The Gold Standard of Economists explains how people turn scarce resources, prices and policy choices into decisions supported by sound evidence. It follows a fictional Singapore transport-voucher pilot and demonstrates why a before-and-after improvement is not enough to prove a policy caused it. The professional standard is a transparent question, justified comparison and clear account of uncertainty.
The eduKate Gold Standard is about dependable management, not a label for superior intellect. This jobs article complements Data Analysts, Actuaries and Teaching Economics without repeating their tasks.
A quick answer: what is an economist’s job?
Economists study how households, organisations and societies allocate limited resources, how incentives change behaviour and what evidence supports a decision. They design comparisons, examine economic statistics, model alternatives, assess trade-offs and communicate limitations so policy and business choices can be better informed.
Worked example: the apparent S$3 policy effect
| Fictional group | Before (weekly spending) | After | Change |
|---|---|---|---|
| Voucher group | S$24 | S$20 | −S$4 |
| Comparison group | S$22 | S$21 | −S$1 |
| Difference in changes | −S$4 − (−S$1) = −S$3 |
All figures are invented. The minus S$3 difference is a descriptive difference-in-differences calculation, not proof of a causal intervention effect. Such an interpretation needs justified comparison, design assumptions, suitable uncertainty analysis and careful treatment of other changes.
Navigate the economic decisions
1. Economists study choices under scarcity, not just money
A fictional Singapore community programme considers helping residents afford essential journeys. The programme manager is excited because transport spending appears to fall after vouchers were introduced. An economist asks whether the same trend also occurred among otherwise comparable people who did not receive vouchers, and whether spending alone describes access to work or healthcare. The professional role is to understand choices, incentives and evidence before treating one changed number as success.
Economics covers households, firms, markets and public institutions. It considers opportunity cost, behaviour, prices and how resources affect wellbeing. Economic analysis is most useful when readers can see whose choices are being studied, which outcome matters and which limits remain. A graph can be statistically neat while telling an incomplete story about real people.
Back to chapters · Next chapter
2. Define the decision before opening the dataset
A request to ‘show whether vouchers worked’ needs a definition. Does worked mean that riders paid less, travelled more, were more likely to attend appointments or felt less financial strain? A single policy can influence several measures differently. The economist should distinguish the decision-maker’s question from whatever variable happens to be easiest to download.
For the fictional voucher pilot, suppose the stated goal is to reduce the financial burden of necessary transport. The economist can specify observed weekly out-of-pocket spending, participant eligibility and survey timing. They should also flag that payment levels alone do not capture missed trips or service accessibility. A credible brief explains its outcome before the analysis produces a tempting result.
Back to chapters · Next chapter
3. Distinguish microeconomics from macroeconomics
Microeconomics can examine household travel choices, incentives or business decisions. Macroeconomics considers broader aggregates such as growth, inflation, unemployment, productivity and external conditions. Economists may specialise in labour, environmental, trade, health, education or other domains and use combinations of theory and empirical analysis.
A voucher evaluation is a microeconomic or programme question, even though it may operate within wider economic conditions. The economist should not invoke a national GDP figure as direct proof that one local programme succeeded. Likewise, economic growth does not by itself tell us how a particular family experienced a transport price. The level of analysis must match the claim.
Back to chapters · Next chapter
4. Begin with the population and eligibility rules
A programme may serve households meeting specific eligibility criteria. If participants differ systematically from non-participants before the programme, their outcomes may also differ for reasons unrelated to the voucher. The analyst needs to know who was eligible, who accepted, who was surveyed and who left the study.
A fictional dataset of 200 users may omit people who could not reach the registration office. The economist should consider that selection before claiming a benefit for all eligible residents. Good analysis names the actual population rather than silently converting a convenient sample into a citywide statement. Administrative records and surveys each have their own coverage limits.
Back to chapters · Next chapter
5. Explain why before-and-after comparisons can mislead
Imagine voucher recipients spent S$24 per week before the pilot and S$20 after it. The S$4 decline is an observed difference, but it does not prove the voucher caused the entire change. Fuel prices, transit patterns, employment, school holidays or other changes could influence spending over the same period.
An economist asks what would probably have happened without the intervention. That counterfactual is not directly observed for the same person at the same moment, which is why study design matters. The profession’s value is identifying which comparison could make a causal argument more credible without pretending that any table can fully reproduce an alternative history.
