Remittance resilience depends on channels and job quality
Formal remittances reached 30,328.80 million USD in FY25 while new overseas departures fell to 1,015,675 [BB Econ 2025, table XVIII]. Those flows describe different populations: remittances come from workers already abroad as well as recent departures. Their divergence is consistent with stronger formal-channel use after exchange rate reform, but does not establish a change in earnings per migrant. Recruitment costs, job quality, destination risk and settlement arrangements will determine how resilient this household income source remains.
Remittances rose as new departures fell
The stored BMET record runs twenty three years and shows a flow that moves in steps, not trends. Departures went from 188,965 in 2001 to 875,055 in 2008, fell to 390,702 in 2010 after the financial crisis, recovered to 1,008,518 in 2017, collapsed to 211,149 in 2020 in the pandemic, then surged to 1,135,873 in 2022 and the record 1,305,453 in 2023 [BMET 2023]. The fiscal series continues the record into the float: 1,197,128 workers in FY24 and 1,015,675 in FY25 [BB Econ 2025, table XVIII], a 15.2 percent fall that is the first demand side test since the surge.
The money half moves the other way. FY25 formal inflows of 30,328.80 million USD equal Tk 366,907.33 crore [BB Econ 2025, table XVIII]. The World Bank calendar series records 27.52 billion USD in 2024 [WB WDI 2026] and its 2025 reading could not be confirmed, the available dataset's mirrored copy of that series holding zero rows and no published World Bank total for that year resolving it; the IMF BPM6 series carried 22,847.6 million USD of secondary income remittances in calendar 2023 [IMF BOP 2023]. Formal remittances rose while new departures fell [BB Econ 2025, table XVIII]. The flows have different denominators: transfers from the existing migrant stock cannot be assigned to that year's departures. Their divergence is therefore not a measure of wages or the amount shifted into formal channels.
The welfare face is the census. 3,950,155 households received remittances at the 2022 count, 9.81 percent of the 40,257,678 households in the country, split 2,999,600 rural against 950,555 urban [BBS Census 2022], and the same census counted 466,666 returned international migrants [BBS Census 2022]. The stock behind the flow is 7.40 million Bangladeshis abroad at the last UN tabulation [UN DESA 2020]; flow and stock are complements, an annual clearance record against a census based estimate, not substitutes. The distribution question is live and belongs to chapter 34's poverty account: remittances arrive disproportionately to households that could afford to send a member abroad, so the poverty effect is smaller than the macro effect.
The macro effect is the anchor fact: 30,328.80 million USD of FY25 inflows [BB Econ 2025, table XVIII] against the 457.9 billion USD FY25 GDP chapter 03 cites [IMF WEO 2026] is a transfer of 6.62 percent of national income, both legs on the fiscal year, nearly as large as the entire formal tax base chapter 04 measures at 7.64 percent of GDP [WB WDI 2026]. The cushion is also the debt anchor: external debt service measured against exports plus remittances stayed inside a 3.9 to 5.8 percent band across FY18 to FY23 and printed 5.8 percent in FY23 [ERD 2023].
No other resource flow transfers purchasing power to households at this scale with no repayment schedule attached.
Departure destinations differ from remittance origins
On departures, the Gulf dominates. Saudi Arabia took 497,674 workers of the calendar 2023 flow, 38.1 percent, and the United Arab Emirates 98,422, together 596,096 workers or 45.7 percent, while Malaysia's 351,683, up from 50,090 in 2022, is the sharpest corridor shift in the record [BMET 2023]. The historical series make the exposure concrete: the UAE corridor fell from 258,348 workers in 2009 to 3,235 in 2018; Malaysia took 175,927 workers in 2018, then 545 in 2019 and 28 in 2021 before the 2023 reopening; Saudi Arabia itself fell 18.7 percent in one year, from 612,418 in 2022 to 497,674 in 2023 [BMET 2023]. Corridors close as fast as they open, on decisions taken in Riyadh, Dubai and Kuala Lumpur, not Dhaka.
