50 Dhaka: megacity, housing and mobility

Chapter 50 of 60 in the Bangladesh 2036 research base. Contents of the series.

The megacity's decade question has stopped being demographic and become institutional

Chapter 14 mapped the urban system, one third of the country counted urban with Dhaka primacy deepening every year, and left the city itself unopened. This chapter goes inside the metropolitan ring, the Dhaka, Gazipur and Narayanganj corridor the census actually measures, and defends a narrower claim: the megacity's binding constraints over the 2026 to 2036 window the chapter 15 scenarios price are institutional, not demographic, because the growth record has changed shape while the stock it sits on has not. The primacy series now shows a plateau, the Dhaka agglomeration's share of the urban population rising just 0.01 points in 2025 after 0.11 in 2024 and about half a point a year through the 2010s [WB WDI 2026], while the agglomeration itself still added 717,212 persons in 2025 to reach 24.65 million [WB WDI 2026]. The stock that plateau rests on is the chapter's finding: a renter city, with 70.46 percent of Dhaka district's 3,737,085 households renting their dwelling against 27.51 percent owning [BBS Census 2022, district tenancy tables]; a two-wheel motorisation record, with motorcycles at 69.1 percent of 2022's new Dhaka registrations [BRTA 2023]; a drainage stock of 82 cumecs of pumping across four stations against a service area above 360 square kilometre [DWASA 2020]; and a state that prints plots faster than occupation, 24,251 Purbachal plots handed over against about 300 approved building plans [RAJUK 2023, Purbachal progress report]. If the institutions that price, record and service this ring, RAJUK's planning perimeter, the land record, the transit accounts, the drainage capital stock, are rebuilt inside the decade the chapter 15 scenarios price, the plateau becomes the dispersal that scenario needs; if they drift, the same stock compounds at ring speed and the megacity's costs, congestion, rent burden and waterlogging, become the decade's output.

The record: primacy is plateauing on top of a housing and mobility stock that has not changed

The population record has two layers that now point in different directions. The agglomeration the World Bank series tracks grew from 14,730,537 persons in 2010 to 24,652,864 in 2025, an addition of about 728,000 persons a year since 2019, with 726,036 added in 2024 and 717,212 in 2025 [WB WDI 2026]. The share of the national urban population that this agglomeration holds rose from 34.91 percent in 2010 through 41.90 in 2022 to 42.22 in 2025, but the increments collapsed: plus 0.60 points in 2022, plus 0.19 in 2023, plus 0.11 in 2024 and plus 0.01 in 2025 [WB WDI 2026]. For the first time in the series, the marginal urban resident is about as likely to be counted outside the Dhaka agglomeration as inside it. Chapter 14's arithmetic, which assumes the 2019 to 2025 pace holds, projects the agglomeration above 34 million persons by 2036, and that is the baseline scenario's Dhaka in chapter 15, not a forecast. Where the growth actually sits is the ring: the census counts 1,515,566 households in Gazipur district and 990,647 in Narayanganj district against 3,737,085 in Dhaka district, where the average household holds 3.62 persons against Chattogram's 4.24, the small household profile of a tenancy city [BBS Census 2022, district household tables]. Dhaka district's density stands at 10,067 persons per square kilometre [BBS Census 2022, district population tables].

