Revealed preference theory as a breakthrough in demand analysis

Samuelson (1938) built demand theory on observed choice instead of introspection. Its main contributions are:

1. From unobservable utility to observable behaviour. Earlier theory rested on cardinal utility, diminishing marginal utility and indifference curves, none of which can be observed. Revealed preference (RP) infers preferences from actual purchases: if a consumer buys bundle A when bundle B was also affordable, A is revealed preferred to B.
2. Law of demand without utility. Using the Weak Axiom of Revealed Preference (WARP), if A is chosen when B is affordable, B cannot be chosen when A is affordable. From this consistency alone Samuelson derived the negative slope of the compensated demand curve, with no need for measurable utility or indifference maps.
3. Testable, positive foundation. Because the assumptions concern observable choices, the theory can be checked against market data (consumer surveys, household budget data such as the Nepal Living Standards Survey). This moved demand theory towards positive science.
4. Wider applications. RP supplies tools for income and substitution effects, index numbers (Laspeyres and Paasche price indices and cost-of-living comparisons), welfare comparisons, and later work on choice under uncertainty.
5. Basis of modern theory. The Strong Axiom (Houthakker) and the Generalised Axiom (Afriat) show that consistent choices can be rationalised by a utility function, so RP links the ordinal approach with empirical demand analysis.

Limitations: it assumes stable tastes, a single consumer and non-satiation, and it cannot reveal indifference between bundles by itself. Even so, its shift to observable choice is why it is regarded as a breakthrough.

Lancasterian Demand Theory (Lancaster's Characteristics Approach)

1. Introduction. Developed by Kelvin J. Lancaster (1966), this theory rejects the traditional view that consumers derive utility directly from goods. Instead, it holds that goods are carriers of characteristics (attributes), and it is these characteristics that yield utility. Consumers demand goods not for themselves but for the characteristics they contain.

2. Core propositions.

• Goods possess characteristics: Every good is a bundle of objective properties — e.g., a car provides transport, comfort, prestige, and safety.
• Utility comes from characteristics, not goods: The consumer's utility function is defined over characteristics:
  $U = U(z_1, z_2, \dots, z_n)$
• Consumption is an activity: A good is consumed as part of an activity that transforms goods into characteristics. The relationship is linear:
  $z_j = \sum_{i} a_{ij} \, x_i$
  where $a_{ij}$ = amount of characteristic $j$ per unit of good $i$, and $x_i$ = quantity of good $i$ consumed.
• Multiple characteristics per good, multiple goods per characteristic: A single good may supply several characteristics, and one characteristic may be supplied by several goods. This creates substitutability among goods at the characteristic level.
• Consumer choice: The consumer maximizes utility subject to the budget constraint $\sum p_i x_i = M$, choosing the combination of goods that gives the most preferred characteristic bundle.

3. Diagrammatic illustration. Characteristics $z_1$ and $z_2$ are measured on the axes. Each good is represented by a ray from the origin whose slope equals the ratio of its characteristics ($a_{1j}/a_{2j}$). The efficiency frontier shows the maximum characteristics attainable from a given expenditure. The consumer's indifference curves over characteristics are convex; the optimum occurs where the frontier is tangent to the highest attainable indifference curve. Corner solutions are common — the consumer may buy only one good if it dominates in the desired characteristic.

4. Significance.

• Explains brand choice, product differentiation, and advertising in terms of attributes rather than the goods themselves.
• Provides a rigorous basis for hedonic pricing and quality measurement.
• Explains why consumers buy bundles and why new goods (new characteristic combinations) alter demand.
• Bridges microeconomic theory with marketing and consumer behaviour analysis.

5. Limitations.

• Characteristics are difficult to measure and quantify objectively.
• The linear goods–characteristics relation is an oversimplification.
• Assumes perfect information about attributes and ignores habit, taste formation, and bounded rationality.
• The number of relevant characteristics may exceed the number of goods, making the model indeterminate.

Conclusion. Lancasterian demand theory shifts the unit of analysis from the good to its characteristics, offering a richer and more realistic explanation of consumer choice than the traditional utility approach.

Given

Demand function (implied):
[ P=120-4Q ] So total revenue: [ TR=120Q-4Q^2 ]

Total cost: [ TC=5+4Q+0.6Q^2 ]

Profit-maximising condition

[ MR=MC ]

[ MR=\frac{dTR}{dQ}=120-8Q ]

[ MC=\frac{dTC}{dQ}=4+1.2Q ]

Set (MR=MC):

[ 120-8Q=4+1.2Q ]

[ 116=9.2Q ]

[ Q=\frac{116}{9.2}=12.6087 \approx 12.61 ]

So profit-maximising output is:

[ \boxed{Q \approx 12.61 \text{ units}} ]

Equilibrium price

Using the demand function:

[ P=120-4Q ]

[ P=120-4(12.6087) ]

[ P=120-50.4348=69.5652 ]

[ \boxed{P \approx Rs\ 69.57} ]

Second-order condition

[ \frac{d^2\pi}{dQ^2}=-9.2<0 ]

So profit is maximised at (Q=12.61).

Answer:
[ \boxed{\text{Equilibrium output } Q \approx 12.61 \text{ units}} ] [ \boxed{\text{Equilibrium price } P \approx Rs\ 69.57} ]

Note: The printed (TR=120-4Q^2) appears to be a misprint. The standard demand form (P=120-4Q), giving (TR=120Q-4Q^2), is used above.

Derivation of the Cost Function from the Production Function

1. The Basic Idea

The cost function expresses the minimum cost of producing any given level of output, given input prices. It is derived from the production function by solving the firm's cost-minimization problem. Formally, for output level $Q$ and input prices $w$ (labour) and $r$ (capital):

$C = C(Q, w, r)$

The production function, on the other hand, shows the maximum output obtainable from given inputs:

$Q = f(L, K)$

2. The Cost-Minimization Problem

The firm minimizes total cost subject to producing a given level of output:

$\min_{L, K} \; C = wL + rK$

$\text{subject to} \quad f(L, K) = Q$

3. First-Order Conditions (Lagrangian Method)

Set up the Lagrangian:

$\mathcal{L} = wL + rK + \lambda \left[ Q - f(L, K) \right]$

First-order conditions:

$\frac{\partial \mathcal{L}}{\partial L} = w - \lambda f_L = 0 \quad \Rightarrow \quad w = \lambda f_L$

$\frac{\partial \mathcal{L}}{\partial K} = r - \lambda f_K = 0 \quad \Rightarrow \quad r = \lambda f_K$

$\frac{\partial \mathcal{L}}{\partial \lambda} = Q - f(L, K) = 0$

4. Tangency (Optimality) Condition

Dividing the first two conditions:

$\frac{w}{r} = \frac{f_L}{f_K} = MRTS_{LK}$

This says the firm equates the input price ratio to the marginal rate of technical substitution. This condition, together with the production constraint, gives the conditional input demand functions:

$L^* = L^*(Q, w, r), \qquad K^* = K^*(Q, w, r)$

5. Obtaining the Cost Function

Substituting the conditional input demands back into the cost equation gives the cost function:

$C(Q, w, r) = w \, L^*(Q, w, r) + r \, K^*(Q, w, r)$

This is the total cost function, which gives the minimum cost of producing output $Q$ at input prices $w$ and $r$.

6. Example: Cobb–Douglas Production Function

Let the production function be:

$Q = A L^{\alpha} K^{\beta}$

Step 1 — Marginal products:

$f_L = \alpha A L^{\alpha-1} K^{\beta} = \frac{\alpha Q}{L}, \qquad f_K = \beta A L^{\alpha} K^{\beta-1} = \frac{\beta Q}{K}$

Step 2 — Tangency condition:

$\frac{w}{r} = \frac{\alpha K}{\beta L} \quad \Rightarrow \quad K = \frac{\beta}{\alpha} \cdot \frac{w}{r} \cdot L$

Step 3 — Substitute into the production function:

$Q = A L^{\alpha} \left( \frac{\beta w}{\alpha r} L \right)^{\beta} = A \left( \frac{\beta w}{\alpha r} \right)^{\beta} L^{\alpha + \beta}$

Step 4 — Conditional input demands:

$L^* = \left[ \frac{Q}{A} \left( \frac{\alpha r}{\beta w} \right)^{\beta} \right]^{\frac{1}{\alpha+\beta}}, \qquad K^* = \left[ \frac{Q}{A} \left( \frac{\beta w}{\alpha r} \right)^{\alpha} \right]^{\frac{1}{\alpha+\beta}}$

Step 5 — Total cost function:

$C(Q, w, r) = w L^* + r K^* = (\alpha + \beta) \left( \frac{Q}{A} \right)^{\frac{1}{\alpha+\beta}} \left( \frac{w}{\alpha} \right)^{\frac{\alpha}{\alpha+\beta}} \left( \frac{r}{\beta} \right)^{\frac{\beta}{\alpha+\beta}}$

For the special case of constant returns to scale ($\alpha + \beta = 1$):

$C = Q \cdot \frac{1}{A} \left( \frac{w}{\alpha} \right)^{\alpha} \left( \frac{r}{\beta} \right)^{\beta}$

which is linear in $Q$ — average cost is constant.

