Grant proposals

Roles of PIs, RMs, and RAs 

While principal investigators (PIs) will write the grant, research staff such as research associates (RAs) and research managers (RMs) may also be involved in writing the proposal, working on the budget proposal, and managing the grant/budget. More details on typical responsibilities are listed below. 

PIs: See this grant proposals checklist, and read sections of this resource as needed. While the institution will likely receive and manage the grant, the PI is ultimately responsible for designing, conducting, and reporting on funded research projects, as well as fiscal management of the sponsored project, including management of the project within funding limitations.

RMs: Read about the budget proposal, grant and budget management, and reporting. The RM may be asked to help plan how the project will be implemented (such as staffing assumptions and timelines), may be heavily involved with grant management, and asked to contribute to the final report. The RM may also be responsible for ensuring all roles (of PIs, RAs, grant managers) are clear. 

RAs: RAs will typically be uninvolved in writing the grant and budget but may be asked to contribute to the grant proposal through a literature review and to check budget assumptions. In practice, RAs are often responsible for managing and executing the grant and should thus understand the grant requirements and expectations. RAs may also be tasked with preparing periodic progress reports, managing the budget for field operations, keeping spending records, and making sure deadlines are met. This means understanding the project timelines, deliverables, budget line items, and payment conditions and cash flow (e.g., tranches, conditions for payment, deadlines for reporting and initiating spending). It is important to know who is the point person for support on these tasks. RAs should read: budgeting principles, budget justification, grant management, and reporting (while RMs and PIs are more likely to write the annual or final report, RAs may be called upon to contribute to the report, such as by producing tables).

Essential terms

• Prime award: The award from the original funder. The prime award also lists the requirements of the grant and timeline for deliverables. Note that some funders, J-PAL included, require that the IRB be approved before funding can be released. 
• Lead organization: Grants will also be made to one “lead” institution, typically where the lead PI is based. If, for example, co-PIs are based at other institutions, funds will be transmitted through those institutions through sub-awards. With J-PAL Initiative grants, MIT is the lead institution, and sub-awards are made to the PI’s university or the implementing J-PAL or Innovations for Poverty Action (IPA) office.
• Sub-awards and sub-contracts: A third party recipient of funds from a prime award. A sub-award is issued from the grantee of a prime award and includes the same elements of a prime award. A sub-award typically goes to a partner or organization that is substantively involved and shares similar project goals (such as the implementing J-PAL or IPA office). A sub-contractor is contracted under a more narrow role and can include, for example, a survey firm that is hired. Other third parties include vendors, or consultants who are contracted for narrow roles (such as a translation service or caterer) and do not necessarily share the same project goals. 
• Fringe: Employee benefits such as health insurance that the employee does not receive as salary but are part of the institution’s personnel costs. Fringe is often included when budgeting for staff (and especially PI or graduate student) time.
• Overhead: The expenses to “keep the lights on” at the office or institution receiving the grant. These expenses include rent, phone/internet bills, administrative costs, and others that are not attributed to a particular project. Institutions typically have a flat overhead rate that is automatically included in the budget of any grant. This rate may be capped by the funder. For example, MIT charges 60.5% for overhead on total direct costs, unless there is a cap by the funder. Overhead for many J-PAL-funded projects is capped at 10% (if going through a university in a high income country) or 15% (if going through a university in a low or middle income country, or an independent non-profit such as IPA).
• Tranches: Funds may not be disbursed all at once, but rather in increments (tranches) at pre-specified times. One thing to note is that different institutions have different policies around tranches and spending. See more in the grant and budget management resource.
• Memorandum of understanding (MoU): A (typically non-legally binding) agreement between two or more parties about the nature of the relationship. The MoU can contain information on the responsibilities of each party. For example, the J-PAL/IPA MoU describes the two organizations’ agreements around communication, office locations, research initiatives, and research staff training.

