Market Trends
Kenya is building affordable housing faster than at any point in its history. Whether it is building a sustainable housing system is a different question, and it is the one our new research publication sets out to answer.
The headline numbers are genuinely impressive. More than 1.31 million Kenyans have registered on the Boma Yangu platform. The construction pipeline runs into the hundreds of thousands of units across all 47 counties. The levy that funds the programme has raised over KES 200 billion since it began. Against that, roughly 8,807 units are recorded as complete and about 14,219 have been allocated.
Those two sets of figures sit uncomfortably together, and most public commentary resolves the discomfort by picking one and ignoring the other. Supporters quote the pipeline. Critics quote the completions. Both are accurate, and neither tells you very much on its own.
Our research takes a different approach. It asks nine questions in sequence, about scale, delivery, demand, allocation, price, finance, operations, lifecycle and what comes next, and follows the evidence wherever it leads. The short version of what we found is this. The Affordable Housing Programme's binding constraints have moved downstream. Construction capacity is no longer the limiting factor. Conversion, allocation, finance and operations are.
Registration is not the same as demand
The most misread number in the programme is the registration total. Registering on Boma Yangu signals interest. It does not establish that a household has saved consistently, qualified on eligibility, been assessed as able to pay, been allocated a unit, financed it, moved in, or taken title.
On the Affordable Housing Board's own disclosures, roughly four in five registered savers have never made a contribution. The path from interest to ownership passes through nine distinct states, and each is measured in a different unit. Registration counts people. Allocation counts units assigned to applicants. Completion counts physical buildings. Occupation counts households. Comparing a figure drawn from one stage with a figure drawn from another produces contradictions that are not really there, and it is the single most common analytical error in Kenyan housing commentary.
This matters commercially, not just editorially. Anyone sizing genuine demand, whether a developer deciding where to build, a lender pricing a portfolio or a supplier planning capacity, needs the number of contributing savers within reach of a specific unit's deposit in a specific location. The registration total is not that number. Treating it as a pipeline of ready buyers will produce a serious misjudgement of the market.
The more useful reading of the attrition between stages is that it is not in itself a failure. It is a measurement. The gap between 1.31 million registrations and 14,219 allocations tells you precisely where the programme's real performance problems sit, provided you are willing to look at the stages rather than only at the endpoints.
Why the published totals differ
Pipeline figures for the programme range from roughly 138,000 units to more than 280,000. It would be easy, and lazy, to present that spread as proof that the data cannot be trusted. The more accurate explanation is that different figures count different things.
In August 2026 the Principal Secretary for Housing and Urban Development stated that 8,807 units were complete and 198,726 under construction. Separately, and in the same period, 23,200 institutional and staff housing units and 37,347 student accommodation units were also reported under construction. Those three programmes together approach 259,000 units, which substantially accounts for the larger figures quoted at industry level. The Kenya National Bureau of Statistics, applying its own classification, recorded 205,311 public sector housing units under construction as at December 2025, of which 138,474 were specifically classified under the Affordable Housing Programme.
Once you know which programmes a given total includes, most of the apparent contradiction dissolves. Affordable housing, institutional and staff housing, and student accommodation are distinct undertakings that happen to be reported both separately and together.
What does not dissolve is the absence of any published reconciliation. No accessible source sets out which units sit in which total, on what date, and under which classification. The reader is left to reconstruct the basis unaided. That is the real disclosure gap, and closing it would cost almost nothing relative to what the programme spends on construction. For a programme funded by a mandatory levy on earned income, a published reconciliation would settle a great deal of unnecessary argument.
What has actually been built
Measured against its own history, the delivery record is strong. On the national statistical record, annual completions rose from 1,655 units in 2024 to 6,738 units in 2025, roughly a fourfold increase, with a further 410 completed by the National Housing Corporation. Cumulative completions of approximately 8,807 by August 2026 are consistent with that trajectory. For context, the predecessor national housing initiative delivered fewer than 3,000 units in total across two projects. The current programme has already passed that several times over.
Measured against need, the picture is different. Independent estimates place Kenya's accumulated housing deficit at around two million units, with annual demand near 250,000 against supply of roughly 50,000. At current output the programme is making a real but modest contribution to an enormous shortfall.
Delivery is also highly concentrated. Working from 130 documented projects in our dataset, the single largest scheme accounts for approximately 26% of all published typology units. The five largest account for close to half, and the ten largest for roughly two thirds. Exclude the flagship estate and the median tracked project falls below 500 units.