The money map is more diversified than the worker map, and this is the chapter's corrected reading of the corridor risk. In FY25 the United States was the largest corridor at 4,732.9 million USD, 15.6 percent of the total, ahead of Saudi Arabia at 4,264.3 million (14.1 percent), the United Arab Emirates at 4,167.9 million (13.7 percent), the United Kingdom at 3,168.5 million (10.4 percent) and Malaysia at 2,804.7 million (9.2 percent); Oman, Italy, Kuwait, Qatar, Singapore and Bahrain follow at 761.1 to 1,652.5 million each [BB RemStat 2025, country-wise workers' remittance inflows]. The six Gulf corridors sum to 13,656.9 million USD, 45.0 percent of FY25, against 7,901.4 million, 26.1 percent, from the United States and the United Kingdom together [BB RemStat 2025, country-wise workers' remittance inflows].
The FY25 growth rates rank the corridors' momentum: the United States corridor grew 59.8 percent from 2,961.6 million, Malaysia 60.8 percent from 1,744.4 million, Saudi Arabia 55.6 percent from 2,741.5 million, while the UAE fell 10.1 percent from 4,635.3 million [BB RemStat 2025, country-wise workers' remittance inflows]. The US corridor's increase is consistent with stronger formal transfers, but the data do not isolate exchange rates from changes in earnings, remittance behaviour or routing; the Gulf remains the volume base and the oil-cycle exposure.
The stock confirms the two geographies. Of 7,401,763 Bangladeshis abroad at the 2020 tabulation, Western Asia held 3,462,362, 46.8 percent, under half, with India the largest single country at 2,488,471, 33.6 percent, a census based count that includes long settled populations; Saudi Arabia held 1,277,624, the United Arab Emirates 1,095,231, Malaysia 415,717, Kuwait 380,046, Oman 316,467, Qatar 261,672, the United States 261,496, the United Kingdom 241,799 and Italy 135,468 [UN DESA 2020]. The Anglophone and European diasporas that chapter 20's service economy and chapter 60's diaspora account build on are a tenth of the stock and the higher income per head end of it; their remittances already arrive through the banking channel. Women are 16.7 percent of the Western Asia stock, 578,010 of 3,462,362 [UN DESA 2020], the demographic fact behind the female migration mechanism below.
Recruitment, settlement and protection
The migration system is a labour supply chain. A worker is recruited locally, pays an agency or broker, is cleared through BMET, flies on a corridor, and remits through a channel chosen at the destination. Each link prices the next: recruitment fees front load the cost, the corridor sets the wage, the settlement channel sets the share that arrives. The available evidence measures the ends, clearances at the origin and remittances at the destination, and the middle is the part it does not price: fee levels paid to agencies and brokers for Gulf and Malaysia placements could not be confirmed from the cited sources, recruitment-cost surveys and corridor-specific BMET records being the relevant evidence, and the qualitative record of fees at several months' wages for low skill placements remains claim, not measurement.
Two stored facts bracket the fee question from either side. First, the state's own credit line prices the upfront cost: Probashi Kallyan Bank lends departing workers up to BDT 3 lakh, unsecured, at 9 percent simple interest, within three days [MoEWO 2022], a ceiling designed for a fee plus travel bill the bank expects to reach lakhs. Second, the fee is a policy variable: the state recruiting agency BOESL sent 13,757 workers abroad at zero migration cost in FY22, 85 percent of them women [MoEWO 2022], proof that the Gulf and Malaysia fee structure is a constructed artefact, not a market constant.
The 2022 to 2023 Malaysia episode is the demonstration at scale: a corridor opened on a government to government arrangement multiplied from 50,090 to 351,683 workers inside two years [BMET 2023], and the recruitment cost inside it is precisely the part the record does not carry.
The formalisation mechanism of 2024 to 2025 is the system's one measured reform. The taka sat at 110 to the dollar through April 2024, moved to 117.7 in May 2024 and 120 by August, then traded 121.6 to 122.9 across 2025 [BIS 2026]. While the official rate sat below the market rate, a dollar sent through hundi bought more taka than a dollar sent through a bank, and that wedge was the informal channel's fee. The float closed the wedge, the banking channel became the best price in town, and the flows moved: the September 2025 print of 2,685.88 million USD sits against a FY25 monthly average of 2,527.4 [BB Econ 2025, table XVIII].
