedule for VillageReach evaluation activities (PDF).
• VillageReach. Vaccine coverage and vaccine and rapid diagnosis tests logistics study: Niassa baseline survey (July 2010) (PDF).

March 16, 2011 - Leadership update

We have discussed the issue alluded to in this VillageReach blog post with VillageReach staff, and do not find it to be cause for concern.

March 7, 2011 - Progress in Mozambique

March 2011 summary

• VillageReach is currently active in Cabo Delgado, Niassa, and Maputo, and has provided a six-month assessment and survey of baseline vaccination coverage for Cabo Delgado.
• The overall expansion is substantially behind the schedule set out in July 2010 (on time in Cabo Delgado but six to twelve months behind in other provinces) for funding-related reasons (though revenue in 2010 closed the entire "stretch funding gap" for the year, the vast majority of that revenue was not received or anticipated until December).
• In Cabo Delgado, data collection appears to have improved, but there are not yet other signs of improvement in the health system's performance. This is consistent with the expected trajectory of VillageReach's three-year involvement in a province, and we do not find it to be a major cause for concern.
• There has been a global supply disruption in a key vaccine that is expected to be resolved by the third quarter of 2011.
• VillageReach had a total of about $2 million in unrestricted revenue in 2010, of which a little over $1 million can be attributed to GiveWell's recommendation.
• We believe that this $1 million has been key to VillageReach's being able to plan to expand its program into the provinces of Gaza and Inhambane.
• We believe that VillageReach has about $1.4 million in room for more funding; these funds would allow it to commit to its full eight-province expansion plan.

Performance since last update

VillageReach is active in Cabo Delgado, Niassa, and Maputo.86 Currently:

• In Cabo Delgado, VillageReach has completed the baseline survey and has had its full program running since mid-2010.
• In Niassa, VillageReach has completed baseline survey and its full program has been running since the beginning of 2011.
• In Maputo, VillageReach is working on baseline survey and other preparations and intends to begin full implementation of its program in mid-2011.

The overall expansion is substantially behind the schedule set out in July 2010 (on time in Cabo Delgado but six to twelve months behind for other provinces).87 Reasons for the delay are partly related to dialogue with the government but appear primarily due to insufficient funding (though revenue in 2010 closed VillageReach's entire "stretch funding gap" for the year, the vast majority of that revenue was not received or anticipated until December).88

VillageReach has provided new information on progress in Cabo Delgado, the first province where it has become active. (It also states that a baseline-coverage survey has been done in Niassa89 but is not yet public.)

• Data collection seems to have improved,90 despite a major data collection issue in October that VillageReach has informed us may have been due to an ongoing election.
• Other indicators, including the particular important "stock out" indicator (VillageReach aims to reduce the frequency with which health centers run out of vital medical supplies), don't show noticeable improvement.91 We don't find this cause for major concern, as VillageReach's successful (in our view) pilot project also didn't show major improvement at the six-month mark.92
  
• A survey has been conducted in Cabo Delgado to assess the existing level of vaccination coverage. Overall immunization has fallen only slightly since the 2008 conclusion of VillageReach's work in this province, but it has fallen significantly for children under the age of 12 months.93 The fact that the overall coverage rate (which includes some people immunized under VillageReach's model) has fallen slightly while the under-12-month coverage rate (only those immunized recently) has fallen a lot is consistent with the hypothesis that the switch away from VillageReach's model has led to inferior immunization coverage (though the previous, more effective system still affects the overall coverage rate).
• There appears to have been a problem with the flow of funds from a funder to the government. In the short run this has resulted in VillageReach's lending money to the government to cover for a delay.94 This problem does not appear to directly involve VillageReach, and VillageReach representatives told us they are unclear on the exact nature of the problem, but they stated that they expect a relatively quick resolution (within a few months) and have included a mention of the issue in the 6-Month Evaluation in the hopes of raising the profile of the problem and raising the odds of its quick resolution.
  
• Supply of pentavalent vaccine (which includes diphtheria-tetanus-pertussis, the vaccinations we focus on in our calculation of lives saved) has become disrupted owing to a global supply problem. VillageReach provided this quote from a GAVI (The Global Alliance for Vaccines and Immunisation) representative:
  
  Yes, there is a global shortage of pentavalent vaccines which is expected to be resolved by Q3 of 2011. Countries may be required to use their buffer stocks this year. UNICEF is trying their best to manage the situation. The shortage has occurred due to withdrawal of WHO pre-qualification status of one of the suppliers last year due to quality issues.