7. Conclusion

The cost function is derived by (i) solving the cost-minimization problem, (ii) using the tangency condition $\dfrac{w}{r} = \dfrac{f_L}{f_K}$ to find conditional input demands, and (iii) substituting these demands into $wL + rK$. The resulting function $C(Q, w, r)$ is the minimum-cost function, and it is dual to the production function — each contains the same technological information.

1. Concept of Peak Load Pricing

Peak load pricing is a pricing technique used mainly by public utilities and service providers—such as electricity, telecommunication, transport, and water supply—where demand fluctuates sharply between peak periods and off-peak periods.

Under this system, a higher price is charged during peak demand periods and a lower price during off-peak periods. The main reason is that during peak periods, demand may exceed existing capacity, forcing the firm to install additional capacity. Therefore, peak-period users should bear not only the marginal operating cost but also part of the marginal capacity cost.

Features

• Demand varies by time: peak and off-peak.
• The product or service is generally non-storable, e.g. electricity.
• Capacity is fixed in the short run.
• Peak users pay a higher price; off-peak users pay a lower price.
• It helps to shift some demand from peak to off-peak, reducing pressure on capacity.

Pricing Rule

If (MC_o) is marginal operating cost and (MC_c) is marginal capacity cost:

[ P_{\text{peak}} = MC_o + MC_c ]

[ P_{\text{off-peak}} = MC_o ]

Advantages

1. Promotes efficient use of existing capacity.
2. Reduces need for additional investment in capacity.
3. Encourages consumers to use services during off-peak hours.
4. Helps utilities recover capacity costs fairly.
5. Improves social welfare by reducing congestion and deadweight loss.

Limitations

1. Requires separate metering for peak and off-peak use.
2. May be difficult to apply where demand cannot be separated by time.
3. Consumers may not respond strongly to price differences.

---

2. Given Demand and Supply Functions

[ Q_d = 4000 - 50P ]

[ Q_s = -1500 + 50P ]

where (P) is price in Rs and (Q) is quantity.

---

a. Equilibrium Price and Quantity

At equilibrium,

[ Q_d = Q_s ]

[ 4000 - 50P = -1500 + 50P ]

[ 4000 + 1500 = 50P + 50P ]

[ 5500 = 100P ]

[ P = 55 ]

Substituting (P = 55) into demand:

[ Q = 4000 - 50(55) ]

[ Q = 4000 - 2750 = 1250 ]

Equilibrium price = Rs 55
Equilibrium quantity = 1250 units

---

b. Consumer’s Surplus and Producer’s Surplus

Inverse demand function

[ Q_d = 4000 - 50P ]

[ 50P = 4000 - Q ]

[ P = 80 - \frac{Q}{50} ]

[ P = 80 - 0.02Q ]

At equilibrium, (Q = 1250), (P = 55).

Maximum willingness to pay at (Q = 0) is (P = 80).

Consumer Surplus

[ CS = \frac{1}{2} \times Q \times (P_{\max} - P_e) ]

[ CS = \frac{1}{2} \times 1250 \times (80 - 55) ]

[ CS = \frac{1}{2} \times 1250 \times 25 ]

[ CS = 15625 ]

Consumer surplus = Rs 15,625

Inverse supply function

[ Q_s = -1500 + 50P ]

[ 50P = Q + 1500 ]

[ P = 30 + \frac{Q}{50} ]

[ P = 30 + 0.02Q ]

Minimum supply price at (Q = 0) is (P = 30).

Producer Surplus

[ PS = \frac{1}{2} \times Q \times (P_e - P_{\min}) ]

[ PS = \frac{1}{2} \times 1250 \times (55 - 30) ]

[ PS = \frac{1}{2} \times 1250 \times 25 ]

[ PS = 15625 ]

Producer surplus = Rs 15,625

Total Surplus

[ TS = CS + PS = 15625 + 15625 = 31250 ]

Total surplus = Rs 31,250

---

c. Impact of Rs 10 Tax per Unit Imposed on Producer

A tax of Rs 10 per unit imposed on the producer shifts the supply curve upward by Rs 10. If (P) is the price paid by consumers, producers receive (P - 10).

New supply function:

[ Q_s = -1500 + 50(P - 10) ]

[ Q_s = -1500 + 50P - 500 ]

[ Q_s = -2000 + 50P ]

At new equilibrium,

[ Q_d = Q_s ]

[ 4000 - 50P = -2000 + 50P ]

[ 6000 = 100P ]

[ P = 60 ]

Substituting (P = 60) into demand:

[ Q = 4000 - 50(60) ]

[ Q = 4000 - 3000 = 1000 ]

New equilibrium price paid by consumers = Rs 60
New equilibrium quantity = 1000 units

Price received by producers:

[ P_s = P - 10 = 60 - 10 = 50 ]

So:

• Price rises from Rs 55 to Rs 60.
• Quantity falls from 1250 units to 1000 units.
• Consumers bear Rs 5 of tax per unit.
• Producers bear Rs 5 of tax per unit.

---

d. Consumer Surplus, Producer Surplus and Deadweight Loss After Tax

New Consumer Surplus

Consumer price (P_c = 60), quantity (Q = 1000).

[ CS = \frac{1}{2} \times 1000 \times (80 - 60) ]

[ CS = \frac{1}{2} \times 1000 \times 20 ]

[ CS = 10000 ]

New consumer surplus = Rs 10,000

New Producer Surplus

Producer net price (P_s = 50), quantity (Q = 1000).

[ PS = \frac{1}{2} \times 1000 \times (50 - 30) ]

[ PS = \frac{1}{2} \times 1000 \times 20 ]

[ PS = 10000 ]

New producer surplus = Rs 10,000

Deadweight Loss

Original total surplus:

[ TS_0 = 31250 ]

New total surplus including tax revenue:

[ CS + PS + \text{Tax Revenue} ]

Tax revenue:

[ T = 10 \times 1000 = 10000 ]

So,

[ TS_1 = 10000 + 10000 + 10000 = 30000 ]

Deadweight loss:

[ DWL = TS_0 - TS_1 ]

[ DWL = 31250 - 30000 = 1250 ]

Alternatively,

[ DWL = \frac{1}{2} \times \text{Tax} \times (Q_0 - Q_1) ]

[ DWL = \frac{1}{2} \times 10 \times (1250 - 1000) ]

[ DWL = \frac{1}{2} \times 10 \times 250 = 1250 ]

Deadweight loss = Rs 1,250

---

e. Tax Revenue of the Government

[ \text{Tax Revenue} = \text{Tax per unit} \times \text{Quantity sold after tax} ]

[ = 10 \times 1000 ]

[ = 10000 ]

Tax revenue = Rs 10,000

---

Final Results

Thus, the Rs 10 per unit tax raises the consumer price, lowers the producer price, reduces quantity traded, and creates a deadweight loss of Rs 1,250.

Production Function

A production function is the technological relationship between inputs (factors of production) and the maximum output that can be produced from them in a given period, at a given state of technology. It is written as:

$Q = f(L, K)$

where $Q$ is output, $L$ is labour and $K$ is capital. It shows the technically efficient combinations of inputs — the maximum output obtainable from each input bundle. Its two central tools are returns to scale (the response of output when all inputs change proportionally) and the marginal rate of technical substitution (the rate at which one input substitutes for another along an isoquant).

CES Production Function and Its Properties

The Constant Elasticity of Substitution (CES) production function, developed by Arrow, Chenery, Minhas and Solow (1961), is:

$Q = A\left[\alpha K^{-\rho} + (1-\alpha)L^{-\rho}\right]^{-\frac{1}{\rho}}$

where $A > 0$ is the efficiency parameter, $\alpha$ ($0<\alpha<1$) the distribution parameter, $\rho > -1$ ($\rho \neq 0$) the substitution parameter, and the elasticity of substitution is $\sigma = \dfrac{1}{1+\rho}$.

Properties:

1. Constant elasticity of substitution: $\sigma = \frac{1}{1+\rho}$ is constant along the whole isoquant, though not necessarily equal to one. This is its defining feature.
2. Linear homogeneity (constant returns to scale): In the above form the function is homogeneous of degree one, so scaling all inputs by $t$ scales output by $t$.
3. Positive but diminishing marginal products: $MP_L > 0$, $MP_K > 0$, with $\partial MP/\partial L < 0$ and $\partial MP/\partial K < 0$ (for $\rho > -1$).
4. Convex, downward-sloping isoquants: The marginal rate of technical substitution diminishes as one moves along an isoquant.
5. Generalises other functions: As $\rho \to 0$ ($\sigma \to 1$) it reduces to the Cobb–Douglas form; as $\rho \to \infty$ ($\sigma \to 0$) to the Leontief fixed-proportion form; as $\rho \to -1$ ($\sigma \to \infty$) to the linear (perfect substitutes) form.
6. Factor intensity governed by $\alpha$: The ratio of the marginal products, and hence relative factor shares, depends on $\alpha$ and on the capital–labour ratio.
7. Non-constant factor shares: Unlike Cobb–Douglas, relative income shares of labour and capital vary with the input ratio unless $\sigma = 1$.

