A bank can announce a rate; it cannot announce the conditions in which the rate will work. Four percent can be an invitation to borrow in one economy and a formidable obstacle in another. The difference depends on what a venture is expected to earn, what households wish to postpone, how prices are expected to change, and what resources are available to meet the resulting expenditure. Wicksell gave this comparison a lasting name: the natural rate, against which the money rate charged by banks could be judged. It was a way to make the monetary quotation answerable to the economy beyond the quotation. Its difficulty has survived its fame. The bank can observe the number it sets and the consequences that follow; the benchmark connecting them has to be inferred, and the consequences arrive amid many other changes.
Wicksell began with a rate on loans that would leave commodity prices without a tendency to rise or fall. He related it to the return that would emerge if capital were lent in goods, without money interposed between those who could postpone their use and those who wished to employ them. His illustration then brings money and banks back into the account. Entrepreneurs borrow to obtain labor, materials, and the means of supporting production until its proceeds arrive. When the terms of the loan correspond to the relevant conditions of production, that financing creates no continuing pressure on the general price level. The benchmark belongs to a relation among returns, costs and available resources. It is not an additional quotation hidden in a bank's books, waiting for a sufficiently diligent official to uncover it.
A maintained discrepancy can give expenditure a continuing push. If banks lend below the relevant normal rate, entrepreneurs find an inducement to enlarge their undertakings. Their bids for workers and goods increase, and, under Wicksell's assumptions, prices rise. The same inducement remains when the next production period begins, so the pressure continues while the discrepancy is maintained. Lending above the normal level works in the opposite direction. This is the cumulative process: an ongoing movement arising from the relation between financing terms and the opportunities they finance. The mechanism explains why a cheap loan can be consequential beyond the fortunes of its immediate borrower. Credit changes the claims made on resources, while the resources themselves must still be produced or drawn away from another use.
The natural rate moves with the circumstances against which the loan is being compared. Wicksell made this qualification central to the argument. Production methods, available capital, labor, land and the wider economic position can change the benchmark; the quoted bank rate can follow in separate, hesitant steps. A rise in the money rate may therefore leave credit relatively easier if the relevant return has risen further. Conversely, the same quotation can become more restrictive as productive opportunities deteriorate. The absolute number supplies too little information. Calling a rate historically high or low compares it with earlier quotations; calling it expansionary or restrictive compares it with a counterfactual for the economy now. The first comparison is readily displayed on a chart. The second requires an account of what else has changed.
Wicksell's banks also operate within arrangements that qualify the image of an unlimited monetary power. An individual bank faces rival banks and withdrawals; a country under an international metallic standard faces movements across its border. A whole system of bank credit presents a different experiment from one bank departing from the rest. Wicksell follows these differences, and at the chapter's close he qualifies even the proposed equality of natural and money rates. Depending on the definitions, risk can require a persistent excess of the capital return over the loan rate. These details matter because the benchmark is often made to do more than the theory warrants. A comparison between a risky venture's expected return and a safe short-term rate cannot be repaired simply by calling one of them natural.
The modern star names a related family of benchmarks, whose meaning depends on the model used to estimate them. Holston, Laubach and Williams define a natural real short-term rate consistent with output at its natural level and constant inflation. Their 2016 model concentrates on persistent movements, estimating that rate together with potential output and its trend growth from output, inflation and interest-rate data. Some transitory disturbances are deliberately left outside their measure of the natural rate. A model concerned with the rate that would stabilize conditions in each individual period can therefore produce a different series without having asked the same question badly. Before an estimate can guide a decision, the decision-maker must know which counterfactual the estimate describes and over what horizon it is intended to hold.
The comparison must first put the quoted rate and the benchmark into the same units. Suppose, in an invented one-year example, that the nominal rate is four percent and the assumed neutral real rate is two percent. If the price level is expected to rise three percent over that year, the projected real return is approximately 0.97 percent; if the expected increase is one percent, it is approximately 2.97 percent. The same quotation falls below the assumed benchmark in the first case and above it in the second. The exact calculation divides the growth of the monetary claim by the projected growth of prices. Merely subtracting expected inflation gives the familiar approximation. Either way, the rate named at a press conference cannot be placed beside a real benchmark without attending to the purchasing power in which the comparison is made.
Estimation then adds a second uncertainty to the uncertainty about future prices. Potential output is itself unobserved, and so is the distinction between a persistent shift in demand and a temporary disturbance. A weak economy can reflect financing that restrains expenditure, a loss of productive capacity, or a combination that changes over time. A statistical model separates those possibilities by imposing relationships and asking how well they explain several observed quantities together. Holston, Laubach and Williams make those choices explicit, including restrictions on the relation between trend growth and the natural rate and on the responses of output and inflation. Their estimates remain imprecise even with hindsight. The star helps organize the inference; attaching it to a variable does not turn the inference into an observation.
Later information can change the apparent location of yesterday's benchmark. An estimate made using data available at the time differs from one permitted to use what happened afterward, and both can depend on parameters estimated over a wider sample. This creates an asymmetry between decision and retrospective judgment. The official must act before all the evidence exists; the later analyst can easily describe the earlier choice as a deviation from a benchmark that was reconstructed after the event. That does not excuse every mistake. It changes what a serious evaluation must examine: the information available, the alternatives considered, the sensitivity of the conclusion to plausible assumptions, and the speed with which policy responded when its account of the economy ceased to fit.
A useful benchmark disciplines policy by making its explanation testable against more than one convenient symptom. Rising prices alone cannot reveal a unique natural rate: imported costs can change, productive capacity can be disrupted, and expectations can move. Nor does weak expenditure by itself identify which part of the economy's capacity remains available. The authority must connect its reading of inflation to employment, production, credit and the likely persistence of the disturbance. A benchmark that explains every possible outcome after the fact would offer little discipline. A model that states its assumptions, exposes its uncertainty and changes its conclusions when the evidence changes can offer more. The public question becomes whether the bank's account survives those comparisons, including the possibilities that would make its chosen course costly or mistaken.
The word natural supplies no independent verdict about who deserves the return. A rate consistent with stable inflation in a model can still operate in an economy with unequal wealth, unequal access to credit and opportunities that some people inherit while others must finance. The model's stability condition takes those circumstances as part of the world it describes. It does not pronounce them fair. Likewise, neutral monetary policy can have different consequences for an indebted household, a deposit holder and a worker seeking employment. Neutrality names a relation to the model's aggregate benchmark; it does not promise that everyone is unaffected or that every claim on future income has been earned. The distinction keeps the volume's moral question alive precisely where a reassuring adjective might otherwise bring it to an early close.
The bank can announce its instrument; the conditions against which it acts must be inferred.
What, then, is a central bank steering by when it steers by the natural rate? By a counterfactual made explicit enough to guide an argument, estimated from imperfect evidence and corrected against the economy's response. The benchmark can prevent the bank from mistaking yesterday's quotation for today's appropriate stance. Its uncertainty requires judgment, and that judgment remains a public act with consequences for people who did not choose the model. Wicksell made the money rate answerable to production and prices; modern estimation makes the proposed answer inspectable without making it certain. The next plate follows the responsibility back to the committee that must decide. An invisible benchmark can constrain a choice. It cannot relieve the chooser of having made it. The argument remains open.