Chapter 1 — The Surplus That Must Be Rebuilt
Opening: What a Society Must Keep Working
A wealthy society can appear, at first glance, to be a collection of valuable things. There are homes, factories, roads, bank accounts, patents and land. Their value can be estimated, recorded and compared. But the use of any one of them depends on people and arrangements that do not fit neatly inside its price. A bridge needs inspection and repair. A worker needs food, sleep and, at some point in life, care from another person. A payment depends on institutions that keep accounts, settle obligations and provide routes to resolve errors. A farm depends on soil and water whose condition may change long before the change appears in an owner’s accounts.
Imagine a hypothetical river town that has prospered for a generation. Its factories sell equipment beyond the region, wages have risen and the local government has built a school and a clinic. A visitor sees new buildings and busy shops. The town’s accounts show an increase in property values. These signs of prosperity do not, by themselves, show how durable or broadly shared it is.
The water system loses a little more water each year because pipes are not replaced. The clinic depends on a small group of experienced nurses approaching retirement. Parents organize care informally when shifts at the factory change; some reduce paid hours because the available arrangements no longer fit. A manufacturer has improved its accounts by postponing maintenance, while a rise in land prices makes it harder for younger households to remain near their work. None of these developments must cause an immediate collapse. Their significance is that each moves a cost or an obligation into another place or another period.
The town also has reasons for confidence. Local firms have trained apprentices, residents have organized a flood-response group, and the municipality has begun to publish a maintenance schedule. A society is not made durable by eliminating all deterioration. Equipment wears out, people move, knowledge becomes obsolete and institutions encounter problems they were not designed to solve. Durability depends on whether losses can be seen, obligations can be assigned and people have the resources and authority to repair or change what no longer works.
This book calls that wider arrangement a wealth civilization. The phrase does not designate a superior culture or imply that societies pass through a fixed sequence toward one model of success. It names a question: how do communities create resources, distribute claims on them, maintain the conditions of production and care, and leave future people the ability to make choices of their own? Different societies answer through different combinations of households, markets, associations and public institutions. Their answers can be productive and unjust at the same time.
The word wealth must therefore be used with care. A financial balance can tell its holder what claims are recorded in an account; it does not by itself establish that every household can obtain housing, that a river remains usable, or that a public promise can be honored. A national measure can combine many assets to reveal trends that an output measure misses, but its categories and valuations decide what becomes visible. Some losses can be repaired with money and time. Others may cross thresholds that cannot be reversed on a useful human schedule. Some transfers make one balance sheet stronger while leaving the underlying system unchanged.
The chapters that follow move from surplus and maintenance through ownership, credit, work and care, infrastructure, ecology and distribution. They then examine cities, knowledge, resilience and obligations to people not yet able to bargain for themselves. Examples are hypothetical unless a source is identified. They are tools for tracing decisions, not disguised claims about a particular place. Sources are described with the limits of what was read and what their methods can support.
There will be no single score at the end. The aim is to make the relationships behind a society’s apparent wealth easier to inspect: who does the work, who holds the claim, what must be maintained, who can challenge a decision and what happens when a source of prosperity also produces a cost. A civilization of wealth is sustained by what people choose to make available and repair, as well as by what they choose to accumulate.
A harvest can be larger than the food eaten immediately after it. A workshop can produce more goods than its workers consume. A city can collect more in taxes than it spends on current services, though the tax itself is an expense to those who pay it. These differences create room for decisions: whether to store, invest, share, repay, celebrate, protect against a bad year or use more now. In ordinary speech that room is called a surplus. But a society cannot know how much is available for a new commitment merely by subtracting this year’s purchases from this year’s receipts.
Some of the resources used to create those receipts have worn out. Machines need parts; soil may need time and treatment; workers need rest and training; records need to be kept; children and older people need care. Some obligations are already owed to people who supplied the current result but have not yet been paid. If these claims are left out, the apparent surplus can include resources required to keep existing commitments. The number may be accurate as a cash balance and misleading as an account of what can safely be promised again.
This does not mean that every item must be restored to its previous condition before a society can invest or consume. A road may be made unnecessary by a change in settlement. A machine may be replaced by a safer design. A community may decide that land used for one form of production should be restored for another purpose. The point is to make the decision visible. A bridge allowed to deteriorate because a new crossing will serve the same users is different from a bridge allowed to deteriorate because its maintenance budget was spent without anyone assessing the consequences.
