Course lexiconFoundations & Diagnosis
Intervention
A deliberate, resourced action designed to change an outcome: a transfer, service, rule or nudge introduced into a system to move it from what is to what works.
Course lexicon
Intervention Design Lexicon
64 terms from the course, defined for practitioners who diagnose problems, design development interventions and carry proven programmes from model to scale. Search, or filter by stage of the design journey.
Foundations & Diagnosis
Theory of change
The explicit causal story linking activities to outcomes through assumed mechanisms and pre-conditions. It names what must be true, not only what will be done.
Foundations & Diagnosis
Mechanism vs activity
An activity is what a programme does (train workers); a mechanism is why that changes behaviour (workers gain status and trust). Designing for the mechanism, not the activity, is what lets an intervention travel.
Foundations & Diagnosis
Results chain
The ordered sequence from inputs and activities to outputs, outcomes and impact that a programme is expected to produce. It separates what you control from what you hope to cause.
Foundations & Diagnosis
Dose-response
How an outcome varies with the intensity or duration of exposure. Many interventions have a threshold below which a thin dose buys nothing.
Foundations & Diagnosis
Problem-driven design
Starting from a precisely defined local problem rather than a favoured solution, so the intervention is fitted to the constraint that actually binds (Andrews, Pritchett & Woolcock 2017).
Foundations & Diagnosis
Binding constraint
The single bottleneck that most limits an outcome, such that relaxing it yields the largest return. Diagnosis finds it before design proposes a fix.
Foundations & Diagnosis
Isomorphic mimicry
Adopting the forms of successful organisations (manuals, logos, best-practice templates) while lacking their function, so reform looks modern but changes nothing (Andrews, Pritchett & Woolcock 2017).
Foundations & Diagnosis
Behavioural diagnosis
Locating the specific psychological or informational friction (present bias, hassle costs, low salience) that stops people acting in their own interest, before designing a nudge (World Development Report 2015).
Foundations & Diagnosis
Premature load-bearing
Asking a fragile system to carry responsibilities beyond its real capacity, so it collapses or fakes compliance — a core failure mode of imported blueprints (Pritchett, Woolcock & Andrews 2013).
Proven Models
Graduation approach
A sequenced, time-bound package (asset transfer, consumption support, coaching, savings) that moves the ultra-poor into self-sustaining livelihoods; BRAC’s model, validated across six countries (Banerjee et al. 2015).
Proven Models
Conditional cash transfer (CCT)
Cash paid to poor households on condition they use health or education services, as in Mexico’s Progresa. It ties income support to human-capital investment.
Proven Models
Unconditional cash transfer (UCT)
Cash given with no strings, trusting recipients to spend well. Evidence such as GiveDirectly shows durable gains without the monitoring cost conditions impose.
Proven Models
Direct Benefit Transfer (DBT)
India’s system of paying subsidies and entitlements straight into bank accounts, cutting intermediaries and leakage in schemes like LPG and the PDS.
Proven Models
JAM trinity
The Jan Dhan accounts, Aadhaar identity and Mobile connectivity stack that makes DBT possible at scale by giving each beneficiary a verifiable, payable identity.
Proven Models
Community health worker
A trained local resident who delivers basic health services and referrals, extending a thin formal system into the last mile at low cost.
Proven Models
ASHA
India’s Accredited Social Health Activist under the NRHM: a village woman, incentivised per task, who links households to public health services. One of the world’s largest CHW cadres.
Proven Models
Self-help group (SHG)
A small savings-and-credit group, usually of women, that pools funds and builds collective agency; the base unit of India’s DAY-NRLM.
Proven Models
SHG federation
Higher-tier structures (village organisations, cluster and block federations) that aggregate SHGs to access bank finance, run enterprises and negotiate with the state (Ostrom 1990).
Proven Models
Producer organisation (FPO)
A farmer-owned collective that aggregates produce and inputs to win the scale and bargaining power smallholders lack individually.
Proven Models
Teaching at the Right Level (TaRL)
Pratham’s approach of grouping children by learning level rather than grade and teaching foundational skills directly; among the most cost-effective education interventions (Banerjee et al. 2017).
Proven Models
Cash-plus
Pairing a cash transfer with a complementary service (coaching, nutrition counselling, market links) because money alone often does not clear every binding constraint.
Proven Models
Nudge / choice architecture
Changing how options are presented (defaults, framing, reminders) to steer behaviour without mandates or new incentives (Thaler & Sunstein 2008).
Proven Models
Big push
A coordinated, large investment across sectors at once, on the argument that complementarities mean only a jump, not a nudge, escapes a poverty trap (Rosenstein-Rodan 1943).
Targeting & Delivery
Proxy-means test (PMT)
Estimating a household’s welfare from observable assets and characteristics when income is unverifiable, to rank who qualifies. Used in India’s SECC-based targeting.
Targeting & Delivery
Community-based targeting
Letting local bodies or the gram sabha identify the poor, exploiting local knowledge but exposing selection to elite capture.
Targeting & Delivery
Self-targeting
Designing benefits so only the needy find them worth claiming, as with the manual-labour requirement in MGNREGA that screens out the non-poor.
Targeting & Delivery
Universalism
Providing a benefit to all rather than a targeted few, trading higher fiscal cost for near-zero exclusion error and lower administrative burden.
Targeting & Delivery
Inclusion error
Giving benefits to people who should not qualify — leakage to the non-poor. One of the two errors every targeting rule trades off.
Targeting & Delivery
Exclusion error
Missing people who should qualify. Often the graver error in anti-poverty programmes, and typically worse under tight targeting.
