Quantitative Techniques in Economics
Economics — Grade 12 | Unit 6 | NEB Nepal
Table Of Contents
Introduction
Unit 6 of NEB Grade 12 Economics revisits quantitative techniques at a more advanced level than Grade 11. While Grade 11 introduced basic statistics and mathematics, Grade 12 extends these tools into more sophisticated applications: advanced index numbers (chain-based, value index, deflation and splicing), regression analysis (the tool that takes correlation to its logical next step — predicting one variable from another), and the mathematics of economic functions including revenue, cost, profit maximization, and elasticity. These are the quantitative foundations of professional economic analysis — the tools used by Nepal Rastra Bank economists, Central Bureau of Statistics researchers, and government planners every day.
1. Advanced Index Numbers
1.1 Review and Extension
According to Irving Fisher, “An index number is a ratio, or an average of ratios, expressing the relative change in price, quantity, value, or some other measure from a base period to a given period.” Grade 12 extends the index number analysis of Grade 11 — introducing chain-based index numbers, value index numbers, and the important practical applications of deflation and splicing.
According to Croxton and Cowden, “Index numbers are statistical devices designed to measure changes in the magnitude of a group of related variables with respect to time, geographic location, or other characteristics.”
1.2 Weighted Index Numbers
Simple unweighted index numbers give equal importance to all items — but in economic reality, different items have different importance. Weighted index numbers assign weights reflecting the relative importance of each item.
Laspeyre’s Price Index (Base Year Weighted):
$$P_{01}^L = \frac{\sum p_1 q_0}{\sum p_0 q_0} \times 100$$
Paasche’s Price Index (Current Year Weighted):
$$P_{01}^P = \frac{\sum p_1 q_1}{\sum p_0 q_1} \times 100$$
Fisher’s Ideal Price Index:
$$P_{01}^F = \sqrt{P_{01}^L \times P_{01}^P}$$
These were covered in Grade 11. Grade 12 extends to chain indices, value indices, and practical applications.
1.3 Chain-Based Index Numbers
A chain-based index (also called a chain index or chain-linked index) calculates each period’s index relative to the immediately preceding period — rather than relative to a fixed base year. The results are then “chained” together to create a continuous series.
According to the Office for National Statistics (UK), “Chain-linked indices are the preferred method for measuring real economic growth — by using the most recent year’s weights in each comparison, they better reflect current economic structure.”
Construction of a Chain Index:
Step 1: Calculate a link relative for each period: $LR_t = \frac{Value_t}{Value_{t-1}} \times 100$
Step 2: Chain the link relatives together by multiplying successive relatives and dividing by 100:
$Chain\ Index_t = \frac{LR_1 \times LR_2 \times … \times LR_t}{100^{(t-1)}}$
Example: Prices of rice (Rs. per kg) in Nepal:
| Year | Price (Rs.) | Link Relative (LR) | Chain Index (Base: Year 1 = 100) |
|---|---|---|---|
| Year 1 | 50 | 100 (base) | 100.0 |
| Year 2 | 55 | (55/50)×100 = 110 | 110.0 |
| Year 3 | 60 | (60/55)×100 = 109.1 | (110×109.1)/100 = 120.0 |
| Year 4 | 63 | (63/60)×100 = 105 | (120×105)/100 = 126.0 |
Advantages of chain indices:
- Always use the most recent weights — reflecting current consumption patterns
- More accurately capture the effect of substitution as relative prices change
- Preferred for measuring real GDP growth (used by CBS Nepal in national accounts)
Disadvantages:
- Cannot be directly compared to earlier periods without tracing the chain
- Computationally more demanding
- Cannot be “spliced” easily without introducing inconsistency
1.4 Value Index Numbers
A value index measures changes in the total monetary value of transactions — combining changes in both price and quantity.
$$V_{01} = \frac{\sum p_1 q_1}{\sum p_0 q_0} \times 100$$
The value index equals 100 × (total value in current period / total value in base period). Unlike price and quantity indices, it reflects both price changes and volume changes simultaneously.
Relationship between price, quantity, and value indices:
$$V_{01} = \frac{P_{01}^F \times Q_{01}^F}{100}$$
This is the factor reversal test — one of the criteria that Fisher’s Ideal Index satisfies. The value index equals the product of Fisher’s price index and Fisher’s quantity index (divided by 100).
Application in Nepal: Nepal’s Central Bureau of Statistics uses value indices to track changes in the total value of Nepal’s exports and imports. If Nepal’s total export value index rises from 100 to 150 between two periods, this means the total value of exports increased by 50% — though this combines both price and volume effects.
