Skip to main content
HSC 2026 published Aug 2026

Economics question styles to look out for

The question styles to have ready — each linked to its evidence and to a question in the practice paper.

Built by a six-model AI panel and backtested against the hidden 2025 papers — how we did it.

About the 2026 exam & how this page was built

Built from a six-model AI analysis of every HSC Economics paper, marking guideline and marking-centre feedback report since 2019 — using the same method we backtested against the real 2025 papers before publishing. First, the housekeeping: the current (2009) Economics syllabus continues to run — 2026 is NOT a final syllabus year, so expect evolution, not a farewell paper. These are styles to prepare for, not guarantees: our backtest showed the examiners keep the skill and twist the format, so practise the skill chain, not a memorised question.

The near-certainties

Section I · 1 mark 6 of 6 models expect this

The Section I calculation battery

The panel's highest-probability single item: derive MPC/MPS from a two-year income-consumption or income-savings table, form k = 1/(1-MPC), apply it to a stated injection change and...

Try it → Q6 Why we expect this
Section I · 1 mark 6 of 6 models expect this

BOP table with one unknown: CA + KAFA = 0 balancing item

A balance of payments table under a floating exchange rate with one entry blank (financial account, exports, or a net income line), solved with the identity CA + KAFA = 0 — the annual...

Try it → Q13 Try it → Q22 Why we expect this
I/II · 1–6 marks 6 of 6 models expect this

Fiscal stance vs budget outcome

Classify a two-year budget-outcome change as expansionary or contractionary, separating discretionary (structural) change from automatic-stabiliser (cyclical) movement — the recurring...

Try it → Q12 Try it → Q24 Why we expect this
On this page
  1. 1. Fiscal policy comes back for the big marks
  2. 2. The case study returns — and it means development, not growth
  3. 3. The Section I calculation battery
  4. 4. The protection diagram pair that ends Section I
  5. 5. Fiscal stance vs budget outcome
  6. Where to spend study time
  7. Practice paper
  8. Check our working

The big five (have these cold)

1. Fiscal policy comes back for the big marks

6 of 6 models expect this — consensus probability 0.50
What it looks like

a 20-mark Section III stimulus question — a budget-balance-as-a-percentage-of-GDP chart showing the recent surplus-to-deficit swing, a Treasury quote, and a verb like evaluate pairing economic growth with a second objective (external stability or distribution).

Why we expect it

all six models call it — fiscal policy hasn't had a dedicated 20-mark question since 2021, the longest Section III absence of any Topic 4 policy, and the budget swing of the last few years is exactly the chart shape examiners like (a reversal you must interpret, not restate).

The traps markers flag (2019, 2021, 2022)

using data from the wrong budget era; restating the stimulus; asserting a judgement once in the conclusion instead of sustaining it across every named objective.

Practise

one fiscal essay plan per week against a different paired objective, always arguing stance from the change in the outcome.

Try it → Q25 6 of 6 models expect this — see the evidence

2. The case study returns — and it means development, not growth

6 of 6 models expect this — consensus probability 0.57
What it looks like

either a 4–6 mark Section II part with a printed "Name of economy: ______" line, or a full Section IV essay on globalisation and development in an economy other than Australia.

Why we expect it

the syllabus mandates the case study and it was examined every year 2019–2024, then skipped in 2025 — the clearest coverage gap on the paper. All six models predict its return.

The traps (2019, 2021, 2024)

quoting GDP growth when the question says development — markers want HDI, life expectancy, literacy and poverty data; choosing an advanced economy (the 2019 feedback named the USA, Canada, England, NZ and Norway as weak selections); no judgement.

Practise

know one developing or emerging economy with five memorised development statistics and two named, evaluated strategies.

Try it → Q27 6 of 6 models expect this — see the evidence

3. The Section I calculation battery

6 of 6 models expect this — consensus probability 0.76
What it looks like

6–8 quantitative multiple-choice items, and the panel's four near-certainties: the simple multiplier from a two-year income/savings table; the balance of payments table with one entry blank (solve with CA + KAFA = 0); the labour force table (unemployment and participation rates); and a terms-of-trade or net-foreign-liabilities index.

The traps

MPS used as MPC and k applied to the wrong base; the capital account left out of the BOP identity; dividing by the working-age population instead of the labour force; hidden unemployment counted as unemployment.

Practise

drill each calculation until the distractor errors feel familiar — the wrong options are built from these exact slips.

Try it → Q6 6 of 6 models expect this — see the evidence

4. The protection diagram pair that ends Section I

5 of 6 models expect this — consensus probability 0.69
What it looks like

the final, hardest multiple-choice items — two questions chained off one demand/supply diagram with a world price: tariff revenue after a change, then the change in foreign producers' revenue or the subsidy size that eliminates imports.

Why we expect it

the 2019/2022/2025 closer pattern, and the 2024 feedback said students could not "accurately interpret the diagram and calculate the economic implications of protection". (The protection essay is rested after 2025 — the diagram arithmetic is not.)

The traps

multiplying the tariff by total consumption instead of imports; valuing foreign producers' sales at the domestic (tariff-inclusive) price instead of the world price.

Try it → Q19 5 of 6 models expect this — see the evidence

5. Fiscal stance vs budget outcome

6 of 6 models expect this — consensus probability 0.63
What it looks like

a multiple-choice or short-answer item where a deficit shrinks — and you must call that contractionary; separating discretionary (structural) changes from automatic-stabiliser (cyclical) movements, possibly with a consequence-of-deficit-financing chain (bond sales → crowding out).

Why we expect it

six of six models; the stance/outcome and cyclical/structural confusions are flagged repeatedly (2022, 2024), and two models make the never-directly-examined cyclical-component calculation a bold call.

