Smoke, Dust, and Two Logics of Governance

Ethiopia is not short on vision. From sectoral strategies to national plans, public institutions are awash in documents promising transformation: universal health care, green industrialization, and clean cooking for all by 2030. The stakes of the latter are acute. More than 60,000 premature deaths annually link to household air pollution. Women and girls bear the brunt of fuel collection; forests continue to degrade. In response, the Ministry of Water and Energy’s (MoWIE) 2024 National Clean Cooking Energy Roadmap sets an ambitious target: universal access by 2030. It rightly recognizes that incremental change is insufficient—demanding transformation of markets, technologies, and institutions is required. The question is not whether these goals matter, but whether the systems pursuing them can deliver.

In the previous piece Ethiopian Airlines Proves Bureaucracy Works—which challenged the assumption that institutional failures stem from scarce resources or training—we argued that structures embedding feedback, accountability, and incentives enable bureaucracies to excel. Here, we test that claim through pointed comparison.

We examine MoWIE’s roadmap not as an outlier, but as emblematic of a broader pattern: planning divorced from adaptation. The flaws we trace—overconfidence in blueprints, accountability without consequence, and resistance to experimentation—recur in the Health Sector Transformation Plan, General Education Quality Improvement Program, and the Ministry of Transport’s National Logistics Strategy, among others. Our focus on clean cooking reflects the author’s familiarity with the sector, not its uniqueness.

Here we critique systems, not MoWIE’s commitment. The roadmap’s ambition is urgent and commendable, authored by dedicated professionals. Yet institutional failure is often structural. Roadmaps prioritize legibility—offering coordination at the cost of local variation, feedback, and adaptation. Even well-intentioned ministries become trapped by their tools.

The comparison to Ethiopian Airlines’ Vision 2035 is deliberate. Governed by a distinct logic, EAL starts from feedback loops—not fixed plans. It treats failures as data, builds mechanisms for correction, and rewards adaptation over compliance. In short: MoWIE reflects a planner’s mindset (sweeping goals, vague metrics, execution divorced from reality); EAL operates as a searcher (iterative, performance-driven, grounded in operational learning).

Holding these systems side by side, we examine not sectors but structures: What makes one brittle, the other adaptive? Why does one translate ambition into accountability, while the other dissolves it into abstraction?

What follows is a focused analysis of three interlinked dynamics: the blueprint trap, the accountability void, and the learning void. The first two represent chronic failures in system design and delivery; the third points to what must replace them. By tracing these dynamics in clean cooking, we surface structural insights applicable across policy domains. The target is not the roadmap, but the planning logic it embodies.

Flaw 1: The Blueprint Trap — Repeating History, Ignoring People

The MoWIE roadmap opens with moral clarity: millions rely on wood, charcoal, and dung— damaging health, forests, and women’s time. Its solution is bold: 35 million households must adopt clean stoves by 2030. But beneath the ambition lies a design flaw: the roadmap is structured for delivery, not discovery. Its logic is linear—diagnose, pick technologies, forecast adoption, scale. Feedback, friction, and failure are treated as afterthoughts. Deviation is a glitch, not a guide.

The result is a planning model blind to the very realities that have derailed past efforts. Biogas systems, for instance, have consistently failed in predictable ways: households in Hadiya Zone lacked enough cattle to sustain them; Halaba lost 62 percent of eligible users due to seasonal variation; Basona Werana had over 60 percent of households falling short of the minimum livestock threshold; Limmu Kossa faced high failure rates from water scarcity, with households walking 15–20 minutes to refill digesters.

These are not anomalies—they are structural signals. Assessments by SNV and others have shown that poor maintenance infrastructure, lack of user training, affordability gaps, and behavioral mismatches routinely undermine clean cooking adoption. Yet the roadmap abstracts adoption into binary metrics, insensitive to mixed-use realities—such as households that boil tea with electricity but still bake injera over firewood. It makes no mention of past distribution missteps, including electric stoves delivered in off-grid areas like Tigray and Afar, many of which were resold for charcoal. By omitting these frictions, the plan hardwires failure into its architecture. When households resist or adapt differently, the roadmap has no way to register that signal. Resistance appears as user failure—not planner error. The system cannot distinguish—it wasn’t built to.

Ethiopian Airlines’ Vision 2035 is as ambitious as MoWIE’s roadmap—perhaps more so. It targets over 270 aircraft, 150 international destinations, 65 million annual passengers, and 1 million tons of cargo. But unlike MoWIE, these are not fixed endpoints—they are hypotheses under continuous refinement. Each goal is scaffolded by systems that learn: new routes begin as provisional trials; fleet expansion is paced by maintenance data and demand shifts; cargo growth builds on pilot initiatives like temperature-controlled freight. Vision 2035 does not treat planning as orchestration but as exploration. Feedback is not a disruption—it is design. Learning isn’t a contingency—it is the infrastructure.

