I've spent more time than I expected on the word “because”.
It often appears in grant applications, partner decks and impact reports as a simple bridge between two facts:
Attendance improved because the programme removed the main barriers to participation.
The sentence is clean. It sounds reasonable. It may even be true.
It is also making a substantial claim.
It says the programme was responsible for the improvement. It asks the reader to accept that attendance would probably have been lower without the intervention, and that other changes during the same period do not explain the result more convincingly.
That is a great deal of work for one small word.
Most impact overclaiming isn't created by invented data. It happens when the language connecting two pieces of real evidence becomes stronger than the evidence itself.
Impact language often skips the middle
A typical results story begins with something the organisation can clearly prove. It delivered workshops. Distributed equipment. Built infrastructure. Reached a large number of people. These are legitimate achievements. They show that the work took place.
They don't yet show what changed because of it.
Different organisations use terms such as activity, output, outcome and impact in slightly different ways. The distinction underneath them matters more than the label: delivering something and changing someone's circumstances are separate claims.
Grant applications compress that difference easily. Space is tight. Funders want a clear account of results. Applicants have to turn monitoring data, participant feedback and programme records into a coherent story.
Under that pressure, “we delivered” can quietly become “we changed”.
The evidence has to match the question
Evidence becomes useful once we're clear about the question it's being asked to answer.
A programme might have reliable records showing that 5,000 households used a service. That's good evidence of use. It cannot, by itself, tell us whether those households became healthier, and even if they did, one question remains: did this intervention cause it? Each step asks more of the evidence.
This is where impact language gets ahead of the research. Several true facts sitting beside each other - people used the service, an outcome improved, published research supports the mechanism - can make an argument more plausible without establishing causation.
The same applies to qualitative evidence. Interviews can explain how people experienced an intervention or why they changed behaviour, which matters for the mechanism. But they answer a different question from a controlled comparison designed to estimate whether the intervention produced the change.
The useful move is to ask what each piece of evidence actually establishes before deciding what sentence it can support. Operational data shows something happened, outcome data shows a condition changed, qualitative research explains why, and a stronger evaluation design shows how much of that change the intervention can reasonably take credit for. They strengthen each other. Their roles stay different.
That matters especially with a word like “because”. It doesn't just connect two facts, it claims a relationship between them. Once a sentence makes that move, evidence that both things happened is no longer enough.
“Because” introduces an alternative world
The central problem in causal claims is easy to describe and difficult to solve.
We can observe what happened after a programme. We cannot observe the same people, at the same time, in an otherwise identical world where the programme did not exist.
That unseen alternative is the counterfactual.
J-PAL explains the logic through impact evaluation: the result we care about is the difference between what happened with the programme and what would have happened without it. Because both outcomes cannot be observed for the same participant, evaluators use comparison groups and other methods to estimate the missing alternative.
Most founders and impact organisations won't have experimental evidence for every claim. Such evidence may be impractical, expensive or unsuitable for the work being assessed.
The counterfactual still matters because it exposes the question hidden inside causal language:
Why do we believe this result would have been different without the programme?
Suppose employment rose among participants after a training initiative.
The training may have improved their skills. The local labour market may also have grown. Another organisation may have provided financial support. The people most likely to find work may have been more likely to enrol or to respond to the follow-up survey.
Several of these explanations can be true at once.
The programme may still have played a meaningful role. The available evidence may support a contribution claim rather than a clean attribution claim.
Contribution is a substantive finding
I've seen organisations treat “contributed to” as a timid substitute for “caused”. But when used properly, it can be a real conclusion.
A credible contribution case explains how the intervention was expected to create change and then tests that explanation against the evidence available. It asks whether the programme was delivered as intended, whether the expected outcomes occurred, whether the proposed mechanism appears in participant or monitoring evidence, and which other influences may have mattered.
The UK government's Magenta Book describes contribution analysis as an evidenced line of reasoning. It is particularly useful where experimental attribution is unavailable or unsuitable.
For a smaller impact organisation, that reasoning might draw on baseline and follow-up data, programme records, participant interviews, external research and a clear theory of change.
No single source has to prove the entire case. Together, they may support a well-founded conclusion that the programme helped produce the result.
That allows the organisation to take credit for the role it can defend while recognising that social, economic and environmental outcomes rarely have one cause.
The verb tells the reader what has been established
Causal verbs are not cosmetic choices made during the final edit.
They tell the reader what relationship the organisation believes the evidence has established.
“Caused”, “produced”, “resulted in” and “led to” usually suggest that the intervention was responsible for the outcome. Those formulations carry a relatively high evidentiary burden. The organisation should be able to explain how other plausible causes were examined.
“Contributed to”, “helped” and “supported” allow other influences to remain present. They still need evidence connecting the intervention to the result.
“Enabled” often sounds cautious. It can carry an equally strong implication. Saying a programme enabled an outcome may suggest that the result could not have occurred without it.
Observational language stays closer to what the evidence directly contains:
- participants reported a change;
- monitoring data showed an increase;
- the outcome was observed among respondents;
- the result improved during the programme period;
- the intervention was associated with the change.
These formulations are not interchangeable. Each tells the reader something different about the source, strength and limits of the claim.
Changing one verb cannot repair weak reasoning. Replacing “caused” with “supported” while leaving the rest of the argument untouched may only make the overstatement less visible.
The sentence needs an evidence trail.
There is real pressure to be confident
Funders, partners and impact investors need clarity. They want to know why the problem matters, whether the intervention is credible and what their support could make possible. A narrative buried in qualifications becomes hard to assess.
Founders also know their claims will be compared with those of organisations willing to speak more boldly. Careful language can feel commercially risky when others are promising transformation.
That pressure is understandable, but it can still create problems later.
An inflated claim may survive an early review and fail during due diligence. It may contradict programme data, create expectations the operating team cannot meet or force the organisation to defend a causal account it never established.
Programme teams understand what was delivered. Monitoring and evaluation colleagues understand the strength and limits of the data. Leadership decides which claims the organisation is prepared to own. Communications has to turn those decisions into language that remains clear under scrutiny.
Precision can make the case stronger
Responsible impact writing does not require a paragraph full of “may”, “could” and “possibly”.
It requires specificity.
Consider this claim:
The programme transformed participants' livelihoods.
It is confident, broad and difficult to examine.
A more useful version might read:
Six months after completing the programme, 62 per cent of surveyed participants reported earning income from at least one of the skills taught. Interviews indicated that access to equipment and peer support also influenced whether participants could apply the training.
The figures are illustrative, but the structure matters.
The second version tells the reader who was assessed, when the assessment took place and what was reported. It separates the observed result from the interpretation of why it occurred. It also leaves room for influences outside the training itself.
The claim has become narrower. The evidence is easier to see, and the result is more informative.
This is often the real work in impact communications: preserving conviction after removing what the evidence cannot carry.
Review the relationship, not only the sentence
Before approving a causal claim, a founder or impact organisation should be able to answer six questions:
- What exactly changed?
- How was the change measured or observed?
- What evidence connects it to the intervention?
- Which other factors may have influenced the result?
- Does the chosen verb describe causation, contribution or association?
- Could the organisation explain that choice if a funder or evaluator asked how it knew?
These questions belong earlier than the final edit.
They affect what the organisation measures, which participant stories it collects and how programme teams document the route from activity to outcome. They also determine whether future claims can become stronger.
Founders and impact organisations do not have to choose between a compelling story and an honest one. They do have to separate what they can demonstrate, what they can reasonably infer and what remains an ambition.
Before “because” reaches the final application, report or partner deck, someone should be able to explain the evidence it is being asked to carry.