Before starting the grant proposal:

1. Ensure the project and research team meet the eligibility criteria. If not, it is better to apply for different grants.
   1. Similarly, understand the priorities of the funder (e.g., thematic areas or priority countries) and the criteria for selection.
2. Check the relevant deadlines. Grants often have multiple parts with different deadlines, e.g. for an expression of interest and a full proposal. 
3. Contact the university office for grants (and relevant office staff, such as the Finance & Operations team if working with J-PAL) to determine who needs to approve what and the internal deadlines. Partners and the lead institution will typically want to review your budget and planned activities in advance, so be sure to account for this in your grant preparation timeline. The time they request will vary by institution; for example, MIT requires a minimum of 5 business days to be able to “provide full and comprehensive proposal review, approval, and submission” of a proposal and proposed budget. However, the grant must be approved by others (such as J-PAL’s Finance & Operations Director, and MIT’s School of Humanities, Arts, and Social Sciences). Be sure to factor in the time required to obtain these additional layers of approval.
4. Plan sub-contracting, sub-awards, MoUs/letters of support, and other supporting materials
   1. The proposal will need to include letters of support from any close partners (such as implementers or any organization receiving a sub-award - NOT consultants or contractors hired for a narrow role). It’s important to have agreements in place with any organization receiving a sub-award or whose collaboration is critical to the projects success. This includes an MoU between the lead institution and implementing organization (such as the local J-PAL or IPA office). 

If available, an example of a successful grant proposal for a grant with similar requirements (particularly in terms of length and thematic area) can give a sense of the writing style, level of detail, and formatting, as well as the actual content of the proposal, that caused the funders to select that proposal.

Research grant proposals oftentimes have specific sections and associated questions you should address. Make sure to identify what those sections are for your grant, ensuring that you answer all components of the prompts and that your narrative speaks to the priorities of the funder and the call to which you are applying. Frequently, proposal prompts include:

• The motivation or problem statement
• The research question(s)
• The contribution of the project to the literature
• A description of the intervention
• The methodology
• A description of your partners 
• Plans to communicate the findings

Some funders include a specific proposal template, with guidelines on the desired length and content of each section. If so, follow that template and be sure to address all points in the guidelines. If the funder does not provide a template, the proposal should nonetheless include the components listed above. It should be geared towards the funder and show that you can ask a clear research question and that your proposed study will be able to answer it.

The motivation or problem statement
Start your proposal with a clear problem statement that motivates the research. It should be succinct, to the point, and convey some level of urgency.

The research question
You should actually state your question and show how it fits into the literature and field. The question itself should be understandable and compelling to researchers outside of your subdiscipline and, if applicable, to the policy community (see Chris Blattman's blog post). It should be narrow and specific enough to answer with your proposed project and should clearly fit in the request for proposals (RFP) scope of work.

Contribution to the literature
As in an academic paper, a literature review in a grant proposal should contextualize your research question and show that your proposed study fills an important gap. You can achieve this by 1) showing that you know the relevant literature on your study’s topic by citing important papers and 2) stating explicitly where the gaps are and how your study will fill them. This can include highlighting methodological shortcomings of the existing research, if applicable.

In determining what studies to include in your literature review, you should consult with top researchers in your field and search peer-reviewed journals, reputable think tanks or other research organizations, and databases (e.g., EBSCO, EconLit, Science Direct). To determine if a study is relevant to your proposal, you should find studies that answer the same or similar question to your study or provide support for your theory or methodology. You should consider the context, including the location, study year, and population to determine the studies’ relevance to your proposal (Hagen-Zanker & Mallett 2013; J-PAL staff and affiliates: see also Sarah Bishop's slides).

Intervention Description and Methodology
Provide sufficient detail about the intervention itself to convince the reader that your study is feasible and that your findings will be reliable. For example, David McKenzie advises that if you are evaluating a cash transfer program you should specify what the eligibility criteria are, what the payment levels are, the duration of the program, etc. You should explain your research design (e.g., method of randomization, number of treatment arms, stratifying variables, clustering, etc.) as well as the details of the program itself, so that the reader understands who is eligible for the program and what they receive and when. Schematics showing how the intervention will be provided, to whom, in what order, and when can be effective for conveying to the reader how complex interventions will operate.

You should also state your key hypotheses and outcomes, and how you will measure them. Describe your mechanisms for how and why you think your intervention will change your final outcomes, and what tests you will conduct to determine whether your intervention was mediated by those mechanisms. You can additionally describe how contextual factors, e.g., levels of rainfall for an agricultural intervention, can modify whether your intervention will prove successful.

