High-Yield Topics — What the Exams Actually Test

How this list was built

Every question in the five available papers — Sample Exam 2026 (20 MCQs), Moed A 2023-24, Moed B 2023-24, Moed A 2025 and Moed B 2025 (each 6 short + 2 long questions, plus a Miluim question in 2025) — was tagged by topic and counted. The counts below are honest: a topic listed at "5/5" genuinely appears in all five papers. The two long questions (32 points each on the Moed papers, ~64% of the grade) are flagged separately, because that's where the exam is won or lost.

The scoreboard

The big four

Two-period consumption, production/TFP, investment/user cost, and labor supply with taxes account for both long questions on every Moed paper — roughly 64 of 100 points each sitting. If time is short, drill these four until you can re-derive them from a blank page. Everything else is short-question material worth 6 points a pop.

---

Tier 1 — appears in every paper

Two-period consumption & the Euler equation — 5/5 papers

The single most-tested topic. Every Moed paper's first long question (32 pts) is a two-period (or three-period) consumption derivation, always the same recipe: write the problem → Lagrangian → FOCs → Euler equation → substitute into the lifetime budget → solve for $c_0$ → comparative statics.

• Trained by: PS2 (both questions — mini life-cycle model, log-utility plan) — plus the goods-market follow-ups in PS4.
• Lecture: Lec_02-Consumption and Saving.
• Know cold: the log-utility signature ($c_0$ = fixed fraction of lifetime wealth, income and substitution effects on $c_0$ cancel exactly) and the CRRA generalisation $\frac{c_{t+1}}{c_t} = \left[\beta(1+r)\right]^{1/\sigma}$.

Production function & TFP accounting — 5/5 papers

Shows up at least once in every paper, usually twice. Two moves cover almost every variant: (1) invert Cobb-Douglas to read TFP off as the residual, $A = Y / \left(K^{\alpha}N^{1-\alpha}\right)$; (2) recover a missing $\alpha$ from a factor share ($\alpha = \frac{RK}{Y}$ or $1-\alpha = \frac{wN}{Y}$).

• Trained by: PS3 (all four questions — Cobb-Douglas proofs, labor shares, implied TFP, growth accounting).
• Lecture: Lec_04-Production.
• Know cold: $MPK = \alpha\frac{Y}{K}$, $MPN = (1-\alpha)\frac{Y}{N}$, and always differentiate the function actually given — the additive-form trap in Moed A 2024 Q4 punishes Cobb-Douglas autopilot.

Investment & the user cost of capital — 5/5 papers

Every paper has a question that reduces to "future MPK = user cost, solve for the one unknown". The examiners rotate which variable is missing: the tax rate, the depreciation rate, future TFP, or the direction of a price effect.

• Trained by: PS4 Q1 (workforce size and optimal investment) and Q3 (taxes and their use); also PS6.
• Lecture: Lec_05-Investment.

---

Tier 2 — a near-guaranteed question

Labor supply & taxes (the static FOC) — 4/5 papers

Missing only from Moed A 2024. The second long question on both Moed B papers is a labor-supply derivation, and the sample exam has two MCQs on it. One condition does everything:

Recipe: plug in the given $U$, substitute the budget constraint, solve for $N(w)$ — then discuss income vs substitution effects.

• Trained by: PS5 Q1–Q3 and PS6 Q1–Q2, Q4–Q5 (the whole set is built on this FOC).
• Lectures: Lec_07-Labor Market, Lec_10-Fiscal Policy.
• Know cold: with log consumption utility, income and substitution effects cancel exactly — that one fact answers three of the six questions above.

Public debt dynamics — 4/5 papers

Missing only from Moed A 2025. Always a short question, in one of two flavours: qualitative (map a shock onto the signs of $d$, $g$, $r$) or numerical forecast — where $r$ and $g$ are never handed to you directly.

• Trained by: PS6 Q3 (simple public debt dynamics).
• Lecture: Lec_10-Fiscal Policy.
• Know cold: the compound-growth back-out $g = \left(\frac{Y_T}{Y_0}\right)^{1/T} - 1$ — it appears every time.

