La Bourse / Volume II / Nº II.08 / Quiz, reflection, essays

A woman and a man seated across from each other in a bright open office, mid-conversation, plants beside them. — Plate II.08 · Quiz, reflection, essaysThe Handshake, and the Ledger Behind It.The handshake is not the opposite of the contract. It is a contract whose enforcement was paid for in advance, over years, by people who are not in the room.

ASSESSMENT · Chapter II.08 — Networks, Trust, and Topology

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. State the two cost curves in this chapter and name the variable each one depends on.

C_t(n) = 150 × (n − 1) for the trust network and C_c(T) = 250 × T for the contract network. One mark for the expressions; one for the observation that the first depends on membership and the second on transactions, which is the whole reason they cross.

2. Give Dunbar's 1992 predicted human group size with its confidence interval, and say what the 2021 reanalysis did to it.

147.8, with a 95 percent confidence interval of 100.2 to 231.1. Lindenfors, Wartel and Lind re-ran it with updated primate data and modern phylogenetic methods and obtained intervals running, across specifications, from fewer than five people to more than five hundred. Full marks require the interval, not the 150.

3. What did Granovetter find about the frequency of contact with the person who supplied a job lead?

Of jobs found through a personal contact, 16.7 percent came from someone seen often, 55.6 percent occasionally and 27.8 percent rarely — so 83.4 percent came from a contact seen less than twice a week.

4. What is the honest status of the trust-and-growth coefficient?

Real in the data we have, smaller than the enthusiastic literature reports, and not securely identified by cross-country regression. Beugelsdijk, de Groot and van Schaik found it robust under extreme bounds analysis in the original 29-country sample and not robust in the extended sample; the strongest causal support is Algan and Cahuc's inherited-trust instrument.

Four on application.

5. A trade association of 300 members meets monthly, maintains a directory, and its members do about eleven pieces of business a year with each other. The chair proposes doubling engagement. Diagnose it.

At the chapter's cost assumptions each member is carrying 150 × 299 = 44,850 a year of maintenance against 2,750 of contracting the same eleven transactions would have cost. The association is on the wrong curve. The stronger answer notes that doubling engagement doubles the cost of the side that is already too expensive, and that the fix is either to raise transaction volume among members or to shrink the trust core and put everyone else on templates.

6. Your head of procurement says the supplier base is "built on trust" and has not changed in six years. What are you looking at, and what two figures would settle it?

A closed network. The two figures are the price of closure to you — the expected-best-of-k gap, about 1.7344 percent of purchase spend on the chapter's assumptions — and the number of counterparties admitted in the last three years. Credit any answer that names both sides; an answer that only names the fairness cost has missed the argument the chapter is making.

7. A manager wants to cut "networking time" because it produces nothing measurable. Give the structural counter-argument in one figure.

Reach lives in ties that do not know each other. Inside a cluster of fifteen, one more tie reaches zero new people; one bridge to a different cluster of fifteen reaches fourteen. Cutting relational time outside the cluster cuts the only part with a return. The stronger answer adds the causal evidence: in the LinkedIn experiment, moderately weak ties produced the most job transmission and the strongest ties the least.

8. Why does the facility in Operationalize This reserve places for counterparties nobody at the firm has met?

Because closure costs the firm about 1.7344 percent of purchase spend in forgone supplier selection — 346,883 a year on a 20,000,000 budget — and the reservation is what keeps the candidate pool wide. It is the profitable half of the design, not the charitable one.

Two that require the arithmetic to be done.

9. A country raises generalised trust by 35 points over a generation. Using the Knack and Keefer slope, what is the effect on income per head after thirty years? Show your working.

The slope is 0.8 pp / 10 pp = 0.080 pp of growth per point of trust. 35 × 0.080 = 2.80 pp a year. Compounded: 1.028^30 = 2.2898, so income per head is about 129.0 percent higher — it has more than doubled. Credit any method landing between about 120 and 140 percent. The point of the question is that the answer is a multiple, and that the same arithmetic run on the Zak and Knack slope gives a visibly smaller one — which is why the chapter quotes both.

