Turning Hidden Underemployment Into Counted, Full-Time Work

Diagnosis

Bangladesh does not have a jobs gap that shows up cleanly in the unemployment line. It has a quality-of-work gap that hides inside employment. The curated assessment puts underemployment at roughly 25 percent of the labor force per the BBS QLFS. That means a quarter of working people are short on hours, stuck below their skill level, or earning too little from the work they do, even though they are recorded as employed. A headline unemployment rate that looks low can sit on top of this and conceal it entirely.

This matters now for three reasons. First, the official statistics do not currently track it in a decision-usable way: the context flags data_status as needing a collector, and current_state is null, so the state has a problem it cannot yet measure month to month. Second, underemployment is a regime-type, medium-horizon problem, not a one-quarter shock, which means it compounds quietly if left unmanaged. Third, the lead body, the Ministry of Labour and Employment (MoLE), owns the policy levers (labor market information, skills, placement) but those levers are only as good as the signal that points them. You cannot close a gap you do not measure.

Recommended actions

1. Stand up an underemployment metric inside the labor statistics system. Owner: MoLE, working with the Bangladesh Bureau of Statistics through the QLFS instrument. Mechanism: add and publish a standing underemployment indicator (hours-based plus a skill-and-income dimension) as a regular QLFS output, and build the collector the context says is missing. Observable signal: a recurring, dated underemployment series replaces the current null state and is reported each round rather than estimated once.
2. Convert skills programmes from supply-led to demand-led. Owner: MoLE with the Bureau of Manpower, Employment and Training (BMET). Mechanism: tie publicly funded training enrolment to employer-verified vacancy lists and apprenticeship commitments, so courses map to jobs that exist. Observable signal: a rising share of trainees placed into full-time roles within a fixed window after completion, tracked by BMET.
3. Use overseas placement to absorb skilled underemployment, not just headcount. Owner: Ministry of Expatriates' Welfare and Overseas Employment with BMET. Mechanism: target BMET pre-departure certification and bilateral placement channels at occupations where domestic workers are stuck below their skill level, and publish destination-occupation matches. Observable signal: migrant placements shift toward certified skilled categories rather than undifferentiated low-wage flows.
4. Make hiring support conditional on verified full-time jobs. Owner: MoLE. Mechanism: structure any wage-subsidy or hiring-incentive budget line so payment is released only against a verified full-time, above-threshold-hours position, not a part-time or nominal one. Observable signal: subsidy disbursements are matched one-to-one to verified full-time placements, with clawback on positions that revert to underemployment.

Sequencing (first 12 months)

Do the measurement first. Until MoLE and BBS publish a real underemployment series, every downstream programme is flying blind and cannot be evaluated. The metric unlocks the rest: it tells BMET which occupations and regions to target, tells the overseas-employment channel where skilled surplus actually sits, and gives the wage-subsidy line a baseline to be judged against. So months one to six: build the collector and ship the first published series. Months six to twelve: point demand-led training and conditional hiring support at the segments the data exposes.

Risks and constraints

The binding constraint is fiscal and administrative, not conceptual. Demand-led training and conditional wage subsidies cost money and require employer-verification capacity that BMET and MoLE must staff and enforce, or the conditionality becomes a formality. The political risk is that a newly published underemployment number, near the roughly 25 percent the context cites, looks worse than the comfortable headline unemployment rate, creating an incentive to soften or delay it. The measurement reform must therefore be insulated and routine, not optional.

Bottom line

A quarter of the labor force is underemployed yet counted as working, and the state cannot currently measure it. MoLE should build the metric first, then aim demand-led skills, skilled overseas placement, and full-time-conditional hiring support at exactly the segments the new data exposes.