Housing is where the institutional record is sharpest. The census tenancy table names three categories, owned, rented and rent free: nationally, chapter 14's weighted sums give 80.79 percent owned, 17.29 rented and 1.92 rent free. Dhaka district inverts the country at 27.51 owned, 70.46 rented and 2.03 rent free, and the ring is nearly as tenancy heavy, Gazipur at 37.64 owned, 61.08 rented, 1.28 rent free, and Narayanganj at 48.37 owned, 50.42 rented, 1.20 rent free, while Chattogram runs 64.26 owned, 34.0 rented, 1.74 rent free [BBS Census 2022, district tenancy tables]. The structures are formal even where the tenure is not: 2,534,764 of Dhaka district's 3,737,085 households, 67.83 percent, live in pucca dwellings, 19.12 percent in semi pucca, 12.85 percent in kancha and 0.20 percent in jhupri, chapter arithmetic on the structure tables [BBS Census 2022, district housing structure tables]. The census slum tables count 206,900 slum dwellings in Dhaka district, 5.13 percent of its 4,035,241 dwellings, plus 5,455 floating urban dwellings [BBS Census 2022, district slum tables], against the 51.51 percent urban deprivation share of the UN-Habitat definition chapter 14 carries [WB WDI 2026]. The survey record fills what the two counts hide: the 2024 informal settlement survey of 960 Dhaka dwellings records a mean household income of 17,557.56 BDT per month, a mean residency of 19.06 years, 70.0 percent tin wall construction, 68.85 percent of dwellings dependent on utility supply water and 8.44 percent of households reporting dengue infections [Figshare 2024, DOI 10.6084/m9.figshare.31641787, *]. Nineteen years of mean residency is the finding: the informal stock is not a way station for new arrivals, it is a settled, rented, generation old city with its own disease and climate exposure.

The affordability question the outline asks, prices and rents versus income, cannot be answered from stored series, and the gap is itself the record: no land price index, no rent index and no mortgage stock series exists in any source available to this chapter, chapter 14's measurement gap standing. What is measured is the construction side, and it acquits construction: the mason's daily wage in Dhaka rose from 496 BDT in 2016 to 590 BDT in 2022, about 2.9 percent a year, a slow input cost for a sector alleged to price people out [BBS Wage 2022]. The scarce inputs are land and title. Bank advances to construction stood at 116,815 crore BDT as of 31 December 2023, 8.1 percent of total advances, a flow that chapter 14 shows finances developers with records rather than households with plots [BB Econ 2024]. The supply channel the state itself operates is RAJUK's: the authority's register of approved developer companies runs to serial 826 [RAJUK 2024, registered developer companies list], and its own flagship delivery is Purbachal, whose plot arithmetic is this chapter's restatement of chapter 14's source: 24,251 of 26,213 residential plots handed over against about 300 approved building plans and 1,495 electricity connections [RAJUK 2023, Purbachal progress report]. Printing plots takes the state a decade; occupation is decided by infrastructure, finance and confidence the state does not print.

Mobility's measured record is a registration flow and an absence. The BRTA series for the Dhaka metro area counts 110,206 newly registered vehicles in 2016, a pandemic dip to 118,254 in 2020, a peak of 174,812 in 2022 and 125,851 in 2023, and the mix is the story: motorcycles were 48.7 percent of the 2016 flow, 69.1 percent in 2022 and 71.8 percent in 2023, while private cars were 8.5 percent of 2022 and 7.7 percent of 2023 [BRTA 2023]. The city motorises on two wheels; no stock series exists to say how large the fleet is. The rail alternative exists as an opening and an absence: MRT Line 6 opened in December 2022 with daily ridership claims the operator publishes and no ridership, fare or cost series stored here, chapter 24's finding, which stands; the operator document available to this chapter is the management system for damaged and blacklisted MRT Pass and single journey tickets, an institutional record, not an account [DMTCL 2023]. The status of the MRT-1 and MRT-5 lines under preparation could not be confirmed; chapter 24's sources would resolve it. Speeds and congestion cost have no series at all and are not established; the Dhaka Transport Coordination Authority's strategic transport plan and BRTA would resolve it. The carbon stake of the mobility mix is measurable and modest: transportation emitted 5.7 of the Dhaka urban area's 84.0 million tonne of CO2 equivalent on a 20 year basis in 2024, 6.8 percent, well behind power and waste [ClimateTRACE 2024].

