The Quasi

Begin with a machine already built. A manufacturer has ordered a press, paid for it, and installed it on the factory floor; the question of whether to build this particular press belongs to the past. Tomorrow's question is whether to run it. The distinction opens Alfred Marshall's great refinement of rent, for the sum already spent and the expense of making another day's output play different parts in that decision. Orders may crowd the factory and leave a handsome surplus after the day's bills are paid. They may dwindle until the press earns scarcely anything beyond the cost of keeping it at work. In either case, its historical purchase price cannot be recovered by wishing the purchase undone. The machine stands there, and the return it can command depends on the uses still available to it. This is the middle country of rent: the things human beings have made, whose quantity cannot quickly increase, and whose income has to be understood across two horizons. The short run begins with the capacity in place. The long run asks what will persuade anyone to place more of it there.

Marshall called the income of such an existing appliance a quasi-rent, rent-like because its supply is fixed for the problem under consideration. The resemblance to land lies in the adjustment that cannot yet be made. A surge in demand for printed matter cannot summon a new press from the manufacturer overnight; while the stock of presses changes slowly, the available ones can command a larger residual. But a fixed number of machines does not imply a fixed number of copies. The owner may add a shift, work the press harder, change the paper, or move it to another job, and these choices bring their own costs. The relevant surplus is what remains after the expenses and alternatives that still bear on the present use have been allowed for. Marshall's account therefore requires attention to the particular margin. The building is in place; the coal is still to be bought. The press exists; the printer can still choose where and how intensively to use it. Fixity places some choices behind the decision-maker while leaving others fully alive.¹

The day's residual can be positive while the investment as a whole remains a loss. Suppose an installed press brings in 100 for a day's work and requires 70 in expenses that can be avoided by leaving it idle. If no better use is available and operation sacrifices nothing else, running it leaves 30 toward the cost of the capacity. The owner may need 40 a day, on average over its life, to recover the original outlay and the ordinary return on the money invested. Today's operation still improves the owner's position by 30 compared with idleness, though it falls short of that longer accounting. A resale price, an alternative lease, accelerated wear, or an avoidable maintenance bill changes the comparison because something can still be gained or lost by the present choice. The useful rule concerns costs that the decision can alter. A cost described as fixed need not be irrecoverable, and a machine's original cost need not be the value sacrificed by using it now. The distinction protects the firm from abandoning worthwhile production merely because the past turned out badly, while refusing to treat every expense attached to a durable asset as safely buried in the past.²

Time alters the question because the next press has not yet been bought. A prospective investor compares the outlay with the expected stream of future residuals, allowing for their timing, the life of the machine, and the risk that orders will disappoint. Those residuals must be expected to repay the investment on acceptable terms or the new capacity will not be supplied. What appears as a quasi-rent on installed equipment can thus be part of the return required to call the equipment into being. Marshall makes this point explicitly: the expectation of future quasi-rents is a condition of investment. It also corrects an easy mistake about competition. If entry and expansion bring exceptional earnings back toward a normal return, the machine's gross residual above operating expenses need not disappear. That residual may still have to cover capital recovery and the normal return over the asset's life. What competition can erode is the excess above the full cost of supplying capacity, under conditions that permit comparable capacity to enter. The quasi-rent is not simply another name for profit above every cost.³

The length of that passage is an economic circumstance, rather than a fixed number of months on a calendar. A spare machine may be brought into service quickly; a new plant may require years; a capacity dependent on a specialized network may be difficult to reproduce at any scale. Demand may also move before the new supply arrives, leaving the expansion to meet a market that has already changed. This is why the clock in Marshall's argument measures the possibilities of adjustment. It does not promise a date on which every profitable position becomes ordinary. Nor does the permanence of a site guarantee a permanent income of unchanged size: demand can leave a fixed location, competing sites can become more accessible, and the work once done there can cease to be wanted. The distinction concerns the supply response available within the chosen horizon. An asset's return may be strongly rent-like today and closely disciplined by replacement tomorrow, while another retains an advantage that no nearby substitute removes. To understand the return, one must name the capacity that is fixed and the adjustment that could loosen it.⁴

The distinction becomes especially consequential when someone proposes to take the residual away. An unexpected levy on a truly fixed stock, leaving every relevant decision about its use unchanged, can initially fall on its owners. Marshall's printing-press example then lets the old presses wear out. When replacement becomes a choice, the expected levy enters the comparison that determines whether another press will be bought, and the adjustment can reach output and its price. The example does not establish that every tax on an asset reduces investment; its result follows from what is taxed, what can change, and what investors expect. It does establish that an income's rent-like character after an investment has been sunk does not by itself settle the consequences of a rule anticipated before the next investment. The same problem arises in a bargain over specialized capacity: the person who has already incurred an irreversible cost can be vulnerable when the terms are reopened. How the return is shared after construction, and what sharing arrangement can support construction in the first place, are connected questions.⁵

The machine's return therefore carries a history that the day's accounts cannot disclose on their own. Some of it may repay a patient and hazardous investment; some may reflect an unexpected shortage; some may survive because rivals are prevented from supplying an alternative. The duration of the surplus helps frame the inquiry, but duration alone cannot tell us which history is true. A long-lived return can finance a long-lived asset, while a brief windfall can arrive without any foresight at all. The task is to follow the return across the investment, the available substitutes, and the conditions of entry, keeping the cost of creation distinct from the subsequent power of possession. This is the refinement Marshall gives the volume's suspicion of rent. A payment can be unnecessary to induce the creation of a machine already standing before us, yet belong to the expected reward by which machines are created. The next plate turns back to the ground and Henry George, where the claim to take the rent rests on a different premise: the underlying site was not waiting for its owner to bring it into existence.