Transformation theory has long operated under a success bias—an implicit assumption that the transformations we label as successes are the ones that ultimately matter. Development institutions measure their portfolios by success rates. Change strategists study successful cases for replicable models. The entire edifice of transformation planning is oriented toward producing outcomes that can be classified, at the moment of completion, as victories.

But what if this framing systematically distorts our understanding of how societies actually change? A growing body of comparative evidence suggests something deeply counterintuitive: transformations that fail by their own stated criteria frequently generate more durable, more equitable, and more adaptive institutional outcomes than those celebrated as successes. The failed democratic transitions that seeded later democratic consolidation. The collapsed economic reforms that inoculated societies against future shock therapy. The aborted revolutions that shifted the Overton window permanently.

This is not a contrarian argument for the sake of provocation. It is a structural claim about the relationship between immediate legibility and long-term systemic impact. When we evaluate transformation by whether it achieved what it set out to achieve, within the timeframe it set for itself, we are applying criteria that may be orthogonal—or even inversely related—to the actual capacity-building, institutional learning, and social reorganization that constitute genuine systemic change. The implications for how we design, fund, and evaluate transformation efforts are profound.

The core analytical problem is what we might call success-outcome divergence—the systematic gap between meeting transformation criteria and producing beneficial long-term outcomes. This divergence arises because the criteria by which transformations are judged successful are typically defined at the outset, shaped by the political economy of the moment, and biased toward measurable, short-term indicators. They capture legibility, not depth.

Consider the structural adjustment programs of the 1980s and 1990s. By the metrics established at inception—fiscal deficit reduction, trade liberalization ratios, privatization completion rates—many were classified as successes. Yet the societies that most faithfully implemented these programs often experienced institutional hollowing, loss of state capacity, and the destruction of social infrastructure that took decades to rebuild. Meanwhile, countries that failed to fully implement adjustment—that resisted, delayed, or selectively adopted reforms—frequently preserved institutional capabilities that proved essential for later, more sustainable transformation.

This pattern recurs across domains. Rapid democratization that meets all procedural benchmarks can produce what Fareed Zakaria termed 'illiberal democracy'—formal success masking substantive failure. Revolutionary movements that seize power but cannot govern often impose more damage than those that fall short of power but permanently alter the political landscape. The success label, once applied, also generates its own distortions: it discourages critical examination, locks in potentially maladaptive institutional arrangements, and creates path dependencies that constrain future adaptation.

The divergence is not random. It is structurally predictable. Transformations optimized for meeting success criteria tend to prioritize speed, visible benchmarks, and the satisfaction of external evaluators over the slower processes of institutional embedding, social learning, and capability development that Amartya Sen's framework identifies as the real substance of development. Success, paradoxically, can truncate the very processes it claims to advance.

What makes this particularly insidious is that success narratives become self-reinforcing. Donor communities, political leaders, and academic institutions all have incentives to maintain the success classification once assigned. The result is an evaluative ecosystem that systematically overweights nominal successes and underweights the generative contributions of nominal failures—creating a distorted knowledge base for future transformation strategy.

Takeaway

The criteria we use to judge transformation success are often designed for short-term legibility rather than long-term impact. When success is defined by the conditions of the present, it may systematically select against the deeper institutional changes that matter most for the future.

If success-outcome divergence explains why successful transformations can produce poor results, we need a complementary framework for understanding why failures produce good ones. The answer lies in what I term generative failure mechanics—the specific pathways through which incomplete, aborted, or defeated transformation attempts contribute to later systemic change.

The first mechanism is institutional inoculation. Failed transformations expose societies to the stresses and contradictions of change without locking them into a single institutional trajectory. This is analogous to how immune systems develop—partial exposure builds adaptive capacity. Russia's failed 1905 revolution, for instance, did not achieve its aims, but it created durable organizational networks, established precedents for collective action, and revealed the structural vulnerabilities of the autocratic system in ways that shaped everything that followed. The failure was, in systems terms, an information-rich event that increased the society's capacity for future transformation.