Back to chapters · Next chapter
6. Calculate a simple difference-in-differences example
In the invented evaluation, recipients’ weekly transport spending falls from S$24 to S$20, a change of minus S$4. A comparison group falls from S$22 to S$21, a change of minus S$1. The difference between changes is minus S$3. Under appropriate assumptions, that may be interpreted as an estimated programme-associated change relative to the comparison trend.
The arithmetic is (20 − 24) − (21 − 22) = −3. But this is a teaching illustration, not real Singapore data or proof of causation. A credible difference-in-differences analysis depends on suitable comparison groups, assumptions about counterfactual trends, absence of important contamination and appropriate uncertainty analysis. The economist should say those requirements plainly.
Back to chapters · Next chapter
7. Test the parallel-trends assumption rather than treating it as magic
Difference-in-differences requires careful thinking about whether the groups would have followed comparable trends without the programme. Similar averages at one time do not establish similar trajectories. Prior observations, institutional knowledge, programme rules and sensitivity analyses may help assess the plausibility of the counterfactual assumption.
If the voucher group was already experiencing rapidly declining travel costs before the pilot, the simple minus S$3 result could be misleading. An economist should look for what the pre-policy data can reveal and explain what cannot be verified. An attractive causal label does not erase design uncertainty.
Back to chapters · Next chapter
8. Remember that transport spending and travel access differ
A person’s transport expenditure could fall because fares became cheaper, because journeys shifted or because they travelled less and missed useful opportunities. A voucher could also increase the number of essential trips while keeping weekly spending unchanged. The economist should match evaluation measures to the intended wellbeing outcome.
For our fictional programme, a richer analysis might consider necessary trips, access to employment or clinics and participants’ reported burden, subject to appropriate privacy and data quality. These are distinct results, not extra figures added merely to make a report longer. Policy evaluation becomes useful when it asks whether the intervention solved the real problem.
Back to chapters · Next chapter
9. Distinguish nominal values from real purchasing power
Money values change meaning when prices change. Comparing household expenditure or earnings across years may require careful treatment of inflation and the appropriate price index. Singapore’s SingStat price statistics provide official measures and metadata relevant to such comparisons.
The analyst should not automatically adjust every cash figure by one broad index and call it a perfect measure of a family’s burden. Different goods, periods and household circumstances matter. A sound report states whether numbers are nominal or adjusted, which index was used and why it suits the question.
Back to chapters · Next chapter
10. Know the source and revision status of official statistics
Official figures have release dates, definitions, preliminary and final states and possible revisions. Singapore’s Data Dissemination Standard describes coverage, timeliness, public access, integrity and data quality principles. An economist should use the correct dataset vintage and link to its metadata.
A copied chart without its date or underlying unit can be misleading even if the numerical values were originally correct. In a changing economy, a revised labour or national accounts series may affect a trend. Professional practice preserves the source and explains material changes rather than pretend an old screenshot is an evergreen fact.
Back to chapters · Next chapter
11. Do not confuse correlation with a policy effect
People who choose an intervention may differ from those who decline. A correlation between voucher use and better outcomes might reflect those differences, a third variable or a genuine programme effect. The analyst should explain the research design and remaining alternative explanations.
A visually compelling scatterplot does not settle whether one factor caused another. Statistical sophistication is useful only when coupled with economic reasoning and suitable evidence. If the available data cannot support a causal claim, the result can still describe who participated, what changed and where further evaluation is needed.
Back to chapters · Next chapter
12. Evaluate benefits alongside opportunity costs
Public money and staff time used for one programme cannot automatically be used elsewhere. An economist helps decision-makers consider the alternative uses of resources and who would gain or lose from different options. That does not mean valuing every social outcome only in dollars or assuming the cheapest intervention is morally best.
Suppose the transport pilot requires more administrative work than anticipated. The economist can report direct costs, implementation effort and potential benefits, while making clear which impacts are measured and which are uncertain or not monetised. A responsible cost-benefit discussion includes assumptions and distribution rather than one simplistic return-on-investment figure.
Back to chapters · Next chapter
13. Treat distribution as distinct from the overall average
Two programmes may have the same average effect but affect households differently. A benefit concentrated among people already able to travel could be less relevant to the original access problem than a smaller average gain for highly constrained users. Disaggregated results need to be interpreted with statistical and privacy safeguards.
