Four plans, one trajectory

Pemex is the most thoroughly documented case of a failure mode that recurs across state-owned enterprises in democratic political systems. Administration after administration launches a credible reform program, international observers take each one seriously, the financial metrics improve for a stretch, and the company’s operational trajectory comes out the far side unchanged. The pattern has held for decades. Each time, the financial press and the political commentary produce a framework for why that particular plan failed, and each explanation stops short of the pattern itself. No account of a single plan can reach it.

The numbers set the terms of the puzzle. Across four presidential administrations, Mexico has launched four Pemex restructuring plans, announced four sets of production and financial targets, and run four rounds of organizational reform. Over the same stretch, crude production has fallen roughly twenty-eight percent in a decade, the 2024 loss came in around thirty billion dollars, and the debt passed one hundred billion at its peak before a state-backed refinancing across 2025 brought it down to an eleven-year low. The production line did not notice. Each plan before the current one set out to reverse these trends, and each was credible enough at launch to draw serious treatment from energy analysts and ratings agencies. None changed the trajectory in any way that survived the political cycle that produced it.

The repetition needs explaining more than the decline does. Resource depletion, aging infrastructure, and capital misallocation are well documented, and they would drag production down under any governance arrangement. What resists ordinary explanation is that four administrations of different political orientations, each calibrating its plan to the operational and financial conditions in front of it, produced outcomes the trajectory data cannot tell apart from each other, or from the path the company would have followed with no plan at all. Had a single plan failed, execution would be the answer, and the next administration would design something more deliverable. When four consecutive plans fail to move a trajectory, the answer is structural, and a structural answer has to explain why every plan, whatever its design, lands in the same place.

The four administrations span two decades. Felipe Calderon opened the regulatory liberalization debate between 2006 and 2012. Enrique Pena Nieto produced the 2013 to 2014 energy reform, which constitutionally reset Pemex’s relationship to private capital and to the Mexican state. Andres Manuel Lopez Obrador reversed key elements of that reform between 2018 and 2024 and recentralized authority over the company’s direction. Claudia Sheinbaum has run her own restructuring since late 2024, in continuity with the operational line her predecessor set. Four orientations, four reform philosophies, and a trajectory that closes each term looking close enough to how it opened that none of the four can claim to have altered it.

The distinction the financial narrative keeps missing, and the one the political narrative actively resists, is the difference between a bad plan and a structural condition that makes good plans unabsorbable. Neither narrative is positioned to name that condition; naming it implicates arrangements neither can address. Pemex needs the structural framing. Only that framing explains how transformation governance behaves inside a state-owned enterprise that answers to democratic politics while running technically demanding operations.

What the financial narrative cannot reach

The financial description of Pemex is accurate at the level it works on, and its accuracy is part of why it dominates the attention the company receives. The debt is real, the production decline is real, the losses are real, and the narrative reads them as the consequences of operational decisions, capital allocation, and policy priorities, which they are. The trouble is what the account leaves out. It describes the symptoms and never examines the decision architecture that produced them.

Pemex’s financial condition is the output of that architecture, so the question worth asking is who decides what the company does. The formal answer is the one any large corporation would give: the board decides, the chief executive executes. The operational answer is where the financial narrative stops. Presidential priorities override board deliberation on anything politically sensitive, and in a national symbol of Mexican economic sovereignty that covers most decisions that matter. The petroleum workers’ union, the STPRM, holds agreements that constrain staffing, job classification, and how labor moves across the company’s footprint. Appointment cycles tied to the sexenio turn over technical leadership before it can accumulate the operational knowledge oil and gas demand. And budget authority for capital deployment runs through the Finance Ministry, so the people who decide where capital goes apply fiscal criteria to operational questions. None of this is incidental. These are the structural features of how the company is governed, and together they generate the pattern the financial trajectory expresses.

Every plan that improved the metrics without touching this architecture improved the reported position and left the source of the decline alone. Financial engineering works on the expression of a structural misalignment, and the misalignment itself sits beyond its reach. The work is necessary, since the expression is a real cost someone has to manage. It is also insufficient. The condition underneath keeps operating, and it keeps producing next year’s number through the same mechanisms that produced last year’s. The 2025 debt reduction is the freshest illustration: an improvement in the reported position, engineered with sovereign support, that says nothing about whether the architecture that generated the debt has changed.

International observers have documented all of this across the same four administrations, accurately and with a consistent blind spot. Ratings actions tracked the deterioration, energy analyses forecast production within reasonable bands, and the balance-sheet reporting has been competent. What the observer community underweights is the structural condition, and the reason sits in the frame it works within. Corporate financial analysis and sovereign credit analysis are separate disciplines, and no standard analytical product assesses the alignment between them. Pemex’s condition lives precisely at that intersection, where neither instrument alone can see it. The documentation of symptoms is voluminous; the architectural diagnosis is missing.

