Kumu / Micro/Macro Econ / Perfect Competition / Stage 1

Stage 1 — Engagement Brief

Delivered so far: the workspace. Nothing has been advised yet — this is where the question gets written down, before anybody looks at an answer.

Objective

One page. State the farm's problem in your own words, and end on a hypothesis you can be shown wrong about — written and committed before the workbook exists.

It is not analysis. It is the condition that makes the analysis worth anything.

Where the numbers are: the scenario data is published on the case page under The givens — bed caps, revenue per bed, hours, fertilizer, the diminishing-returns rates, and the labor formula. Your brief restates the problem, not the numbers.

Before you start: your portfolio repo needs to exist, with docs/briefs/ in it. That is Stage 0 — the course's opening step rather than part of this case — and it takes about forty minutes.

Learn — why a hypothesis has to come first

A prediction written before the model runs is falsifiable. The same sentence written afterwards is a summary of what the model said, and it teaches you nothing, because you can no longer tell the difference between "I understood the economics" and "I read the answer cell." Stage 3 asks you to hold your prediction up against what the model found. That comparison only exists if you commit to a number now.

This is not a classroom conceit. It is why a consultant writes an engagement brief before opening the data, and why "we always thought so" is the least trustworthy sentence in business. A wrong hypothesis, precisely reasoned, is worth exactly as much as a correct one — and far more useful to you than a lucky one.

What makes a hypothesis genuine

"I expect roughly 15 tomato beds, 20 carrot, 25 mesclun, because tomatoes earn four times what carrots do per bed and I doubt the labor penalty eats that gap." — specific, wrong in an interesting way, and testable. Compare: "I expect a balanced mix because diversification reduces risk." That sentence would survive any outcome, which is exactly what disqualifies it.

Do — write the brief

Create docs/briefs/perfect-competition-brief.md. Read the case scenario first — the assumptions, the prices, the caps are all on the case page and in the stage brief your instructor handed out. Then, before touching Solver or the workbook, write half a page to a page:

  1. The problem in your own words. What is the farm deciding? What is fixed, what is chosen, and what limits the choice? If you can't state it without re-reading the case, you don't have it yet — and that's useful information about where to spend the next hour.
  2. Your hypothesis. "I expect the optimal mix to be X because Y." Real numbers — beds of tomatoes, carrots, mesclun — and real reasoning: prices per bed, labor intensity, the diminishing-returns percentages, the caps. Name the mechanism you think will decide it.

The shape of a brief — the four sections, and the frontmatter that goes at the top — is on Deliverable templates. Copy the structure; the content is yours.

Commit and push with a message that says what changed: Add perfect-competition brief with mix hypothesis. Not update.

AI boundary for this stage: AI may explain the case's economics or push back on your reasoning. It may not write the brief. A hypothesis you didn't generate can't be honestly compared against results in Stage 3 — and the comparison is the whole point.

Written it? Now have it attacked

The most useful thing a model can do with a brief is refuse to be impressed by it. Once your draft is written and committed, paste it in and ask for the holes. The instruction not to rewrite is the load-bearing half — without it you get back a better brief, and you have crossed the boundary while trying to respect it.

Here is my engagement brief. Do not rewrite it and do not suggest
replacement wording.

1. Name every assumption I left implicit.
2. Name every claim I have not supported.
3. Ask me the three questions a client would ask.

Is my hypothesis falsifiable — could this model actually show it wrong,
or would any outcome be consistent with it?

That last question is the one worth paying for. A hypothesis that survives every possible result is the most common defect in this stage, and it is much easier to see in someone else's sentence than in your own. The critique loop is the general version of this habit.

Deliverable

WhatWhere it goes
The briefdocs/briefs/perfect-competition-brief.md

Check yourself