Interdependencies
Also written Interdependency · Interdependence of outcomes
The problem that one intervention's outcomes depend on other interventions or on external factors, so the change observed cannot be credited to the project alone.
In plain language
A tutoring programme raises exam results. Did the tutoring do it?
Partly. Some of the credit belongs elsewhere — to a parent who made space for homework, a school that stayed open, a mid-day meal that kept a child in class. The workbook has a name for this problem: interdependencies.
Its definition is short. Interdependencies are the outcomes of one or a series of interventions that depend on other interventions, or on other external factors.
Its own example is a peer tutoring programme. Better school results are the outcome. But that outcome also depends on the family support given to the beneficiary. Remove the family support and the same tutoring produces less.
This is listed as one of the challenges of conducting social impact assessment, and it is a hard one. The project's own records cannot show it. Nothing in the tutoring log mentions a parent.
It is also why the fully integrated model captures social, economic and environmental indicators at the same moments. Read together, they show links that any one of them alone would hide.
How it works
The definition (Chapter 7, section 7.4). Interdependencies refer to the outcomes of one or a series of interventions that are dependent on other interventions or other external factors. The workbook's illustration: the outcome of improved educational levels through a peer tutoring programme is also dependent on the family support given to the beneficiary.
Where it sits. It is one of seven challenges the workbook lists for social impact assessment in section 7.4:
| Challenge | The difficulty |
|---|---|
| Lack of standard use of terminologies | Outcome, output and impact are used differently by different organisations — the workbook notes Acumen Fund describes BACO as cost per outcome when it actually focuses on output |
| Lack of standardisation of reporting framework | Multiple national and international frameworks; donors accept only their own formats, so a team publishes several reports |
| Lack of common measures within a sector | Some use cost effectiveness, some benefit-cost ratio, some qualitative methods, so programmes in the same sector cannot be compared |
| Unavailability of accurate and reliable data | Smaller organisations keep no proper database; primary-stakeholder data can be biased by conflict of interest |
| Assumptions | Projections rest on assumptions, such as that every child in an education programme will take up higher education or a job |
| Interdependencies | Outcomes depend on other interventions and external factors |
| Timeframes | Health or microfinance programmes may show impact in 3–5 years; environment or behaviour-change programmes in 15–20 years, so no standard timeframe fits all |
Why it is not the same as an assumption. An assumption is a condition outside the project's control that must hold true for the project to succeed, stated up front in the logframe. An interdependency is about crediting the result afterwards: the change did happen, and something other than this project helped cause it. One is a planning entry; the other is an attribution problem at assessment time.
How the workbook handles it. Three published answers sit in the paper:
- Ask the counterfactual. Chapter 5's review process requires the assessor to ask what would have happened in the absence of the project, and how much of the evidenced change the project actually contributed.
- Measure contribution, not just outcome. The Impact Management Project's CONTRIBUTION dimension exists for this, comparing the effect against what would likely occur anyway in both depth and duration. Its ten impact risk factors include contribution risk by name.
- Capture indicators together in time. The fully integrated model's arrangements cover environmental, economic and social indicators captured at similar points in time, specifically so that interdependencies among the varied indicators are captured rather than lost.
No figure attached. Section 7.4's interdependencies bullet carries no number, no weighting and no formula for splitting credit. The 3–5 and 15–20 year figures in the same list belong to the timeframes challenge, not to this one, so there is nothing numerical to attach to interdependencies itself.
A worked example
Illustrative programme and figures, built on the workbook's own peer-tutoring example.
Vidyarthi Mitra, a Bhopal NPO, runs peer tutoring for 1,200 children in classes 6 to 8 across 15 government schools. Its reported outcome after two years: average marks in mathematics up from 41% to 58%.
An assessor tests how much of that 17-point gain the tutoring can claim.
| What else changed in the same two years | Effect |
|---|---|
| The state added 2 trained maths teachers to 9 of the 15 schools | Classroom teaching improved independently of the tutoring |
| A mid-day meal upgrade cut absenteeism from 22% to 11% | More children present for both tutoring and class |
| A separate NGO ran parent literacy classes in 4 of the 15 schools | Homework supervision at home improved in those 4 |
What the assessor does with it. She splits the 15 schools:
| School group | Marks gain | Reading |
|---|---|---|
| The 4 with parent literacy classes too | 41% to 64% | Tutoring plus family support |
| The 2 with neither new teachers nor parent classes | 41% to 49% | Closest to tutoring alone |
| The remaining 9 | 41% to 57% | Tutoring plus new teachers |
The 8-point gain in the last group is the least entangled figure available. The 23-point gain in the first group is real, and it is not Vidyarthi Mitra's alone.
