Every research-performing institution has to decide, structurally, where the people who prepare proposals, negotiate awards, monitor spending, and manage compliance actually sit in the org chart. There are two dominant answers — departmental (distributed) research administration, where staff are embedded in academic departments or schools close to the principal investigators (PIs) they support, and a central sponsored programs office (commonly abbreviated SPO or OSP, for Office of Sponsored Programs), a single institution-wide unit that administers awards consistently across the whole institution. Most real institutions run some blend of the two rather than a pure form of either. This guide covers what each model actually does, the genuine trade-offs between them, how a hybrid (“hub-and-spoke”) structure tries to capture the benefits of both, and — a distinction that gets conflated with this one surprisingly often — how the departmental-vs-central organizational question relates to, but is not the same axis as, the pre-award/post-award functional split in grant administration.
The two dominant organizational models
Departmental (distributed) research administration
In a departmental model, research administration staff — often called departmental research administrators (DRAs) — are hired by, report to, and sit physically within an academic department, school, or college, rather than a central office. The National Council of University Research Administrators (NCURA), the main U.S. professional association for this field, describes DRAs as distinguished from other research-support roles specifically because they are “intimately involved with all facets of the research administration process, including daily interactions with faculty” — and typically manage the full lifecycle for their unit’s grants, not just one phase: proposal development, funded-award management, compliance monitoring, and closeout.
That proximity is the model’s defining advantage. A DRA who works down the hall from a principal investigator, understands the discipline-specific quirks of that department’s funders, and has an ongoing working relationship with the same faculty year after year can move faster and give more tailored guidance than a generalist in a central office handling requests from dozens of departments at once. The trade-off is standardization: department-by-department administration means practices, documentation habits, and even compliance interpretation can vary meaningfully across a single institution, and small departments may not have the volume of awards to justify (or fund) a dedicated, well-trained DRA at all — a real capacity/consistency risk that shows up disproportionately at smaller or lower-research-volume units.
Central sponsored programs office (SPO/OSP)
In a central model, one institution-wide office — variously called the Office of Sponsored Programs, Office of Research Administration, Sponsored Programs Office, or similar — handles award administration for the whole institution (or a large share of it), with standardized procedures, centralized institutional signature authority, and dedicated compliance and financial-reporting specialists. NCURA’s own peer-review program for central offices describes two common reporting structures: a combined, single reporting line (typically up through academic/research leadership, occasionally through finance or advancement) or a split operation with multiple reporting lines, most commonly dividing responsibility between the academic and financial sides of the institution.
Centralization’s advantage is consistency: one set of procedures, one institutional signature authority, one place where regulatory changes (a new Uniform Guidance revision, a funder policy update) get implemented once rather than department-by-department, and — because compliance specialists handle high volumes of similar transactions — deeper regulatory expertise than most individual departments can build or afford on their own. NCURA’s peer-review criteria for central offices explicitly assess risk management/compliance, operational efficiency, business-process consistency, and responsiveness to sponsors, which is a reasonable proxy for what the model is actually optimized for. The trade-off is responsiveness and specificity: a central office serving an entire institution has less bandwidth per PI, less exposure to any one department’s discipline-specific norms, and can become a bottleneck when proposal or award volume peaks (grant deadlines cluster heavily around a handful of dates each year).
The hybrid model
Because neither pure model is obviously superior, most research-intensive institutions run some form of hybrid — often described using the “hub-and-spoke” framing common in broader higher-education administration: a central “hub” retains institution-wide functions that most benefit from standardization and concentrated expertise — regulatory compliance, institutional signature authority, financial reporting, systems administration, effort reporting, indirect-cost recovery — while departmental “spokes” retain the PI-facing, discipline-specific, high-touch functions: proposal development support, day-to-day budget monitoring, and the ongoing relationship with faculty. Exactly where the line falls between hub and spoke functions is an institutional design choice, not a fixed formula, and it’s common for large research universities to have a different balance in a medical school or engineering college (often more departmental capacity, given research volume and complexity) than in smaller humanities or social-science departments (often served almost entirely by the central office).
Trade-offs at a glance
| Dimension | Departmental (distributed) | Central (SPO/OSP) |
|---|---|---|
| Speed / responsiveness to a given PI | Faster — dedicated staff, ongoing relationship | Slower — shared capacity across many PIs, deadline clustering creates bottlenecks |
| Consistency of practice and compliance interpretation | Lower — varies department to department | Higher — one set of procedures applied institution-wide |
| Discipline-specific / funder-specific expertise | Higher — staff specialize in their department’s typical funders and award types | Lower per-transaction, but broader regulatory depth overall |
| Cost efficiency / economies of scale | Lower — redundant staffing and training across departments; small units may be under-resourced | Higher — shared systems, training, and specialist compliance staff |
| Institutional risk visibility | Fragmented — hard to see systemic risk across departments | Consolidated — one office can track compliance risk institution-wide |
| Resilience to staff turnover in a given unit | Lower — a single DRA’s departure can leave a department without institutional knowledge | Higher — coverage doesn’t depend on one person |
How this connects to the pre-award/post-award functional split
Departmental-vs-central is an organizational axis: where in the institution a function sits. The pre-award/post-award distinction is a functional axis: which part of the grant lifecycle a given piece of work belongs to — pre-award covering opportunity identification through proposal submission and award negotiation, post-award covering everything from the start of the period of performance through expenditure monitoring, reporting, and closeout. These two axes are independent of each other, and conflating them is a common source of confusion: an institution can organize pre-award work departmentally and post-award work centrally, or vice versa, or run either phase as a hybrid.
