Chapter 1: What Rome Could Provide
Prologue: The Good Life Has an Address
The retired woman in this imagined town has three letters on her table. Her pension statement shows a payment she can live on. The building manager wants a contribution toward replacing the elevator. The service that helps her husband at home cannot yet offer the additional visits she requested. None of the letters is especially surprising. Together they make the next year difficult to plan.
She owns the apartment. Its value has risen, and an adviser could reasonably describe her household as wealthier than it was. Selling would release money, but she would have to find another home, arrange the move, and rebuild the local connections through which much of her husband's help arrives. Keeping the apartment means accepting the elevator contribution and the uncertainty of the work. The household has assets and income. It also has needs that cannot be settled by adding the two figures together.
The elevator is shared. A neighbor who rarely uses it would rather defer replacement. Another needs it every day but cannot easily meet the proposed charge. The manager can obtain a quotation; she cannot turn that quotation into consent, contributions, and a completed installation. The woman can pay her share and still have to wait.
This is not a story about a household deprived of every resource. Its relative comfort makes another question visible. How far do the resources people possess give them dependable command over the life they are trying to lead? Additional money would help. It would not immediately produce a trained home-care worker or resolve the building's disagreement.
Some needs can be met through purchases. Some require public services or cooperation with neighbors. Others depend on unpaid work and promises whose fulfillment cannot be demanded at a counter.
A life does not arrive as an average
The 2009 commission chaired by Joseph Stiglitz, with Amartya Sen and Jean-Paul Fitoussi, urged greater attention to household income and consumption, their distribution, and the wealth supporting them. It also separated questions about present well-being from the resources needed to sustain it. These were recommendations about measurement, not proof that one policy could deliver a good life everywhere.[0-1]
That distinction leaves room for a more demanding inquiry. What can people actually obtain with the resources they command? Which necessities remain uncertain even when their income rises? Who supplies the work that makes another person's apparent independence possible?
For the woman at the table, her husband's care is not a philosophical objection to material comfort. It is part of what makes that comfort usable. If she supplies more help herself, she gives up other activities and may become exhausted. A paid helper could relieve that burden if one becomes available on terms the household can sustain. A public service could do so too, but someone must recruit, train, and retain the people delivering it. Assigning responsibility to a household, a company, or a government does not remove the work.
Nor are all the town's residents in the same position. A landlord may gain from rising rents. A tenant may lose room in the budget. An older homeowner may possess a valuable asset while struggling to pay for repairs. A young renter may earn more and own little. Their circumstances are not adequately described by sorting them along a single line from poor to rich. Income, wealth, obligations, and access matter together.
An average can tell us something essential about a town while describing none of its households particularly well. We need the average, but we also need to know what it conceals.
Who can depend on whom
Prosperity becomes tangible when people can make plans that reach beyond the next emergency. A shopkeeper can replace worn equipment without gambling the household's rent. A worker can refuse dangerous conditions without immediately losing shelter. Someone recovering from illness can expect help without having to prove a new personal connection each morning. These are different achievements. They require different resources and forms of cooperation.
Dependability is not the promise that circumstances will never change. Such a promise would be expensive to honor and often impossible. Her husband's needs may increase; a neighbor may leave; the town may discover that an earlier service plan was mistaken. A society capable of sustaining prosperity needs ways to revise commitments as well as reasons to trust them. People should be able to learn without treating every disappointed expectation as a betrayal, and to challenge a decision without placing their whole livelihood in jeopardy.
There are real conflicts here. A cheaper apartment for one household can mean a lower return for its owner. More generous care can require taxes, forgone consumption, or workers drawn from another service. Protecting a neighborhood from change can preserve valued relationships while preventing newcomers from joining it. These choices cannot be settled merely by naming the system that makes them. We have to ask whose resources are committed, which alternatives remain, and whether those carrying the cost have a defensible place in the decision.
The chapters begin with evidence from an ancient economy, then move toward poverty, resource revenues, property, trust, remedies, enterprise, cities, and finance. They end with the difficulties created by success itself. An institution that once opened opportunities can become a way to reserve them for established participants. A useful protection can grow difficult to revise. A prosperous generation can leave impressive claims on paper and too few people prepared to fulfill them.
There is no inevitable decline in that sequence. There are choices whose consequences may be difficult to recognize while the beneficiaries are still celebrating.
At the table, the woman does not need a perfect society before making a decision about the apartment. She needs an intelligible account of what she can depend on, what she must contribute, and which uncertainties she still has to carry. Her pension statement is part of that account. So are the other two letters. Prosperity deserves to be judged where people have to live with it.