How CES Differs from Cobb–Douglas

Grounds of difference in brief: The CES function relaxes the restrictive Cobb–Douglas assumption of unitary elasticity of substitution. Empirically, $\sigma$ is often found to differ from one (e.g. less than one in many industries), and CES captures this while Cobb–Douglas cannot. CES is therefore a more general and flexible specification, of which Cobb–Douglas is a special case ($\rho = 0$). The cost of this generality is greater estimation complexity and the loss of the simple, constant factor-share interpretation.

CES (constant elasticity of substitution) production function

Arrow, Chenery, Minhas and Solow (1961) proposed $Q=A\left[\delta K^{-\rho}+(1-\delta)L^{-\rho}\right]^{-\nu/\rho},\qquad \rho\ge-1,\ \rho\ne0,$ where $A>0$ is the efficiency parameter, $0<\delta<1$ the distribution parameter, $\rho$ the substitution parameter and $\nu$ the returns-to-scale parameter. (With $\nu=1$ the exponent is $-1/\rho$.)

Properties

1. Constant elasticity of substitution: $\sigma=\dfrac{1}{1+\rho}$, constant at every point of the isoquant and for every input ratio.
2. Homogeneity: it is homogeneous of degree $\nu$. With $\nu=1$ it shows constant returns to scale; $\nu>1$ gives increasing and $\nu<1$ decreasing returns.
3. Positive, diminishing marginal products: $MP_K,MP_L>0$ with $MP$ falling, so isoquants are downward-sloping and convex.
4. Factor shares depend on $\sigma$ and the input ratio, unlike Cobb–Douglas where shares are constant.
5. Marginal rate of technical substitution: $MRTS=\dfrac{\delta}{1-\delta}\left(\dfrac{L}{K}\right)^{1+\rho}$.
6. Nests several functions as special cases:
   • $\rho\to0$ ($\sigma=1$): Cobb–Douglas;
   • $\rho\to\infty$ ($\sigma=0$): Leontief fixed proportions;
   • $\rho=-1$ ($\sigma=\infty$): linear function with perfect substitutes.

How it differs from Cobb–Douglas

Given

[ Q_d=1800-8P,\qquad Q_s=-200+4P ]

---

i) Equilibrium price and quantity

At equilibrium: [ Q_d=Q_s ] [ 1800-8P=-200+4P ] [ 2000=12P ] [ P^*=\frac{500}{3}=166.67 ]

Quantity: [ Q^=-200+4\left(\frac{500}{3}\right) ] [ Q^=\frac{1400}{3}=466.67 ]

Equilibrium price = Rs 166.67; equilibrium quantity = 466.67 units.

---

ii) Consumer’s surplus and producer’s surplus

Inverse demand: [ P=225-\frac{Q}{8} ] Maximum demand price = Rs 225.

Consumer’s surplus: [ CS=\frac12(225-P^)Q^ ] [ CS=\frac12\left(225-\frac{500}{3}\right)\left(\frac{1400}{3}\right) ] [ CS=\frac12\left(\frac{175}{3}\right)\left(\frac{1400}{3}\right) ] [ CS=\frac{122500}{9}=13611.11 ]

Inverse supply: [ P=50+\frac{Q}{4} ] Minimum supply price = Rs 50.

Producer’s surplus: [ PS=\frac12(P^-50)Q^ ] [ PS=\frac12\left(\frac{500}{3}-50\right)\left(\frac{1400}{3}\right) ] [ PS=\frac12\left(\frac{350}{3}\right)\left(\frac{1400}{3}\right) ] [ PS=\frac{245000}{9}=27222.22 ]

Consumer’s surplus = Rs 13,611.11; producer’s surplus = Rs 27,222.22.

---

iii) Impact of Rs 12 per unit tax

Let consumer price be (P_d) and producer price be (P_s). With tax: [ P_d=P_s+12 ]

Demand: [ Q_d=1800-8P_d ]

Supply: [ Q_s=-200+4P_s ]

At new equilibrium: [ 1800-8(P_s+12)=-200+4P_s ] [ 1800-8P_s-96=-200+4P_s ] [ 1704-8P_s=-200+4P_s ] [ 1904=12P_s ] [ P_s=\frac{476}{3}=158.67 ]

Consumer price: [ P_d=P_s+12=\frac{476}{3}+12=\frac{512}{3}=170.67 ]

New quantity: [ Q_t=-200+4\left(\frac{476}{3}\right) ] [ Q_t=\frac{1304}{3}=434.67 ]

After tax:

• Price paid by consumers rises to Rs 170.67
• Price received by producers falls to Rs 158.67
• Quantity falls from 466.67 units to 434.67 units

Tax burden: [ \text{Consumer burden per unit}=170.67-166.67=Rs\ 4 ] [ \text{Producer burden per unit}=166.67-158.67=Rs\ 8 ]

So consumers bear (1/3) and producers bear (2/3) of the tax.

---

iv) Consumer surplus, producer surplus and deadweight loss after tax

After tax: [ Q_t=\frac{1304}{3} ]

Consumer surplus: [ CS_t=\frac12\left(225-\frac{512}{3}\right)\left(\frac{1304}{3}\right) ] [ CS_t=\frac12\left(\frac{163}{3}\right)\left(\frac{1304}{3}\right) ] [ CS_t=\frac{106276}{9}=11808.44 ]

Producer surplus: [ PS_t=\frac12\left(\frac{476}{3}-50\right)\left(\frac{1304}{3}\right) ] [ PS_t=\frac12\left(\frac{326}{3}\right)\left(\frac{1304}{3}\right) ] [ PS_t=\frac{212552}{9}=23616.89 ]

Deadweight loss: [ DWL=\frac12 \times \text{tax} \times (Q^*-Q_t) ] [ DWL=\frac12 \times 12 \times \left(\frac{1400}{3}-\frac{1304}{3}\right) ] [ DWL=\frac12 \times 12 \times 32 ] [ DWL=192 ]

After tax:

• Consumer surplus = Rs 11,808.44
• Producer surplus = Rs 23,616.89
• Deadweight loss = Rs 192

---

v) Tax revenue of the government

[ \text{Tax revenue}=\text{Tax per unit}\times Q_t ] [ =12\times \frac{1304}{3} ] [ =4\times 1304 ] [ =5216 ]

Government tax revenue = Rs 5,216.

---

Final Results

Hypothetico-Deductive Method in Economics

The hypothetico-deductive method is the standard scientific method used in economics (and other social sciences) in which a theory is built by deducing testable conclusions from a set of initial assumptions or hypotheses. It proceeds in a logical sequence: assumptions → deductions → predictions → verification → acceptance or rejection of the hypothesis. It is also called the deductive method or scientific method in economics.

Steps of the Method

1. Observation and identification of the problem – The researcher observes an economic phenomenon (e.g., rising inflation) and defines the problem clearly.
2. Formulation of hypotheses – Tentative explanations are framed. For example: "An increase in money supply causes inflation."
3. Specification of assumptions – Simplifying assumptions are made (e.g., ceteris paribus, rational consumers, perfect competition) so that the analysis remains manageable.
4. Deduction of implications – Logical conclusions are drawn from the assumptions. If money supply rises and output is constant, then the price level must rise (Quantity Theory of Money: MV = PT).
5. Testing/verification – The deduced predictions are tested against real-world data using econometric techniques, statistical tools, or historical evidence.
6. Acceptance, rejection or modification – If facts support the hypothesis, it becomes a theory or law; if not, it is rejected or revised.

Examination of the Method

Merits:

• It is logical and systematic, giving economics the character of a positive science.

• It allows prediction and policy formulation (e.g., monetary and fiscal policy).
• It is objective, as conclusions follow from assumptions rather than personal bias.
• It permits controlled reasoning through ceteris paribus, which is difficult in laboratory experiments.

Demerits:

• Economic phenomena are complex and interdependent, so isolating variables is difficult.
• Human behaviour is not fully rational or predictable, unlike physical particles.
• Assumptions may be unrealistic (e.g., perfect competition), reducing the validity of deductions.
• Verification is difficult because controlled experiments are rarely possible in economics.
• Conclusions are often conditional and may not hold across different times and countries.

Conclusion

The hypothetico-deductive method provides economics with a rigorous, logical framework for theory building and testing. However, because of the complexity of human and social behaviour, its assumptions and predictions must be applied with caution. It is best used in combination with the inductive and historical methods to make economic analysis more realistic and reliable.