The trap

reading stance from the sign of the balance instead of the change in it.

Try it → Q12 Try it → Q24 6 of 6 models expect this — see the evidence

Also on the radar

  • Headline vs underlying inflation — assign the cause to the correct measure: a cyclone or excise change moves headline; broad wage pressure moves underlying. Don't call disinflation "deflation" (flagged 2021).
  • The KAFA → NPY chain — explain how financial-account inflows create later primary-income debits. The marking centre has flagged this twice (2023, 2025); a restated accounting identity scores nothing.
  • Micro/labour reform chains — a named reform traced through efficiency → productivity → unit labour costs → aggregate supply → cost inflation. Name a real product or factor market; "the government deregulated" is a description, not a chain.
  • Market-based environmental policy vs regulation — the most repeated environment complaint (2021, 2024). Market-based means a price incentive (tax, ETS, deposit scheme, tradeable permit) — a ban or target is not one. Know the externality diagram: the corrective tax is the vertical gap at the socially optimal quantity.
  • External stability as the synthesis essay — CAD and net foreign debt as % of GDP, the savings–investment gap, valuation effects, and a judgement (four models rate it the overdue 20-marker; two name the Pitchford thesis as the top-band discriminator).
  • TWI vs bilateral rates — the AUD can rise against the USD and fall on the TWI: it's a trade-weighted index and the US dollar's weight is single-digit. Confused with the CPI or bilateral rates in three feedback reports (2019, 2021, 2022).

Where to spend less time

The panel's clearest rested call: distribution of income and wealth took the 2025 Section III essay (after a 2023 10-marker) and five of six models rest it — expect at most a Gini/Lorenz multiple-choice item. The protection essay, monetary policy and unemployment also give up their big-mark slots after heavy 2025 treatment: all three keep their Section I staples, so know the mechanics — everything is examinable — but don't build your revision around another 10- or 20-marker there.

The honest fine print

When we backtested this method on the 2025 papers (Chemistry and Maths Extension 1 — the two subjects with scored holdouts), our high-confidence topic calls went 37/37, and about half of our specific question predictions recognisably appeared — but the examiners inverted formats, migrated written staples into multiple choice, and broke two long streaks. Economics adds its own wrinkle: the choice structure of Sections III and IV means even a correct essay-theme call may appear as the option you don't pick, and current-year statistics (cash rate, inflation, budget balance) date quickly — check the latest RBA and ABS figures before quoting them. Prepare the skill chains above, not memorised questions, and you're covered either way.

Want to check our working? every call above, with each model's own prediction
6 of 6 models expect this — consensus probability 0.50 Fiscal policy returns to Section III with a budget-balance chart 6 of 6 models expect this
eco-q1-fiscal-siii-stimulus Section III 20 marks discriminator consensus 0.50

The panel's unanimous essay-theme call: a 20-mark Section III stimulus question evaluating fiscal policy against economic growth plus one paired objective (external stability or distribution), with a budget-balance-as-%-of-GDP chart showing the 2022-24 surpluses swinging back to deficit, plus a Treasury/budget-paper quote. The 2019 Q25 / 2021 Q26 stimulus convention, unused since 2021 — the longest Section III absence of any Topic 4 policy. Five of six models make this their leading extended-response prediction; the sixth (opus) still rates it 0.30 as a stimulus question and predicts fiscal takes a full Section II block instead.

Each model's own prediction
  • Claude Fable 5: growth + distribution/external stability pairing, monetary as the paired option (p 0.35)
  • Claude Opus 5: budget-balance chart + Treasury quote, growth + one objective (p 0.30); fiscal SII block p 0.55
  • GPT-5.6 Sol: stance interpretation + deficit-financing evaluation, bold call (p 0.42)
  • Grok 4.6: growth and external stability with crowding-out and CAD/NPY channel (p 0.55)
  • Gemini 3.1 Pro: growth and external stability, cash balance + net foreign debt chart (p 0.65)
  • DeepSeek V4: 10-year budget/debt/GDP stimulus, stabilisers calculation embedded (p 0.70)

Marker-feedback lineage: 2019 Q25 (wrong budget era, stimulus restated), 2021 Q25 (judgement not sustained across objectives), 2022 Q21 (financing consequences underdeveloped)

A note on the numbers behind this call

The specific budget figures models quote (2022-23 surplus ~+$22bn, 2024-25 deficit ~-\$27bn) are NOT in the source pack — external-knowledge anchor (verify currency before publication). The chart SHAPE (surplus-to-deficit reversal) is the prediction; the numbers are world knowledge.

Pending human verification

This panel cites real legislation, cases or statistics from model knowledge rather than the NESA evidence pack. We're checking each is current and correctly named before treating it as fact.

In the practice paper: Q25

6 of 6 models expect this — consensus probability 0.57 The 'economy other than Australia' case study returns after its 2025 absence 6 of 6 models expect this
eco-q2-case-study-economy-returns Section II/IV 4–20 marks discriminator consensus 0.57

All six models call the return of the mandated Topic 1 case study — examined 2019, 2021, 2022, 2024 (Section II part with a printed 'Name of economy' line) and 2023 (Section IV essay), then absent in 2025 for the first time in the window: the clearest coverage gap in the pack. The panel splits on realisation: fable, opus and gpt-5.6-sol predict the 4-6 mark Section II part; grok, gemini and deepseek predict a full Section IV essay evaluating globalisation/development strategies in a named economy. Either way the marked skill is DEVELOPMENT evidence (HDI, life expectancy, literacy, poverty rates), not GDP growth.