Notes and Sources
[0-1] Commission on the Measurement of Economic Performance and Social Progress, Joseph E. Stiglitz, Amartya Sen, and Jean-Paul Fitoussi, *Report* (2009), executive summary, recommendations on material living standards and sustainability, https://ec.europa.eu/eurostat/documents/8131721/8131772/Stiglitz-Sen-Fitoussi-Commission-report.pdf. The opening town and household are hypothetical.
Chapter 1: What Rome Could Provide
The Baths of Neptune at Ostia have outlasted the arrangements that kept them warm. The archaeological park describes cold and heated rooms, a place for exercise, mosaics, and dwellings above with their own entrances. The complex was rebuilt under Hadrian, completed under Antoninus Pius in 139, and restored into the fourth century.[1-1] These are specific findings about a particular building. They establish a material achievement without telling us how every inhabitant of Ostia lived.
That gap is a useful place to begin a book about prosperity. A bath building could bring an amenity within reach of people unable to reproduce it in their own homes. Its construction assembled skills, materials, money, and authority. Its continued use depended on less conspicuous work. The floors that remain are evidence of the first accomplishment; the service once delivered through them depended on many more conditions than masonry.
We should take the achievement seriously. Human comfort does not become imaginary because the society providing it was unjust. Nor does an impressive amenity settle the question of whose comfort counted. Rome presents both temptations with unusual force. A visitor can mistake the monumental for the typical. A critic can make coercion explain away every useful thing. Either response loses something essential: societies can produce real benefits while allocating security, choice, and exposure to harm very unevenly.
Prosperity concerns the practical opportunities that people can depend upon. Output supplies much of what makes those opportunities possible. But there is further work between producing a surplus and allowing a household to live with some freedom from fear. Someone must be able to obtain the food, use the service, retain the benefit of a day's work, and expect that a setback will not destroy everything. The presence of resources answers only part of that inquiry.
Reading what survives
A building is not a household survey.
The same is true of a finely made object, an inscription honoring a donor, or an account of a wealthy person's possessions. Survival and visibility favor some activities over others. Durable construction leaves a record that ordinary maintenance may not. A commission can preserve the name of the person who paid while leaving the people who did the work almost anonymous. Asking what a monument proves requires asking what it had a chance to preserve.
Imagine finding a modern concert hall after most of a city's records have vanished. Its acoustics and construction could demonstrate technical competence. Its scale might suggest the ability to mobilize substantial resources. Yet the building alone could not establish the disposable income of a cleaner, the admission policy, or the reliability of local medical care. A charitable owner, a wealthy audience, public subsidy, and substantial inequality could all be consistent with the same surviving shell. The example is an analogy about inference, not an assertion that Roman civic life worked like a modern concert hall.
Archaeology becomes more informative when different kinds of evidence constrain one another. Distribution across sites can help distinguish an exceptional possession from a broadly available product. Wear and repair can reveal use after construction. Domestic remains can complicate a story told by civic architecture. The result still depends on where excavation occurred and what survived. A missing object does not always indicate deprivation; a surviving object does not tell us that its owner had an adequate diet all year.
There is no need to choose between credulity and silence. Evidence can support a bounded claim. The baths at Ostia establish that an elaborate shared facility existed and underwent restoration. They do not establish universal access or comprehensive household welfare. Keeping the claim within that boundary strengthens the achievement rather than diminishing it. We can admire what was built without requiring it to carry a conclusion it cannot sustain.
Documentary evidence presents a different kind of selection. An administrator writes about matters that need a decision, fit his responsibilities, or allow him to present his conduct in a favorable light. Routine success may never enter the correspondence. Neither may hardship that the office does not treat as its business. Pliny's proposal concerning the baths at Prusa makes these issues concrete. In a letter to Trajan, he describes the existing bath as old and neglected and favors a new building. He identifies money being recovered from private individuals and a redirection of expenditure otherwise used for an oil distribution. He also invokes the city's dignity and the emperor's reputation.[1-2] The letter gives us a request and proposed financing, not proof of completion or a measurement of the resulting benefits. Its author had a position in the system, rather than a complete ledger of social costs.
It also leaves a recognizable question: what should a community give up to improve a service? Recovered money may look like newly available wealth from the office's perspective, but recovering a claim is different from producing additional labor or materials. Redirecting one expenditure can finance another while withdrawing a benefit from its previous recipients. A public improvement can be desirable and still have a cost that somebody experiences immediately. The mention of civic prestige adds another reason to ask whether the people choosing a project and the people benefiting most from it are the same.
