Index Numbers & Percent Change — Index Numbers and Percent Change (Intro) - Business Statistics

SECTION 1

Historical Context & Motivation

Long before spreadsheets and dashboards, merchants, policymakers, and economists wrestled with a deceptively simple question: how much have prices actually changed? Comparing the cost of a single commodity across two dates is trivial, but comparing the aggregate cost of a basket of goods — bread, fuel, textiles, housing — demands a systematic framework. That framework is the index number, a ratio that re-expresses a complex data series relative to a fixed reference point called the base period. Paired with percent change — the proportional shift between any two values — index numbers have become indispensable tools in modern business analytics, finance, and economic policy.

1707 • Early Price Comparisons

English bishop William Fleetwood publishes Chronicon Preciosum, comparing the purchasing power of money over centuries — one of the earliest systematic attempts to track price changes across a basket of goods.

1864 • Laspeyres Index

German economist Étienne Laspeyres proposes a weighted aggregate index using base-period quantities as weights, establishing a formula still used in the Consumer Price Index (CPI) today.

1874 • Paasche Index

Hermann Paasche introduces an alternative that weights prices by current-period quantities, enabling analysts to capture substitution effects as consumers shift spending patterns.

1922 • Fisher Ideal Index

Irving Fisher proposes the geometric mean of the Laspeyres and Paasche indices, recognized as the 'ideal' index because it satisfies several desirable mathematical properties, including the time-reversal test.

1996–Present • Modern Applications

The Boskin Commission recommends chained indices for the U.S. CPI, and stock market composites (S&P 500, DJIA) become ubiquitous business benchmarks, extending index-number methodology well beyond price measurement.

Throughout this evolution, the core challenge has remained constant: how do we condense a multidimensional data set into a single, interpretable number that faithfully represents change over time or across categories? This lesson introduces the foundational concepts — simple index numbers and percent change — and equips you with the computational toolkit to apply them in real business contexts.

SECTION 2

Core Principles & Definitions

Before diving into calculations, it is essential to internalize the foundational principles that govern how index numbers and percent-change metrics work. These principles recur across every industry application, from tracking commodity prices to benchmarking portfolio returns against the S&P 500.

Base Period

A reference point in time (or place) against which all other values are compared. The base period's index value is set to 100 (or sometimes 1.00), creating a universal benchmark for interpretation.

Simple Index Number

The ratio of a current-period value to a base-period value, multiplied by 100. It answers: 'What is today's value as a percentage of the base?' An index of 120 means a 20% increase from the base.

Percent Change

The proportional difference between two values expressed as a percentage. It captures the magnitude and direction of change, serving as the most intuitive measure of relative movement in business data.

Directionality & Sign

Percent change is signed: positive values indicate growth, negative values indicate decline. The asymmetry of percentages means a 50% drop requires a 100% gain to recover — a critical insight for financial analysis.

Comparability

Index numbers enable apples-to-apples comparison of series measured in different units or magnitudes. Revenue in dollars and output in tons can both be re-expressed on a common 'base = 100' scale.

✦

KEY TAKEAWAY

Think of an index number as a speedometer needle that has been re-zeroed. When you calibrate your speedometer so that 60 mph reads as '100,' every future reading instantly tells you how your speed compares to that benchmark — 110 means 10% faster, 90 means 10% slower. Percent change is simply reading the difference between two needle positions. The power lies in normalization: it does not matter whether the underlying data is in dollars, units, or hours — the index converts everything into a common language of relative magnitude.

SECTION 3

Visual Explanation — From Raw Data to Index

The diagram below illustrates the transformation from raw revenue figures (in thousands of dollars) to an index series with 2020 as the base year. Notice how the base-year value maps to exactly 100 on the index axis, and every subsequent year's index value communicates its position relative to that anchor.

Left panel: raw revenue in thousands of dollars. Right panel: the same data re-expressed as an index series with 2020 as the base year (= 100). The amber arrow represents the transformation formula: divide each year's value by the base-year value and multiply by 100. Notice the shaded horizontal band at 100 on the right axis — every point above it indicates growth relative to the base.