An economist should examine meaningful subgroup patterns where the study design and sample permit them. Small cells may be unreliable or expose sensitive information. Fair policy analysis asks who receives value, who bears inconvenience and whether the evidence actually supports the apparent difference, without overclaiming from noisy subgroup charts.
Back to chapters · Next chapter
14. Understand incentive effects and unintended responses
People respond to prices, rules and opportunities, but not always in the simple direction assumed by a policymaker. A subsidy might change timing, provider choice or application behaviour. Those changes can be beneficial, neutral or problematic depending on programme purpose and constraints.
For the fictional voucher scheme, a household might move journeys to more convenient times or increase essential trips. A programme manager who expected only lower expenditure could misinterpret this behaviour. Economists help institutions ask how rules interact with people rather than regard unexpected behaviour as noncompliance by default.
Back to chapters · Next chapter
15. Choose models appropriate to the question
Economic models simplify reality to clarify relationships. Demand curves, regression models, structural systems and simulations each have suitable purposes and limitations. A more complex model does not automatically produce more credible insight. The result depends on assumptions, identification, data and validation.
A decision-maker may ask for one precise impact number from a dataset too small or biased to support it. The economist should explain the limitation, test defensible alternatives and avoid imposing a sophisticated method merely to make the answer look authoritative. Professional judgement includes knowing when a transparent descriptive analysis is more honest.
Back to chapters · Next chapter
16. Assess uncertainty and sensitivity explicitly
Estimates vary with sampling, assumptions, measurement and model choice. An economist should communicate meaningful uncertainty in a way that readers can act on, using appropriate statistical methods when warranted. A result may be compatible with several plausible effects, and that range can materially change the policy recommendation.
If the apparent minus S$3 effect weakens when an appropriate comparison group is changed, that sensitivity belongs in the report. Hiding it to protect a successful story undermines confidence. Good economic advice makes robustness visible and identifies what additional evidence would reduce uncertainty.
Back to chapters · Next chapter
17. Separate forecasting from causal evaluation
Forecasting asks what might happen under specified future conditions; causal evaluation asks what an intervention changed relative to an alternative. Both are useful, but their claims are different. A model that predicts next quarter’s aggregate demand accurately may tell us little about the effect of one voucher programme.
An economist should label scenarios as scenarios and the actual observational comparison as an observational comparison. It is misleading to report a forecast as though it were a tested result or convert one local pilot into a certain national projection. Decision-makers need to understand the strength and purpose of each output.
Back to chapters · Next chapter
18. Keep policy data confidential and proportionate
Household and administrative datasets may contain sensitive personal information. Analysts should comply with applicable law, institutional approvals and secure data arrangements, avoiding unnecessary identifiers in reports or presentations. Public benefit does not remove the responsibility to protect participant privacy.
For a small neighbourhood pilot, publishing exact participant characteristics could make people recognisable even without names. The economist should use appropriate disclosure safeguards and collaborate with data stewards. Access permissions and limits on linking datasets should be settled before data are combined.
Back to chapters · Next chapter
19. Make code and assumptions inspectable
Spreadsheets, statistical scripts and dashboards can contain mistakes or unexplained transformations. An economist should preserve the analysis specification, input sources, software versions and material revisions under suitable review arrangements. A convenient graph exported from a notebook is not a complete audit trail.
Singapore’s Statistical Best Practices handbook emphasises documenting concepts, processes and methods so data remain interpretable. The professional result should be reproducible by an authorised reviewer rather than dependent on one analyst’s private spreadsheet.
Back to chapters · Next chapter
20. Present trade-offs in language decision-makers can use
A minister, business leader or community organisation may need to decide what to do next rather than learn an entire statistical method. Lead with the question, current evidence, key assumptions, who is affected and what remains unresolved. Technical appendices can hold detailed estimation work without contradicting the accessible summary.
A useful short finding from the fictional voucher study would say that treated spending fell more than comparison spending, but the illustrative difference alone cannot establish causation. It would identify the additional checks needed before expansion. Clear economic communication does not obscure limitations to make an executive summary more cheerful.
Back to chapters · Next chapter
21. Recognise the professional roles of policy economists
Singapore’s Economist Service, as described by the Ministry of Energy, Trade and Industry (METI), is a professional public-sector scheme for rigorous policy-oriented research. The 14 October 2026 career page mentions quantitative analysis, communication and rotation opportunities.
The exact career requirements and organisational structure can change, and economists also work in universities, businesses, financial firms and consultancies. A candidate should evaluate the actual team’s research work, training and ethical expectations. Holding the title economist does not confer exclusive legal authority over public-policy decisions.
