Chapter 17 — Read economic data and budget claims like an examiner

Exam fit: common foundation. Level: basic to intermediate.

Economic claims often sound precise because they contain a large number. A reliable reader asks what was measured, for whom, over what period, in which units, against which comparison, and at what revision stage. The skill is useful in examinations and in daily life: it prevents a correct number from being used to support the wrong conclusion.

17.1 The six-part label

A claim audit checks period, unit, denominator, base and source. | Original explanatory diagram; schematic, not a measured data chart.

Before interpreting a statistic, write six labels beside it: indicator, unit, population or coverage, period, price or statistical base, and status/source. “Inflation is 5 per cent” is incomplete without the index, month or year, comparison period and release. “The economy grew by 8 per cent” needs a measure (GDP or GVA), a real or nominal basis, a financial year or quarter, and an advance/provisional/revised status. A budget number also needs the department or account, year, and whether it is BE, RE or actual.

The six-part label is a checklist, not a demand to memorise every note. If the question stem omits one dimension, decide whether it matters to the comparison. A one-year nominal revenue comparison may be appropriate for cash receipts; a real output-growth comparison requires a constant-price measure. Two CPI rates can be compared only if the index, period-over-period convention and relevant series are compatible.

17.2 Stock, flow and rate

A stock is measured at a point in time: public debt on 31 March, bank deposits on a date, or a population count. A flow is measured over an interval: annual tax revenue, expenditure, exports or income. A debt stock divided by annual GDP is a ratio of a point-in-time liability to an annual flow; it is useful, but the numerator and denominator have different time dimensions. A rate expresses change or proportion relative to a defined base. Always inspect the denominator. A five per cent rise in a small base can be smaller in rupees than a two per cent rise in a large base.

Percentage change is `(new value − old value) ÷ old value × 100`. If a value moves from 200 to 230, the increase is 30, and the growth rate is 15 per cent. If it then falls from 230 to 200, the fall is about 13.04 per cent, not 15 per cent: the denominator changed. This asymmetry is a frequent source of misleading “reversal” claims.

17.3 Nominal, real and per-capita comparisons

Current-price (nominal) values combine changes in quantity and prices. Constant-price (real) values attempt to isolate quantity changes using a specified reference price structure. A deflator is used to convert between them in the relevant accounting framework; do not divide by an unrelated index unless the question explicitly provides that simplifying assumption. Per-capita values divide an aggregate by population. They are averages and can move because the aggregate changes, the population estimate changes, or both. None of these calculations shows how income is distributed among people.

For a clean comparison, match indicator to question. Nominal tax receipts answer a cash-revenue question. Real GDP growth is used for volume growth. Per-capita real income may help compare average output or income over time, but it does not establish median household welfare. A high average can coexist with low wages, insecure work or unequal access to public services.

17.4 Read a budget table before drawing a conclusion

First locate the table title, department or fund, financial year columns and footnotes. Identify whether the figure is revenue or capital, receipt or expenditure, and whether borrowing or public-account items appear. Then identify the accounting stage. Actuals are recorded outcomes for a completed period (subject to audit and later accounting revisions); RE is a revised estimate for the current year; BE is a proposed estimate for the coming year. A BE is a plan and forecast, not money already spent.

Budget execution can differ from BE for many reasons: revenue may be higher or lower, a project may be delayed, a transfer may be released late, procurement may take time, or a disaster may change priorities. To assess execution, compare compatible actuals and estimates and ask which line items account for the gap. Low capital spending may indicate implementation constraints, but the headline alone cannot tell whether projects were delayed, reclassified, or funded through another channel. Follow the detailed statement and outcome information.

Worked case: comparing a budget estimate with a result

Suppose a department has a BE of ₹1,000 crore for a year and actual expenditure of ₹840 crore. The shortfall is ₹160 crore. Relative to the BE it is 16 per cent. This arithmetic does not show that the department wasted 16 per cent, nor that every service fell by 16 per cent. You would need the expenditure classification, revised estimate, releases, commitments, outcomes and reasons for variance. The safe inference is only: actual recorded spending was ₹160 crore below the original estimate in the table being compared.

17.5 Base effects, revisions and breaks in series

Growth rates depend on the base. If a prior-year value is unusually low, a subsequent normal value may produce a large percentage increase, called a base effect. This does not make the new growth number false; it changes what the rate means. Read the level and the rate together, and compare with a suitable pre-shock period when the question permits.

Statistical series are periodically rebased or improved. A change of base, weights, survey design, definition or data source can affect comparability. India’s national accounts and CPI updates therefore need clear series labels; the chapter on inflation explains the CPI transition and this book’s source register notes the national-accounts series change. Do not splice old and new series into one trend unless the statistical authority provides a compatible back series or a transparent linking method. Revisions are normal statistical work, not proof that one release is automatically unreliable.

