A fundamental hypothesis of the project is that the MVP package of interventions can be delivered at a modest cost. The needs assessment conducted by the UN Millennium Project estimated that achieving the MDGs would require local service delivery and community-based investments of approximately $US 120 per person per year (in 2005 USD) during the 10-year period from 2005 to 2015 (Sanchez et al., 2005; World Health Organization, 2003).
It is important to understand that the $120 is not an increment above a baseline level of spending. Rather it is the estimated total cost of the MDG package of interventions, some part of which is in place without the MVP. The MVP is therefore providing a financial “top up” to existing funds, with the aim of reaching a total of around $120 per person per year. The incremental “cost” of the MVP, is therefore not the full $120, but only the top up. Will we not know how big the top up is relative to the full $120, unless we do costing in the comparison areas, which at the moment is not included in our evaluation budget. We do know that the roughly $60 per person per year that the project spends is part of the top up.
It is also important to underscore that this $120 annual per capita figure does not reflect the entire cost of the MVP project. The $120 includes the costs of service delivery, im- plementation, and on-site management, including estimated values of in-kind donations. Off-site costs, comprising salaries and overhead for all scientific and support staff at the Earth Institute and Millennium Promise staff based in New York and at the regional MDG Centers in Dakar and Nairobi, are excluded from this cost assessment. These excluded
off-site staff are primarily involved in project design, implementation research, monitor- ing and evaluation, logistics, and fundraising. They are not involved in direct operations, so their costs should be considered a one-time cost to design and operate the project, rather than an ongoing cost of running an MVP-style project in a scale-up context.
4.10.1
Methodology
Costs are collected for the entire project area (MV1 and MV2), but a distinction between MV1 and MV2 costs will be drawn, allowing for the costs in the research village (MV1), where investments have been more heavily concentrated, to be distinguished and ana- lyzed separately. Due to the varying degree of detail in external stakeholders expendi- ture records, as well as the spillover effect of certain investments and the difficulty of isolating beneficiary groups, it will not be possible to distinguish perfectly between MV1 costs and MV2 costs for every intervention. In cases where estimations are necessary, all assumptions made will be recorded and clearly outlined in the final evaluation.
Of the estimated $120 per capita annual cost, the project (Millennium Promise) supplies $60 per capita, while national and local governments, external donors (including NGOs, multilateral organizations, and private donors), and the local community (mainly in kind as labor and material inputs) supply the rest (Figure 4.2) (UN Millennium Project, 2005). Understanding these inputs is critical to evaluating the success of the project in relation to the $120 per capita annual project target, and to assess scalability of project interventions (Sanchez et al., 2007). The nature and intensity of inputs is likely to differ substantially between clusters due to community needs, local disease profile, and local economic base. A full economic costing assessment, in line with established methods of social and health policy interventions (Catterall, 1985; Ahren, 1976; Pushpangadan, 1997; Rahman and Alam, 1987; Hutchinson, 1969) is underway in each project cluster. The aim of the as- sessment is to document the annual on-site costs of the project by site, stakeholder (see Figure 4.2), sector (see Figure 4.3), year, and within the MV1 only as well as the entire
Figure 4.2: Costing model by stakeholder (2005 USD) (Sanchez et al., 2005).
cluster.
Core project investments and external stakeholder investments are tracked via two com- plementary mechanisms. Core project expenditures (those made with funds that flow through the Millennium Promise bank account) are tracked and reported quarterly via the projects internal cost-tracking system. Expenditures made by external stakeholders (government, community, and other donors) within the sectors of education; health; agri- culture, animal husbandry, business development, and environment; and infrastructure, are collected periodically by local site team members.
A series of data collection templates have been created for each stakeholder operating within each project cluster (including both MV1 and MV2). There are approximately 20 government, donor, and community stakeholders per project cluster. All costs within the defined project sectors (see above) are collected. The costing data collected via these individual stakeholder templates are amalgamated with the core project costing data col- lected via the internal tracking system to form a single comprehensive costing database for each project cluster.
For contributions made in kind, all prices are documented using the standard cost impu- tation method recommended for multi-center interventions (Grieve et al., 2009; Schulen- burg, 2000; Wordsworth et al., 2005). This method involves establishing local unit costs for each in-kind contribution (e.g. daily wage rate in the case of labor contributions). These unit costs are then used along with qualitative data collected during key informant interviews to calculate a total cost for each contribution (e.g. daily wage rate × number of laborers × number of days worked) (Grieve et al., 2009; Schulenburg, 2000; Wordsworth et al., 2005).
4.10.2
Data management and analysis
After the costing data from all stakeholders have been collected and aggregated, the data will be archived and available for analysis. For the purposes of cross-site comparison, compatibility with core project expenditures, and measurement against the MVP cost- ing model, all expenditure amounts will be converted to 2005 US dollars using average annual exchange rates for each project year. Annual per capita expenditures will be cal- culated for each project cluster, research village (MV1), sector, and stakeholder, using the total cluster population (MV1 and MV2) as well as the research village population (MV1 only).
Analysis of the data will be focused around questions of sustainability, replicability, and scalability. Per capita costs by sector and stakeholder are essential to planning any project scale-up or replication.