Chapter 23: Collaborative and Multi-Institutional Proposals — When One PI Isn't Enough

Most of this book has assumed a single applicant: one investigator, one organization, one voice. But some of the most important work — and some of the largest grants — cannot be done by one person or one institution. A clinical trial across several health systems, a research center spanning departments, a training program drawing faculty from many labs, an international development consortium (Chapter 21), a multi-site service program — these require teams, and they require a different kind of proposal. When one PI isn't enough, you enter the world of collaborative and multi-institutional proposals, where the science or the program is only half the challenge and the collaboration itself — its structure, its leadership, its credibility — is the other half.

That second half is what most applicants underestimate, and it gives us the chapter's threshold concept: reviewers fund teams that have clearly already learned to work together. A collaborative proposal is judged not only on the quality of the science and the strength of the individuals but on whether the collaboration is credible — whether these people and institutions can actually function as a team, coordinate across distance and difference, divide labor and credit without conflict, and deliver something none of them could alone. Reviewers have seen too many "collaborations" that were really a lead investigator and some names on a page, and they have watched funded teams fracture. So they scrutinize the collaboration as hard as the content — and the proposals that win are the ones that demonstrate, not merely assert, that the team is real and ready.

This chapter teaches you to build and represent that team. We'll cover when a proposal genuinely needs multiple PIs and institutions (and when it doesn't), how to structure a multi-PI or consortium application and choose a leadership model, how to write the leadership and coordination plan that convinces reviewers the team can work, how to budget and administer subawards, and the hardest craft of all — making many authors speak with one voice. Our anchors grow naturally: Dr. Hernandez's diabetes trial, which expanded to a second clinical site in her A1 (Chapter 22), now scales into a genuinely multi-site, multi-PI study; and Lighthouse Community Services, which partners with other organizations to deliver a reentry program no single agency could run alone. (As always, multi-PI rules and subaward mechanics vary by funder and change; verify specifics with your funder and your grants office.)

23.1 When You Actually Need a Team

Before assembling a team, ask the harder question: do you actually need one? Collaboration is powerful but costly — it adds coordination overhead, administrative complexity, and risk — so a team should be assembled because the work requires it, not because more names seem more impressive.

You genuinely need a multi-institutional team when the work demands capabilities, populations, or resources that no single institution has: a trial that needs more patients than one site can enroll (Hernandez's situation as she scales); expertise spanning disciplines that don't coexist in one department; a program serving a population that several organizations reach collectively but none reaches alone (Lighthouse's situation); equipment, data, or facilities held at different institutions; or a scale of effort beyond one organization's capacity. In these cases, the collaboration adds real value — it makes possible work that otherwise couldn't happen.

There is also a class of grants that require collaboration by design — center grants, multi-site trials, training consortia, and many large federal and international mechanisms are structured specifically to fund teams, and a single-institution application to such a mechanism would be non-responsive. And funders increasingly use collaboration as a deliberate strategy: pooling expertise across institutions, building research networks, fostering interdisciplinary work, or requiring that national resources be shared. So part of the "do I need a team?" question is reading the funder's expectation — some mechanisms reward or demand collaboration, others fund individual investigators, and applying with the wrong structure (a lone PI to a center mechanism, or an over-built consortium to an individual-investigator program) misreads the funder as surely as any mismatch in Part III. Match your structure both to what the work requires and to what the mechanism expects.

You do not need a team when a single capable investigator or organization could do the work, and adding partners would only dilute focus, complicate administration, and split the budget without adding genuine capability. Reviewers can tell the difference between a collaboration driven by the work and one assembled for appearances, and the latter — a "team" of loosely connected names with no real interdependence — is a weakness, not a strength. The first discipline of collaborative proposals is honesty about whether you need one — and the courage to keep a proposal lean when the work doesn't demand a crowd.

🧩 Productive Struggle: Before reading on, consider why a stronger-looking team — more famous names, more institutions, more breadth — might actually make a proposal weaker. Jot your thinking. The resolution, developed through this chapter, is that reviewers fund working teams, not impressive rosters: every added partner increases coordination risk and administrative complexity, and a team assembled for its names rather than its necessary, complementary contributions signals exactly the kind of collaboration that fractures after funding. A tight team of genuinely interdependent partners who clearly work together beats a glittering roster of loosely connected stars — because the reviewers are betting on whether the collaboration will function, not on whose names are attached.

