Fiscal policy comes up again and again in H2 Economics 9570, and it is the topic where I see the widest gap between what students write and what the mark scheme rewards. This post sets out the method I teach: how to build a fiscal policy description that earns analysis marks, and how to build an evaluation that earns AO4 credit instead of restating advantages and disadvantages.
Two things in the official syllabus explain why this matters so much.
First, evaluation sits inside most of the marks. In the SEAB scheme of assessment, about 60% of the marks in each paper come from questions that test evaluation (AO4) alongside knowledge, application and analysis. Evaluation is not a bonus line at the end of an essay. It is built into the majority of both papers.
Second, the syllabus tells you which evaluative angles it wants. In the Additional Information column of Theme 3.2.3, four statements use the words "is required". Three of them bear directly on fiscal policy:
- that governments focus on different macroeconomic objectives "depending on the state of the economy and the level of development of the country"
- "an awareness of the desirability for a government to maintain fiscal sustainability over the long term"
- "an understanding of how transfer payments can improve income distribution and help achieve inclusive growth"
Those are the evaluative angles for fiscal policy, published in advance. Very few students I meet have read them.
The principle I use throughout is chains, not checklists. You are not going to memorise an acronym and deploy it. You are going to follow the economic logic step by step until it reaches the variable the question asks about, and then question that chain where it is weakest. That is what evaluation is. If you want the general version of this idea across subjects, I wrote about it in how evaluation works in IGCSE, IB and H2 Economics.
One syllabus point to settle first: discretionary only
The 9570 syllabus describes fiscal policy as "how discretionary fiscal policy can influence the level of economic activities and living standards through government spending and taxation", and lists "discretionary fiscal policy" under Concepts and Tools of Analysis.
The words "automatic" and "stabiliser" do not appear anywhere in the syllabus document. If your notes spend pages on them as a core tool, that space is not earning marks. Mentioning them will not cost you anything, but building a whole evaluation paragraph around them spends time where no marks sit. Everything here is about discretionary fiscal policy: deliberate, announced decisions to change government spending or taxation.
Part 1: Description, or building a chain that reaches the end
Why most descriptions stop short
Here is the chain most students write:
Government increases spending, so AD increases, so national income increases.
Three links. Nothing in it is wrong, and it earns very little, because it skips every step where the economics happens. The real question is by how much, through what, under what conditions and with what side effects.
Here is the same chain, completed:
- Identify the instrument.
- Identify which component of AD it enters.
- Size the initial injection.
- Work the multiplier.
- Shift AD.
- Read the outcome off the AS curve at the relevant range.
- State the effect on the specific variable the question asked about.
Seven links. The sixth is the one that separates a strong answer from an average one, and it is the one I see missing most often.
Step 1: Name the instrument precisely
"The government uses expansionary fiscal policy" is not precise enough to analyse. Fiscal policy works through two instruments, and each has sub-types with different effects.
Government spending (G)
- Government consumption: direct purchases of goods and services such as healthcare, defence and public sector wages. This enters AD directly as G.
- Government investment: infrastructure and R&D. This enters AD as G in the short run and also raises long-run aggregate supply. That dual effect is the most useful distinction in the whole topic.
- Transfer payments: vouchers, rebates and CPF top-ups. These do not enter AD as G. They raise household disposable income and reach AD indirectly through C, and only to the extent that recipients spend them.
Taxation (T)
- Direct taxes on income and profit: these change disposable income and therefore C, and change the after-tax return on investment and therefore I.
- Indirect taxes such as GST: these affect prices and real purchasing power directly, and they fall on income groups very differently from direct taxes.
The transfer payment point is where I see students go wrong most often with Singapore Budget material. Vouchers are not an increase in G; they raise disposable income, and their effect on AD depends on how much recipients spend. Writing "CDC Vouchers increase G and therefore AD" is a mechanical error that a marker spots immediately.
Step 2: Size the injection, then work the multiplier
The multiplier is where the size of the effect is decided, and it is the first real piece of analysis.
k = 1 / MPW, where MPW = MPS + MPT + MPM
The value of the multiplier depends on how much of each extra dollar of income leaks out of the circular flow through saving, taxes and imports instead of being spent again at home. The larger the leakages, the smaller the multiplier.
This is where Singapore is a special case, and where you can stand apart from a generic answer. Singapore is a small and very open economy with a high marginal propensity to import. A large share of any extra domestic spending leaks abroad quickly, because so much of what is consumed here is imported. A high MPM means a high MPW, which means a small multiplier.