The second mechanism is capability preservation through incompleteness. When transformations fail to reach their stated endpoints, they often leave intact institutional structures that would have been dismantled by success. This is Sen's insight applied dynamically: capabilities that survive a failed transformation become resources for future change. East Asian developmental states that resisted full liberalization—failing by Washington Consensus standards—preserved state capacities that enabled their subsequent heterodox development strategies.

The third mechanism is narrative reframing. Failed transformations generate counter-narratives that challenge dominant paradigms and expand the range of thinkable alternatives. They become reference points for future movements—cautionary tales, inspirational fragments, unfinished projects that invite completion. The Prague Spring failed militarily but succeeded epistemically, permanently embedding the possibility of 'socialism with a human face' into political imagination across the Eastern Bloc.

These mechanisms operate on different timescales and interact in complex ways. A single failed transformation may simultaneously inoculate institutions against shock, preserve capabilities by preventing their premature restructuring, and seed narratives that mature over decades. The compounding effect means that the cumulative impact of a sequence of failures can exceed the impact of any single success—a finding that radically challenges linear models of transformation progress.

Takeaway

Failed transformations are not empty events. They inoculate institutions, preserve capabilities that success would have dismantled, and seed narratives that reshape what future societies believe is possible. Progress often moves through a sequence of productive failures rather than a single decisive victory.

If the conventional success-failure binary distorts our understanding of transformation, we need alternative evaluative frameworks that capture what actually matters. This requires moving from outcome-centered to impact-centered evaluation—a shift that changes not just how we judge past transformations but how we design future ones.

An impact-centered framework evaluates transformation attempts along three dimensions that conventional success criteria typically ignore. First, capability expansion: did the transformation attempt, regardless of its immediate outcome, expand the substantive freedoms and institutional capacities available to the affected population? This draws directly on Sen's capability approach but applies it to the transformation process itself rather than just its endpoints. A reform that fails but builds organizational capacity, civic skills, and institutional knowledge has expanded capabilities even in defeat.

Second, systemic learning: did the transformation generate knowledge—about structural constraints, about intervention dynamics, about social responses—that improves the quality of future transformation attempts? Failed transformations are often richer sources of systemic learning than successful ones, precisely because failure forces deeper interrogation of assumptions. Success, by contrast, can produce false confidence and spurious causal narratives.

Third, adaptive resilience: did the transformation attempt leave the system more or less capable of responding to future pressures and opportunities? This is the most counterintuitive dimension, because transformations that succeed in imposing a new order can actually reduce adaptive resilience by eliminating institutional diversity, suppressing alternative organizational forms, and creating rigid path dependencies. Incomplete transformations, by contrast, tend to increase systemic heterogeneity—the very condition that enables future adaptation.

Adopting these criteria has radical implications for transformation strategy. It suggests that the optimal approach is often not to pursue singular, decisive transformation but to cultivate portfolios of transformation attempts—some of which are expected to fail in conventional terms—that collectively expand capabilities, generate learning, and build adaptive resilience. This is not an argument against ambition. It is an argument for a more sophisticated understanding of how ambition translates into lasting change across the complex temporalities of social systems.

Takeaway

Evaluating transformation by whether it achieved its stated goals is like judging a seed by whether it became the tree you planned. The better question is whether it changed the soil—expanding capabilities, generating learning, and increasing a society's capacity to transform itself in the future.

The success bias in transformation theory is not merely an academic problem—it actively shapes how billions of dollars in development funding are allocated, how political movements design their strategies, and how societies understand their own histories. By systematically privileging legible, short-term success over the deeper mechanics of systemic change, we have built an evaluative infrastructure that is, in important respects, inversely calibrated to what matters.

Recognizing the generative power of failure does not mean celebrating it or engineering it. It means building evaluative frameworks sophisticated enough to capture the full temporality of social transformation—the slow sedimentation of capabilities, the long maturation of institutional learning, the decades-long arcs through which defeated movements reshape the possible.

The most important transformations in human history were rarely recognized as such when they occurred. Many were classified as failures. The question for transformation strategists is whether we can learn to see impact where our current frameworks see only defeat—and to design for resilience rather than for victory.