Drainage and heat close the record, because they are the stock the next storm tests. The Dhaka Water Supply and Sewerage Authority's drainage assets in FY20 were 385 kilometre of storm sewer, 80 kilometre of open canal, 10.5 kilometre of box culvert and four permanent pumping stations, Kalyanpur at 20, Dholaikhal at 22, Rampura at 25 and Kamalapur at 15 cumecs, 82 cumecs combined, against a service area above 360 square kilometre, 0.23 cumecs of pumping per square kilometre by this chapter's arithmetic [DWASA 2020]. The utility's own accounts show what carrying this stock costs: revenue of 2,303.46 crore BDT in FY25, 10 percent above FY24, receivables of 807 crore BDT at 30 June 2025, and an interest expense that rose to 173.89 crore BDT in FY25 from 120.40 crore BDT in FY24 as the surface water borrowing loads [DWASA 2026]. Heat raises the intensity side of the same equation: Dhaka's April to May 2024 mean daily maximum of 42.7 degrees Celsius ranks fourth in the 44 year city grid record, behind three readings from the 1990s [NASA POWER 2024, chapter 11's tabulation]. The receiving side of the city's growth is measured too: IOM's Displacement Tracking Matrix counted 795,798 internally displaced persons in Dhaka division of 4,955,527 nationwide on 31 October 2025 [IOM DTM 2025], and chapter 49's relative burden reading, 1.8 percent of the division's population against Barishal's 6.5 percent, disciplines the migration narrative: the megacity grows on economic in-migration first, and disaster displacement chains into it through the labour market, not through direct flight.

Mechanism: a renter city, a two-wheel equilibrium and a pumping ceiling are three prices of the same missing institutions

The renter mechanism is chapter 14's land price mechanism read from the demand side. Chapter 14 shows restricted supply capitalising into land prices inside the planned area, because the Detailed Area Plan restricts conversion while demand explodes. The census adds what that does to tenure: households buy ownership where title is bankable, the rural family plot, and rent where it is not, the apartment and tin-wall city, which is how 70.46 percent of Dhaka district rents while the country holds 80.79 percent owner occupancy [BBS Census 2022, district tenancy tables]. The formal construction sector and the informal tenancy stock are the same market's two faces: 67.83 percent pucca structures house a district where seven households in ten rent, and the 19.06 years of mean slum residency [Figshare 2024, DOI 10.6084/m9.figshare.31641787] shows the rental equilibrium is stable, not transitional. Nothing in the record prices this: the renter majority transacts on unrecorded rents, the developer register to serial 826 [RAJUK 2024] is the only licensed supply channel, RAJUK's Purbachal prints public plots at administrative speed, and the financial system holds no mortgage stock to speak of because chapter 14's title machine, the cadastre, is unfinished. Housing policy is conducted without a price, a rent or a loan series, which is why every affordability claim in the public conversation is not established and why the instrument gap is the finding.

The two-wheel mechanism is queueing theory applied by households. In a district of 10,067 persons per square kilometre [BBS Census 2022, district population tables], road space is rationed by time in the queue, so the marginal motorised household buys the vehicle that queues smallest and cheapest: motorcycles rose from 48.7 to 69.1 percent of new registrations between 2016 and 2022 while cars fell to 8.5 percent [BRTA 2023]. The metro was the intended exit from this equilibrium, and line 6 opened in December 2022 [chapter 24]; but the decision that would test the exit, expanding the network, proceeds without the accounts, because the ridership, fare and cost series are unpublished [chapter 24's finding, standing]. Chapter 24's induced demand record applies to the road alternative: lanes added fill. The carbon cost of getting the equilibrium wrong is small, transport at 6.8 percent of the city inventory [ClimateTRACE 2024]; the productivity cost is the congestion no series measures, and the safety cost sits in a police and hospital record no source available to this chapter holds a city series for; it could not be confirmed, and those institutions would resolve it.