💡 Key Insight: Every collaborator you add should pass a simple test: what specific, necessary capability does this partner contribute that the team otherwise lacks? If you can answer crisply — "the second site doubles our recruitment and reaches a different patient population," "this partner holds the community relationships we need to enroll participants," "this co-investigator brings the statistical genetics expertise our analysis requires" — the partner belongs. If the answer is vague — "they add prestige," "we've worked with them before," "it strengthens the application" — the partner may be dead weight that adds coordination cost without capability. Build your team from necessary, complementary contributions, and you can articulate each partner's role with the clarity reviewers want; build it from names and relationships, and you'll struggle to justify the structure you've created.

📊 From the Field: Lighthouse shows the test applied in the nonprofit/service world (composite). Its reentry workforce program needs more than Lighthouse alone can provide: people returning from incarceration need workforce training (Lighthouse's strength), but also stable housing, behavioral health support, and employer relationships — and Lighthouse does not do all of these well. So a strong collaborative proposal pairs Lighthouse with a housing organization, a behavioral-health provider, and an employer-partnership intermediary. Run the test on each: the housing partner contributes a necessary capability (participants who are unhoused can't sustain employment) Lighthouse lacks; the behavioral-health partner addresses a documented barrier Lighthouse isn't licensed to treat; the employer intermediary holds relationships that convert training into jobs. Each passes — each adds a necessary, complementary capability, and together they deliver a wraparound program none could alone. Contrast a partner that fails the test: a well-known national nonprofit invited only because its name might impress funders, with no distinct role in the actual service model. Lighthouse leaves that one out. The collaboration is built from what the work requires — a coalition of genuinely interdependent providers — which is exactly what a funder of reentry services wants to see, because it mirrors how the problem actually gets solved.

23.2 Structuring the Team and Choosing a Leadership Model

Once you know you need a team, you must structure it — and the central decision is leadership. Several models exist, and funders have specific mechanisms for them.

Single-PI with collaborators and subawards. The traditional structure: one principal investigator at a lead institution holds the grant, with co-investigators and partner institutions contributing through subawards (Section 23.4). The PI leads; others contribute defined pieces. This is the simplest model and fits when there is a clear single leader and the partners play supporting (if essential) roles.

Multiple-PI (MPI). Many funders, notably the NIH, allow a multiple-PI model in which two or more investigators share leadership as equals (or near-equals), jointly responsible for the project. The MPI model fits genuinely co-led projects where no single person is "the" leader — where the work integrates two or more domains of equal weight. The NIH MPI model requires a leadership plan (Section 23.3) explaining how the multiple PIs will govern the project. Choosing MPI is a real decision: it signals true shared leadership, distributes credit, and changes governance — so choose it when the project is genuinely co-led, not as a courtesy.

Center, program-project, and consortium structures. Larger mechanisms — research centers, program-project grants, training programs, multi-site consortia — have their own architectures, often organized into cores (shared resources serving the whole project, like an administrative core, a biostatistics core, a community-engagement core) and components/projects (individual research projects or program elements under the center's umbrella). These are complex applications with their own rules, usually a single overall PI or director plus project/core leaders, and they demand sophisticated coordination structures. (Hernandez's scaled multi-site trial, if large enough, might be organized with a coordinating center and site-specific components.)

The structural choice should follow the work's real shape: a clearly-led project with supporting partners → single-PI with subawards; a genuinely co-led project → MPI; a large, multi-component effort → a center/consortium structure. And whatever the structure, the lead institution (the prime) carries overall fiduciary and administrative responsibility for the award, including for funds passed to partners — a responsibility that, as Chapter 21 stressed for international consortia, is real and heavy.

