Forty years of commitment, six years of sight
In the same quarter, three unrelated public decisions made the same move: they extended the horizon of what the State promises without extending the horizon of what the State oversees. This is not bad faith or improvisation. It is a pattern, and for organizations that depend on public infrastructure it changes the risk calculation.
On 9 April 2026 Mexico published the Strategic Infrastructure Investment Act. Article 120 authorizes contracts of between 4 and 40 years for private capital to finance, build and operate public infrastructure. The same act amends Article 32 of the Federal Budget Act so that these multi-year commitments are reported in the federal budget for only five years beyond the current fiscal year: six in total.
Forty years of obligation. Six years of reporting.
In parallel, in Mexico City, a mechanism created in July 2025 as an administrative agreement — revocable in a day — to accelerate housing permits became, within twelve months, a bill amending three separate laws. Committees approved it on 3 August 2026. That same mechanism cut the structural review from 225 calendar days to 15.
And in the north, while nearshoring investment grew 39% year over year, industrial vacancy multiplied sevenfold and 91% of firms in industrial parks reported difficulty securing power. The institutional response was a standing committee, and a 400 kV transmission line built to serve a single data center.
What pushes is real and not cynical: attracting private capital to strategic infrastructure requires offering decades of certainty, and responding to a housing crisis requires permits that no longer take two years. None of the three decisions is arbitrary.
What resists is the architecture: a six-year term with no re-election, where no administration owes anything to the one that signed before it. Mexico has already lived this scenario — Texcoco, 2018 — and the new law does not fully shield against it.
The common move is subtler than "the State overpromises." It is a temporal mismatch: the duration of the commitment grew, the duration of the oversight did not. In infrastructure, the reporting window stayed at six years. In housing, the weight of technical review was reduced one institutional step at a time, each defensible on its own. In energy, coordination is organized project by project instead of building surplus capacity.
Whoever structures the deal today wins. Whoever has to enforce it in 2035 pays.
How much of what your organization committed to for the long term depends on being honored by someone who did not sign it? And the follow-up, which is the uncomfortable one: do you have any mechanism that outlives the person who signed, or does your continuity rest on that person still being there?
For an organization that contracts with the State or depends on public infrastructure, the relevant risk is not cancellation — that is rare and litigable. It is the visibility window. A forty-year commitment reported over six years is legally real and fiscally invisible from year seven onward. That does not show up in a risk analysis that only reads the termination clause.
Where this becomes a decision available today is energy. Power availability is already deciding industrial location ahead of labor cost and ahead of logistics, and institutional coordination proceeds project by project. Securing firm supply is a decision that gets more expensive the longer you wait, and one that can still be made today.
The corollary for anyone planning ten years out: certified self-generated power has stopped being a sustainability differentiator and is becoming a condition of entry.
In your own long-term commitments, separate two categories that are probably mixed together today: what is protected by contract, and what is protected only by the continuity of whoever signed it.
Give the second category a review date. Not out of distrust, but because a commitment that lives only inside a personal relationship is not an organizational commitment: it is an agreement between two people who will not always be in those chairs.
That separation can be done in one session. What it opens afterward — which instrument sustains the unprotected category — is design work.
The signal that doesn't fit the reading. It's published anyway: a note that only shows what confirms it isn't a reading, it's an argument.
- Strategic Infrastructure Investment for Development with Wellbeing ActCámara de Diputados · Diario Oficial de la Federación · Apr 9, 2026
- Mexico City Congress advances Norm 26 amendment to ease social housing constructionCongreso de la Ciudad de México · Aug 3, 2026
- CFE Bulletin 123/26: Standing Committee on Electric Service with the Government of QuerétaroComisión Federal de Electricidad · Jul 14, 2026
The three core facts were checked against primary sources: the full text of the federal act on the Chamber of Deputies site (Articles 120 and 32 read in full), the official communiqué from Mexico City's Congress published the same day as the committee vote, and the bulletin downloaded directly from CFE.
What is NOT primary data, and is flagged as such: industrial vacancy figures (Prodensa via Industry & Energy Magazine) and nearshoring FDI growth (BBVA Análisis) are specialized secondary sources. A figure on cancelled projects in Querétaro and Guanajuato cited by Comce could not be independently verified and was excluded from this reading.
The connection between the three facts is our interpretation, not a link declared by any source. None of the three decisions cites the other two.
This reading weakens if the oversight mechanism proves sufficient. The act does create a Strategic Planning Council chaired by the Executive and made up of eight agencies: it is not that oversight does not exist, it is that its reporting horizon is far shorter than the commitment's. If that Council is installed within the 120 days the law specifies and operates normally, and if the Finance Ministry issues the regulation within 180 days, the argument loses force.
Mexico also has more than a decade of multi-year schemes — PPPs since 2012, concessioned highways — that survived presidential transitions without mass cancellation. Texcoco is the documented exception, not the rule, and building a reading on the exception is exactly the error this archive tries not to make.
And there is a signal in our own archive that pushes back: the May 2026 voice-cloning reform shows the same State producing pioneering, fast technical regulation when an organized guild demands it. Institutional capacity is not uniformly thinned; it is unevenly distributed according to who pushes. If a case appears in 2027 where long-term oversight was strengthened at the same pace as the commitment, this reading will be corrected and recorded.
Written before the counter-evidence appears.