Each model's own prediction
  • Claude Fable 5: SII 4-6 marks, due after 2025 skip (p 0.60); SIV globalisation essay (p 0.35)
  • Claude Opus 5: SII part with printed 'Name of economy' line (p 0.60)
  • GPT-5.6 Sol: compact SII evaluation, one trade + one social strategy (p 0.30)
  • Grok 4.6: SIV 'evaluate globalisation's impact... economy other than Australia', near-reprise of 2023 Q27 (p 0.58, bold call)
  • Gemini 3.1 Pro: SIV development-strategies essay (p 0.60)
  • DeepSeek V4: SIV case-study essay, Asian economy, inequality + environment angles (p 0.75)

Marker-feedback lineage: 2019 Q22(c), 2021 Q21(c), 2023 Q27, 2024 Q24(c) — growth/development conflation flagged in every one

In the practice paper: Q27

6 of 6 models expect this — consensus probability 0.76 Simple multiplier MC from a two-year income/savings table 6 of 6 models expect this
eco-q3-multiplier-mc Section I 1 marks mid range consensus 0.76

The panel's highest-probability single item: derive MPC/MPS from a two-year income-consumption or income-savings table, form k = 1/(1-MPC), apply it to a stated injection change and report the new equilibrium income. An annual fixture in all seven papers 2019-2025. Distractor logic consistently predicted: MPS used as MPC, the change in income returned instead of the level, k applied to the wrong base, multiplier omitted entirely.

Each model's own prediction
  • Claude Fable 5: eighth straight year, paired-table format (p 0.85)
  • Claude Opus 5: phrased from the table, not a stated MPC (p 0.85)
  • GPT-5.6 Sol: p 0.76
  • Grok 4.6: p 0.70
  • Gemini 3.1 Pro: stated-MPC routine variant (p 0.80)
  • DeepSeek V4: embedded in fiscal SII chain: MPC → new Y → crowding out (p 0.60)

Marker-feedback lineage: 2019 Q15, 2022 Q17, 2023 Q18, 2025 Q7 lineage; persistent calculation weaknesses flagged 2019-2022

In the practice paper: Q6

6 of 6 models expect this — consensus probability 0.73 BOP table with one unknown: CA + KAFA = 0 balancing item 6 of 6 models expect this
eco-q4-bop-balancing-item-mc Section I 1 marks mid range consensus 0.73

A balance of payments table under a floating exchange rate with one entry blank (financial account, exports, or a net income line), solved with the identity CA + KAFA = 0 — the annual Section I fixture (2021 Q18, 2022 Q11, 2024 Q12, 2025 Q13). Distractors encode the right magnitude with the wrong sign, the capital account omitted from the identity, or only the goods lines summed. Gemini and deepseek realise the same skill as a Section II current-account calculation opener.

Each model's own prediction
  • Claude Fable 5: p 0.80
  • Claude Opus 5: p 0.85
  • GPT-5.6 Sol: classification variant: aid/remittances/profits as debits and credits (p 0.72)
  • Grok 4.6: p 0.68
  • Gemini 3.1 Pro: SII: calculate CA balance from a BOP table (p 0.75)
  • DeepSeek V4: SII: trade balance + CA calculation parts (p 0.60)

Marker-feedback lineage: 2025 Q25 feedback (inverse CA/KAFA relationship not understood); 2023 Q22 same gap

In the practice paper: Q13 Q22

5 of 6 models expect this — consensus probability 0.74 Labour force calculation MC with the wrong-denominator trap 5 of 6 models expect this
eco-q5-labour-force-calculation-mc Section I 1 marks mid range consensus 0.74

Unemployment and/or participation rate from a table of working-age population, employed and unemployed persons — often back-solving for part-time employment or the participation rate (2019 Q14, 2021 Q17, 2024 Q6). The dominant predicted distractor: dividing by working-age population instead of the labour force, and treating a discouraged worker (hidden unemployment) as officially unemployed. Deepseek rests unemployment but still predicts this calculation survives as MC.

Each model's own prediction
  • Claude Fable 5: due after missing 2025 (p 0.75)
  • Claude Opus 5: back-solve variant with one rate given (p 0.70)
  • GPT-5.6 Sol: discouraged-worker second leg (p 0.68)
  • Grok 4.6: p 0.72
  • Gemini 3.1 Pro: p 0.85

Marker-feedback lineage: 2020 Q22(a) and 2025 Q23(b): participation and unemployment rates conflated; hidden unemployment misclassified

In the practice paper: Q9

5 of 6 models expect this — consensus probability 0.69 Paired protection-diagram MC: tariff revenue and foreign producers' revenue 5 of 6 models expect this
eco-q6-protection-diagram-pair Section I 1–2 marks discriminator consensus 0.69

The late-Section-I closer: two chained MC items sharing one domestic demand/supply diagram with a world price — compute government revenue after a tariff change, then the change in FOREIGN producers' revenue (valued at the world price, not the tariff-inclusive price) or the per-unit subsidy that eliminates imports. The 2019 Q19-20 / 2022 Q19-20 / 2025 Q17-18 lineage, always among the hardest MC items. Distractors multiply the tariff by total consumption, use pre-change quantities, or read producer revenue at domestic output. Gemini is the dissenter, resting protection to at most one MC after the 2025 Q28 essay.

