What Scotland tested
For funders and delivery partners designing cash assistance, Scotland's Cash-First Programme offers something rare: an official evaluation that names its own weak points. The lesson is not that cash works or fails. It is that a programme built on short-term, non-standardised funding cannot prove durable impact, however well it is delivered.
The programme sat under the Scottish Government's “Cash-First: Towards Ending the Need for Food Banks in Scotland” plan, nine actions over three years to reduce reliance on food parcels. Action 1 funded new local partnership approaches, committing up to £1.8 million (about US$2.4 million) over three years across eight delivery partner areas: Aberdeen City, Aberdeenshire, City of Edinburgh, Fife, Glasgow City, North Lanarkshire, Orkney and West Dunbartonshire.
It ran for 30 months, from October 2023 to March 2026. The final evaluation, published on 29 September 2026, covers activity from April 2024 to December 2025. Each area received an allocation of up to £200,000.
The results, as reported
Delivery partners reported 10,611 interactions and distributed £526,533 (about US$697,000) directly to individuals. Because spending was flexible, roughly a third of the funding went out as cash or shopping vouchers while two-thirds paid for staffing, partnership development, referral pathways and data processes. The evaluators describe the programme as “resource intensive”.
The findings are mixed but honest. Service users preferred cash or cash equivalents, valued the choice, and reported respectful interactions. Over half said they worried less about running out of food, and self-reported food-bank use fell. But impacts on food insecurity itself were inconsistent, and the report states plainly that evidence of sustained or transformational change “was limited, reflecting short term funding”.
Impacts were strongest where emergency income was combined with wraparound support: welfare rights advice, benefits checks, and help with housing or energy costs. Over half of service users received that wider support; around two-fifths did not receive it, or did not recognise it, which the report links to a reduced chance of preventing future hardship.
The data gap that undercut the evaluation
This is the finding funders should read twice. The absence of standardised data-sharing protocols, and inconsistent data collection — especially across food banks and informal provision — limited the programme's ability to track repeat use, attribute outcomes and evidence changes in food-bank demand. It also slowed information-sharing between services, so some people had to re-tell their story to each new worker.
Why this matters: the programme's headline goal was to reduce food-bank reliance, yet its own design could not measure that at population level. Self-reported declines in food-bank use are the weakest form of that evidence. They record what a participant says, not whether demand shifted where they live.
Data sharing did improve in some areas and strengthened coordination, but progress was constrained by the time needed to negotiate GDPR-compliant agreements and by limited frontline capacity to embed new systems.
Why short funding capped the impact
Structural drivers — in-work poverty, rising living costs and debt — often limited longer-term change, and the pilot was too small and too short to move them. Both service users and delivery partners were less confident about longer-term positive impacts. Some resistance to cash-first persisted among partners on safeguarding grounds, and views on the role of food banks differed. Awareness among potential users stayed uneven, partly because most partners deliberately limited promotion to manage referral volumes.
A pilot that must spend its energy building referral pathways and data agreements from scratch has little room left to embed the change it was funded to make.
What to build in from day one
Treat this evaluation as a design checklist, not a verdict.
First, agree data-sharing terms before the first payment, not after. Budget time and legal capacity for GDPR-compliant agreements between the cash provider, food banks and informal providers; without them, repeat use and demand shifts stay invisible.
Second, define up front what “reduced food-bank reliance” means and how it will be measured across the whole local system, not only among participants. Record a baseline before you start.
Third, fund the wraparound support as core, not optional. Two-fifths missing it is a design failure with a measurable cost in crises never prevented.
Fourth, be explicit about duration. If the goal is change that outlasts the funding, the funding has to outlast the pilot.
The honest conclusion for funders: this pilot demonstrated added value in dignity, choice and short-term wellbeing, and it demonstrated its own limits in evidence. Build the measurement in first, or plan to learn less.
Caveats
The figures here are the Scottish Government's own evaluation totals, not an independent audit. Dollar equivalents use the European Central Bank euro reference rates published 12 October 2026 (EUR/USD 1.1181; EUR/GBP 0.84478, giving about US$1.32 per £1), rounded for readability and applied as a common rate to both sums; they communicate scale, not historical dollar spending or purchasing power. The evaluation covers activity from April 2024 to December 2025, while the value of spend reported covers the whole programme.
Sources
Cash-First Programme: final evaluation report — Scottish Government, 29 September 2026. https://www.gov.scot/publications/cash-first-programme-final-evaluation-report/
Euro reference exchange rates — European Central Bank, 12 October 2026. https://www.ecb.europa.eu/stats/eurofxref/eurofxref-daily.xml
About the author
Vincent Dania is a social protection and programme delivery specialist based in Abuja, Nigeria. He leads Hyrax, the AI enablement company, and is a PhD researcher in social protection at the Institute for Social Protection Studies, Nnamdi Azikiwe University, Awka, where his work examines AI and labour-market risk among informal workers. Profile, publications and consultancy: https://vincentdania.name.ng