Duality in consumer theory refers to the existence of two equivalent ways of describing the same consumer behaviour: the primal problem (maximising utility subject to a budget constraint) and the dual problem (minimising expenditure subject to a given utility level). Both approaches describe the same underlying preferences and yield the same optimal choices, so either can be used to analyse consumer behaviour.

The two problems

Primal problem: $\max_{x_1, x_2} U = U(x_1, x_2) \quad \text{subject to} \quad p_1 x_1 + p_2 x_2 = M$

This gives the Marshallian (ordinary) demand functions $x_1^m(p_1, p_2, M)$ and $x_2^m(p_1, p_2, M)$, and the indirect utility function $V(p_1, p_2, M)$, which shows the maximum utility attainable at given prices and income.

Dual problem: $\min_{x_1, x_2} p_1 x_1 + p_2 x_2 \quad \text{subject to} \quad U(x_1, x_2) = \bar{U}$

This gives the Hicksian (compensated) demand functions $x_1^h(p_1, p_2, \bar{U})$ and $x_2^h(p_1, p_2, \bar{U})$, and the expenditure function $E(p_1, p_2, \bar{U})$, which shows the minimum expenditure needed to reach utility level $\bar{U}$.

Key duality relationships

1. Inverse relationship between the value functions: $E(p_1, p_2, \bar{U}) = M \iff V(p_1, p_2, M) = \bar{U}$ That is, $E$ and $V$ are inverses of each other in the income/utility argument.

2. Shephard's Lemma (from the expenditure function): $\frac{\partial E}{\partial p_i} = x_i^h(p_1, p_2, \bar{U})$

3. Roy's Identity (from the indirect utility function): $x_i^m(p_1, p_2, M) = -\frac{\partial V/\partial p_i}{\partial V/\partial M}$

4. Slutsky equation links the two demand systems: $\frac{\partial x_i^m}{\partial p_j} = \underbrace{\frac{\partial x_i^h}{\partial p_j}}_{\text{substitution effect}} - \underbrace{x_j^m \frac{\partial x_i^m}{\partial M}}_{\text{income effect}}$

Significance

• It provides two equivalent analytical routes to the same result; whichever is mathematically convenient can be chosen.
• It allows welfare measurement (compensating and equivalent variations) through the expenditure function.
• It establishes the integrability of demand functions: from observed Marshallian demand, one can recover preferences.
• It is the foundation of consumer surplus, price indices, and welfare economics.

In short, duality means that utility maximisation and expenditure minimisation are mirror images of each other, and the Marshallian and Hicksian demand functions, along with $V(\cdot)$ and $E(\cdot)$, contain the same information about consumer preferences.

The hypothetico-deductive method is the standard methodology of positive economics. A hypothesis is proposed, testable implications are deduced from it, and these are checked against observed data. The hypothesis is then accepted, rejected or revised. A hypothesis is a tentative proposition, and Friedman argued it should be judged by the accuracy of its predictions rather than the realism of its assumptions.

Steps

1. Observation and hypothesis. A regularity is noticed and a causal hypothesis is stated.
2. Deduction. A specific, testable prediction is logically derived from the hypothesis plus stated initial conditions.
3. Testing. The prediction is compared with data using statistical or econometric methods.
4. Verification or falsification. If the prediction holds, the hypothesis is provisionally accepted. If it fails, it is rejected or modified. Popper stressed that a hypothesis can never be finally proved, only falsified.

Example: the law of demand

• Hypothesis: Other things equal, the quantity demanded of a good varies inversely with its price.
• Deduction: If the price of rice in Kathmandu rises from Rs 60 to Rs 80 per kg, quantity demanded will fall.
• Testing: Using household survey or market data, estimate $Q_d = a - bP$. The hypothesis predicts $b > 0$, that is, a negative slope of the demand curve.
• Decision: If $\hat{b}$ is positive and statistically significant, the hypothesis is supported. If $\hat{b}$ is zero or negative, so that quantity rises with price, the hypothesis is contradicted. The ceteris paribus assumption or the model specification is then re-examined.

Merits and limitations

• It gives economics a scientific and testable character and separates positive from normative statements.
• It supports prediction and policy evaluation.
• Controlled experiments are rare, data are non-experimental, ceteris paribus seldom holds, and value judgements can enter the choice of problems and variables.

Duality in Consumer Theory

Meaning. Duality is the idea that a consumer's choice problem can be looked at in two equivalent ways — either as maximising utility subject to a given budget, or as minimising expenditure to reach a given level of utility. Both views describe the same underlying preferences and yield the same optimal choices. The first is the primal problem; the second is the dual problem.

The primal problem (utility maximisation).

$\max_{x_1, x_2} U = U(x_1, x_2) \quad \text{subject to} \quad p_1x_1 + p_2x_2 = M$

Its solution gives the Marshallian (ordinary) demand functions $x_i^m = x_i(p_1, p_2, M)$ and the indirect utility function $V(p_1, p_2, M)$, which shows the maximum utility attainable at given prices and income.

The dual problem (expenditure minimisation).

$\min_{x_1, x_2} E = p_1x_1 + p_2x_2 \quad \text{subject to} \quad U(x_1, x_2) = \bar{U}$

Its solution gives the Hicksian (compensated) demand functions $x_i^h = x_i(p_1, p_2, \bar{U})$ and the expenditure function $E(p_1, p_2, \bar{U})$, the minimum money needed to attain utility $\bar{U}$.

The dual relationship. The two problems are mirror images:

$V(p_1, p_2, M) = \bar{U} \iff E(p_1, p_2, \bar{U}) = M$

That is, the indirect utility function and the expenditure function are inverses of each other. Substituting one into the other recovers the original variable:

$E\big(p_1, p_2, V(p_1, p_2, M)\big) = M, \qquad V\big(p_1, p_2, E(p_1, p_2, \bar{U})\big) = \bar{U}$

Recovering demands. The two demand systems are linked by the Slutsky equation:

$\frac{\partial x_i^m}{\partial p_j} = \underbrace{\frac{\partial x_i^h}{\partial p_j}}_{\text{substitution effect}} - \underbrace{x_j^m \frac{\partial x_i^m}{\partial M}}_{\text{income effect}}$

Also, by Roy's identity, Marshallian demand is obtained from the indirect utility function:

$x_i^m = -\frac{\partial V/\partial p_i}{\partial V/\partial M}$

and by Shephard's lemma, Hicksian demand is obtained from the expenditure function:

$x_i^h = \frac{\partial E}{\partial p_i}$

Why duality matters. It allows the same preference structure to be analysed from whichever side is convenient — for example, welfare and price-index problems are easier with the expenditure function, while market demand is easier with the Marshallian function. It also provides the theoretical basis for measuring compensating and equivalent variations in welfare economics.

Let $u = x_1^{\alpha}x_2^{\beta}$ with $\alpha+\beta=1$, and let $\bar u$ be the target utility.

(a) Hicksian demands

$\min_{x_1,x_2}\; p_1x_1+p_2x_2 \quad \text{s.t.}\quad x_1^{\alpha}x_2^{\beta}=\bar u$

$\mathcal L = p_1x_1+p_2x_2+\mu\left(\bar u - x_1^{\alpha}x_2^{\beta}\right)$

First-order conditions:

$p_1 = \mu\alpha\frac{u}{x_1},\qquad p_2=\mu\beta\frac{u}{x_2}\;\Rightarrow\;\frac{p_1}{p_2}=\frac{\alpha x_2}{\beta x_1}\;\Rightarrow\;x_2=\frac{\beta p_1}{\alpha p_2}x_1$

Substituting in the constraint, $\bar u = x_1\left(\dfrac{\beta p_1}{\alpha p_2}\right)^{\beta}$, so

$h_1(p,\bar u)=\bar u\left(\frac{\alpha p_2}{\beta p_1}\right)^{\beta},\qquad h_2(p,\bar u)=\bar u\left(\frac{\beta p_1}{\alpha p_2}\right)^{\alpha}$

(b) Expenditure function

$E(p,\bar u)=p_1h_1+p_2h_2=\bar u\,p_1^{\alpha}p_2^{\beta}\left[\alpha^{\beta}\beta^{-\beta}+\beta^{\alpha}\alpha^{-\alpha}\right]$

Since $\alpha^{\beta}\beta^{-\beta}=\alpha\cdot\alpha^{-\alpha}\beta^{-\beta}$ and $\beta^{\alpha}\alpha^{-\alpha}=\beta\cdot\alpha^{-\alpha}\beta^{-\beta}$, the bracket equals $(\alpha+\beta)\alpha^{-\alpha}\beta^{-\beta}=\alpha^{-\alpha}\beta^{-\beta}$. Therefore

$E(p,\bar u)=\bar u\left(\frac{p_1}{\alpha}\right)^{\alpha}\left(\frac{p_2}{\beta}\right)^{\beta}$

(c) Indirect utility function

Set $E(p,V)=M$ and solve for utility:

$V(p,M)=\frac{M}{\left(\dfrac{p_1}{\alpha}\right)^{\alpha}\left(\dfrac{p_2}{\beta}\right)^{\beta}}=M\,\alpha^{\alpha}\beta^{\beta}\,p_1^{-\alpha}p_2^{-\beta}$

(d) Marshallian demands

Use $x_i(p,M)=h_i\big(p,V(p,M)\big)$:

$x_1=V\left(\frac{\alpha p_2}{\beta p_1}\right)^{\beta}=M\alpha^{\alpha}\beta^{\beta}p_1^{-\alpha}p_2^{-\beta}\cdot\alpha^{\beta}\beta^{-\beta}p_2^{\beta}p_1^{-\beta}=\frac{\alpha M}{p_1}$

and similarly

$x_2=\frac{\beta M}{p_2}$

(Roy's identity, $x_1=-\dfrac{\partial V/\partial p_1}{\partial V/\partial M}=\dfrac{\alpha M}{p_1}$, gives the same result.) The consumer spends the fixed shares $\alpha$ and $\beta$ of income on the two goods.