Each model's own prediction
  • Claude Fable 5: Q17-20 slot, two chained items (p 0.80)
  • Claude Opus 5: tariff revenue (p 0.78) + foreign-producer revenue as final MC (p 0.50)
  • GPT-5.6 Sol: p 0.70
  • Grok 4.6: paired diagram closer, Q19-20 (p 0.78)
  • DeepSeek V4: SII consumer/producer-surplus welfare variant (p 0.40)

Marker-feedback lineage: 2024 Q21(a): could not 'accurately interpret the diagram and calculate the economic implications of protection'; 2022/2025 final-MC placement

In the practice paper: Q19

6 of 6 models expect this — consensus probability 0.63 Fiscal stance vs outcome: automatic stabilisers and the structural/cyclical split 6 of 6 models expect this
eco-q7-fiscal-stance-stabilisers Section I/II 1–6 marks mid range consensus 0.63

Classify a two-year budget-outcome change as expansionary or contractionary, separating discretionary (structural) change from automatic-stabiliser (cyclical) movement — the recurring trap being that a SHRINKING deficit is contractionary even while the budget stays in deficit (2019 Q16, 2021 Q4/Q5, 2023 Q8, 2025 Q11). Four models also predict a Section II escalation: given budget data with an unemployment change, identify the cyclical component and explain a consequence of deficit financing (crowding out via bond sales). Gemini and deepseek make the never-directly-examined structural/cyclical calculation a bold call.

Each model's own prediction
  • Claude Fable 5: MC stance/stabiliser classification (p 0.70); surplus-use coverage gap SII (p 0.40)
  • Claude Opus 5: MC (p 0.75); SII 2/3/5 ladder on financing + stance (p 0.55)
  • GPT-5.6 Sol: SII: stance vs outcome + stabilisers off budget data (p 0.58)
  • Grok 4.6: MC discretionary vs automatic (p 0.72); SII fiscal chain (p 0.62)
  • Gemini 3.1 Pro: SII: cyclical component + crowding out (p 0.55); structural-vs-cyclical calculation bold call (p 0.40)
  • DeepSeek V4: stabilisers calculate-explain-justify bold call (p 0.50)

Marker-feedback lineage: 2022 Q21(b)-(c): one financing method/factor given where the stem used plurals; 2024 feedback: cyclical/structural confusion

In the practice paper: Q12 Q24

5 of 6 models expect this — consensus probability 0.62 Headline vs underlying inflation: assign the cause to the correct measure 5 of 6 models expect this
eco-q8-headline-underlying-inflation Section I 1–2 marks discriminator consensus 0.62

A table or scenario where headline and underlying inflation diverge: pick the row assigning a volatile one-off (cyclone hitting fruit supply, fuel excise change, oil shock) to headline and a persistent driver (broad wage pressure, minimum wage rise) to underlying — the 2020 Q15 / 2024 Q16 / 2025 Q14 design. Distractors swap the assignments or treat a falling inflation rate as deflation (disinflation confusion, flagged 2021). Deepseek escalates it to a Section II part explaining the divergence off a weighted CPI table.

Each model's own prediction
  • Claude Fable 5: cause-pairing MC (p 0.60)
  • Claude Opus 5: opposite-directions table row-select (p 0.60)
  • GPT-5.6 Sol: with disinflation-vs-deflation second leg (p 0.65)
  • Grok 4.6: cyclone/fruit-supply variant (p 0.68)
  • DeepSeek V4: SII: explain divergence + justify policy response (p 0.56)

Marker-feedback lineage: 2021 Q23(b): low inflation within the band confused with deflation; 2024 Q16 / 2025 Q14 lineage

In the practice paper: Q8

5 of 6 models expect this — consensus probability 0.53 Financial account inflows generate later net primary income debits 5 of 6 models expect this
eco-q9-kafa-npy-link Section II 3–5 marks discriminator consensus 0.53

The marking centre's most repeated BOP complaint turned into a Section II discriminator: explain the causal chain from a capital and financial account surplus (foreign investment/borrowing credits) to servicing costs that appear as net primary income debits in later periods — off a compact BOP table, demanding cause and effect rather than a restatement of the accounting identity. Flagged as not understood in 2023 Q22(b) and again at essay length in 2025 Q25.

Each model's own prediction
  • Claude Fable 5: sustained financial-account surpluses → servicing costs (p 0.45)
  • Claude Opus 5: supplied-table variant, 3-5 marks (p 0.45)
  • GPT-5.6 Sol: goods surplus + NPY deficit table (p 0.40)
  • Gemini 3.1 Pro: CA calculation then the structural KAFA/NPY link (p 0.75)
  • DeepSeek V4: trade balance, CA, then the FA-surplus/NPY-deficit link (p 0.60)

Marker-feedback lineage: 2023 Q22(b) (cause and effect, not description), 2025 Q25 (only the CA side understood)

In the practice paper: Q22

6 of 6 models expect this — consensus probability 0.51 Micro/labour reform traced to productivity, aggregate supply and cost inflation 6 of 6 models expect this
eco-q10-micro-productivity-chain Section II 4–20 marks discriminator consensus 0.51

All six models predict the supply-side chain discriminator: a NAMED microeconomic or labour-market reform (competition policy, deregulated product market, skills/training program, enterprise bargaining) traced through efficiency type → productivity → unit labour costs → aggregate supply → cost inflation or structural unemployment. Fable hangs it on Australia's stalled productivity (GDP-per-hour flatline) as the 2026 contemporary hook; grok and gemini/deepseek escalate to a 20-mark (SIII micro-reform-on-employment-and-inflation reprising 2019 Q26, or an SIV efficiency essay). The chain, not the policy description, is what scores.