Targeting & Delivery
Delivery chain
The full sequence of actors and steps between a policy decision and a citizen receiving the service. Interventions fail in its links, not in their intent.
Targeting & Delivery
Last mile
The final, hardest stretch of delivery to the remotest or poorest beneficiary, where cost per person is highest and systems are thinnest.
Targeting & Delivery
Missing middle
Households too well-off for welfare yet too poor or informal for tax-based benefits, who fall through the design of most targeted schemes.
Targeting & Delivery
Leakage
The share of resources that never reaches intended beneficiaries, lost to ghost claimants, diversion or corruption. DBT and Aadhaar aim to squeeze it.
Targeting & Delivery
State capacity
The administrative ability to actually implement, monitor and enforce a policy. Weak capacity turns good designs into paper (Pritchett 2013).
Targeting & Delivery
Implementation fidelity
How faithfully a programme as delivered matches the programme as designed. Low fidelity, not a wrong model, sinks many scaled interventions.
Incentives & Cost
Incentive compatibility
A design in which acting in one’s own interest also serves the programme’s goal, so no one must be monitored into doing the right thing.
Incentives & Cost
Principal-agent problem
When the party doing the work (agent) has different goals and better information than the party who wants it done (principal), inviting shirking or gaming.
Incentives & Cost
Multitasking problem
When paying for one measurable task diverts effort from unmeasured but valued tasks; incentivising quantity can crowd out quality (Holmstrom & Milgrom 1991).
Incentives & Cost
Cost-effectiveness
Outcome achieved per unit of spending, letting very different programmes be compared on a common yardstick (e.g. learning gain per rupee).
Incentives & Cost
Cost per outcome
The all-in cost of producing one unit of the result that matters (a life saved, a child reading), not merely one unit of output delivered.
Incentives & Cost
Elite capture
Local powerholders diverting a programme’s benefits or governance to themselves, a recurring risk in community-based delivery and targeting.
Incentives & Cost
Free-riding
Enjoying a collective benefit without contributing to it, which erodes provision of public and club goods (Olson 1965), though communities often solve it through local rules (Ostrom 1990).
Incentives & Cost
Cream-skimming
Providers selecting the easiest-to-serve beneficiaries to hit targets, leaving the hardest cases unserved and inflating apparent success.
Incentives & Cost
Willingness to pay
What beneficiaries would give up for a good, used both to gauge real demand and to debate whether charging even a token price screens out the poor.
Scaling
Voltage effect
The tendency of an intervention’s effect to shrink as it scales, as the conditions that made the pilot work fail to replicate (List 2022).
Scaling
Scalable evidence
Findings that hold not just in a trial but under the delivery, staffing and selection conditions of a full-size programme — the bar policy needs, above internal validity alone.
Scaling
External validity
Whether a result found in one place, population or scale transports to another. High internal validity does not guarantee it.
Scaling
Pilot-to-scale collapse
When a result proven at pilot size vanishes on rollout because charismatic founders, motivated staff or a favourable site cannot be reproduced.
Scaling
Negative economies of scale
Rising per-unit cost or falling effect as a programme grows, from stretched supervision, thinner talent and administrative strain.
Scaling
General-equilibrium / spillover effects
Effects that appear only at scale, when a programme shifts prices, wages or norms for everyone, not just its direct recipients (Muralidharan & Niehaus 2017).
Scaling
Government adoption
Scaling by having the state absorb an intervention into its own systems and budgets, gaining reach but risking dilution through bureaucratic constraints.
Scaling
NGO replication
Scaling by an organisation extending its own delivery to new sites, keeping fidelity higher but growth slower and costlier than government routes.
Scaling
Functional scaling
Growing the function or outcome rather than copying a fixed programme, adapting the form to each new context while holding the mechanism constant.
Scaling
Scaling readiness
A structured assessment of whether an innovation and its enabling conditions (supply chains, finance, policy) are mature enough to scale, and what is still missing (Sartas et al. 2020).
Adaptation & Practice
Problem-Driven Iterative Adaptation (PDIA)
An approach that solves locally defined problems through rapid cycles of experiment, feedback and adjustment, building capability by doing rather than importing blueprints (Andrews, Pritchett & Woolcock 2017).
Adaptation & Practice
Adaptive management
Steering a programme by continuously testing assumptions against evidence and revising the plan, rather than executing a fixed design to completion.
Adaptation & Practice
Blueprint vs process design
Korten’s contrast between a fixed plan imposed from above and a learning process that fits the programme to people and place as it goes (Korten 1980).
Adaptation & Practice
Developmental evaluation
Evaluation embedded in real time to inform an evolving intervention, rather than judging a finished one after the fact (Patton 2011).
Adaptation & Practice
Positive deviance
Finding the few in a community who already achieve better outcomes with the same resources, and spreading their practices as a home-grown solution.
Adaptation & Practice
Learning loop
The cycle of act, measure, reflect and adjust that turns delivery into evidence and lets an intervention improve while it runs.
Adaptation & Practice
Deliverology
A centre-of-government method of setting targets, tracking data and running delivery routines to drive frontline results, pioneered in the UK and adapted in several states (Barber 2007).
Adaptation & Practice
Do no harm
The obligation to anticipate and avoid the unintended damage (to markets, conflict, dignity) an intervention can cause, weighing costs against intended good.
Adaptation & Practice
Fidelity-adaptation trade-off
The tension between delivering a proven model exactly and adjusting it to local context: too rigid and it does not fit, too loose and its active ingredient is lost.
No terms match. Try a broader search or a different filter.