Subadditivity Effect

The subadditivity effect is a judgment pattern in which estimates for separate components sum to less than an estimate for the whole. It has been studied in probability judgments and does not mean that arithmetic itself changes.

A forecaster estimates a 60 percent chance that a shipment will be late, then assigns 15 percent to weather, 20 percent to staffing, and 10 percent to equipment. The causes may overlap or omit other delays, so their sum cannot be treated as a clean partition without checking the categories.

Suppose people estimate the chance that an event will occur, then estimate the chances of several possible causes separately. If the separate estimates add to less than the whole, some probability may have been left unassigned or the judgments may reflect different reasoning.

The reverse can happen in another setting: component estimates can sum to more than the whole. Such differences show why related judgments should be checked for consistency rather than assumed to fit together automatically.

The effect can arise when people consider categories one at a time and fail to include all possibilities, or when the whole and parts are judged with different information. The exact process depends on the task and how questions are framed.

For example, a planner may estimate the likelihood of delays from weather, staffing, and equipment, then compare their sum with the overall chance of a delay. If the categories overlap or omit causes, the arithmetic needs careful interpretation.

A public-health team estimates the chance of an outbreak, then estimates separate routes by which it might occur. If routes overlap, adding the figures double-counts cases; if a route is missing, the sum understates the whole. The category structure matters.

The concept does not establish that one particular estimate is correct. The total and components may each be uncertain, and a mismatch can reflect incomplete categories, overlapping events, or inconsistent judgment.

Define the whole event clearly and list the component categories. Check whether they are mutually exclusive when adding probabilities and whether they cover the relevant possibilities.

Use the same timeframe, assumptions, and evidence for the total and its parts. If estimates come from different sources or questions, note that before interpreting a gap as a meaningful effect.

For high-stakes forecasts, write down the assumptions behind each estimate and revise them when new information appears. A structured probability tree or table can make missing and overlapping cases easier to spot.

Consistency is a useful check, not a demand for false precision. The aim is to see whether the parts and whole tell a coherent story and to identify what remains unknown.

Psychological perspective

A whole event and its component explanations can prompt different mental tasks. People may focus on a few plausible categories, overlook others, or answer the questions using different assumptions.

Subadditivity is a pattern in estimates, not evidence that a specific forecast is false or that people cannot reason mathematically. Interpretation depends on how events are defined, whether categories overlap, and what information respondents received.

Thinking clearly

Define the event and its time period, then list components that are both relevant and clearly distinguished. Check whether the list leaves out plausible cases or counts one case more than once.

Use consistent evidence and assumptions when estimating the whole and its parts. Treat a mismatch as a prompt to inspect definitions before changing the numbers.

For consequential forecasts, show the categories and uncertainty explicitly. Use a probability tree or table when it helps, and avoid implying precision that the evidence cannot support.

What category might be missing or overlapping when you break a broad forecast into parts?