1.5 Deflation of Index Numbers
Deflation is the process of removing the effect of price changes from a value series — converting nominal (current price) values into real (constant price) values that reflect only changes in physical volume.
$$Real\ Value = \frac{Nominal\ Value}{Price\ Index} \times 100$$
According to Paul A. Samuelson, “Deflation of economic series is essential for understanding real changes in output, income, and living standards — nominal growth that merely reflects inflation is economically meaningless.”
Application to Nepal’s GDP:
If Nepal’s nominal GDP grows from Rs. 100 billion to Rs. 110 billion but the price level (as measured by the GDP deflator) rises from 100 to 108:
$$Real\ GDP = \frac{110}{108} \times 100 = 101.85\ billion$$
Real growth is only 1.85% — not the 10% suggested by nominal growth. Most of the nominal increase was price inflation, not real output growth.
Deflating wages to find real wages: If nominal wages rise 8% but inflation is 6%, real wages have risen by approximately 2% — workers can buy 2% more goods, not 8% more.
Practical examples in Nepal:
- Nepal Rastra Bank deflates nominal credit growth by the inflation rate to find real credit growth
- Nepal’s annual budget documents present both nominal and real expenditure figures
- Nepal’s agricultural statistics deflate crop value indices to separate price effects from yield changes
1.6 Splicing of Index Numbers
Splicing is the technique of joining two overlapping index number series — computed with different base years or different methodologies — into a single continuous series.
This is necessary when: the base year is changed, the index basket is updated, or the methodology is revised — creating a break in the series that would otherwise make comparison impossible.
Splicing procedure:
If the old series runs from Period 0 to Period T (with base = Period 0) and the new series starts from Period S (where S < T) with a new base:
$$Spliced\ index\ for\ period\ t = Old\ Index_t \times \frac{New\ Index_T}{Old\ Index_T}$$
The overlap period (Period S to T) is used to calculate the linking factor — the ratio that converts old series values to the new series scale.
Example: Nepal’s Consumer Price Index (CPI) has been periodically rebased — from base year 2005/06 to 2018/19. When the base is changed, CBS Nepal splices the old and new series together so that a continuous inflation rate can be calculated across the entire period.
According to the International Monetary Fund (IMF), “Splicing is essential for maintaining long-run comparability of economic indices when methodological changes make direct comparison impossible — without splicing, historical analysis of price trends would be disrupted by every base year revision.”
2. Regression Analysis
2.1 Meaning and Purpose
According to Sir Francis Galton, who coined the term “regression” in his 1886 study of heredity, regression analysis measures the statistical relationship between variables in a way that allows prediction. (Galton noticed that tall parents tended to have children who “regressed” toward the mean height — giving the technique its name.)
According to Croxton and Cowden, “Regression analysis is a statistical technique that establishes the mathematical relationship between a dependent variable and one or more independent variables — enabling prediction of the dependent variable for given values of the independent variable(s).”
According to Jan Tinbergen, the pioneer of econometrics, “Regression analysis is the fundamental tool of quantitative economic analysis — it enables economists to estimate the magnitude of economic relationships from observed data, rather than merely asserting that relationships exist.”
Purpose of regression analysis in economics:
- Estimating demand functions (relationship between price and quantity demanded)
- Measuring the determinants of income, wages, and growth
- Forecasting future economic variables (GDP, inflation, trade)
- Evaluating the impact of policies (effect of a minimum wage on employment)
- Testing economic theories against real-world data
2.2 Difference Between Correlation and Regression
According to A.L. Bowley, “Correlation measures the degree and direction of association between two variables — regression specifies the nature of that relationship and enables prediction.”
| Basis | Correlation | Regression |
|---|---|---|
| Purpose | Measures degree of association | Estimates the mathematical relationship; enables prediction |
| Direction | Symmetric: r(X,Y) = r(Y,X) | Asymmetric: regression of Y on X ≠ regression of X on Y |
| Output | A single coefficient (r) between −1 and +1 | An equation: Y = a + bX |
| Causality | Implies no direction of causation | Implies direction: X influences Y |
| Number of lines | One correlation coefficient | Two regression lines (Y on X, and X on Y) |
| Application | Describing relationship | Predicting one variable from another |
2.3 Lines of Regression
For any bivariate data set (X, Y), there are two regression lines:
i. Regression of Y on X (predicts Y from X):
$$Y = a + bX$$
Where the regression coefficient b (slope) and intercept a are calculated from:
$$b_{yx} = \frac{N\sum XY – \sum X \sum Y}{N\sum X^2 – (\sum X)^2}$$
$$a = \bar{Y} – b_{yx}\bar{X}$$
This line is used to estimate Y for a given value of X.
ii. Regression of X on Y (predicts X from Y):
$$X = c + dY$$
$$b_{xy} = \frac{N\sum XY – \sum X \sum Y}{N\sum Y^2 – (\sum Y)^2}$$
$$c = \bar{X} – b_{xy}\bar{Y}$$
This line is used to estimate X for a given value of Y.