A margin that belongs to several people
Consider a food-processing cooperative. At the end of the year it has money left after paying wages, suppliers, interest and taxes. The members propose a distribution. An engineer reports that the refrigeration system will need a major component within three years; the accountant says there is no immediate bill. The cooperative could distribute the whole balance and borrow later. It could retain enough to purchase the part. Or it could reduce production and rent space from another operator when the present equipment fails.
Each choice divides a risk. A distribution gives members money now, which they may need for their own debts or household expenses. Retention leaves less available to them but reduces the amount the cooperative must find later. Borrowing preserves the current distribution if a lender is willing to lend on affordable terms when the component is needed. Renting avoids ownership of a replacement but makes the cooperative dependent on capacity that another firm may also need. None of these options is automatically prudent without knowing the cooperative’s contracts, cash needs, access to credit and alternatives.
The engineer’s estimate is uncertain too. A part may last five years or fail next month. Setting aside the full estimated replacement cost today can tie up resources that would have produced a greater benefit elsewhere. Waiting until the warning is visible to everyone can make repair expensive or impossible. A useful decision describes the range of outcomes, identifies who absorbs each one and revisits the estimate as new information appears. It also records the decision so that a later manager can distinguish an informed risk from an obligation that was simply forgotten.
The same reasoning applies beyond an enterprise, although a public institution cannot always refuse the people who depend on it. A utility that postpones replacing a water main may maintain low bills for current customers. When the main breaks, households, businesses, emergency services and the utility can all bear costs. If poorer neighborhoods have fewer alternatives to tap water, the same failure does not impose the same burden on everyone. A budget saving is not a complete measure of the consequence.
A backlog is a signal, not an invoice
The United States Government Accountability Office examined the Department of Defense’s installation maintenance and reported an estimated backlog for fiscal year 2025. The report describes funding below identified maintenance needs, workforce concerns and inconsistent information about work orders. It also notes that standards vary across military services. The headline estimate is a warning about a large set of reported needs; it is not a bill that could be paid once to make every facility equally useful or eliminate all future maintenance. [1-1]
GAO analyzed policies, budgets, maintenance information and other documents, and interviewed officials and maintenance customers at eight selected installations. Those sites were not selected to represent every installation statistically. The report describes, among other cases, an aircraft parking area where repeated smaller repairs continued while a larger replacement proceeded in phases. Officials estimated resources spent on repairs and described the competing priorities that slowed replacement. That account illustrates how a sequence of constrained decisions can preserve an asset in the short run while making the longer decision harder. It does not establish that every deferred project should be replaced immediately or that a different budget would have resolved every constraint. [1-1]
Backlog figures can rise for several reasons. More work may be left undone. The cost of materials or labor may increase. An agency may inspect more assets, change its estimate, include work it did not previously record or discover that a facility is no longer fit for its purpose. A larger reported figure can therefore reflect a real deterioration, better information or both. Comparing two numbers without understanding their definitions can make a manager who has identified hidden problems look worse than one who has not looked for them.
A backlog can also contain work that should not be done. A building may be empty because the service it once housed has moved; restoring it to an earlier standard may consume funds needed elsewhere. Before disposing of it, however, an agency may need to remove hazardous material, consult a community or fulfill a legal obligation. The choice is not between a costless deletion from the register and a costless repair. It is a comparison of future uses and responsibilities, including the costs of ending a use.
How a saving becomes someone else’s expense
In one hypothetical municipality, maintenance on a bus fleet is delayed. The operating budget records fewer repair payments this year. The buses remain in service, so the saving appears to support another priority. Over time, breakdowns become more frequent. Passengers wait longer, mechanics work overtime, and a private contractor is hired at short notice. An employer may lose a shift because workers cannot arrive on time. A parent may have to pay for another form of transport or miss an appointment.
The municipality’s accounts may eventually record the emergency contractor and overtime. They may not record the passenger’s lost time or the appointment that could not be rescheduled. The expense has not vanished; part of it has moved outside the account used to justify the original decision. This movement is not necessarily intentional. Different institutions measure different responsibilities, and some costs are difficult to observe. It is nevertheless a reason to ask who must adapt when an asset is not maintained.
There are cases where delay is justified. A municipality may face an immediate flood that makes an uncertain future bus repair less urgent. The fleet may be near retirement, and a reliable replacement may arrive soon. A scheduled repair may require a scarce component that is more valuable in an ambulance. Treating all deferral as negligence would conceal these choices. The discipline is to identify the affected service, the evidence for delay, a review date and the person authorized to change course if the forecast proves wrong.
A decision record should also distinguish maintenance from preservation of the status quo. A city may replace diesel buses with a different system, but the transition creates obligations: charging or fueling facilities, trained workers, compatible routes, spare parts and a plan for vehicles that are no longer used. A new vehicle is not a completed service until the rest of the system can support it. Nor is an old vehicle worth keeping merely because money was spent on it before.
