The Economics of Interruption
When infrastructure fails, costs multiply. Kristin set off the shock; the bill built up over a succession of low-pressure systems.
The first thing to say about Storm Kristin and the two weeks of storms that followed is not economic. It is human. There were fatalities and thousands of people spent days without basic services. During Kristin’s initial passage, Civil Protection reported five deaths and more than 8,000 incidents. Across the full “train of storms” that has affected the country since 28 January, the reported death toll rose to 16.
But that is precisely why economic analysis matters. When critical infrastructure fails, the cost is not measured in euros alone. It is measured in suffering, risk and lost trust. And economics, when practised well, is also a discipline of prevention: clearly bounded figures help reduce the likelihood that the next bill will once again be paid in lives.
What made this episode economically destructive was not the weather alone. IPMA recorded a peak gust of 48.9 metres per second (175.9 km/h) at Monte Real Air Base and says higher values may have occurred elsewhere, given the interaction between airflow and terrain. In public comments cited by the press, REN referred to “winds of around 200 km/h” during Kristin. What amplified the shock was the cascading failure (of energy, mobility and productive capacity) and, above all, its persistence.
In electricity distribution, the peak was immediate: at 6 am on 28 January, E-REDES recorded around one million customers without supply. That figure alone is consistent with a severe but recoverable incident. What turned the event into a prolonged economic disruption was the time taken to restore service: on 10 February, 35,000 customers in the worst-affected areas were still without electricity, including 26,000 in Leiria.
When an interruption persists, it stops being a “blackout” and becomes a “regime”. Companies do not lose hours; they lose production cycles, contractual deadlines, inventory and reputation. The marginal cost of the first day is high. The marginal cost of the tenth day is qualitatively different: it opens the way to cascading breakdowns, supplier replacement and investment decisions that do not return to their starting point when electricity comes back.
REN reported 774 kilometres of lines out of service, 61 pylons toppled and others damaged. Restoration would take weeks. Despite the damage, it distinguished this from supply interruptions attributable to its network, limited to the Zêzere substation area. Transmission-grid damage should not be confused with local distribution failures.
Transport amplified the shock in a country where traffic flows are concentrated along a few routes. When simultaneous failures compress redundancy, logistics do not merely slow down; they become unpredictable. And unpredictability is an invisible tax on an exporting economy: contractual penalties, urgent shipments, wasted production and, ultimately, lost market share when confidence breaks down.
That risk has already emerged on the ground. Companies in the Central region reported limited capacity, delays and uncertainty over future orders, alongside public references to investment being reassessed when physical damage and interruptions persist.
The cost, in three layers (with explicit boundaries)
There are three layers that must not be confused.
First, physical damage and the reconstruction effort. On 6 February, the Government estimated that losses caused by severe weather already exceeded €4 billion. That estimate should not be confused with a measure of lost GDP.
Second, the economic activity lost during the interruption. This includes production stoppages, energy not supplied, logistics disruption and extraordinary operating costs. The crucial point is that direct “energy costs” are only part of the problem: in the real economy, losses quickly shift to undelivered output and failed contracts.
Third, second-order effects. This is the most insidious channel and the least measurable in the short term: customers diversifying their suppliers, higher implicit risk premiums, deferred investment and relocation being considered. What is lost here is not one day of GDP; it is predictability. Predictability is a macroeconomic asset: the inverse of risk and uncertainty.
Reconstruction is not prosperity
Portugal cannot accept one trap: confusing reconstruction with growth. Reconstruction adds to GDP, but it is defensive GDP. It replaces destroyed capital; it does not increase productive capacity. And it occupies capital, teams and time that could otherwise be used to raise the country’s productivity and competitiveness.
The problem is not out in the field. It is in the design.
The exposed fragility cannot be explained by operational execution alone. It is also a matter of incentives. Regulatory systems tend to reward efficiency under normal conditions and treat resilience as an externality, because “good weather” is measured more frequently and carries greater weight in performance indicators. The result is predictable: networks that are efficient day to day but fragile under stress.
As climate change increases the frequency and intensity of extreme events, “stress” stops being an exception and begins to approach a new normal. And once it is no longer exceptional, resilience stops being a cost and becomes an investment.
Three priority reforms (principles, not rhetoric)
Portugal faces a choice: restore things as they were, or make resilience a design principle. The second path costs more immediately, but is economically superior over the life cycle.
First, selective hardening along critical corridors. The serious question is not whether to “bury everything” (“all the cables”), but where paying a resilience premium makes sense, based on risk and the economic cost of each hour of failure: industrial areas, logistics hubs and essential services.
Second, operational continuity in critical clusters. Create the technical and regulatory conditions for islanded operation where the cost of interruption is disproportionate: distributed generation, storage and intelligent switching. The aim is not permanent self-sufficiency; it is continuity under stress.
Third, service-level agreements and penalties proportionate to economic cost, not merely to the number of customers affected. Set different maximum restoration times according to criticality and align incentives with the economic value destroyed per hour of interruption. Without this, the system tends to optimise the easiest indicator, not the most relevant cost.
In the end, the question this cycle of storms leaves us is not only “How much did it cost?” It is: how much does it cost, year after year, to live without redundancy until the day the climate sends the bill? And how many more people must spend days without basic services before we accept that resilience is not an engineering luxury, but a condition of dignity and competitiveness for a country?
Portugal can choose to return to normal. Or it can choose to build differently. The first option is cheaper this year. The second is cheaper this decade.
Original bibliographic details (Portuguese)
Ensaio de Miguel Sousa Soares, publicado originalmente no LinkedIn a 12 de fevereiro de 2026. Esta página é o arquivo pessoal do autor.
Miguel Sousa Soares is an economist and analyst of geoeconomics and artificial intelligence, with more than twenty years’ experience in strategy and corporate finance. He writes for Visão, Jornal de Notícias, Expresso and Correio da Manhã. He comments on artificial intelligence on NOW on Fridays.