Universal · 9 min read

Grant Budgets Explained: Direct, Indirect, FEC, and How Reviewers Read Them

A grant budget tells reviewers two things: can this team execute the proposed work? and does this team understand what the work actually costs? Most under-budgeted proposals are penalised as much as over-budgeted ones. This guide covers the universal budget categories, the rules around indirect costs, and the most common mistakes.

1. Direct costs vs indirect costs

Direct costs are spent specifically on the project: salaries of project staff, consumables, equipment, fieldwork, conferences, publications.

Indirect costs (also “overheads”, “F&A” in the US, “FEC uplift” in the UK) cover the institution’s shared services that enable the project — libraries, IT, HR, building utilities, research office. Funders apply different indirect-cost rates:

• NIH: institutional rate, negotiated with US government. Typical 50–70% of direct costs.
• NSF: same negotiated rate as NIH.
• Horizon Europe: flat 25%.
• UKRI (FEC): 80% of full economic costs — institution covers the remaining 20%.
• FCT: 25% of direct costs.
• DFG: 22% of direct costs (project lump sum).

2. Personnel: the biggest line

Personnel typically accounts for 60–80% of direct costs. Universal rules:

• PI salary: usually not chargeable to most national bottom-up grants (Sachbeihilfe, Discovery Projects, AAPG, FCT IC&DT). Personal-fellowship grants (ERC, FLF, Vidi) do cover PI salary.
• Postdoc salaries: at the host institution’s standard rate. Some funders set minimum (NIH NRSA scale, Marie Curie unit cost).
• PhD students: scholarship + bench fees. Typical 3 years for European/UK; 4–5 years for North America.
• Technicians and research assistants: at standard rate. Effort percentage must match workplan.

3. Equipment

Most funders cap equipment at a percentage of total budget (typically 25–50%). Equipment over a threshold (often €25,000 or $50,000) requires explicit justification with quotes.

Reviewers ask three questions about each equipment line:

• Is this needed for the project? (Map to specific aim/method.)
• Is it not already available at the host institution?
• Is the model and pricing realistic? (Padded equipment is flagged.)

4. Consumables

Consumables include lab reagents, software licences, computing time, fieldwork supplies, sequencing services. The simplest format: tabulated by aim or work package, with cost-per-unit and total quantity. Lump-sum “consumables: €30,000” without breakdown invites scrutiny.

5. Travel

Most funders allow conference travel, fieldwork travel, and collaboration travel. Typical caps:

• Conference: 1 international per project-year for the PI; 1 per year for trainees on the project.
• Fieldwork: justified by work package.
• Collaboration: short-term visits to consortium partners.

6. Publications and dissemination

Open-access fees should be budgeted explicitly. Typical €2,000–€3,000 per article. Some funders (Wellcome, Plan S) have specific OA policies. Workshops, summer schools, and patient/public engagement also fit here.

7. External services

Sequencing, mass spectrometry, animal model generation, statistical consulting. Capped at 20–30% in most funders. Each external-service line should name the provider and explain why this is more efficient than in-house work.

8. Indirect costs: the structural difference

Different countries treat indirect costs differently. Three patterns:

• Negotiated rate (US): each institution negotiates its rate with the federal government. The same project at MIT vs at a state university can have very different total budgets for the same direct costs.
• Flat rate (EU, FCT): 25% on direct costs, regardless of institution. Predictable but sometimes inadequate for full cost recovery.
• FEC (UK): full economic cost calculated, then funder covers 80% (UKRI), institution covers 20%. Encourages institutional buy-in.

9. The five most common budget mistakes

• Personnel without workpackage anchor — a postdoc listed but no aim depends on them.
• Equipment without justification — a sequencer listed in budget; no methodology line explains why.
• Salary uplift errors — institutional pay scale increases not factored in for multi-year budgets.
• Consumables as lump sum — no breakdown by category or by aim.
• OA fees missing — no budget line for open-access publication, even though the project promises X publications.

10. Tips

• Build the budget after the workplan, not before. Each budget line should map to a workplan task.
• Use the institution’s standard rates — don’t round to nicer numbers.
• Double-check the indirect-cost rate against the funder’s rule for your country.
• Get the host institution’s research office to validate the budget 2 weeks before deadline.
• Consider salary uplift of 2–3% per year for multi-year budgets.

Track every funder’s call automatically

Grant Radar tracks 24 funders worldwide and emails you the calls that match your ORCID profile, with links to the official budget rules. Free during the public beta.