Concentration transmits risk. A delay, dispute or servicing failure at one large estate moves the national numbers materially. It also concentrates the operating challenge, because the first genuine test of estate management at scale will happen at a handful of sites rather than being spread thinly across the country.
One further point deserves stating plainly. National completion totals are published without a project-level breakdown. Working only from named schemes in the public record, we could independently trace 1,971 completed units, principally at Mukuru, King's Sapphire in Nakuru, Machakos Town and Homa Bay. That is a floor, not an estimate of total completions, and the distance between it and the national figure reflects the absence of project-level reporting rather than any suggestion that the national figure is wrong.
Completion is not occupation
A completed home and an occupied home are not the same thing, and the difference has direct financial consequences.
Reporting in 2026 described completed estates waiting for occupants while other sites ran behind on contractor payments, with the Board recording more than 15,000 applicants still to be matched to units. At the same time, more than 20 projects were reported to have passed 70% completion.
A finished but empty unit generates no receipts. It continues to accrue holding costs, security costs and maintenance obligations, and it is exposed to deterioration and vandalism. For an owner, a lender or an asset manager, the meaningful sequence therefore runs from completion to allocation, then occupation, then sustained payment, then title. Measuring the programme at the completion stage alone overstates its effective output, and measuring it at the allocation stage is not much better.
Who gets the homes, and how
A good deal of public commentary treats allocation, deposits and tenant purchase as undefined. They are not. The Affordable Housing Regulations, 2025 set out eligibility criteria, electronic application, Board verification of documents and assessment of ability to pay, allocation according to the applicant's income category, notification with reasons and return of deposits for unsuccessful applicants, deposit assistance for the lowest income band, provision for a one-time change of allocated unit, reallocation following default, restrictions on disposal, and continuing obligations after allocation including maintenance.
The deposit illustrates the point. The regulatory position is settled at up to 5% of the purchase price, reduced from a previous 10% requirement. What remains unsettled is the information environment around it. Superseded figures are still published online and still repeated by third-party sites, so a prospective buyer researching the programme may encounter a requirement materially higher than the current rule. The regulatory risk here is low. The consumer information risk is real, and buyers should verify terms against the regulations and confirm project-specific offer terms in writing before parting with any money.
The harder question is not whether the framework exists but how consistently it operates, and that is where the flagship Mukuru estate has become the first real test. In August 2026 residents whose homes had been demolished to make way for the development protested publicly, reporting that they had registered, paid deposits and received confirmation, only to be told later that units were unavailable. Similar complaints were raised at the earlier handover in May 2025.
The Board has rejected claims of irregularity, maintaining that beneficiaries passed through enumeration, registration, verification and allocation. It has also confirmed that it is verifying the complaints and will reconsider approximately 300 verified applicants alongside 164 previously enumerated households.
We draw no finding of wrongdoing, and the available evidence does not establish one. The more useful reading is that the allocation system is under pressure rather than that it has failed. Mukuru is being tested against exceptional conditions: demand vastly exceeding supply, and households holding a prior claim arising from displacement. The operative question for the programme is not whether one estate handled it perfectly. It is whether the verification, appeal and reallocation machinery can scale to hundreds of estates at once. No allocation register is published by project, so that question cannot currently be answered from outside.
What the homes cost, and what "price" actually means
The most striking commercial characteristic of the published pricing is that it is national rather than local. Almost 90% of priced typologies in our dataset sit on just eight fixed national price points. A studio in a small county town carries the same published price as a studio in a major urban centre.
Across the tiers, social units run from roughly KES 640,000 to KES 1.28 million, affordable units from KES 1.0 million to KES 4.62 million, and market units from KES 1.8 million to KES 10 million. The median published typology price across 784 priced observations is approximately KES 1.73 million.
The consequence of national pricing is that location differentiation is materially weaker than in conventional private-market housing. Location has not disappeared from the cost base, because it still affects land, servicing, unit mix and amenity provision. It is simply far less visible in the published price than a private developer's pricing would show. In practice this means units in high-value urban nodes look inexpensive relative to the open market, while units in lower-value towns look comparatively expensive, with obvious consequences for uptake in each case and for any eventual resale.
