Read first: two TICGL studies that frame this research
This report asks who pays when a project must repay lenders and also serve low-income users. These two companion studies ask why Tanzanian projects fail to reach financial close in the first place, and how the capital stack that sits above the equity should be built.
Tanzania's investment gap stems mainly from weak project preparation, limited long-term debt and shallow capital markets, not from a lack of investor interest. Only about 22% of the USD 7.7 billion of FDI registered in 2024 was disbursed.
Read the study →The shortfall comes mainly from the inability to assemble the debt and blended-finance layers above equity. It points to pension-fund infrastructure allocations, a viability gap fund and a long-tenor infrastructure lender, the same levers this report tests for LGA projects.
Read the study →Prepared by TERI for research and strategic-information purposes only. It is not investment, legal, tax or financial advice. The markets and bus terminals discussed are case studies of a wider LGA problem, and the empirical illustrations use an anonymised pre-feasibility case; project, place and respondent names are withheld. Construction-cost, yield and operating-cost inputs are explicit assumptions (listed under Sources & Assumptions) and must be replaced with quantity-surveyor and financial-model figures for any real project. The case evidence is small, non-random and self-reported. Press and public-domain evidence is quoted as reported and is not always like-for-like with a PPP.
00 — OverviewExecutive Summary: Bankable on Paper, Unaffordable in the Market
Tanzania's decentralisation reforms aim to give LGAs more fiscal autonomy, and the current policy push adds private financing and PPPs for revenue-generating assets across sectors, including housing, waste, tourism, public transport, markets, shop complexes and bus terminals. The logic is attractive: the private sector brings capital and management, the LGA keeps the land and earns a share of revenue. The difficulty is structural. When the asset earns only from charges, the only cash flows available are the rents, fees and tariffs paid by the very people the project is meant to serve. TICGL/TERI tested whether bankability and affordability can be met at once, using markets and bus terminals as case studies, with public evidence on Tanzanian and regional experience and data from one illustrative project.
The same arithmetic applies to any LGA project planned or operated on user charges: required revenue is set by capital cost, cost of money and operating cost, while affordable revenue is set by what users can pay. Markets and terminals are used here because their rents, fees and bypass behaviour can be observed and counted. The nine-point screen in Tab 9 and the pass-through test in Tab 7 are written to be reused for other LGA projects.
Rent Needed vs Rent Users Can Afford
Yield on Cost: What Affordable Rent Earns vs What Capital Demands
Key findings
- The gap is large and robust to method. Three independently built models find that on commercial terms the rent needed to repay the building is 3.4 to 7 times what the median respondent says they can pay (median about TZS 8,300 per lettable m² per month; upper quartile TZS 16,700).
- Affordable rents carry only a fraction of the building. At stated affordability, rent supports roughly 14–39% of construction cost and yields 2–5% on cost, against about 10.8% on a 15-year government bond. A 50% capital grant plus cheaper long-tenor debt still leaves required rent at 1.3–2.6 times what traders say they can pay.
- Cross-subsidy has a hard ceiling. If 50–80% of space is let at affordable rents, anchor tenants would have to pay TZS 105,000–247,000 per m² per month. Half of prospective tenants (8 of 16) can pay more than TZS 200,000, which helps but does not close the gap.
- The gap does not disappear; it moves. Whoever closes it (traders, consumers, taxpayers or investors) bears it. Making a project bankable by user charges transfers the cost to traders and, where they can, to the citizen.
- For small traders, pass-through is arithmetically impossible. Raising rent from TZS 75,000 to TZS 300,000 would need a 90% price rise at TZS 250,000 of sales (45% even at the top of the band); at TZS 520,000 rent, 178%. In practice traders exit, sub-let or go informal.
- For high-volume operators and commuters, pass-through is feasible but regressive. A restaurant adds roughly TZS 50–135 per customer (1–3% of spend); a TZS 1,000–2,000 stand fee is about TZS 35–135 per passenger per trip, and 9 of 12 surveyed customers visit daily.
- Tanzania's own terminals show the risk. The TZS 54.6 billion Magufuli terminal (about 93% central-grant funded) struggles with bypass, municipal revenue losses and TZS 249 million of rent arrears, while Morogoro's Msamvu terminal shows what demand-led siting achieves, though its reported takings equal only about 7% of phase-one cost.
- The way forward is unbundling, explicit public support and a screening test. Separate the public-good core from the commercial layer, fund the core through VGF, grants or concessional capital, protect sitting traders with a protected tier inside tiered rents, and apply a nine-point affordability–bankability screen at pre-feasibility stage.
Bankability does not remove the subsidy; it moves it from the budget to the wananchi. The question for every LGA market or terminal is who should carry the viability gap: taxpayers through a visible grant, or small traders and commuters through rents, prices and fees. The second route is regressive and is rarely shown in feasibility studies. The next ten tabs show the policy setting, the evidence, the numbers and the options.
01 — Policy & Fiscal ContextA Thin Local Revenue Base, a New PPP Framework and Expensive Money
LGA own-source revenue (OSR) remains a small part of public finances, which is why revenue-generating assets are attractive to councils. At the same time the capital used for the recent generation of terminals has come mainly from the centre, and the PPP pipeline is large but converts slowly. This tab sets out the fiscal and legal backdrop against which the bankability test in Tabs 5–6 should be read.
1.1 A thin local revenue base
| Budget | LGA own-source revenue (TZS trillion) | Total budget (TZS trillion) | LGA OSR share of budget |
|---|---|---|---|
| 2024/25 | 1.36 | 49.35 | ≈ 2.7% (reported as "about 3%") |
| 2026/27 | 1.97 | 62.33 | ≈ 3.2% |
Sources: Ministry of Finance Budget Frame 2024/25 (TZS 1,356,341 m of TZS 49,345,688 m); 2026/27 budget (TZS 1.97 trillion within TZS 62.33 trillion); shares are TICGL/TERI calculations. A UNDP-supported workshop report (June 2025) notes OSR has stayed low, at about 15.5% across three fiscal years, despite electronic collection systems.
LGA Own-Source Revenue and Its Share of the National Budget
The PPP Pipeline vs Signed Deals
Self-reliance is therefore a direction of travel rather than a fiscal reality. Completed terminals, including Magufuli at Mbezi Luis, Dodoma's terminal and Msamvu in Morogoro, are presented in official and press reports as a strategy to reduce councils' dependence on central government, yet their capital came mainly from the centre (Tab 3).
1.2 The PPP framework: what changed in 2023
| Feature of the PPP (Amendment) Act No. 4 of 2023 and GN 838A | Why it matters for LGA markets and terminals |
|---|---|
| "Public funding" defined as support creating fiscal commitments or contingent liabilities | Makes the cost of closing a viability gap visible and countable; it cannot be hidden in optimistic revenue forecasts. |
| Explicit provision for tax benefits, guarantees and Viability Gap Funding (VGF) | The law already recognises that some socially valuable projects cannot recover their cost from users. VGF exists for exactly this case. |
| Concept-note stage removed; pre-feasibility studies due at the start of each budget cycle | Pre-feasibility is the right moment to run an affordability–bankability screen (Tab 9). |
| Contracting authority may hold part ownership of a project company | Enables council-plus-patient-capital structures such as the Msamvu joint company (Tab 3). |
The PPP Centre reported in September 2025 that 410 projects had been identified across 184 LGAs in sectors including housing, waste, tourism and public transport, with only a handful at implementation stage. By June 2026 nine PPP contracts had been signed nationally (DART phases, mandatory vehicle inspection, the Kariakoo commercial building, TAZARA and the Dar es Salaam port), mostly national-scale flagships. The gap between the local pipeline and closed deals suggests many local concepts are not yet bankable. An older World Bank-supported municipal PPP presentation (undated; to be checked against current regulations) reached a similar diagnosis: projects were economically attractive but fiscally unaffordable, needing viability-gap or availability payments most municipalities outside Dar es Salaam could not fund.
