Tool/Mental Model/No. 0069
Backcasting
Backcasting is a planning method that starts with a desired future and works backward to find the steps needed to reach it. John B. Robinson named energy backcasting in 1982. Unlike forecasting, which asks what is likely, it asks what must happen for a chosen outcome.
- Evidence
- Useful, modest evidence
- Read
- 6 min
- Links
- 11 connections
01You've seen this when…
- in life
You want to graduate next June. Checking backward from graduation reveals that your final course requires a prerequisite offered only this fall. The prerequisite’s fall-only schedule makes registration the next move, however hard you study.
- at work
Your team wants new customers to finish setup without calling support. Before that can happen, customers need to import their own data. Before that, someone must untangle the permissions that currently require an administrator.
- out in the world
A town wants children to be able to walk safely to every school. Planners start with the routes children would need, then identify the crossings, street changes, and budget decisions that must come first.
02The idea
A graduation date can look comfortably distant until you discover a prerequisite that must be completed this month. Backcasting makes those hidden dependencies visible before they become emergencies.
Start by describing a future condition you want. Then ask what must already be in place immediately before it can exist. Repeat the question for each prerequisite until the chain reaches something you can do now. Finally, walk forward through the chain to check whether the route holds together.
The direction matters. Planning forward from today often produces a better version of what already exists. Starting at the destination gives you permission to consider changes that current habits and budgets leave out. It can loosen status quo bias, without making every ambitious destination attainable.
Backcasting is not forecasting. A forecast asks what is likely to happen; backcasting asks what would have to happen to reach a chosen outcome. The methods complement each other. Forecasts help test whether your proposed route is plausible.
It also differs from scenario planning, which explores several possible futures, and a pre-mortem, which starts with an imagined failure. Backcasting starts with a desired condition and builds routes toward it. The result is a conditional plan. Predictions about success and claims that your actions will cause it need separate evidence.
03How to use it
- Describe an observable finish line. Set a date for reaching your desired condition within the constraints you must preserve. Specify the change in what people can do or experience. A new training platform is an output; employees performing a task competently is an outcome.
- Sketch more than one route. There may be several ways to reach the same condition. Separate the goal from your favorite solution. Use constraint relaxation to ask which restrictions are physical or legal and which are merely customary.
- Identify the last necessary conditions. Ask what must be true immediately before success. Distinguish prerequisites from things that would merely help. Record assumptions about customers, partners, regulations, or technology that you do not control.
- Keep stepping backward. For each condition, identify its own prerequisites. Continue until you reach actions available today. Branch where work can happen in parallel. Mark long lead times and the critical path, the dependency chain that sets the earliest possible completion date.
- Check the route against reality. Walk forward through it. Test costs, capacity, sequencing, and external dependencies. Use reference-class forecasting to compare timing with similar efforts. Keep inconvenient estimates intact; if the route cannot fit, change the date, the approach, or the goal.
- Commit to first moves and review points. Give the next actions owners and dates. For major assumptions, specify what evidence would support them and when you will check. Keep later stages conditional so new information can change the route. Where possible, preserve options by deferring some commitments.
A practical output is a one-page map, with dependencies linking today’s actions on the left to the destination on the right.
04A worked example
Consider a fictional transit agency that wants all 60 buses on its busiest routes to be electric by the end of 2032, without reducing service. Its initial plan is to replace diesel buses as they retire.
What it looks like A purchasing problem. Once the agency chooses a model and finds the money, it can place orders over several years.
What’s actually going on Buying buses is only one dependency. Working backward exposes the rest:
- By the end of 2032, the fleet must sustain regular service. The agency needs suitable vehicles and reliable charging, with trained drivers and mechanics to run the fleet and enough operating reserve for breakdowns.
- During 2030–2032, deliveries and training must be phased together. The pace of replacement is limited by the agency’s ability to charge and maintain the new buses.
- Before the first deliveries in 2030, the depot must be ready. That requires electrical capacity, charging equipment, permits, and a construction window that keeps current operations running.
- In 2027, the grid work may need to begin. The utility estimates a three-year lead time for the upgrade. The date on the map rests on that estimate, which needs confirmation.
The agency now identifies a utility study as its earliest urgent action, giving it priority over a bus order. A small route trial can run alongside that study to test winter range and charging requirements before the full fleet commitment.
What made it work The destination included a service requirement, so success required both owning 60 electric buses and sustaining service. The backward chain exposed infrastructure work that ordinary replacement planning missed. Walking forward then separated work that could run in parallel from work that had to wait.
05When to reach for it
06When it misleads
- The destination is mistaken for a forecast. Estimating a preferred future’s likelihood requires evidence beyond a coherent story about reaching it. Keep separate estimates of what you want and what current evidence makes plausible.
- The chain contains wishes disguised as prerequisites. A plan that requires affordable technology or supportive legislation is conditional on those developments. Give such assumptions tests, alternatives, and decision points.
- The goal is too narrowly specified. Electrifying buses while cutting service could undermine the original purpose. Add guardrails and examine second-order effects, including costs shifted onto other people.
- One route becomes the only imaginable route. A neat backward chain can hide alternatives. Compare routes before treating any sequence as necessary.
- The map becomes more precise than the knowledge. Detailed dates ten years out can create false confidence. Specify near-term actions closely and distant stages more loosely. Revisit both the destination and the schedule when needs or conditions change.
07Roots
In the early 1980s, Canadian researcher John B. Robinson was examining energy futures. Energy planners could project rising demand and calculate the power stations needed to meet it. People wanted energy because it kept rooms warm and enabled transportation and useful work. Different ways of delivering those services could produce very different futures.
Robinson described energy backcasting in a 1982 paper in Energy Policy. Planners could specify a desirable future energy system independently of projected demand and analyze the changes needed to reach it. Earlier normative forecasting had also worked from preferred outcomes; Robinson gave the backward approach a name and a policy method.
The technique spread through sustainability planning, where extending current trends often leads somewhere nobody wants to go. Later researchers brought stakeholders into the process and broadened the task of defining the desired future beyond a single analyst. Residents and businesses could join public agencies in defining the destination and challenging each other’s assumptions. Backcasting became a way both to draw a route and to expose disagreements about where that route should lead.
08How solid is this?
Backcasting has a substantial methodological and case-study literature, especially in energy and sustainability planning. The literature supports its use as a planning framework. A general claim that backcast plans succeed more often than other plans would require comparative evidence.
09Connections
- Often confused with Scenario Planning, Pre-Mortem, Inversion
- Helps counter Status Quo Bias
- IncludesWorking Backwards
- See alsoReference-Class Forecasting, Theory of Change, Critical Path, Constraint Relaxation, Second-Order Thinking, Real Options
+ 1 more in the list
10Origin and sources
John B. Robinson named and described energy backcasting in 1982, building on earlier normative-forecasting approaches that began with preferred futures.
- [1]Robinson, J. B. (1982). Energy backcasting: A proposed method of policy analysis. Energy Policy, 10(4), 337–344.
- [2]Dreborg, K. H. (1996). Essence of backcasting. Futures, 28(9), 813–828.
- [3]Quist, J., & Vergragt, P. (2006). Past and future of backcasting: The shift to stakeholder participation and a proposal for a methodological framework. Futures, 38(9), 1027–1045.
Suggest an edit· Updated 2026-10-02