Backcasting

Describe the future you actually want in terms of what people are doing, then step backwards in stages to today. The first condition that is not yet true is where the plan starts. It is the only futuring move that gives a direction, which makes it the most useful and the easiest to fake.

Futuring: the four moves

Move 3 of 4 · Turns a preference into a route

Name the future you want, then work back to what must be decided now.

Time: 2 hours 30 minutes
People: 4–10, including people who can make the first decisions
You need: A chosen destination scenario, a long wall or table for the ladder, the two worksheets
You leave with: A dated chain of conditions back to today, and two or three decisions with owners and dates

Needs some practice

The idea

What the move is, and why it works

What the move is

Forecasting starts from today and asks what is likely. Backcasting starts from a future you want and asks what would have to have happened to get there. You write the destination, then step back two or three years at a time, writing at each stage the conditions that must already be true. When you reach a condition that is not true today, you have found the start of the plan.

The output is a route, not a set of options. That is why it is worth doing, and why it needs the other moves around it. A route that only works in one future is a bet.

The idea behind it

Most plans extend what is already happening. Each year’s plan is last year’s plus a little. Backcasting breaks that by starting somewhere the present cannot reach by extension. Working backwards makes the gaps visible, especially the dull ones: the policy nobody has written, the meeting nobody has held.

Writing conditions, not actions, is the discipline. “Teachers moderate portfolios across departments” is a condition. “Run a training day” is an action. Conditions let the group see what has to be true. Actions come last, and only once the conditions are clear.

Why it is easy to fake

The commonest failure is a destination that is the current plan, achieved. The route then confirms the plan, and the method has decorated a decision already made. The test is simple. If the first step is something the school was going to do anyway, the destination was not far enough away.

The second failure is a destination written in values: students who are “responsible”, “future-ready”, “resilient”. You cannot work back from an adjective. Ask what a visitor would see people doing.

When

When to reach for it, and when not to

Use it when

The group agrees on where it wants to be, but the plan only extends what is already happening.
A vision statement exists and nobody can say what the school should do differently next term because of it.
You have scenarios and one of them is clearly the one people would choose.

Skip it when

The group does not agree on the destination. Run scenarios first, or name the disagreement.
The horizon is under three years. Ordinary planning is better and cheaper.
Nobody in the room can make any of the first decisions. You will produce a route for someone else.
The destination is fixed in advance by someone not in the room and cannot be questioned.

Before

Before anyone is in the room

Choose the destination scenario, and be ready to say why this one and not the others.
Set the end date and a decide-by date, the point by which the first decisions must be made.
Collect any existing vision or strategy statement, so you can see where the backcast departs from it.
Invite at least one person who controls a budget, a timetable or a policy.
Mark out a ladder on a long wall, from the end date on the left to today on the right.
Have the other three scenarios to hand for the final check.

Steps

How to run it

7 steps, about 2 hours 30 minutes in all. The questions in quotation marks are meant to be read out as they are.

