Part 1 · The project's core idea
Core Concept & Design Rules
One supervised, credit-bearing semester of real government service inside every undergraduate degree — the sixteen operating rules, the scope decision, the role limits that protect students and offices alike, and the decisions that remain open for government.
The idea in one page
Every eligible undergraduate in Bangladesh would complete one full semester of direct government service inside their degree — real, useful, authorized work in a government office, school, hospital support service, or public programme — before graduation. The service semester is substitutive, not additive: it carries academic credit that replaces existing elective or practicum requirements, so programme duration stays neutral and no student's graduation is delayed by design (RSK-01/02 controls). Students continue paying their existing university fees; universities bear programme costs; the government contributes hosting offices, supervisors and work — not money.
Government institutions declare how many students they can usefully absorb and on what tasks; the system publishes these as service seats; universities — not individual students — apply for seat allocations and book their own eligible students through a national digital platform; completion is verified triangulatedly (host supervisor + university + platform records) and becomes a graduation requirement. The pipeline, capacity and affordability maths are tested in the Feasibility pages of this site; the recommendation that emerges is a phased programme starting at pilot scale and rising to a ~130,000/year ceiling under the baseline funding rule, with the 300,000 target held open only as a conditional question for a later parliamentary decision (PL-15, UFM-19).
Objectives and the results chain
The programme is designed around a disciplined results chain, so that claims stay proportionate to evidence. The distinction matters: outputs are directly countable products of the programme; outcomes are changes in institutional behaviour that can be measured against baselines; impact claims (better public service, stronger accountability, reduced corruption) are held to the standard of independent evaluation — and are never promised in advance.
Outputs — directly measured
- Students placed, serving and completing per cycle (19-indicator M&E framework, CSV 22)
- Government registers digitized, records organized, surveys completed
- Verified digital completion records on the national platform
- Trained supervisors certified (target pipeline: 5,000–15,000)
Outcomes — measured against baselines
- Service-delivery process improvements documented by host offices (AC-01/02)
- Record completeness improvements in non-sensitive registers (AC-03)
- Graduate employability and civic-learning indicators (pilot evaluation)
- Platform-mediated transparency: allocation and utilization publicly reported
Impact — evaluated, not promised
- Any claim about reduced corruption requires independent evaluation design (Phase 3 report, Section 12)
- International evidence (Uganda capitation, Brazil audit audits, Indonesia KKN) supports plausibility, not proof
- Zero corruption-reduction savings are counted in the financial model (ASM-22, UFM-23)
Bangladesh's civil service has 521,922 vacant posts (26.3% of sanctioned strength) as of 30 June 2025, while BCS recruitment recommends only 1,457–2,805 candidates per cycle against roughly 40,000 annual retirements. Students on authorized, supervised tasks — records digitization, register completeness, citizen help desks, survey support — add capacity at the margin without displacing any employee. The design rule is explicit: students must never replace paid staff, and hosting density is capped by supervisor availability, not by vacancy counts (see the methodological note in the capacity model).
The sixteen baseline operating rules
These rules were fixed at project start and every design choice in this blueprint traces back to them. Where research found a rule could not be implemented as assumed, the finding is linked.
Scope and eligibility
Scope follows the affordability analysis (CSV 30) rather than ambition. The table below is the recommended initial compulsory scope under the baseline funding rule — National University colleges and the Open University are excluded from compulsion at entry, with voluntary pathways for later phases.
| Segment | Institutions | Final-year pool (derived) | Initial scope | Basis |
|---|---|---|---|---|
| Public direct-teaching universities | 49–53 | 52,651 | In — conditional on UGC budget-line circular | AFF-01, PL-10 |
| Private universities — top & mid tiers | ~70 | ≈55,000 of 77,527 | In with entry criteria (financial capacity screening) | AFF-02/03, PL-11 |
| Private universities — lower tier (<1,500 enrolled) | ~46 | ≈22,500 | Deferred / opt-in — compulsory scope would force closures or illegal fee rises | AFF-04 |
| National University affiliated colleges | ~2,500 | majority of national cohort | Excluded from compulsion — 64.3% of NU's UGC transfer; voluntary opt-in later | AFF-05, PL-15 |
| Bangladesh Open University (distance) | 1 | NA | Excluded — separate distance-service model required | AFF-06 |
What students may and may not do
Role limits are a first-order design rule, not an afterthought. The following boundaries apply everywhere, in every sector, regardless of discipline:
Students may — under named supervision
- Organize, index and digitize non-sensitive public records and registers
- Staff citizen help desks: form-filling assistance, information, queue management
- Collect authorized field and survey data (agriculture, environment, statistics)
- Support after-school tutoring, library organization, science-fair activity
- Perform data-quality checks and documentation for approved projects
- Observe service delivery against published citizen charters and log structured findings
Students may never
- Exercise statutory or inspection powers; make binding decisions on any citizen matter
- Handle personal, classified or confidential files (patient records, case files, personal land records)
- Participate in procurement evaluation, audit adjudication, or enforcement of any kind
- Perform clinical procedures, structural/engineering sign-off, or licensed legal work
- Confront or evaluate staff — individual blame assignment is outside programme scope (AC-02)
- Replace paid employees — hosting caps and task-mix rules enforce this (RSK-04, AC-06)
When a student observes a suspected irregularity, the only permitted route is the programme's protected referral pathway: written form → programme protection officer → department head / Anti-Corruption Commission / Public Interest Information Act (PIDA 2011) disclosure where applicable. Students never confront staff, and retaliation against a reporting student is a stand-alone programme offence (RSK-07, RSK-16). Full architecture on the Integrity page.
Open policy decisions for government
The research deliberately presents options rather than inventing rules. Decisions that remain open — each with the evidence needed to take it — include:
- Graduate-level repeat service — recognize a prior undergraduate completion, or require a second (shorter) service block for graduate degrees? (Programs page, Option A/B/C)
- The 300,000 question — whether to legislate for a scale that structurally requires NU college participation, or phase to the ~130,000 baseline-feasible ceiling and revisit (PH-4 gate).
- Compulsion timing — legislate the compulsory element before the pilot (risk: legislating ahead of evidence) or after Gate 1 (risk: voluntary pilot under-participation).
- NU college pathway — voluntary opt-in during PH-2, or a separately funded national-service track outside the university-funded baseline?
- Seat-approval authority — sectoral ministries, a central national programme office, or a mixed model (options on the Seats page).
- Stipend policy — the no-stipend baseline is legally defensible (LEG-18) but carries equity risk (RSK-03); the need-based fund covers ~30% of participants at BDT 1,500–3,000/month (ASM-25). A universal stipend is a deliberate departure from the baseline and must be decided openly.