A free interactive teaching and learning resource for secondary economics students and their teachers.
Economics Pad exists because no existing resource offered what teaching this subject actually requires: a dynamic, interactive space in which the relationships between economic variables can be explored and tested, rather than simply labelled on a static diagram. This site is an attempt to build that space.
The tools are designed primarily for students studying CAIE 0455 iGCSE Economics and CAIE 9708 AS and A-Level Economics, though they are relevant across syllabuses and levels.
Three interactive tools are available now — free, browser-based, no login or installation required.
An interactive market equilibrium model, first encountered in 0455 iGCSE. Adjust the determinants of supply and demand and observe real-time effects on price and quantity.
An interactive macroeconomic model of Keynesian aggregate demand and aggregate supply. The presentation of short and long run AS is particularly useful for observing the effects of transitions in spare capacity.
Build a diagram from scratch. Choose curves, label intersections, and export as a standalone interactive file. Intended primarily for educators building scenario-specific diagrams to give to students as interactive models rather than static ones.
The following are in development. Content is not yet published, but these are the directions the platform is heading.
Pre-built interactive diagrams showing economic scenarios for 0455 and 9708 students, designed to develop A01, A02 and A03 skills.
Guided exercises in which students move curves in response to economic events, then check their reasoning against a model answer.
Topic-by-topic resources and interactive models tailored to the CAIE 0455 and 9708 specifications, developed and refined with student feedback.
Economics Pad was founded and built by Frederick Graham, a teacher of CAIE 0455 iGCSE and 9708 A-Level Economics at the British International School of Bucharest. Frederick holds a First Class degree in Economics (University of Buckingham), a Postgraduate Diploma in Applied Economics with Merit (University of Nottingham), and a Postgraduate Diploma in Education with QTS (University of Birmingham).
The platform grew out of classroom practice and practitioner research into the use of dynamic interactive tools in secondary education. It is intended as a resource for teachers building interactive teaching materials, and for students who want to engage with economics as a living subject.
Move the sliders for the non-price determinants of demand and supply to shift the curves, click and drag either curve, or grab points A and B to change a curve's slope. Previous positions are kept as faded shadows.
Move the sliders below to see how each determinant affects the steepness of the demand curve. A steeper curve = more inelastic demand.
Move the sliders below to see how each determinant affects the steepness of the supply curve. A steeper curve = more inelastic supply.
If you used this in a lesson or for studying, please let me know how you used it so I can work on improvements.
Open feedback form →Move the sliders for the components of AD and the factors of production for AS to shift the curves. The previous positions are kept as faded grey shadows so you can compare.
If you used this in a lesson or for studying, please let me know how you used it so I can work on improvements.
Open feedback form →Click chart to place. Hover over a dot and click × to remove it.
Position curves freely, then click Set to start recording ghost shadows.
Standalone interactive HTML file.
Pre-built interactive diagrams with exam-style questions. Each scenario includes A01, A02 and A03 practice questions.
Guided diagram exercises with model answers. Download the interactive diagram and work through the questions independently.
Topic-by-topic resources for Cambridge IGCSE (0455), Cambridge A-Level (9708) and other syllabi.
Economics Pad is a tool built out of classroom practice. The ideas on this page are things being actively thought about, not things that exist yet. They are published here in the spirit of transparency, and because the questions they raise are interesting, at least to me.
If you have thoughts on any of the ideas, the contact details are on the Home tab.
It’s nice having a space to ramble ideas, I may move this to a separate “blog” space.
The problem as I see it.
Generic AI programs are a useful and very powerful tool. They are also limited in ways that matter for economics education specifically. They are not inherently good teachers, but they can often become good teachers.
They do not reliably understand the scope of a syllabus. They do not understand the nuances of how arguments are formed and assessed within a specific specification. They do not reliably understand what “correct” looks like on a CAIE 0455 or 9708 mark scheme outside variations of indicative content given on the MS, or why a curve moves in one direction and not another without potentially going far beyond the scope of the syllabus in question.
These are not small gaps. For a student trying to check whether their diagram is correct, or whether their argument is well-formed, a generic AI response is at best partially useful and at worst actively misleading.
One direction I’m interested in exploring is the development of what might be called an AI SKAE Lab — AI Subject Knowledge and Application Enhancement.
The core idea: rather than asking a generic AI to teach economics, give a carefully constructed knowledge file — a SKAE pack — to an AI at the start of a session. That pack begins with the syllabus scope, correct diagram logic, common student errors, mark scheme language, and how arguments should be structured for a specific topic. It then goes through much the same process as a student does with a teacher, exploring scenarios, answering questions, receiving feedback and being instructed in meta-cognition for future responses. After some history of being “taught” by a human teacher, the pack would be given to a “fresh” iteration of the same AI and AfL and AoL of its performance would be measured in practice, and again fed back to it.
Once broadly capable, the SKAE pack would be published to be given to an AI for a student to use with the same area of subject matter within the same syllabus. Crucially, any response the AI gives would carry a clear disclaimer distinguishing between knowledge and advisable teaching behaviour drawn directly from the SKAE pack and inferences made by the AI beyond it. Students would know, at all times, which is which.
Economics Pad may be a useful canvas for this, especially for keeping initial experiments with this small-scale and within a testable environment. An SKAE pack is made for questions and materials within a syllabus and within concepts shown on a particular diagram (the AI is effectively taught how to teach as described above).
A student works through a diagram interactively, exports an image of their work, and submits it alongside a written explanation or question. The AI evaluates the diagram against the SKAE pack, and following the lessons “learned” by the base AI when it was taught the SKAE pack, responds to the student.
The visual interactive element of Economics Pad addresses something in particular that current AI programs struggle with: spatial reasoning about curve movements and equilibrium shifts. Students will use AI, but I keep seeing my students using malinformed AI for purposes such as these. This also makes developing and testing this concept of exportable SKAE packs to give AI teaching skills more realistic for starting concepts in a micro setting (no pun intended).
This is a self-study tool in concept — not a replacement for teaching, but a resource for independent practice outside the classroom to help address the situation that AI is already being used widely in this way.
One question worth addressing directly: the SKAE pack does not need to be human-legible. The pack is optimised for the AI that reads it, not for a teacher or student. We use oil for energy, but we don’t drink it.
This removes a constraint that was only there for the wrong reasons. The disclaimer system is behavioural — the AI is instructed to flag in its responses to students in terms of what is drawn from the pack and what it is inferring beyond it. That distinction is preserved in the output, not in the pack itself.
An idea that makes this more interesting throughout: at the end of a teaching session the AI that was taught condenses its own learning — in its own terms, optimised for a fresh instance of itself to read. A human writing “when a student does X, say Y” is an approximation of this; an AI writing it would do it precisely. The pack then contains not just rules but a record of worked reasoning, which is closer to how humans learn from case studies than from textbooks. Whether this produces meaningfully more consistent behaviour in the fresh AI is an open and testable question and the core of my idea here.
The intention is to start with something small and bounded. Supply and demand interactions for example (not supply and demand in general, but a specific set of interactions within a specific part of a specific syllabus). There are a finite number of determinants, a finite set of correct curve movements, a defined set of student errors, and a clear marking framework. These constraints make it testable. You can actually evaluate whether the AI’s responses are correct or not, which is much harder in an open-ended topic.
If that works, the natural progression is to extend to broader but still bounded areas in the syllabus (for example — from supply and demand movements to these movements including elasticity of supply and demand).
Each stage tests whether the pack architecture scales, not just whether it works in one case. Whether learned pedagogical behaviour is topic-specific or partially generalisable is one of the more interesting questions this could eventually shed light on.