Definition of Production Functions

Homogeneous Production Function

A production function $Q = f(L, K)$ is said to be homogeneous of degree $n$ if, when all inputs are multiplied by a positive constant $t$, output increases by $t^n$:

$f(tL, tK) = t^n f(L, K)$

If $n = 1$, the function is linearly homogeneous (constant returns to scale); if $n > 1$, increasing returns to scale; if $n < 1$, decreasing returns to scale. Homogeneity is a purely mathematical property that permits the use of Euler's theorem to explain factor payments.

Linear Production Function

A linear production function expresses output as a linear combination of inputs:

$Q = aL + bK$

where $a$ and $b$ are positive constants representing the marginal products of labour and capital. It is homogeneous of degree one. The isoquants are straight lines, and the marginal rate of technical substitution (MRTS) is constant ($b/a$), implying perfect substitutability between inputs. Factors are perfect substitutes, so the producer may use only one factor.

Leontief Production Function

The Leontief (fixed-proportions) production function assumes inputs must be combined in rigidly fixed ratios:

$Q = \min\left(\frac{L}{a}, \frac{K}{b}\right)$

where $a$ and $b$ are the required input coefficients per unit of output. It is homogeneous of degree one. The isoquants are L-shaped (right-angled), and the elasticity of substitution is zero — inputs are perfect complements with no substitutability. Any extra amount of one input beyond the required ratio adds nothing to output.

Major Properties of the Cobb-Douglas Production Function

The Cobb-Douglas production function is:

$Q = A L^{\alpha} K^{\beta}$

where $A > 0$ is the efficiency parameter and $\alpha, \beta$ are output elasticities. Its major properties are:

1. Homogeneity of degree $(\alpha + \beta)$: $f(tL, tK) = t^{\alpha+\beta} f(L, K)$. If $\alpha + \beta = 1$, returns to scale are constant; if $> 1$, increasing; if $< 1$, decreasing.

2. Constant output elasticities: The elasticity of output with respect to labour is $\alpha$ and with respect to capital is $\beta$, both constant. That is, $\frac{\partial Q}{\partial L}\cdot\frac{L}{Q} = \alpha$ and $\frac{\partial Q}{\partial K}\cdot\frac{K}{Q} = \beta$.

3. Positive but diminishing marginal products: $MP_L = \alpha A L^{\alpha-1}K^{\beta} > 0$ and $\frac{\partial MP_L}{\partial L} < 0$ (similarly for capital), provided $0 < \alpha, \beta < 1$.

4. Unit elasticity of substitution: The elasticity of substitution between labour and capital is exactly one ($\sigma = 1$).

5. Factor shares are constant: Under perfect competition, labour's share of output is $\alpha$ and capital's share is $\beta$. The function is thus consistent with the observed stability of relative factor shares.

6. Euler's theorem (for $\alpha + \beta = 1$): Output is exactly exhausted by factor payments: $Q = L\frac{\partial Q}{\partial L} + K\frac{\partial Q}{\partial K}$.

7. Log-linear form: Taking logarithms gives $\ln Q = \ln A + \alpha \ln L + \beta \ln K$, which is linear in parameters and easily estimable by ordinary least squares.

8. Isoquants are smooth, convex, and negatively sloped, with MRTS diminishing along an isoquant.

Difference Between Cobb-Douglas and CES Production Functions

The CES (Constant Elasticity of Substitution) production function is:

$Q = A\left[\delta L^{-\rho} + (1-\delta)K^{-\rho}\right]^{-\frac{1}{\rho}}, \quad \sigma = \frac{1}{1+\rho}$

Conclusion: The Cobb-Douglas function is a convenient special case of the CES function with $\sigma = 1$. The CES function is superior in flexibility because it allows the elasticity of substitution to differ from unity, but it is harder to estimate and interpret.

Rationalism and Empiricism in Economic Methodology

1. Meaning of the two positions

Rationalism is the epistemological view that reason is the primary source and ultimate test of knowledge. In economic methodology, a rationalist approach holds that economic knowledge is derived by logical deduction from self-evident axioms or first principles about human behaviour, and that the conclusions so derived are valid independently of (or prior to) empirical observation. Empiricism, by contrast, holds that all substantive knowledge comes from experience — observation, measurement and experiment — and that claims which cannot be tested against observed facts are empty of scientific content.

2. Basis of knowledge

• Rationalism: Knowledge is innate or a priori. The starting point is a set of axioms, e.g. "human beings act purposively to maximise their objectives subject to constraints." From such axioms, theorems are derived by pure logic. The classical and Austrian traditions illustrate this: Lionel Robbins defined economics as "the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses," a definition resting on logical deduction rather than measurement. Ludwig von Mises's praxeology is the purest rationalist programme — economic laws are held to be logically necessary truths about human action, not empirical generalisations.
• Empiricism: Knowledge is a posteriori. The starting point is data — prices, quantities, incomes, employment. The method is induction: regularities are observed and generalised into laws. The German Historical School (Roscher, Schmoller) and American institutionalism (Veblen, Mitchell) exemplify this, insisting that economic laws must be derived from historical and statistical records of actual economies rather than from abstract axioms. Modern applied economics — randomised controlled trials, natural experiments, national accounts — is empiricist in spirit.

3. Method: deduction versus induction

Rationalism proceeds deductively: axioms → logical inference → prediction. The prediction is a theorem whose truth follows from the axioms, so falsification of the prediction is taken to indicate an error in the auxiliary assumptions rather than in the axioms themselves. Empiricism proceeds inductively: observation → pattern → hypothesis → further observation. For the empiricist, a proposition is meaningful only if it can be confronted with evidence; for the rationalist, logical consistency and coherence with the axioms are the criteria of validity.

4. Role of assumptions and realism

Rationalism tolerates "unrealistic" assumptions, since what matters is the predictive or logical entailment of the model. Milton Friedman's "as if" methodology is a bridge here: he argued that the realism of assumptions is irrelevant and only predictive success counts — a position that borrows the rationalist's deductive structure while submitting it to empirical testing. Empiricism, on the other hand, demands that assumptions be observationally grounded; hence the institutionalist critique that the perfectly competitive model assumes away the very institutions (firms, contracts, customs) that need explanation.

5. Examples

• Rationalist example: The law of demand. Given the axiom of diminishing marginal utility and rational choice, the inverse price–quantity relation is deduced without any need to observe a single market. Similarly, the Ricardian theory of comparative advantage is derived from a simple two-country, two-commodity model by logic alone.
• Empiricist example: The estimation of the consumption function. Kuznets's data showed that the average propensity to consume was stable in the long run while falling in the short run — a finding that could not be deduced from Keynes's psychological law and forced the permanent-income and life-cycle revisions. Similarly, the Phillips curve was accepted as a stable empirical regularity until the stagflation of the 1970s falsified it, illustrating the empiricist's insistence on testing.

6. Strengths and limitations

Rationalism gives economics its deductive rigour, generality and internal consistency, but risks producing elegant models with no contact with reality. Empiricism keeps economics accountable to facts and generates testable knowledge, but induction can never establish a law with certainty (the problem of induction), and data are often inadequate, non-experimental and subject to measurement error. Pure empiricism also struggles to distinguish correlation from causation without theoretical structure.

7. Conclusion

The two positions are best seen as complementary rather than mutually exclusive. Modern economic methodology — from Samuelson's operationalism to the credibility revolution in applied microeconomics — combines rationalist deduction (theory and formal models) with empiricist testing (identification and data). Theory supplies the hypotheses; evidence decides between them. Neither alone can produce reliable economic knowledge.

Falsification in Economics

Falsification is a criterion of scientific method, associated mainly with the philosopher Karl Popper, which holds that a statement or theory is scientific only if it is capable of being refuted (proved false) by empirical evidence. A theory that cannot, even in principle, be contradicted by any conceivable observation is, on this view, not scientific but metaphysical.