  Unfortunately, this supplier has not been successful in regaining that status. New suppliers have been pre-qualified by WHO this year but they require lead time to produce the volume necessary to match the demand. So, in short, Mozambique is not the only country affected but all GAVI countries are affected by this.

  This issue may be partly responsible for the particularly high stock-out rates for this particular vaccine.95

Notes on the presentation and quality of the materials provided by VillageReach:

• Overall, we found the provided materials to be exceptionally clear and straightforward about VillageReach's progress, struggles, and plans.
• The six-month report does not present a chart of each of VillageReach's core indicators; for example, rather than showing the percentage of health centers with a stock out of at least one vaccine (which is VillageReach's official definition of its stock-out indicator96), it separates stock outs by individual vaccine and does not include data on non-vaccine supplies.97 VillageReach made the case to us over the phone that the charts shown are more informative at this stage of the project, while still making it clear that there has not been progress on key indicators to date. We find this a reasonable argument, but we have requested that in the future VillageReach chart all standard indicators along with whatever other charts it finds to be informative.

Plans from here and room for more funding

• VillageReach is currently active in the provinces of Cabo Delgado, Niassa, and Maputo.98
• It is currently in discussions with the Gaza province to begin work there.99
• VillageReach plans to begin talks with the Inhambane province once it has completed a key hire.100 VillageReach states that "in order to commit to starting work in Inhambane in early FY2012 (ie. Jan-March 2012 time frame), we would like to recognize funding by December 2011 that supports the program for the subsequent 12 months."101 According to our calculations, this would require that VillageReach receive about $200,000 in unrestricted 2011 revenue,102 which seems likely: VillageReach raised about $1 million in calendar year 2010 from non-GiveWell sources,103 and its forward-looking projection is about $850,000 per year in revenue from non-GiveWell sources.104
• Our understanding is that VillageReach's work in Gaza and Inhambane has been made possible by donations coming as a result of GiveWell's recommendation.105 VillageReach received about $2 million in unrestricted revenue in 2010, of which a little over $1 million can be attributed to GiveWell's recommendation.106
• VillageReach hopes to expand to three more provinces, beginning in late 2012.107 Doing so will require about $3.6 million in additional funding;108 incorporating conservative projections of incoming revenue, we estimate an overall funding gap of about $1.4 million.109 VillageReach states that additional revenue of $500,000-$1,000,000 would accelerate discussions with these additional provinces.110
• We believe that VillageReach has "room for more funding" of approximately $1.4 million in above-expectations revenue in 2011,111 and we would cease recommending it to donors if it reached that threshold.

Other notes

In discussions with VillageReach, we have learned more about the factors that go into expanding to a new province, aside from funding. In addition to funding its own operations and getting sign-off from the local health authorities, VillageReach must ensure that there is sufficient government funding earmarked for helping to carry out its program112 and that it has sufficient capacity of its own, which can involve a hiring process.113 VillageReach representatives have told us that these factors are relatively straightforward to address when funding on its own end is available, but these issues are worth keeping in mind as potential non-financial bottlenecks to expansion.

March 2011 sources

• Beale, John. VillageReach Director of Strategic Development. Email to GiveWell, March 4, 2011.
• GiveWell. VillageReach.
• VillageReach. Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011). VillageReach shared this document with GiveWell, but the document is not yet cleared for public consumption. We expect it to be cleared in the future.
• VillageReach. Mozambique budget (XLS).
• VillageReach. Mozambique budget with GiveWell additions (XLS).
• VillageReach. Mozambique funding gap (XLS).
• VillageReach. Mozambique project plan (XLS).
• VillageReach. Revenue attribution (2010) (XLS).
• VillageReach. "Room for more funding" analysis (DOC).
• VillageReach. Vaccination coverage baseline survey for Cabo Delgado Province (PDF).

• 1.

  VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 1.

• 2.

  VillageReach, "Health System Strengthening in Mozambique: Update (November 2011)," Pg 3.
  "A very unfortunate consequence of the funding challenges and interruptions to distributions is an interruption to the data. The data collected as part of the DLS relies on visits to the health center, and without distributions there are no visits." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 3.

• 3.

