The John Locke Essay Competition is one of the most competitive essay contests for high school students. 2,000 words. Three questions. Under 1% win rate. What follows is a guide to the economics content you actually need — the frameworks, the primary literature, and the analytical moves that separate shortlisted essays from rejected ones.
The 2026 Questions
Q1. Should we fear a cashless society? Q2. Technology now allows personalised pricing. If this came to be widely used, what effects should we expect? Q3. Did Jeff Bezos get rich at the expense of his customers, his employees, neither or both?
Pick based on which one you can handle technically, not which sounds most interesting.
Q1 — Cashless Society
What the question is actually asking
Two sub-questions are buried here: (1) what functions does physical cash perform that digital payment doesn’t, and (2) are those functions valuable enough that their loss constitutes harm?
Most entrants collapse these into “cashless = surveillance = bad.” That’s a conclusion, not an argument.
The economic functions of cash
Cash performs at least five functions that electronic money does not:
- Anonymity — a bearer instrument with no transactional metadata
- Offline settlement — no infrastructure dependency
- Hard budget constraint — psychologically and mechanically
- Monetary sovereignty hedge — a direct central-bank liability outside the commercial banking system
- Financial inclusion backstop — no account, credit check, or ID required
Pick two or three and examine them seriously. Covering all five at 2,000 words produces a list, not an argument.
Technical frameworks to deploy
Payment systems economics. Rochet & Tirole on two-sided markets explains why card networks charge what they do and why merchant-side fees that look exploitative are the equilibrium outcome of platform competition. “Cashless” usually means “card/digital dominant,” which means the interchange fee structure becomes the de facto payment tax.
CBDC design tradeoffs. A wholesale CBDC (bank-to-bank) is economically trivial. A retail CBDC raises the real questions: account-based or token-based? Intermediated or direct? These aren’t technical details — they determine whether the central bank now sees every transaction. BIS working papers (Auer, Cornelli, Frost) are the primary literature.
Bank disintermediation risk. If retail CBDC is attractive, deposits flee commercial banks during stress, amplifying runs. Brunnermeier & Niepelt (2019) formalises this. Any serious essay on Q1 needs to engage with it.
The Diamond-Dybvig lens. Bank runs are coordination failures. A frictionless digital payment system makes coordination easier — runs faster and more contagious. SVB in 2023 lost $42bn in hours, largely via mobile banking apps.
Evidence worth citing
- Sweden’s Riksbank reports on cash decline (physical cash ~1% of GDP)
- ECB digital euro consultation papers
- MAS 2021 Payments Landscape study (Singapore-relevant)
- Agur, Ari & Dell’Ariccia (2022) on CBDC design and welfare
- Chodorow-Reich et al. on the 2016 India demonetisation — the closest thing to a natural experiment on forced cashlessness
Where most essays fail
Treating “privacy” as self-evidently valuable. It isn’t, in welfare-economic terms, unless you can specify the externality. Why does transaction privacy matter beyond preference? Candidates: chilling effects on legal-but-stigmatised spending, political expression, protection against state abuse in non-democratic regimes, protection against commercial surveillance and subsequent price discrimination (linking Q1 to Q2). Pick one and argue it.
Q2 — Personalised Pricing
The most technically demanding question. The economics is well-developed, the empirical literature is current, and most entrants will butcher it.
The central result you must know
First-degree price discrimination maximises total surplus. It eliminates deadweight loss entirely because every consumer whose willingness to pay exceeds marginal cost gets served. All surplus goes to the producer, but the pie is maximally large.
This is counter-intuitive to most students, who assume price discrimination is “unfair” and therefore welfare-reducing. Efficiency and distribution are separate questions. An essay that conflates them fails on technical grounds in the first page.
The correct welfare question is: does personalised pricing approximate first-degree discrimination, and if so, is the distributional outcome acceptable?