Work that does not arrive as an invoice
A society also reproduces its capacity through work performed without a wage. Preparing meals, helping a person take medicine, arranging transport to a clinic and watching a child while another adult works all take time. Some tasks can be purchased as services; others are provided within households or communities. The absence of a payment does not make the work free to the person performing it. Time used for care cannot simultaneously be used in precisely the same way for paid employment, study, rest or another commitment, even when several activities overlap.
The International Labour Organization’s 2018 analysis used available time-use surveys from 64 countries to estimate the volume of unpaid care work. The surveys were conducted in different years and covered countries representing about two thirds of the world’s working-age population. The report’s estimates show substantial differences in reported time between women and men. They are not a contemporaneous count of every person in every country, and surveys can miss or classify simultaneous and intermittent activities differently. Their value here is to expose an amount and distribution of work that a wage bill alone does not show. [1-2]
A hypothetical household illustrates why the distinction matters. An older person lives with an adult daughter who has a paid job. A public clinic changes its hours, so the daughter takes unpaid leave twice a month to accompany her parent. The clinic may show lower staffing costs. The household now carries a scheduling and income cost. If the daughter leaves employment entirely, the household loses earnings and the employer may lose an experienced worker. If a neighbor helps instead, the cost is redistributed again, perhaps willingly, perhaps with an expectation of future assistance. Calling any one of these arrangements efficient requires specifying whose time and choices are included.
Care is not merely a substitute for paid work. People may value caring for relatives and prefer to provide it themselves. A paid service can be inadequate, intrusive or unavailable at the hours required. Equally, an apparently voluntary arrangement can reflect a lack of alternatives or unequal power within a household. A useful policy gives people choices: reliable services, leave that can actually be taken, work schedules compatible with responsibilities, and recognition of those who provide care. Each choice requires public resources, employer cooperation or both. It cannot be achieved by adding the estimated monetary value of unpaid hours to an aggregate and assuming that the work has become easier.
Making care visible can also create administrative burdens. A benefit tied to documented hours may require carers to keep records of intimate activities. A service designed around a narrow definition of dependence may exclude people with fluctuating conditions. Paying a family member may provide income while reducing contact with services that could help the person receiving care. The design must balance recognition, privacy, quality and the ability of the recipient to direct their own life. The relevant surplus is partly a surplus of time and attention, not only a financial balance.
When a replacement cannot replace the loss
Another hypothetical community earns income by harvesting timber. It can build a road, sell more wood and use the proceeds to buy equipment. If the forest regrows at a rate consistent with the harvest, the community may be able to continue both production and the services the forest provides. If harvesting removes soil, fragments habitat or changes water flows, planting the same number of trees elsewhere may not restore the original functions. The question is not whether a tree has any monetary value. It is whether the particular functions affected can be recreated where and when people and other living systems need them.
The World Bank’s 2024 comprehensive wealth overview explicitly discusses the difficulty of substituting other assets for natural capital. Its aggregate method treats assets as highly substitutable, while the overview warns that this assumption can fail as resources become scarce or critical thresholds are approached. The report’s account is useful because it identifies an assumption that can be hidden inside an apparently comprehensive number. It does not supply a universal threshold for every forest or establish that all natural losses are irreversible. [1-3]
A community may have good reasons to accept a particular environmental change. A new water reservoir can displace habitat while making a town less vulnerable to drought. A solar installation can require land and materials while reducing dependence on another energy source. The choice requires more than comparing a project’s monetary gain with a single estimated loss. Decision makers need to know which effects are reversible, who experiences them, what uncertainty remains and whether another design could provide the service at lower cost to the affected people and ecosystems.
Thresholds are often uncertain. Waiting for conclusive evidence of irreversible damage may allow damage to occur before anyone can respond. Declaring every disturbance unacceptable can prevent housing, energy or transport improvements that people also need. A practical approach is to define limits where evidence justifies them, monitor indicators that can reveal an approaching problem, and retain the ability to change or stop the activity. Monitoring is useful only if someone has authority to act on the result and the budget to do so.
The calendar of promises
An institution can have a positive annual balance while its future commitments exceed the resources available to meet them. This is not unique to governments. A business can sell a product with a warranty, a household can borrow for a house, and a cooperative can promise retirement payments. The timing of receipts and obligations matters. A promise due in ten years may be manageable if income arrives as expected and difficult if a customer leaves or an asset fails early.