Each model's own prediction
  • Claude Fable 5: SII 4-5 marks, productivity-stagnation hook (p 0.50); full essay (p 0.30)
  • Claude Opus 5: SII with printed 'Product market' line (p 0.50); labour chain part (p 0.50)
  • GPT-5.6 Sol: reform → efficiency → AS chain in a named market (p 0.51)
  • Grok 4.6: SIII micro reform on employment + inflation, 2019 Q26 reprise (p 0.48)
  • Gemini 3.1 Pro: SIV efficiency/structural-change essay (p 0.55)
  • DeepSeek V4: SIV productivity + competitiveness essay, ONE product + ONE factor market (p 0.50)

Marker-feedback lineage: 2020 Q24(b) (no product market identified), 2023 Q24(c) (macro/micro confusion), 2024 Q22(c) (more than one labour policy required), 2025 Q23(d) (NAIRU vs cyclical policies)

A note on the numbers behind this call

Fable's GDP-per-hour-worked flatline (~0-1% since 2016) and the 2025 productivity-roundtable framing are external-knowledge anchors (verify currency before publication); the 2024 Q25 productivity stimulus table is in the pack.

Pending human verification

This panel cites real legislation, cases or statistics from model knowledge rather than the NESA evidence pack. We're checking each is current and correctly named before treating it as fact.

In the practice paper: Q24

6 of 6 models expect this — consensus probability 0.51 Market-based environmental policy vs regulation, with the externality diagram 6 of 6 models expect this
eco-q11-env-market-based-vs-regulation Section II 4–10 marks discriminator consensus 0.51

A Section II environment part demanding ONE named market-based policy (price-incentive mechanism: ETS/safeguard mechanism, container deposit, water trading, carbon/per-unit tax) explicitly contrasted with regulation — the single most repeated environment feedback item (2021, 2024). Four models attach the negative-externality diagram: draw or complete the social-cost curve, and/or the satellite MC identifying the corrective per-unit tax as the vertical gap between social and private cost at the socially optimal quantity (2023 Q14, 2025 Q20 — fable 0.55, opus 0.40, gpt-5.6-sol 0.58, grok 0.50). Grok and deepseek make it a full 10-mark SII with limitations of achieving ESD.

Each model's own prediction
  • Claude Fable 5: assess ONE market-based policy with named Australian example (p 0.50); externality MC (p 0.55)
  • Claude Opus 5: named-scheme-with-data variant (p 0.45); classification MC (p 0.65)
  • GPT-5.6 Sol: compare instrument vs regulation vs target vs agreement (p 0.39)
  • Grok 4.6: 1+4+5 SII: draw social cost, market-based policy, ESD limitations (p 0.52)
  • Gemini 3.1 Pro: graph-completion + market-based vs regulatory comparison (p 0.65)
  • DeepSeek V4: coal-power externality: draw curve, per-unit tax, evaluate vs regulation (p 0.55)

Marker-feedback lineage: 2021 Q22(b) and 2024 Q23(c) (market-based confused with regulation); 2023 Q14 / 2025 Q20 (wrong vertical gap)

A note on the numbers behind this call

Named Australian schemes beyond the pack (safeguard mechanism, container-deposit/Return and Earn appears once, EV incentives) are external-knowledge anchors (verify currency before publication).

Pending human verification

This panel cites real legislation, cases or statistics from model knowledge rather than the NESA evidence pack. We're checking each is current and correctly named before treating it as fact.

In the practice paper: Q23

4 of 6 models expect this — consensus probability 0.49 External stability returns as the synthesis 20-marker (CAD, NFL, terms of trade) 4 of 6 models expect this
eco-q12-external-stability-synthesis Section III/IV 20 marks discriminator consensus 0.49

Four models converge on external stability as the overdue big-mark T3 issue — last anchoring an extended response in 2020 Q25 (2025's SIII was BOP composition, a T2 question, not the sustainability issue): a stimulus essay off CAD and net-foreign-debt-as-%-of-GDP series (grok, gemini, deepseek SIII; gpt-5.6-sol SIV without stimulus), demanding the savings-investment gap, valuation effects, NPY servicing, and a cyclical-vs-structural judgement. Gemini and deepseek both push the Pitchford 'consenting adults' thesis as the top-band discriminator. Pairs naturally with eco-q1 as the second Section III option.

Each model's own prediction
  • GPT-5.6 Sol: SIV synthesis: ToT, competitiveness, saving, borrowing, valuation (p 0.34)
  • Grok 4.6: SIII CAD + NFD stimulus, Pitchford-style judgement, bold call (p 0.50)
  • Gemini 3.1 Pro: SIII with ToT + NFL stimulus; Pitchford/NPY bold call (p 0.45-0.50)
  • DeepSeek V4: SIII commodity-shock transmission: NFL calc, KAFA/NPY link, ToT deterioration (p 0.60)

Marker-feedback lineage: 2020 Q25 (causes explained instead of impacts), 2019-2022 (shallow external-stability components)

In the practice paper: Q26

Watch list — worth having ready (10)

  • TWI vs bilateral rate divergence (AUD up against USD, down on the TWI) — five models, formats too scattered to cluster: fable SII 0.40, opus MC-weights-table 0.45, grok SII 3-marker 0.48, gemini SII 0.60, deepseek MC 0.55. The renminbi ~31% / yen 13% / euro 9% weights table IS in the pack (past MC) — corpus-grounded

  • Net foreign liabilities calculation MC — NFL = net debt + net equity, % of GDP, judgement second leg (fable 0.60, opus 0.60, gpt-5.6-sol 0.55, grok 0.65); biennial rhythm due again 2026

  • International organisation role MC as the Section I opener — IMF vs World Bank vs WTO mandate swap (fable 0.75, opus 0.80, gpt-5.6-sol 0.57, grok 0.75)

