Production Functions, Costs & Efficiency: Short/Long Run Analysis

TAKEAWAYS

• Production functions show how inputs (labor, capital) combine to create output, with returns to scale (increasing/decreasing/constant) determining long-run growth potential.
• Short-run costs are fixed (e.g., factory rent) + variable (e.g., wages), while long-run costs are all variable, leading to U-shaped average cost curves and marginal cost driving output decisions.
• Optimal factor ratios (e.g., labor/capital mix) depend on factor prices and marginal productivities—firms hire until MP_L/wage = MP_K/rental rate.
• Economies of scale (e.g., bulk discounts, specialization) lower LAC, while diseconomies (e.g., coordination costs) raise it—explaining why some firms dominate markets (e.g., Ncell vs. smaller telcos).
• Profit maximization occurs where MR = MC (short run) or P = MC (perfect competition), but fixed costs force shutdown rules (P ≥ AVC to avoid losses).
• Real-world trade-offs: Firms like Daraz (e-commerce) use automation (capital) to cut labor costs, while Pathao (ride-hailing) relies on labor-intensive matching algorithms—both optimize factor ratios.

---

1. Production Functions: How Inputs Become Output

Key Definitions

• Production Function: $Q=f(L,K)$ Output depends on labor (L) and capital (K) inputs.
• Total Product (TP): Total output from all inputs.
• Marginal Product (MP): Additional output from one extra unit of input, holding others constant. $M{P}_L=\frac{\Delta Q}{\Delta L}$, $M{P}_K=\frac{\Delta Q}{\Delta K}$.
• Average Product (AP): Output per unit of input. $A{P}_L=\frac{Q}{L}$, $A{P}_K=\frac{Q}{K}$.

Law of Diminishing Marginal Returns

As more of a variable input (e.g., labor) is added to fixed inputs (e.g., machinery), MP eventually falls. Why? Workers crowd machinery, efficiency drops. Example: A Nepali brick kiln with 10 workers produces 1000 bricks/day. Adding a 11th worker boosts output to 1500 (MP=500), but a 21st worker adds only 100 bricks (MP=100). TP rises but at a decreasing rate.

Types of Returns to Scale

Worked Example: Daraz’s Warehouse Efficiency Daraz’s production function for order fulfillment: $Q=200L^{0.6}{K}^{0.4}$, where:

• $L$ = labor (workers),
• $K$ = capital (automated sorting robots),
• $Q$ = orders fulfilled/hour.

Questions:

1. If Daraz hires 100 workers and installs 50 robots, what’s the marginal product of labor (MP_L)?
2. If wages rise by 20%, how should Daraz adjust its factor ratio?

Solution:

1. $M{P}_L=\frac{\partial Q}{\partial L}=200\times 0.6\times L^{-0.4}{K}^{0.4}$. At $L=100$, $K=50$: $M{P}_L=120\times 100^{-0.4}\times 50^{0.4}\approx 120\times 0.30\times 2.71\approx \ast \ast 10.4orders/worker\ast \ast$.
2. Optimal factor ratio: $\frac{M{P}_L}{wage}=\frac{M{P}_K}{rental}$. If wages rise, $M{P}_L/wage$ falls → hire more robots (capital) to restore equality.

---

2. Cost Analysis: Short Run vs. Long Run

Short-Run Costs (Fixed Plant)

In the short run, at least one input is fixed (e.g., factory size). Costs are:

• Total Cost (TC): $TC=TFC+TVC$.
• Average Fixed Cost (AFC): $AFC=\frac{TFC}{Q}$ (falls as Q rises).
• Average Variable Cost (AVC): $AVC=\frac{TVC}{Q}$ (U-shaped due to diminishing MP).
• Marginal Cost (MC): $MC=\frac{\Delta TC}{\Delta Q}$ (cuts AVC at its minimum).

Key Insight: The MC curve intersects AVC and AC at their lowest points. If price $P<\text{min AVC}$, the firm shuts down (e.g., a small hotel in Pokhara during monsoon).

Long-Run Costs (All Inputs Variable)

• Long-Run Average Cost (LAC): Envelope of all short-run AC curves.
• Economies of Scale: LAC falls as Q rises (e.g., Nepal’s cement industry—larger plants like Shivam Cement have lower per-unit costs).
• Diseconomies of Scale: LAC rises due to coordination issues (e.g., NTC’s bureaucracy slowing decision-making).

Real-World Example: Nepal’s Banking Sector

• Global IME Bank (small) has high per-customer costs (manual processing).
• NMB Bank (large) uses automation and economies of scale to offer lower interest rates on loans.

---

3. Factor Intensity and Optimal Input Mix

Factor Intensity

• Labor-intensive: High $L/K$ ratio (e.g., handloom weaving in Nepal).
• Capital-intensive: High $K/L$ ratio (e.g., Daraz’s automated warehouses).
• Optimal Ratio: Achieved when $\frac{M{P}_L}{wage}=\frac{M{P}_K}{rental}$.

