LESSON 12.1 — Population Projection Methods (Linear, Exponential, Modified Exponential, Gompertz, Comparative, Ratio)
A. Standard Map
| Topic | Governing Source | Exam Focus |
|---|---|---|
| Projection vs forecast | Definitions; uncertainty | Distinction |
| Linear method | Arithmetic growth; constant absolute change | Formula + use |
| Exponential (geometric) method | Constant percentage rate | Formula + use |
| Modified exponential | Saturating growth; asymptote | Formula + use |
| Logistic / Gompertz curve | S-curve with saturation | Formula + use |
| Comparative method | Compare to similar city | Approach + caveats |
| Ratio method | City as share of region/nation | Approach |
| Component method (cohort-component) | Births, deaths, migration components | Concept + use |
| Choice of method | Time horizon, data, growth stage | Decision logic |
B. Why It’s Used
Paper II §12 of the TGPSC syllabus opens with “Population projection methods, linear model, exponential curve, modified exponential, Gompertz growth curve, comparative method and ratio method.” Every Master Plan, every infrastructure design (water supply, sewerage, drainage, transport), and every housing-demand estimate begins with a population projection. A projection that is too high produces over-built infrastructure and wasted capital; too low produces premature breakdown and shortages. The exam tests method identification (which method fits which growth pattern), formulae (linear, exponential, modified exponential, Gompertz), and worked computations. Planners use these methods routinely — they are the technical core of the planner’s quantitative skill.
C. Mechanism in Words
-
A population projection is a calculation of what the future population will be if specified assumptions about fertility, mortality, and migration hold. It is distinct from a forecast (which asserts what the population will be) and an estimate (which calculates current population between censuses). A projection says: “if current trends continue, the population in year X will be Y.” Planners use projections to size infrastructure — the design population for a water supply system is the projected population at the design horizon (typically 20–30 years). Because projections rest on assumptions about the future, they carry uncertainty; good practice is to prepare high, medium, and low variants to bracket the range.
-
The Linear (Arithmetic) method assumes the population grows by a constant absolute amount per unit time. Formula: P_t = P_0 + (B × t), where P_t is population in year t, P_0 is the base-year population, B is the absolute growth per year, and t is years from base. The graph is a straight line. This method is simple and intuitive but rarely fits real populations — populations do not usually grow by the same absolute number every year. It is appropriate only for short-horizon projections in slow-growing areas, or where the data clearly shows linear growth.
-
The Exponential (Geometric) method assumes the population grows at a constant percentage rate per unit time. Formula: P_t = P_0 × (1 + r)^t, where r is the annual growth rate (as a decimal). Equivalently, in continuous form: P_t = P_0 × e^(rt). The graph is a curve that grows steeper over time. Exponential growth is appropriate for populations with a roughly constant growth rate — many developing-country cities during phases of rapid urbanisation. The exponential method is the most commonly used simple projection method in Indian practice.
-
The Modified Exponential method assumes the population approaches a saturation level asymptotically. Formula: P_t = S − (S − P_0) × e^(−kt), where S is the saturation (maximum) population, P_0 is the base-year population, k is a growth-rate parameter, and t is years. The graph is a curve that starts growing exponentially and then flattens as it approaches S — a saturating curve. This method is appropriate for cities where growth is constrained by physical or economic limits — an island city, a city with no land for expansion, a city approaching its carrying capacity.
-
The Logistic / Gompertz curve produces the classic S-curve of population growth. The Logistic curve formula: P_t = S / (1 + m × e^(−kt)), where S is the saturation population, m and k are parameters, and t is years. The curve starts slow, accelerates, then slows as it approaches saturation — like the demographic transition itself. The Gompertz curve is similar but asymmetric: P_t = S × e^(−b × e^(−kt)). Both produce S-curves; the Gompertz curve rises faster initially and slows more gradually than the logistic. These methods are appropriate for cities or regions with a clearly identifiable growth cycle: slow early growth → rapid growth → maturation.
-
The Comparative method projects a city’s population by analogy with another, more developed city. The planner identifies a “similar” city that is 20–30 years ahead (often in a developed country or in an earlier-developed part of India), observes how its population grew at the equivalent stage, and applies that pattern to the subject city. For example, projecting Warangal’s growth might use Hyderabad’s growth at an earlier stage as a template. The method is intuitive but heavily assumption-laden — the choice of “comparable” city is critical and contested.
-
The Ratio method projects a city’s population as a share of a larger region’s projected population. The city’s share (e.g., Hyderabad = 35% of Telangana urban) is held constant or projected separately; the region’s population is projected by a primary method (usually cohort-component); the city’s projection = ratio × region. This method leverages the fact that regional data is often more robust than city-level data.