1.3 The cost of money
Private capital is priced off the government yield curve. The 15-year Treasury bond yield stood at about 10.79% in the week to 2 October 2026, while 20- and 25-year yields were above 14% during 2025. A property investor in a secondary-town market will require a premium over these risk-free rates for construction, letting and regulatory risk. The report uses a base required yield on cost of 16%, with 12% as a concessional-finance case and 18% as a stressed case, and cross-checks with a lender-style model using a debt-service-cover test.
Required Yield on Cost vs Government Bond Yields
02 — Why the Two Logics CollideTwo Roles, a Triangle, a Loop and a Formula
A single LGA project asks the council to be two things at once: a service provider, whose job is access, equity, safety and public health, and an investor and asset owner, whose job is to earn a return and recover the cost of capital. These roles use different customers, different success measures and different views on price.
2.1 Two roles in one project
| LGA as service provider | LGA as investor / asset owner | |
|---|---|---|
| Objective | Access, equity, safety, public health | Own-source revenue; recover cost of capital |
| The "customer" | Small trader, commuter, resident | Tenant able to pay a market rent |
| Success measure | People reached, accidents avoided, sanitation, jobs | Cash flow, debt-service cover, investor return |
| Pricing view | Affordable or subsidised | Cost-reflective rents and fees |
| Main risk | A project that costs but does not earn | A project that excludes those it should serve |
2.2 The triangle: public value, affordability, bankability
Market and stand projects try to deliver three things at once. In practice any two can be achieved, but the third breaks without additional support.
| Combination | What is possible | What is lost |
|---|---|---|
| Public value + affordability | Good facility, low charges, funded by public capital | Bankability: no cash flow to repay commercial debt |
| Bankability + affordability | A small, cheap, basic facility | Public value: safety, sanitation and infrastructure are stripped out |
| Bankability + public value | A modern facility that repays its debt | Affordability: rents and fees exclude small users |
2.3 The vicious loop
A lender also prices as cost the objectives an LGA must retain: police post, CCTV, drainage, toilets, inclusion of sitting traders, fee restraint. They are legitimate, but they raise capital cost and common-area ratios while earning little rent.
2.4 The bankability arithmetic
Rent = [ (Construction cost per gross m² ÷ Net-to-gross efficiency) × Required yield on cost ] ÷ [ (1 − operating-cost ratio) × (1 − vacancy) × 12 ]
// Base case
(2,200,000 ÷ 0.70) × 16% ÷ (0.80 × 0.90 × 12) ≈ TZS 58,200 per m² per month
Every term other than construction cost is a lever that bankability pushes up: higher required yield, lower efficiency (more common areas), higher vacancy and operating costs. Lenders also test debt-service cover (commonly 1.2–1.4x), which raises required revenue further. Try your own assumptions below, and see which lever moves the answer most.
Required-rent calculator (Model A)
Change the inputs; results update instantly. Capital contribution scales required rent by (1 − contribution), as in Table 15 of the report.
Which Lever Moves the Required Rent Most?
2.5 Conservation of the gap
The viability gap is conserved: the difference between what the facility must earn and what users can pay does not vanish when a project is structured as a PPP; it is assigned to someone. There are only four candidates.
| Who pays | Mechanism | Typical consequence |
|---|---|---|
| Traders (tenants) | Higher rents, deposits and service charges | Exit, under-letting, arrears, informal trading outside the building |
| Consumers / commuters | Higher prices, parking, toilet and stand fees passed through | Regressive burden; reduced footfall |
| Taxpayers / the LGA | VGF, capital grants, land contribution, guarantees | Fiscal commitments and contingent liabilities |
| Investors / lenders | Lower returns or higher risk accepted | Project does not reach financial close |
2.6 Other mechanisms that bite
| Mechanism | How it bites |
|---|---|
| Price-setting limits | Rents, stand fees and levies are set by by-law and political judgement, not cost recovery; leaders resist fees that hurt small traders. A cost-reflective tariff may be legally or politically unavailable. |
| Avoidance and bypass | If the fee is the product, users avoid the facility when alternatives exist (Tab 3). |
| Optimism bias | To look bankable, forecasts of vehicle numbers, occupancy and ability to pay are inflated; when they fail, the LGA is left with direct or contingent liabilities. Support for a project (95% in the case) is not the same as ability to pay. |
| Long contracts, changing policy | Directives on fees or rent relief can arrive mid-concession; investors price protection into the deal. Clear tariff-indexation and compensation rules are part of bankability. |
| Maintenance neglect | Without a ring-fenced sinking fund, quality decays and revenue follows. |
| Displacement | Sitting traders fear losing customers during relocation and higher rent on return; without priority-of-return and rent protection, the project displaces rather than strengthens small business. |
2.7 Financial versus economic returns
A project can have a low financial return (FIRR) yet a high economic return (EIRR) once avoided accidents, sanitation, time savings, jobs and council revenue are counted. Where that is so, forcing it to "pay for itself" means redesigning the project to suit the lender rather than designing support that matches its public value.
| High EIRR | Low EIRR | |
|---|---|---|
| High FIRR | Standard PPP: private sector can carry the debt; regulate tariffs | Purely commercial; the LGA need not participate except as landowner |
| Low FIRR | Deserves support: VGF, concessional loans, availability payments, or public build with private operation | Do not build; reconsider design, site or need |
03 — Tanzanian & Regional EvidenceWhat Terminals and Markets Have Actually Done
The public record is mixed in financing structure, so these cases are not like-for-like PPP comparisons. They share the same revenue and affordability logic, and they show what happens when a facility must earn from the people it serves.
3.1 Magufuli terminal, Dar es Salaam: a grant-funded warning
The Magufuli terminal at Mbezi Luis was built at about TZS 54.6 billion, of which TZS 50.9 billion (93%) came from a central-government grant and the balance (about TZS 3.7 billion) from the Ubungo council's own revenue. It is a grant-funded project, not a PPP, but its revenue logic is the one a bankable project would rely on. Press reports describe 70 to 90 buses a day using other stops and a municipal loss first put at more than TZS 50 million a month and later reported as more than TZS 600 million a month in government discussions. The two figures are not reconciled; the earlier is consistent with a report that at least TZS 1.4 million a day was being missed. Bus owners cite distance from the centre, passenger convenience, private terminals with better facilities, and a mandatory entry requirement that conflicts with the regulator's approval of private terminals.
Passengers to Kigamboni, Mbagala, Temeke and the city centre often pay more for onward transport, and the Prime Minister has himself warned against forcing buses into terminals where this would raise operating costs and fares.
In September 2026 the Ubungo council was pursuing over TZS 249 million of rent arrears from businesses in the passenger building. Traders cited low passenger flow, switched-off lifts and mixed uses, and asked for rent relief.
