The Missing Middle Between Good Intentions and Real Change

Jul 27, 2026

10 min read

0 views

78%

A familiar disappointment: why so many good ideas fail in the real world

Most organizations are not short on good intentions. They are short on proof.

A new strategy is announced. A social program launches. A foundation funds a promising intervention. A company rewrites its values on the wall and calls it transformation. Yet months later, the same uncomfortable question returns: did anything actually change, and if so, because of what?

That question is more than a bureaucratic annoyance. It is the fault line between aspiration and reality. We live in a time when everyone wants impact, but very few systems are designed to trace the path from action to outcome. The result is a strange modern contradiction: we have more measurement than ever, yet often less understanding of causality.

This is where a deeper tension emerges. On one side sits the ambition to build a better economy, a better institution, a better society. On the other side sits the discipline of asking, with humility, what actually produced the change. Put differently, vision without causal thinking becomes wishful thinking, and causal thinking without vision becomes empty bookkeeping.

The most valuable work happens in the missing middle, where theory meets evidence, and where moral purpose meets analytical discipline.

The real question is not whether something worked, but how change happened

A surprising number of debates about progress collapse into a binary: success or failure. But real systems do not behave like light switches. They behave like weather patterns, ecosystems, and economies, full of interacting forces, delays, feedback loops, and unintended effects.

That is why the most important question is rarely, “Did the program work?” It is, “What was the causal pathway?”

A causal pathway is the chain of events that connects an intervention to a result. If a training program leads to more hiring, the pathway might include improved skills, stronger confidence in interviews, better employer perception, and access to networks. If a community initiative reduces violence, the pathway may include trust building, changes in social norms, increased informal monitoring, and shifts in local opportunity. Without a pathway, a result is just a number. With a pathway, a result becomes an explanation.

This matters because action in the world is always mediated. We do not directly “create” outcomes. We perturb systems and observe how those systems respond. A policy can succeed for the wrong reason, fail for the right reason, or appear to work while masking hidden tradeoffs. Impact is not a trophy. It is an inference.

The deepest form of accountability is not proving that something happened, but understanding why it happened well enough to repeat, adapt, or abandon it.

That shift changes everything. It moves organizations away from vanity metrics and toward learning systems. It also changes the nature of leadership. The question is no longer whether a leader can declare success, but whether a leader can map causality under uncertainty.

Contribution analysis: a practical answer to an impossible standard

In many complex settings, the dream of perfect attribution is seductive but unrealistic. Real change is usually produced by multiple actors, overlapping initiatives, and outside shocks. A nonprofit improves youth outcomes, but so do schools, families, labor markets, and local policy. A company reduces turnover, but so do market conditions and leadership changes. A country sees better health indicators, but only partly because of one program.

This creates a trap. If we demand absolute proof that one intervention caused one outcome, we may end up proving nothing at all. Yet if we give up on causality, we drift into storytelling without evidence. The useful middle ground is contribution analysis.

Contribution analysis asks a different kind of question: given the context, the theory of change, and the observed evidence, how plausible is it that this intervention contributed to the outcome? It does not pretend to isolate every variable. Instead, it builds a reasoned case through multiple forms of evidence, including:

A clear theory of how change was supposed to happen
Evidence that the expected steps actually occurred
Consideration of alternative explanations
Comparison between what was expected and what was observed
Iterative refinement of the story as new evidence appears

The power of this approach is philosophical as much as methodological. It accepts that in complex systems, certainty is often the wrong goal. What we need instead is disciplined confidence. Not blind faith, not fake precision, but warranted belief.

This is a crucial insight for anyone trying to build a better capitalism, a better public institution, or a better civic initiative. Systems change cannot be judged only by outcome snapshots. It must be assessed by whether the underlying mechanisms are strengthening or weakening. A beautiful theory with no observable pathway is fantasy. A noisy intervention with a visible pathway may be the seed of durable change.

A useful analogy is navigation. A sailor does not need perfect certainty about every current to reach port. But the sailor does need repeated fixes, a map, and a working model of how wind and tide interact. Contribution analysis is that kind of navigation tool for social change.

Why the measurement debate is really a debate about power, not just method

It is tempting to treat evaluation as a neutral technical exercise. In practice, it is also political.

Who gets to define success? Who decides which outcomes matter? Who bears the burden of proof when a system changes slowly? These are not small questions. A corporation may celebrate quarterly earnings while ignoring worker insecurity. A government may optimize for visible outputs while neglecting dignity or trust. A funder may demand measurable results that are easiest to count, not those most meaningful to human lives.

This is why evaluation and capitalism should not be separated. Any system that claims to create value must also answer a harder question: value for whom, and by what route? If an organization says it is improving society, it should be able to explain the mechanism by which that improvement occurs. Otherwise, it risks mistaking correlation for contribution, and contribution for virtue.