01
Choose the destination and say why
15 min
Name the scenario or future you are backcasting from. Ask the group to say what they distrust about it as well as what they want from it. The preferred future deserves the most scepticism, because everyone says they want it.
Ask the room
- Why this future and not the others?
- What does this future cost, and who pays?
- Is there anything in it we would quietly not want?
Watch for Instant agreement. It usually means the destination is too vague to disagree with.
02
Describe the destination as behaviour
25 min
Rewrite the destination as things people are doing, seen by a visitor on an ordinary day. Students, teachers, parents, someone from outside. Strike out every value word and replace it with an action you could watch.
Ask the room
- What would a visitor see a student doing at 10 a.m.?
- What does a teacher spend their time on that they do not now?
- How would we know, without asking anyone?
Watch for Adjectives. “Students use AI responsibly” cannot be worked back from. “Secondary 4 students explain, unprompted, where AI helped and where it misled them” can.
03
Set the end date and the decide-by date
10 min
Put a year on the destination, usually five to ten years out, far enough that the present is not a constraint. Then set a decide-by date two to three years from now: the point by which the first decisions must have been made.
Ask the room
- How far out does the present stop being a constraint?
- By when would we have had to decide, for this to be possible?
Watch for An end date only two years away. Backcasting needs room. Below three years, use ordinary planning.
04
Step back in stages
40 min
From the destination, step back two or three years. Ask what had to be true by then. Write each answer as a condition, not an action, on the backcast ladder. Repeat until you reach the decide-by date.
Ask the room
- What had to be true two years before this?
- Who made that true?
- What would we have seen at that point, if it were happening?
Watch for Actions creeping in. “Train teachers” is an action. Ask what the training would have made true.
05
Find the first condition not true today
20 min
Walk back along the ladder to today. Mark each condition true, partly true or not true. The first one that is not true, reading from today outwards, is where the plan starts. Circle it.
Ask the room
- Which of these is true in our school now?
- Which is the earliest one that is not?
- Is that something we were already planning to do?
Watch for A first step the school was going to take anyway. If so, the destination was too close. Push it further out.
06
Turn it into decisions with owners
25 min
For the first untrue condition, list the two or three decisions that would make it true. Give each an owner and a date before the decide-by date. Write what you will stop doing to make room.
Ask the room
- What has to be decided for this to become true?
- Who can decide it, and by when?
- What do we stop doing to make room?
Watch for Decisions with no owner, or owned by “the leadership team”. Name a person.
07
Check the route in the other worlds
15 min
Take the first decisions and read them against the other three scenarios. Does the route still make sense if a different world arrives? Note where it would do harm, not only where it would fail.
Ask the room
- In which world would this first step backfire?
- What would we change about it so it does no harm in that world?
Watch for Skipping this step because the group is tired. A route that only works in one world is a bet, and you should know you are making it.

Worksheets

Two sheets to print and fill in

Each one prints on its own sheet of A4. The grey line under each is an example from the worked situation.

Futuring · Backcasting · Worksheet

Backcast ladder

Steps back from the destination to today in stages, so the first untrue condition stands out.

Destination

End date

Decide-by date

Date

	Year	What has to be trueA condition, not an action	Who has to make it true	What you would see	True today?Yes, partly or no
The destination
One stage back
Two stages back
Three stages back
The decide-by date
Today

Example 2028 | The school has decided who owns a student’s portfolio and who can see it by default | Senior leaders, with students | A written default, set to private | No

geraldajam.com/work/toolkit

Futuring · Backcasting · Worksheet

First decisions

Turns the first untrue condition into decisions with owners and dates, and checks them against the other worlds.

Destination

Decide-by date

Date

The destination, as things people do

What a visitor would see. No value words.

The first condition that is not true today

Decisions that would make it true, each with an owner and a date

What we will stop doing to make room

How the first decisions fare in the other three worlds

Example First condition not true today: no two departments describe a good piece of portfolio work in the same way.

geraldajam.com/work/toolkit

Worked example

Backcasting from the Flourishing Techno-human City to 2028

A horizon scan for Singapore’s schools

The flourishing scenario is the one everyone says they want, so it gets the most scepticism and is the only one of the four with a direction to work back from. The group took its picture of assessment: mastery portfolios and learner-owned records, portable credentials, AI helping with feedback but final judgement left human and spoken. We set the destination at 2035 and asked what an ordinary secondary school would have to have decided by 2028.