An economy without national accounts
Ancient output is harder to reconstruct than an individual building. There are no comprehensive Roman national accounts waiting to be translated into modern categories. Estimates must bring scattered evidence into a model. Population assumptions affect the denominator. Judgments about subsistence, prices, income, and the relation between recorded wealth and annual returns affect the numerator and its distribution. A precise answer printed in a table can hide uncertainty inherited from several stages of construction.
Walter Scheidel and Steven Friesen's study of the second-century imperial economy makes this modeling task explicit. It compares approaches to aggregate output and builds bounded income distributions using ancient evidence and comparative assumptions. Its reconstruction places a large majority near subsistence while allowing a smaller middling sector and a much smaller elite. The authors emphasize controlled conjecture and the disproportionate visibility of more affluent people in surviving evidence.[1-3] These are model-based conclusions, not results from a representative ancient census of household income.
For our purpose, publishing an exact Roman GDP would offer the wrong kind of confidence. What matters is the relationship that a consistent account must explain. An economy that supports concentrated high consumption must also support the people supplying necessities. If average resources are limited, one cannot indefinitely enlarge both the affluent group's consumption and everyone else's command over goods without increasing the assumed total. Distribution places constraints on the story even when the total is uncertain. The unit of account can mislead as well. Expressing resources in a staple equivalent may help compare purchasing power, but it does not mean people consumed that quantity of the staple. A household needs shelter, fuel, clothing, tools, and many other things. The conversion is a measuring device. Once the device is mistaken for a lived budget, apparent clarity becomes distortion. What an ancient income could buy also varied with location, season, obligations, and access to nonmarket resources.
Annual averages conceal timing. A household can produce enough over a year and still lack food before its next receipt. A regional estimate can conceal a district facing interruption. An individual's apparent consumption can depend on someone else's unpaid or coerced labor. Even a well-supported estimate of output per person would need additional evidence before it could describe how dependable an ordinary life was.
A mean is useful because it compresses information. It is dangerous when the compression is forgotten.
None of this makes aggregate output irrelevant. Very low production limits the material possibilities of redistribution. Dividing inadequate supplies fairly does not make them adequate. The more interesting question is how much of an economy's productive achievement becomes a usable improvement for households, through what channels, and with what exclusions. That inquiry needs aggregate and household perspectives together. It cannot be completed by replacing one with the other.
A signal carried in ice
Evidence about Roman production has also arrived from far outside the Roman world. Joseph McConnell and colleagues used high-resolution measurements of lead in Greenland ice, together with atmospheric transport modeling, to reconstruct changes in European emissions associated with ancient mining and smelting. They report high emissions during the early imperial period and sharp declines coincident with major epidemics. The conversion from deposition to emissions involved substantial uncertainty, including assumptions about atmospheric transport.[1-4]
This is unusually valuable evidence because it is independent of an elite author's wish to praise a ruler or complain about a province. But independence from literary motives does not make a proxy a complete economic account. Lead emissions are most directly evidence about activities that emitted lead. Moving from those activities to all production requires further reasoning. Moving from all production to household welfare requires more reasoning again. No chemical concentration measures whether a worker could refuse an employer's demand.
A higher signal can indicate more material activity while leaving unresolved the desirability of its full consequences. Production may supply useful goods, generate income, expose workers to dangerous conditions, and leave environmental damage. The balance cannot be read from the production indicator alone. This point applies even where evidence is much better than it is for antiquity. A measure can perform its intended function accurately while remaining silent on another question that matters greatly. Coincidence in time also has to be handled carefully. An epidemic can affect labor, demand, mobility, fiscal receipts, and the organization of production. A decline in a proxy near an epidemic is consistent with disruption, but a single series does not identify each channel or its size. A regional sector may recover differently from another. Describing the alignment is warranted; treating it as a full causal decomposition would go beyond what has been observed.
The ice record therefore adds an important discipline to the discussion without supplying a final verdict. It makes it harder to treat all apparent material change as an illusion of surviving monuments. At the same time, its limits make it harder to treat production as a direct reading of the good life. The analyst has to hold both ideas at once: meaningful economic evidence exists, and its meaning has boundaries.
Who could keep the benefit?
A society's ability to organize labor does not tell us the terms on which labor was supplied. This becomes unavoidable in a society that recognizes slavery. Gaius's Institutes distinguishes free people from slaves and describes the master's authority, including the acquisition for the master of what a slave acquires. The same text records restrictions on excessive cruelty under Antoninus.[1-5] It is a legal exposition, not a survey of enforcement or an exhaustive description of lived experience.