  • CPI/real-GDP-growth deflation calculation (fable MC 0.85, opus MC 0.70, gpt-5.6-sol SII 0.48, deepseek SII per-capita variant 0.65) — annual fixture, format uncertain

  • Monetary policy Section III option — fable's minority call (35% chance): cash-rate round-trip chart 0.10% → 4.35% → ~3.60% evaluated across BOTH phases. The rate path is an external-knowledge anchor (verify currency before publication) — none of these figures are in the pack. Panel majority rests monetary after 2025 Q24 (grok/gemini/deepseek MC-only)

  • Inflation stakeholder essay — individuals/firms/government frame applied to inflation, the unused T3 fit for the 2022/2024/2025 SIV craft (grok 0.52; fable and opus predict the frame in question_type_trends)

  • 1-mark sketch/graph-completion returns after vanishing in 2024-25 (opus bold 0.35 — Lorenz/social-cost/subsidy shift; grok subsidy-to-zero-imports sketch 0.35; gemini graph-modification trend; deepseek externality part (a))

  • 2025 US tariff escalation as a contemporary Section II hook on Australia's trade flows (fable SII 0.45 + bold 0.35, opus partner-protection part 0.40) — external-knowledge anchor (verify currency before publication): the pack ends at the 2025 paper and does not contain the tariff round

  • RBA ample-reserves system / interest-rate corridor implementation MC (opus bold 0.25, deepseek corridor-diagram 0.55, gemini transmission 0.70) — 'Ample Reserves system (2025)' is in the pack's 2025 marking guidelines

  • Panel's rested calls: distribution is the unanimous rest (5/6 rested votes; 2025 SIII Q26 essay + 2023 Q23 — expect at most a Gini/Lorenz MC); the protection ESSAY rests after 2025 Q28 (topic contested overall, stddev 0.27: gemini/deepseek rest it entirely, fable/opus/grok keep the MC pair); monetary and unemployment keep MC but lose their 10-mark blocks after 2025 Q24/Q23

Where to spend your study time

How likely each topic is to appear this year.

T4: Fiscal policy 94% likely

Chance of a big question (4+ marks) here: 78%

Question types predicted here stimulus based ×5 short answer ×5 multiple choice ×3 essay ×1 extended response ×1

What each model expects

  • DeepSeek V4: A 20-mark stimulus essay on fiscal policy and the trade-off between growth and inflation, incorporating multiplier and NAIRU concepts.
  • Claude Fable 5: Fiscal returns to Section III with a surplus-to-deficit budget chart, paired against the monetary option, echoing the 2021 twin-macro-policy Section III.
  • Gemini 3.1 Pro: A major Section III extended response evaluating fiscal policy's impact on growth and external stability, alongside a Section II question on deficit financing.
  • GPT-5.6 Sol: Fiscal policy is the leading candidate for a 2026 stimulus essay, centred on interpreting stance and evaluating short-run stabilisation against longer-run financing and capacity constraints.
  • Grok 4.6: A 10-mark Section II fiscal chain (definition, deficit financing, current stance) is the highest-conviction SII item, with a genuine chance of a budget-stimulus SIII pairing growth and external stability.
  • Claude Opus 5: Fiscal policy takes a full 10-mark Section II question in 2026 for the first time since 2022, most likely built on financing a deficit or the structural/cyclical split of the budget balance.
T1: Global economy, globalisation, case study 91% likely

Chance of a big question (4+ marks) here: 72%

Question types predicted here essay ×4 short answer ×4 multiple choice ×4 extended response ×1

What each model expects

  • DeepSeek V4: A case-study essay in Section IV, focusing on the effects of globalisation on development indicators in an Asian economy, with emphasis on income inequality and environmental degradation.
  • Claude Fable 5: The case-study evaluation returns to Section II at 4-6 marks after its 2025 absence, and one Section IV option is a globalisation-and-development essay.
  • Gemini 3.1 Pro: Section IV extended response focusing entirely on the global economy case study, specifically evaluating development strategies.
  • GPT-5.6 Sol: Globalisation will appear through an organisation-role distractor and possibly a compact case-study evaluation, but not another broad international-integration essay.
  • Grok 4.6: Section IV revives the case-study evaluate on an economy other than Australia, last used as a 20-marker in 2023 Q27.
  • Claude Opus 5: The case study of an economy other than Australia returns as a Section II part worth 4–6 marks with a printed 'Name of economy' line, after being absent from the 2025 paper for the first time in this window.
T3: Inflation 90% likely

Chance of a big question (4+ marks) here: 67%

Question types predicted here multiple choice ×6 short answer ×5 essay ×1 extended response ×1

What each model expects

  • DeepSeek V4: A short-answer requiring calculation of a weighted inflation rate and a policy-recommendation justification based on the acceleration of underlying inflation.
  • Claude Fable 5: Two quantitative inflation MC items (a CPI calculation plus a headline/underlying cause pairing) and a Section II effects question framed around disinflation back to the target band.
  • Gemini 3.1 Pro: Section II multi-part question requiring a CPI calculation followed by an explanation of inflation's transmission to external competitiveness.
  • GPT-5.6 Sol: Inflation will return beyond multiple choice through a cause-to-policy chain, with the highest marks awarded for matching the policy instrument to demand or supply-side inflation.
  • Grok 4.6: Inflation is the leading candidate for the Australia-side Section IV essay, reviving the individuals–firms–government frame last used for inflation in 2022 Q27.
  • Claude Opus 5: Inflation returns as a Section II component chained to either labour market policy or international competitiveness, with a headline-versus-underlying MC that requires assigning causes to the correct measure.
T3: Economic growth 89% likely

Chance of a big question (4+ marks) here: 58%

Question types predicted here multiple choice ×8 short answer ×3 stimulus based ×1