The drainage mechanism is a stockflow mismatch with a zoning co-conspirator. Pumping capacity of 82 cumecs over 360 square kilometre [DWASA 2020] is a fixed stock; the city above it adds roughly 728,000 persons a year [WB WDI 2026] and the wetlands that would hold the water are being filled, chapter 30's wetland account, under flood flow zone protections chapter 14 shows on paper in the 2022 plan. Chapter 28 documents where the money for the water side goes instead: Saidabad phase 3, replanned to a June 2029 target at 16,014.83 crore BDT with 7.65 percent spent, is a supply project, not a drainage one, and the utility's rising interest bill [DWASA 2026] shows the borrowing is already loaded. The heat record sharpens the timing: pre-monsoon convective rain over a filled flood plain is the regime that tests 0.23 cumecs per square kilometre, and the 42.7 degrees Celsius reading of 2024 [NASA POWER 2024] is the intensity marker chapter 11's slow hazard account points at. The mechanism's cruelty is distribution: waterlogging lands first on the low rent, low lying rented stock the survey counts, tin walls at 70.0 percent [Figshare 2024, DOI 10.6084/m9.figshare.31641787], so the drainage ceiling is a regressive tax collected in kind.

The displacement mechanism explains why the megacity keeps growing without being the country's refuge. The IDP stock in Dhaka division, 795,798 persons, is 16.1 percent of the national total against the division's 26.8 percent population share, chapter arithmetic on [IOM DTM 2025] and the census counts chapter 14 cites; disaster displacement lands on the coastal divisions first, chapter 49's reading, and reaches Dhaka as poorer, slower, labour market migration. The city's pull is chapter 08's wage gap and the ring's garment and construction demand, and its absorbing institution is the informal rental stock: the same 70.46 percent renter city, densifying one sublet at a time. Every constraint the city fails to service, drainage, transit, title, does not stop the inflow; it lowers the quality of the stock that receives it.

The decade ahead: four decisions inside the ring, each with a named author

The planning perimeter decision, authored by RAJUK with the two Dhaka city corporations, is whether the widened map gains enforcement. The 2022 Detailed Area Plan targets a 2035 horizon with flood flow and agricultural zones, chapter 14's account, and RAJUK's zone and subzone boundary map, approved at the authority's meeting recorded as 09/2024, reorganises the planning area into eight zones spanning Ashulia, Savar and Uttara in the northwest, Purbachal, Khilkhet and the thanas beyond in the east, Tarabo, Demra and Matuail in the southeast, and Pagla, Fatullah, Chasara and Narayanganj City Corporation in the south [RAJUK 2024, zone and subzone area map]. The drawing is the institutional admission that the corridor city chapters 14 and 30 describe is now the planning object; the decision is whether zoning binds inside it, and the revealing indicators are the flood flow zone conversion record and the building plan approvals the authority publishes.

The Purbachal completion decision, authored by RAJUK with the Ministry of Housing and Public Works, is the occupancy test. The plot arithmetic the progress report carries, beyond chapter 14's headline: 26,213 residential plots split 11,209 at 3 katha, 10,361 at 5, 2,618 at 7.5 and 2,025 at 10 katha, alongside 3,563 non residential plots of which 472 administrative and 1,033 commercial, 7,864 plots allocated to original residents, 315 of 365 kilometre of road built, 42 of 52 bridges, 28 of 43 kilometre of lake, 144 acres of forest retained, the 13.33 kilometre Kuril link road over eight bridges, four schools built and operating, and water supply under a public private partnership whose first phase was inaugurated on 30 January 2023, with 30 acres allocated to police establishments, 26 acres to the Bangladesh-China Friendship Exhibition Centre, 37 acres to the cricket board and 150 bigha to Dhaka University [RAJUK 2023, Purbachal progress report]. What remains is 50 kilometre of road, 10 bridges, 15 kilometre of lake and the institutional plot litigation the report itself lists as a delay cause. This chapter's arithmetic treats the ratio of building plans to handed over plots, 300 against 24,251, as Purbachal's single performance metric: absorption there is what keeps demand off the flood plain and out of the informal densification path.