So demand-side fiscal policy is structurally less powerful in Singapore than in a large, less open economy: a dollar of stimulus buys less extra domestic output here than it would in the United States. That helps explain why Singapore's fiscal policy leans towards supply-side measures and redistribution, and it lets you explain the shape of the Singapore Budget instead of listing its contents.
Step 3: Read the outcome off the AS curve
This is the step that decides your grade, and the one most often missing.
A rise in AD does not have one outcome. It has three, depending on where the economy is operating:
- Keynesian (horizontal) range. Lots of spare capacity and high unemployment. AD rises, real output rises a lot and the price level barely moves. Expansionary fiscal policy works hardest here.
- Intermediate (upward-sloping) range. Some spare capacity, but tightening. AD rises, and both real output and the price level rise. There is a real trade-off between growth and price stability.
- Classical (vertical) range. Full employment. AD rises, real output does not change and the price level rises. Expansionary fiscal policy here is purely inflationary, so the policy that was right in the first case does harm in this one.
The point you can take into an exam: how well fiscal policy works depends on the state of the economy, not on the policy itself. This is the point the syllabus flags when it requires an understanding that governments focus on different objectives "depending on the state of the economy".
If a question gives you information about unemployment, capacity use or the output gap, that information is there because the AS range is part of the answer.
Step 4: Land on the variable the question asked about
Chains that stop one step short are the most common reason answers stall in the middle bands. If the question asks about living standards, you are not finished at "real output rises". You need the next links: real output rises, so real income per head rises, so the material standard of living rises, subject to how that income is distributed and what happens to non-material well-being.
If the question asks about unemployment, take it to the labour market: AD rises, the derived demand for labour rises and demand-deficient unemployment falls. Name the type. Fiscal stimulus deals with demand-deficient unemployment. It does not deal with structural or frictional unemployment, which the syllabus names separately. That distinction alone is a ready-made evaluative point.
If the question asks about inclusive growth or income distribution, take it to the distributional outcome, which is where transfer payments belong.
Part 2: Evaluation, or questioning your own chain
Evaluation is not a paragraph you bolt on at the end. It is going back to your chain and asking, at each link: what would have to be true for this to hold, and what if it is not? Generic evaluation names a limitation. Contextual evaluation shows why it matters, or does not, in this case. I give students one test:
Would this sentence survive unchanged in an essay about a different country or a different policy? If yes, it is not evaluation yet. It is a general fact about fiscal policy, and it earns like one.
Below are the six chains that do the work for fiscal policy. They are not a checklist. In a 25-mark essay you will develop two or three properly instead of listing all six.
Chain 1: Magnitude, or how big is the effect really?
The chain: size of policy, then the multiplier, then MPW, then openness, then the actual change in output.
The questioning: a headline stimulus figure tells you the injection, not the outcome. In Singapore the high marginal propensity to import shrinks the multiplier, so the same headline number produces less domestic output than it would elsewhere.
Also question which instrument is used. A dollar of transfer payments has a smaller effect on AD than a dollar of government consumption, because part of every transfer is saved, and the share saved rises with the income of the recipient. A voucher paid to every household therefore has a smaller AD effect per dollar than a means-tested payment to lower-income households, who spend more of each extra dollar. That point has real bite when a question involves payouts that are not targeted.
Chain 2: Context, or is the economy in the right state for this?
The chain: output gap, then the AS range, then the trade-off between objectives.
The questioning: expansionary fiscal policy near full employment produces inflation, not growth. Contractionary policy in a downturn deepens it. The syllabus requires awareness that the right objectives vary with the state of the economy, so an evaluation that treats fiscal policy as always effective or always weak has missed a required point.
This is also where conflicts between macroeconomic objectives come in, which the syllabus names at 3.2.3(c). Expansionary policy for growth and jobs may worsen price stability and, by raising import demand in a small open economy, may worsen the balance of trade. Naming a conflict is knowledge. Weighing which objective should come first, given the context in the question, and saying why, is evaluation.
Chain 3: Time, or will it arrive when it is needed?
The chain: recognition lag, then decision lag, then implementation lag, then impact lag.
The questioning: fiscal policy is slow. Data confirming a downturn arrives late, a Budget follows a fixed parliamentary timetable, infrastructure spending is paid out over years, and the multiplier works through rounds of spending over time. The risk: stimulus can arrive at the wrong time, landing after recovery and adding to inflation instead of closing an output gap.
Singapore's 2020 response is the strongest counter-example, and using it as a counter-example is what makes it evaluative. Four Budgets were delivered in a little over three months: Unity on 18 February, Resilience on 26 March, Solidarity on 6 April and Fortitude on 26 May. That is not the speed the standard lag critique describes. So the evaluative point is conditional: the lag argument is strong in normal times and weak in an acute crisis, when a government has both the institutional capacity and the fiscal room to act quickly.