17.6 Correlation, causation and attribution

When two variables move together, that is correlation. It does not by itself show that one caused the other. If public investment and output both rise, possible explanations include a direct investment effect, improved private confidence, favourable rainfall, export demand, population changes or common policy factors. Causal evaluation requires a credible comparison, timing and control for alternative explanations.

Use careful verbs. “Rose alongside” describes a pattern. “Contributed to” may be justified by supporting evidence. “Caused” is a stronger claim and requires evidence that competing explanations were addressed. In an exam, a basic conceptual question may ask whether a policy can affect demand; answer the theory. A question about a specific programme’s measured impact needs the named evaluation and date.

17.7 Evidence ladder for claims

For a dynamic fact, prefer the issuing statistical authority or department and the original table. Next use a clearly dated official summary that points to the source table. A speech or press note may explain policy intent but should not replace the underlying dataset when definitions matter. Secondary analysis can be useful for interpretation, but check its period, method and source. Record the release date as well as the reference period: a 2025–26 survey released in 2026 is not a 2026–27 measurement.

Claim-check worksheet

1. Copy the exact claim, without strengthening its wording.

2. Identify its indicator and units.

3. Find the official table and note its reference period and release date.

4. Check whether the figure is nominal or real, stock or flow, and its estimate/revision status.

5. Recalculate any percentage, ratio or per-capita figure from the stated inputs.

6. Check whether the comparison uses compatible definitions and series.

7. Separate observed fact, explanation and forecast.

8. State the conclusion at the strength the evidence supports.

Worked application and examiner distinction

A headline says “revenue rises 20%.” Obtain the two values. If the old value was ₹500 crore and the new value ₹600 crore, the arithmetic checks: an increase of ₹100 crore is 20% of ₹500 crore. Next ask whether the values are actuals, BE or RE, whether the same account line is compared, and whether a one-off receipt is included. If one is an estimate and the other an actual, the comparison can still be informative, but its meaning must be stated accurately.

Another headline says “prices fall 10%.” Determine whether it means one commodity, the CPI index, a month-on-month rate, year-on-year inflation or the index level. A fall in inflation from 6% to 4% means prices are still rising more slowly, not that the price level fell. A 10% fall in one item's price does not imply a 10% fall in the cost of every household basket.

When a chart has a truncated vertical axis, inspect the scale before judging visual magnitude. When a per-capita figure is presented, identify the population estimate and price basis. When a state ranking is stated, ensure all states use the same period and data vintage. Write down one uncertainty that could change the conclusion. This disciplined check is faster than memorising an incorrect interpretation.

Second worked calculation and interpretation

A group of 50,000 is 2% of a population of 25 lakh. If a report uses a population estimate of 24 lakh, the share is about 2.08%. The numerator is unchanged, but the denominator changes the rate. This matters in employment, poverty, tax collection, per-capita output and programme coverage. Never copy a percentage without its population definition and base year.

For a growth rate, recompute with the old value as denominator. For a share, divide component by total. For a ratio of debt to GDP, record whether both are measured under compatible accounting definitions. Check rounding and units. If the published table gives rounded figures, your recomputation may differ slightly; report the difference rather than silently forcing equality.

Applied examiner workshop

Suppose a chart shows revenue rising from 100 to 110 but its vertical axis starts at 95. The visual bars may make the rise look dramatic, although the numerical change is 10%. Inspect axis origin and scale, read values directly and compute the percentage. A truncated axis is not automatically fraudulent, but its design can exaggerate visual difference if readers do not notice it.

Two reports give different GDP figures for the same year. Compare release date, series, current or constant prices, estimate stage and revisions. One may be advance and one revised; one may use an older series. The correct response is not to choose the larger figure but to identify the publication and measure. Preserve the original vintage if evaluating what was known at a particular time.

When a source is a press release, locate the linked table and methodology. When a source is a budget speech, check the budget document. A secondary article may summarise correctly but compress status labels. Record reference period separately from publication date; a release published in 2026 may measure an earlier fiscal or survey year.

Chapter recall

• A statistic needs an indicator, unit, coverage, period, base and source/status.
• Stocks are measured at a date; flows cover a period; rates need a denominator.
• A percentage increase and an equal-sized percentage decrease do not cancel because their bases differ.
• BE is a plan, RE is a revised estimate, and actuals are recorded outcomes.
• Series changes and revisions require labels; avoid unverified splicing.
• Correlation is not sufficient evidence of causation.
• A claim should not be stronger than its underlying table and method.

Quick retrieval

1. What six labels would you request before comparing two growth figures?

2. Why does a 15% rise followed by a 15% fall not return a value to its start?

3. What can and cannot be inferred from actual expenditure below BE?

4. What is a base effect, and how should levels help interpret it?

5. Which official record should you inspect before repeating a current statistical claim?

Chapter 17 practice MCQs

1. Which information is needed to interpret an inflation claim responsibly?

• A. only a rupee symbol
• B. only the speaker's name
• C. the population of another country
• D. index, reference period and comparison convention

Answer: D. The measured index and period establish what the rate compares.