The two regression lines are different — one minimizes vertical (Y) deviations, the other minimizes horizontal (X) deviations. They intersect at the point $(\bar{X}, \bar{Y})$ — the means of X and Y.
When r = ±1 (perfect correlation), both lines coincide — there is only one line and perfect prediction is possible. When r = 0 (no correlation), the lines are perpendicular — knowing X tells you nothing about Y, and vice versa.
2.4 Regression Coefficient and Its Properties
The regression coefficient b measures the change in Y for a unit change in X:
- $b_{yx}$ = regression coefficient of Y on X = how much Y changes per unit increase in X
- $b_{xy}$ = regression coefficient of X on Y = how much X changes per unit increase in Y
Properties of regression coefficients:
i. Both have the same sign: $b_{yx}$ and $b_{xy}$ are both positive or both negative — matching the sign of the correlation coefficient r.
ii. Geometric mean equals |r|: $\sqrt{b_{yx} \times b_{xy}} = |r|$
This is a key formula that directly connects regression coefficients to the correlation coefficient.
iii. Both cannot exceed 1 in absolute value simultaneously: If $|b_{yx}| > 1$, then $|b_{xy}| < 1$.
iv. Units differ: $b_{yx}$ is measured in units of Y per unit of X; $b_{xy}$ is measured in units of X per unit of Y — they have different units and cannot be directly compared.
v. Affected by scale but not location: Changing the origin (subtracting a constant) does not change regression coefficients; changing the scale (multiplying by a constant) does change them.
2.5 Worked Example
Problem: Nepal Rastra Bank wants to study the relationship between money supply growth (X, % per year) and inflation (Y, % per year). Data for 8 years:
| Year | X (Money supply growth %) | Y (Inflation %) |
|---|---|---|
| 1 | 12 | 7 |
| 2 | 15 | 8 |
| 3 | 18 | 10 |
| 4 | 20 | 11 |
| 5 | 14 | 8 |
| 6 | 22 | 13 |
| 7 | 25 | 14 |
| 8 | 17 | 9 |
Step 1: Calculate required sums:
| X | Y | X² | Y² | XY |
|---|---|---|---|---|
| 12 | 7 | 144 | 49 | 84 |
| 15 | 8 | 225 | 64 | 120 |
| 18 | 10 | 324 | 100 | 180 |
| 20 | 11 | 400 | 121 | 220 |
| 14 | 8 | 196 | 64 | 112 |
| 22 | 13 | 484 | 169 | 286 |
| 25 | 14 | 625 | 196 | 350 |
| 17 | 9 | 289 | 81 | 153 |
| ΣX=143 | ΣY=80 | ΣX²=2687 | ΣY²=844 | ΣXY=1505 |
N = 8, $\bar{X}$ = 143/8 = 17.875, $\bar{Y}$ = 80/8 = 10
Step 2: Calculate $b_{yx}$:
$$b_{yx} = \frac{N\sum XY – \sum X \sum Y}{N\sum X^2 – (\sum X)^2} = \frac{8(1505) – (143)(80)}{8(2687) – (143)^2} = \frac{12040 – 11440}{21496 – 20449} = \frac{600}{1047} = 0.573$$
Step 3: Calculate intercept a:
$$a = \bar{Y} – b_{yx}\bar{X} = 10 – 0.573(17.875) = 10 – 10.24 = -0.24$$
Regression equation (Y on X): Y = −0.24 + 0.573X
Interpretation: For each 1% increase in money supply growth, inflation increases by approximately 0.573 percentage points — consistent with the monetarist view that money supply growth drives inflation (though the coefficient less than 1 suggests other factors also matter).
Prediction: If money supply grows by 20% next year: Y = −0.24 + 0.573(20) = −0.24 + 11.46 = 11.22% predicted inflation
2.6 Applications of Regression in Nepal’s Economy
i. NRB monetary policy: Nepal Rastra Bank uses regression models to estimate the relationship between money supply, credit growth, and inflation — informing its monetary policy decisions.
ii. Agricultural forecasting: CBS Nepal uses regression to forecast crop production from planted area, rainfall, and input use data.
iii. Tourism demand forecasting: Nepal Tourism Board estimates the relationship between tourist arrivals and factors like India-Nepal connectivity, airfare prices, and GDP of source countries.
iv. Poverty analysis: Regression analysis identifies the determinants of household income and consumption — quantifying the contribution of education, land ownership, remittances, and geography to poverty outcomes.
v. Public finance analysis: Nepal’s fiscal authorities use regression to forecast tax revenue from GDP growth projections — estimating tax buoyancy (the elasticity of tax revenue with respect to GDP).