Timeline planning
Though you may not include it in the grant proposal, it can be helpful to create a Gantt chart, detailing when key activities will take place over the course of the project life cycle. You may need to work backwards to figure out when activities should take place in order to meet certain hard deadlines, like the end of grant funding. For other interventions, key activities may coincide with times of the year (e.g., harvest) or by subject (e.g., when a child is two years old).

Detailed calendars of activities can be quite helpful. When considering when different activities will take place you should keep in mind both research and implementation-related concerns, such as:

• IRB deadlines
• Your university’s deadlines, particularly if you anticipate needing a full board review. Allow time for the IRB protocol to be approved.
• University schedule/academic year: PIs (and graduate student RAs) will have more time for research during summer months.
• Time needed to recruit, hire, and train RAs and field staff.
  • Local labor laws vary and are important to keep in mind. The local J-PAL or IPA office (or survey firm if you are not working with J-PAL or IPA) should help you ensure any hiring is done according to local regulations. They will also be able to help you navigate any regulations around your proposed activities. 
• Local seasons or calendars: for example, you don’t want to do an intervention to increase teacher attendance when school is not in session, or to collect harvest data before harvest. See also Survey logistics.
  • Logistical challenges at certain times of the year, such as washed out roads during the rainy season, could delay surveys or program implementation
• Events when conditions are unsafe or affect data quality: for example, elections in many contexts can be contentious 
• Local holidays, including bank and government holidays 

Sample size/power
Proposals should provide a sufficient level of detail for the reader to reproduce the power calculations described in the document. Specifically, you should include the effect size you expect for your study and why it is appropriate/sufficient, take up/compliance, variance, number of clusters, observations per cluster, and intra-cluster correlation coefficient. You should also include your expectations around attrition and survey response in your power calculations. You may want to consider including adjustments for multiple hypothesis testing to your power calculations. If you plan on doing sub-group analysis, include information on strata and how this affects your power calculations.

You should state all assumptions that you make and the data you are using to perform the analysis. Proposals that use data from a pilot study, or a study similar to the proposed study, are more convincing than information gleaned from other studies and contexts. Finally, you should state how the results vary with changes in key parameters, especially when evidence on the size of those parameters is not reliable. It can be useful to convey the results of power calculations using a graph or table. Visuals can summarize clearly and concisely how power varies with changes in key parameters.

Resources on conducting basic power calculations can be found below. Further guidance can be found in our power calculations resource:

• J-PAL Evaluating Social Programs Power case study and slides - a primer on power calculations
• EGAP Power Calculator - an intuitive, platform-independent calculator, though with limited functionality
• 3ie power calculator (.xls direct download) - less intuitive but with greater functionality

A description of your partners
You want to convey that your study is feasible, which means providing background on your partners, particularly any partners that are critical to the project’s success (such as implementers) and their commitment to the project. These partners will also likely need to provide a letter of support that will be submitted with the proposal. For guidance on how to evaluate potential research partnerships, see the Research Practice Partnerships guide to structuring a partnership, or our resource on assessing the viability of research partners. 

Plans to communicate the findings
Plans to communicate the findings can include academic papers and presentations, reports, policy briefs, and more. The funder may have specific requirements regarding deliverables; be sure to check these and describe how you will meet them.

Additional tips

• Use the right language for the audience: if you know that your proposal will be read by experts in the field (e.g., a J-PAL Initiative grant), you can use jargon. If writing for an audience of non-experts (or a layperson), write for an intelligent reader who is unfamiliar with jargon and other specifics of the field (and avoid jargon). 
• Use the active voice wherever possible (e.g., avoid phrasing such as “the bednets will be distributed beginning on March 1” in favor of “our partners will distribute the bednets beginning on March 1”). 
• Be sure to proofread your proposal before submission. 
• If possible, have someone who was not involved in the grant writing read the proposal, both for spelling/grammatical issues, and for content (particularly areas that may be unclear to someone unfamiliar with the proposed project). If the proposal is going to a non-expert, this proofreader/second set of eyes is ideally not an expert either and will be able to point out areas that need additional explanation for a layperson.
• Consider addressing risk in your proposal. Risks can include financial risks (which can be implicitly or explicitly addressed in the budget), as well as events that might prevent the completion of the study (e.g., pandemics, wide-scale social unrest, etc.). If working in a context where these events are probable, consider developing a plan for mitigating the risks if possible, and include this in your grant proposal.
• If your project will conduct phone surveys or remote monitoring, you may need to modify your budget proposal. Relative to in-person surveys, phone surveys often require additional days of training. Estimates of surveyor productivity may need to be adjusted, as enumerators may need to contact respondents multiple times before reaching them. Research teams may also consider budgeting a supervisor to oversee multiple enumerators by checking call logs, listening to calls, etc. Certain costs will change with remote surveys--research teams might not need transportation allowances, but instead might need data packs or air time. For more information, see our blog post on adapting to phone surveys.  