  "The main finding of the Cabo Delgado evaluation was that there was a willingness and capacity to implement the DLS in the province, but they faced major funding problems in 2011 that led to the disruptions since June 2010...The arrangement in Cabo Delgado was that the DLS would be implemented with 3 partners: DPS provided funds management and implementation, funds from EGPAF [Elizabeth Glaser Pediatric AIDS Foundation], and technical assistance from VillageReach. However, the flow of funding from EGPAF to DPS was halted due to certain conditions in the sub-agreement between DPS and EGPAF." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pgs 1-2.

• 4.

  "We decided to allocate some backup funds for per diems in the province. These funds are to be used when EGPAF funds aren't available and are managed by VillageReach. This cost-sharing approach still demands that DPS fund the costs of fuel, which is 38% of the monthly operational cost. Monthly per diems cost $2,856 and fuel costs $1,800. DPS continues to pay the staff and indirect costs." VillageReach, "Status of VillageReach Work Following 1-Year Evaluations (January 2012)," Pg 2. 2856/(2856+1800) = 61.3%

• 5.

  "DPS Cabo Delgado happily accepted the cost-sharing approach and have been using VillageReach funds for the last four months of distributions, which happened to all health centers and on time." VillageReach, "Status of VillageReach Work Following 1-Year Evaluations (January 2012)," Pg 2.

• 6.

  "There continue to be challenges with collecting data and entering it in vrMIS. The result of these problems is continued limited analysis of the system in the province." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 3.

• 7.

  "To improve activities in the health centers, we worked with DPS to implement a pilot one-on-one health worker training in 2 districts...Verbal reports of the training indicate that it has been very successful, we plan to continue the training again later this month.
  Another response to the quality problems in Niassa has been our hiring of additional support for the province." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 3.

• 8.

  "In 2011, there were interruptions to distributions in Niassa due to funding constraints...Throughout the year, the distributions have continued to be only to the health centers in 6 out of the 16 districts because of limited funding...
  Our agreement is that if DPS doesn't have the funds to cover all districts and all health centers, they can ask VillageReach to pay for the per diems and they will pay for the fuel. This maintains their current financial commitment while expanding to all provinces. This arrangement started in December. The cost of this arrangement is $2,665/month for per diems paid by VillageReach and $2,415/month for fuel." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pgs 3-4.

• 9.

  "The distributions started in November for Gaza and December for Maputo. In both provinces the distributions went to only one zone..." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 4.

• 10.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities (August 2011)."

• 11.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities (January 2012)."

• 12.

  "The distributions started in November for Gaza and December for Maputo. In both provinces the distributions went to only one zone..." VillageReach, "Status of VillageReach Work Following 1-­Year Evaluations (January 2012)," Pg 4.

• 13.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities (August 2011)."

• 14.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities (January 2012)."

• 15.

  See GiveWell, "Changes in VillageReach Budget (February 2012)."

• 16.

  VillageReach, "Mozambique Budget Change (February 2012)."

• 17.

  John Beale and Leah Barrett, phone conversation with GiveWell, January 12, 2012.

• 18.

  GiveWell, "Changes in VillageReach Budget (February 2012)."

• 19.

  VillageReach, "Health System Strengthening in Mozambique: Update (November 2011)," Pg 4.

• 20.

  See GiveWell, "Changes in VillageReach Budget (February 2012)."

• 21.

  GiveWell, "Changes in VillageReach Budget (February 2012)."

• 22.

  Mozambique National Institute of Statistics, "Multiple Indicator Cluster Survey (2008)."

• 23.

  VillageReach, "Budget Variance Report (FY2011)."

• 24.

  John Beale, email to GiveWell, February 10, 2012.

• 25.

  John Beale, email to GiveWell, February 10, 2012.

• 26.

  John Beale, email to GiveWell, March 4, 2011.

• 27.

  John Beale, phone conversation with GiveWell, February 16, 2012.

• 28.

  John Beale, email to GiveWell, March 4, 2011.

• 29.

  VillageReach, "Mozambique Budget with GiveWell Additions (November 2011, updated February 2012)."

• 30.

  Data from GiveWell's internal donation records.

• 31.

  VillageReach, "Mozambique Budget with GiveWell Additions (November 2011, updated February 2012)," Cell G24.

• 32.

  See GiveWell, "Changes in VillageReach Budget (February 2012)."

• 33.

  VillageReach, "Mozambique Budget with GiveWell Additions (November 2011, updated February 2012)," Cell E23.

• 34.