Chain 4: Sustainability, or can it be repeated? (syllabus-required)
The chain: policy, then the budget position, then financing, then the long-run constraint.
The syllabus requires "an awareness of the desirability for a government to maintain fiscal sustainability over the long term". This chain is named content, not decoration.
The standard version: persistent deficits need borrowing, debt builds up, interest payments crowd out future spending, and government borrowing may crowd out private investment through higher interest rates.
The Singapore version is where the marks are, because Singapore's fiscal position changes this chain, not just softens it. Singapore's headline budget balance is not simply revenue minus spending. One of the items between them is the Net Investment Returns Contribution (NIRC): the share of returns on the reserves, managed by GIC, MAS and Temasek, that the Government may spend. It is capped at up to 50% of the expected long-term real returns, with the rest kept to preserve the reserves.
From the Ministry of Finance's Analysis of Revenue and Expenditure for FY2026 (Table 3.1a):
| Item | FY2020 | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|---|---|---|
| Primary balance | -$18.99bn | -$12.31bn | -$13.84bn | -$1.87bn | +$3.24bn | +$6.40bn | -$2.57bn |
| NIRC | $18.24bn | $20.37bn | $22.38bn | $22.97bn | $24.07bn | $27.53bn | $28.48bn |
| Overall fiscal position | -$51.57bn | +$1.88bn | +$1.72bn | -$2.55bn | +$6.13bn | +$15.10bn | +$8.54bn |
Read FY2026 across. An underlying primary deficit of $2.57 billion becomes a headline surplus of $8.54 billion. Special transfers and top-ups to funds pull the balance down, and NIRC, at $28.48 billion, is the large item that pulls it back up. NIRC is bigger than the estimated collections from GST ($22.26bn) or Personal Income Tax ($21.80bn). Among the individual taxes, only Corporate Income Tax ($37.77bn) is larger.
Two evaluative uses follow, and they point in opposite directions, which is what makes this good material.
The strengthening use. Because Singapore funds a large share of spending from returns on its reserves instead of borrowing, the debt and crowding-out constraints that bind most governments bind much less tightly here. That strengthens the case for fiscal policy, especially long-horizon supply-side spending, compared with a similar economy without reserves.
The qualifying use. The same structure creates a different dependency. The headline fiscal position is now partly a matter of investment returns, not only of tax and spending decisions. A long run of weaker returns would weaken the reported position with no change in policy at all, and this sits alongside spending pressures from an ageing population.
Run both directions from the same evidence, then commit to which dominates in the question's context.
Chain 5: Distribution, or who actually gains? (syllabus-required)
The chain: measure, then incidence across income groups, then income distribution, then inclusive growth.
The syllabus requires "an understanding of how transfer payments can improve income distribution and help achieve inclusive growth", and it lists income inequality and the Gini coefficient under Concepts and Tools of Analysis. This is examinable content, not background.
The questioning runs two ways.
Growth is not automatically fair. An expansionary package that raises real output may leave income distribution unchanged or make it worse, depending on who receives the spending and who owns the factors whose returns rise. "Growth improves living standards" is incomplete until you say whose.
Indirect taxes and their offsets. GST is regressive in incidence: lower-income households spend a larger share of their income, so a single rate takes a bigger share of a poorer household's income. But regressivity describes the tax on its own, not the fiscal system as a whole. Singapore's two-step GST rise from 7% to 9% (to 8% in January 2023, then 9% in January 2024) came with an Assurance Package that was enlarged at successive Budgets and with the permanent GST Voucher scheme. If the offsetting transfers are means-tested and large enough, the combined package can be progressive even though the tax alone is not.
The evaluation sits in the qualification: the GST Voucher is permanent, but the Assurance Package is time-limited, so the longer-run outcome depends on decisions not yet made.
One outcome measure is worth knowing: the Budget 2025 statement reported income inequality after government taxes and transfers at its lowest since 2000. That measures the combined effect of the whole system, which is stronger than listing schemes. Its limitation: it measures income flows, not wealth.
Chain 6: Comparison, or is fiscal policy the right tool here?
The chain: objective, then the available policies, then their relative suitability, then a judgement.
The syllabus lists fiscal, monetary and supply-side policies together at 3.2.3(b), which signals that comparison is expected. A question asking whether fiscal policy is the most effective way to reach some objective is asking you to rank tools, not describe one.