Budgeting principles

• Your costs can be split into survey costs (salaries for field staff, transportation, materials, etc.), miscellaneous field costs such as renting a local office, university-based personnel costs (e.g., PI travel), and indirect costs (overhead and fringe). More details on these are below.
  • For J-PAL staff and affiliates: See J-PAL's sample budget template for more guidance. See also J-PAL South Asia's project timeline template that can help guide planning.
• Build in buffer days and flexibility in the budget, as unforeseen costs will inevitably arise. That is, overestimate costs and time needed to conduct field activities where possible.
  • Factor in extra time when the concept is relatively unknown (for example, piloting a new method), if you are in a new geographic location where you or the survey team or implementing partner has not worked previously, or if you are working with new partners (both survey team or implementing partner).
  • Budget for extra equipment for when things break (they will)
• In general, you should scrutinize expenses that occur far enough off in the future that their costs may change. For example, it is likely that survey activities will not begin for several months after writing the budget for the proposal. During this time, staff salaries may rise (e.g., due to cost of living adjustment). 
  • If you are budgeting for both a baseline and endline survey, and there is a significant time lapse between the two, you may want to factor inflation into your endline cost estimates. In the US, 3% inflation is standard for projection purposes. The inflation rate will vary by country and should be confirmed with the grant manager in advance.
• Be conservative in estimating foreign exchange rates: don’t use the current exchange rate, but a long term (e.g., 6 month) average. Your institution likely has an exchange rate calculator, or you can use OANDA or see the links at MIT's foreign exchange help page.
• The budget should reflect what is described in the scope of work and the proposal. 
• If your project will be implemented by J-PAL or IPA, the implementing office will help with budgeting. Your budget will typically follow either the template provided by the funder or a J-PAL/IPA template (.xls direct download) if applicable.
• Be aware of procurement and payment restrictions by your institution. Many institutions require procurement (an open bidding process) when making a large (such as over $10,000) purchase of goods or contract of services. There may also be restrictions on payment for certain vendors, car insurance, international travel, and more. These restrictions vary by institution, so it is important to check while planning out the project budget and timeline.
• The first version of a budget should be very detailed. This helps you budget accurately and ensure that you’re including all costs. This also makes it easier to adjust assumptions and prices in a way where you can see exactly what is changing and why. This budget should help the RA/RM/PIs understand everything that will go into running a project. Once you have budgeted for the entire project, you can collapse categories to go into the donor budget templates. For example, in the detailed internal budget you may want to separate transport costs into vehicle hire, fuel costs, insurance, and parking fees, but in the donor template, just have one category for transport.
  • This will also help you think through whether there will be extra fees for some purchases, such as admin fees for hiring cars, delivery fees for some purchased goods, etc.
• Most projects will apply for funding from multiple donors. Set up a system that allows for tracking which expenses are allocated to different funding proposals. If you need to do more fundraising in the future or if some proposals are rejected, this will help you see where the shortfall is without having to review each proposal individually. This can be done using J-PAL Africa's template budget tracker (.xls direct download). 
• When starting a budget, take time to make sure that it is very simple to edit and minimize the chance of human error. 
  • If referencing a value from a different sheet, use lookup, sumif or index match formulas rather than manually referencing a cell. For an example of how to create a budget in Excel, see Cornell's Sponsored Budget Template.
  • Put in checks to make sure you’re summing all line items in a subcategory
  • If you are linking to an assumptions page, make sure that you only edit those pages and not the full budget template. 