🎓 Going Deeper — the contact PI and how MPI actually works: The multiple-PI model is widely misunderstood, so a few specifics help. Under the NIH MPI model, two or more PIs share genuine scientific leadership and all are recognized as PIs — but one is designated the contact PI, who serves as the primary point of communication with the funder and handles administrative coordination. Critically, the contact PI is not "the real PI" or the senior partner: the role is administrative, and the contact PI does not have greater scientific authority by virtue of the designation (the leadership plan defines actual authority). This matters because choosing MPI is sometimes mistakenly seen as demoting yourself or sharing credit you'd rather keep — but for genuinely co-led work, MPI accurately represents the leadership, gives all PIs full credit and recognition (important for career advancement, Chapter 27), and is viewed favorably by reviewers precisely because it's honest about how the work is led. The mistake is using MPI when the project is really single-PI (confusing reviewers about who leads) or avoiding MPI when the project is genuinely co-led (understating a partner's role and risking resentment). Choose MPI when leadership is truly shared, designate a contact PI for the administrative role, and let the leadership plan specify who actually decides what. (Rules and the MPI option vary by funder and mechanism — confirm before structuring.)

📊 From the Field: Watch Hernandez choose a model as her trial scales (composite). Her A1 added a second clinical site to prove recruitment feasibility (Chapter 22). Now, funded and planning a larger study, she needs three sites across different health systems to enroll a definitive sample. The sites aren't interchangeable subcontractors — each brings a distinct patient population and a local investigator whose clinical and community knowledge is essential. Hernandez weighs the models. A pure single-PI-with-subawards structure would work, but it understates the site investigators' genuine intellectual leadership. Because one site PI is a behavioral-intervention expert co-leading the science (not just running a site), Hernandez opts for a multiple-PI model with that colleague, and brings the other site leaders in as co-investigators with subawards. The structure now mirrors the real division of leadership — and her leadership plan can describe a governance that reviewers will find credible because it matches how the team actually works. The lesson: let the structure follow the true shape of the collaboration, not the other way around.

23.3 The Leadership and Coordination Plan

Here is where collaborative proposals are won or lost, and where the threshold concept lives. The leadership plan (required for NIH MPI applications, and wise for any complex collaboration) is the document in which you convince reviewers that this team can actually function. It is not boilerplate; it is, often, the single most scrutinized element of a collaborative proposal, because it is the evidence that the collaboration is real.

A strong leadership and coordination plan addresses:

• Governance and decision-making. Who leads what? How are decisions made, especially when PIs disagree? A credible plan names a decision process — not "we'll all agree" (teams don't always agree), but a real mechanism (e.g., the contact PI has final say on X, the steering committee votes on Y, a defined process resolves disputes).
• Roles and responsibilities. Who is responsible for which aims, cores, or sites? Clarity here prevents the gaps and overlaps that sink real collaborations and signals that the team has thought it through.
• Communication and coordination. How will the team stay aligned across institutions and distance? Regular meetings, a coordinating center, shared data systems, defined reporting lines — the machinery of staying coordinated.
• Conflict resolution. What happens when there's a serious disagreement or a partner underperforms? A plan that anticipates conflict and names a resolution process reassures reviewers far more than one that pretends conflict won't occur.
• The evidence of a working relationship. Crucially, the plan (and the proposal) should show that this team has worked together before or has concretely prepared to — prior collaborations, joint publications, pilot work together, a planning process already underway. This is the single most powerful signal that the collaboration is real.

🚪 Threshold Concept: Reviewers fund teams that have clearly already learned to work together. A collaborative proposal asks reviewers to bet not just on people but on a relationship — on whether these investigators and institutions will actually coordinate, share, defer, and deliver across the friction of distance, ego, and institutional difference. That bet is risky, and reviewers know it: they've seen funded teams fracture. So they look for evidence that the team has already learned to work together — prior joint work, a concrete coordination structure, a leadership plan that anticipates real conflict, a division of roles that fits how the people actually operate. Cross this threshold and you stop treating the leadership plan as administrative box-checking and start treating it as the core argument that your collaboration is fundable. The science can be brilliant and the individuals stellar, but if the collaboration isn't credible, reviewers won't fund the risk — and if it visibly is, you've answered the question they most worry about. This is why two proposals with identical science and equally distinguished investigators can diverge so sharply at review: the one that demonstrates a working collaboration clears the risk the reviewers fear, while the one that merely asserts a team leaves that risk unaddressed — and unaddressed risk, in a competitive field, is enough to sink an otherwise-excellent application.