What the Terminals Cost
Magufuli: Two Reported Monthly Revenue Losses
3.2 Dodoma and Morogoro: location and partner structure
Dodoma's terminal, built at over TZS 24 billion, has been reported as under-used, with empty shelves and deteriorating fittings. By contrast, Msamvu in Morogoro, developed by the council and the LAPF pension fund through a joint company, cost about TZS 16 billion for its first phase, and reported takings rose from about TZS 350,000 to about TZS 3 million a day after expansion. Even that success equals roughly TZS 1.1 billion a year, about 7% of phase-one cost before operating expenses and tax. It shows how hard it is for a terminal to repay capital from fees alone, and why siting on a busy corridor and a long-horizon partner matter.
Msamvu: Daily Takings Before and After Expansion
Msamvu: A Year of Takings vs Phase-One Cost
3.3 Regional market redevelopments
Kampala (Uganda)
- Traders left new stalls because rents were high relative to what they paid before; stalls were sold on after tax holidays ended.
- Authorities cut rents in some markets, yet many stalls reportedly stayed vacant.
- On 30 September 2026 traders again raised high shop rents with the city authority, which had 550 vacant spaces in government markets.
Kumasi, Kejetia (Ghana)
- Traders said store rents were too high; the Assembly arranged bank loans so traders could pay premiums over five years.
- Members alleged premiums were inflated to service the construction loan.
- Some phase-one shops bought for GH¢25,000 were resold for GH¢100,000–120,000: subsidised space captured by better-connected holders.
Takoradi (Ghana)
- A University of Bergen study of a proposed PPP-financed market redevelopment found likely implications for informal traders' access to trading space.
- Lesson: displacement is a design risk to be planned for, not an afterthought.
Kejetia, Kumasi: Allocation Price vs Resale Price of a Phase-One Shop
3.4 Lessons
| Case | What happened | Lesson |
|---|---|---|
| Magufuli, Dar es Salaam | Large grant-funded terminal; bypass, municipal revenue losses, tenant arrears | Fee-based design fails when users have alternatives; siting, feeders and fee structure matter as much as the building |
| Msamvu, Morogoro | Council–pension-fund joint company; strong corridor location; takings up about 9x | Demand-led siting and patient capital help, but fees alone still recover only a small share of capex |
| Kumasi and Kampala | Premiums, rents or stalls priced above traders' means; vacancy, resale and capture | Cost-recovery pricing without allocation rules displaces target users |
| World Bank-supported urban projects (Tanzania) | Results frameworks target 70–80% recovery of bus-station O&M cost from fees | Even the donor benchmark aimed at operations and maintenance, not full capital recovery |
04 — Case: What Users Can PayEvidence from an Illustrative Case: Strong Support, Thin Wallets
The case is a proposed mixed-use complex combining shops and a mini-bus stand in a fast-growing urban ward, used here as a case study of an LGA project that must earn from its users. It draws on structured interviews with institutions and police, prospective tenants, existing traders, and customers and commuters, plus restaurant and market leaders. Respondents' monthly rent ranges were converted to bucket mid-points and divided by the floor area they said they would need. The evidence is small, non-random and self-reported, so results indicate direction and order of magnitude. The full survey detail is available in the full report (see Request the full report).
4.1 Support is high; the issue is price, not demand
All 21 existing traders said they would move into a new building (13 "strongly support", 8 "support"), and 11 of 12 customers said they would use the facility regularly. Local leaders and departments backed the concept and stressed priority for sitting traders, fair compensation, sanitation and security. Demand is not the constraint; the constraint is what that demand can pay.
4.2 Affordability
Maximum Monthly Rent Respondents Say They Can Pay
The Affordability Curve: Share Who Can Pay No More Than X
| Indicator | Value | Comment |
|---|---|---|
| Existing traders able to pay at most TZS 100,000 / month | 11 of 21 (52%) | 20 of 21 (95%) cap at TZS 200,000 |
| Prospective tenants capped at TZS 200,000 | 8 of 16 (50%) | 6 in the 200–300k band; 2 above 300k (a financial outlet, a hardware store) |
| Respondents with usable size and rent data | 37 | Median stated maximum TZS 150,000; 76% (28 of 37) at or below TZS 200,000 |
| Median stated rent per m² | TZS 8,300 / month | Interquartile range 8,300–16,700; highest 17,100 |
| Existing traders: current rent (median) | TZS 75,000 / month | Mean stated maximum ≈ 1.4x current rent |
| Existing traders with sales under TZS 500,000 / month | 76% (16 of 21) | One preferred not to say |
| Rent-to-sales at top of sales band (median) | ≈ 15% | Rent of 100–200k is 20–40% of top-of-band sales for the lowest band |
Table 7. Affordability indicators. Per-m² figures use assumed unit sizes (kiosk 6, small 9, medium 18, large 35, restaurant 40 m²).
4.3 Three observations that matter for design
1 · Unit mix drives revenue density
- Small units (under 12 m²) show the highest stated rent per m² (median about TZS 16,700).
- Medium units show about TZS 8,300.
- A mix weighted towards small units raises revenue per m².
2 · Haircut stated rents
- Low reported sales combined with rents already near 15% of sales suggest under-reported turnover or optimistic stated willingness.
- Revenue forecasts built on stated rents should be discounted.
3 · Non-revenue needs are large
- 11 of 21 traders and 8 of 12 customers named missing or dirty toilets and water among main complaints.
- Officials and police asked for a police post, CCTV, lighting, separate gates, waste collection, wastewater treatment, drainage and public toilets.
- All add capital cost without adding rent.
Advance Rent Prospective Tenants Could Pay
Support, Concerns and Complaints
05 — The Bankability TestAffordable Rents Earn 2–5% on Cost; Capital Needs 12–18%
Affordable rents deliver a yield on cost of only about 2.3% (median), 3.0% and 4.6% (upper quartile), all below the 10.8% government bond yield. At stated rents the building would be a worse investment than lending to the Government.
Required Rent Rises With Cost and Yield, Far Above What Users Can Pay
How Much of the Building Can Rent Support?
Across the affordable range of TZS 8,300–17,100, rent supports roughly 14–29% of construction cost at a 16% required yield and 19–39% at 12%. The balance, 61% to 86% of capex, is the viability gap in this illustration.
5.1 Required rent by construction cost and yield
| Cost per gross m² (TZS m) | 12% yield | 14% yield | 16% yield | 18% yield |
|---|---|---|---|---|
| 1.2 | 23,800 | 27,800 | 31,700 | 35,700 |
| 1.6 | 31,700 | 37,000 | 42,300 | 47,600 |
| 2.0 | 39,700 | 46,300 | 52,900 | 59,500 |
| 2.2 (base) | 43,700 | 50,900 | 58,200 | 65,500 |
| 2.4 | 47,600 | 55,600 | 63,500 | 71,400 |
| 2.8 | 55,600 | 64,800 | 74,100 | 83,300 |
Table 8. Required rent (TZS per lettable m² per month). No cell falls within the stated affordability range (median 8,300; highest 17,100). Cell shading: amber under 35,000; red 35,000–55,000; dark red above 55,000.