3. Mathematics for Economics
3.1 Total Revenue, Average Revenue, and Marginal Revenue Functions
If demand is a linear function: P = a − bQ (downward-sloping demand curve)
Total Revenue: TR = P × Q = (a − bQ)Q = aQ − bQ²
Average Revenue: AR = TR/Q = a − bQ = P (the demand curve itself)
Marginal Revenue: MR = dTR/dQ = a − 2bQ
Key result: When demand is linear (P = a − bQ), MR has the same intercept (a) but twice the slope (−2b) compared to the demand/AR curve. MR falls twice as fast as AR.
Example: If P = 100 − 5Q:
- TR = 100Q − 5Q²
- AR = 100 − 5Q
- MR = 100 − 10Q
At Q = 8: P = AR = 60; MR = 100 − 80 = 20; TR = 60 × 8 = 480
Price elasticity and TR/MR relationship:
$$MR = P\left(1 + \frac{1}{e_d}\right) = P\left(\frac{e_d + 1}{e_d}\right)$$
Where $e_d$ = price elasticity of demand (negative for downward-sloping demand)
Implications:
- When |$e_d$| > 1 (elastic): MR > 0; TR increases as price falls
- When |$e_d$| = 1 (unitary elastic): MR = 0; TR is maximized
- When |$e_d$| < 1 (inelastic): MR < 0; TR decreases as price falls
3.2 Total Cost, Average Cost, and Marginal Cost Functions
Typical cubic total cost function: TC = a + bQ − cQ² + dQ³
Where a = fixed cost; the cubic form generates the S-shaped TC curve and U-shaped AC and MC curves.
Fixed Cost: FC = a (constant term)
Variable Cost: TVC = TC − FC = bQ − cQ² + dQ³
Average Fixed Cost: AFC = FC/Q = a/Q (continuously declining)
Average Variable Cost: AVC = TVC/Q = b − cQ + dQ²
Average Total Cost: ATC = TC/Q = a/Q + b − cQ + dQ²
Marginal Cost: MC = dTC/dQ = b − 2cQ + 3dQ²
Finding minimum ATC and AVC:
Set dATC/dQ = 0 and dAVC/dQ = 0 respectively to find the output levels at which average costs are minimized.
Key property: MC = AVC at minimum AVC; MC = ATC at minimum ATC. This can be proved mathematically:
If AVC = TVC/Q, then: d(AVC)/dQ = [Q·(dTVC/dQ) − TVC]/Q² = (Q·MC − TVC)/Q²
Setting d(AVC)/dQ = 0: Q·MC = TVC → MC = TVC/Q = AVC
So MC = AVC at minimum AVC. ✓
Example: If TC = 100 + 10Q − 3Q² + Q³
- FC = 100
- TVC = 10Q − 3Q² + Q³
- AVC = 10 − 3Q + Q²
- ATC = 100/Q + 10 − 3Q + Q²
- MC = 10 − 6Q + 3Q²
Minimum AVC: dAVC/dQ = −3 + 2Q = 0 → Q = 1.5 (minimum AVC point) AVC at Q = 1.5: 10 − 3(1.5) + (1.5)² = 10 − 4.5 + 2.25 = 7.75
Verify MC = AVC at Q = 1.5: MC = 10 − 6(1.5) + 3(1.5)² = 10 − 9 + 6.75 = 7.75 ✓
3.3 Profit Function and Profit Maximization
Profit (π) = Total Revenue − Total Cost = TR − TC
Profit maximization condition (first-order condition): dπ/dQ = 0
Since π = TR − TC: $$\frac{d\pi}{dQ} = \frac{dTR}{dQ} – \frac{dTC}{dQ} = MR – MC = 0$$
Therefore: MR = MC (profit maximization condition)
Second-order condition (to confirm it is a maximum, not minimum): $$\frac{d^2\pi}{dQ^2} = \frac{d(MR)}{dQ} – \frac{d(MC)}{dQ} < 0$$
This means: the slope of MR must be less than the slope of MC — MC must be rising faster than MR (or MR must be falling faster than MC) at the profit-maximizing point.