🗣️ From the Review Panel: (A reviewer on a multi-PI application reflects.) With a team application, I'm always asking one question behind all the others: will these people actually work together, or will this fall apart in year two? I've seen it fall apart — brilliant investigators who couldn't share a decision, sites that drifted out of coordination, a "team" that was really one person and some subcontractors who barely spoke. So I read the leadership plan hard, and I look for tells. A plan that says "the PIs will communicate regularly and resolve issues collegially" tells me they haven't thought about it. A plan that names who decides what, how disputes get resolved, how the sites stay coordinated — and that's backed by evidence these people have already collaborated successfully — tells me this is a real team. Show me you've already learned to work together, and show me a governance structure that survives disagreement, and I can fund the collaboration. Make me guess, and the risk is too high.

To make it concrete, here is a condensed sketch of Hernandez's leadership plan for her three-site MPI trial (paraphrased):

Governance. Drs. Hernandez and Okafor [the behavioral-intervention co-lead] serve as Multiple PIs with shared authority over scientific direction. Dr. Hernandez serves as contact PI, responsible for communication with the funder and final authority over budget and timeline. Major scientific decisions are made jointly by the two MPIs; a Steering Committee (the two MPIs plus the three site investigators) meets monthly and decides protocol questions by majority, with the contact PI breaking ties.

Roles. Dr. Hernandez leads the trial design and primary analysis; Dr. Okafor leads the intervention and fidelity; each site investigator leads recruitment and delivery at their site. A Coordinating Center (administrative and biostatistics functions) supports all sites.

Communication and coordination. Monthly Steering Committee calls; weekly operational calls among site coordinators; a shared data and tracking system; quarterly in-person (or virtual) all-team meetings.

Conflict resolution. Scientific disagreements escalate to the Steering Committee; unresolved disputes go to the two MPIs, then, if needed, to an external advisory board member named in advance. Underperformance by a site triggers a defined remediation process.

Track record. Drs. Hernandez and Okafor have co-authored two prior studies; the team conducted joint pilot work across two of the three sites during the prior award [Chapter 22's second site].

Notice what this plan does: it names who decides what (not "we'll agree"), it specifies real coordination machinery, it anticipates conflict and underperformance with a process, and it offers evidence the team has already worked together. A reviewer reading it can believe this collaboration will function. That belief is the whole point.

🪞 Learning Check-In: If you're assembling a team, notice any temptation to treat the leadership plan as a formality to dash off after the science is done. That instinct is exactly backwards for a collaborative proposal. For a team application, the collaboration is part of the science's feasibility — a brilliant plan that the team can't execute together is worth nothing, and reviewers know it. So the leadership plan deserves real thought, real honesty about how decisions and conflicts will be handled, and real evidence of a working relationship. If writing it surfaces uncomfortable questions — who actually decides? what if two PIs disagree? has this team really worked together? — that discomfort is valuable: it's surfacing the risks reviewers will probe, while you can still address them.

23.4 Budgeting and Administering Subawards

When partners at other institutions contribute, money flows to them through subawards (also called subcontracts or subgrants), and the mechanics matter both for the budget and for the administration.

A subaward is a portion of the grant passed from the lead institution (the prime) to a partner institution to perform part of the work. Each subaward partner prepares its own budget — its personnel, its costs, and crucially its own indirect costs — which is then incorporated into the overall proposal budget. Several points trip up first-time team builders:

• Each partner has its own indirect rate. A subaward partner applies its negotiated indirect rate to its direct costs, so a multi-institution budget contains multiple indirect rates (Chapters 11–12). The prime then applies its own indirect rate, but typically only to a limited portion of each subaward (often the first portion of each subaward, per funder rules) — not to the whole sub. The result is a layered budget that must be built carefully and correctly.
• Each subaward needs documentation. Funders typically require, for each subaward partner, a budget, a budget justification, a scope of work, and a letter of commitment or consortium/subaward agreement signed by an authorized official — committing the partner institution, not just the individual investigator (Chapter 13's letter-of-commitment lesson, applied to subs).
• The prime carries administrative responsibility. The lead institution is accountable to the funder for the whole award, including monitoring subrecipients, ensuring they comply with the rules (2 CFR 200 for federal awards, Chapter 19), and managing the flow of funds and reports. This is real administrative work that must be planned and, often, resourced. Many large collaborative awards explicitly fund an administrative core or coordinating function precisely because managing a multi-institution project — tracking subrecipients, consolidating reports, coordinating meetings, ensuring compliance across institutions — is a substantial job in its own right. Budgeting for that coordination is not padding; it is acknowledging the genuine cost of running a team, and a proposal that pretends the coordination will happen for free signals either naïveté or a plan that will quietly fail under the administrative load.

📋 Template — assembling a multi-institution budget (verify your funder's and institutions' rules): For the lead institution: its personnel, direct costs, and indirect rate. For each subaward partner: a complete sub-budget (personnel, direct costs, the partner's own indirect rate) plus a scope of work, budget justification, and a signed letter of commitment / subaward agreement. Then integrate: sum direct and indirect across all parties; apply the prime's indirect to the allowable portion of each subaward per funder rules; confirm every total reconciles and every partner's commitment is documented. Build coordination/administration (meetings, a coordinating function, subrecipient monitoring) into the budget as real, funded work — not an unfunded afterthought. Start early: gathering budgets and signed commitments from multiple institutions takes far longer than a single-institution budget.

⚠️ Common Pitfall: Underestimating the time and coordination a multi-institution budget and submission require. A single-institution proposal you can assemble on your own timeline; a multi-institution proposal depends on every partner's grants office producing a budget, a justification, a scope of work, and signed commitments — each on their internal timeline, with their internal deadlines, routed through their officials. If you start gathering these a week before the deadline, you will miss it, because one partner's signature will be stuck in one partner's office. The fix is to start the subaward-gathering process weeks earlier than you would a single-institution submission, give every partner a clear internal deadline well ahead of the real one, and treat the coordination of partners' paperwork as a major project task with an owner. The most common way collaborative proposals fail is not bad science — it's a partner's documentation arriving too late to submit.

🔄 Check Your Understanding: A three-institution proposal has a lead university and two subaward partners. A first-time PI builds the budget by applying the lead university's single indirect rate to the entire combined direct-cost total. What's wrong, and how should the indirect costs actually work?

Answer

The error is treating the multi-institution budget as if it had one indirect rate. In fact each institution applies its own negotiated indirect rate to its own direct costs — so the budget is layered: each subaward partner computes its indirect on its direct costs, and the lead university applies its indirect rate only to a limited, allowable portion of each subaward (per funder rules — often just the first portion of each sub), not to the partners' full amounts. Applying the lead's single rate to the whole combined total both misstates the indirect costs and signals the PI doesn't understand subaward budgeting. The fix: build each partner's sub-budget with its own indirect rate, then integrate per the funder's rules for prime indirect on subawards (Chapters 11–12).

✅ Best Practice: Form the team — and start the paperwork — far earlier than feels necessary, because collaboration runs on other people's timelines, not yours. The moment you decide a proposal will be collaborative, do four things: confirm each partner's genuine commitment (not just enthusiasm, but their grants office's willingness to produce a subaward package by your internal deadline); give every partner a written internal deadline well ahead of the funder's; designate one person to own the coordination of all partner documents; and begin the relationship and any required consortium agreement immediately. Multi-institution proposals routinely require four to six weeks of lead time just for the administrative assembly — budgets, justifications, scopes of work, and authorized signatures from several institutions — entirely separate from writing the science. The teams that submit smoothly are the ones that treated partner coordination as a project with an owner and an early start; the ones that scramble (or miss the deadline) are the ones that assumed partners' paperwork would materialize on demand. As with the single-institution lesson of Chapter 4, the most preventable failure is starting too late — and with multiple institutions, the late-start penalty multiplies.