5.2 Three models, one conclusion
To test robustness, the gap was estimated with three independently built models: (A) the yield-on-cost model above; (B) a simple capital-recovery model (capital charge over 15 years at 16%, 85% occupancy, no operating costs, so a floor); and (C) a lender-style model requiring 1.3x debt-service cover with 70% debt at 17% over 12 years, operating costs and a maintenance reserve. Models B and C were re-computed independently and reproduce the figures in the working drafts.
Monthly Rent Needed for an 18 m² Shop, Three Models
How Many Times the Median Trader's Affordability?
| Model and capital structure | Rent per m² / month (TZS) | Rent for 18 m² shop (TZS) | Multiple of median affordability (8,300/m²) |
|---|---|---|---|
| A. Yield-on-cost, commercial (2.2m/gross m², 16%) | 58,200 | 1,048,000 | 7.0x |
| B. Capital recovery, commercial (1.6m/lettable m², 16%, 15 yrs) | 28,100 | 506,000 | 3.4x |
| C. Lender-style, commercial (2.0m/lettable m², DSCR 1.3) | 44,600 | 803,000 | 5.4x |
| A. with 50% capital contribution and 12% hurdle | 21,800 | 393,000 | 2.6x |
| C. with 50% grant + cheaper debt (10%, 20 yrs) | 19,300 | 347,000 | 2.3x |
| C. with fully public capital (rent covers O&M and reserve only) | 8,800 | 159,000 | 1.1x |
Table 9. Triangulation of the viability gap. Model A's cost is per gross m² (≈ TZS 3.1m per lettable m²), so it sits at the high-cost end; Models B and C use cost per lettable m².
The conclusions agree. On commercial terms, required rent is 3.4–7 times the median stated affordability. A 50% capital grant with cheaper, longer debt still leaves rent 2.3–2.6 times the median. Only when public capital funds the building does rent fall into the range traders say they can pay (about TZS 159,000 for an 18 m² shop, against TZS 100,000–200,000). The "3x versus 7x" difference between drafts reflects model design, not the data.
06 — Cross-Subsidy & Stand FeesWhy Anchor Tenants and Gate Fees Cannot Close a Gap This Size
6.1 The limits of cross-subsidy
A common response is to cross-subsidise: let anchor tenants (banks, pharmacies, supermarkets, restaurants) pay premium rents so small traders can pay less. The case data show why this has limited reach. The highest stated rents, from large-format tenants, are about TZS 14,000–17,000 per m², not the TZS 58,000 needed. If a social share of space is let at TZS 11,000 per m², the commercial rent required to average out at the Model A rent is shown below.
Anchor Rent Needed to Cross-Subsidise Social Units
| Share of space at social rent | Rent anchors would need (TZS/m²/month) | Multiple of highest stated rent |
|---|---|---|
| 50% | 105,000 | 6.1x |
| 70% | 168,000 | 9.8x |
| 80% | 247,000 | 14.4x |
Table 10. Anchor rent required to cross-subsidise social units to the base-case bankable average.
Cross-subsidy and mixed use raise revenue, but they cannot close a gap of this size. Competing private buildings in the same corridor also cap the premium anchors will pay. Half of prospective tenants (8 of 16) can pay more than TZS 200,000, so a protected tier funded by a stronger tier is real, but it is a partial answer.
6.2 Stand economics
The transport officer and regulator indicated 300–600 public vehicles a day and a possible stand fee of TZS 1,000–2,000 per trip, with a proposal for a toll gate. Gross fees would be roughly TZS 110 million (300 vehicles, TZS 1,000) to TZS 438 million (600 vehicles, TZS 2,000) a year. After 15% collection leakage and 30% operating cost, net income is about TZS 65–261 million, which under commercial terms (70% debt at 17% over 12 years, 1.3x cover) supports only about TZS 0.4–1.4 billion of total investment. The terminals actually built cost TZS 16–55 billion.
Stand-fee calculator
Defaults are the low case from the report (300 vehicles, TZS 1,000). Set 600 vehicles and TZS 2,000 for the high case. Investment supported uses 70% debt at 17% over 12 years with 1.3x debt-service cover.
What Stand Fees Can Fund vs What Terminals Cost
Debt Service on a Terminal Component vs Gross Stand Fees
A stand fee alone rarely carries the capital cost. Stands rely on cross-subsidy from commercial space, parking and advertising, and survey commuters accept only small charges (Tab 7).
07 — Pass-Through to CitizensFrom the Trader to the Citizen: Who Really Pays
If a facility is made bankable by cost-reflective rents and fees, the cost does not stay with the trader or operator. It travels along the chain below, and the final burden falls on the citizen as consumer, commuter or household member.
| Channel | Who is charged first | How it reaches the end user | Constraint on pass-through |
|---|---|---|---|
| Shop and stall rent | Trader | Higher prices, smaller product range, or a shift to higher-margin goods | Competing complexes; customers' low spend; informal street trade outside the building |
| Service charges, deposits, advance rent | Trader | Same as above; also reduces working capital and stock | Tenants' capacity to pay 1–6 months in advance |
| Stand and toll fees | Daladala / bajaji operators | Higher fares or lower service quality | Fares and stand fees are regulated; operators absorb first |
| Parking, toilets, entry fees | Customer / commuter directly | Direct out-of-pocket cost | Stated willingness to pay is low and mostly daily-use |
Table 11. Channels of cost pass-through.
7.1 Trader level: how much would prices have to rise?
To keep income unchanged, a trader who absorbs a rent increase through prices must raise sales by the same amount. To a first approximation (ignoring any fall in volume), the required price rise equals the rent increase divided by monthly sales. The baseline is the existing traders' median current rent of TZS 75,000; sales for the lowest band are taken at its mid-point, TZS 250,000.
Price Rise Needed to Keep Income Unchanged
The Pass-Through Curve
| Monthly sales (TZS) | Rent → 150,000 | Rent → 300,000 | Rent → 520,000 | Rent → 1,050,000* |
|---|---|---|---|---|
| ≈ 250,000 (under-500k band) | +30% | +90% | +178% | +390% |
| ≈ 750,000 (500k–1m band) | +10% | +30% | +59% | +130% |
| ≈ 2,000,000 (1m–3m band) | +4% | +11% | +22% | +49% |
Table 12. Price rise required to offset a rent increase from TZS 75,000. TZS 520,000 is the simple break-even rent for an 18.5 m² shop at TZS 1.6m per m² (Model B); *TZS 1.05m is Model A's rent for an 18 m² shop. At the top of the lowest sales band (TZS 500,000) the first three columns fall to +15%, +45% and +89%.
Pass-through calculator
Enter a trader's current rent, a proposed new rent and monthly sales to see the price rise needed to keep income unchanged.
Pass-through is not realistic for most existing traders. At 76% reporting sales below TZS 500,000, even doubling rent to TZS 150,000 requires a 30% price rise (15% at the top of the band) in a market where 42% of customers spend under TZS 5,000 a visit and street trade is cheaper. At a bankable rent the arithmetic breaks entirely: rent exceeds monthly sales. The consequence is not that prices rise by hundreds of percent; it is that those traders leave, sub-let or stay informal. The facility is then let to whoever can afford it: larger-format retailers, banks, pharmacies and electronics outlets. This composition effect keeps the building full and the project "bankable" but changes whom it serves. Traders' own worries ("loss of customers" 29 of 37, "rent increase after return", "no guarantee of space") are early signals of it.