-
The Component (Cohort-Component) method is the gold standard for population projection. It breaks the population into age-sex cohorts and projects each cohort forward by applying age-specific fertility rates, mortality rates, and migration rates. The demographic equation (Lesson 1.1) is applied at each timestep: cohort(t+1) = cohort(t) + births − deaths + net migration. The method is data-intensive (requires age-specific fertility, mortality, and migration rates) but produces the most realistic projections — and the most useful for planning (because age structure matters for schools, hospitals, workforce, etc.). The United Nations Population Division, the Office of the Registrar General of India, and the Census Commissioner use cohort-component for official national and state projections.
-
Choice of method depends on time horizon, data availability, and growth stage. Short horizon (5–10 years) — linear or exponential methods are adequate. Medium horizon (10–20 years) — exponential, modified exponential, or comparative methods. Long horizon (20–50 years) — modified exponential, logistic, or cohort-component (especially when age-structure detail matters). Data-poor situations — comparative or ratio methods. Data-rich situations — cohort-component. The stage of demographic transition also matters — exponential fits cities in early transition (rapid growth), modified exponential fits mature cities approaching saturation, logistic fits cities passing through full transition.
D. Core Concept Explanations
C1. The six projection methods
| Method | Formula | Curve shape | Best for |
|---|---|---|---|
| Linear | P_t = P_0 + B × t | Straight line | Short horizon; very slow growth |
| Exponential | P_t = P_0 × (1 + r)^t | Accelerating curve | Constant-rate rapid growth |
| Modified exponential | P_t = S − (S − P_0) × e^(−kt) | Saturating curve | Cities with growth constraints |
| Logistic | P_t = S / (1 + m × e^(−kt)) | S-curve | Full demographic transition |
| Gompertz | P_t = S × e^(−b × e^(−kt)) | Asymmetric S-curve | Same as logistic, faster early growth |
| Comparative | P_t = comparison with a similar city | Varies | Analogous cities |
| Ratio | P_city = ratio × P_region | Varies | Cities as share of a larger region |
C2. Linear vs Exponential — illustrated
A city with P_0 = 1,000,000 grows at 2% per year (exponential) or by 20,000 per year (linear equivalent at year 0):
| Year | Linear (P_0 + 20,000 × t) | Exponential (P_0 × 1.02^t) |
|---|---|---|
| 0 | 1,000,000 | 1,000,000 |
| 5 | 1,100,000 | 1,104,081 |
| 10 | 1,200,000 | 1,218,994 |
| 20 | 1,400,000 | 1,485,947 |
| 30 | 1,600,000 | 1,811,362 |
| 50 | 2,000,000 | 2,691,588 |
After 50 years, exponential gives 35% higher population than linear at the same initial growth rate. The choice matters enormously at planning horizons of 20–30 years.
C3. Logistic curve — how it saturates
For a city with S = 5,000,000 saturation, m = 4, k = 0.05, t in years:
| Year (t) | P_t |
|---|---|
| 0 | 5M / (1 + 4) = 1,000,000 |
| 10 | 5M / (1 + 4 × e^(−0.5)) = 5M / (1 + 4 × 0.6065) = 5M / 3.4261 = 1,459,376 |
| 30 | 5M / (1 + 4 × e^(−1.5)) = 5M / (1 + 4 × 0.2231) = 5M / 1.8925 = 2,641,979 |
| 60 | 5M / (1 + 4 × e^(−3.0)) = 5M / (1 + 4 × 0.0498) = 5M / 1.1991 = 4,169,626 |
| 100 | 5M / (1 + 4 × e^(−5.0)) = 5M / (1 + 4 × 0.0067) = 5M / 1.0270 ≈ 4,868,778 |
The curve starts slow (year 0–10), accelerates (10–30), and saturates (60+) — the classic S-curve.
E. Worked Numericals and Parameter Tables
E1. Linear projection
A town had population 200,000 in 2011 and 240,000 in 2021 (Census years). Linear growth rate B = (240,000 − 200,000) / 10 = 4,000/year. Projection for 2031: P_2031 = 240,000 + (4,000 × 10) = 280,000.
E2. Exponential projection
Using the same data: r = (240,000/200,000)^(1/10) − 1 = 1.2^0.1 − 1 ≈ 1.84% per year. Projection for 2031: P_2031 = 240,000 × (1.0184)^10 = 240,000 × 1.20 = 288,000. (Slightly higher than linear; the gap grows over longer horizons.)