23.5 One Voice from Many Authors

A collaborative proposal is written by many people, and it usually reads like it — until someone makes it not. The final craft challenge of this chapter is making a multi-author proposal speak with one voice, because a proposal that reads as a stitched-together patchwork of different authors' sections signals exactly the disorganization reviewers fear in a team.

The problem is real: different co-investigators write in different styles, at different levels of detail, with different assumptions, and a draft assembled from their contributions is jarring — inconsistent terminology, repetition, gaps between sections, tonal whiplash, and a logic model (Chapter 5's coherence) that doesn't quite connect because no one owned the whole. The reader feels the seams on every page. Reviewers reading such a proposal don't just notice the seams; they infer that the team is as disjointed as its prose.

The solution is ownership and integration. Someone — usually the lead PI or a designated writer/coordinator — must own the whole document and integrate the contributions into a single coherent voice: harmonizing terminology and style, eliminating repetition, filling gaps, ensuring the sections connect, and making the aims, narrative, and budget cohere (Chapter 5's one-argument discipline, now across authors). This is not optional polish; it is what makes the proposal read as the product of a coordinated team rather than a loose confederation. Contributors draft their pieces; one owner makes them one coherent document that speaks with a single voice.

How does the owner actually achieve one voice without alienating co-authors who wrote the pieces? A few practices make it work. First, agree on the structure and key terminology up front, before anyone drafts — a shared outline and a short glossary of the project's central terms prevent the worst inconsistencies at the source. Second, give contributors clear, bounded assignments ("write the 1,200-word recruitment section covering X, Y, Z") rather than vague ones, so the pieces fit together by design. Third, the owner rewrites freely for voice and coherence — not changing anyone's science, but harmonizing style, terminology, and transitions — and explains to the team that this integration is how the proposal reads as one team's work, so no one takes the editing personally. Fourth, circulate the integrated draft so co-authors can check their science survived the harmonizing intact. This is itself a small exercise in the collaboration the proposal describes: a team that can produce one coherent document through clear assignments, generous integration, and good-natured editing is demonstrating, in miniature, the coordination it promises to bring to the funded work. Handled well, the one-voice process doesn't just improve the proposal; it builds the team.

🔍 Why Does This Work?: Why does a single coherent voice signal a fundable team, when the reviewers know many people wrote it? Because the proposal is the team's first deliverable, and how the team produces it is evidence of how the team will produce everything else. A proposal that reads as one coherent argument demonstrates that the team can integrate its members' contributions into a unified whole — exactly the capability the actual project will require. A patchwork proposal demonstrates the opposite: that the members worked in silos and no one integrated them, foreshadowing a project where the sites drift, the cores don't connect, and the whole is less than its parts. So the one-voice discipline isn't cosmetic; it's a live demonstration of the team's coordination capacity, read by reviewers as a preview of the collaboration itself. Make the proposal cohere, and you've shown the team coheres.

📊 From the Field: One delicate aspect of team proposals deserves naming directly: dividing credit and responsibility. Behind every collaboration are real people with careers, egos, and institutional pressures, and the questions of who is PI versus co-investigator, whose institution is the prime, how authorship and recognition are shared, and how the budget is split can quietly poison a collaboration before the work even begins — or after. Experienced team-builders handle this explicitly and early, before submission: they have the candid conversations about roles, leadership, credit, and money while the proposal is being built, rather than papering over the questions and letting resentment fester. It feels easier, under deadline pressure, to leave these questions vague and "sort them out later" — but later, with real money and real recognition at stake, is exactly when vagueness becomes conflict. The leadership plan (Section 23.3) is partly a record of those agreements — who leads what, who decides, how credit is recognized — made concrete. A collaboration where the credit-and-responsibility questions were avoided is a collaboration waiting to fracture, and reviewers, who have seen this, are reassured by a proposal whose division of leadership is clear and evidently agreed. The uncomfortable conversations you have before submission are far cheaper than the conflicts you'll have after funding if you skip them. Treat dividing credit and responsibility as essential team-formation work, not an awkward topic to avoid — and let the leadership plan reflect the genuine agreement you reached.