EAL embeds iteration structurally. Its performance is tracked through dense operational data flows—from route profitability to fleet reliability—reviewed weekly by cross-functional teams. Proposals are field-tested before being scaled. Performance shortfalls trigger design reviews, not blame. Even its passenger experience and crew workflows are shaped through continuous refinement loops, informed by staff feedback, frontline conditions, and customer data. Vision 2035 is a bet on disciplined trial-and-error.

Critically, EAL accepts friction as the price of success. It doesn’t plan around ideal conditions but designs around constraints: crew availability, fuel shocks, political disruptions, airport logistics. Where MoWIE erases variance, EAL builds for it. One treats complexity as noise to eliminate; the other treats it as a signal to amplify. One plans to build graveyards of unused stoves; the other builds adaptive fleets.

The deeper cost: a design blind to friction breeds a bureaucracy blind to failure. Abstraction becomes impunity—no one answers when things collapse.

Flaw 2: The Accountability Void — Counting Inputs, Not Impact

MoWIE’s blindness to failure is structural. The roadmap lacks scaffolding to track outcomes, assign responsibility, or enable correction. Success is measured by inputs: stoves delivered, policies launched, workshops held. What happens after handoff—whether stoves are used, sustained, or deliver benefit—lies outside its vision.

It proposes no usage audits, verification triggers, or accountability for abandonment. Past patterns predict impact: Sidama’s 30 percent biogas failure prompted no consequences; Afar’s stove resale met silence, not redesign. The roadmap assumes adoption reflects willingness, not design fit. Problems are framed as downstream—”community reluctance” or “market immaturity”—never upstream error.

Incentives entrench this. Promotions reward donor alignment, not outcomes. Reports track completion, not correcton. Staff are asked if interventions were delivered, not if they worked. It’s a system optimized for plausible deniability: outputs substitute for impact and paperwork displaces feedback.

Ethiopian Airlines structures accountability into its core operations. Vision 2035 explicitly links strategic goals to measurable performance indicators—on-time arrivals, safety records, customer satisfaction, and employee productivity. Route managers are not judged by effort but by impact: profitability, adaptability to shocks, and responsiveness to real-time data. Promotion, resource allocation, and retention are tied to these outcomes—not to seniority or checklist completion.

This is operationalized through an enterprise-wide Performance Management System where managers are held accountable for measurable outcomes—not inputs. Failures trigger correction, not concealment. When COVID collapsed passenger demand, EAL didn’t just weather the storm— it reassigned operational responsibility, restructured incentives, and repurposed the fleet for cargo. The pivot wasn’t ad hoc; it was tracked, measured, and reinforced—producing USD 2.3 billion in freight revenue by 2023. Similarly, the costly outsourcing of maintenance didn’t just get flagged—it triggered a strategy shift, with leadership accountable for building an in-house MRO (Maintenance, Repair, and Overhaul) unit. That system now serves external clients like Lufthansa and RwandAir, generating USD 60 million annually. These are not just success stories—they’re accountability in motion: failure → data → redesign → outcome.

Until clean cooking institutions link responsibility to results—financially, institutionally, professionally—they will mistake motion for progress. At EAL, accountability functions as a pressure valve: it converts institutional challenge into energy—not punishment.

Flaw 3: The Learning Void — When Certainty Silences Evolution

MoWIE’s roadmap suffers not just from rigid design or weak accountability—but from a deeper malfunction: an institutional hostility to uncertainty. If poor planning blinds the system (Flaw 1) and poor accountability paralyzes it (Flaw 2), the absence of adaptive learning ensures it stays broken.

MoWIE treats adoption as guaranteed and scale as a given. Devices are rolled out nationally without site-specific trials. Failures are framed as user problems, never as flaws in design, fit, or strategy. No space is reserved for pilot testing, phase-ins, or iterative redesign. This is not just a missed opportunity—it is deliberate policy.

The result: error is replicated. Hadiya’s failed biogas systems in 2021 reappear in Halaba in 2023— same model, same flaws, no redesign. Grassroots innovators are locked out: local stove artisans can’t access funds to refine or test their own models. Inflation doubled stove prices post-launch, but the roadmap had no provision to recalibrate or adapt. This isn’t oversight—it’s design. Institutions that can’t question themselves build failure into the blueprint.

Ethiopian Airlines runs on a different logic: it assumes uncertainty and builds systems to absorb it. Vision 2035 treats change—not certainty—as the operating environment. Performance data drives decisions; low-performing routes are trimmed, rerouted, or redesigned. New routes launch as time-bound pilots. Shocks become inputs, not excuses. When a fleet surplus emerged in 2018 due to accelerated Dreamliner deliveries, EAL responded by launching the “Africa by Demand” charter service, leasing idle aircraft to African governments and UN agencies for elections and peacekeeping. It also preemptively converted four passenger 777s into cargo-in-cabin aircraft— anticipating the rise of e-commerce logistics. These moves generated USD 140 million in ancillary revenue even before COVID, and, critically, positioned EAL to dominate African cargo when the pandemic arrived. This wasn’t improvisation—it was institutionalized flexibility. What others saw as overcapacity, EAL treated as a proving ground.

