Why a Central Bank Needs More Than One Forecast

Central-bank alternative-scenario analysis is the subject here: the Federal Reserve's practice of putting conditional paths beside a baseline forecast so policymakers can reason about risks that the most likely path necessarily compresses. The central question is not whether the Federal Reserve can forecast the economy, but whether one forecast can carry the whole policy problem. This series will give readers a practical way to distinguish a baseline from an alternative scenario, follow the mechanisms that connect a risk to economic outcomes, and judge what the evidence does—and does not—show.

The problem with a single center line

A forecast is often presented as if it were a single answer. In monetary policy, that appearance is useful but misleading. The Federal Reserve's baseline is a modal forecast: the staff's considered view of the most likely path for variables such as economic growth, unemployment, and inflation. It is not a casual guess. The historical record describes it as the product of a large team, multiple models, substantial resources, and a detailed account of the economy.

That concentration is a strength. Policymakers need a coherent reference point against which to discuss interest rates and the economy. Without a baseline, every possibility would have equal visual status, and deliberation could become a catalogue of fears rather than a judgment about the most plausible outlook.

But monetary policy is also risk management. An interest-rate decision has consequences if the economy follows a less likely path: an unexpected collapse in demand, a supply disruption, a financial-market shock, or a sudden shift in inflation expectations. The policy question is therefore two-part. What seems most likely? And which departures from that path would matter enough that policymakers should understand them before they occur?

An alternative scenario answers the second question in a disciplined form. It is a structured point forecast built under different assumptions about how the economy evolves. The assumption might be a sharp slowdown in productivity, a rise in energy prices, a housing slump, or a change in financial conditions. The scenario is then carried through an economic model, with judgment where the model cannot settle the issue, to show how output, inflation, and sometimes the policy rate might differ from the baseline.

This does not make the alternative a rival forecast in the ordinary sense. Federal Reserve staff scenarios were generally shorter and less fully developed than the baseline, often relying on a single model supplemented by judgment. Their purpose was to illustrate a vulnerability or trade-off around the central projection, not to provide a second complete description of the economy.

The distinction matters because a central bank can be wrong in more than one way. It can misjudge the most likely path. It can also understand the likely path reasonably well while failing to prepare for a consequential departure. A baseline is designed to organize the first problem. A set of alternatives keeps the second problem in view.

There is another reason the center line is an incomplete container: the consequences of being wrong are not necessarily symmetric. A forecast error that moves inflation a little above target may call for a different response from an error that combines persistent inflation with falling output. A broad interval around the baseline can show that uncertainty is high, but it may not show which combinations of outcomes are connected by a common cause. For policy, the shape of the mistake matters, not only its distance from the center.

This is why alternatives belong inside deliberation rather than in a footnote added after a forecast is finished. They force a question about exposure. If the baseline depends on continuing confidence in financial markets, stable productivity, or well-anchored expectations, what would happen if that condition failed? The answer need not be a prediction. It can still change how policymakers assess the robustness of today's decision and what signals would matter tomorrow.

The first branch: gold, confidence, and payments

The logic becomes clearer in the Federal Reserve's earliest publicly documented use of the phrase “alternative scenarios.” At the April 1968 meeting, after the collapse of the Gold Pool and amid efforts to manage the private-sector price of gold, Robert Solomon presented the FOMC with two possible paths. One involved a return to international monetary stability if confidence in the existing gold price could be restored. The other involved renewed instability if the United States failed to correct its balance-of-payments problem.

The value of the example is not that policymakers could assign a precise probability to each outcome. The value is that the same present situation contained different mechanisms and different policy implications. A return of confidence would produce one monetary environment; a failure to resolve the payments problem would produce another. Describing only a central path would have hidden the condition that separated them.

This is the basic grammar of a scenario:

  1. Identify a change or risk that could alter the economy.
  2. State the conditions under which it occurs.
  3. Trace how it moves through spending, production, prices, financial conditions, or expectations.
  4. Compare the resulting path with the baseline.
  5. Ask what the difference means for policy.

The first two steps are essential. Without a specified risk, the exercise is merely a vague warning that “something could go wrong.” Without stated conditions, a reader cannot tell whether the scenario is a plausible branch, a stress test, or an intentionally extreme illustration. The last three steps turn a story into a decision-relevant comparison.

A conditional path is best understood as a counterfactual with its assumptions left visible. It does not say, “This is what will happen.” It says, “If this source of change materializes, and if these transmission relationships are a reasonable approximation, this is how the main variables could move relative to the baseline.” That wording preserves two kinds of uncertainty at once: uncertainty about whether the event occurs and uncertainty about how strongly the economy responds if it does.

