Laspeyres, Paasche and Fisher
Index Numbers — Chapter 5, GSEB Class 12 Statistics
Why Weighted Index Numbers?
In a simple average, every item gets equal importance. But in reality, rice matters more than saltin a family's budget. Weighted index numbers assign weights (quantities consumed) to reflect this importance.
Laspeyres Index
Uses base year quantities as weights. Easy to compute because q₀ stays constant. But since it uses old consumption patterns, it ignores that people may have shifted to cheaper substitutes.
Example
If 2018 basket = 100 kg rice + 50 L milk, Laspeyres always uses this same basket — even if in 2024 people buy less rice and more noodles.
Real Life
Government CPI-L uses Laspeyres approach. When inflation is reported as 6%, the actual feeling may be lower because people substitute expensive items.
Paasche Index
Uses current year quantities as weights. Reflects current consumption but understates inflation because people naturally shift toward cheaper goods when prices rise.
Example
If rice price doubled, you buy less rice and more noodles. Paasche uses the new (lower) rice quantity, making the index appear lower than the true price increase.
Real Life
Used in GDP deflator. When GDP growth is reported, the deflator using Paasche may understate the actual price rise consumers experience.
Fisher Index
Geometric mean of Laspeyres and Paasche. Balances the overstatement of I_L with the understatement of I_P. Called the “ideal index” because it satisfies both time reversal and factor reversal tests.
Example
If I_L = 120 (overstates) and I_P = 110 (understates), I_F = √(120 × 110) = √13200 ≈ 114.89 — a balanced middle ground.
Real Life
Used by RBI for some inflation measures. International agencies prefer Fisher for cross-country comparisons.
Comparison of All Three
| Feature | Laspeyres (I_L) | Paasche (I_P) | Fisher (I_F) |
|---|---|---|---|
| Weights Used | Base year quantities (q₀) | Current year quantities (q₁) | Both (geometric mean) |
| Formula | Σ(P₁q₀) / Σ(P₀q₀) × 100 | Σ(P₁q₁) / Σ(P₀q₁) × 100 | √(I_L × I_P) |
| Bias | Overstates inflation | Understates inflation | Ideal (balanced) |
| Ease of Computation | Easy (q₀ constant) | Harder (needs current data) | Moderate (needs both) |
| Time Reversal Test | ✗ Fails | ✗ Fails | ✓ Satisfies |
| Factor Reversal Test | ✗ Fails | ✗ Fails | ✓ Satisfies |
Key Takeaways
Key Takeaways
- Laspeyres (I_L) uses base year quantities (q₀) as weights — easy to compute but overstates inflation.
- Paasche (I_P) uses current year quantities (q₁) as weights — reflects current patterns but understates inflation.
- Fisher (I_F) = √(I_L × I_P) — the "ideal index" that balances both biases.
- I_F satisfies both time reversal and factor reversal tests — I_L and I_P fail both.
- By × bxy = r² for regression coefficients; I_F = √(I_L × I_P) for index numbers.
- In GSEB exams, compute all three when asked — show the formulas and identify the bias direction.