E3. Modified exponential
A city approaching saturation S = 3,000,000; P_0 = 1,500,000 (in 2020); k = 0.04. Projection for 2040 (t = 20): P_2040 = 3,000,000 − (3,000,000 − 1,500,000) × e^(−0.04 × 20) = 3,000,000 − 1,500,000 × e^(−0.8) = 3,000,000 − 1,500,000 × 0.449 = 3,000,000 − 674,000 = 2,326,000. The city is approaching but not yet at saturation.
E4. Ratio method
A state’s urban population is projected to be 25 million in 2031. The subject city currently (2021) is 12% of state urban. Holding the ratio constant: P_city(2031) = 0.12 × 25 million = 3 million. If the ratio is projected to rise (city gaining share) to 15%: P_city = 0.15 × 25 = 3.75 million.
E5. Doubling time check
At exponential growth rate r = 3% per year, doubling time = ln(2)/ln(1.03) ≈ 23.4 years. A city of 1 million in 2020 doubles to 2 million by 2043–44 at this rate.
F. Design Criteria
| Parameter | Standard / Typical value | Source |
|---|---|---|
| Typical Master Plan horizon | 20 years | URDPFI 2015 |
| Typical Perspective Plan horizon | 20–50 years | URDPFI 2015 |
| Census interval | 10 years | Census Act 1948 |
| Doubling time formula (continuous) | ln(2) / r ≈ 0.693 / r | Mathematics |
| Doubling time formula (discrete) | ln(2) / ln(1+r) | Mathematics |
| Rule of 70 | Doubling time ≈ 70 / (r in %) | Mental arithmetic |
| Standard projection variants | High, Medium, Low | Demographic convention |
G. Application Zones
- Master Plan population chapter — every plan opens with a projection.
- Infrastructure design — water supply, sewerage, drainage, roads sized for design population at horizon.
- Housing demand estimation — projected households × affordability distribution.
- School / hospital planning — age-structure (from cohort-component) drives demand.
- Comparative planning studies — analogous-city benchmarks.
H. Common Confusions
| Confusion | Reality |
|---|---|
| “Linear and exponential give the same answer.” | No — they diverge significantly over 20+ year horizons. |
| “Projection = forecast.” | No — projection is conditional (“if these trends continue”); forecast asserts what will happen. |
| “Gompertz and logistic are identical.” | No — both are S-curves, but Gompertz is asymmetric (faster early growth, slower late). |
| “Modified exponential has no saturation.” | False — modified exponential approaches S (saturation) asymptotically. |
| “Comparative method is more rigorous than cohort-component.” | False — comparative is intuitive but assumption-laden; cohort-component is the gold standard. |
| “Ratio method requires age-structure data.” | No — ratio method uses aggregate populations; only cohort-component needs age structure. |
| “Linear method is appropriate for cities in rapid transition.” | No — linear usually underestimates; exponential or modified exponential fit better. |
I. Compare & Contrast
I1. Linear vs exponential vs modified exponential
| Method | Shape | Use case | Long-horizon behaviour |
|---|---|---|---|
| Linear | Straight line | Slow growth; short horizon | Linear → often underestimates |
| Exponential | Accelerating curve | Constant-rate growth | Exponential → often overestimates (no upper limit) |
| Modified exponential | Saturating curve | Growth with constraints | Approaches S |
I2. Logistic vs Gompertz
| Method | Curve shape | Growth phase | Saturation phase |
|---|---|---|---|
| Logistic | Symmetric S-curve | Steepens at midpoint | Slows gradually |
| Gompertz | Asymmetric S-curve | Rises faster initially | Slows more gradually |
J. Memory Hooks
- “L-E-M-L-G” — Linear, Exponential, Modified exponential, Logistic, Gompertz — five projection methods.
- “P_t = P_0(1+r)^t” — exponential formula.
- “S / (1 + m × e^(−kt))” — logistic formula.
- “70 / r = doubling time” — quick mental math.
- “Comparative = analogous city; Ratio = share of region” — two contextual methods.
- “Cohort-component = gold standard” — uses age-sex cohorts + fertility/mortality/migration.
- “20-year Master Plan; 20–50-year Perspective Plan” — typical horizons.