7.2 High-volume operators: feasible, but it reaches the consumer
For businesses with many customers, a rent increase is spread thinly. A restaurant with 200 customers a day (surveyed range 150–300) and average spend of about TZS 4,500 adds roughly TZS 50 per customer if rent rises by TZS 300,000, and TZS 133 if it rises by TZS 800,000: between 1% and 3% of the bill.
Restaurant: Extra Cost per Customer
What Customers Accept to Pay for Parking
Pass-through is easiest for the tenants who could already pay, and reaches consumers invisibly in small increments. A bankable complex therefore tends to be sustained by high-traffic, higher-income-serving uses while the small-ticket trade that justified the public investment is crowded out.
7.3 Direct charges on commuters and shoppers
Stand, parking and toilet fees fall on users directly. Nine of 12 customers visit daily, so a TZS 1,000 charge is about TZS 26,000–30,000 a month for a daily user, and for the five of 12 who spend under TZS 5,000 a visit, a TZS 1,000 fee is at least a fifth of spending. A stand fee of TZS 1,000–2,000 per vehicle equals about TZS 35–135 per passenger per trip if a vehicle carries 15–30 passengers (assumption). For a daily commuter making two trips on 26 days a month, that is roughly TZS 1,800–7,000 a month, charged mostly to low-income users. Because fares and stand fees are regulated, operators absorb the fee first and pass it on through lower frequency or quality.
Stand Fee per Passenger per Trip
1. Absorption. Traders keep prices flat and eat the rent. Margins collapse, arrears rise and vacancies appear (as in the Magufuli passenger building). The gap returns to the project as revenue risk.
2. Substitution. Rents are paid by higher-turnover tenants; low-income traders move to the periphery or the street. The gap is closed, but the public purpose is diluted.
3. Pass-through. Prices, fares and fees rise. Burdens land on low-income households, with daily users bearing the most, constrained by competition and fare regulation.
Bankability does not remove the subsidy; it moves it from the budget to the wananchi. The question for each project is who should carry the cost: taxpayers through a visible viability-gap grant, or small traders and commuters through prices and fees. The second route is regressive and is rarely shown in feasibility studies. Every pre-feasibility report should therefore include a pass-through table like Table 12.
08 — Closing the GapOptions and Their Numerical Effect: Who Should Pay, and How
Because the gap is conserved, the policy question is not how to eliminate it but who should pay it, and how. There is no single fix; the options below can be combined.
| Option | How it works | Advantage | Risk / trade-off |
|---|---|---|---|
| A. Full user-pay | Rents and fees set at break-even | Bankable on paper; no public subsidy | Exclusion, vacancy, bypass; political resistance |
| B. Tiered rents with cross-subsidy | Anchors and parking pay premium rates; protected tier for sitting small traders | Keeps traders in; uses stronger tenants' ability to pay (8 of 16 prospects can pay above 200k) | Needs enough high-paying demand; ceiling shown in Table 10 |
| C. Explicit VGF for public-good components | Government or partner funds drainage, toilets, police post, access roads as grant or availability payment | Separates social cost from commercial cash flow; fits the 2023 PPP framework | Needs a funded LGA window; contingent-liability discipline |
| D. Unbundle the project | Council builds the public-good core; commercial space let on lease or BOT | Lowers capital the operator must recover; clearer risk allocation | Interface risk; procurement complexity |
| E. Operator lease with minimum guaranteed rent | Council leases the complex for base rent plus revenue share, with rent bands and service standards | Transfers collection and maintenance risk; predictable council income | Base rent still limited by ability to pay; needs tariff governance |
| F. Demand-side design | Site on commuter flows, feeder links, separate gates, pre-lets and passenger surveys | Raises volume rather than price | Bypass persists if enforcement is weak (Tab 3) |
| G. Lower capital cost | Standardised, modest designs; phased construction | Directly lowers break-even rent | May compromise quality and amenities |
Table 14. Policy options.
8.1 What combined support does to required rent
Required Rent Falls Only With Large Public or Concessional Capital
| Capital contribution (grant / land / concessional) | Required yield | Required rent (TZS/m²/month) | × median stated | × upper quartile |
|---|---|---|---|---|
| 0% | 16% | 58,200 | 7.0x | 3.5x |
| 30% | 16% | 40,700 | 4.9x | 2.4x |
| 40% | 14% | 30,600 | 3.7x | 1.8x |
| 50% | 12% | 21,800 | 2.6x | 1.3x |
| 60% | 12% | 17,500 | 2.1x | 1.05x |
| 70% | 12% | 13,100 | 1.6x | 0.8x |
Table 15. Effect of public/concessional capital contribution and a lower required yield on required rent (Model A, base cost TZS 2.2m per gross m²).
Affordability is reached only at roughly 60–70% public or concessional contribution with a ~12% hurdle, consistent with Model C's finding that a 50% grant plus cheaper debt still leaves rent about 2.3 times the median. At that level the project is, economically, publicly funded infrastructure with private operation, and the relevant question becomes value for money rather than bankability.
8.2 Design principles
1 · Separate the social and commercial layers from the start
Anchor shops, parking and advertising can carry debt; toilets, security, wastewater and small-trader units need grants or availability payments.
2 · Use EIRR to justify support
Quantify avoided accidents, time, health and jobs and use them as the basis for public contribution, rather than hiding the gap behind optimistic revenue forecasts.
3 · Protect sitting traders contractually
Priority of return, a temporary relocation plan and a cap on rent increases during a transition period; allocate subsidised space by published rule to avoid capture and resale.
4 · Build for voluntary demand, not compulsion
Location, services and user cost should attract users; compulsion should be a last resort, not the revenue base. Commission independent demand checks and use verified, lower-bound revenue in the financial model.
5 · Write tariff governance into the contract
A council- and regulator-approved tariff band, indexation rules, an independent review period and compensation rules.
8.3 Raising revenue density without burdening small traders
- Unit mix. More small units (higher rent per m²), fewer large low-yield bays; upper floors for offices or services.
- Ancillary income. Advertising, private-vehicle parking, logistics and loading space, events, telecom masts, priced for those who can pay.
- Collection efficiency. Electronic collection (the council accountant cited 80–95% collection) raises net revenue without raising tariffs.
- Advance rent as working capital. 7 of 16 prospective tenants could pay one month in advance, 7 could pay three months and 2 six months.
- Mixed income and a ring-fenced maintenance fund from year one, with a tariff review rule.
The Msamvu experience also points to a financing route: a council–pension-fund joint company combines patient long-horizon capital with a commercially managed asset, an arrangement worth testing for market and stand projects with a credible commercial layer.
