Why statistics in economics: from data to policy

Detail

1. What economics needs facts about

• Economics has three core areas:
• Consumption: how people spend their income on goods and services.
• Production: how goods and services are made.

• Distribution: how income from production is shared among wages, rent, interest and profit, and among people.

• Economics also needs facts on special problems:

• Poverty: people who cannot meet basic needs.
• Disparity: gaps in income and wealth between people or regions.
• Unemployment: people who want work and are available for it, but find none.

• Cost of disasters: tsunamis, earthquakes, and animal diseases such as bird flu.

• Economic data are economic facts expressed in numbers.

• They are collected to understand a problem and the causes behind it.
• Example: "Unemployment is high" is an opinion. "The unemployment rate is 4.8%" is a data point that can be compared and acted on.

2. The chain from data to policy

The four links come in a fixed order:

Step	What happens	Example: poverty
1. Data	Facts are collected first	How many people are poor, and where they live
2. Economic analysis	The problem is explained by its causes	Poverty is caused by unemployment, low productivity and backward technology
3. Economic policy	A measure is chosen to solve the problem	Employment schemes and technology upgrades
4. Evaluation	Statistics checks whether the policy worked	New data show whether poverty fell

• NCERT's evaluation example (Class 11, Introduction): has family planning checked population growth?

• Only data collected before and after the policy, such as Census population counts, can answer this.

• Key rule: without data there is no analysis, and without analysis there is no policy.

• The chain is a loop, not a straight line. Evaluation produces new data, and the new data start the next round of analysis.
• How often data arrive changes how fast policy can react.
• The redesigned Periodic Labour Force Survey (PLFS) changed its sampling method from January 2025. It now produces labour data every month [8][9].
• The first Monthly Bulletin covered April 2025 [8].
• It gives three indicators every month, for both rural and urban India [8]:
  • LFPR (Labour Force Participation Rate): the share of people who are working or looking for work.
  • WPR (Worker Population Ratio): the share of people who are working.
  • UR (Unemployment Rate): the share of the labour force that has no work.
• UR for persons aged 15 years and above: 4.7% (November 2025) and 4.8% (December 2025) [10][11].
• Earlier, most of these numbers came only once a year. A policy maker had to wait many months to see whether a jobs policy was working.

3. What "statistics" means

• Statistics has two meanings:

Sense	Meaning	Example
Plural: the data themselves	Numerical facts	"India's rice output figures for 50 years"
Singular: the method	A science with four stages	Collection → organisation and presentation → analysis → interpretation

• The four stages of the method: 1. Collection: getting the facts, through surveys, the Census or records. 2. Organisation and presentation: classifying the facts and showing them in tables and graphs. 3. Analysis: working out summary measures (mean, dispersion, correlation). 4. Interpretation: drawing conclusions for decisions.

Quantitative vs qualitative data

• Quantitative data can be measured in numbers, such as income, output, age or price.
• NCERT's example: rice output rose from 39.58 million tonnes (1974-75) to 106.5 MT (2013-14).
• Worked example, growth over the period:
  • Growth % = (New − Old) ÷ Old × 100
  • = (106.5 − 39.58) ÷ 39.58 × 100 = 66.92 ÷ 39.58 × 100 ≈ 169%
  • So output became about 2.7 times larger in 39 years.

• Latest data: rice output reached about 150 MT (1,501.84 lakh tonnes) in 2024-25, a record [2][3]. (NCERT: 106.5 MT, 2013-14.)

  • An earlier estimate for 2024-25, the Third Advance Estimate, put it at 1,490.74 lakh tonnes, against 1,378.25 lakh tonnes in 2023-24 [2].
  • This shows that official estimates are revised in stages: advance estimates come first and the final estimate comes later. Always quote the stage and the year.

• Qualitative data record attributes (qualities that cannot be measured in numbers), such as gender, religion or marital status.

• They are sometimes recorded in degrees (ranked levels):
  • unskilled / skilled / highly skilled
  • sick / healthy
• Such data can be counted (for example, 40% of workers are skilled), but the quality itself cannot be measured.