Why personalised pricing isn’t actually first-degree PD
Real-world personalised pricing is estimated, not observed. Firms infer willingness to pay from proxies — browser, device, location, purchase history. This introduces:
- Estimation error — some consumers mispriced, some excluded, some overcharged relative to true WTP
- Strategic consumer behaviour — VPNs, incognito mode, cart abandonment, account switching. Sophisticated consumers gain; naive ones pay for it.
- Adverse selection — your data profile reflects the kind of person who leaves data trails
- Ratchet effects — paying a high price today reveals high WTP, raising future prices. Rational consumers under-consume.
Each is a distortion from the first-degree benchmark. Strong essays model at least one formally.
Key literature
- Varian (1985, 1989) — foundational on price discrimination welfare
- Shiller (2014) — “First-Degree Price Discrimination Using Big Data” — the paper that put a number on it (14.6% revenue increase for Netflix-style personalisation)
- Dubé & Misra (2023) — field experiment at Ziprecruiter showing personalised pricing raised both firm profit and consumer surplus. Surprising result. Engage with why.
- Bonatti & Cisternas (2020) on ratchet effects in data-driven pricing
- Acquisti, Taylor & Wagman (2016) — survey on the economics of privacy
- Calvano et al. (2020) on algorithmic collusion
- FTC (2024) report on surveillance pricing
Mechanisms worth building an essay around
Pick one, not five:
- Search costs and obfuscation — personalised pricing may reduce efficiency by making prices unlearnable (Ellison & Ellison)
- Two-sided information — when consumers know firms know them, they behave strategically (ratchet effect)
- Algorithmic collusion — personalisation infrastructure may tip oligopolistic markets into tacit collusion
- Heterogeneous consumer naivety — personalisation redistributes from naive to sophisticated consumers (DellaVigna & Malmendier)
- Dynamic effects on entry — if incumbents can identify and pick off marginal customers, entrants face harder unit economics
Where most essays fail
Three specific traps:
- Treating “price discrimination is unfair” as a premise rather than a conclusion requiring argument. Economists don’t share this intuition by default. Earn it.
- Ignoring the Dubé-Misra result because it’s inconvenient. Engage with it — specify the conditions under which personalised pricing raises consumer surplus and whether those conditions generalise.
- Hand-waving about “consent” and “trust.” These are real variables but need to be specified economically — trust as a reduction in search costs, consent as a contracting problem under bounded rationality.
Q3 — Jeff Bezos
Thousands of entrants will write this. Most will produce one of two templates:
- Template A: Amazon created consumer surplus → Bezos earned his wealth → neither
- Template B: Amazon exploited labour/killed competitors → both
Both are weak. Three moves that aren’t:
Move 1: Force a definition of “at the expense of”
In standard welfare economics, “A got rich at B’s expense” means a transfer, not a Pareto improvement. If Amazon’s existence made every consumer and employee better off than the counterfactual, Bezos got rich at no one’s expense, regardless of the size of his fortune. This is the pecuniary externality point — displaced competitors aren’t a real welfare cost because their losses are offset by consumer gains.
The interesting question is when the framework breaks down:
- Monopsony in labour markets (Azar, Marinescu & Steinbaum 2022 on Amazon’s labour market power)
- Monopoly rents extracted from marketplace sellers (Khan 2017 on platform market structure)
- Externalities on non-transacting third parties (local retail, tax bases, logistics externalities)
Move 2: Actually model the counterfactual
“Did Bezos get rich at someone’s expense” is a counterfactual claim. It requires a counterfactual to compare against. What would prices, wages, variety, and market structure look like without Amazon? Most entrants don’t attempt this. If you can build even a rough model — pre-Amazon retail margins as a benchmark, or international comparators where Amazon is weak — you’re doing work the median entrant isn’t.
Move 3: Distinguish the three parties properly
The question asks about customers, employees, neither, or both. Don’t lump them together.
- Customers: almost certainly net beneficiaries. Consumer surplus literature on Amazon is unambiguous. Argue this cleanly and move on.
