Three Case Studies in Maintenance: What Lasts, What Fails, and Why

Ise Grand Shrine, Mie Prefecture, Japan

We begin with the one that worked.

The Ise Grand Shrine in Mie Prefecture, Japan, has stood—in some form—for over two millennia. Not because it was built to last, but because it was built to be rebuilt. Every twenty years, the entire shrine is dismantled and reconstructed, a practice that began in 690 CE and continues today. The result is a structure in perfect condition, maintained not by repair but by renewal.

Two other systems came close to this kind of longevity. Neither made it. The contrast between what Ise got right and what Rome and Venice got wrong is the lesson.

---

Ise Grand Shrine: The One That Endured

The Inner Shrine was said to have been built around 4 BCE; the Outer Shrine around 500 CE.

The unique features that enabled this shrine to remain in perfect condition today include:

• The Tokowaka Concept (Eternal Youth and Renewal). Total reconstruction every 20 years, starting from 690 CE. By the 14th century, the original forest next to the shrine was depleted, and the shrine management had to source timber from elsewhere. In 1923, a 200-year management plan was established for the cypress forest; upon its completion, all lumber for rebuilding will be locally harvested. In the current 20-year rebuilding cycle—the 63rd, launched in May 2025—some of the timber will come from the shrine's own forest.

• A built-in knowledge transfer. The 20-year cycle allows apprentice carpenters to watch and practice hands-on alongside master carpenters. The apprentices will be ready in the next cycle as masters. A synchronicity between craftsmen and forest.

• Financing independent of government. The Jingu Shrine Management uses the 20 years between each rebuilding to raise funds: offerings from 8 to 14 million annual visitors, sales of amulets, private functions like corporate prayers, weddings, and blessings. Corporate donations make up a large percentage. The estimated budget for the current 63rd cycle is 57 billion yen.

---

Those That Almost Made It

The Familiae Aquarum of Rome

Rome had Ise's engineering ambition—but not its funding discipline. The difference shows in what one auditor found when he actually looked.

Frontinus didn't just suspect corruption—he measured it. When he took charge of Rome's water supply in 97 CE, he compared the volume of water entering the aqueducts at their sources against what actually arrived at the distribution tanks. The gap was enormous: roughly 10,000 quinariae unaccounted for across the system. The cause was fraud, and it ran both ways. His own water-men—the aquarii—were diverting public water into private pipes for a fee. "We have detected," he wrote, "the drawing of waters from public conduits for use by private persons." But the deeper rot was structural: private villas, shops, even "all establishments of unwholesome pleasure" had tapped the lines illegally, and the supervisors meant to stop it were either complicit or negligent. Frontinus documented every deficiency in De Aquaeductu, but his manuscript was a diagnosis, not a cure. The water kept flowing—until the funding, the workers, and the will to maintain it all ran dry.

The deficiencies are instructive:

• The water tax was meant to fund maintenance specifically, but it was collected by the state Treasury and never held in a separate "lock-box."

• Private villas bribed workers to puncture and divert water meant for public facilities—fountains and baths meant for all Romans.

• Because the aqueducts were exceptionally well built, the early years required little maintenance. Budgets for maintenance and recruitment of new apprentices were neglected. The experienced workers died off.

• · Emperors mandated "hereditary trades," producing unmotivated new workers.

• · In 330 CE, Constantine moved the capital from Rome to Constantinople, further reducing the population and skilled workforce.

• · In 410 CE, the Visigoths cut the aqueducts—already clogged from decades of neglect—bringing an end to a magnificent gravity-fed water system.

At its peak during the Pax Romana (27 BCE–170 CE), Rome's population exceeded a million people. The water system supplied an estimated 500,000 to 1 million liters of fresh water daily.

A curious legacy: the only aqueduct still functioning today feeds the Trevi Fountain—a site most tourists have visited without knowing it.

---

The Venetian Arsenal

Venice had Ise's skilled workforce—but not its renewal. The difference shows in who actually showed up for work.