Temporal arbitrage through the sexenio

Mexico’s six-year presidential term creates a specific timing problem for any state-owned enterprise under presidential influence, and the timing is what makes architectural redesign politically irrational for any single administration to attempt. An incoming president has every reason to announce transformation early: the announcement separates the new administration from the old one, signals capacity for change, and gives the base something visible to read as reform commitment. An outgoing president has every reason to declare progress at the end of the term whatever the underlying trajectory shows: the political cost of the eventual failure lands on the successor. The closing phase of a sexenio is when an administration harvests the return on having launched a transformation. The return on having finished one arrives on someone else’s watch.

The incentive to launch and the incentive to complete run on different clocks, and the gap between those clocks is where the arbitrage lives. A restructuring plan with a new organizational design, a fresh leadership team, and a set of production targets pays political returns within twelve to eighteen months, well inside the horizon the administration operates on. The redesign that would make any plan durable pays out well beyond it. That redesign would have to change the union’s role in operational decisions, the budget-authority relationship with the Finance Ministry, the appointment cycle that determines who holds technical leadership, and the constitutional and statutory provisions defining the company’s relationship to the state. It takes years to execute and delivers its operational results long after the president who authorized it has left office. A successor collects the return, often a successor of a different political orientation, while the administration that started the work pays its full political cost.

The rational response to that structure is to launch the visible plan and defer the redesign, and all four administrations did exactly that. The trajectory is the accumulated result of four consecutive deferrals, each producing a plan whose ambitions outran the architecture it had to operate inside, each followed by another deferral driven by the same incentive that produced the last. The arbitrage is rational for any single administration and expensive for the enterprise that absorbs all four.

The four constraints every plan works around

Pemex’s decision architecture holds four features that every reform has acknowledged and none has reached. The consistency of that acknowledgment without resolution deserves attention on its own. It marks the line between what conventional reform can address and what it cannot.

Presidential authority over operations runs through budget allocation, leadership appointments, and the political direction that overrides the formal independence management nominally holds. The authority is legitimate as much as it is real: Pemex is a national symbol as much as an energy company, and the power governing it reflects that weight. Its operational cost is just as concrete. Decisions that require presidential alignment cannot move at the speed a major oil and gas operation demands, and the lag compounds across the operating cycle in ways the political narrative never surfaces.

Union control over hiring and job classification predates the current misalignment by decades and has outlasted every restructuring cycle, including the cycles that set out to change it. The STPRM agreement makes operational flexibility contingent on labor negotiation, in a political context where aggressive labor negotiation costs any administration far more than it can expect to recover. So every plan flags the constraint as material, every plan manages it at the margin, and no plan reaches it. Reaching it would demand political capital the sponsoring administration does not have to spend.

Appointment cycles tied to the sexenio starve the company of the institutional knowledge technical operations require. Leadership that turns over with the political cycle cannot build the multi-year understanding complex fields and infrastructure depend on; that understanding comes from years inside the company’s specific assets, and transferable credentials do not substitute for it. Each administration’s appointments are defensible on their own terms, and the appointees are usually capable. The accumulated effect on institutional knowledge is the invisible cost the metrics never capture. No plan has designed around it, and the reason is a regress: designing around it would take an appointment cycle independent of the sexenio, which would take the political authority to constrain the sexenio’s appointment power, which is the one authority no administration is positioned to surrender.

Budget authority for capital deployment sits with the Finance Ministry, and that routing decides which considerations shape the investments that set production. The Ministry’s criteria are fiscal: legitimate concerns for any government, and the wrong instrument for choosing which fields to develop, which joint ventures to pursue, and which infrastructure to build. Capital tracks the federal fiscal cycle while the assets run on an operational cycle of their own, so deployment never matches what the company is actually trying to operate.

Sovereign alignment debt

Alignment debt shows up at Pemex in a form qualitatively different from the private-sector version, and the difference changes what kind of intervention could ever resolve it. Inside a corporation, alignment debt builds between formal governance and the operational reality of a single organization, and the corporation can in principle pay it down by redesigning its own internal governance: a hard exercise, but one it has the authority to run. At Pemex the debt sits among three governance logics that no single actor has the authority to redesign together. The enterprise logic requires operating on technical and commercial criteria to stay operationally adequate. The state logic requires responding to political direction and public policy: the state owns the company, and the company serves state purposes. The political logic rewards each administration for visible initiative over durable structural change, which is what the electoral system pays for.