Stated Rent per m² by Unit Size
09 — Screen & RecommendationsA Nine-Point Affordability–Bankability Screen for LGA Markets and Stands
TICGL/TERI proposes the following screen for LGA market and stand concepts at pre-feasibility stage. Thresholds are proposals for discussion and should be calibrated on a larger set of projects. Projects should also be judged on economic return (EIRR) alongside financial return (Tab 2, Section 2.7).
| Test | Question | Indicative threshold |
|---|---|---|
| 1. Affordability | Share of target tenants whose stated maximum rent is at or above the proposed rent | At least 70% for the commercial layer; required rent ≤ upper-quartile stated affordability |
| 2. Rent-to-sales | Proposed rent as share of verified monthly sales | Around 15% or less for small traders |
| 3. Capital-recovery gap | Share of annualised capital cost recovered from rent at observed ability to pay | Report the gap explicitly; fund through a named VGF or grant source |
| 4. Public-good cost | Cost of non-revenue components as share of total capital | Separate budget line; not loaded into rent |
| 5. Bypass and demand risk | Alternatives available to operators and traders; enforcement capacity | Pre-lets, letters of intent and passenger surveys; mitigation plan for terminals |
| 6. Pass-through | Price or fare increase needed to cover proposed rent and fees, by sales band and per passenger | Pass-through table in every pre-feasibility report; flag if above 10% for small traders |
| 7. Displacement | Plan for relocation, rent protection, priority of return | Documented, with allocation rule published before approval |
| 8. Sustainability | Maintenance sinking fund and tariff-review rule | Funded from year one |
| 9. Fiscal exposure | Contingent liabilities and VGF commitment as share of LGA OSR | Disclosed in the concept note and within PPP Centre limits |
Table 16. Proposed nine-point affordability–bankability screen.
Quick self-check: how many tests does your project pass?
Tick each test your concept already passes with documented evidence. This is a conversation starter, not a substitute for a full feasibility study.
9.1 Recommendations by stakeholder
| Stakeholder | Recommendation |
|---|---|
| PPP Centre and Ministry of Finance | Issue guidance for markets and stands that separates social and commercial layers; set clear VGF eligibility rules for LGA projects; require EIRR with FIRR and a pass-through table at pre-feasibility. |
| PO-RALG and councils | Stop using "self-financing" as the only decision criterion; classify projects by FIRR/EIRR; budget capital and maintenance for non-revenue components; create ring-fenced maintenance funds. |
| Lenders and development finance institutions | Offer long-tenor, concessional funding for the commercial layer; accept blended revenue; use guarantees to lower the cost of capital. |
| Advisers and appraisers | Test ability to pay, not just support; present sensitivity and downside cases; show the bankability gap and options to close it. |
| Traders and communities | Take part early in unit design, rent bands and relocation sequencing; use associations to negotiate priority of return. |
Table 17. Recommendations by stakeholder.
Separate the public-good core from the commercial layer; fund the core through a ring-fenced VGF window or grant; let the commercial layer carry tiered rents with a protected tier for sitting traders; and consider public build with private operation where tests 1–3 cannot be met even after realistic public contribution.
TICGL/TERI, The Bankability–Affordability Dilemma, October 20269.2 Limitations and further research
- Small, non-probability case evidence: results indicate the mechanism, not the outcome for every LGA project.
- Stated, not revealed, willingness to pay: brackets converted to mid-points and unit sizes assumed; sales probably under-reported (making affordability look worse) while stated rents may be optimistic.
- Simplified financial models: scenario costs, not a quantity-surveyor estimate; no taxes, land value, construction-period interest, inflation, indexation or phasing; one cost of capital.
- Pass-through assumes unchanged volumes and full recovery through prices and 15–30 passengers per vehicle.
- Mixed sources: news reports and an undated programme presentation; legal details such as concession limits for LGA PPPs should be verified against current regulations.
- Compare completed Tanzanian market and stand projects: actual revenue versus feasibility-study forecasts (optimism bias).
- Quantify the economic value of the social layer (safety, sanitation, avoided accidents) using recognised SCBA methods.
- Design availability-payment and VGF structures suited to LGAs with small own-source revenue.
- Track displacement after modern markets open: rents, sales and share of original traders retained.
- Analyse contingent liabilities that long-term PPP contracts create for LGAs.
10 — TICGL & Your ProjectFrom Bankability Findings to Feasibility, Market-Entry and Market-Research Decisions
TICGL (Tanzania Investment and Consultant Group Ltd), with its research arm TERI, builds the tests in this report into the studies it delivers: feasibility studies built to withstand a lender, a co-investor or a PPP Centre review, market-entry and setup support, and market research with a named data architecture. The bankability–affordability question is where all three meet, because a market or terminal only works if the financial model, the demand evidence and the entry structure agree. TICGL has delivered feasibility studies for PPP projects including the Boko Dawasa bus terminal and the Msasani Fish Market redevelopment.
10.1 From finding to study test to deliverable
| Finding in this report | What a TICGL study tests | Deliverable |
|---|---|---|
| Required rent is 3.4–7x stated affordability | Financial model triangulated across yield-on-cost, capital-recovery and lender-style (DSCR) methods; required-rent frontier by cost, yield, efficiency, opex and vacancy; NPV, FIRR and EIRR | Bankability frontier table, break-even rent, viability-gap size in TZS |
| Stated affordability is not revealed affordability | Structured willingness-to-pay survey with sampling frame, verified sales, haircut on stated rents, comparison with nearby informal and private rents | Verified affordability distribution and a lower-bound revenue case |
| Cross-subsidy has a ceiling | Tenant-mix and unit-mix optimisation, anchor demand testing, pre-lets and letters of intent, competing-corridor rent comparables | Tiered rent schedule with a protected tier and tested anchor premium |
| Pass-through lands on citizens | Pass-through table by sales band and per passenger; fare and stand-fee regulation check; elasticity assumptions stated | Pass-through table and a displacement-risk statement |
| Terminals fail through bypass | Vehicle and passenger counts, bus-owner and operator interviews, alternative-stop mapping, enforcement capacity, regulator position | Demand and bypass-risk assessment with mitigation plan |
| The public-good core is not bankable | Unbundling of social and commercial capex; EIRR/SCBA of avoided accidents, health, time and jobs; VGF or availability-payment sizing; contingent-liability disclosure against OSR | VGF case and PPP Centre-ready submission |
| Sitting traders can be displaced | Relocation sequencing, priority-of-return rules, rent-cap during transition, published allocation rule | Displacement and allocation plan |
Source: TICGL/TERI analysis; service lines as described on ticgl.com/what-we-do/.
Feasibility studies
- Market and demand, technical and operational, financial and economic, legal and regulatory, environmental and social, and risk assessment.
- DSCR, NPV and IRR models with downside cases.
- PPP structuring: VGF, availability payments, BOT or lease.
Market entry & setup
- Entry-mode strategy: council joint company, BOT, lease or operator concession.
- Company registration, TIC incentive certificates, sector and local approvals.
- Partner and site set-up.
Market research
- Primary and secondary data, competitive landscape and demand analysis.
- Willingness-to-pay and ability-to-pay surveys.
- ARIMA forecasting, regression and trade-data analysis.
Due diligence
- Financial, legal, commercial, operational and counterparty checks before a joint venture or new partner.
- ESG and land checks.
For LGAs and PPP authorities
Pre-feasibility with the nine-point screen, a pass-through table and a named source for the viability gap, so a concept note can stand up to PPP Centre review.
For developers and operators
Required-rent frontiers, tenant-mix design and pre-let strategy before committing capital, plus entry-mode and approvals support.
For lenders and DFIs
Independent demand checks, lower-bound revenue cases and DSCR stress tests, so blended and concessional structures are sized on evidence.
Talk to TICGL/TERI
Request a feasibility, market-entry or market-research scoping call, or a briefing on this report's findings for your market or terminal project.
What We Do →TICGL Markets →TERI Research →Book a consultation →✉️ economist@ticgl.com11 - Full ReportRequest the Full Report
Want the complete research report?