4. What statistics does: four functions

(a) Precision

• Statistics turns vague words into exact numbers.
• "310 people died in the Kashmir earthquake" is a statistical fact. "Hundreds died" is not.
• Precision matters for policy. Relief funds, compensation and rebuilding budgets need exact counts.

(b) Condensing

• Condensing means reducing a mass of data to a few summary measures.
• Mean (average) = sum of all values ÷ number of values.
• Variance = the average of the squared gaps of each value from the mean.

• Standard deviation (SD) = √variance. It shows how widely the values are spread around the mean.

• Worked example:

• Incomes of 5 families (Rs '000/month): 10, 12, 15, 18, 45.
• Mean = 100 ÷ 5 = Rs 20,000.
• You cannot remember thousands of incomes, but you can remember one average income.
• Note: 4 of the 5 families earn less than the mean, because one rich family pulls the average up. This is why the mean is always read together with a measure of spread.

(c) Finding and testing relationships

• Statistics checks whether two variables move together, and how strongly. This is correlation.
• NCERT's pairs:
• Price and demand: price up → quantity demanded usually down. This is a negative relationship.
• Income and consumption: income up → consumption up. This is a positive relationship.

• Government spending and the price level: more spending → more demand → prices may rise.

• Theory claims a relationship. Statistics tests whether the data support it.

(d) Statistical prediction

• Statistical prediction means applying statistical tools to past or survey data to estimate future values for planning.
• NCERT's examples:
• Deciding in 2017 how much to produce in 2020.

• Deciding how much oil India should import in 2025.

• Why it matters:

• Forecast too low → shortages and price rise.
• Forecast too high → waste, idle stock and money blocked.

5. The limit: statistical methods are no substitute for common sense

• The story: a family of four knew the river's average depth. It was less than the family's average height, so they crossed, and the children drowned.
• Worked example with numbers:
• Depth at 5 points across the river (ft): 2, 2, 8, 2, 2 → average depth = 16 ÷ 5 = 3.2 ft.
• Heights of the family (ft): 6, 5.5, 3, 3 → average height = 17.5 ÷ 4 ≈ 4.4 ft.
• 3.2 < 4.4, so crossing "looks safe".

• But at the 8 ft point, the two 3 ft children are fully under water.

• The lesson:

• The average hid the spread. A range of 2–8 ft was hidden behind one number, 3.2.
• Always ask two more questions: what is the dispersion (spread), and what is the extreme (worst) value?

• Section 6 comes back to this story (measures of dispersion).

• The global rule on misuse: the UN Fundamental Principles of Official Statistics has 10 principles. Principle 4 is "Prevention of Misuse": statistical agencies may comment when their numbers are wrongly read or used [6][7].

• The Principles were first adopted by the Conference of European Statisticians in 1991 and by the UN Statistical Commission in 1994. The UN General Assembly endorsed them on 29 January 2014 in resolution A/RES/68/261 [6][7].

6. The legal backbone: who may collect data, and why people answer truthfully

• The chain from data to policy works only if people give true answers and official data are trusted.
• The Collection of Statistics (CoS) Act, 2008 is a major step to improve the legal framework of India's official statistical system [4]. It:
• lets the government collect statistics on economic, demographic and social subjects;

• protects the confidentiality of each person's answers.

• The Collection of Statistics Rules, 2024 were notified in December 2024. They apply the Act as amended by the Jan Vishwas (Amendment of Provisions) Act, 2023 [5].

• The Jan Vishwas Act is a law that removed or softened small criminal penalties in several Acts, to reduce the fear of prosecution for minor lapses.

• The link to policy: respondents answer honestly only if they trust that their data will stay confidential. Honest answers give accurate data, and accurate data give sound policy.