The ethical consequence does not depend on knowing the exact contribution of enslaved labor to every sector. A person can contribute to measured production while lacking control over the use of that contribution and over fundamental aspects of their own life. Providing food or an amenity within such a relationship cannot make the coercion disappear. Prosperity defined solely as the availability of objects would miss what is being done to the person who helps produce them.
At the same time, the presence of slavery does not establish that every productive advance required slavery or that every free household benefited equally from it. Those are separate causal claims needing separate evidence. We should resist the habit of turning a justified moral judgment into a universal economic explanation. Condemning ownership of human beings is clear. Assigning all technical achievement to that institution is a much more ambitious historical proposition.
One practical question reaches across these distinctions: who has the power to retain an improvement? If more efficient work lowers the time needed for a task, the saved time can become rest, additional production, a larger return to an owner, or a reduced charge to a customer. Technology does not choose among those outcomes by itself. Contracts, authority, competition, and the worker's ability to leave help decide who gains. The same physical advance can therefore support different experiences of prosperity.
This is why income, autonomy, and access belong in the same inquiry without being collapsed into a single number. A person may have higher consumption with less freedom. Another may gain legal independence while facing severe material insecurity. We need enough dimensions to describe the change honestly. Adding them into one score would require value judgments about substitution: how much insecurity offsets an additional good, or how much consumption compensates for loss of control. Such judgments should not hide inside an index. Distribution is also more complicated than a division between a rich class and a poor mass. A small trader, tenant, landowner, official, dependent family member, and enslaved worker may have different combinations of assets, obligations, and authority. The categories are not interchangeable merely because several people consume similar quantities. One has an asset to sell; another depends on continued permission; another has income that ends when health fails. Their exposure to the same setback differs before it happens.
Keeping an amenity in service
Return to the baths. The first question is how a society could build them. The second is how it could keep them usable. Construction is a concentrated effort with a visible result. Operation is a sequence of recurrent demands. Staff must appear. Repairs compete for funds. Supplies have to arrive. Someone must notice failures and have both the authority and incentive to correct them. A society can perform the first task and neglect the second.
This is an analytical account of what a shared physical service entails, rather than a reconstruction of the management of the Baths of Neptune. The archaeological record cited here identifies restoration over time; it does not provide the operating accounts needed to establish each arrangement. The broader inference is modest: a facility's service life cannot be separated from renewed contributions of labor and resources. Past investment supplies an asset, not an exemption from future work.
Access has its own sequence. A facility can exist yet be unavailable to someone because of price, schedule, location, status, or a competing obligation. These possibilities should be investigated rather than automatically assumed for a particular Roman building. Their analytical importance is straightforward. Measuring the stock of facilities is easier than measuring the opportunity to use them. The two can move together, but the connection must be established.
Suppose, as a purely modern hypothetical, that a shared exercise center is physically accessible but open only during a cleaner's paid shift. Its existence improves the city's amenity count while doing little for that cleaner. Extending opening hours could create a real benefit with no new building. Yet it would require more staffing or a change in duties, and perhaps a higher contribution from users or taxpayers. An improvement in practical access can depend less on another capital project than on the organization of ordinary work. The counterexample is equally relevant. Perfectly convenient hours cannot compensate for a facility that lacks safe equipment or adequate capacity. Organizational insight is no substitute for necessary material resources. Prosperity requires attention to the chain because any sufficiently important break can prevent use. Which break deserves priority is an empirical question about the people and place concerned. There is no universal instruction to spend more on construction or more on management.
Prestige can sometimes help finance a useful amenity. A donor's desire to be remembered may align with the community's need. But the alignment is partial. A new wing can carry a name more visibly than maintenance of an existing room. A politician can claim credit for an opening ceremony more easily than for preventing gradual deterioration. These are possible incentives, not findings about every ancient benefactor. They explain why visible investment deserves to be examined alongside less visible upkeep.
To see the budget problem more closely, keep the exercise center hypothetical and give its managers a repair decision. A contractor reports that one changing room needs work. Closing it now would inconvenience users; waiting risks a longer closure later. The board can finance the repair from its reserve or seek a grant. Applying for the grant takes staff time and may postpone work. Using the reserve leaves less protection against a different failure. A simple instruction to maintain the asset does not decide which risk is acceptable or who has authority to accept it.
The managers first need a usable assessment of the defect. They then need a decision date, because an indefinitely pending application is itself a decision to keep operating under the present conditions. If they close the room, users who cannot change elsewhere may stop coming. If they keep it open, they need a credible basis for judging that use safe. After authorizing work, they must arrange access for the contractor, adjust staffing, notify users, and check completion before reopening. Paying an invoice is one stage in restoring the service. It is not the whole result.
