Skills training and enterprise support: the weakest part of most LRPs

Vocational training is the default livelihood intervention for non-farm displacement, and it has the weakest evidence base of anything in a resettlement plan. Tailoring, welding, hairdressing, motorcycle repair, delivered as a three-month course to a cohort selected from the affected population, with a kit at the end.

Some of these work. Most produce a certificate, a sewing machine and a household earning less than it did before, in a market that did not need another tailor.

Why it fails so consistently

• Supply-driven trade selection. Trades are chosen because a training provider offers them, not because local demand exists.
• Market saturation by the programme itself. Training forty people from one community in one trade creates forty competitors in a market that supported two.
• No working capital. A graduate with skills, a kit and no cash for materials cannot trade.
• No premises. Particularly acute for displaced households, whose location is exactly what the project removed.
• Wrong participant. The training place goes to a young household member while the person whose livelihood ended is a fifty-year-old farmer.[2]

Training is an input to a livelihood, not a livelihood. The question a design has to answer is who will pay this person for this skill, and where.

What to require before funding another round

1. A market assessment identifying demand that is unmet locally, with the number of new entrants the market can absorb stated explicitly.
2. An employer or buyer identified in advance — a firm that will hire, a contractor that will subcontract, a buyer that will purchase — rather than an assumption that self-employment follows certification.
3. A capital component, staged and tied to the business actually operating.
4. Premises addressed, whether through allocated space, rent support or siting within a market.
5. Follow-up over at least a year, since the failure point is month four, not the end of the course.[3]

Employment linkage, where the project is the employer

On projects with a large construction workforce, the most direct intervention is employment on the project itself. It has a real advantage — the demand is certain — and two real limits: construction jobs are temporary, and they end when the project does, which is precisely when the livelihood programme should be concluding.[1]

To be worth anything, project employment for displaced households needs a reserved proportion, training sequenced before recruitment rather than after, and a transition plan for what happens at demobilisation. Otherwise it defers the livelihood problem by three years and returns it intact.

Who should deliver it

Livelihood implementation is a specialist function, and engineering-led project units are poorly placed to run it. Contracting an organisation with a track record in the specific livelihood area — an agricultural NGO, an enterprise development body — usually produces better results than an in-house team.

The procurement has to be started early, because in most public agencies engaging such a partner takes the better part of a year, and the programme is supposed to begin before displacement.[4]

Measure income, not attendance

The indicators that matter are net income against baseline, businesses still trading at twelve and twenty-four months, and the proportion of participants whose total household income has recovered. Participation counts, kits distributed and courses completed are inputs, and reporting them as results is the specific reason this component so rarely survives a completion audit.