The cash incentive is the second half of the package: the government has paid an incentive on formal inflows since 1 July 2019, starting at 2 percent and raised to 2.5 percent [MoEWO 2022]. At the FY25 flow the incentive is worth roughly BDT 9,172.7 crore a year, 2.5 percent of Tk 366,907.33 crore, arithmetic rather than a budget line; the outlay actually booked in chapter 04's accounts could not be confirmed here, the Finance Division budget documents being the resolving source. The chapter's reading stands: this was a price reform, and its persistence depends on the price staying honest.
A re-peg would reopen the informal channel on the same arithmetic that closed it.
The female migration mechanism is the system's least recorded and most sensitive part. The ministry's own tabulation puts FY22 departures at 861,263 men and 105,242 women, 966,505 workers in all, a 10.9 percent female share, against BMET's January to December 2022 count of 1,135,873 on its different year basis [MoEWO 2022]. The qualitative record carries concentration in a small number of corridors and occupations, domestic work above all, with the weakest wage protection of any category, and the BOESL zero cost channel's 85 percent female share [MoEWO 2022] shows the instrument works where it is applied. The chapter treats female migration as the test case for the recruitment cost and wage protection decisions: if the system protects the most vulnerable corridor, it protects them all, and if it cannot, no skill upgrade elsewhere compensates.
The vulnerability mechanism is concentration times skill. A cushion financed by a two to one less skilled majority [BMET 2023] in corridors that are 45 percent Gulf by money and 46 percent Gulf by stock sits exposed to three shocks at once: an oil revenue shock cuts Gulf hiring, a nationalisation or wage protection enforcement shock repatriates workers, and an automation shock devalues the specific skill band. No cited series projects Gulf labour demand over the horizon to 2036, and the chapter projects no number of its own; the figure is not established, the World Bank migration outlooks and the Gulf development plans being the named authors that could support a forecast.
The skill history sharpens the point: the skilled share reached 43.1 percent in 2017, then the Gulf and Malaysia less skilled surge, 811,261 less skilled workers in 2022 alone, drove it to 24.8 percent by 2023 [BMET 2023]. The departure mix shifted toward less-skilled categories. These counts do not establish changes in wages per worker or isolate the contribution of channel formalisation to remittance growth.
The hundi mechanism closes the settlement account. The parallel channel priced itself as the gap between the official and the market rate, and the enforcement question changed shape when the float closed that gap: chasing couriers is the old strategy, keeping the official rate honest is the new one, and the indicator to monitor is the same exchange rate series [BIS 2026]. The institutional record is not empty: the national anti money laundering strategy names hundi and hawala as a typology and stands a Central and Regional Task Force for Preventing Illegal Hundi Activities [BFIU 2019], while conviction statistics against hundi networks are not in the available evidence and enforcement is treated here as complementary to the price, never a substitute for it.
Decisions on skills, corridors and returnees
The forces the FY36 scenarios assume are four. First, the Gulf demand cycle: chapter 15's stall scenario assumes Gulf labour demand softens with oil revenue and the flow erodes, the baseline holds remittances near the FY25 level of 30,328.80 million USD while the reform case assumes better jobs support further growth [BB Econ 2025, table XVIII]. Second, corridor diversification: Malaysia's reopening [BMET 2023] shows corridors can be built by policy and closed by policy; Japan, Korea and Europe are the higher wage candidates whose arrival depends on the language and skills systems chapter 32 examines, and the US and UK corridors at 26.1 percent of FY25 money [BB RemStat 2025, country-wise workers' remittance inflows] are the base it would build on.
Third, settlement normalisation: the formalisation either holds at the post float level, making the 2.5 percent incentive [MoEWO 2022] the cheapest export subsidy in the budget, or reverses with a rate regime change. Fourth, the returnee question: the census counted 466,666 returned international migrants [BBS Census 2022], and the reintegration machinery has measured scale, not just intent: Probashi Kallyan Bank disbursed BDT 841.58 crore in loans to 37,111 pandemic returnees in FY22, and its standing reintegration product lends men 9 percent and women 7 percent, BDT 3 lakh unsecured and up to BDT 50 lakh secured [MoEWO 2022].