Budget justification

Funders will also ask you to include a budget justification, where you describe any assumptions made for each budget line item. Some tips here are:

• Provide sufficient detail in your budget estimations so that it is clear how the totals were constructed and how budget amounts would change if necessary (e.g., sample size increased). For example, for enumerator costs, include in your budget justification the daily rate for enumerators, number of enumerators, and number of days of work for each enumerator.
  • Include ALL costs - budgeting should be done alongside survey implementation planning.
  • Your justification should tell a story of the project that an auditor (or funder) can read without needing to ask you for clarification.
• Think about what is allowable and reasonable. Donors may raise questions if the budget is too exorbitant or cost justifications are unclear; basically, they want to know if their money would be well-spent. Cost justifications should be sufficiently detailed but not too long. If you are working with a J-PAL or IPA office, the office often has standard language on fringe or overhead rates.
  • Local staff, particularly RAs and RMs, should read over the budget assumptions to ensure they are reasonable (for example, that the expected number of surveys to complete per day is reasonable, given survey length, travel time, etc.).

Individual budget sections

Survey costs
Survey costs (including for implementation monitoring and quality control) should include the costs for enumerator salaries, materials, and transportation--see Survey logistics for more on this. These costs will depend on assumptions such as:

• Total number of surveys
• Number of surveyors and number of supervisors (including the number of supervisors per surveyor)
• Number of days spent in each location
• Number of surveys/person/day
• Staff training time - this requires the same considerations and assumptions around the number of surveyors and supervisors, as well as days spent in each location. Costs will also depend on the location of the training site and how easily it is accessed by public transportation
• Number of debrief days or sessions
• Number of contingency days for days when there is inclement weather or other events (such as protest action)
• Leave days or bonus pay for surveyors
• Survey piloting costs
• Computer- vs. paper-based data collection

Survey costs may include:

• Field team salaries, based on the assumptions listed above. This will likely comprise the majority of your costs. Note that some donors will not allow for international staff costs.
• Electronics such as tablets for surveyors (for digital data collection); GPS devices; laptop for the RA
• Data entry (if using paper-based surveys)
• Transportation costs for the survey team
• If renting vehicles or motorcycles: Rental costs, fuel costs, driver salaries, insurance, repair costs, helmets (for motorcycles). Note that these will increase in areas that are not easily accessible by public transportation, or where public transportation is unsafe at certain hours.
• Any rented or purchased equipment, such as scales for health surveys or other measurement devices
• Materials such as printed lists, umbrellas, pens, helmets if using motorcycles, branding if surveyors will get t-shirts or vests that identify them as surveyors. 
  • Always budget more that you think you will need for stationery and printing - there are often things you need that you can’t foresee ahead of time (and pens go missing very easily).
• Quality control: Translation and back translation personnel costs; revisit and tracking costs (for missing respondents); quality control checks covered under J-PAL Best Practices, such as back-checks on 10% of surveys, double data entry for paper surveys, etc.
• Monitoring costs (both for surveys, such as through spot-checks, back-checks, etc., and implementation monitoring)
• Census listing costs (if needed): Will include wages for surveyors, transportation, etc.
• Qualitative work such as focus groups
• Respondent compensation (use the word compensation rather than gift, which some funders object to). See also the discussion of compensation in Define intake and consent process
• Communication costs (such as phone credit)
• Implementation: There may be other costs associated with the implementation of your specific project. For example, you may incur a cost if, as part of your project, you need to send SMS messages or provide incentives. You should budget for those accordingly. General program implementation costs, such as those that implementers would incur regardless of the study (e.g., for schools to run an educational program) are not typically covered by a grant.
• SurveyCTO or similar survey programs: Ensure that you budget for several months before and after fieldwork so that you have enough time to program surveys and can have access to the raw data following the survey and allow for extra data collection. 

Miscellaneous field costs

• Renting a field office, which may also require field office repairs and maintenance costs
• Local office staff: If you are working with a J-PAL or IPA office, they will budget for personnel to help manage your project, such as an RA, a local research manager, a finance officer, and in-country supervision/oversight. Note that for salaried employees, you will need to include their annual salary and time allocated to the project (% full-time equivalent, or FTE). If the project will take place over an extended period of time, you may want to budget for promotions or cost of living adjustments.
• Bank charges (for example, withdrawing large sums from a foreign account)
• Local office meeting expenses
• Health and safety: First aid kits and training, pepper spray if surveyors will be in dangerous areas. 
• Local assistance: You may need to work with local neighborhood watch groups to ensure the safety of field staff, or pay to have someone on standby in the local community to assist if there are any safety concerns. You may want help from local community members for other tasks, such as taking soil samples or helping locate households. Compensation may be small but can add up.