A hypothetical town plans a new sports center. The construction grant pays for the building, and the opening ceremony attracts donations. The grant does not cover heating, cleaning, staff, insurance or roof replacement. The council expects membership fees to pay those costs. Before committing, it can examine several scenarios: high use, low use, a severe winter and a major repair. The exercise will not reveal the future. It can show which assumptions must hold for the promise to be affordable and what the council will do if they do not.
A small community may rationally build a center that never covers its operating expenses because it provides a place for youth programs, emergency shelter or social connection. The subsidy is not necessarily a failure. It should be a decision made with an account of the benefit sought and the recurring resources required, rather than a surprise discovered after the capital grant is spent. A society that refuses every project without a direct financial return could underprovide valuable public services. One that accepts every attractive project without identifying its operating obligations can make promises it cannot sustain.
The person who approves construction may not be the person responsible for the annual operating budget. A school district can receive a building while a separate authority pays for transport and staff. A ministry can fund equipment that a local clinic must service. A grant can expire before the first major repair comes due. These divisions of authority make it possible for each office to follow its own rules while the combined arrangement remains incomplete. Coordination requires more than publishing a total cost: the parties must agree who pays, who can change the service and who answers when the expected funds do not arrive.
A maintenance reserve can help, but a reserve is not a substitute for judgment. Money may lose value if costs rise faster than expected; a protected fund can be raided in an emergency; a restriction on using it may prevent a better replacement. In a poor household or local government, reserving cash for an uncertain future repair can be difficult when a present need is acute. Credit, insurance, shared facilities and public support can spread the burden, but each creates its own claims and exclusions. There is no accounting device that removes scarcity.
Which repair comes first?
Even a well-informed institution cannot maintain everything at once. A local government may have an unsafe bridge, a leaking school roof and a treatment plant that needs new controls. Each project has advocates. Each can become more expensive if delayed. A ranking based only on the size of the repair bill may favor the smallest job, even if the treatment plant’s failure would interrupt service to the whole town. A ranking based only on the number of users may overlook a bridge that is the only access route for a small community.
A useful comparison starts with the service and its alternatives. How likely is failure within the period being considered? What would happen if it occurred? Can users reach another facility, and at what cost? Can a temporary repair reduce risk while a larger project is designed? The answers will be uncertain, and different people may reasonably disagree about the weight to give them. The decision should nevertheless make the disagreement inspectable. An official can explain why a school roof was repaired before the bridge, and residents can point out a condition the official missed.
Consider a rural district with two bridges. One carries many vehicles but has a detour adding ten minutes to most journeys. The other carries fewer vehicles but is the only route for an ambulance to reach a village in heavy rain. If the district counts only traffic, the first bridge will dominate. If it counts only the worst conceivable event, the second may consume the entire budget even when the risk is remote. A credible assessment examines condition, exposure, feasible alternatives and the consequences for particular users. It may still choose either bridge. The reason matters because the district will need to defend the choice and revise it if conditions change.
An emergency repair can also be the wrong response to a long-term change. If a settlement is moving away from an eroding shoreline, rebuilding every road in its original location may lock people into repeated losses. Relocation is not a simple technical correction: households may lose access to work, burial sites, neighbors or land they cannot replace at a similar price. The question of whether to maintain, adapt or withdraw must include those costs and the authority of affected people to contest the plan. A balance sheet that records only the replacement value of the road will miss much of the decision.
Information has to travel back
A maintenance system depends on reports from people close to the work. A nurse may notice that a refrigerator for medicines fails whenever power fluctuates. A driver may hear a new noise before a bus breaks down. A resident may see a drain blocked before officials receive a formal complaint. If reporting creates blame or extra paperwork without producing a response, people may stop reporting. The institution then sees fewer complaints and may conclude that conditions have improved.
The reverse problem also occurs. A new reporting channel can produce a sudden increase in recorded faults because people now have a way to report them. A manager measured by the number of faults may discourage the channel; a manager measured only by the speed of closing tickets may mark a problem resolved after a temporary workaround. A meaningful measure pairs the number with other information: whether users can still obtain the service, whether a repair recurs and whether the people who reported it agree that it has been addressed.
The organization needs a route from observation to action. The mechanic who knows that a component is failing may not be allowed to order a replacement. The purchasing office may need three bids for a part available from only one supplier. The supplier may require advance payment while the public body pays only after delivery. Each rule may have a reason: limiting favoritism, preventing waste or protecting public funds. Together, they can prevent a modest repair until a larger breakdown makes an emergency exception available. Changing the outcome may require revising an approval threshold, maintaining a list of qualified suppliers or holding a spare part, not simply telling workers to be more attentive.