K. Revision Ladder
| Order | Item | Time |
|---|---|---|
| 1 | Memorise the 5 primary projection methods with formulas | 45 min |
| 2 | Memorise linear vs exponential vs modified exponential distinction | 30 min |
| 3 | Memorise logistic vs Gompertz curve shapes | 30 min |
| 4 | Memorise comparative and ratio methods with use cases | 20 min |
| 5 | Practise linear and exponential projection arithmetic | 45 min |
| 6 | Practise modified exponential and logistic arithmetic | 45 min |
| 7 | Memorise cohort-component method concept | 20 min |
| 8 | Memorise the doubling-time rule (Rule of 70) | 10 min |
| 9 | Map Telangana projection practice (HMDA HMDP, DTCP town plans) | 30 min |
L. Exam Traps
| Trap | Correct response |
|---|---|
| Question pairs linear with “constant percentage growth.” | False — linear is constant absolute growth; exponential is constant percentage. |
| Question lists modified exponential as having no upper limit. | False — it approaches S (saturation) asymptotically. |
| Question pairs Gompertz with symmetric S-curve. | False — Gompertz is asymmetric. |
| Question pairs cohort-component with “uses no age data.” | False — cohort-component is built on age-sex cohorts. |
| Question lists projection as synonymous with forecast. | False — projection is conditional; forecast asserts the future. |
| Question asks doubling time at 5% (Rule of 70). | 70 / 5 = 14 years. |
| Question pairs ratio method with “needs fertility data.” | False — ratio method uses aggregate populations only. |
M. Answer-Writing Cues
- For method questions, give method + formula + curve shape + use case: “The exponential (geometric) method, P_t = P_0 × (1+r)^t, assumes a constant percentage growth rate; it is appropriate for cities in rapid urbanisation phases but tends to overestimate over long horizons because it has no saturation limit.”
- For numerical questions, show the formula, the substitution, and the units.
- For choice-of-method questions, justify based on time horizon, data, and growth stage.
N. PYQ Integration
Pattern questions only:
Pattern question 1 — Method identification
Q. The population projection method that assumes the population approaches a maximum saturation level asymptotically is:
– (A) Linear
– (B) Exponential
– (C) Modified exponential ✓
– (D) Comparative
Ans: (C). Modified exponential = S − (S − P_0) × e^(−kt).
Pattern question 2 — Formula
Q. The exponential (geometric) population projection formula is:
– (A) P_t = P_0 + B × t
– (B) P_t = P_0 × (1 + r)^t ✓
– (C) P_t = S / (1 + m × e^(−kt))
– (D) P_t = S − (S − P_0) × e^(−kt)
Ans: (B). (A) is linear; (C) is logistic; (D) is modified exponential.
Pattern question 3 — Doubling time
Q. Using the Rule of 70, the doubling time at 3.5% annual growth rate is approximately:
– (A) 14 years
– (B) 20 years ✓
– (C) 35 years
– (D) 70 years
Ans: (B). 70 / 3.5 = 20 years.
Pattern question 4 — MSQ
Q. Which of the following are standard population projection methods in planning?
– (A) Linear ✓
– (B) Exponential ✓
– (C) Gompertz ✓
– (D) Trial and error
Ans: (A), (B), (C). Trial and error is not a projection method.
Pattern question 5 — Numerical
A city of 200,000 in 2021 is projected using the exponential method at 2% per year. The 2031 population (10 years) is approximately:
– (A) 220,000
– (B) 240,000
– (C) 244,000 ✓
– (D) 280,000
Ans: (C). 200,000 × (1.02)^10 = 200,000 × 1.219 ≈ 243,800.
O. Mini-Check — Lesson 12.1
- Distinguish projection, forecast, and estimate.
- State the linear projection formula.
- State the exponential projection formula.
- State the modified exponential formula and the meaning of S.
- State the logistic curve formula.
- Distinguish logistic from Gompertz curves.
- State the comparative method’s approach.
- State the ratio method’s approach.
- Which method is the gold standard, and why?
- Compute doubling time at 5% growth using the Rule of 70.
Answers:
1. Projection = conditional calculation (“if trends continue”); forecast = asserts what will happen; estimate = current population between censuses.
2. P_t = P_0 + B × t (B = absolute growth per year).
3. P_t = P_0 × (1 + r)^t (r = annual growth rate).
4. P_t = S − (S − P_0) × e^(−kt). S = saturation (maximum) population.
5. P_t = S / (1 + m × e^(−kt)).
6. Both produce S-curves, but logistic is symmetric and Gompertz is asymmetric (faster early growth, slower late).
7. Project a city’s population by comparing it to a more-developed, analogous city at an earlier stage.
8. Project a city’s population as its share of a larger region’s projected population (the ratio is held constant or projected separately).
9. Cohort-component — uses age-sex cohorts with age-specific fertility, mortality, and migration rates; produces realistic projections and age-structure detail essential for schools, hospitals, workforce planning.
10. 70 / 5 = 14 years.
Next: Lesson 12.2 — Regional Surveys, Cluster/Factor Analysis, I-O Techniques & Spatial Analysis.