The Singapore comparison is the strongest version of this chain. MAS runs monetary policy through the exchange rate instead of the interest rate, which the syllabus names as the case of Singapore. A small, very open economy that manages its exchange rate and allows free movement of capital cannot also set its own domestic interest rate.
The consequence: Singapore does not have conventional interest rate policy for managing domestic demand, so fiscal policy carries more of that load than in most economies. That is a much better answer to "why does Singapore use fiscal policy so actively?" than the size of the reserves.
Then question it the other way. Because the multiplier is small (Chain 1), fiscal policy is both more relied on and less powerful per dollar in Singapore than elsewhere. Holding those two facts together and reasoning from the tension between them is the synthesis AO4 asks for. The same comparison logic applies to trade policy, which I work through in my guide to evaluating protectionism and tariffs.
Part 3: Building the judgement
AO4 asks candidates to "synthesise economic arguments to arrive at well-reasoned judgements and decisions". A conclusion that says "fiscal policy has advantages and disadvantages, so it depends" has synthesised nothing and decided nothing.
A judgement needs three things.
1. A criterion, stated. Closing an output gap, sustainability over a decade, or distribution? Different criteria give different verdicts from the same evidence, so name yours. "Judged by its effect on long-run productive capacity rather than short-run demand..." is a strong opening because it tells the marker what standard you are using.
2. A ranking, with a reason. Which of your evaluative points dominates, and why in this context? Not every limitation matters equally. Time lags matter a lot for short-term stabilisation and hardly at all for long-term infrastructure investment. Say which one this question is about.
3. A condition under which you would change your mind. It costs one sentence: "This judgement would reverse if the economy were operating near full capacity, in which case..."
Worked contrast
Weak conclusion (generic, no criterion, no ranking):
In conclusion, fiscal policy can be effective in promoting economic growth but it also has limitations such as time lags and the risk of a budget deficit. Therefore the government should use it carefully alongside other policies.
Strong conclusion (criterion stated, ranked, conditional):
Judged by its capacity to raise potential growth rather than actual growth, fiscal policy in Singapore is more defensible as a supply-side tool than as a demand-management one. The multiplier is the binding limitation: with a high marginal propensity to import, each dollar of demand stimulus leaks abroad quickly, so the domestic output gain per dollar is modest. Spending on infrastructure, education and R&D avoids that constraint, because it works through long-run aggregate supply rather than through the multiplier, and because Singapore's reserve-backed fiscal position means the sustainability concerns that would normally limit such spending bind less tightly here. This judgement would reverse in a sharp demand-side shock. In 2020, with the economy contracting and a large negative output gap, the case for immediate demand support was overwhelming despite the weak multiplier, because the alternative was lasting damage to productive capacity through business failures and long-term unemployment.
The second version does not contain more facts than the first. It contains a criterion, a ranking, a reason and a stated condition for reversal.
Part 4: The evidence bank
Evaluation needs evidence, and generic evidence produces generic evaluation. These are Singapore Budget measures from 2020 to 2026 that attach to each chain, so you can draw on specifics instead of gesturing at "the Budget". For a fuller set of this year's measures, see my post on Singapore Budget 2026 examples for economics essays.
| Chain | Evidence | Key figures |
|---|---|---|
| Magnitude | Transfer payments vs government investment | SG60 Vouchers ($600 or $800, 2025) as transfers; a new national semiconductor R&D fabrication facility (Budget 2025) as investment |
| Context | 2020 vs 2025 as opposite AS positions | 2020: a 5.4% contraction (MTI's estimate at the time) and a large negative output gap. 2025: growth of about 5% against an initial forecast of 1% to 3% |
| Time lags | 2020's four Budgets as a counter-example | Unity 18 Feb, Resilience 26 Mar, Solidarity 6 Apr, Fortitude 26 May 2020; $92.9bn combined, about 19.2% of GDP |
| Sustainability | NIRC and the primary vs headline gap | FY2026: primary deficit $2.57bn, NIRC $28.48bn, overall surplus $8.54bn |
| Distribution | GST rise and its offsets | 7% to 8% (Jan 2023), 8% to 9% (Jan 2024); Assurance Package enlarged to over $10bn; income inequality after taxes and transfers lowest since 2000 (Budget 2025) |
| Comparison | MAS exchange rate policy vs fiscal policy | The syllabus names exchange rate management as the case of Singapore |
| Supply side | The structural thread | SkillsFuture mid-career $4,000 top-up (2024); Progressive Wage Credit Scheme co-funding raised to 30% (2026); Enterprise Innovation Scheme extended to AI spending at 400% deduction (2026) |
A caution on quoting fiscal figures. Singapore reports each year's budget balance more than once: an estimate at the February Budget, a revised figure a year later, and then the actual outturn. The figures above are MOF's current series and are not always the numbers reported in the press at the time. FY2020 is the clearest case. The Fortitude Budget projected an overall deficit of $74.3 billion, but MOF's current figure for FY2020 is $51.57 billion. Quote the current series, and say which set of accounts it comes from.