The Arsenalotti's privileges were supposed to secure loyalty. Instead, they created a workforce that showed up on paper but not in the yards. By the late 16th century, the gap was staggering: in 1559, 2,183 masters were enrolled on the Arsenal's rosters, but a head count during a single week found only 960 actually came to work—less than 44 percent. Many masters spent their prime years in private shipyards where wages were higher, then returned to the Arsenal in old age to claim what amounted to a state pension. The records describe men in their eighties and nineties, "crippled or even completely blind," still being carried to the shipyards so they could collect their pay. The Republic had built a system of lifetime security. What it got was a system of lifetime entitlement—and a shipyard that increasingly existed on paper.

Just as Rome had its familiae aquarum, the Republic of Venice relied on the Arsenalotti—an elite, state-funded guild of specialized shipwrights, caulkers, and engineers housed inside the Venetian Arsenal. At its peak in the 15th and 16th centuries, the Arsenal employed up to 16,000 highly trained workers. The system was a marvel of proto-industrial efficiency: a proto-assembly line producing hull, rigging, oars, and weapons—one complete warship daily. The workforce received special privileges and lifetime employment, free housing, pensions, and tax-free wine. During off-hours, they provided bodyguard services, night watch, and fire-fighting.

The deficiencies that failed the system:

• Their privileges made them a politically powerful hereditary caste; later generations suffered from absenteeism.

• Theft of materials and Venetian economic downturn exacerbated the situation.

• Maritime trade shifted to the Atlantic.

• They resisted new ship designs to avoid changing their assembly line setup.

• The Serrata (Lock-in) of 1297 fixed the existing families as the hereditary class of elites, resulting in loss of new capital, new ideas, and new leaders.

• Political stability came at a cost: elites became landlords rather than merchants, losing the adventurous capital that had built the Arsenal.

• · By 1797, the arrival of Napoleon was met with a simple handover—a total surrender. The Great Council dissolved the Most Serene Republic of Venice.

---

The Budgeting Gap

When we propose a new hospital, we can estimate—from the maintenance records of hospitals with the same design and layout—what the roof, the operating theaters, and the HVAC will cost to maintain, starting in year three. The data exists.

What doesn't exist is the line item.

Instead, we wait until something breaks. A pane of floor-to-ceiling glass falls in a departure hall at KKIA, and then we inspect—what, all the glass walls?

This is not a knowledge problem. We know what maintenance costs. It is a budgeting problem: the money for upkeep is committed after failure rather than before it. New projects are visible and ribbon-cuttable; maintenance is invisible until it is urgent. And because the information is available and the line item is not, the gap is not one of ignorance.

"Perhaps it's time for a new airport?" a classmate remarked. It is a reasonable question, given how we currently budget. The observation is not the problem; it is a symptom of how we fund things.

The deeper fix is simple and hard: write the maintenance line item before the building opens.

---

So What Did We Learn?

Notwithstanding the differences in circumstance among the three case studies, several features mattered for the longevity and durability of infrastructure and services.

Funding and governance

• · Maintenance funding must be protected from general treasury raids.

• · Independent oversight and audit are non-negotiable.

Human capital

• Apprenticeship systems preserve tacit knowledge.

• Recognition of master craftsmen matters.

• Hereditary castes, however privileged, breed complacency.

Supply chains

• Secure raw material sources before you need them.

Culture

• The belief that "no breakdown is acceptable" must be shared by government and public alike.

---

A Modest Proposal

One structural problem remains: how do you reward people for preventing something that never happens? Ise solves this through religious devotion and a 1,300-year tradition. Most institutions cannot replicate that.

One possible approach to the danger of complacency—the "we'll fix it when it's required" attitude—is a form of Reverse Bonus Payment: a yearly bonus for no breakdown, increasing each year to a large payout at regular intervals of five years. This could begin after the initial "expected no repairs" period.

Just an idea. It could be applied to all jobs where "nothing happened" is what society wants: fire brigades, security services, and perhaps even public health.

Ise shows that renewal can be institutionalized. The question for our own infrastructure is whether we can build that discipline without a 1,300-year tradition behind us.