📐 Project Checkpoint — If collaborative, draft a leadership plan and subaward sketch: If your project needs a team, (1) justify each partner with the Section 23.1 test — what necessary, complementary capability does each contribute? (2) Choose a leadership model (single-PI with subawards, MPI, or center/consortium) that fits the real shape of the collaboration. (3) Draft a leadership and coordination plan: governance and decision-making, roles, communication, conflict resolution, and the evidence your team has worked together. (4) Sketch the subaward structure: who gets a subaward, for what scope, and the documentation (budget, scope of work, letter of commitment) each needs — and note the earlier timeline multi-institution gathering requires. (5) Name the one owner who will integrate all contributions into a single voice. Save it in your "My Proposal" document. If your project is single-PI, read this anyway — you may join others' teams, and understanding subawards and leadership plans makes you a better collaborator.

23.6 Strategy: Build the Team the Reviewers Can Believe

Pull the threads together. Succeeding with collaborative proposals means: assemble a team only when the work genuinely needs one, with each partner justified by a necessary, complementary capability; structure it with a leadership model that fits the real division of leadership (single-PI, MPI, or center/consortium); write a leadership and coordination plan that demonstrates — with governance, roles, conflict resolution, and evidence of prior collaboration — that the team can actually work together; budget and administer subawards correctly and on an earlier timeline; and integrate the many authors into one coherent voice. Above all, hold the threshold concept: reviewers fund teams that have clearly already learned to work together, so the credibility of the collaboration is as central as the science.

Set the collaborative proposal beside the single-applicant proposals of earlier chapters:

The deepest connection is to Chapter 21: an international consortium is a collaborative proposal with borders added, and everything here — the leadership plan, the subawards, the one-voice discipline, the credibility of the collaboration — applies there too, intensified. Master the collaboration, and you can lead the team efforts that tackle the problems no single investigator or organization can.

It's worth naming why this skill matters more every year. The largest and most consequential problems — in health, in science, in social services, in development — increasingly exceed what any single investigator or organization can address, and funders have responded by creating more team-based, center, network, and consortium mechanisms to fund work at that scale. The grant writer who can only operate solo is limited to the shrinking set of problems one person can own; the grant writer who can build and lead a credible team can pursue the larger, more ambitious, more impactful work that defines a serious funding career (Part VI). Learning to make a collaboration fundable — and then to make it function — is therefore not a niche skill for occasional big grants; it is a core capability for anyone who wants to work at the scale the hardest problems demand. The team is not a complication to tolerate; it is the vehicle for the most important work.

🔄 Check Your Understanding: Two multi-PI proposals describe equally strong science and equally distinguished investigators. One is funded and one is declined. Name two collaboration-specific factors, from this chapter, that could decide it.

Answer

Several are valid. Two clear ones: (1) The leadership/coordination plan and evidence of a working relationship — the funded team demonstrated a credible governance structure (who decides what, how conflict is resolved) backed by evidence they'd already collaborated successfully, while the declined team offered a vague "we'll coordinate collegially" with no evidence, leaving reviewers unconvinced the collaboration would function. (2) The coherence of the proposal itself / one voice — the funded proposal read as one integrated argument (signaling a coordinated team), while the declined one read as a patchwork of different authors' sections (signaling silos and foreshadowing coordination failure). Both reflect the threshold concept: reviewers fund teams that have clearly already learned to work together, so the collaboration's credibility — not just the science or the names — decides it. (Also acceptable: a team assembled for necessary complementary capabilities vs. assembled for prestige; correctly structured subawards and leadership model vs. a mismatched one.)

Spaced Review

Retrieve these from earlier chapters without looking back, then check against the collapsed answers.

1. (From Chapter 22) A collaborative proposal is declined with feedback that the team seemed disjointed. How does the resubmission skill apply, and what specifically would you strengthen?
2. (From Chapter 13) How does the organizational-capacity and letter-of-commitment craft of Chapter 13 extend to a multi-institution team and its subaward partners?
3. (From Chapter 11) How do indirect (F&A) costs work in a multi-institution budget, and why does a layered subaward budget contain multiple indirect rates?