The transit account decision, authored by DMTCL with the Road Transport and Highways Division and the Finance Division, is publication. Chapter 24 sets the first threshold, that the ridership, fare and cost accounts appear at all, and the fare recovery threshold of 50 percent that marks operational viability; this chapter adds the land value side of the return. The value capture instrument already exists in the state's hands: RAJUK holds commercial plots in Purbachal's central business district and contracted an iconic tower consortium with Kajima Corporation [RAJUK 2023, Purbachal progress report], and station corridor land in a 24.65 million person agglomeration [WB WDI 2026] is the financing base that does not depend on fares. The decision that pairs with it is the MRT-1 and MRT-5 financing design, whose status could not be confirmed, chapter 24's sources resolving it; the fiscal risk is the availability payment commitment chapter 24 prices into chapter 04's account.

The housing finance decision, authored by Bangladesh Bank with the Ministry of Housing and Public Works and the Ministry of Land, is the instrumentation gap stated as a decision. The mortgage market chapter 14's cadastre would enable, 12.7 million digitised khatians with the Bangladesh Digital Survey begun in Chattogram in August 2023 [DLRS 2023, annual report 2022-23], is downstream of land administration; the rent and price indices are downstream of nothing but publication. This chapter's reform scenario assumes a published land price index, rent index and mortgage stock series inside the decade the chapter 15 scenarios price, because every other instrument, holding tax reform, tenant protection, mortgage credit, compensation for acquisition, prices off those series; the stall scenario assumes the gap holds, and the 70.46 percent renter city absorbs the next 728,000 arrivals a year through unmeasured rents and informal densification. The decentralisation question the outline names is owned by chapter 49 on the fiscal base and chapter 14 on the secondary city instrument; this chapter's contribution is the receiving side: the share plateau is the first measured opportunity to redirect growth toward the ring and the secondary cities, and it stays open only while the ring's institutions are actually built.

Three risks print in utility and registry records, and the upside is the plateau itself

Risks. First, the two-wheel lock-in: the motorcycle share of new registrations passes 75 percent while the transit accounts stay unpublished, the queue equilibrium hardens, and the MRT expansion decision is taken on estimates; the revealing indicators are the BRTA registration mix [BRTA 2023] and the DMTCL publication event. Second, drainage saturation: a pre-monsoon event over filled wetlands tests 82 cumecs of pumping [DWASA 2020] while flood flow zones convert inside the new perimeter, and the revealing indicators are the authority's pumping stock against its FY20 baseline and chapter 28's delivery clocks, with the Saidabad phase 3 June 2029 target the nearest dated one. Third, the rent trap: no rent index, no mortgage stock and a census renter share climbing from 70.46 percent, with the informal stock absorbing in-migration through densification until the next shock, disease or flood, prices it; the revealing indicators are the next census tenancy and slum tables against the 206,900 slum dwellings and 5,455 floating dwellings of 2022 [BBS Census 2022, district slum tables].

Upside. First, the plateau dividend: if the largest city share keeps printing increments near zero, the reform scenario's dispersal is already under way through the ring, and the policy task narrows to servicing Gazipur and Narayanganj rather than stopping Dhaka; the revealing indicator is the share series [WB WDI 2026]. Second, the cadastre to mortgage pipeline: chapter 14's digital survey converts a title record into a lending base, and the first published mortgage stock series would be the housing market's birth certificate; the revealing indicator is that series' existence. Third, the value capture dividend: transit corridors through printed land, Purbachal's commercial plots and the station catchments, can finance completion without fare dependence; the revealing indicators are RAJUK's commercial plot allotments and the first DMTCL annual account.