In practice, a recognizable pattern does show up across many institutions: post-award work — with its concentrated compliance, audit, and financial-reporting risk (effort reporting, indirect cost recovery, subaward monitoring, federal reporting deadlines) — is somewhat more likely to be centralized than pre-award work, which benefits more directly from a DRA’s proximity to the PI during proposal development. But this is a tendency, not a rule, and plenty of well-run institutions centralize both phases, or decentralize both, depending on their size, research volume, and risk tolerance. Anyone mapping their institution’s own structure should treat “who does pre-award vs. post-award work” and “is that person departmental or central” as two separate questions with two separate answers, not one combined variable.
What actually drives the choice
There’s no single correct model — NCURA’s own peer-review methodology for both departmental and central-office reviews explicitly incorporates “the institution’s size, mission, and culture” as part of the assessment, rather than benchmarking every institution against one ideal structure. Factors that tend to push an institution toward more centralization: high award volume relative to administrative staff capacity, a need for consistent compliance posture across many departments (especially after an audit finding or major compliance lapse), and cost pressure that makes redundant departmental staffing hard to justify. Factors that tend to push toward more departmental capacity: research-intensive units (a medical school or large engineering college, for example) with award volume and complexity that justifies dedicated staff, disciplines with unusual funder requirements that a generalist central office struggles to serve well, and a strong institutional preference for close PI-administrator relationships. Most institutions land somewhere on a spectrum between the two poles rather than at either extreme, and the position on that spectrum often shifts over time — a compliance incident tends to push an institution toward more centralization; PI frustration with slow turnaround tends to push it back toward more departmental capacity.
Why the organizational model matters beyond org-chart design
The choice between departmental, central, and hybrid administration has a direct, practical consequence for anyone working on data standards and interoperability in research administration — CASRAI’s own area of focus. Whichever model an institution uses, the underlying data about awards, organizational units, and personnel still has to be represented consistently if it’s going to move cleanly between a departmental grants-management spreadsheet, a central office’s financial system, a CRIS (current research information system), and a funder’s own reporting portal. A distributed model, in particular, multiplies the number of people entering or touching award data by hand, which raises the practical value of a shared organization unit model, consistent identifiers, and controlled vocabularies that don’t depend on every department independently reinventing the same fields — the same interoperability problem CASRAI’s Dictionary and CERIF-aligned data model exist to solve. A central office, by contrast, has an easier time enforcing one system and one set of field definitions, but still benefits from the same standards when exchanging data externally with funders, sponsors, and other institutions.
Frequently asked questions
What is the difference between departmental and central research administration?
Departmental (distributed) research administration means dedicated staff sit within an academic department or school and serve that unit’s principal investigators directly, across the full award lifecycle. Central research administration means one institution-wide office (an SPO/OSP) administers awards for the whole institution using standardized procedures. Most institutions use some blend of both rather than a pure form of either.
What does a sponsored programs office do?
A sponsored programs office (also called an Office of Sponsored Programs, OSP, or Office of Research Administration) is the central unit that administers externally funded research awards institution-wide — typically holding the institution’s signature authority for proposals and awards, applying consistent compliance procedures, and often specializing in financial reporting, indirect-cost recovery, and regulatory compliance (for example under the Uniform Guidance at 2 CFR 200).
Is research administration usually centralized or decentralized?
Neither is universal — it depends heavily on institution size, research volume, and risk profile. Most research-intensive universities run a hybrid, with some functions (proposal support, day-to-day budget monitoring) handled by departmental staff and others (compliance, institutional signature authority, financial reporting) handled centrally. Larger, higher-research-volume units within an institution (a medical school, for example) more often have dedicated departmental capacity than smaller units, which are more likely to rely entirely on the central office.
What is a hybrid research administration model?
Often described with the “hub-and-spoke” framing: a central “hub” handles functions that most benefit from standardization and specialist expertise (compliance, signature authority, financial and effort reporting), while departmental “spokes” retain PI-facing functions that benefit from proximity and discipline-specific familiarity (proposal development support, ongoing budget monitoring). The exact split is an institutional design choice, not a fixed formula.
Who handles pre-award vs. post-award activities?
That depends on the institution’s organizational model, not on the pre-award/post-award distinction itself — the two are separate axes. Either phase can be handled departmentally, centrally, or through a hybrid split. In practice, post-award work (with its concentrated compliance and financial-reporting risk) is somewhat more often centralized than pre-award work, but this is a common pattern, not a rule.
Related CASRAI resources
- Pre-award phase — dictionary definition and worked examples
- Post-award phase — dictionary definition and worked examples
- Organization unit — how institutional structures are modeled in CASRAI’s data standards
- Principal investigator (PI)
- Uniform Guidance (2 CFR 200): The Governing Framework for Federal Research Grants
- Effort Reporting Methodologies
- FDP Subaward Templates
- Research Integrity & Compliance — cluster overview