2. A stock is usually measured:

• A. at a point in time
• B. only over a decade
• C. as a percentage change
• D. only in constant prices

Answer: A. Debt on a date is a stock; tax collection over a year is a flow.

3. Annual tax receipts are generally a:

• A. population denominator
• B. stock at a single instant
• C. flow
• D. price index base

Answer: C. Receipts accrue over a specified financial period.

4. A value rises from 200 to 230. Its percentage increase is:

• A. 30%
• B. 15%
• C. 13.04%
• D. 50%

Answer: B. The rise is 30; 30 divided by the starting value 200 equals 15%.

5. A rise from 200 to 230 followed by a fall from 230 to 200 means the fall is approximately:

• A. 30%
• B. 13.04%
• C. 15%
• D. 43.04%

Answer: B. The fall is 30 divided by 230, or about 13.04%; the denominator changed.

6. A ratio of debt measured on 31 March to annual GDP combines:

• A. two daily prices
• B. a stock numerator and a flow denominator
• C. two population stocks
• D. only revenue flows

Answer: B. Debt is a point-in-time stock, while annual GDP is measured over a year.

7. A current-price output increase may reflect:

• A. tax rates alone
• B. both price and quantity changes
• C. population alone
• D. quantity alone in every case

Answer: B. Nominal values combine current prices with quantities.

8. Per-capita output is obtained by dividing an aggregate by:

• A. tax revenue
• B. population
• C. imports
• D. the inflation rate

Answer: B. “Per capita” means per person, using the relevant population estimate.

9. Which is the correct statement about per-capita output?

• A. It equals the poverty rate.
• B. It is a guaranteed minimum wage.
• C. It directly measures income equality.
• D. It is an average and does not describe each person's income.

Answer: D. An aggregate average hides distribution.

10. A BE in a budget is best understood as:

• A. audited actual expenditure
• B. a household survey
• C. an estimate for the coming financial year
• D. final cash spent already

Answer: C. Budget estimates are plans/forecasts, not recorded outcomes.

11. RE means:

• A. revenue entitlement
• B. revised estimate
• C. real expenditure only
• D. recorded employment

Answer: B. RE revises the estimate for the current year.

12. Actual expenditure is below BE. The safest immediate conclusion is:

• A. recorded spending was below the original estimate in that comparison
• B. all programme outcomes fell by the same percentage
• C. fraud necessarily occurred
• D. the economy has no debt

Answer: A. A variance is an observation; its causes and service effects need more evidence.

13. A low base can make the following period's growth rate:

• A. appear high even when the level has only partly recovered
• B. always zero
• C. impossible to calculate
• D. equal to inflation by definition

Answer: A. A small denominator can magnify the percentage change.

14. A statistical series changes base year. A careful user should:

• A. concatenate values without labels
• B. check definitions and whether a compatible back series is available
• C. assume every past figure was false
• D. ignore the change

Answer: B. Methodological changes may alter comparability.

15. Statistical revisions are:

• A. only used for budgets
• B. proof that every release is invalid
• C. the same as a tax increase
• D. a normal way to incorporate improved data or methods

Answer: D. Estimates can be revised as better inputs become available.

16. Two variables move upward together. This establishes:

• A. that the first caused the second
• B. that no other factor matters
• C. correlation, but not necessarily causation
• D. that the data are nominal

Answer: C. Association alone does not identify causal direction or rule out alternatives.

17. “Output rose alongside public investment” is an example of:

• A. proof of causal impact
• B. a population estimate
• C. a budget receipt
• D. a descriptive association

Answer: D. The wording reports co-movement without asserting that one caused the other.

18. For an official statistic, the strongest first check is usually:

• A. a quote without a period
• B. the source agency's dated table and its notes
• C. an unattributed social-media image
• D. a chart with no units

Answer: B. The issuing agency's table provides definitions, period and method.

19. A survey for 2025–26 released during 2026 measures:

• A. the reference period stated by that survey, not automatically 2026–27
• B. all events up to the release date
• C. only 2026–27 conditions
• D. the next financial year by definition

Answer: A. Release date and reference period are distinct.

20. A claim about a programme's causal impact needs:

• A. a named evaluation with a credible method and period
• B. only a before-and-after headline
• C. a policy announcement alone
• D. a budget estimate alone

Answer: A. Causal attribution requires evidence beyond a coincident change.

21. If the comparison's denominator is not stated, a percentage claim is:

• A. always false
• B. incomplete and potentially misleading
• C. always a stock
• D. automatically a real rate

Answer: B. The denominator determines the meaning and size of a rate.

22. The best final step in evaluating a claim is to:

• A. strengthen the claim to make it persuasive
• B. state only what the evidence supports
• C. omit the source period
• D. substitute a different indicator

Answer: B. Conclusions must match the data and method.