La Bourse / Volume IV / Nº IV.07 / Quiz, reflection, essays

A woman reading in a rattan chair beside a tall white window, light falling across her and the floor. — Plate IV.07 · Quiz, reflection, essaysThe Second Room.Everyone draws a supply chain as a line because a line fits on a page. The thing itself is a population, and it is lit from a room you have not entered.

ASSESSMENT · Chapter IV.07 — Supply Chains as Ecologies

Three instruments: a ten-point quiz, eight reflection questions, five essay prompts. The quiz checks comprehension rather than recall. The reflections are private and first-person. The essays are arguable from more than one side.

THE QUIZ — ten points

Four on recall.

1. Write the population expression for tier n and define each term.

N_n = N₁ · b^(n-1) — your direct supplier count multiplied by the branching factor raised to the tier below the first. N₁ is the number of firms you hold a contract with; b is the number of suppliers behind each of them. One mark for the expression, one for stating that b is the term that makes the word "chain" wrong.

2. Define the exposure ratio and say what it does to a supplier ranking.

E = the value of downstream output a node gates, divided by the node's own revenue from it. It is close to the inverse of spend share, so a list ranked by spend sorts exposures in nearly the reverse of the order that matters. Credit for naming Kraljic's bottleneck quadrant as the same finding in two-by-two form.

3. State the common-mode failure expression and say what q represents.

P(all m fail) = q + (1 − q)·p^m. q is the probability that every source fails together for a shared reason: one region, one port, one sub-tier supplier, one certification body, one sanctions regime. It is a floor that no number of sources passes.

4. What are the two levers that increase enforcement depth n, and which one is not* available by spending more on audits?

Lower the audit cost per node c — by buying inspections jointly with other buyers — and raise the leverage retention λ — by concentrating share, contract length and joint engineering in fewer relationships. Neither is an audit. Raising the audit budget alone moves C₁ in the wrong direction.

Four on application.

5. Your risk committee asks for "full visibility to tier four" across all sixty families. Answer them with a number and a counter-proposal.

Tiers one to four cost $44,928,000 a year — 12.48 percent of external spend — and return 21.6 percent of the fourth tier, because disclosure decays at roughly sixty percent a tier. The counter-proposal is to map the top decile by exposure ratio to exhaustion and the rest to tier one, and to state the denominator in the paper so the committee knows what is not covered. Full marks require naming the uncovered remainder explicitly; a coverage claim without a denominator is the failure this chapter is built against.

6. A colleague argues that because your critical family has four qualified suppliers, the risk is "effectively eliminated." What is missing?

The common-mode term. With p = 0.05 and q = 0.015, four sources reach 0.015006 — which is 0.000006 above the floor and 1.5 percent in absolute terms, not zero. The right question is what q actually is for that family: are the four in one region, one port, one sub-tier, one certification scheme? A strong answer notes that the fourth source bought 0.012 percentage points and cost $1,080,000 a year.

7. Two families. Family A: a $3.50 microcontroller, stoppage costs $40 million, four suppliers, all in one seismic zone. Family B: a $400,000 casting, stoppage costs $2 million, one supplier, twenty-year relationship. Which is mismanaged and how?

Both, in opposite directions. A has redundancy that does not address its real risk — q is high because the four share a geography — and the money would buy more by moving one source out of the zone than by adding a fifth inside it. B is fine on source count (a $2 million event is far below the $23 million breakeven for a second source) and its long relationship is the resilience mechanism, but it should be checked for recruitability: could anyone else make the casting if asked? Credit for noticing that the expensive part is the lower risk.

8. Your competitor's procurement director proposes sharing factory inspections. Your general counsel is nervous. Give the commercial case in three sentences.

Eight buyers sharing one inspection take the cost per node from $6,000 to $750, which moves the enforceable depth of our standard from 1.605 tiers to 2.198 — more than half a tier, bought with a phone call. We are currently paying separately for the same inspection of the same factory and accepting the same findings. The mature versions already exist and are joined, not invented: the Accord, the Responsible Minerals Initiative, Sedex, amfori BSCI.

Two that cannot be answered without doing the arithmetic.

9. A firm buys 80 part families, dual-sourced, with a branching factor of 4. Audits cost $6,000 a node. It values full compliance at $9,000,000 a year and retains leverage λ = 0.30 per tier. How deep does its code of conduct reach? Show the working.