University-based personnel costs

• PI salary and benefits: Some J-PAL Initiative funds do not allow you to include PI salaries in the budgets. For other grants, it is common to include a month of PI salary in the budget. You may also be able to budget for benefits/fringe. See note about FTE above. As above, if the project will take place over an extended period of time, you may want to budget for promotions or cost of living adjustments.
• PhD student or university-based RA: If you are not working with J-PAL or IPA, you should budget for personnel who can help manage the project. This is often a PhD student who will help ensure the successful launch of surveys, e.g. by overseeing enumerator trainings, and who will ensure the quality of the data collection. It is common to include costs for the graduate research assistant (salary plus tuition costs) in the budget. As above, you may need to budget for promotions (less likely) or cost of living adjustments (more likely). See note about FTE above.
• Travel: You should budget in the costs of PI, and potentially graduate student RA, travel into the budget. Include in the budget details on the number of trips, flight/travel costs, length of trips, and accommodation and per diem rates. Check the donor guidelines for policies about international travel, etc. For example, unless part of the original approved budget, USAID does not allow for international travel without written permission, so it is useful to budget for such travel early on. Moreover, grants funded by the US government (including USAID) are subject to the Fly America Act, which stipulates that international travel funded by that grant must be done on a US carrier (with certain exceptions). This can restrict travel flexibility and increase costs, so be sure to budget accordingly.
• J-PAL/IPA research staff training (if applicable)
• Transport and communication: For RA and RM for travelling to meetings, communicating with partners.
• Subscriptions: Dropbox, Stata licences, Zapier, etc. for RA and RM. 

Indirect costs
These cover general overhead and administration costs, such as office rent and utilities, which are often shared across projects. Factoring in overhead costs can get complicated, and the relevant finance team should be heavily involved in this stage of the budgeting process.

Different funders will allow you to charge different amounts in indirect costs. The maximum rate is typically dependent on the type of institution (e.g. university vs. non-profit) and where it is located (e.g., developed or developing country-based). For example, many J-PAL Initiatives cap indirect costs at 10% for universities in high-income countries and 15% for universities in developing countries or non-profits regardless of location, though some initiatives may have different caps. If you are not using a grant template provided by the funder, be sure to factor these costs in early, as even indirect costs on the lower end of the spectrum (e.g., 10%) can add up quickly. Without the cap, overhead rates can run quite high. Overhead rates may also vary depending on whether the institution is receiving a sub-award (versus being lead institution). If your overhead exceeds the maximum allowable by the donor (e.g., if the donor caps overhead at 10% but the implementing office’s overhead is fixed at 15%), you may need to budget some of the overhead costs as direct costs. This should be done in consultation with the relevant finance staff member or grant manager at the implementing organization.

Remote surveys require different devices and a modified survey timeline compared to in-person surveys. These differences should be addressed in the budget proposal. In addition to the information presented above, when budgeting for a remote survey, research teams should consider the following: 

• Devices: In addition to the devices used in in-person surveys, with remote surveys, the budget should include headphones, tablets and sim cards, and phone recharges. Research teams can consider employing a number masking platform, especially if enumerators conduct calls with their personal phones. 
• Surveyor training: Relative to in-person surveys, remote surveys often require additional days of training. For more information, see J-PAL South Asia’s budgeting for surveys during the Covid-19 outbreak resource.  Research teams may also consider budgeting additional supervisors to oversee multiple enumerators by checking call logs, listening to calls, etc., and budgeting for additional research manager time for any extra surveyor training.
• Productivity: Estimates of surveyor productivity may need to be adjusted. While phone surveys typically are shorter than in-person surveys, enumerators may need to contact respondents multiple times before reaching them. Your estimate would need to take into account whether you’re trying to target a specific respondent (caregiver, household head) to complete the survey with, which might mean more attempts to get that specific individual on the line. In addition, enumerators might have lower success rates during certain times of the day, causing the timeline to extend. See J-PAL South Asia’s budgeting for surveys during the Covid-19 outbreak resource for more information. 

Last updated February 2021.