The ratchet below the street

How infrastructure inequality compounds without anyone deciding it should

The Shades generator fails on a Tuesday. The repair crew arrives on Friday. They are not slow; they have four jobs ahead of this one, all of them in districts whose aldermen have already sent letters. By Friday the fault has been running for three days, which means the distribution panel connected to the generator has been cycling under abnormal load for three days, which means the wiring in the distribution panel has experienced thermal stress it would not have experienced if the repair had arrived on Tuesday. The crew fixes the generator. They do not inspect the distribution panel: they are three days behind schedule and the panel is not on the work order. The panel is noted as “monitor” in the visit log and left.

The Nap Hill generator fails on a Thursday. The repair crew arrives on Friday morning, before the Merchant Quarter has noticed anything is wrong. The fault has been running for eighteen hours. The distribution panel has experienced no abnormal load. The crew fixes the generator and, while they are on site, inspects the panel as routine. They find a minor wear issue and replace the relevant component. Total time on site: four hours. The panel is logged as “serviced.”

The generators are the same model, installed the same year. In three years’ time, the Shades distribution panel will fail. The Nap Hill distribution panel was serviced last Friday.

No individual in this chain has decided that the Shades should have worse infrastructure than Nap Hill. The repair crew prioritised by the order in which jobs arrived, which reflects the order in which complaints were made, which reflects the political urgency of the districts affected. The result is structurally identical to a deliberate decision to widen the infrastructure gap between districts. It just happened without one.

The moving parts

Most discussions of infrastructure inequality focus on investment: which districts receive more money for maintenance and improvement, and how those allocations reflect political power. The investment question is real and important. But it misses a mechanism that compounds the inequality independently of investment levels, and that operates through something much more granular: the speed of repair.

Infrastructure that fails and is repaired quickly is in a different condition after the repair from infrastructure that fails and is repaired slowly. This is true in three reinforcing ways.

The first is load transfer. When infrastructure fails and the failure is not immediately addressed, connected systems absorb the load it would have carried. A water pump offline for three days puts more pressure on adjacent pumps than one offline for eight hours. Adjacent pumps that have run above design load for three days are closer to their own failure threshold than pumps that ran above load for eight hours. The slow repair does not just leave the repaired component slightly worse. It leaves the surrounding network slightly worse too.

The second is repair quality under pressure. A crew arriving promptly, with adequate time allocated, does a thorough job: proper sealing, secondary inspection, routine servicing of adjacent components. A crew arriving late, already behind schedule, does what is necessary to restore service and moves on. The prompt repair removes the underlying cause. The late repair patches the symptom. Infrastructure that has been patched fails again sooner than infrastructure that has been properly repaired. When it fails again, it will receive another late repair.

The third is inspection frequency. Infrastructure repaired promptly is inspected at the time of the repair, which generates information about the condition of adjacent components. Infrastructure repaired slowly is inspected later, which means adjacent wear is identified later, which means adjacent failures are caught at a more advanced stage when they do occur. The pattern of quick response in maintained districts is also a pattern of higher inspection frequency, which is also a pattern of smaller and cheaper failures. The pattern of slow response in neglected districts is a pattern of lower inspection frequency, which is a pattern of larger and more expensive failures.

Through each of these mechanisms, every cycle of the repair queue leaves the infrastructure in neglected districts slightly worse, relative to maintained districts, than the cycle before. The gap is not a level. It is a direction, and it has no natural reversal.

The Ratchet

The mechanism is a ratchet because it clicks in one direction and does not naturally click back.

Widening the gap requires nothing active. It requires only that repair scheduling reflects existing complaint pressure, that complaint pressure reflects existing political influence, and that political influence reflects existing wealth and social organisation. These conditions are present by default in any city that has ever existed. The gap widens as a by-product of ordinary operations.

Closing the gap requires something active and explicitly political: a decision to prioritise districts that generate less complaint pressure, that have less political influence, that do not have aldermen writing letters to the Palace. This decision must be made against the active resistance of the districts that currently benefit from their priority status, who will correctly identify any equalisation policy as something that delays their own repairs to accelerate someone else’s.

The political case for closing the gap is also progressively harder to make as the gap widens. In year one, the infrastructure in the Shades is slightly worse than the infrastructure in Nap Hill: the case for equalisation is a case for marginal adjustment. In year thirty, the infrastructure in the Shades is substantially worse, the technical capacity to maintain it is lower because its failure frequency has kept the repair budget perpetually depleted, and the cost of bringing it up to standard requires a large explicit investment that would never have been necessary if equalisation had begun in year one. The cost of intervention rises with each year of inaction. The political will required to close the gap rises with the cost. The gap widens because closing it tomorrow is harder than it was yesterday, which means it is deferred until tomorrow.

The Mezzogiorno

The infrastructure gap between northern and southern Italy has been documented since national unification in 1861. It has been the subject of policy interventions, national development programmes, European structural funds, and political debate for more than a century and a half. It has not closed.

The ratchet mechanism is visible in the water sector in particular. Southern Italian municipalities, in Calabria, Sicily, Basilicata, and parts of Campania, operate water distribution networks that are older, more prone to failure, and more expensive to maintain than their northern equivalents. They are more expensive to maintain partly because they fail more often, and they fail more often partly because failures are repaired more slowly. The slower repair is not primarily a result of lower funding, though that is a factor. It is also a result of lower local technical capacity: the engineering workforce available for maintenance in a Calabrian comune is smaller and less specialised than in a Milanese municipalità, partly because the employment opportunities that would attract and retain engineering talent are fewer, partly because a history of unreliable infrastructure has not generated the institutional knowledge base that sustained good infrastructure produces.