What each model expects

  • DeepSeek V4: A calculation-based short-answer combining real GDP growth, per capita adjustments, and a critique of GDP as a welfare measure.
  • Claude Fable 5: The multiplier MC appears for an eighth straight year; growth features in Section II data work rather than as an essay after the 2024 Section III pairing.
  • Gemini 3.1 Pro: Multiplier calculation in multiple choice; growth concepts integrated into macro extended responses rather than standalone.
  • GPT-5.6 Sol: Section I will contain a multiplier or real-growth calculation, while Section II will require a causal link from an AD or AS change to output and another economic objective.
  • Grok 4.6: A multiplier MC is near-certain; growth itself is examined as an objective inside a policy essay rather than as a stand-alone 20-marker after 2024 Q25.
  • Claude Opus 5: A multiplier or MPC calculation will appear in Section I, most likely between Q7 and Q18, phrased from a two-year income/savings table rather than from a stated MPC.
T4: Microeconomic and labour market policies 88% likely

Chance of a big question (4+ marks) here: 66%

Question types predicted here short answer ×5 essay ×3 multiple choice ×3 stimulus based ×1 extended response ×1

What each model expects

  • DeepSeek V4: A Section IV essay integrating micro reform effects on productivity and NAIRU, with explicit reference to labour market deregulation and national competition policy.
  • Claude Fable 5: Productivity stagnation is 2026's supply-side hook: a Section II discriminator ties a named reform to aggregate supply, with a live chance of a full micro/labour essay.
  • Gemini 3.1 Pro: Section IV essay requiring students to link specific microeconomic reforms to different types of efficiency and structural change.
  • GPT-5.6 Sol: The microeconomic discriminator will require an explicit reform-to-efficiency-to-aggregate-supply chain, probably in a named labour or product market rather than a generic policy description.
  • Grok 4.6: Microeconomic reform on employment and inflation is the second SIII option, recycling the 2019 Q26 pairing that has not returned as a 20-mark.
  • Claude Opus 5: Microeconomic or labour market policy takes the Topic 4 slot in Section II with a named-market answer line, after being displaced entirely by monetary policy in 2025.
T2: Trade, balance of payments, foreign investment 88% likely

Chance of a big question (4+ marks) here: 49%

Question types predicted here short answer ×6 multiple choice ×6

What each model expects

  • DeepSeek V4: A Section II data-response question that tests the link between capital flows and the net primary income deficit, using recent Australian BOP data.
  • Claude Fable 5: The annual balancing-item MC calculation appears again, and a Section II part probes the CA/KAFA or NPY link; no BOP essay after 2025's Section III Q25.
  • Gemini 3.1 Pro: Section II calculation of the Current Account, followed by an explanation of the savings-investment gap driving the NPI deficit.
  • GPT-5.6 Sol: Balance-of-payments coverage will retreat from the 2025 essay to a debit-credit classification or missing-balance calculation, probably followed by a short causal explanation.
  • Grok 4.6: BOP is calculation/classification MC only after 2025 Q25 already asked trends in size and composition since 2016.
  • Claude Opus 5: Section I will carry two balance-of-payments numeric items — one solving the BOP identity for a missing component and one on net foreign liabilities — and Section II will use a BOP table as the stem for a 2/3/5 ladder.
T3: Environmental sustainability 86% likely

Chance of a big question (4+ marks) here: 62%

Question types predicted here short answer ×5 multiple choice ×5 stimulus based ×2 extended response ×1

What each model expects

  • DeepSeek V4: A multi-part environment question in Section II that requires drawing a negative-externality graph and comparing a carbon tax with a direct-emission regulation, referencing Australian policies.
  • Claude Fable 5: The environment essay cycle (2020, 2023) points to a 2026 Section III net-zero option; failing that, a market-based-policy Section II discriminator is near-certain.
  • Gemini 3.1 Pro: Section II question requiring the modification of a market failure graph and a comparison of market-based vs regulatory environmental policies.
  • GPT-5.6 Sol: Environmental coverage will move beyond another isolated market-based instrument and instead test the comparative operation and limitations of different policy categories.
  • Grok 4.6: Environment is a full 10-mark SII (social-cost diagram plus market-based policy plus limitations) rather than a 20-mark after 2023 Q26 and 2024–25 SII parts.
  • Claude Opus 5: Environment appears at reduced weight in 2026 — one Section I policy-classification item plus a 2–5 mark Section II part — rather than the full 10-mark block it held in 2024 or the Section III essay of 2023.
T2: Exchange rates and their effects 85% likely

Chance of a big question (4+ marks) here: 44%

Question types predicted here multiple choice ×6 short answer ×4 extended response ×1

What each model expects

  • DeepSeek V4: One demanding MC item requiring analysis of a parallel shift in demand for AUD and its disparate effect on the TWI versus the USD bilateral rate.
  • Claude Fable 5: Exchange rates retreat to MC plus at most a small Section II part; no 20-mark treatment after the 2024 Q28 essay and 2025 Q22 10-marker.
  • Gemini 3.1 Pro: A 4-mark Section II question linking exchange rate depreciation to imported inflation and distinguishing the TWI.
  • GPT-5.6 Sol: After the 2025 depreciation chain, 2026 will narrow exchange-rate assessment to intervention or graph mechanics rather than another broad effects question.
  • Grok 4.6: Exchange rates appear as MC plus a 3–5 mark SII part on TWI versus bilateral rates, not as a 20-mark after 2024 Q28 and 2025 Q22.
  • Claude Opus 5: Exchange rates will appear in Section I at least twice but, after taking the whole of Section II Q22 in 2025, will not occupy a full 10-mark Section II question again in 2026.
T3: External stability 84% likely may rest