Worked example: A firm faces demand P = 200 − 8Q and TC = 100 + 12Q²
TR = (200 − 8Q)Q = 200Q − 8Q² MR = 200 − 16Q MC = 24Q
Setting MR = MC: 200 − 16Q = 24Q → 40Q = 200 → Q = 5*
Price: P* = 200 − 8(5) = Rs. 160
Profit: TR = 160 × 5 = Rs. 800 TC = 100 + 12(25) = Rs. 400 π = 800 − 400 = Rs. 400
Second-order check: d²π/dQ² = d(MR)/dQ − d(MC)/dQ = −16 − 24 = −40 < 0 ✓ (confirmed maximum)
3.4 Price Elasticity from Demand Functions
From a linear demand function P = a − bQ (or Q = (a/b) − (1/b)P):
$$e_d = \frac{dQ}{dP} \times \frac{P}{Q}$$
Since dQ/dP = −1/b (slope of demand with respect to price):
$$e_d = -\frac{1}{b} \times \frac{P}{Q} = -\frac{P}{bQ}$$
From the demand function Q = f(P):
$$e_d = \frac{dQ}{dP} \times \frac{P}{Q}$$
Example: If demand is Q = 100 − 2P, find price elasticity at P = 30:
At P = 30: Q = 100 − 2(30) = 40
dQ/dP = −2
$$e_d = -2 \times \frac{30}{40} = -1.5$$
Interpretation: Demand is elastic at P = 30 — a 1% price increase reduces quantity demanded by 1.5%.
Point elasticity formula for non-linear demand: For Q = aP^b:
$$e_d = b$$
The elasticity equals the exponent directly — a constant elasticity demand function.
Income elasticity from an Engel function Q = f(Y):
$$e_y = \frac{dQ}{dY} \times \frac{Y}{Q}$$
Cross elasticity from a demand system Q_A = f(P_A, P_B):
$$e_{AB} = \frac{\partial Q_A}{\partial P_B} \times \frac{P_B}{Q_A}$$
(partial derivative, since both prices affect quantity demanded of A)
4. Quantitative Techniques in Nepal’s Economic Analysis
i. Central Bureau of Statistics (CBS): CBS Nepal applies index number theory daily — calculating and publishing the Consumer Price Index (CPI), the National Accounts (GDP deflated from nominal to real), and the National Living Standards Survey poverty statistics. Understanding chain indices and deflation is essential to interpreting CBS publications.
ii. Nepal Rastra Bank (NRB): NRB economists use regression analysis to estimate the determinants of inflation, the impact of monetary policy on credit growth, and the relationship between remittances and the exchange rate. NRB’s Monetary Policy framework is informed by quantitative models of the Nepali macroeconomy.
iii. National Planning Commission: Nepal’s planners use revenue and cost analysis — marginal cost of infrastructure provision, cost-benefit ratios of development projects — to allocate the development budget across competing priorities. Quantitative project appraisal is a core planning function.
iv. Development research: Nepal’s universities, think tanks (Institute for Integrated Development Studies, South Asia Watch on Trade Economics and Environment), and international research institutions (World Bank, ADB) use econometric regression to analyze the determinants of poverty, inequality, and growth in Nepal — generating the evidence base for policy advocacy.
v. Business decision-making: Nepal’s banks use regression-based credit scoring models. Manufacturing firms estimate demand functions. Trekking companies forecast tourist arrivals. In each case, the mathematical tools of Unit 6 are the practical instruments of economic decision-making.
Conclusion
Quantitative techniques are the bridge between economic theory and economic reality — they enable economists and policymakers to move from “prices and quantities are related” (qualitative) to “a 1% increase in money supply growth increases inflation by 0.573 percentage points” (quantitative). This precision is not pedantry — it is the difference between economic analysis that can guide policy and economic intuition that can only describe it.
As Jan Tinbergen observed, “Econometrics — the application of statistical and mathematical methods to economic problems — has transformed economics from a science of principles into a science of measurement.” For Nepal’s Grade 12 economics students, mastery of quantitative techniques completes the analytical toolkit that makes economics both intellectually rigorous and practically powerful.
As Paul Samuelson noted, “Good economic analysis requires both clarity of theory and precision of measurement — neither is sufficient alone.” The quantitative techniques in Unit 6 provide the precision that transforms clear economic thinking into evidence-based economic advice — the foundation of informed policy in Nepal and everywhere else.
Prepared for NEB Grade 12 Economics — Unit 6: Quantitative Techniques in Economics Aligned with the National Curriculum Framework 2076, Curriculum Development Centre, Sanothimi, Bhaktapur