Each slow repair leaves the network slightly worse. The 2023 Sicilian water crisis revealed distribution infrastructure that had been deteriorating for thirty years: pipe networks running at leakage rates that would trigger immediate regulatory intervention in northern Italy, pumping stations operating on equipment that had exceeded its service life by years, pressure inadequate to reach upper floors without household storage. The gap had not been created by thirty years of explicit decisions to neglect Sicilian water infrastructure. It had been created by thirty consecutive cycles in which the repair allocation was insufficient to close the gap opened by the previous year’s failures, combined with thirty consecutive cycles in which failures were resolved more slowly than in better-resourced regions, leaving each iteration of the network slightly worse than the last.

The ratchet does not require malice. It requires that prioritisation reflects existing capacity, and that existing capacity reflects existing inequality. These conditions are permanently in place. The ratchet turns continuously.

Seine-Saint-Denis

In the Paris metropolitan area there are observable socioeconomic and energy‑service disparities between central arrondissements and the suburban département of Seine‑Saint‑Denis. Seine‑Saint‑Denis, immediately north of Paris, is one of the poorest and most densely populated parts of metropolitan France, with a proportion of social and low‑income housing substantially above the national average.

Electricity distribution in Seine‑Saint‑Denis is operated by the same national distributor, Enedis, under a uniform regulatory framework that does not explicitly permit differential service quality based on income or wealth. What does vary are the broader conditions that surround energy use and complaints: energy vulnerability and precarity rates are higher in Seine‑Saint‑Denis than in wealthier neighbouring départements, and local diagnostics commissioned by the département identify energy hardship as a significant social issue.

There are also longstanding indicators of infrastructure age and maintenance challenges in many social‑housing districts across Île‑de‑France. Older equipment and distribution networks that have not been replaced as recently as in wealthier suburbs reflect investment patterns that prioritise urgent, politically salient needs. That does not imply intent to disadvantage one area, but rather that investment cases in higher‑income zones are frequently more immediately visible and politically legible.

During the 2022 European energy crisis, French authorities formally recognised précarité énergétique, the vulnerability of households to energy cost and supply pressures. Policy responses emphasised commercial and economic continuity while also expanding social support measures for vulnerable households. Within that context, the combination of higher energy vulnerability, older infrastructure and unequal complaint channels contributed to observable differences in service experience between Seine‑Saint‑Denis and neighbouring wealthier départements.

What the gap costs, and who pays it

The infrastructure inequality ratchet is a cost transfer mechanism as much as an equity problem. The costs created by the ratchet do not disappear. They are paid somewhere.

The direct costs are paid by the residents of neglected districts: in service interruption, in the time and money spent on workarounds, in the health consequences of inadequate utilities, in the reduced economic opportunity of a location that is known for infrastructure unreliability. These costs are real and large. They do not appear in the infrastructure budget.

The indirect costs are paid by the whole city when the accumulated neglect produces a failure large enough to require emergency intervention. The emergency contract to restore a water network that has been deteriorating for thirty years costs more than thirty years of adequate maintenance would have. The political crisis generated when a neglected district’s infrastructure fails publicly costs more, in trust, legitimacy, and emergency expenditure, than the equalisation investment that would have prevented it. These costs appear in the emergency budget, not the maintenance budget, and in the political cost of crisis, not the investment cost of prevention. The connection between the decades of ratcheting and the emergency cost at the end is visible in retrospect and almost never in prospect.

The people who benefit from the ratchet, residents of maintained districts whose repair priority is protected by their political influence, do not bear any of these costs directly. Their infrastructure is maintained. The emergency costs when the neglected districts’ infrastructure fails are spread across the public budget. The private benefit of priority maintenance is retained. The public cost of accumulated neglect is socialised.

Reversing a ratchet

A ratchet that turns by default can only be reversed by deliberate counter-pressure. In infrastructure terms, this means scheduling systems that explicitly prioritise repair speed in historically underserved areas; maintenance contracts with service level commitments that are enforced by district rather than averaging across districts; inspection frequency requirements that are higher for older and more failure-prone infrastructure rather than uniform; and investment decisions that treat the cost of closing the existing gap as a separate and explicit line, visible as a legacy cost of previous ratcheting rather than hidden in the general maintenance budget.

Each of these interventions is technically straightforward. Each of them requires an explicit political decision to prioritise districts that do not generate the political pressure that would make prioritising them the path of least resistance. That decision must be sustained not for one budget cycle but for enough cycles to reverse the accumulated degradation of previous cycles. In a system where budget decisions are made annually and the compounding effect of slow repairs operates over decades, the political commitment required to reverse the ratchet is of a substantially different order from the political indifference that allowed it to run.

The same street, thirty years later

In Ankh-Morpork, the infrastructure of the Shades and the infrastructure of Nap Hill are not different because different investments were made in them at the beginning. They are different because, at every fork in the thirty years between the beginning and now, the repair crew went to Nap Hill first.

The generator on Nap Hill was replaced four years ago, on schedule, before any failure occurred, because the maintenance contract specified it and the aldermen had ensured the contract was honoured. The generator in the Shades is on its third service extension because each time replacement came up for discussion, the budget was committed elsewhere, which was always possible because the generator was, technically, still running.

When the Shades generator fails this winter, it will be repaired on Friday. The distribution panel, which the crew did not have time to check when they were there in the autumn, will begin to cause problems in the spring. The spring problem will be patched in summer, when the crew can fit it in. The summer patch will hold until next winter.

No individual has decided that the Shades should have worse infrastructure than Nap Hill. The gap has simply been maintained, one repair queue at a time, for thirty years.

It will be maintained for thirty more unless someone decides, explicitly and at some political cost, that it should not be.