A better plan cannot reconcile the three, as the tension lives outside any plan, in the relationship between an enterprise, its state owner, and the political system governing both. Every administration that restructured the debt, adjusted capital allocation, and launched a new production program managed the expression of sovereign alignment debt and left the debt itself intact. The trajectory across four administrations is the cumulative record of management without resolution.

The concept travels. It applies wherever the three logics stay in tension, wherever the political calendar runs against operational timescales, and wherever the instruments of financial and operational reform cannot reach the political architecture in charge. Pemex is the most visible current case. Petrobras carries the condition: the Lava Jato investigation exposed the governance pathologies that political integration with the state had produced, and successive reform attempts kept hitting the same architectural limit. Eskom carries it too, with a two-decade trajectory tracking the repeated-reform-without-redesign pattern the framework predicts. The condition surfaces in the national oil companies of several OPEC states as well, in institutionally different dress. Political systems, ownership structures, and operational characteristics vary in ways that change how the three logics interact, so the framework does not apply uniformly. The condition still recurs often enough to give the category diagnostic leverage.

Naming the condition does one useful thing: it moves the analytical question from why a specific plan failed to whether any plan, under the prevailing political conditions, could succeed. The first question produces the plan-failure-plan cycle that has defined state enterprise reform in many countries. The second produces a harder conversation about whether the political conditions themselves can change, and what changing them would take. That conversation reaches architecture the people in the room often cannot challenge, which is why it rarely happens. It is also the conversation the condition requires.

Easy to specify, extraordinary to enact

The requirements for structural transformation at Pemex are easy to name and politically extraordinary to implement, and the gap between naming and implementing is what sovereign alignment debt looks like in operational terms. The requirements have to address the four features together. Each reinforces the others, and the untouched part reabsorbs any partial reform.

The partial reforms already run make the point. The 2013 to 2014 energy reform opened the constitutional possibility of private capital in the upstream sector, a change no previous reform had reached, while the operational governance of the company stayed largely as it was in the dimensions this piece describes. The 2018 reversal restored the centralization the constitutional reform had partly relaxed, which shows how durable the centralized architecture is even against constitutional change aimed at loosening it. The architecture absorbs reform in the dimensions the reform reaches and keeps operating in the dimensions it does not, and the trajectory reflects the untouched dimensions more than the changed ones.

Decision authority independent of the sexenio would take a board with genuine independence from the sitting administration, technical appointment processes that span presidential terms with explicit insulation from the cycle, and budget authority that keeps operational decisions below a defined threshold out of the Finance Ministry. Each element is easy to specify and demanding to enact: each transfers authority the political system currently holds to a structure it would not directly control, and that transfer is the concession no administration has been willing to make. Union governance that permits operational flexibility would take a renegotiation of the STPRM agreement that has defined the company’s constraints for decades; the union’s stake in the current arrangement is real, and the political cost of the renegotiation has exceeded every administration’s appetite. Capital allocation on operational criteria would take separating the investment decisions that set the production trajectory from the political decisions that currently govern them, so that the company makes those calls on technical and commercial grounds and the fiscal cycle governs what a fiscal cycle should.

Each change is politically costly and structurally necessary, and the space between that necessity and the conditions that would make it possible is the architecture of sovereign alignment debt. The question the current plan invites is whether it redesigns the architecture that made every previous plan unabsorbable, or stacks new ambitions on unchanged foundations. The medium term will answer, through the trajectory the next administration inherits and the next plan sets out to reverse.

The architecture beneath the plan

The most useful reading of any Pemex plan sets aside its production targets, investment program, and debt approach. Those three are the plan’s surface: they describe what the plan intends to deliver without describing the conditions under which delivery is possible. The useful reading asks what the plan does about the appointment cycles, the union constraints, the budget-authority routing, and the presidential horizon that rewards visible initiative over durable change.

A plan that addresses those conditions, even in part, is something different from its predecessors and earns analytical attention on its own terms. A plan that improves the metrics and the targets while leaving the architecture alone joins the sequence its predecessors defined, and the trajectory will tell the story in the next sexenio the way it has told it in each of the last four. The ambition of the targets is beside the point. The question is whether anyone has redesigned the architecture beneath them to make them deliverable, and the plan’s specific choices about the four features answer it empirically.

Sovereign alignment debt does not yield to better plans. It yields to a structural redesign of the governance relationship among enterprise, state, and political system, and that redesign is the work no administration has done. The evidence is the consistency itself: the particular failures vary from plan to plan, and the trajectory does not. The architecture is the constant no one has touched. Until someone touches it, the trajectory already tells you what the current plan delivers.