Chance of a big question (4+ marks) here: 64%

Question types predicted here multiple choice ×4 stimulus based ×3 essay ×1 extended response ×1

What each model expects

  • DeepSeek V4: A structured stimulus essay unpacking the transmission of a commodity-price shock through the Balance of Payments, with calculation of net foreign liabilities and analysis of the savings–investment gap.
  • Claude Fable 5: External stability stays quantitative: a ToT index item and likely an NFL calculation in Section I, with essay treatment rested after 2025's BOP-trends Q25.
  • Gemini 3.1 Pro: Section III extended response focusing on external stability, specifically demanding analysis of the savings-investment gap and net foreign liabilities.
  • GPT-5.6 Sol: A broad external-stability essay is a plausible 2026 return, requiring measures beyond the current account and explicit distinction between cyclical improvement and structural exposure.
  • Grok 4.6: External stability is the largest T3 coverage gap and is the leading SIII companion to fiscal, using CAD/NFD charts last seen as a 20-mark in 2020 Q25.
T4: Monetary policy 81% likely

Chance of a big question (4+ marks) here: 42%

Question types predicted here multiple choice ×8 short answer ×3 stimulus based ×1

What each model expects

  • DeepSeek V4: A few multiple-choice questions testing the mechanics of monetary policy implementation, with one using a corridor diagram.
  • Claude Fable 5: A Section III stimulus essay evaluating monetary policy across the 2022-25 hike-then-cut cycle, offered against a fiscal alternative, mirroring the 2021 Q25/Q26 pairing.
  • Gemini 3.1 Pro: Multiple choice focus on the transmission mechanism and the RBA's interest rate corridor.
  • GPT-5.6 Sol: Monetary policy will be rested from another full Section II chain and instead tested through transmission or a concise stance-and-limitation problem.
  • Grok 4.6: Monetary policy is MC-only in 2026 after 2025 Q24 already examined purpose, the corridor and a cash-rate cut.
  • Claude Opus 5: Monetary policy holds Section I and may anchor one Section III stimulus question on the current easing cycle, but will not repeat as the whole of a Section II 10-mark question after 2025.
T3: Unemployment 75% likely may rest

Chance of a big question (4+ marks) here: 44%

Question types predicted here multiple choice ×6 short answer ×3 extended response ×1

What each model expects

  • Claude Fable 5: Unemployment is confined to a Section I calculation and a small typology part; the 10-mark slot rests after 2025's NAIRU question and 2024's full-employment essay.
  • Gemini 3.1 Pro: Labour force calculations in multiple choice, with distractors targeting the inclusion of discouraged workers.
  • GPT-5.6 Sol: A labour-force calculation will be paired with a discriminator requiring students to separate cyclical unemployment, participation changes and policies that genuinely lower NAIRU.
  • Grok 4.6: Unemployment is MC-plus-short-SII after 2024 Q26 (full employment policies) and 2025 Q23 (NAIRU/participation); no third consecutive substantial item.
  • Claude Opus 5: Unemployment holds two Section I items — one labour force calculation and one participation rate causal item — but does not repeat the full 10-mark Section II block it took in 2025.
T2: Free trade, protection, agreements 73% likely models split may rest

Chance of a big question (4+ marks) here: 37%

Question types predicted here short answer ×5 multiple choice ×5 extended response ×1

What each model expects

  • DeepSeek V4: A protection diagram question requiring computation of welfare effects and explanation of deadweight loss.
  • Claude Fable 5: Two chained protection-diagram MC items close Section I, and global tariff escalation supplies a Section II context; the 20-mark protection essay rests after 2025 Q28.
  • GPT-5.6 Sol: Protection will be limited chiefly to a late Section I graph calculation, with no protection essay after its dominant 2025 Section IV treatment.
  • Grok 4.6: Protection is MC-diagram only in 2026 after the 2025 Q28 methods-of-protection essay; expect a tariff/quota revenue pair, not another 20-mark.
  • Claude Opus 5: At least one, and probably two, multiple-choice items will require a numeric answer read off a protection diagram, with a tariff-revenue or foreign-producer-revenue calculation among the last five MC items.
T3: Distribution of income and wealth 55% likely may rest

Chance of a big question (4+ marks) here: 40%

Question types predicted here short answer ×1

What each model expects

  • DeepSeek V4: A short-answer item on interpreting a movement in the Lorenz curve and proposing redistribution policies.

How likely each topic is to appear. Open a topic for the question types to practise there.

Practise with the paper

Economics practice paper

100 marks · 28 questions

Every question is traceable to the consensus prediction behind it — open the web version and each question carries a “why this question” link into the evidence. All questions are original Intuition compositions in NESA style.

Read it on the web Download PDF Solutions (PDF)

Exam Success Program

Book our ESPs now — online or in person

A paper on your desk is one thing — the clock running is another. Sit Economics practice exams under real conditions and get them marked with real feedback. Exam times run seven days a week, right up until the HSC begins.

Book your ESP →

Places are still open for 2026

Intu AI

One paper isn't enough? Generate more

Intu AI builds unlimited practice questions for Economics in these styles, marks your working, and explains what you missed — aligned to your syllabus.

Practise with Intu AI →

Share & save

Practice paper (PDF)

Every style card and paper question has its own link — hover any card and use its copy-link icon to share exactly the thing you mean. Printing this page gives a clean copy too.

Published Aug 2026, before the exams. In November 2026 we score these predictions publicly against the real paper — per-model calibration and question-level hit rates, the same harness as the 2025 backtest. How we did it.