One distinction deserves emphasis, particularly from a firm whose work includes valuation. A published programme price is an administered figure. It is not a valuation and should not be read as evidence of open-market value. Where a published price sits materially below what a comparable private unit would command in the same location, that difference represents programme subsidy and allocation advantage, not a demonstrated market discount. Establishing the true relationship between published prices and assessed market values would require transaction evidence that is not yet publicly available.
What it costs to live there, not just to buy
The Affordable Housing Act defines affordable housing as housing that is adequate and costs no more than thirty percent of a person's monthly income. That is the programme's own statutory yardstick, and it is the right one to apply.
Reported tenant purchase instalments run from approximately KES 4,000 a month for a studio to approximately KES 23,000 for a three-bedroom unit, over terms of up to thirty years. Buyers occupy the unit while paying and receive sectional title on completion of payment. On its face this is a genuinely powerful mechanism, because it brings ownership within reach of households who would never qualify for a conventional mortgage.
But those figures describe the cost of acquiring a home. They do not describe the cost of living in it. A household's real recurring housing burden also includes the group life insurance premium provided for in the regulations, the service charge, water, electricity, waste collection, security, any sinking fund contribution, and the transport cost implied by where the unit is located. Tested against the statutory thirty percent threshold, it is the total of these that determines whether a home is genuinely affordable, not the instalment in isolation.
A household can meet the monthly instalment comfortably and still be unable to sustain occupation once everything else is added. Lenders underwriting tenant purchase exposure face exactly the same blind spot.
We have deliberately not published an illustrative total. No verified service charge schedule exists for any estate in the programme, and publishing a figure without a single verified component would create the appearance of evidence where none exists. The absence is itself the finding. Establishing a credible occupancy cost benchmark for affordable housing estates is, in our view, the most useful piece of work the sector could commission next.
Can it be financed?
The programme has two financing questions, and they are usually discussed separately. The first is whether the levy that funds construction is collected reliably. The second is whether households can finance acquisition. Both now point in the same direction.
On the first, the levy has raised more than KES 200 billion since inception, including approximately KES 73.2 billion in the 2024/25 financial year, which exceeded target. Against that, an estimated KES 100 billion has been either unremitted or evaded, concentrated among non-compliant employers and in the informal sector. The Finance Act, 2026 responded by empowering the Kenya Revenue Authority to recover unpaid levies as though they were unpaid tax, a significant strengthening of enforcement.
That development is instructive beyond its immediate effect. It demonstrates that collection performance improves when it is measured and given legal machinery. It also points to an uncomfortable symmetry: the programme has a remittance problem at the top of its financial architecture that closely resembles the service charge collection risk it will face at the bottom, once tens of thousands of households are contributing to the running costs of their estates. Both are remittance disciplines, not construction problems.
On the second question, Kenya's mortgage market remains strikingly shallow, with only around 26,000 active mortgage accounts nationally against a stated ambition of one million. The sector's own centre of gravity has visibly shifted toward demand-side finance, with industry discussion in 2026 focused on reaching informal-income, SACCO and micro-entrepreneur households historically excluded from mortgage markets. That is the same diagnosis our research reaches from the data. The programme can build faster than Kenyan households can currently finance.
There is also a longer-term financial idea taking shape. Government has discussed recycling receipts from completed and occupied homes to fund subsequent phases. That is sound portfolio logic: capital funds construction, construction produces units, units are allocated and occupied, occupation generates collections, collections fund operations and maintenance, and asset performance determines whether capital can be recycled into the next phase. The difficulty is that every stage after occupation depends on measurement the programme does not currently publish. Collections cannot be assessed without service charge and arrears data. Maintenance adequacy cannot be assessed without condition data. Capital recycling cannot be underwritten without both.
The question nobody is measuring
This brings us to the part of the programme that is least documented and, we would argue, most consequential.
No programme-level service charge schedule, collection record or arrears data is published anywhere. This is not regulatory silence, because the 2025 Regulations do impose post-allocation obligations including maintenance, and the Sectional Properties Act supplies the eventual ownership and management structure through a corporation of unit owners. The gap is evidential. A framework that exists on paper but produces no published operating data cannot be assessed for adequacy.
There is also an unresolved question about the interim. While tenant purchase buyers are still paying, they hold no title, so ownership has not yet transferred. The public record does not establish who bears management responsibility during that period, or how service charges are set, collected and accounted for between handover and final payment. For households who may spend up to three decades in that state, this is a material unanswered question, and it is not resolved simply by the eventual existence of a management corporation.