Back to chapters · Next chapter
22. Build the profession through economics, mathematics and judgement
Useful foundations include microeconomics, macroeconomics, statistics, econometrics, programming and clear writing. An aspiring economist should learn to explain why a chosen comparison group is plausible and what a dataset cannot reveal, rather than memorise technical terms without understanding their assumptions.
A portfolio could use the fictional transport-voucher example to show the raw changes, difference-in-differences calculation and limitations. Strong work identifies alternative explanations and what additional evidence might be worth collecting. The ability to communicate uncertainty is as important as the ability to fit a model.
Back to chapters · Next chapter
23. Judge economic work by better decisions, not flattering predictions
A government or firm may dislike a result that complicates its preferred policy. An economist should resist adjusting assumptions or selecting a convenient measure to produce an attractive conclusion. The profession is valuable when results remain defensible under independent scrutiny and useful even if they recommend caution.
Return to the fictional voucher pilot. The gold-standard economist does not simply declare a S$3 causal benefit. They explain the actual comparison, test its assumptions and consider who benefited, which costs were incurred and what remains unknown. Economic evidence should improve decisions, not decorate them.
An additional counterfactual test: why the same spending trend can mean different things
Suppose the pilot changed just as an employer introduced flexible hours for many participants. Some people might travel less frequently for reasons unrelated to the voucher. If such a change affected the voucher group differently from its comparison, the minus S$3 estimate could capture more than the programme. The economist should seek relevant information about changes and understand whether the available data allow a meaningful correction.
Now imagine the pilot enabled previously unaffordable clinic trips. Out-of-pocket spending could remain the same despite a genuine improvement in access, because participants used more journeys. This possibility reinforces the need for outcomes beyond a single cost average. A robust evaluation might look at meaningful travel access and wellbeing alongside financial burden, with appropriate privacy controls.
An economist’s contribution is not choosing the most favourable interpretation. It is defining which claims the current data can support, which assumptions remain speculative and what study design would help distinguish them. That reasoning may recommend more evidence, a changed programme or an explicitly conditional next step.
Professional economist quality checks
- Is the decision and target population stated clearly?
- Are outcomes, units, time periods and denominators consistent?
- Are the sources, publication dates and possible revisions of official statistics visible?
- Are treated and comparison populations genuinely suitable for the question?
- Are confounding, pre-trends and plausible alternative explanations examined?
- Are opportunity costs, distribution and unintended effects considered?
- Can another competent analyst reproduce the important calculations?
- Does the executive conclusion match the strength of the evidence?
Frequently asked questions about economists
Do economists only work for banks?
No. They work across government policy, universities, international institutions, consulting, finance, businesses and research organisations.
What is the difference between a data analyst and an economist?
Both may analyse quantitative evidence. Economists use economic theory, policy context and methods tailored to behaviour, incentives, scarcity and causal questions.
Does a price drop prove that a subsidy worked?
Not necessarily. Other conditions and trends may also change. The relevant design and counterfactual assumptions must be examined.
What is difference-in-differences?
A method comparing changes over time between groups. Causal interpretation requires assumptions about how the groups would have evolved without the intervention.
Can GDP growth show whether every household is better off?
No. GDP is an aggregate measure and does not, by itself, describe the distribution of gains or any one household’s experience.
What official Singapore statistics can economists use?
SingStat provides national accounts, prices, employment and other official series, with definitions, release schedules and metadata.
How can someone enter Singapore’s public Economist Service?
The Ministry of Energy, Trade and Industry (METI) provides current recruitment, scholarship and research assistant programme information, updated in October 2026.
What is economic professionalism?
Methodological care, reliable data, clear uncertainty, fair treatment of trade-offs and advice that remains useful even when it complicates a preferred decision.
Official data and relevant eduKate articles
- METI: Singapore Economist Service — public service economist career
- METI: Join the Economist Service — current careers information, 14 October 2026
- SingStat: Data Dissemination Standard — integrity and quality
- SingStat: Statistical Best Practices — documentation and data analysis standards
- SingStat: Official economic and financial indicator metadata
- eduKate: Data Analysts
- eduKate: Actuaries
- eduKate: Teaching Economics
An economist’s gold standard is not one smooth curve or a persuasive prediction. It is a defensible explanation of choices, evidence and trade-offs that helps people make better decisions with the uncertainty still visible.