This isn’t “learning from failure.” It’s learning before failure ossifies. Uncertainty becomes productive when institutions plan for it, fund it, and reward it.

Ethiopia’s clean cooking sector will not improve by refining static plans. It needs institutional redesign that treats doubt as fuel. Pilot tests must be reinstated as legitimate policy tools, not signs of indecision. A meaningful share of program budgets—say, 30 percent—should be set aside for structured experimentation across agro-ecological zones. Promotions must be tied not to distribution counts but to sustained adoption verified in the field. And failures, like the resale of stoves in Afar, must serve as immediate triggers for design revision—not as afterthoughts buried in reports.

MoWIE’s triad of flaws forms a closed loop: it can’t see reality, won’t own failure, and refuses to look. EAL breaks the loop by making doubt its design partner. Until uncertainty is formalized into how policy is made and revised, MoWIE will continue mistaking action for progress. EAL’s true innovation isn’t aviation—it’s its ability to say “we don’t know yet” and build from there. That humility is Ethiopia’s most underutilized resource.

Clockwork vs. Ecosystem: Why Institutional Logic Determines Destiny

The roadmap’s failure lies not in its ambition, but in a fatal misalignment: its institutional logic is fundamentally mismatched to the living complexity it claims to address. Rigid design, diffuse accountability, and suppressed learning are not isolated flaws—they are interconnected symptoms of an operating system engineered for a world that does not exist. To grasp why MoWIE perpetuates failure while EAL transcends it, we must move beyond surface metrics and confront the deeper paradigms governing how institutions engage with chaos.

MoWIE embodies clockwork logic—a worldview inherited from the high-modernist playbook. It treats society as a static machine: complexity is an obstacle to be eliminated through central control; knowledge is deemed complete at the planning stage; outcomes are engineered via budgets, coordination, and compliance. In this paradigm, deviation signals error, not information. Citizens become passive recipients of technical fixes; human behavior is reduced to input-output equations. The roadmap’s silence on past failures (Afar’s resold stoves, Sidama’s biogas collapse) is not oversight—it is ideology. When reality intrudes, the system defaults to blame or silence.

EAL operates by ecosystem logic—a humility toward uncertainty forged in competitive markets. Here, society is a dynamic web: complexity is navigated, not suppressed; knowledge is dispersed and emergent; outcomes evolve through disciplined adaptation. Lomé’s cargo hub emerged not from a decree, but from probing West African freight demand; COVID’s devastation became a laboratory for reinvention. Deviation is metabolized as data. This is not improvisation, but structured respect for uncertainty: provisional actions, continuous sensing, and recursive learning.

These paradigms forge divergent realities:

Clockwork logic (MoWIE) builds for predictability but shatters against friction. Informal markets, gendered labor dynamics, and seasonal income swings—all erased in blueprints— resurface as systemic shocks. Without feedback loops, failures recycle endlessly: Halaba repeats Hadiya’s biogas mistakes; stove distributions ignore inflation’s toll on affordability. The system’s response? Blame users (“resistance”) or retreat into reports. Ecosystem logic (EAL) builds for adaptation and thrives on friction. Market volatility becomes cargo opportunities; maintenance gaps spawn USD 60 million MRO revenue; route failures inform network redesign. Accountability is wired into practice: managers own outcomes, not activities; data triggers redesign, not punishment.

Conclusion – When Adaptation Is the Policy

Ethiopia’s clean cooking crisis lays bare an uncomfortable truth: ambition without adaptive logic is performance, not progress. The divide between MoWIE’s roadmap and Ethiopian Airlines’ trajectory stems not from resources or commitment, but from their opposing relationships to uncertainty.

EAL treats planning as hypothesis—tentative, friction-aware, continuously updated. Vision 2035 is not a script but a system for learning: cargo pivots during crisis, trial routes tested with leased aircraft, maintenance systems built from gaps—not despite uncertainty, but through it.

MoWIE treats planning as fate. Its roadmap is static, blind to friction, insulated from feedback. Biogas failures repeat, stoves are resold, mixed-fuel use persists—not because these are invisible, but because the system is designed not to see them. Deviation is error, not intelligence.

Clean cooking demands more than better technologies—it demands institutions fluent in complexity. Will families bake injera on electric stoves? Can policy navigate the informal ecosystems surrounding LPG vendors, charcoal traders, and firewood distributors? Who actually benefits when fuel collection time drops—women, children, or the labor market? What happens when inflation doubles stove prices after rollout? These aren’t “implementation challenges” to be ironed out—they are the terrain. Institutions that reduce them to cells in a spreadsheet fail by design. Institutions that observe, adapt, and evolve—like Ethiopian Airlines build durable, context- aware solutions.

Blueprints don’t cook meals. Adaptive institutions do. Until Ethiopia retools its public sector to learn with the same intensity it plans, clean cooking will remain a well-funded aspiration, smothered by the smoke of its own certainty. Visions like “universal access by 2030” will not inspire—they will indict.

Tsegaye Nega is a professor emeritus at Carleton College in the United States and founder and CEO of Anega Energies Manufacturing.

Contributed by Tsegaye Nega