N₁ = 80 × 2 = 160. C₁ = 160 × 6,000 = $960,000. V₁/C₁ = 9,000,000 / 960,000 = 9.3750, so ln(V₁/C₁) = 2.2380. b/λ = 4 / 0.30 = 13.3333, so ln(b/λ) = 2.5903. n = 1 + 2.2380 / 2.5903 = 1.864 tiers — one full tier and 86 percent of the second. The reasoning mark is for saying what that means operationally: tier two is reachable for the larger relationships within it and not for the rest, so the firm should be selecting inside tier two rather than pretending to cover it.*

10. Using p = 0.05, q = 0.015, 60 families and $18,000 per supplier a year: should a firm whose stoppage costs $50,000,000 add a third source? Compute it.

P(2) = 0.015 + 0.985 × 0.05² = 0.017462. P(3) = 0.015 + 0.985 × 0.05³ = 0.015123. ΔP = 0.002339. Expected saving = 0.002339 × 50,000,000 = $116,969 a year. Extra admin = 60 × 18,000 = $1,080,000 a year. Net −$963,031 — do not take it. The breakeven event cost is $461,661,769. Full marks require the breakeven figure, because that is the number that tells the firm what would have to change for the answer to flip — and it is an event cost, not a supplier count.

REFLECTION — eight questions, for one person and a pen

These are not for a room. Write the answers by hand if you can; the slowness is the point.

Name the last time you were stopped by a company you had no contract with. What did you learn about your own map that day, and what did you do with what you learned?

Which of your relationships would survive the Aisin test — if their factory burned tonight, who would start work before the terms were agreed? Write the names. Then ask what you did to earn each one.

Where have you added a supplier to feel safer rather than because a computation asked for it? What did that decision cost in influence, and would you make it again knowing the subtraction?

What do you already know about your second tier that has never been written into any system? How did you come to know it, and who else in your organisation knows something similar and has never been asked?

Think of a supplier who is small enough that you could end them, and important enough that they could stop you. How do you behave toward them, and is it the behaviour you would want from your own largest customer?

Which part of your work is a q — the single shared dependency that would take out everything you have carefully separated? Answer about your own life, your own attention and your own health before answering about a business.

When did a number change a decision you had already made emotionally? What was it about the number that got through when the argument did not?

If you could hand your successor one page about where things actually come from, what would be on it? Write the page now, and notice what you cannot yet fill in.

ESSAY PROMPTS — five

Each is arguable from more than one side. Each requires at least one source the chapter cites and at least one it does not.

1. Diversity against influence. The chapter computes that quadrupling the source count costs 1.049 tiers of enforcement reach, so the resilient configuration is often the one with less power over upstream conduct. Argue either that this trade should be resolved in favour of resilience — because a stopped line helps no worker anywhere — or in favour of leverage, because conduct upstream is the thing a buyer is uniquely able to change. Use Locke's The Promise and Limits of Private Power and one empirical study of audit effectiveness the chapter does not cite.

2. Is the Aisin recovery repeatable? Sixty-two firms made a part they had never made, in five days, without a contract. Argue either that this is a general property of dense industrial networks and can be deliberately cultivated, or that it was specific to the Toyota group's history, geography and ownership structure and cannot be transplanted. Use Nishiguchi and Beaudet (1998) and at least one source on industrial districts or clusters — Piore and Sabel, Saxenian or Becattini — that the chapter does not cite.

3. The waist and the mouth. The Responsible Minerals Initiative audits roughly three hundred smelters rather than thousands of mines, at a 96.25 percent saving. Argue either that enforcing at the network's waist is the correct and general answer, or that it displaces scrutiny from where harm actually occurs and buys assurance at the cost of seeing anything. Use the RMI's own conformance documentation and one field study of artisanal mining the chapter does not cite.

4. Whose number is the event cost? Every source-count decision in this chapter turns on the cost of one stoppage, which is estimated inside the firm by people with an interest in the answer. Argue either that this number can be made honest — through pre-registration, external review, or post-event calibration — or that it is structurally unknowable and the arithmetic therefore dresses a judgement as a computation. Use Simchi-Levi, Schmidt and Wei (2014) and one source on forecasting bias or reference-class forecasting the chapter does not cite.

5. Regulation or arithmetic. Mandatory due-diligence law now requires many firms to identify and address risks in their supply chains. Argue either that regulation is the only instrument that can force V₁ high enough to make deep tiers economic — the chapter computes that reaching tier three from the base case would need $1.6 billion a year of compliance value — or that regulation raises C₁ faster than V₁ and the real gain comes from shared inspection and financing spreads. Use the ILO (2016) brief on purchasing practices and one primary legal text — the German Lieferkettengesetz or the EU corporate sustainability due diligence directive — that the chapter does not cite.