The training pipeline is the skill decision's capacity check, and it is bigger than the record usually states. Sixty four Technical Training Centres and six Institutes of Marine Technology trained 856,305 people in FY22, 38,579 of them women, across 55 trades, with forty upazila level TTCs added and a one hundred TTC construction project under way [MoEWO 2022]. The pipeline exists; what no cited series measures is employer take up and wage outcomes for its graduates, which is the number that would connect training spending to the skilled share of departures.
The main decisions. The recruitment cost decision, whether fees are capped, disclosed and enforced under the Overseas Employment and Migrants Act and the Recruiting Agents Licence and Conduct Rules 2019 [MoEWO 2022], is the largest single welfare gain available to the bottom 40 percent of migrant households; the Act's enforcement record in numbers could not be confirmed here, the Wage Earners' Welfare Board and embassy labour wings being the resolving sources. The skill decision, pre departure training scaled to the corridor mix, decides the wage ceiling. The settlement decision, holding the float and the incentive together, decides the cushion's recorded size.
The reintegration decision, credit and recognition for the 466,666 returned [BBS Census 2022], decides whether return is a retirement or a redeployment, and the welfare machinery already carries scale: BDT 402.39 crore of assistance to 1.11 million workers and families in FY22, including about BDT 180.10 crore of death compensation and wage arrears to 6,000 families, BDT 13.25 crore of corpse transport and burial to about 3,800 families, scholarships above BDT 9 crore to 5,000 children, and mandatory migrant insurance at a BDT 990 two year premium for BDT 200,000 of cover, half paid by the Board, with 169 claims settled in FY22 [MoEWO 2022].
Risks and opportunities
Risks. First, a Gulf demand break: a quarter under 2,400 million USD of monthly remittances [BB Econ 2025, table XVIII] or departures below 800,000 a year signals the shock, and the indicators to monitor are the monthly series and the BMET clearances. Second, a protection failure: a wage arrears scandal in a dominant corridor would move workers and remittances together, and the indicator to monitor is the embassy complaint record, which is not stored in the cited data. Third, the formalisation reversal: a return to rate management reopens the informal channel on the arithmetic that closed it, and the indicator to monitor is the exchange rate path [BIS 2026].
Upside. First, the skill premium: lifting the skilled share back toward the 43.1 percent 2017 peak [BMET 2023] raises the flow's value per worker permanently, and the training pipeline's 856,305 annual completions [MoEWO 2022] is the capacity to do it with. Second, the corridor build: Malaysia's reopening from 28 workers in 2021 to 351,683 in 2023 [BMET 2023] shows policy can open corridors at scale, and higher wage corridors are version two of the same instrument. Third, the recorded cushion dividend: every billion moved from informal to formal channels is a billion the banking system can intermediate, which chapter 06's deposit base and chapter 04's tax base both price, and the 26.8 percent FY25 step [BB Econ 2025, table XVIII] is the size of the last such move.
Testing the policy case
BMET and Bangladesh Bank should publish recruitment and remittance evidence by corridor while keeping migrant stocks and new clearances distinct. Better settlement prices without stronger formal flows would challenge a price-only explanation. Higher transfers accompanied by rising recruitment debt would not constitute an unqualified welfare improvement.
What to watch
The thresholds below are author-proposed monitoring markers, not official targets or estimated policy effects. Interpret them alongside the source dates and definitions.
- Remittances, Bangladesh Bank fiscal year basis. Latest cited value 30,328.80 million USD in FY25 [BB Econ 2025, table XVIII]. Threshold: a fiscal year below 28,000 million USD marks the cushion eroding; above 36,000 million USD meets chapter 15's reform-path assumption, without by itself proving a skill upgrade. The World Bank calendar series is a different basis and is not a substitute in this indicator.