This ordinary sequence gives substance to the word capacity. A reserve without technical judgment may be spent on the wrong repair. Judgment without authority may leave a known defect untouched. Authority without an account of the decision can expose users to convenience disguised as expertise. The finished building embodies past coordination; maintaining its usefulness continually tests present coordination. These relationships explain what we should look for in evidence about service continuity, even where an ancient record cannot supply every part of the answer.
Vulnerability inside achievement
Dependability changes the value of a benefit. A service that ordinarily works but disappears during a crisis may still be valuable. Its failure at that moment, however, can have a larger cost than its average availability suggests. A household with reserves can seek an alternative. A household already using all its resources may have none. The same interruption can thus widen differences that were easy to overlook during ordinary conditions.
Reserve capacity appears wasteful if judged only against the day's demand. Unused staff time, spare materials, and alternative routes all have costs. Keeping them available means forgoing another use. Yet maximizing utilization everywhere can leave no room for a disruption. The appropriate margin depends on the consequences of failure and the available alternatives. The argument is not that every service should maintain enormous reserves. It is that apparent efficiency must be assessed against the risks being accepted.
Ancient evidence offers no warrant for a simple story in which one institutional flaw eventually undid every accomplishment. Different kinds of disruption can reinforce one another, and the surviving records do not let us assign a universal weight to each. Nor should later history be used to deny earlier benefits. A society can improve some people's lives for a long period without securing those improvements forever. Duration is evidence of accomplishment; it is not proof of immunity. There is a reverse error too: assuming that a system's persistence demonstrates that ordinary people accepted its distribution as fair. Persistence may reflect consent, limited alternatives, coercion, divided interests, or the difficulty of organizing change. Several can coexist. Describing how an order reproduced itself does not establish that it deserved to reproduce itself. Explanation and approval answer different questions, even when the same evidence informs both.
For an account of prosperity, that separation creates room for better judgment. We can ask what made cooperation possible, what it cost, and whom it served. We can identify a useful capability without treating the whole social order as a package that must be endorsed. We can also recognize a severe injustice without pretending that removing it would automatically solve every coordination problem. Moral clarity should make the practical inquiry more exacting.
The household beyond the monument
Rome cannot supply a ready-made design for a good society. Its circumstances, evidence, and institutions are too distant and too varied for that use. It can teach a more demanding way of asking questions. Material provision is real. Its distribution is an additional fact. Control over one's life is another. The ability to maintain services across shocks is still another. None becomes unnecessary because one of the others is impressive.
The most consequential unit of observation is often the person who is least visible in the record of achievement. What did the service allow that person to do? What could they expect next month? Could they retain a gain, refuse a demand, or recover from a loss? Where evidence cannot answer, the proper response is an acknowledged gap rather than an invented ancient household. An unanswered question still helps reveal what our preferred measures leave out.
Prosperity seen from the ground requires this movement between scale and experience. A society needs productive capacity large enough to make better lives possible. Households need dependable ways of reaching that capacity. The mosaic can survive when the opportunity has vanished. The task is to understand the work that once connected the two, and the terms on which people were permitted to share in it.
Notes and Sources
[1-1] Parco Archeologico di Ostia Antica, “Terme di Nettuno,” official archaeological site interpretation. Layout, construction history, and restoration. https://ostiaantica.cultura.gov.it/pannelli-didattici-e-mappe/larea-a-servizio-del-fiume/terme-di-nettuno/
[1-2] Pliny the Younger, Letters 10.23, to Trajan, Latin text and public-domain English translation, Roman Letters. A proposal concerning the baths at Prusa. https://romanletters.org/letters/pliny_younger/10023/
[1-3] Walter Scheidel and Steven J. Friesen, “The Size of the Economy and the Distribution of Income in the Roman Empire,” Journal of Roman Studies 99 (2009), 61–91, especially the methodological introduction and income-distribution discussion. A reconstruction using ancient evidence and comparative assumptions. https://www.cambridge.org/core/services/aop-cambridge-core/content/view/S0075435800000071
[1-4] Joseph R. McConnell et al., “Lead Pollution Recorded in Greenland Ice Indicates European Emissions Tracked Plagues, Wars, and Imperial Expansion during Antiquity,” PNAS 115 (2018), 5726–5731. See provenance, conversion uncertainty, historical linkages, and methods. https://archaia.yale.edu/sites/default/files/files/McConnell%20Lead%202018.pdf
[1-5] Gaius, Institutes 1.9 and 1.52–53, English text at The Latin Library. Legal categories and restrictions on masters' conduct; the text does not establish enforcement rates. https://www.thelatinlibrary.com/law/gaius1.html