Strip the destination to behaviour. The scenario was written in warm words. We rewrote it as a scene: a Secondary 4 student moving to a new school brings six pieces of work she owns. A teacher there reads two, questions her on one in person, and accepts the rest. Every piece carries a note on what she would change and a comment from someone the work affected.
2035 back to 2032. For the scene to be ordinary in 2035, schools had to trust each other’s judgements of a piece of work by 2032. That needs teachers in different schools to describe good work in a shared way, and some cross-school moderation to be routine.
2032 back to 2030. Before shared judgement, each school needed students to own their records, with the school holding a copy rather than the key. Any analytics feature had to pass an ethics review with sharing set to private by default. The scenario itself shows a parent finding one bad default, a reflection shared with a whole cohort, and getting it changed.
2030 back to 2028. By 2028 the school must have decided four things. Which few capacities a portfolio is meant to show. That some pieces are witnessed in person, through a short spoken defence. Who owns the data and what is visible by default. And where the teacher time comes from, which meant deciding what to stop marking.
The first untrue condition. Reading from today, the first thing not true was the most boring. No two departments described a good piece of portfolio work the same way. So the first step was not buying a platform. It was a meeting of heads of department with three pieces of student work on the table.
The awkward part. The group runs no school, and much of the route depended on decisions above any school: national examinations, admissions, national data rules. The backcast was useful for showing which conditions a school controls and which it does not. The school-level ones were smaller and duller than the scenario suggested.

What came out of it

A ladder from 2035 back to today and four decisions due by 2028, the first of which is a moderation meeting. Checked against the other three worlds, the decision about data ownership turned out to matter most. In Velvet Cage the same portfolio, held by the system rather than the student, becomes a tracking record.

Signs

How to tell it is working, and when it has gone hollow

Working

The first step is something the school was not already planning to do.
The destination is written as scenes a visitor could watch.
Someone with a budget or a timetable leaves owning a decision with a date on it.
The route includes something the school will stop doing.

Gone hollow

The destination is the current plan, finished.
The ladder is a list of actions with years attached.
Every first step is a purchase or a training day.
Nobody checks the route against the worlds the group did not choose.

Variations

Other ways to run it

In 45 minutes

Skip the scenario. Ask the group for one scene they would want to see in five years, written as behaviour. Step back twice, find the first untrue condition and name one decision with an owner. It is rough, but it reliably shows the gap between the vision statement and next term’s plan.

On your own

Pick one thing you want to be ordinary in your classroom or department in five years. Fill in the backcast ladder alone, then show the first untrue condition to a colleague and ask if they agree it is untrue. Their answer is usually the interesting part.

With Three Horizons

Draw Sharpe’s three curves: today’s system declining, the future system rising, and the messy middle where the transition happens. Put the destination on the third curve and the first decisions in the middle. It helps a group see that some of today’s practice has to be let go, not just added to.

What to do with what you made

Take the first decisions to whoever owns them, with the ladder attached, and put the decide-by date in the diary. Before anything is approved, run the first decisions through the stress-testing move. A route that only works in one world should be chosen knowingly.

Stress-testing Run the first decisions through all four worlds before committing.The four Singapore scenarios Including the flourishing world this backcast starts from.Reading the disposition Check whether the route grows from what the school is already inclined to do.

Tools

Tools, templates and guides worth having

Published elsewhere, and worth the time.

Toolkit
The Futures Toolkit (Government Office for Science, 2017)
Includes a step-by-step backcasting template and guidance on running it with a group.
Method
Three Horizons (Sharpe, 2013)
One diagram that connects today’s system, the future one and the transition between them. Very usable in a staffroom.
Method
The Natural Step’s ABCD process
A widely used backcasting process from sustainability planning: awareness, baseline, a compelling vision, then steps down to now.

Reading

Where it comes from

Start with the first on the list.

Robinson, J. B. (1990). Futures under glass: a recipe for people who hate to predict. Futures, 22(8), 820–842.
The founding account. The title alone is worth quoting to a sceptical group.
Dreborg, K. H. (1996). Essence of backcasting. Futures, 28(9), 813–828.
When backcasting is the right tool and when it is not.
Sharpe, B. (2013). Three Horizons: The Patterning of Hope. Triarchy Press.
The three curves, and how to work in the transition between them.

Singapore