- Employees: genuinely contested. Amazon pays above minimum wage but the monopsony literature suggests it pays below the competitive wage in local labour markets where it dominates. Warehouse injury rates exceed industry norms. This is where your essay earns marks.
- Third parties not in the question: sellers, competitors, municipalities. You can argue the question is mis-specified — “neither customers nor employees, but sellers and competitors” is a defensible thesis.
Literature for Q3
- Khan (2017), “Amazon’s Antitrust Paradox,” Yale Law Journal
- Azar, Marinescu, Steinbaum (2022) on labour market monopsony
- Hsieh & Rossi-Hansberg (2023) on the industrial revolution in services
- Amazon annual reports and 10-K filings for wage, employment, and margin data
Where most essays fail
Moralising. “Billionaires shouldn’t exist” and “he earned every dollar” are both weak starting points. Judges want economic reasoning about mechanisms — not normative priors dressed up as conclusions.
Choosing Your Question
Not which interests you most. Which one you can execute.
| Q1 Cashless | Q2 Personalised Pricing | Q3 Bezos | |
| Core framework | Monetary economics, payment systems | Price theory, IO, info economics | Welfare econ, antitrust, labour |
| Literature access | Moderate (policy, BIS) | High (academic + accessible) | High, but crowded |
| Competitive field | Medium | Medium-high | Saturated |
| Technical demand | Medium | High | Medium |
| Originality ceiling | High | High | Low (unless novel angle) |
Q2 is our default for students with genuine microeconomic grounding. Q1 for policy or monetary interest. Q3 only if you have a specific, non-obvious thesis.
Structural Requirements
At 2,000 words, every paragraph must earn its place.
- Opening (150 words). Question, thesis, roadmap. No ramp-up.
- Definitions (200 words). Define your terms rigorously. Most essays skip this and collapse.
- Core argument (900–1,000 words). One mechanism, developed properly. Theory → evidence → worked example → implication.
- Strongest counter-argument (300–400 words). Not a strawman. The objection that would most damage your thesis. Given its due, then answered.
- Qualification (150 words). When does your argument not hold? Signals intellectual honesty and is heavily weighted.
- Conclusion (100–150 words). Restate the thesis, now defended. No summary. No “further research needed.”
Citation Standards
In-text citations in APA count toward the 2,000-word limit. Bibliography and endnotes do not.
Minimum: 8–12 sources, of which at least 4 should be peer-reviewed papers or working papers from recognised institutions (NBER, CEPR, BIS, IMF, central banks). Textbook citations are weak. Blog post citations are weaker. Wikipedia is disqualifying.
If you cite a paper, you should have read it — not just the abstract. Misrepresented citations are worse than missing ones.
Technical Errors That Sink Essays
- Confusing efficiency with distribution (nearly universal)
- Using “monopoly” when you mean “market power” or “dominant firm”
- Treating pecuniary externalities as welfare-relevant
- Citing “the deadweight loss” without specifying it
- Invoking “market failure” without naming the specific failure (externality, information asymmetry, public good, market power)
- Using behavioural arguments loosely — specify the bias, the evidence, the welfare implication
- Aggregating consumer and producer surplus without weighting — if your argument depends on interpersonal utility comparisons, state it
- Missing the counterfactual — “X caused Y” requires specifying what would have happened without X
AI Policy (2026)
AI tools are permitted; the work must remain your own. Practical interpretation: use AI for source discovery and grammar, not for generating arguments or prose. AI-generated economic writing has a distinctive register — hedged, balanced, technically imprecise — that experienced readers detect.
Submission
- Deadline: 31 May 2026, 23:59 BST
- File name: FirstName-LastName-Economics-QuestionNumber.pdf
- No name inside the document
- No footnotes; endnotes only
- Referee email required (a teacher familiar with your written work)
- Submit a day early. Servers fail at deadlines.
For students who want their thesis stress-tested, their argument pushed, or their sources examined — this is the work we do. Book a complimentary consultation at qconsult.org, or write to enquiries@qeducation.sg.