7. The project cycle (Class 11, Use of Statistical Tools)

• A project report follows fixed steps: 1. Identify the problem (the objective). 2. Choose the target group. 3. Collect data:

  • Primary data: collected first-hand by the researcher, for example through a questionnaire.
  • Secondary data: already collected by someone else, for example Census or NSS reports.
  • A project may use either or both. 4. Organise and present the data in tables and graphs. 5. Analyse the data: averages, dispersion, correlation. 6. Conclude, with predictions and suggestions. 7. Bibliography: list the sources used.

• Target group means the set of people the study focuses on. It follows from the objective:

Study	Target group	Why
Cars	Middle and high-income groups	Only they can afford cars
Soap	All rural and urban consumers	Almost everyone uses soap
Safe drinking water	Rural and urban people	Everyone needs it

Worked example: the toothpaste project

• Sample: 100 households, 67% urban.
• Mean monthly family income: Rs 18,000 (SD Rs 9,000).
• Mean spending on toothpaste: Rs 104 per household per month (SD Rs 35.60).
• Top brands: Pepsodent, Colgate, Close-up. Main media influence: television.
• Going one step further with the NCERT numbers:
• Coefficient of variation (CV) = SD ÷ Mean × 100. It measures spread relative to the average, so two series can be compared even when their units or sizes differ.
• Income: CV = 9,000 ÷ 18,000 × 100 = 50%.
• Toothpaste spending: CV = 35.60 ÷ 104 × 100 ≈ 34%.
• Reading: households differ much more in income than in toothpaste spending, because toothpaste is a necessity.

• Share of income spent on toothpaste = 104 ÷ 18,000 × 100 ≈ 0.58%.

• This note follows the same cycle: collection (§2-4) → organisation (§5) → analysis (§6-7) → India's data system (§8-12).

• Cross-ref: Marshall's definition, scarcity and choice → economic-problem-systems.

Prelims Hooks

• Statistics means both the data (plural) and the method (singular): collection → organisation and presentation → analysis → interpretation.
• Qualitative data (gender, skill level, health status) cannot be measured in numbers. They can only be counted or ranked in degrees. Trap: "skilled / unskilled" is qualitative, not quantitative.
• The correct order of the chain is data → economic analysis → economic policy → evaluation.
• The coefficient of variation = SD ÷ Mean × 100. It compares spread across series. Toothpaste example: income CV 50%, spending CV about 34%.
• The river-crossing story teaches that an average hides dispersion. Statistics is no substitute for common sense.
• Rice output was about 150 MT (1,501.84 lakh tonnes) in 2024-25, a record [2][3]. (NCERT: 39.58 MT in 1974-75 → 106.5 MT in 2013-14.)
• The Collection of Statistics Act, 2008 is India's main law for official data collection. Its Rules were notified in December 2024, in line with the Jan Vishwas Act, 2023 [4][5].
• The UN Fundamental Principles of Official Statistics has 10 principles. It was endorsed by the UNGA on 29 January 2014 (A/RES/68/261). Principle 4 is Prevention of Misuse [6][7].
• The PLFS was revamped from January 2025. The first monthly bulletin was for April 2025. It reports LFPR, WPR and UR for rural and urban India [8][9].

Mains Points

• Evidence-based policy needs timely data.
• Monthly PLFS indicators from 2025 [8] cut the gap between a labour-market shock and the policy response.

• Slow or old data (for example, a Census that is delayed) weaken the "evaluation" link. Schemes then run on outdated numbers.

• Trust in data is a governance issue (GS-II link).

• The CoS Act, 2008 promises confidentiality [4], and the UN Principles call for impartiality and prevention of misuse [6].

• Without trust, people give wrong answers, and bad data then lead to bad targeting of subsidies and welfare.

• The limits of averages in policy.

• A rising average income, or record rice output [2], can hide inequality between regions, classes and genders. This is the river-crossing lesson.

• Policy should use distribution measures (SD, CV, the lowest decile) and not the mean alone.

• Evaluation closes the loop.

• NCERT's family-planning question shows why schemes need built-in data and outcome measures.
• Output counts, such as money spent, do not show impact. Only data on outcomes can show whether a policy worked.