What to watch: five indicators whose thresholds mark the regime

1. Dhaka share of the urban population. Current value 42.22 percent in 2025, with increments of plus 0.11 points in 2024 and plus 0.01 in 2025 [WB WDI 2026]. Threshold: increments held near zero through 2030, which this chapter's dispersal scenario assumes, mark the opening; a print above 45 percent is chapter 14's lock-in regime.
2. Dhaka district renter share. Current value 70.46 percent of households in 2022 [BBS Census 2022, district tenancy tables]. Threshold: a reading above 75 percent at the next census confirms the rent-trap regime; a fall toward 60 percent with rising Purbachal building plans marks the ownership delivery turn.
3. Motorcycle share of new Dhaka vehicle registrations. Current value 71.8 percent in 2023, from 69.1 percent in 2022 and 48.7 percent in 2016 [BRTA 2023]. Threshold: a share above 75 percent is the two-wheel lock-in; a fall below 60 percent alongside a published metro account marks the transit turn.
4. Metro fare recovery ratio. Current value not established; DMTCL reports would resolve it, and chapter 24 sets account publication as the first threshold. Threshold: a ratio above 50 percent marks operational viability; continued non publication through the MRT-1 and MRT-5 financing decisions is the monument risk.
5. Dhaka drainage pumping capacity. Current value 82 cumecs across four permanent stations in FY20 [DWASA 2020], with later readings not confirmed pending the authority's reports. Threshold: a published capacity programme tied to canal recovery marks the drainage regime; a stock unchanged at the next annual report while chapter 28's project clocks slip confirms the saturation risk.

Sources used

[WB WDI 2026] World Bank World Development Indicators via bdpolicy parquet, series: EN.URB.LCTY, EN.URB.LCTY.UR.ZS, EN.POP.SLUM.UR.ZS.

[BBS Census 2022] Population and Housing Census 2022 district tables via bdpolicy parquet: household and tenancy tables, housing structure tables, slum and floating dwelling tables, household size tables, district population and density tables.

[BBS Wage 2022] BBS daily wage rates for construction trades, mason in Dhaka, series: bbs_wage_mason_dhaka via bdpolicy.db.

[BB Econ 2024] Bangladesh Bank monthly economic trends via bdpolicy.db, series: bb_bank_advances_construction, bb_bank_advances_total, from chapter 14.

[BRTA 2023] Bangladesh Road Transport Authority newly registered motor vehicles, Dhaka metro area, total, motorcycle and private car, calendar 2016 to 2023, via bdpolicy.db series brta_dhaka_*.

[DMTCL 2023] Dhaka Mass Transit Company Limited, management system for damaged, ineffective and blacklisted refundable MRT Pass and single journey tickets, via ocr_text/dmtcl.

[ClimateTRACE 2024] Climate TRACE coalition, CO2e 20 year GWP emissions for the Dhaka urban area by sector, calendar 2024, via the hdx/climate-trace/bgd-climate-trace CSVs.

[DWASA 2020] Dhaka Water Supply and Sewerage Authority annual report 2019-20 via ocr_text/dhaka_wasa, storm drainage assets, permanent pumping stations and service area.

[DWASA 2026] Dhaka Water Supply and Sewerage Authority audited financial statements FY2024-25 via ocr_text/dhaka_wasa, revenue, receivables and interest expense.

[NASA POWER 2024] NASA POWER city grid reanalysis, Dhaka April to May 2024 mean daily maximum temperature, from chapter 11's tabulation.

[IOM DTM 2025] International Organization for Migration Displacement Tracking Matrix, internally displaced persons total and Dhaka division as of 31 October 2025, via HDX HAPI in bdpolicy.db, series: hapi_idps_*.

[RAJUK 2023] Rajdhani Unnayan Kartripakkha, Purbachal New Town progress report and Detailed Area Plan documents via ocr_text/rajuk.

[RAJUK 2024] Rajdhani Unnayan Kartripakkha, zone and subzone area map approved at the authority's 09/2024 meeting and registered developer companies list, via ocr_text/rajuk.

[Figshare 2024] Figshare, Dhaka informal settlement household survey of 960 dwellings, DOI 10.6084/m9.figshare.31641787, via bdpolicy.db series dhaka_slums_*.

[DLRS 2023] Department of Land Records and Survey annual report 2022-23, digitised khatians supplied for online upload and the Bangladesh Digital Survey start in Chattogram, via ocr_text/govtwin_min_land, from chapter 14.

---

Verified line by line against primary sources: 133 claims checked, 4 corrected.

Previous: 49 Local government and regional inequality

Next: 51 Land records and property rights

All sixty chapters and the ten narrative chapters.