Goods-market equilibrium (S and I curves) — 4/5 papers

Rarely a standalone question, but it's the closing act of most long questions and gets two dedicated shorts. The drill: shift each curve separately, then read off which of $r$, $S$, $I$ is pinned down and which is ambiguous.

• Trained by: PS4 Q2–Q3, PS6 Q1.
• Lecture: Lec_06-Equilibrium in the Goods Market.
• Know cold: the meta-rule — curves shifting in opposite directions pin the price ($r$) and leave quantities ambiguous; shifting the same way pin the quantity and leave $r$ ambiguous.

Unemployment flows & steady state — 4/5 papers

Missing only from Moed B 2024. Pure plug-in territory once you dodge the classic denominator traps.

• Trained by: PS5 Q4 (steady-state unemployment flows).
• Lecture: Lec_08-Labor Market Data, Participation & Unemployment.

---

Tier 3 — shorter odds, cheap points

Inflation measurement: CPI vs deflator — 3/5 papers

A three-good table (one good never consumed), compute both measures, explain the gap. Moed A 2024 Q6 (robots drag the deflator negative while the CPI rises), Moed A 2025 Q6 (PCE deflator vs CPI: changing vs fixed basket), Sample Exam 2026 Q6. Rules: strip non-consumption goods out of the CPI (but not the GDP deflator); deflator = changing quantities, CPI = frozen basket. Trained by PS1 Q2, Q3 and Q7; lecture Lec_01-Data Review.

Inequality, polarization & SBTC — 2/5 papers + Miluim

Pure "discussed in class" material, zero algebra: Moed A 2024 Q5 (90/10, 90/50, 50/10 ratios; polarization = rising 90/50 with falling 50/10) and the Moed B 2025 Miluim question (SBTC is monotone, so it explains the college premium but not U-shaped polarization). Trained by PS5; lecture Lec_09-Inequality & Polarization.

Task-based production & robots — 2/5 papers

Moed A 2025 Q2 (immigration lowers robot demand — $[R+N]$ makes them substitutes) and Sample Exam 2026 Q2 and Q17 (productivity vs displacement effects). The move: check the sign of the cross-partial. Trained by PS5 Q5; lecture Lec_09-Inequality & Polarization.

GDP vs GNP vs NFP — 2/5 papers

Moed A 2025 Q4 (Ukrainian emigration: GDP falls, GNP flat, NFP rises) and Sample Exam 2026 Q4. Anchor: $GNP = GDP + NFP$ — GDP is where production happens, GNP is who earns. Trained by PS1; lecture Lec_01-Data Review.

PPP comparisons — 2/5 papers

Moed B 2024 Q6 (PPP GDP per capita growth and catch-up) and Sample Exam 2026 Q10. Formula: $GDP^{PPP} = \frac{GDP^{local}}{P^{local}} \times P^{US}$, then per-capita, then growth. Trained by PS1 Q5; lecture Lec_01-Data Review.

Liquidity constraints & the MPC — 2/5 papers

Moed B 2024 Q1 and Sample Exam 2026 Q9: constraints fall with income; constrained households have the highest MPC; the top-quintile exception is the "wealthy hand-to-mouth". Lecture Lec_02-Consumption and Saving.

Efficient factor allocation across firms — 1/5 papers

Sample Exam 2026 Q8 only, so far: allocate a fixed factor across firms by equalising marginal products. Trained by PS7 (both questions — optimal allocation of capital, and two factors under constant returns); lecture Lec_04-Production. New material since the last problem set, so treat it as live for 2026 despite the thin past-paper record.

Business cycles: trend vs cyclical — 1/5 (Miluim only)

Moed A 2025 Miluim: trend = expected path, cyclical = deviations; non-durables/services consumption is smoother than income, evidencing consumption smoothing. Low priority unless Miluim-eligible.

---

Reading the format

The four Moed papers are open-notes (formula sheet attached — see the Macro Equation Sheet) with 6 shorts (6 pts) + 2 longs (32 pts). The Sample Exam 2026 is the 20-question MCQ dress rehearsal for the 2026 format — same toolkit, but distractors punish sign errors and dropped denominators rather than rewarding partial derivations. Revise for the long-question derivations first; MCQ speed comes free once the derivations are automatic.

Next step: turn these frequencies into a schedule — see the Macro Revision Plan.