What the report says. Not "our tutoring raised marks 17 points". Instead: marks rose 17 points across the programme; in schools without concurrent interventions the gain was 8 points; the outcome is interdependent with family support, teacher additions and attendance, and the assessor states this in the limitations and in the contribution discussion.
Why the honest version is the more useful one. A funder reading "17 points, ours" will fund more tutoring. A funder reading the split will fund tutoring and parent literacy classes together — which is where the 23-point number came from.
Why NISM asks about it
Chapter 7 (Social Impact Assessment Reporting, 11% weightage), section 7.4, defines interdependencies and gives the peer-tutoring example. Section 7.2.3 refers to it again as something the fully integrated model is designed to capture, and Chapter 5's review questions attack the same problem from the assessor's side.
Chapter 7's fourth sample question lists interdependencies, assumptions, timeframes and standardisation of reporting framework and asks which are challenges of social impact assessment — the answer is all four. So the high-value revision is the seven-item list in 7.4, not the definition alone. Expect also a question giving the tutoring-and-family-support example and asking which challenge it illustrates.
Common exam traps
- Interdependencies are not assumptions. An assumption is a condition that must hold for success, stated in advance. An interdependency is about how much of an achieved outcome this project can claim. Both appear in the same 7.4 list as separate challenges.
- Nor are they constraints on impact. A constraint limits what the project can achieve; an interdependency means something else helped achieve it.
- The dependency can be on another intervention or on an external factor. The definition names both, so a question describing a second NGO's programme and a question describing a government scheme both count.
- The workbook's example is peer tutoring and family support. Learn it as given; it is the version that appears in question stems.
- Seven challenges in section 7.4, and Chapter 7's sample question tests them as a set. Interdependencies, assumptions and timeframes are the three most often offered together.
- The answer is to state it, not to solve it. The workbook offers no formula for splitting credit. What it requires is the counterfactual question, the CONTRIBUTION dimension, and disclosure in the report's limitations.
- Do not confuse it with contribution. Contribution is the measurement dimension used to handle the problem; interdependency is the problem.
Check yourself
1.An impact report shows that improved exam results from a peer-tutoring programme were strongest among children whose families supported them at home. Which SIA challenge does this illustrate?
- a)Assumptions
- b)Interdependencies
- c)Lack of standard use of terminology
- d)Timeframes
Show the answer
Answer: (b) Interdependencies
Interdependencies are outcomes that depend on other interventions or external factors. The workbook's own example is peer tutoring depending on family support.
Assumptions are projections built in advance (for example, that all children will take up jobs). Terminology is about outcome and impact being used for outputs. Timeframes concern how long impact takes to appear.
Where this is taught
Free preparation for NISM Series XXIIIRelated terms
- ContributionOne of the five Impact Management Project dimensions: how an enterprise's reported effect compares to what would likely have happened anyway, measured through its depth and duration data categories.
- AssumptionAn external condition a project's causal pathway depends on holding true — set out alongside the risk that would follow if it does not — as a distinct column in the Theory of Change table.
- BACO RatioAcumen Fund's ratio, used in the Fully Integrated Model, for judging whether a social enterprise investment is a viable use of capital based on financial, technological and social leverage.
- Constraint on impactAn external factor — such as a market or demand-side mismatch — that limits how much impact a project can achieve, distinct from a challenge in conducting the social impact assessment itself.
- Fully integrated modelThe organisational model in which impact assessment is built into a programme and run continuously rather than at the end, used mainly by social venture capital firms and impact investors.
- Impact Management ProjectThe framework, used by the GIIN's IRIS+ system, that describes impact in five dimensions — What, Who, How Much, Contribution and Risk — with defined data categories for each.
- Impact reportingCommunicating to stakeholders the difference an organisation has made; it may be an independent impact report, or part of an annual report or a sustainability report.