Part 5: Applying the method in each paper
Paper 2 (essays): 75 marks, 60%, 2 hours 30 minutes
You answer three of six questions: at least one from Section A (microeconomics), at least one from Section B (macroeconomics) and the third from either. Each is worth 25 marks, which gives you about 50 minutes per essay including planning.
Fiscal policy questions usually sit in Section B, typically in two parts with the second asking for evaluation. About 40% of Paper 2 marks come from questions testing AO1 and AO3 only, and about 60% from questions that also test evaluation. An essay that spends 80% of its length on description and adds two evaluative sentences is badly balanced before anyone reads it. Build the chain efficiently, then spend real length questioning it.
Paper 1 (case studies): 60 marks, 40%, 2 hours 30 minutes
There are two compulsory case studies of 30 marks each. In each, about 12 marks go to data response questions and about 18 to higher-order questions. Each case study gives you a few pages of data in text, tables or charts, followed by several part-questions.
The distinctive requirement here is AO2, interpreting and evaluating information. It asks you to interpret economic information in textual, numerical or graphical form and to make valid inferences from the information and its limitations.
That last part is often ignored and easy to satisfy. If an extract gives a figure for government spending, ask: is it nominal or real, a forecast or an outturn, and does it cover the period the question asks about? One sentence noting that a fiscal figure is a Budget-day estimate is exactly the inference with limitations that AO2 rewards.
For the higher-order questions, run the same six chains, but take your evidence from the extracts and use your own knowledge to explain and evaluate it.
I teach H2 Economics at Higher Nucleus Learning Studio. Class details are on the Higher Nucleus H2 Economics page.
Sources
- Singapore Examinations and Assessment Board, 9570 Economics GCE A Level H2 syllabus (2026): Scheme of Assessment and Theme 3.2.3
- Ministry of Finance, Analysis of Revenue and Expenditure, Financial Year 2026: Table 3.1a and revenue tables
- Ministry of Finance, Budget Statements and Ministerial Statements, FY2020 to FY2026
- Ministry of Trade and Industry, GDP releases and growth forecasts, 2020 to 2026
Questions people ask
How do I evaluate fiscal policy in H2 Economics?
Build the analytical chain first, then question it at its weakest links. The six that do the work are the size of the multiplier, the state of the economy and the relevant AS range, time lags, fiscal sustainability, distribution, and comparison with monetary and supply-side policy. Develop two or three properly instead of listing all six, and finish with a judgement that states its criterion.
Are automatic stabilisers in the H2 Economics 9570 syllabus?
No. The syllabus specifies discretionary fiscal policy only, and the words "automatic" and "stabiliser" do not appear in it. Mentioning automatic stabilisers will not lose you marks, but building answers around them spends time where no marks sit.
Why is fiscal policy less effective in Singapore than in other countries?
Singapore is a small and very open economy with a high marginal propensity to import, so a large share of any demand stimulus leaks abroad instead of circulating at home. That gives a small multiplier, so each dollar of demand-side stimulus produces less extra domestic output than it would in a larger, less open economy. It is a major reason Singapore's fiscal policy leans towards supply-side and redistributive measures.
Why does Singapore rely on fiscal policy if its multiplier is small?
MAS runs monetary policy through the exchange rate, not the interest rate, so Singapore does not have conventional interest rate policy for managing domestic demand. Fiscal policy therefore carries more of that load than it does elsewhere, even though each dollar does less.
What is NIRC and why does it matter in a fiscal policy essay?
The Net Investment Returns Contribution is the share of returns on Singapore's reserves that the Government may spend, capped at up to 50% of expected long-term real returns. It means a headline surplus can sit alongside an underlying primary deficit: in FY2026, a $2.57 billion primary deficit becomes an $8.54 billion overall surplus. It supports arguments in both directions on fiscal sustainability, which the syllabus requires you to be aware of.
How do I make my evaluation contextual instead of generic?
Apply one test to every evaluative sentence: would it survive unchanged in an essay about a different country or a different policy? "Fiscal policy has time lags" survives, so it is generic. "A stimulus announced at the February Budget and paid out over the following year risks landing after an economy already recovering from a mild slowdown" does not, so it is applied.
This article is also in Chinese: 阅读中文版.