10. A buyer with a purchase budget of 8,000,000 could qualify 60 suppliers but buys only from the 10 inside its network. Using the chapter's expected-best-of-k model, what does the closure cost per year?

E[best of k] = 0.9 + 0.2/(k+1). Open: 0.9 + 0.2/61 = 0.903279. Closed: 0.9 + 0.2/11 = 0.918182. The gap is 1.4903 percent of spend, or 119,225 a year. The stronger answer notes that the gap narrows as k rises — the first few additional candidates do nearly all the work — and that this is an argument for a ramp of two or three places a year rather than for throwing the network open.

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 three people whose trust in you currently carries the most economic weight. When did each of them decide, and what did you do that week?

Who introduced you to the thing you now do for a living? How well did you know them at the time, and what does that suggest about where your next turn comes from?

Which relationships are you maintaining out of habit rather than exchange? Not to cut them — to notice that you are paying for them, and to decide whether you meant to.

Where have you been the outsider paying the newcomer's tax? What did it cost you, in money or in years, and who eventually let you in?

Where are you the insider? Name a specific person your network is currently keeping out, and the reason you have given yourself for it.

What would somebody be able to read about your reliability if you were not there to explain it? If the honest answer is nothing, what is the first record you could start keeping this month?

Think of a time a handshake did work that a contract could not have done. What had been paid in advance, by whom, and over how long?

Which of your ties is the one bridge between two worlds that otherwise do not touch? What would it take to build a second one?

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. Is the trust dividend causal? The chapter reports the Knack–Keefer and Zak–Knack slopes, the Beugelsdijk et al. robustness failure in the extended sample, and the Algan–Cahuc inherited-trust instrument. Argue either that generalised trust has a identifiable causal effect on growth large enough to justify policy, or that the correlation is largely the residue of institutions, schooling and history and that trust is a symptom rather than a cause. Engage Algan and Cahuc directly, and at least one source on the measurement validity of the World Values Survey trust item that the chapter does not cite.

2. The Maghribis and the Genoese, re-argued. Greif reads the Maghribi coalition as a reputation equilibrium and the Genoese as a legal one. Edwards and Ogilvie dispute the reading of the Geniza evidence. Write the case that a reputation-based trading network is a stage that economies grow out of — then write the case that it is a permanently available alternative that was displaced by politics rather than by efficiency. Use Greif and Edwards and Ogilvie, and one source on modern extralegal commercial order that the chapter does not cite.

3. Dunbar's number as design input. The chapter insists on the interval — 100.2 to 231.1 in 1992, far wider in the 2021 reanalysis. Argue either that the layered structure is robust enough to build organisations on, or that the whole quantity is an artefact of regression on small primate samples and that organisational designers should abandon it. Use Dunbar's 1992 paper and Lindenfors, Wartel and Lind, and one empirical study of organisational unit size that the chapter does not cite.

4. The price of closure. The chapter computes closure as costing insiders about 1.7344 percent of purchase spend and outsiders 8.1 percent of revenue, and concludes that opening is a procurement decision rather than an ethical one. Argue the counter-case: that reducing exclusion to its efficiency cost concedes the moral argument to whoever finds the efficiency cost small, and that some networks should be opened even when the arithmetic says to keep them shut. Use Portes and Fafchamps, and one source on discrimination or referral hiring that the chapter does not cite.

5. What a settlement layer replaces. M-Pesa did not build trust between strangers; it made trust unnecessary for one act. Distributed ledgers make a stronger version of the same claim. Argue whether trustless settlement infrastructure substitutes for social trust, complements it, or erodes it by removing the occasions on which it would otherwise have been built. Use Suri and Jack, and one source on payment infrastructure or distributed ledgers that the chapter does not cite.