Worked Example: Nepali Brick Kiln

• Production function: $Q=50L^{0.7}{K}^{0.3}$.
• Wage (w): Rs. 500/worker/day.
• Rental (r): Rs. 2000/machine/day.
• Current inputs: 20 workers, 5 machines.

Questions:

1. Is the kiln using the optimal factor ratio?
2. Should it hire more workers or buy more machines?

Solution:

1. Calculate $M{P}_L$ and $M{P}_K$: $M{P}_L=50\times 0.7\times L^{-0.3}{K}^{0.3}=35\times 20^{-0.3}\times 5^{0.3}\approx 35\times 0.45\times 1.62\approx \ast \ast 25.7bricks/worker\ast \ast$. $M{P}_K=50\times 0.3\times L^{0.7}{K}^{-0.7}=15\times 20^{0.7}\times 5^{-0.7}\approx 15\times 6.96\times 0.38\approx \ast \ast 40bricks/machine\ast \ast$.
2. Compare ratios: $\frac{M{P}_L}{w}=\frac{25.7}{500}=0.0514$, $\frac{M{P}_K}{r}=\frac{40}{2000}=0.02$. Since $0.0514>0.02$, the kiln should hire more workers to equalize ratios.

---

4. Profit Maximization and Shutdown Rules

Short-Run Profit Maximization

• Rule: Produce where $MR=MC$.
• Perfect Competition: $P=MR$, so $P=MC$.
• Monopoly/Oligopoly: $MR<P$, so $MR=MC$.

Worked Example: Nepal’s Small Dairy Farmer

• Demand: $P=100-2Q$.
• Cost: $TC=50+10Q$.
• Find: Profit-maximizing output, price, and profit.

Solution:

1. Total Revenue (TR): $TR=P\times Q=(100-2Q)Q=100Q-2Q^2$.
2. Marginal Revenue (MR): $MR=\frac{d(TR)}{dQ}=100-4Q$.
3. Marginal Cost (MC): $MC=\frac{d(TC)}{dQ}=10$.
4. Set $MR=MC$: $100-4Q=10$ → $Q=22.5$.
5. Price: $P=100-2(22.5)=55$.
6. Profit: $\pi =TR-TC=(55\times 22.5)-(50+10\times 22.5)=1237.5-275=\ast \ast Rs.962.5\ast \ast$.

Shutdown Rule:

• If $P<\text{min AVC}$, shut down (lose only TFC).
• If $\text{min AVC}\le P<\text{Average Total Cost (AC)}$, operate at a loss.

---

In the Real World

1. Daraz’s Algorithm-Driven Warehouses

   • Idea Used: Capital-intensive production (robots > labor).
   • How: Daraz’s automated sorting systems reduce labor costs by 30% while increasing order fulfillment speed. The optimal $L/K$ ratio is set by comparing the marginal product of robots (faster sorting) vs. wages (cheaper in Nepal but slower).

2. Pathao’s Driver-Labor Trade-off

   • Idea Used: Labor-intensive marginal product and diminishing returns.
   • How: Pathao’s driver supply follows the law of diminishing returns—adding the 100th driver in Kathmandu increases rides by 5%, but the 1000th driver adds only 0.5%. Pathao adjusts driver incentives (wages) to optimize $M{P}_L$.

3. Nepal’s Hydropower Plants (Capital vs. Labor)

   • Idea Used: Fixed vs. variable costs and economies of scale.
   • How: Small hydropower projects (e.g., 1 MW plants) have high AFC (dam construction) but low AVC (labor). Large projects (e.g., West Seti, 756 MW) achieve economies of scale, reducing AC per unit of electricity by 40%.

---

Exam Tip

1. Memorize the Shapes:

   • MC cuts AVC/AC at their minimum.
   • LAC is the envelope of SAC curves.
   • TP rises at a decreasing rate (diminishing MP).

2. Profit Maximization Shortcuts:

   • Perfect Competition: $P=MC=MR$.
   • Monopoly: $MR=MC$, then find $P$ on demand curve.
   • Shutdown: $P<\text{min AVC}$.

3. Factor Intensity Tricks:

   • If $w$ rises, substitute toward capital (e.g., Nepali farms use more tractors when labor costs rise).
   • If $r$ (rental cost) falls, substitute toward capital (e.g., Daraz adds more robots).

4. Numerical Questions:

   • Always label axes in graphs (e.g., "Quantity (kg)").
   • For production functions, use logarithmic differentiation if exponents are decimals.
   • For cost minimization, set $\frac{M{P}_L}{w}=\frac{M{P}_K}{r}$.

5. Common Pitfalls:

   • Long run ≠ infinite time: Just enough to change all inputs.
   • AFC never touches MC: AFC falls toward zero but never meets MC.
   • Profit ≠ Revenue: Always subtract total cost!

---

Final Visual Summary: Interpretation: The firm produces 100 units where the isoquant is tangent to the isocost line, ensuring $\frac{M{P}_L}{w}=\frac{M{P}_K}{r}$.