Answers

1. Read the feedback analytically (Chapter 22): "team seemed disjointed" is a collaboration-credibility critique, so the resubmission should strengthen the leadership and coordination plan (clearer governance, decision-making, conflict resolution), add or foreground evidence the team has worked together, and — often overlooked — integrate the proposal into one coherent voice so it no longer reads as a patchwork. It's an agree-and-fix about the collaboration itself, not the science. 2. Chapter 13's capacity case (proving the team can deliver) extends to proving the whole multi-institution team can deliver, and its letter-of-commitment lesson becomes the requirement that each subaward partner provide a commitment signed by an authorized institutional official (binding the institution, not just the investigator) plus a scope of work — capacity and commitment, multiplied across partners. 3. Each institution applies its own negotiated indirect rate to its own direct costs, so a multi-institution budget layers multiple indirect rates; the prime applies its indirect rate only to a limited, allowable portion of each subaward (per funder rules), not to the whole sub — producing the layered structure that must be assembled carefully and reconciled.

Chapter Summary

Key Takeaways

• Some work needs a team — a trial across sites, a center, a training program, a consortium, a multi-org service program. When one PI isn't enough, the collaboration itself becomes half the proposal.
• Threshold concept: reviewers fund teams that have clearly already learned to work together. The collaboration's credibility — governance, coordination, evidence of a working relationship — is judged as hard as the science.
• Assemble a team only when the work needs one. Each partner must pass the test: what necessary, complementary capability does this partner contribute? A roster assembled for prestige is a weakness, not a strength.
• Choose a leadership model that fits the real division of leadership: single-PI with subawards, multiple-PI (MPI) for genuinely co-led work, or center/consortium structures (cores and components) for large efforts. The prime carries overall responsibility.
• The leadership and coordination plan is the most scrutinized element: governance and decision-making, roles, communication, conflict resolution, and evidence the team has worked together. Not boilerplate — the core argument that the collaboration is real.
• Subawards pass funds to partners, each with its own indirect rate, its scope of work, and a signed letter of commitment; the prime monitors subrecipients. Multi-institution budgets are layered and take far longer to assemble — start weeks early.
• One voice from many authors: one owner must integrate contributions into a single coherent proposal, because a patchwork signals a disjointed team and a coherent voice signals a coordinated one.

Action Items

1. Justify each partner with the necessary-complementary-capability test; cut dead-weight collaborators.
2. Choose the leadership model (single-PI+subs / MPI / center-consortium) that matches how the team really works.
3. Write a leadership and coordination plan with real governance, conflict resolution, and evidence of prior collaboration.
4. Build the layered subaward budget (each partner's own indirect rate, scope, signed commitment) on an earlier timeline.
5. Name the one owner who integrates all contributions into a single coherent voice.

Common Mistakes

• Assembling a team for prestige rather than necessary, complementary capability.
• Treating the leadership plan as boilerplate — vague "we'll coordinate collegially" instead of real governance and conflict resolution.
• Offering no evidence the team has worked together — leaving reviewers to bet on an unproven collaboration.
• Botching the subaward budget (indirect rates, the prime's limited indirect on subs) or gathering partner documents too late.
• Submitting a patchwork proposal in many voices, signaling a disjointed team.

Decision Framework — "Do I need a team, and how do I make it fundable?"

1. Does the work genuinely require multiple institutions? → Each partner must add a necessary, complementary capability; if not, stay single-PI.
2. What leadership model fits? → Clearly led → single-PI+subs; genuinely co-led → MPI; large multi-component → center/consortium.
3. Can I demonstrate the team works together? → Write a real leadership/coordination plan with governance, conflict resolution, and evidence of prior collaboration.
4. Are the subawards built and documented correctly, and early? → Layered indirect rates, scopes, signed commitments; start weeks ahead.
5. Does the proposal read as one voice? → Assign one owner to integrate all contributions into a coherent whole.

🔁 Carry this forward: Collaboration is the second of Part IV's cross-cutting skills. Next, grant writing with AI (Chapter 24) confronts a fast-changing tool that is reshaping how proposals are drafted, researched, and reviewed — with real power and real pitfalls. The team-coordination and one-voice disciplines you built here matter there too, as AI becomes another contributor whose output a human owner must integrate, verify, and make truly the team's own.