  "VillageReach has also contracted with other organizations to improve logistical systems in other countries, but generally hasn't used, and doesn't plan to use, funding from individuals for projects outside of Mozambique (with a few exceptions discussed below)." See our April 2011 review of VilllageReach.

• 35.

  VillageReach, "Contract Engagements (2011)."

• 36.

  John Beale, phone conversation with GiveWell, February 16, 2012.

• 37.

  John Beale, phone conversation with GiveWell, February 16, 2012.

• 38.

  John Beale, email to GiveWell, February 10, 2012.

• 39.

  VillageReach, "Mozambique Budget with GiveWell Additions (November 2011, updated February 2012)," Cell C34.

• 40.

  "The total sample size for the vaccine coverage survey was 211 children (planned 210 children) between the age of 12 months and 23 months." VillageReach, "Vaccination Coverage Baseline Survey for Cabo Delgado Province," Pg 11.

• 41.

  "In the 2010 survey only the 12-23 month age group was covered as they represent an age group that have been potentially vaccinated since the end of the first EPI programme supported by VillageReach with the DPS Cabo Delgado." VillageReach, "Vaccination Coverage Baseline Survey for Cabo Delgado Province," Pg 17.

• 42.

  World Health Organization, "Panacea Biotec DTP-Based Combination and Monovalent Hepatitis B Vaccines Delisted from WHO List of Prequalified Vaccines."

• 43.

  John Beale and Leah Barrett, phone conversation with GiveWell, January 12, 2012.

• 44.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 45.

  John Beale and Leah Barrett, phone conversation with GiveWell, July 15, 2011.

• 46.

  VillageReach expects to conduct the evaluation in July and says that reports "would not be anticipated until at least one month following data collection activities."

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 47.

  John Beale and Leah Barrett, phone conversation with GiveWell, May 12, 2011.

• 48.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 2.

• 49.

  "In the Cabo Delgado baseline, we selected the sample to represent the provincial level coverage rate. In the Niassa baseline, DPS wanted to get district-level (ie. sub-province) coverage rates but also had some restrictions in their own funding, limiting a larger study…We could only fit 6 districts in the budget. We consulted with Mark Kane (an epidemologist and vaccination expert who originally advised us on our initial program evaluation) who determined that it meets our provincial level baseline needs. DPS selected the 6 districts to represent 3 well-performing and 3 poorly-performing districts. The program is operating in 3 of the districts from the baseline and 3 districts not included in the baseline." John Beale, email to GiveWell, June 29, 2011.

• 50.

  "The survey was undertaken in six districts. In each of the districts a sample of ten hoursholds and ten villages that were selected on a random basis using probability proportional to size. The sample is representative of the districts covered under the survey." VillageReach, "Vaccine Coverage and Vaccine and Rapid Diasgnosis Tests Logistics Study: Niassa Baseline Survey (July 2010)."

• 51.

  John Beale, email to GiveWell, June 29, 2011.

• 52.
  • "The program is operating in 3 of the districts from the baseline and 3 districts not included in the baseline." John Beale, email to GiveWell, June 29, 2011.
  • "Due to lack of various resource—human, financial and material that does not allow a distribution of quality, Dedicated Logistics System should be expanded to the remaining 10 districts that are currently doing the distributions themselves." VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 16.
• 53.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 2.

• 54.
  • "The percentage of health units reporting data from September through December was less than 20% of health centers in Niassa. This has serious implications for the validity of the data and does not provide a representative sample of health centers served by DLS." VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 8.
  • "This reflects the fact that some of the health units have and have complete forms, the quality of the data overall is very poor. Each health unit had differences in the numbers provided in the register, A01 form, and A03 form." VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 7.
• 55.

  "In four of the six months studied, field coordinators left for the distributions without receiving per diems and were forced to use their own money to cover expenses. In January, the delay in the availability of funds was so great that no vaccines were distributed during the month." VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 5.

• 56.

  John Beale, email to GiveWell, June 29, 2011.

• 57.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 2.

• 58.

  John Beale, email to GiveWell, July 29, 2011.

• 59.