The full TICGL/TERI report includes the complete survey detail, all assumptions, the three financial models and the reconciliation notes. Email us and we will send it, or discuss how to apply the nine-point screen to your own LGA project.
✉️ economist@ticgl.comAbout the author
Amran Bhuzohera authored this research for the TICGL/TERI Economic Research Division.
Read the author profile. Contact: economist@ticgl.com · Mbezi Beach Africana, Dar es Salaam.
12 - Sources & AssumptionsMethod, Assumptions and Reconciliation Notes
Source report. TICGL/TERI, The Bankability–Affordability Dilemma: Why markets and bus terminals struggle to be both "bankable" and publicly useful in Tanzania's Local Government Authorities, and who ends up paying the difference, comprehensive research report, October 2026, consolidating three TICGL/TERI working drafts. Case data. Anonymised case survey; full detail in the full report on request (economist@ticgl.com). Calculations. Required rent follows the Model A formula in Tab 2; figures on this page marked "TICGL/TERI calc." (cumulative shares, sensitivity bars, anchor-rent curve, pass-through curve, LGA OSR shares) are arithmetic on the report's inputs. Limitations. See Tab 9, Section 9.2.
Key assumptions (Annex A)
| Item | Assumption | Note |
|---|---|---|
| Model A: construction cost | TZS 2.2m per gross m² (range 1.2–2.8m) | Illustrative; replace with QS estimate |
| Model A: efficiency, opex, vacancy, yield | 70%; 20% of gross rent; 10%; 16% base (12%, 14%, 18% tested) | Yield compares with ≈10.8% 15-year bond (Oct 2026) |
| Model B | TZS 0.8–3.0m per lettable m²; 16%; 15 years; 85% occupancy | Capital recovery only; ignores opex (a floor) |
| Model C | TZS 1.5/2.0/2.5m per lettable m²; 70% debt at 17%, 12 years; DSCR 1.3; opex TZS 60,000/m²/yr; reserve 1.5% of cost/yr; occupancy 85% | Concessional: 50% grant, debt at 10% over 20 years; fully public: rent covers opex and reserve |
| Rent brackets → mid-points (TZS 000) | <100: 75; 100–200: 150; 200–300: 250; 300–500: 400; >500: 600 | Stated maximum monthly rent |
| Current rent mid-points (TZS 000) | <50: 35; 50–100: 75; 100–200: 150 | Existing traders; three pay none |
| Unit sizes (m²) | Kiosk 6; small 9; medium 18; large 35; restaurant 40 | From size categories in survey |
| Sales mid-points (TZS 000) | <500: 250; 500–1,000: 750; 1,000–3,000: 2,000 | Pass-through analysis; top of lowest band (500) shown as sensitivity |
| Pass-through formula | Price rise = rent increase ÷ monthly sales | First-order; ignores volume response |
| Restaurant example | 200 customers/day, 30 days, TZS 4,500 spend | Mid-range of survey brackets |
| Stand | 300–600 vehicles/day; TZS 1,000–2,000; 365 days; 85% collection; 30% costs | Officers' and regulator's estimates |
| Passengers per vehicle | 15–30 | Assumption |
Reconciliation notes and editorial checks
| Issue | Treatment in this report |
|---|---|
| Size of the affordability gap (≈3x vs ≈7x) | Differences reflect model design, not data. Model B has no operating costs and uses TZS 1.6m per lettable m²; Model A uses TZS 2.2m per gross m² (≈3.1m per lettable m²) with opex and vacancy. All three are shown (Table 9). |
| Sales mid-point for the lowest band | Earlier draft used TZS 400,000; the band mid-point of TZS 250,000 is now used, with the top of the band as a sensitivity. |
| Magufuli municipal revenue loss | Press figures differ: more than TZS 50 million a month (older reports) and more than TZS 600 million a month (later government talks). Both are cited; they are not reconciled. |
| Magufuli cost | Reports range from about TZS 50 billion to 54.6 billion; the report uses TZS 54.6 billion (TZS 50.9 billion central grant). |
| Kejetia financing | A phase-two loan figure could not be verified and is omitted. |
| World Bank municipal PPP programme | Source is an undated presentation; used as supporting diagnosis only. |
| Passengers per vehicle | 15–30 used instead of 25. |
| Share of cost rent can support (this page) | The report text quotes 19–38% at a 12% yield; on this page 17,100 ÷ 43,700 gives about 39% at the top of the stated range, so the web text says 19–39%. |
| Dodoma terminal name | The report names Dodoma's terminal both Nzuguni and Nane Nane; this page says "Dodoma terminal" (TZS 24 billion or more) pending confirmation of the name. |
- TICGL/TERI, The Bankability–Affordability Dilemma, October 2026, and underlying working drafts; case survey data (anonymised).
- Ministry of Finance, Budget Frame 2024/25; TanzaniaInvest, "Budget proposals 2026/27"; Clyde & Co, "Tanzania's National Budget Speech 2026"; UNDP Tanzania INFF workshop for local governments, 10 June 2025.
- Clyde & Co, PPP Act amendments (July 2023) and PPP Regulations Amendments (2023); ICE/ITA, "Tanzania scales up PPP projects", 16 September 2025; TanzaniaInvest, "Tanzania Signs Nine PPP Contracts for DART, Ports and TAZARA Railway" (June 2026).
- World Bank PPP Support Programme, "Tanzania's Municipal PPPs" (undated); World Bank results frameworks P159489 and P148974; TanzaniaInvest, "Tanzania Financial Markets Week to 2 October 2026".
- Daily News, The Citizen, The Guardian (IPP Media) reporting on the Magufuli, Nane Nane/Dodoma and Msamvu terminals (2018–September 2026); Daily Monitor (Uganda) on Kampala markets; Ghanaian Times, MyJoyOnline, Graphic Online on Kejetia; University of Bergen study on Takoradi.
13 - Quick AnswersFrequently Asked Questions
What is the bankability-affordability dilemma in Tanzania's Local Government Authorities?
To attract private capital, an LGA project (illustrated here by markets and bus terminals) must produce revenue that repays lenders and rewards investors. But its users are often low-income people who cannot pay cost-reflective charges. Public value, affordability and bankability can be achieved two at a time; the third breaks without extra public support.
How much higher is the bankable rent than what traders can afford?
In the TICGL/TERI illustrative case, the rent needed to repay the building on commercial terms is 3.4 to 7 times the median stated affordability of about TZS 8,300 per lettable square metre per month. The base case needs about TZS 58,200 per square metre per month; the highest stated rent in the sample was TZS 17,100.
Can cross-subsidy from anchor tenants close the gap?
Only partly. If 50 to 80 percent of space is let at an affordable social rent of TZS 11,000 per square metre, anchor tenants would have to pay TZS 105,000 to 247,000 per square metre per month, 6.1 to 14.4 times the highest rent any respondent said they could pay.
Who ends up paying the viability gap?
The gap is conserved: it is paid by traders (higher rents, exits, arrears), consumers and commuters (higher prices and fees), taxpayers or the LGA (viability gap funding, grants, guarantees) or investors and lenders (lower returns or no financial close). Making a project bankable through user charges moves the cost from the budget to the citizen.
Can small traders pass a rent increase on to customers?
Not realistically. With 76 percent of existing traders reporting sales under TZS 500,000 a month, raising rent from TZS 75,000 to TZS 300,000 would need a 90 percent price rise at TZS 250,000 of sales. Traders instead exit, sub-let or go informal, and the space goes to higher-turnover tenants.