In economics, falsification means that an economic hypothesis or theory must generate testable predictions that can be checked against real-world data. If observation contradicts the prediction, the theory is rejected or revised; if the prediction survives repeated testing, the theory is corroborated (though never finally "proved" true). For example, the hypothesis "a rise in the price of a good, other things equal, reduces the quantity demanded" is falsifiable because data on prices and quantities could, in principle, show the opposite. By contrast, a claim such as "all economic behaviour is ultimately driven by invisible psychological forces" cannot be tested and is therefore non-falsifiable.

Which method of reasoning is based on falsification?

Falsification is based on the deductive method of reasoning. In the deductive (hypothetico-deductive) approach, the economist starts from general premises or assumptions, derives a specific, testable conclusion (hypothesis) by logical deduction, and then confronts that conclusion with empirical facts. Falsification operates at this testing stage: a single observation inconsistent with the deduced prediction is enough to falsify the general theory. Thus deduction supplies the testable proposition, and falsification judges it.

The inductive method, by contrast, moves from particular observations to general conclusions. It relies on verification — accumulating confirming instances — rather than on refutation. Since no finite number of confirming cases can logically establish a universal law (the problem of induction), induction cannot provide a decisive test; hence falsification is not its basis. Falsification belongs, therefore, to the deductive (hypothetico-deductive) method.

Fit the linear cost function $Y = a + bX$ by least squares, where $X$ is output and $Y$ is cost.

With $n=9$, the normal equations are

$\sum Y = na + b\sum X \;\Rightarrow\; 337 = 9a + 214b$

$\sum XY = a\sum X + b\sum X^2 \;\Rightarrow\; 8906 = 214a + 5868b$

Solving:

$b = \frac{9(8906) - 214(337)}{9(5868) - 214^2} = \frac{80154 - 72118}{52812 - 45796} = \frac{8036}{7016} \approx 1.1454$

$a = \frac{337 - 214(1.1454)}{9} \approx 10.21$

The estimated cost function is $\hat Y = 10.21 + 1.1454X$.

At $X = 52$:

$\hat Y = 10.21 + 1.1454(52) \approx 10.21 + 59.56 = 69.77$

The estimated cost of producing 52 units is about 69.77 cost units. Since 52 lies outside the observed range (10 to 40), this is an extrapolation and should be treated with caution.

Chamberlin's Model of Monopolistic Competition

Edward Hastings Chamberlin presented his theory of monopolistic competition in his book The Theory of Monopolistic Competition (1933). It explains a market structure that lies between perfect competition and monopoly — a blend of both, hence the name "monopolistic competition."

Meaning and Features

Monopolistic competition is a market where many sellers sell differentiated but close substitute products. Its main features are:

1. Large number of buyers and sellers — each firm is small relative to the market, so its price–output decision does not provoke retaliation from rivals.
2. Product differentiation — the core of the model. Firms differentiate through brand name, quality, design, packaging, location, or service. This gives each firm a monopoly over its own product, hence a downward-sloping demand curve.
3. Close substitutes — because products are similar, each firm faces competition from rivals, so demand is highly elastic.
4. Free entry and exit — in the long run, firms can enter or leave the "product group," eliminating abnormal profit or loss.
5. Selling costs — advertising, sales promotion and other selling costs become important tools of competition.
6. Imperfect knowledge of prices and quality among buyers.

Short-Run Equilibrium

In the short run, each firm behaves like a monopolist on its own demand curve. It maximizes profit where MR = MC, with MC rising. The firm may earn supernormal profit (if AR > AC), normal profit (AR = AC), or incur loss (if AR < AC). Since entry is not yet complete, abnormal profit can persist in the short run.

Long-Run Equilibrium

Abnormal profit attracts new firms into the product group. As new entrants offer close substitutes, the demand curve of each existing firm shifts leftward (and becomes more elastic). This process continues until excess profit is wiped out, i.e., until the firm's demand curve becomes tangent to its average cost curve.

The long-run equilibrium conditions are:

• MR = MC (profit maximization)
• AR (P) = AC (zero abnormal profit, since free entry)
• The demand curve is tangent to the AC curve.

Because the demand curve is downward-sloping, tangency occurs to the left of the minimum point of the AC curve. Hence:

$P > MC \quad \text{and} \quad P = AC > \text{minimum AC}$

Excess Capacity Theorem

At equilibrium the firm produces less than the least-cost output (the output at minimum AC). The gap between the least-cost output and the actual output is excess capacity. This is the "excess capacity theorem" of Chamberlin — a direct result of product differentiation and downward-sloping demand. Firms operate with spare capacity, charge a price above marginal cost, and do not produce at the lowest possible cost.

Diagrammatic Illustration

In the long-run diagram:

• The DD (AR) curve is tangent to the AC curve at point T.
• The MR curve cuts the MC curve at the same output level OQ.
• Price OP equals average cost QT, so profit is zero.
• Since OQ < OQ* (least-cost output), excess capacity Q*Q exists.

Critical Evaluation

Merits:

• Realistic — reflects actual markets (toothpaste, soap, restaurants, clothing).
• Explains product differentiation, advertising and selling costs, which perfect competition ignores.
• Shows that firms may operate with excess capacity yet survive.

Demerits:

• The concept of a "product group" is vague and difficult to define.
• Tangency solution is indeterminate — the exact point of tangency cannot be uniquely determined.
• Selling costs are treated vaguely; their effect on demand is hard to measure.
• It is more a descriptive than an analytical model.

Conclusion

Chamberlin's model is a major contribution to microeconomic theory. It bridges the gap between the unrealistic assumptions of perfect competition and the extreme case of monopoly, and provides a realistic framework for analysing markets where firms compete through product differentiation and selling efforts rather than price alone.

a) Positivism and Constructivism

Positivism is a philosophical position, associated with Auguste Comte and later with the Vienna Circle, which holds that genuine knowledge can only come from observable, measurable facts and the logical analysis of them. Its core assumptions are:

• Ontology: There is a single, objective reality that exists independently of the observer (realism).
• Epistemology: The researcher can study this reality from outside, as an objective outsider; knowledge is discovered, not created.
• Methodology: It favours quantitative methods — hypothesis testing, statistical inference, controlled experiments — aiming at general laws, prediction and causal explanation. Values and meanings are excluded as unscientific.

Constructivism (interpretivism) rejects this. Its assumptions are:

• Ontology: Reality is socially constructed; there are multiple, subjective realities shaped by experience, language and culture (relativism).
• Epistemology: The researcher cannot stand outside the subject matter; knowledge is co-created through interaction with participants.
• Methodology: It favours qualitative methods — interviews, observation, case studies — aiming at understanding (Verstehen) meanings and context rather than universal laws.

In economics, positivism underlies neoclassical modelling, econometrics and the fact–value distinction of Friedman's positive economics. Constructivism underlies heterodox and institutional approaches, feminist economics and much development fieldwork. The two differ fundamentally on whether economic reality is a single measurable system or a socially constructed one. Most modern practice is mixed-method, accepting elements of both.

---

b) Consumption Duality

Consumption duality is the proposition in consumer theory that the consumer's utility-maximisation problem and the expenditure-minimisation problem are two equivalent ways of describing the same behaviour. The two problems are "dual" to each other.

Primal problem — maximise utility subject to a budget constraint:

$\max_{x_1,x_2} U(x_1,x_2) \quad \text{s.t.} \quad p_1x_1 + p_2x_2 = M$

This yields Marshallian (ordinary) demand functions $x_i = x_i(p_1,p_2,M)$ and the indirect utility function $V(p_1,p_2,M)$.

Dual problem — minimise expenditure to achieve a given utility level:

$\min_{x_1,x_2} p_1x_1 + p_2x_2 \quad \text{s.t.} \quad U(x_1,x_2) = \bar{U}$

This yields Hicksian (compensated) demand functions $h_i = h_i(p_1,p_2,\bar{U})$ and the expenditure function $E(p_1,p_2,\bar{U})$.

Key duality results:

1. Identity: $E\big(p_1,p_2,V(p_1,p_2,M)\big) = M$ and $V\big(p_1,p_2,E(p_1,p_2,\bar{U})\big) = \bar{U}$.
2. Shephard's lemma: $\dfrac{\partial E}{\partial p_i} = h_i(p_1,p_2,\bar{U})$.
3. Roy's identity: $x_i(p_1,p_2,M) = -\dfrac{\partial V/\partial p_i}{\partial V/\partial M}$.
4. Slutsky equation: $\dfrac{\partial x_i}{\partial p_j} = \underbrace{\dfrac{\partial h_i}{\partial p_j}}_{\text{substitution effect}} - \underbrace{x_j\dfrac{\partial x_i}{\partial M}}_{\text{income effect}}$.