  Examples of copied text:

  • The "lessons learned" sections in the two reports are nearly identical, with the only difference being one additional lesson listed in the Cabo Delgado report. In light of the implementation problems reported in Niassa, it is not clear whether is appropriate to conclude for both provinces that they have "demonstrated the capacity to implement the vaccine logistics system." (VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 16 and VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 14.)
  • The report indicates that there was some checking of data for consistency across reporting forms and for complete reporting, but the description of what they found and error rates are the same in the two reports, bringing into question whether this actually occurred in both provinces.
    • "Through observation of these forms, the field coordinators rarely completely filled out each section of the form with only approximately 10% of the forms completely filled in....Some sections of the health visit form are consistently and accurately filled in. These include the medical supply inventory table (Section 3) and the EPI vaccination tables (Section 9, 10, and 11). There is no indication that this data may be inaccurate, and therefor this data is the most valuable for analysis." VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 5 and VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 5.
    • "This reflects the fact that some of the health units have and complete the forms, while the quality of the data overall is very poor. Each health unit had difference in the numbers provided in the register, A01 form, and A03 form." VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 8 and VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 7.
  • In two examples, the report is internally inconsistant, reporting more positive results in the narrative text (copied from the Cabo Delgado report) and less positive results in charts/tables (which we assume contain the actual data from Niassa). (Follow page references are from VillageReach, "Niassa Dedicated Logistics System Six-Month Process Evaluation (February 2011).")
    Example 1: (we later found out that this was an error made during translation)
    • Same as Cabo Delgado report: "The data completeness measured as health units reporting data by month was: 83%, 52%, 60%, 90%, 22%, and 99% from June through November, respectively." Pg 4.
    • The chart on the next page of "% of Health Centers Reporting Data" shows data for September to January: 29%, 16%, 28%, 4%, and 0%.

    Example 2:

    • Same as Cabo Delgado report: "Distributions of vaccines to health facilities are occurring on a monthly schedule with minimal delays." Pg 4.
    • Other parts of the report indicate that delays were in fact substantial: "In January, the delay in the availability of funds was so great that no vaccines were distributed during the month." Pg 5. The average delay over the period covered by the report was 10.33 days. Pg 4.
• 60.

  John Beale, email to GiveWell, July 29, 2011.

• 61.

  John Beale and Leah Barrett, phone conversation with GiveWell, July 15, 2011.

• 62.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 3.

• 63.

  VillageReach expects to conduct the evaluation in August and says that reports "would not be anticipated until at least one month following data collection activities." VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 64.

  Dos Santos, et al. 2010.

• 65.

  John Beale and Leah Barrett, phone conversation with GiveWell, May 12, 2011.

• 66.

  Dos Santos, et al. 2010, Pg 13. (This document is in Portuguese. We used the Google translation service to examine it.)

• 67.

  In its plan for the project, VillageReach based its estimates of "incremental children vaccinated" on two assumptions: (1) "Children vaccinated with current system is based on estimates from coverage rate studies for Cabo Delgado and Niassa. All other provinces [including Maputo] based on national rate of 72%"; (2) "Total children forecasted to be vaccinated with new system is 80% in the first year, 85% in the second year, and 90% in the third year." VillageReach, "Health System Strengthening in Mozambique," Pg 7.

• 68.

  VillageReach believes there is room for improvement in Maputo in:

  1. "The capacity of health systems to distribute commodities effectively such that vaccines are administered correctly, per required regimens, to achieve proper efficacy (as an example, despite the high vaccination coverage rates noted, the Maputo DPS [baseline] study also noted that only a minority of children were vaccinated correctly, at the correct age – we saw similarly insufficient vaccination quality in the baselines we conducted for Niassa and Cabo Delgado)."
  2. "The cost effectiveness of the distribution of commodities. As we cited in our original cost analysis in 2009, our methodology provides improvements in health worker capacity and reduces costs of transport."
  3. "Support for the distribution of other commodities. As an example and as we have cited previously, we’re also adding RDTs [rapid diagnostic tests] to distributions and will be tracking impact of their inclusion in the program."

  John Beale, email to GiveWell, July 29, 2011.

• 69.

  VillageReach expects to conduct the evaluation in December 2011 and says that reports "would not be anticipated until at least one month following data collection activities." VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 70.

  John Beale and Leah Barrett, phone conversation with GiveWell, July 15, 2011.
  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 4.

• 71.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 4.

• 72.

  VillageReach, "VillageReach, "Mozambique Budget with GiveWell Additions (February 2011, updated July 2011)."

• 73.
  • VillageReach sent revenue data that covered through February 9, 2011, but we mistakenly added revenues that we tracked since January 1, 2011. Therefore, we double counted some revenues between January 1 and February 9, 2011.
  • VillageReach provided an estimate of expected revenues for the Mozambique project in fiscal year 2011 as of February 9, 2011. We added these expected revenues to a projection that already included expected revenues for 2011.
• 74.