What does the Magufuli bus terminal show about fee-based terminals?
The TZS 54.6 billion Magufuli terminal at Mbezi Luis was about 93 percent funded by a central-government grant, yet buses bypass it, the municipality loses revenue, and in September 2026 the Ubungo council was pursuing over TZS 249 million of tenant rent arrears. Morogoro's Msamvu terminal, built with a pension-fund partner on a main corridor, performs better but its reported takings equal only about 7 percent of its phase-one cost.
How much public capital makes rents affordable?
In the base case, required rent falls to the range traders say they can pay only at about 60 to 70 percent public or concessional capital contribution with a hurdle rate near 12 percent. At that point the project is economically publicly funded infrastructure with private operation, and the question becomes value for money.
What should an LGA test before taking a project to PPP?
TICGL/TERI proposes a nine-point screen: affordability, rent-to-sales, capital-recovery gap, public-good cost, bypass and demand risk, pass-through, displacement, sustainability and fiscal exposure, with a pass-through table and both financial and economic returns shown at pre-feasibility stage.
Muhtasari kwa Kiswahili
Mradi uwe wa kibenki (bankable) na bei iwe nafuu kwa wananchi: Je, vyote viwili vinawezekana? Halmashauri (LGAs) zinahimizwa kupanga na kuendesha miradi inayozalisha mapato, kama masoko, vibanda vya biashara na vituo vya mabasi, ili kuongeza mapato ya ndani na kupunguza utegemezi kwa Serikali Kuu. Ili kuvutia mitaji binafsi, miradi hiyo lazima ilipe madeni na kutoa faida kwa wawekezaji. Lakini watumiaji wake, kama wafanyabiashara wadogo, madereva wa daladala na wanunuzi wa kila siku, wengi wana kipato kidogo. Hali hii inahusu miradi mingi ya halmashauri (nyumba, taka, utalii, usafiri wa umma, masoko na vituo vya mabasi). Utafiti huu wa TICGL/TERI (Oktoba 2026), ulioandikwa na Amran Bhuzohera, unapima kama mahitaji hayo mawili yanaweza kutimizwa kwa pamoja, kwa kutumia masoko na vituo vya mabasi kama mifano (case studies), pamoja na ushahidi wa Tanzania na kanda.
Pengo ni kubwa: Kodi inayohitajika kulipa jengo kwa masharti ya kibiashara ni mara 3.4 hadi 7 ya kiasi ambacho mfanyabiashara wa kati anasema anaweza kulipa (kodi ya kati ya TZS 8,300 kwa mita ya mraba kwa mwezi, dhidi ya TZS 58,200 inayohitajika katika hali ya msingi). Kwa kodi wanazoweza kulipa, mapato ni asilimia 2.3 hadi 4.6 tu ya gharama ya ujenzi, wakati hatifungani ya miaka 15 inalipa takriban asilimia 10.8.
Ruzuku ya sehemu haitoshi: Hata ruzuku ya asilimia 50 ya mtaji pamoja na mkopo nafuu wa muda mrefu bado inaacha kodi inayohitajika mara 1.3 hadi 2.6 ya uwezo wa wafanyabiashara. Kodi huwa nafuu tu pale mtaji wa umma au wa masharti nafuu unapofikia asilimia 60 hadi 70.
Ruzuku ya ndani kwa ndani ina kikomo: Ikiwa asilimia 50 hadi 80 ya nafasi itapangishwa kwa kodi nafuu, wapangaji wakubwa (benki, maduka makubwa, maduka ya dawa) wangelazimika kulipa TZS 105,000 hadi 247,000 kwa mita ya mraba kwa mwezi, jambo lisilowezekana sokoni.
Pengo halipotei, linahamia: Anayelimaliza pengo ndiye anayebeba mzigo: wafanyabiashara, wateja na abiria, walipakodi, au wawekezaji. Mradi ukifanywa wa kibenki kwa ada za watumiaji, gharama inahamia kwa mwananchi. Kwa mfanyabiashara mdogo, kodi ikipanda kutoka TZS 75,000 hadi 300,000 bei zingepaswa kupanda asilimia 90 (mauzo ya TZS 250,000); kwa hiyo wengi huondoka, hupangisha kwa wengine au hurudi mitaani, na nafasi huchukuliwa na wenye mauzo makubwa. Abiria hulipa TZS 35 hadi 135 kwa safari kwa kila ada ya kituo ya TZS 1,000 hadi 2,000.
Mifano ya Tanzania: Kituo cha Magufuli (TZS bilioni 54.6, takriban asilimia 93 ni ruzuku ya Serikali Kuu) kinakabiliwa na mabasi kukwepa, hasara ya mapato ya manispaa na malimbikizo ya kodi ya zaidi ya TZS milioni 249 (Septemba 2026). Msamvu (Morogoro), kilichojengwa na LAPF kwenye barabara kuu, kinafanya vizuri zaidi lakini mapato yake ya mwaka ni takriban asilimia 7 tu ya gharama ya awamu ya kwanza. Kuna miradi 410 ya PPP katika halmashauri 184, lakini mikataba 9 pekee ndiyo imesainiwa kitaifa hadi Juni 2026.
Mapendekezo: (1) Tenganisha kiini cha huduma za umma (vyoo, mifereji, usalama) na sehemu ya kibiashara; (2) fadhili kiini hicho kwa Ufadhili wa Pengo la Uwezekano (VGF), ruzuku au mtaji nafuu; (3) tumia kodi za ngazi tofauti na linda wafanyabiashara waliopo kwa haki ya kurudi na ukomo wa ongezeko la kodi; (4) jenga kwa mahitaji halisi, si kwa kulazimisha; (5) andika utawala wa ushuru kwenye mkataba; (6) tumia kipimo cha vipengele tisa (uwezo wa kulipa, kodi dhidi ya mauzo, pengo la kurejesha mtaji, gharama za huduma za umma, hatari ya kukwepa, mwitikio wa bei, uhamishaji, uendelevu na mzigo wa kifedha) wakati wa upembuzi wa awali, pamoja na jedwali la mwitikio wa bei. TICGL/TERI hutoa upembuzi yakinifu (feasibility), ushauri wa kuingia sokoni na utafiti wa soko.
Hitimisho: Uwezo wa kibenki hauondoi ruzuku; unaihamisha kutoka bajeti kwenda kwa wananchi. Swali kwa kila mradi wa halmashauri ni nani abebe pengo: walipakodi kupitia ruzuku inayoonekana wazi, au wafanyabiashara wadogo na abiria kupitia bei na ada.
- Kodi inayohitajika: mara 3.4 hadi 7 ya uwezo wa kati · TZS 58,200 dhidi ya 8,300 kwa m² kwa mwezi
- Mapato kwa kodi nafuu: asilimia 2.3 hadi 4.6 ya gharama · hatifungani ya miaka 15: asilimia 10.8
- Mtaji wa umma unaohitajika kufikia unafuu: asilimia 60 hadi 70 · mtaji wa ruzuku ya asilimia 50 bado unaacha mara 2.3 hadi 2.6
Chanzo: TICGL/TERI, The Bankability–Affordability Dilemma (Oktoba 2026), mwandishi: Amran Bhuzohera. Kwa ripoti kamili: economist@ticgl.com.