Significance: Duality lets us recover preferences from observed demand (or expenditure) data, decompose price effects into income and substitution effects, and derive welfare measures — compensating variation and equivalent variation — from the expenditure function. It is the foundation of applied welfare economics and consumer surplus analysis.

Compensated vs Uncompensated Demand Curve: Price Decomposition Method

1. The Price Decomposition (Slutsky) Identity

When the price of a good $x$ changes, the total effect on quantity demanded splits into two parts:

$\underbrace{\frac{\partial x}{\partial p_x}}_{\text{Total (Marshallian)}} = \underbrace{\frac{\partial x}{\partial p_x}\Big|_{U=\bar{U}}}_{\text{Substitution effect}} \;-\; \underbrace{x\,\frac{\partial x}{\partial M}}_{\text{Income effect}}$

where $M$ is money income. The uncompensated (Marshallian) demand curve $x(p_x, M)$ reflects the total effect; the compensated (Hicksian) demand curve $x^c(p_x, \bar{U})$ reflects only the substitution effect, because income is adjusted to keep utility constant.

2. Slopes of the Two Curves

Compensated (Hicksian) curve: By the fundamental result of consumer theory, the substitution effect is always negative:

$\frac{\partial x^c}{\partial p_x} = \frac{\partial x}{\partial p_x}\Big|_{U=\bar{U}} < 0$

So the compensated demand curve is downward sloping.

Uncompensated (Marshallian) curve: For a normal good, $\dfrac{\partial x}{\partial M} > 0$, so the income-effect term $-x\,\dfrac{\partial x}{\partial M} < 0$ reinforces the substitution effect:

$\frac{\partial x}{\partial p_x} = \underbrace{\frac{\partial x^c}{\partial p_x}}_{(-)} \;-\; \underbrace{x\,\frac{\partial x}{\partial M}}_{(+)} \;<\; \frac{\partial x^c}{\partial p_x} < 0$

Both slopes are negative, but the Marshallian slope is more negative — i.e. larger in absolute value — than the Hicksian slope. Hence the uncompensated curve is flatter and the compensated curve is steeper.

3. Elasticity Comparison

Writing elasticities, $\varepsilon = \dfrac{\partial x}{\partial p_x}\cdot\dfrac{p_x}{x}$. Multiplying the Slutsky identity through by $p_x/x$:

$\varepsilon_{\text{uncompensated}} = \varepsilon_{\text{compensated}} - \frac{p_x}{x}\cdot x\,\frac{\partial x}{\partial M} = \varepsilon_{\text{compensated}} - \underbrace{\frac{p_x x}{M}}_{s_x}\cdot\underbrace{\frac{\partial x}{\partial M}\frac{M}{x}}_{\eta_M}$

since $\frac{p_x}{x}\cdot x\frac{\partial x}{\partial M}=p_x\frac{\partial x}{\partial M}=\frac{p_x x}{M}\cdot\frac{M}{x}\frac{\partial x}{\partial M}$.

Since both slopes are negative, taking absolute values:

$|\varepsilon_{\text{uncompensated}}| = |\varepsilon_{\text{compensated}}| + \underbrace{s_x\,\eta_M}_{>\,0}$

where $s_x = \dfrac{p_x x}{M}$ is the budget share and $\eta_M = \dfrac{\partial x}{\partial M}\cdot\dfrac{M}{x}$ is the income elasticity of demand. Since the second term is positive for a normal good,

$\boxed{|\varepsilon_{\text{uncompensated}}| > |\varepsilon_{\text{compensated}}|}$

4. Diagrammatic Illustration

 p_x
  |        S (uncompensated, flatter)
  |       /
  |      /   H (compensated, steeper)
  |     /
  |    /
  |___/________________ x

At the initial equilibrium $E_0$, both curves intersect. When $p_x$ rises to $p_x'$:

• The compensated curve moves only from $E_0$ to $E_1$ (pure substitution effect, small fall in $x$).
• The uncompensated curve moves from $E_0$ to $E_2$ (substitution plus income effect, larger fall in $x$).

Because the quantity response is smaller along the compensated curve for the same price change, it is steeper and hence less elastic.

5. Conclusion

The uncompensated demand curve is flatter (more elastic) because it absorbs both the substitution and income effects of a price change. The compensated demand curve, holding utility constant, captures only the substitution effect and is therefore steeper (less elastic). The gap between the two elasticities equals the budget-share-weighted income elasticity of demand, $s_x\eta_M$, and vanishes only when the income effect is zero (e.g., a good with zero budget share).

Given: Production function: (Q = 100 K^{0.5} L^{0.5}) Wage rate: (w = 150) Interest rate: (r = 40) Product price: (P = \text{Rs. } 2)

---

a) Marginal product of labour and capital

[ MP_L = \frac{\partial Q}{\partial L} = 100 \times 0.5 , K^{0.5} L^{-0.5} = 50 \left(\frac{K}{L}\right)^{0.5} ]

[ MP_K = \frac{\partial Q}{\partial K} = 100 \times 0.5 , K^{-0.5} L^{0.5} = 50 \left(\frac{L}{K}\right)^{0.5} ]

---

b) Efficient condition for producing 1118 units of output

The efficient (least-cost) condition requires:

[ \frac{MP_L}{MP_K} = \frac{w}{r} ]

Now,

[ \frac{MP_L}{MP_K} = \frac{50 (K/L)^{0.5}}{50 (L/K)^{0.5}} = \frac{K}{L} ]

So,

[ \frac{K}{L} = \frac{150}{40} = 3.75 ]

[ \Rightarrow K = 3.75 L ]

Also, the output constraint is:

[ Q = 100 K^{0.5} L^{0.5} = 1118 ]

[ \Rightarrow 100 \sqrt{KL} = 1118 ]

[ \Rightarrow \sqrt{KL} = 11.18 ]

[ \Rightarrow KL = (11.18)^2 \approx 125 ]

Substitute (K = 3.75 L):

[ 3.75 L^2 = 125 ]

[ L^2 = \frac{125}{3.75} = 33.33 ]

[ L \approx 5.77 ]

[ K = 3.75 \times 5.77 \approx 21.65 ]

Efficient input combination: [ \boxed{K \approx 21.65,\quad L \approx 5.77} ]

---

c) Profit and interpretation

Total Cost:

[ TC = wL + rK = 150(5.77) + 40(21.65) ]

[ TC = 865.5 + 866 = \text{Rs. } 1731.5 \approx \text{Rs. } 1732 ]

Total Revenue:

[ TR = P \times Q = 2 \times 1118 = \text{Rs. } 2236 ]

Profit:

[ \pi = TR - TC = 2236 - 1732 = \text{Rs. } 504 ]

[ \boxed{\text{Profit} \approx \text{Rs. } 504} ]

Interpretation: The firm earns a positive economic profit (supernormal profit) of about Rs. 504. Since price (P = 2) is greater than average cost (AC \approx 1.55), the firm is profitable. In a perfectly competitive market, such positive profit will attract new firms in the long run, increasing market supply and lowering price until (P = AC) and economic profit becomes zero. In the short run, the firm should continue production as long as (P \ge AVC). However, because the production function exhibits constant returns to scale (sum of exponents (= 0.5 + 0.5 = 1)), the long-run profit-maximizing output is theoretically infinite when (P > AC). Thus, the given output of 1118 units is not the profit-maximizing level; the positive profit signals an incentive for expansion or entry of new firms.

Price Discrimination under Monopoly

Meaning. Price discrimination exists when a monopolist sells the same product to different buyers, or in different markets, at different prices, where the price difference is not justified by differences in cost. It requires three conditions: (i) the seller must have monopoly power (a downward-sloping demand curve), (ii) markets must be separable — resale between buyers must be impossible or prohibitively costly (no arbitrage), and (ii) buyers must have different price elasticities of demand.

Degrees of price discrimination (Pigou's classification).

• First degree (perfect) discrimination: the monopolist charges each consumer exactly his reservation price, i.e. the maximum he is willing to pay for each unit. All consumer surplus is extracted and becomes monopoly profit. Output equals the perfectly competitive level, but distribution is highly unequal. In practice it is rarely possible because the seller cannot read every buyer's mind.
• Second degree discrimination: prices vary by quantity or block purchased, not by buyer identity — e.g. electricity tariffs, mobile data packs, "buy 2 get 1 free". The seller captures only part of consumer surplus (self-selection by buyers).
• Third degree discrimination: the market is divided into separate sub-markets (by age, sex, location, income, time) and a different price is charged in each. This is the most common form — e.g. student and senior-citizen rates, cheaper matinee shows, dumping in foreign markets.

Profit-maximising rule (third degree). The monopolist treats each sub-market as a separate monopoly and applies the marginal rule:

$MR_1 = MR_2 = \dots = MC$

Using the relation between marginal revenue and elasticity, $MR = P\left(1 - \dfrac{1}{e}\right)$, equalising $MR$ across markets gives:

$\frac{P_1}{P_2} = \frac{\left(1 - \dfrac{1}{e_2}\right)}{\left(1 - \dfrac{1}{e_1}\right)}$

Hence the market with the lower elasticity of demand (less elastic) is charged the higher price, and the more elastic market the lower price. If elasticities were equal, discrimination would not pay.