Those remedies also create risks. A larger purchasing threshold gives local managers room to act and creates more discretion that can be abused. A list of approved suppliers speeds orders and can exclude a capable new entrant. A spare part in storage costs money and may become obsolete. The right arrangement depends on the frequency and consequence of failure, the market for the part and the institution’s ability to audit its choices. The comparison is between imperfect procedures whose costs fall at different moments.
The productive use of uncertainty
It is tempting to demand that every liability be known before any new spending occurs. Complete knowledge is unavailable. An inspection can miss a hidden defect, estimates can change and a new technology can make an old replacement plan unnecessary. Some experiments will fail, and an organization that never accepts uncertainty may preserve a deteriorating arrangement because the alternative cannot be guaranteed.
The answer is to distinguish uncertainty that can be reduced from uncertainty that must be borne. A utility can inspect a sample of pipes, test pressure and compare failures by location. The results may narrow the range of likely repair costs without revealing the condition of every buried section. It can then choose an initial program, monitor results and revise the schedule. If the first streets reveal far more corrosion than expected, the governing board should have an agreed route to reconsider the budget. If they reveal less, resources can move to another need.
A contingency is useful when it is linked to a decision. A fund for unexpected repairs gives an institution some room to act, but the rules must say who may draw on it and how the use is reported. An emergency exception can speed a response, but repeated use for predictable problems suggests that the ordinary process is not working. A reserve without a plan can become an attractive target for another expenditure; a plan without accessible funds can become a list of obligations that nobody can meet.
Keeping the choice available
A society’s capacity to make a future choice is itself a result of present decisions. If a town keeps records of its pipes, trains technicians and retains several suppliers, it has more options when a failure occurs. If it allows those skills and records to disappear, it may have to accept the terms of the only company able to restore service. The second town may have reported lower costs for years. Its saving purchased consumption or another investment, but it also reduced its room to negotiate later.
That observation does not require every community to duplicate every capability. A small town may sensibly share an engineering team with neighboring districts. A hospital may contract for specialized maintenance that it could not afford to provide internally. Dependence is not the same as helplessness. The relevant questions are whether the arrangement has alternatives, whether its terms can be understood, and whether the people relying on the service have a way to respond if the provider fails.
Surplus is therefore not a pile of money discovered after the important work is done. It is the margin remaining after a set of judgments about what must be preserved, what can change, who has already contributed and who will bear the consequences of a promise. Those judgments can be disputed. They should be, because the people who receive a distribution may not be the people who maintain the system that makes it possible.
The next chapter turns to the rules that make such claims enforceable. Ownership can protect the person willing to maintain an asset, but it can also place a cost on people who depend on it without holding a title. A society’s wealth depends on how those rights and obligations are arranged, contested and revised.
Notes and Sources
[1-1] U.S. Government Accountability Office, Installation Maintenance: Better Information on Risks and Challenges Needed to Improve Oversight of DOD Facility Conditions, GAO-26-107255, published and publicly released August 21, 2026, https://www.gao.gov/products/gao-26-107255; full report https://files.gao.gov/reports/GAO-26-107255/index.html, accessed 2026-10-31. Read Fast Facts, highlights, the scope-and-methods summary and the full-report passages on the funding model and Minot’s phased pavement repair. GAO analyzed DOD documents and interviewed officials and maintenance customers at eight selected installations that do not statistically represent all installations. Backlog and repair costs are estimates under specified methods. The full report, underlying work orders and subsequent changes were not independently reviewed.
[1-2] International Labour Organization, Care work and care jobs for the future of decent work (2018), chapter 2, section 2.1.2, https://www.ilo.org/sites/default/files/wcmsp5/groups/public/%40dgreports/%40dcomm/%40publ/documents/publication/wcms_633135.pdf, accessed 2026-10-31. Read the chapter summary and the passages describing time-use surveys, coverage and unpaid care as a primary activity. The 64-country estimate draws on surveys from different years and covers countries representing 66.9 percent of the world’s working-age population; simultaneous and intermittent care may be difficult to capture. The rest of the report, survey microdata and harmonization procedures were not reviewed.
[1-3] World Bank, The Changing Wealth of Nations 2024, publication overview, https://www.worldbank.org/en/publication/the-changing-wealth-of-nations, accessed 2026-10-31. Read the overview’s sections on comprehensive wealth, the chained Törnqvist volume index and limited substitutability. The overview states that the current aggregate method assumes high substitutability and discusses why that assumption can fail. The complete report, underlying data and detailed asset estimation methods were not reviewed for this passage; the overview does not establish a threshold for a particular ecosystem.