The same logic applies inside institutions. If promotions rise after a leadership training, was it the training, or a broader cultural shift? If emissions fall after a sustainability pledge, was it the pledge, or market pressure? If employee engagement improves, did the new policy matter, or did people simply adjust to a new cycle? The causal lens does not cynically dismiss progress. It protects progress from self deception.

The deeper danger is not failure. It is illusory success. When organizations cannot distinguish between theater and transformation, they may scale the wrong thing. They may pour resources into visible outputs that do not drive outcomes. They may reward stories that sound good instead of mechanisms that actually work.

A strong causal practice therefore becomes a moral practice. It asks institutions to be honest about limits, to distinguish aspiration from evidence, and to keep learning after applause has arrived.

A framework for thinking in pathways, not slogans

The most useful shift is mental, not managerial. To think causally, stop asking whether an initiative is “good” in the abstract, and start asking where it sits in the chain of change.

Here is a simple framework for doing that.

1. Name the mechanism

What, exactly, is supposed to change? Not just the outcome, but the intermediate behavior or condition that makes the outcome possible.

For example, a financial literacy workshop does not reduce poverty by magic. It may improve budgeting habits, increase uptake of beneficial products, or reduce costly mistakes. If none of those intermediate shifts happen, the workshop’s outcome story is weak.

2. Look for evidence at each step

Do not wait until the final outcome to judge whether the intervention is taking hold. Check whether the pathway is functioning.

A workforce program might track not only job placement, but interview performance, employer callbacks, and skill retention. A community initiative might track trust, participation, and network density before waiting for long term crime data.

3. Test rival explanations

Every result has competitors. Economic growth, seasonality, policy changes, leadership turnover, selection effects, and plain luck all matter. The stronger your alternative explanations, the more carefully you must interpret the evidence.

This is not a reason for paralysis. It is a reason for humility.

4. Use the theory to improve the intervention

A good causal model is not just a report. It is a design tool. If the pathway is weak, repair the weak link. If the wrong population is being reached, adjust targeting. If the mechanism depends on trust, invest in trust before investing in scale.

5. Treat learning as part of the intervention

In complex environments, the first version is rarely the best version. An organization that cannot adapt its theory of change is not really learning, it is just persisting.

The goal is not to be right once. The goal is to become less wrong in public.

This framework works because it respects both ambition and uncertainty. It does not ask organizations to prove omniscience. It asks them to become legible to themselves.

What better capitalism looks like when it is forced to explain itself

If capitalism is to deserve trust, it must do more than generate activity. It must generate accountable value.

That means the most serious reform is not simply moral language or larger budgets. It is causal discipline. A healthier economic system would reward institutions that can explain how their products, policies, and investments produce durable benefits. It would discourage systems that extract short term gains while hiding long term harm in the fog of complexity.

Imagine a business that can show not just revenue growth, but the pathway by which it improved worker mobility, reduced waste, deepened customer well being, or strengthened community resilience. That is a richer story than “we made money.” It is also a more defensible one.

The same principle applies to philanthropy and public policy. A funder should not only ask whether an initiative scaled. It should ask whether the causal mechanism became stronger as scale increased. A government should not only ask whether a program reached more people. It should ask whether the pathway remained intact under pressure.

This is where the intersection of capitalism and contribution analysis becomes especially powerful. Both are, at their best, systems for allocating scarce resources toward desired outcomes. Both therefore need a way to distinguish genuine contribution from convenient narrative. When that discipline is absent, capital chases optics, institutions chase legitimacy, and change becomes performative.

When that discipline is present, something better becomes possible: institutions can argue about values while agreeing on evidence. That is how mature systems evolve.

Key Takeaways

Stop asking only whether something worked. Ask what causal pathway linked action to outcome, and whether that pathway is visible in the evidence.
Use contribution analysis when attribution is impossible. In complex systems, the goal is often a credible case for contribution, not perfect isolation of causality.
Track intermediate signals, not just final outcomes. Intermediate changes reveal whether the mechanism is functioning before the final result appears.
Treat measurement as a moral practice. Clear causal thinking protects institutions from self deception and keeps them accountable to real value.
Build systems that learn. The best interventions are not only effective, they become more effective because they are designed to be revised.

The new standard: not proof of perfection, but explanations that earn trust

The deepest lesson here is that progress should not be judged by confidence alone, nor by metrics alone, but by the quality of the explanation connecting the two.

We do not need institutions that claim to control change. We need institutions that can trace change honestly. That means accepting complexity without surrendering rigor, and pursuing ambition without hiding behind vague narratives. It means making peace with the fact that in the real world, influence is often partial, distributed, and messy.

That is not a weakness. It is the actual texture of change.

The future belongs to organizations that can answer a harder question than “Did it work?” They must answer, “How do we know, what else could explain it, and what does that teach us about the next step?” The moment we adopt that standard, evaluation stops being an afterthought. It becomes the engine of credible transformation.

And perhaps that is the most hopeful idea of all: the path to a better system is not only through better outcomes, but through better reasons for believing those outcomes are real.