Diagrammatic illustration. Draw two panels side by side, one for each sub-market, with the common $MC$ curve. In the inelastic market the $MR$ curve cuts $MC$ at a low output and a high price; in the elastic market it cuts at a larger output and a lower price. Total output under discrimination is generally greater than under a single-price monopoly (because the low-price market is now served), though still below the competitive output. Profit is higher than under uniform pricing, since the monopolist gains by selling extra units in the elastic market without cutting the price in the inelastic one.

Effects / welfare implications.

• Gain: output expands, so allocative efficiency improves relative to simple monopoly; consumers in the elastic (poorer) market get the good at a lower price; the firm earns higher profit which may finance investment.
• Loss: consumer surplus is transferred to the monopolist; equity worsens; "predatory" or exclusionary discrimination may drive out rivals; consumers in the inelastic market are exploited.

Relevance to Nepal. Examples include NEA's time-of-day and industrial vs. domestic tariffs, Nepal Telecom's off-peak call rates, air fares differing by season and booking time, and differential doctor's fees. Such practices are common where resale is difficult and buyers differ in sensitivity to price.

Conclusion. Price discrimination is a device by which a monopolist converts consumer surplus into profit. It is feasible only under monopoly power, market separation and differing elasticities; it raises the monopolist's profit and usually total output, but at the cost of equity.

Deductive and Inductive Methods of Analysis

Deductive method (hypothetico-deductive / abstract method). It proceeds from the general to the particular. The analyst begins with a general premise — an axiom, assumption or established principle — and by logical reasoning derives a particular conclusion or prediction. Classical and neo-classical economics (Ricardo, Marshall, Walras) relied heavily on it. Its steps are: (i) state the general premise or assumption, (ii) reason logically from it, (iii) arrive at a particular conclusion, and (iv) test the conclusion against facts. For example, from the general law "other things being equal, quantity demanded falls as price rises," one deduces that raising the price of rice will reduce its demand.

Inductive method (empirical or historical method). It proceeds from the particular to the general. The analyst collects and observes particular facts, classifies and analyses them, and from the observed uniformities formulates a general law or theory. It was emphasised by the German Historical School and by institutionalists. Its steps are: (i) observation and collection of facts, (ii) classification and analysis of facts, (iii) formulation of a general hypothesis or law, and (iv) verification of the law. For example, by observing price and consumption data for many households over time, one may induce the general law of demand.

Key differences

Conclusion. Neither method is self-sufficient. Deduction gives logical rigour and generality but needs facts for verification; induction gives factual realism but needs theory for direction and interpretation. Modern economics therefore uses them as complementary — hypotheses are deduced from theory and then tested inductively against data.

The von Neumann–Morgenstern (N-M) utility index is a cardinal index of utility under risk, built from a person's choices between gambles. It is constructed as follows.

1. Fix two reference outcomes. Choose the best outcome $A$ and the worst outcome $B$ in the range of interest, and assign arbitrary values, e.g. $u(B)=0$ and $u(A)=1$ (or 0 and 100).
2. Take an intermediate outcome $C$ (e.g. a sure Rs 50,000, with $B$ = Rs 0 and $A$ = Rs 1,00,000).
3. Offer a lottery giving $A$ with probability $p$ and $B$ with probability $1-p$. Vary $p$ until the person is indifferent between the lottery and receiving $C$ for certain. This $p^*$ is the indifference point (the certainty equivalent of the lottery is $C$).
4. Assign the index. By the expected utility hypothesis, $u(C)=p^*\,u(A)+(1-p^*)\,u(B)=p^*$ so if the person is indifferent at $p^*=0.7$, then $u(C)=0.7$.
5. Repeat for many outcomes, then plot outcome against index to obtain the utility curve. A concave curve means risk aversion, linear means risk neutrality and convex means risk loving.

Key property. The index is unique only up to a positive affine transformation, $v=a+bu$ with $b>0$. Origin and unit are arbitrary, but ratios of utility differences are meaningful, and the ranking of gambles by expected utility is unchanged.

Requirements: the axioms of completeness, transitivity, continuity and independence must hold.

Firm's Equilibrium under One Variable Input Case

Meaning

A firm is said to be in equilibrium when it maximizes its profit. In the one-variable input case, the firm uses one variable input (say labour, (L)) along with fixed inputs (capital) in the short run. Since capital is fixed, the firm can attain equilibrium only by adjusting the quantity of the variable input.

Conditions of Equilibrium

The firm maximizes profit where the marginal revenue product of the variable input equals its marginal factor cost:

[ MRP_L = MFC_L ]

where:

• (MRP_L = MP_L \times MR) (marginal physical product of labour times marginal revenue)
• (MFC_L = \dfrac{\Delta TC}{\Delta L}) (addition to total cost from hiring one more unit of labour)

Under perfect competition in both product and factor markets, (MR = P) and (MFC_L = w) (wage rate), so the condition becomes:

[ MRP_L = P \times MP_L = VMP_L = w ]

That is, the value of marginal product of labour equals the wage rate.

Why This Condition Ensures Equilibrium

• If (MRP_L > MFC_L): the revenue from hiring one more unit of labour exceeds its cost, so the firm can increase profit by employing more labour.
• If (MRP_L < MFC_L): the cost of the last unit of labour exceeds the revenue it generates, so the firm can increase profit by reducing employment.
• Only when (MRP_L = MFC_L) is profit maximized, and the firm has no incentive to change the input level.

Second-Order Condition

For profit maximization (not minimization), the (MRP_L) curve must be falling at the point of intersection with the (MFC_L) line, i.e.:

[ \frac{d(MRP_L)}{dL} < 0 ]

This follows from the law of diminishing marginal returns.

Diagrammatic Representation

In the figure:

• The (MRP_L) curve slopes downward due to diminishing marginal product.
• The (MFC_L) line is horizontal under perfect competition in the factor market (equal to wage (w)).
• Equilibrium occurs at point (E), where (MRP_L = MFC_L), giving the optimal employment level (OL^*).

Conclusion

Thus, in the one-variable input case, a firm attains equilibrium when the marginal revenue product of the variable input equals its marginal factor cost (or, under perfect competition, when (VMP_L = w)), provided the (MRP_L) curve is falling at that point. At this level of employment, the firm maximizes its profit.

Given: $TC = 200 + 70Q - 10Q^2 + Q^3,\quad Q=40$

TFC = fixed part of TC = 200

TVC = variable part of TC: $TVC = 70Q - 10Q^2 + Q^3$ At $Q=40$: $TVC = 70(40) - 10(40)^2 + (40)^3$ $= 2800 - 16000 + 64000 = 50800$

TC at $Q=40$: $TC = TFC + TVC = 200 + 50800 = 51000$

AFC: $AFC = \frac{TFC}{Q} = \frac{200}{40} = 5$

AVC: $AVC = \frac{TVC}{Q} = \frac{50800}{40} = 1270$

MC: $MC = \frac{dTC}{dQ} = 70 - 20Q + 3Q^2$ At $Q=40$: $MC = 70 - 20(40) + 3(40)^2$ $= 70 - 800 + 4800 = 4070$

Answer

So, at 40 units of output: TFC = 200, TVC = 50,800, TC = 51,000, AFC = 5, AVC = 1,270 and MC = 4,070.

Derivation of the Short-Run Supply Curve of a Firm under Perfect Competition

A perfectly competitive firm is a price taker: it faces a horizontal demand (average revenue) curve at the market price $P$, so $P = AR = MR$. Its short-run supply curve is derived from the profit-maximising condition and the shut-down condition.

Step 1: Profit-maximising condition

Profit is maximum where marginal revenue equals marginal cost, with MC rising:

$MR = MC \quad \text{and} \quad MC \text{ rising at the point of intersection}$

Since $MR = P$ under perfect competition, the condition becomes:

$P = MC$

Step 2: Shut-down condition

The firm produces only if price covers average variable cost. The shut-down point is the minimum of AVC, where $P = \min AVC$. Below this price the firm loses more by producing than by shutting down (it would lose only fixed cost).

Step 3: Deriving the curve

• At each market price $P_1, P_2, P_3, \dots$, the firm reads off the output $Q_1, Q_2, Q_3, \dots$ from the rising portion of its MC curve where $P = MC$.
• Plotting these price–quantity pairs gives the firm's supply curve.
• The curve starts only from the minimum point of AVC; the portion of MC below $\min AVC$ is not part of supply because the firm shuts down there.

Conclusion

The short-run supply curve of a perfectly competitive firm is that portion of its short-run marginal cost curve which lies above the minimum point of average variable cost. It is upward sloping, showing that a higher price induces a larger quantity supplied, and it becomes vertical (zero output) at any price below $\min AVC$.