The comparison also gives policymakers a common language for disagreement. One participant may think the event is unlikely but severe; another may think it is more likely but easier to absorb. They can still examine the same conditional path and debate its assumptions, magnitude, and policy consequences. A single forecast tends to compress those disagreements into one number. A scenario keeps the disagreement attached to the part of the reasoning where it actually belongs.

The 1968 case also shows why alternative scenarios are not limited to statistical uncertainty. A fan chart can show that outcomes tend to spread around a forecast. It cannot, by itself, explain why one branch would emerge from a crisis of confidence while another would emerge from a balance-of-payments failure. Those are different stories about the world, and therefore potentially different stories about monetary policy.

What an alternative scenario adds

The central bank's toolkit contains several ways to discuss an uncertain future, and they should not be collapsed into one another. A fan chart summarizes general uncertainty around a forecast, often using the institution's history of forecast errors to display intervals. A qualitative risk discussion names particular risks and explains their possible consequences without necessarily quantifying them. An alternative scenario gives one specified risk a quantitative, conditional path.

The difference becomes especially important when the future has distinct, plausible branches rather than one smoothly widening cloud of outcomes. Consider an inflation shock that could prove temporary or persistent. A fan chart centered on an average path may communicate the overall range, yet leave unclear why the two possibilities would lead to different decisions. The same problem appears when a central bank is weighing a demand shock against a supply shock: both may raise inflation, but only one may bring stronger output with it. The scenario is valuable because it preserves the identity of the shock while the model translates it into consequences.

A qualitative discussion can sometimes do this better than a numerical table when the mechanism is too novel or difficult to quantify. But qualitative language alone may not reveal how large the output or inflation trade-off could become. The tools are complements. A central bank can use a broad display to communicate the usual range, a written risk discussion to name what is unusual, and an alternative path to make one important mechanism concrete.

That division of labor makes the role of scenarios more precise. They do not replace a general measure of uncertainty, and they do not claim to enumerate every outcome. They make a selected mechanism visible. In a demand scenario, stronger spending might push output and inflation higher together. In a supply scenario, weaker productivity might raise inflation while reducing output. Both can be inflationary, but they create different trade-offs for a central bank.

The same principle applies to policy communication. A central bank may describe the expected direction of future rates, publish its own projected policy path, or show how rates would respond under a particular scenario. These devices answer different questions about the policy outlook. A general projection can clarify the central bank's current assessment; a scenario with a policy response can show how that assessment might change if a specified event occurred. Neither should be read as an unconditional promise.

A useful scenario portfolio is therefore not a long list of alarming possibilities. It is a small set of conditional comparisons chosen because each one reveals a material risk, a different transmission mechanism, or a policy trade-off that the baseline does not show by itself. Its value lies in the relationship between the branch and the baseline, not in the drama of the branch's title.

Selection is unavoidable. A central bank cannot publish a separate path for every factor that might matter, and including a risk gives it prominence that an omitted risk does not receive. The discipline is to explain why this branch deserves attention and what it is meant to illuminate. A scenario close to the baseline may help show a realistic way the outlook could drift; a more severe departure may reveal the cost of a risk that would otherwise be easy to underestimate. Neither calibration is automatically the right one. The reader needs to know what information the chosen distance is meant to provide.

The Federal Reserve's long historical record is useful precisely because it lets this distinction be studied over time. It contains scenarios that were modest variations on the central projection and others designed to expose a material break in normal relationships. Looking back can show whether the resulting set contained paths that resembled later developments. It cannot remove the uncertainty that existed when the scenarios were first prepared.

flowchart LR B["Modal baseline forecast"] --> J["Monetary-policy judgment"] R["Identifiable or structural risk"] --> S["Conditional alternative path"] S --> T["Economic trade-off"] T --> J B --> T

This relationship also sets the boundary for what follows. A scenario is conditional: if the stated assumptions hold, the model and accompanying judgment imply a path. It is not a prediction that the assumptions will hold, a probability assigned to the path, or a commitment to take a particular action. The Federal Reserve's historical study evaluates scenarios retrospectively, but even that evaluation asks how plausible their paths were after outcomes became known; it does not turn them into forecasts that were individually knowable in advance.

The practical judgment supported by this distinction is modest but important. When reading a central-bank forecast, do not ask only whether the center line looks reasonable. Ask what important branches are being considered, which mechanisms make them different, and what would change for policy if one of them emerged. The Federal Reserve uses alternatives because a baseline is the right anchor for collective judgment, but it is the wrong container for every consequential uncertainty.

The next question is how to tell general uncertainty from a named risk—and why that distinction determines whether a central bank reaches for a fan chart, a qualitative discussion, an alternative scenario, or a conditional policy path.

References