Kumu / Micro/Macro Econ / Imperfect Competition / Stage 2

Stage 2 — Analysis, Memo, Prompt Log

Delivered so far: a hypothesis on the record, an audited model of both markets. The client can now see the eighty-fold gap. What they cannot see is why it exists, what it costs everybody else, or what to do about it.

Objective

Optional · ungraded · self-paced

This case is an optional extension. Nothing here is submitted and nothing is due — self-paced, for students who want to run the argument out to its end.

Explain the gap in economics, using your own model's numbers. Put a dollar figure on what the pricing power costs society. Then write the recommendation to somebody who has to act on it.

Your model shows the same company earning $106.0M in one market and $8.785B in the other, from the same crop and the same fixed costs. That is the fact. The analysis is the reason.

Learn — the four things the analysis has to establish

Why marginal revenue sits below price

One more bag sold lowers the price on every bag, so the revenue the firm actually keeps from that sale is the price minus the discount it just gave every other buyer. For linear demand that works out to exactly twice the slope. The price taker's extra bag changes nothing at all, so its marginal revenue simply is the price: MR = P = $120.

Why the price comes off demand

MR = MC picks the quantity — 34,025,974 bags. Demand tells you the price buyers will pay for that quantity — $297.03. The test of your explanation is whether it would fix a classmate who just priced at $69.05. If it only restates the definitions, it would not.

What markup and Lerner measure

Price at 4.30 times marginal cost; roughly 77% of the price is margin over cost. The non-GMO row is 1.0× and zero. These are gauges of market power, not of efficiency or quality — and the patent is what keeps the demand curve downward-sloping in the first place.

What the deadweight loss means

About $2.99 billion a year, and it is not a transfer. Profit is a transfer: buyers pay it, the firm receives it, somebody still has it. Deadweight loss is the surplus from trades that never happen — farmers who would have paid more than marginal cost but less than $297. The monopoly withholds roughly 26M bags to hold the price up.

The result most students walk past

If the patent vanished and price fell to marginal cost, GMO seed would sell about 60.23M bags at roughly $121.46 — almost exactly the $120 that commodity non-GMO seed sells for. Strip the patent and GMO seed is just seed. That near-coincidence is the case's sharpest result, and an analysis that reports the deadweight-loss number without noticing it has done the arithmetic without reading the answer.

Do — write the analysis

Create analysis/imperfect-competition-analysis.md. Open by reproducing your Stage 1 hypothesis unedited, followed by an honest verdict: where it landed against the model, and what you misjudged. A prediction that missed by ten times with a sharp account of why beats a lucky guess with no account.

Then the four questions above, from your own workbook's numbers, plus a setup paragraph in your own words on why the firm is a price taker in one market and a price maker in the other, and what the patent has to do with it. Export at least two figures into analysis/figures/ — the D/MR/MC chart is the obvious first — and reference each in the text.

Cite the case's own figures rather than inventing statistics, and do not bring in outside numbers without a source.

Do — take a position on the patent

One paragraph, a position, defended. The material to argue with:

Either side is defensible. "There are arguments on both sides" is the one answer that earns nothing — the paragraph exists to make you weigh roughly $3 billion a year of vanished surplus against the incentive that produced the trait at all.

Do — write the memo and update the log

The memo goes in docs/decisions/imperfect-competition-memo.md — half a page to somebody who has to act, whether that is a regulator weighing the patent or an executive deciding what the pricing power is worth defending. Recommendation, the reasoning that drives it, the judgment call where the evidence ran out, and what would change your answer. The memo template has the structure.

Then update prompt-log.md at the repository root with a dated section covering both stages — the generation session and the formula debugging from Stage 1 belong here too — and close with a reflection of 300 words or fewer that names one AI error you caught and how you verified it. "The AI was great and made no mistakes" reads as not having looked.

Finally, update capabilities/pricing-power/README.md so its "exercised in:" line points at the analysis and the memo as well as the brief.

AI boundary for this stage: paste your draft and ask the model to attack the argument. Do not paste the questions and ask for prose — a model that has not seen your workbook writes about the textbook instead of your numbers, and it shows.

Deliverable

WhatWhere it goes
The evidence, with figuresanalysis/imperfect-competition-analysis.md · analysis/figures/
The recommendationdocs/decisions/imperfect-competition-memo.md
Sessions and reflectionprompt-log.md at the repo root

Check yourself