- Formal monthly remittances. Latest cited value 2,685.88 million USD in September 2025 against a FY25 monthly average of 2,527.4 [BB Econ 2025, table XVIII]. Threshold: a quarter under 2,400 million a month signals erosion; sustained prints above 2,800 signal the formalisation regime holding. Corridor rider: a Gulf share above the FY25 mark of 45.0 percent [BB RemStat 2025, country-wise workers' remittance inflows] means oil cycle coupling is deepening even if the total holds.
- Overseas departures. Latest cited value 1,015,675 workers in FY25 [BB Econ 2025, table XVIII]. Threshold: below 800,000 a year signals a Gulf labour demand shock; a new peak above 1.4 million with the skilled share above 30 percent would be consistent with the quality upgrade regime.
- Skill mix of departures. Latest cited value 24.8 percent skilled in 2023, 323,993 skilled against 654,781 less skilled [BMET 2023]. Threshold: the skilled share crossing 40 percent, a band the record reached in 2017 at 43.1 percent, marks wage upgrading; falling below 20 percent marks the low skill trap deepening.
- Remittance-channel use. The cited aggregates do not measure the formal share of all transfers [BB Econ 2025, table XVIII]. Do not divide total remittances by new departures to estimate earnings or channel performance. Proposed assessment: compare formal-channel volumes, transfer costs and corridor-specific evidence. A formal share requires an estimate of total transfers, including informal channels; the available aggregates do not provide it.
Sources used
[BB Econ 2025] Bangladesh Bank Monthly Economic Trends, October 2025, tables IB and XVIII, via the bb/remittances lake: workers' remittances and persons left for overseas employment, fiscal series.
[BB RemStat 2025] Bangladesh Bank Statistics Department, Monthly Report on Workers' Remittance Inflows, July 2025, Annexure III country-wise workers' remittance inflows FY2016-17 to FY2025-26, held in the cited data at bb/remittances.
[BMET 2023] Bureau of Manpower, Employment and Training series via bdpolicy.db, series: bmet_overseas_employment_total, bmet_overseas_employment_skilled, bmet_overseas_employment_less_skilled, bmet_workers_ksa, bmet_workers_uae, bmet_workers_malaysia.
[MoEWO 2022] Ministry of Expatriates' Welfare and Overseas Employment annual report 2021-22, via ocr_text/expat: incentive history, female and male departures, TTC throughput, Probashi Kallyan Bank loans, Wage Earners' Welfare Board assistance, BOESL zero cost placements, migrant insurance.
[UN DESA 2020] UN International Migrant Stock tabulation via the humanitarian/undesa_migrant_stock_bd parquet, Bangladesh origin stock by destination and sex.
[BBS Census 2022] BBS Population and Housing Census 2022 via bdpolicy.db, series: census2022_remittance_recipient_households_nos, census2022_total_households_nos, census2022_returned_migrants_total_nos.
[WB WDI 2026] World Bank indicator snapshot via bdpolicy.db, personal remittances received BX.TRF.PWKR.CD.DT via the hdx/wb_combined mirror, tax revenue GC.TAX.TOTL.GD.ZS; the bdpolicy.db series wb_remittance_inflows holds zero rows and is not used.
[IMF WEO 2026] IMF World Economic Outlook database, GDP estimate chapter 03 cites.
[IMF BOP 2023] IMF Balance of Payments statistics via bdpolicy.db, series: imf_bop_secondary_income_remittances_usd_million.
[ERD 2023] Economic Relations Division via bdpolicy.db, series: erd_debt_service_export_remittance_pct, debt service to exports plus remittances.
[BIS 2026] Bank for International Settlements USD/BDT series via bdpolicy.db, series: bis_usd_bdt_monthly_eop.
[BFIU 2019] Bangladesh Financial Intelligence Unit, National Strategy for Preventing Money Laundering and Combating Financing of Terrorism 2019 to 2021, via ocr_text/bfiu.
[Not confirmed, chapter 21] Recruitment fee levels, Gulf labour demand projections, embassy wage complaint records, Overseas Employment and Migrants Act enforcement statistics, remittance incentive outlay booked in budget accounts; resolving sources World Bank remittance price surveys, Gulf development plans, Wage Earners' Welfare Board, embassy labour wings, Finance Division budget documents.