  We assume this based on a scheduled 6-month progress evaluation scheduled for March 2012. VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 75.

  VillageReach, "Tentative Schedule for VillageReach Evaluation Activities."

• 76.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 5.

• 77.

  VillageReach, "Comments on GiveWell's VillageReach update (July 2011)," Pg 3.

• 78.

  John Beale, email to GiveWell, July 29, 2011.

• 79.

  John Beale and Leah Barrett, phone conversation with GiveWell, May 12, 2011.

• 80.

  John Beale and Leah Barrett, phone conversation with GiveWell, July 15, 2011.

• 81.

  "VillageReach has also contracted with other organizations to improve logistical systems in other countries, but generally hasn't used, and doesn't plan to use, funding from individuals for projects outside of Mozambique (with a few exceptions discussed below)." See our April 2011 review of VilllageReach.

• 82.

  VillageReach, "Contract Engagements (2011)."

• 83.

  John Beale and Leah Barrett, phone conversation with GiveWell, May 12, 2011.

• 84.

  John Beale and Leah Barrett, phone conversation with GiveWell, May 12, 2011.

• 85.

  "We don't have that breakdown currently, as we recently revised our budget formats." John Beale, email to GiveWell, June 29, 2011.

• 86.

  VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 87.

  See VillageReach, "Mozambique Project Plan," which directly compares the previous and current schedules.

• 88.

  "VillageReach is active in Cabo Delgado, Niassa, and Maputo. The initial start last year in these provinces was slower than had been anticipated, primarily due to concerns about funding and some delays in dialogue with the government , but we now have good progress in all three provinces ... In the last six weeks of 2010 VillageReach received funding from individual donors well beyond expectations. Until this time it had not been clear what level of activity would be possible for the Mozambique program in 2011.

  Because of this lack of visibility, we had not pursed any new DPS discussions late least year, except for Gaza, where [a funder] had already made it clear that they can provide funding for the Gaza provincial governments activities. Note that it can take some months of discussion with a provincial government before permission to work in the province is received and finances and personnel are allocated by the government." VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 89.

  "Baseline assessments have been completed in both Cabo Delgado and Niassa and distributions are well underway." VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 90.

  See chart in VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 5.

• 91.

  See charts beginning at VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011).," Pg 11. The chart on stock outs is particularly important to the question of whether VillageReach is succeeding in improving the efficiency of the local health system.

• 92.

  See our report on VillageReach.

• 93.

  "There has been a reduction in vaccine coverage from 2008 to 2010 (children below 12 months of age) of nearly 18 percentage points, although the vaccine coverage including children above 12 months of age is only slight lover than the 92.65 in 2008, 89% in 2010." VillageReach, "Vaccination Coverage Baseline Survey for Cabo Delgado Province," Pg 31.

• 94.

  "A key requirement of the Dedicated Logistics System is that the monthly funding mechanism to support the field coordinators and the distribution schedule are in place and operating without the need for external support. The distribution costs are covered by a partnership with [a funder], which results in full funding for the previously identified operational needs. However, discussions during the evaluation revealed that there are some challenges in the flow of funds from [a funder] to DPS related to the larger agreement between the partners (i.e. not limited to Dedicated Logistics System activities) that affected the availability of per diems for the November distribution. Additionally, in February the funds were not available for the distributions and DPS relied upon a loan from VillageReach to support the costs. Prior to this problem, the funding of the monthly distributions was working well. Without fixing the problem, there is a significant risk to the operations and outcomes of the Dedicated Logistics System." VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 5.

• 95.

  VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 11.

• 96.

  "Stock outs: Reports the number and percentage of health units having a stock level of 0 (termed a stock-out event) of any particular vaccine, EPI-related supplies, or rapid diagnostic tests. This indicator measures how well the logistics system is functioning to supply the health unit with vaccines and associated commodities to meet communities’ need for a full-month. The target for stock outs is 0% of the health units reporting, but 5% is an acceptable level. We expect stock outs to increase up to 40% at the start of the implementation of the dedicated logistics system as demand for vaccines increases (due to increased supply). After 6 months of implementation, we expect stock outs to decrease to 5% or below." VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 21.

• 97.

  See chart at VillageReach, "Cabo Delgado Dedicated Logistics System Six-Month Process Evaluation (February 2011)," Pg 11.