Field reporting has already given early signals. Occupation at several otherwise complete schemes has been slowed by water shortages, inadequate waste disposal arrangements and missing amenities. Residents at one commissioned estate praised the homes while noting the absence of a children's playground. These are not peripheral complaints. They are the operating indicators that determine whether a completed estate becomes a functioning neighbourhood or a deteriorating one. Water availability reflects utility coordination and cost recovery. Waste management reflects service contracting and budget adequacy. Maintenance response reflects whether asset records and contractor arrangements exist at all.
A completed building is not yet a functioning estate. The transition from construction to sustained operation requires records, structures, budgets and accountability that have to be established deliberately rather than assumed. Where that transition is unmanaged, estates deteriorate long before their design life expires, and the value created by construction is quietly eroded.
Our publication sets out a professional framework for this phase, organised around nine pillars: estate and asset information; handover and mobilisation; occupancy and estate administration; facilities and property management; maintenance and lifecycle management; compliance, risk and governance; financial and commercial management; performance and portfolio intelligence; and continuous improvement, which feeds operating experience back into asset records, maintenance planning and budgets. It is deliberately generic, applicable to any large residential estate rather than to a single scheme.
Sequence matters as much as content. Asset information captured at handover is inexpensive. Reconstructed years later, after documentation has dispersed and warranties have expired, it is costly and incomplete. The same applies to compliance records and defect close-out. Decisions taken, or deferred, in the first months of an estate's operating life determine its cost trajectory for decades. On the evidence we reviewed, the programme is at precisely that point for its first cohort of completed estates.
What this means, depending on where you sit
For government and the Board, the principal exposure is disclosure and allocation integrity rather than construction. A published reconciliation of programme figures and an auditable allocation register would strengthen confidence more efficiently than further delivery announcements. Establishing the post-handover operating framework now, while completed estates are still few, is far cheaper than retrofitting it later.
For developers and contractors, genuine opportunity exists where sites are urban and well serviced. Risk concentrates in payment timing, in thin project-level demand outside the flagship schemes, and in servicing and amenity obligations that delay occupation. Due diligence has to descend below the national headlines to the individual estate.
For lenders and investors, the binding constraint is finance rather than units. Tenant purchase exposure cannot be fully underwritten while occupancy costs and arrears performance remain unpublished. The shallow mortgage base and the sector's demand-side pivot indicate where the next phase of value will be created.
For buyers and households, the scheme genuinely brings ownership within reach, and the regulatory position is clearer than commentary suggests. The practical risks are superseded information still circulating publicly, unquantified occupancy costs, and allocation uncertainty where demand greatly exceeds supply.
For estate managers and property professionals, the emerging requirement is operational: handover discipline, asset information, compliance systems, service charge structuring and portfolio performance reporting for a rapidly growing national residential portfolio. Capability built now will be scarce and valuable within two years.
Why this matters now
The programme's construction achievement is real and, measured against its predecessors, substantial. Its vulnerabilities are administrative, financial and operational rather than physical: conversion, disclosure, allocation, collection, and the management of what has already been delivered.
These are solvable problems. Solving them is considerably less expensive than the alternative, which is a national portfolio of physically completed estates that decline for want of an operating model.
The question is no longer whether Kenya can build affordable homes. It is whether Kenyans can finance them, whether the right households receive them, and whether the estates created can be sustained once built. Building the home is the beginning of the asset lifecycle, not the end of the housing programme.
Read the full research
This article summarises the principal findings of our research publication, *Kenya's Affordable Housing Programme: From Housing Delivery to Long-Term Asset Performance*.
The full report examines all nine tests in detail, with the underlying data, charts, tables, source register and a confidence assessment setting out what the evidence supports and what remains unknown. It draws on our proprietary dataset of 130 documented projects across 41 counties, comprising 816 unit typologies and 784 priced observations, together with Kenyan statutory instruments, official statistics, government statements, court records and national reporting current to September 2026.
Every statistic in the report carries its sample size. Where credible sources differ, both values are retained and the reason for the difference is examined rather than averaged away. Where the evidence does not support a conclusion, we say so.
The complete publication is available to download here > https://jwrealty.ke/market-intelligence/publications
JW Realty & Consulting Ltd is an integrated property advisory firm based in Nairobi, providing Valuation, Agency, Property & Facilities management, Lease advisory, Technical survey and Research services across Kenya.