• 98.

  VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 99.

  "We had not pursed any new DPS discussions late least year, except for Gaza, where [a funder] had already made it clear that they can provide funding for the Gaza provincial governments activities. Note that it can take some months of discussion with a provincial government before permission to work in the province is received and finances and personnel are allocated by the government." VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 100.

  "After Gaza, we will work on Inhambane next, with outreach to the government there expected once we have hired a new program manager; we have identified a strong candidate. We estimate the start of those discussions with Inhambane Province in early summer 2011." VillageReach, "'Room for More Funding' Analysis," Pg 2.

• 101.

  VillageReach, "'Room for More Funding' Analysis," Pg 2.

• 102.

  Costs for 2011-2012 including only the provinces of Cabo Delgado, Niassa, Maputo, Gaza, and Inhambane were a total of $1,771,717. "Mozambique Funding Gap" shows about $1,592,491 available, implying a remaining gap of $179,226. See VillageReach, "Mozambique Budget with GiveWell Additions," Cell H17 and VillageReach, "Mozambique Funding Gap."

• 103.

  VillageReach, "Revenue Attribution (2010)."

• 104.

  "We estimate approximately $850 in non-GiveWell unrestricted revenues/and or Moz-restricted revenues [ie. some foundations have this requirement, others do not] for this fiscal year, assuming $350-400k is from individuals and $450-500k comes from foundations." John Beale, email to GiveWell, March 4, 2011.

• 105.

  VillageReach states,

  "You asked us for our thinking as to what funding amounts trigger what levels of activity …

  • If VillageReach had raised $500,000 less in late 2010:
    • VillageReach would still be active in Cabo Delgado, Niassa and Maputo.
    • VillageReach would still be planning to pursue expansion in Gaza when the opportunity arises, probably a few months from now.
    • VillageReach would not be planning on pursuing expansion beyond Gaza until/unless significant further funding came in.
  • If VillageReach had raised $1,000,000 less in late 2010:
    • VillageReach would still be active in Cabo Delgado, Niassa and Maputo.
    • VillageReach would not be planning on any further expansion unless significant further funding came in and the program would have been at serious risk of ending prematurely."

  This is consistent with our analysis of VillageReach's budget, which is similar to the analysis we show in "Mozambique Budget with GiveWell additions." VillageReach, "'Room for More Funding' Analysis," Pg 2.

• 106.

  VillageReach, "Revenue Attribution (2010)."

• 107.

  "Note we would also expect to start work in Manica Province before the end of FY2012. Sofala and Tete would follow at the beginning of FY2013 (ie., late CY2012/early 2013)." VillageReach, "'Room for More Funding' Analysis," Pg 2.

• 108.

  VillageReach, "Mozambique Budget with GiveWell Additions," Cell H18.

• 109.

  VillageReach, "Mozambique Budget with GiveWell Additions," Cell G22.

• 110.

  "If we raise up to $500k more than is forecast by the end of CY2011: VillageReach will likely still pursue the program as presented in the project plan, although there might be some modest acceleration.

  If we raise $500k - $1M more than is forecast by the end of CY2011: VillageReach will likely accelerate planned discussions with the 7th and potentially 8th provinces.

  If we raise any amount beyond $1M more than is forecast by the end of CY2011:
  VillageReach may be able to accelerate discussions with the 8th province." VillageReach, "'Room for More Funding' Analysis," Pg 2.

• 111.

  In summary, VillageReach projects approximately $1 million in unrestricted, annual, non-GiveWell funding in the future. It requires $500,000 of this for general operations and $500,000 of this for its Mozambique project. Given this funding, VillageReach's Mozambique project would require a total of $1 million in addition to expected funding to fully fund the Mozambique expansion.

• 112.

  For example: "We had not pursed any new DPS discussions late least year, except for Gaza, where [a funder] had already made it clear that they can provide funding for the Gaza provincial governments activities. Note that it can take some months of discussion with a provincial government before permission to work in the province is received and finances and personnel are allocated by the government." VillageReach, "'Room for More Funding' Analysis," Pg 1.

• 113.

  "After Gaza, we will work on Inhambane next, with outreach to the government there expected once we have hired a new program manager; we have identified a strong candidate. We estimate the start of those discussions with Inhambane Province in early summer 2011." VillageReach, "'Room for More Funding' Analysis," Pg 2.