The key insight from this diagram is that the absolute magnitude of the raw data becomes irrelevant once indexed. Whether the base-year revenue was $300K or $3 billion, the index value would still be 100, and 2023 would still read 150 — indicating a 50% cumulative increase. This normalization is precisely what makes index numbers so powerful for cross-series comparison: a product line generating $50K in base-year revenue and another generating $5M can be plotted on the same indexed chart, making their relative growth trajectories directly comparable.

SECTION 4

Mathematical Framework

The mathematical machinery behind index numbers and percent change is straightforward, but precision in application matters. Below are the core formulas you will encounter throughout this course and in professional practice.

SIMPLE INDEX NUMBER

Iₜ = (Vₜ / V₀) × 100

Where Iₜ = index value at time t, Vₜ = observed value at time t, and V₀ = observed value in the base period. The result is unit-free: an index of 130 means the value is 130% of the base, i.e., a 30% increase.

PERCENT CHANGE

% Change = ((V_new − V_old) / V_old) × 100

Where V_new is the more recent observation and V_old is the earlier observation. A positive result indicates growth; a negative result indicates decline. The denominator is always the starting value — the reference against which change is measured.

PERCENT CHANGE FROM INDEX VALUES

% Change₍ₜ₁→ₜ₂₎ = ((Iₜ₂ − Iₜ₁) / Iₜ₁) × 100

Because index numbers are proportional to underlying values, you can compute percent change directly from two index values without reverting to raw data. This is especially useful when working with published indices such as the CPI or a stock index.

REBASING AN INDEX

I'ₜ = (Iₜ / I_new_base) × 100

Sometimes you need to shift the base period from one year to another. Divide every index value in the series by the index value of the desired new base year, then multiply by 100. The new base year's index becomes 100, and all other values are rescaled proportionally.

⚠

Asymmetry Warning

Percent changes are not symmetric. If a stock drops 25% (from $100 to $75), recovering to $100 requires a gain of ($100 − $75) / $75 × 100 = 33.3%, not 25%. This asymmetry has direct implications for portfolio return calculations and loss recovery analysis. Always keep the correct denominator in mind — percent change is always computed relative to the starting value of the interval in question.

SECTION 5

Types of Index Numbers & Business Applications

Index numbers come in several varieties, each suited to different analytical questions. At the introductory level, the distinction between simple (unweighted) and composite (aggregate) indices is the most important. The diagram below maps the taxonomy and connects each type to a common business use case.

This taxonomy shows how index numbers branch into simple (single-item) and composite (aggregate) forms. Composite indices further subdivide based on weighting scheme. This lesson focuses on simple indices and percent change; weighted composites (Laspeyres, Paasche) are covered in subsequent lessons.

Index Type	Formula Approach	Best For
Simple Price Index	(Pₜ / P₀) × 100 for a single commodity	Tracking one item over time (e.g., gasoline, gold)
Simple Quantity Index	(Qₜ / Q₀) × 100 for a single item's volume	Measuring output changes (e.g., units shipped)
Simple Value Index	(Vₜ / V₀) × 100 where V = P × Q	Total revenue or expenditure comparisons
Percent Change	((V_new − V_old) / V_old) × 100	Period-over-period analysis (MoM, QoQ, YoY)

Summary of introductory index number types and their primary business applications

SECTION 6

Worked Example — Quarterly Sales Analysis

Suppose you manage a regional retail chain. Headquarters has provided quarterly revenue figures (in $thousands) for four quarters and asked you to (a) construct an index series using Q1 as the base, and (b) compute quarter-over-quarter percent changes. The data is: Q1 = $800K, Q2 = $920K, Q3 = $860K, Q4 = $1,040K.

Constructing Index Numbers and Percent Changes

1
Step 1 — Identify the Base Period

The problem designates Q1 as the base period. Therefore V₀ = $800K, and the base index value I₁ is automatically 100. All subsequent index values will be expressed as a percentage of $800K.

V₀ = $800K → I_Q1 = 100.0
2
Step 2 — Compute Index Values for Each Quarter

Apply the formula Iₜ = (Vₜ / V₀) × 100 for each quarter. For Q2: I = ($920K / $800K) × 100 = 1.15 × 100 = 115.0. For Q3: I = ($860K / $800K) × 100 = 1.075 × 100 = 107.5. For Q4: I = ($1,040K / $800K) × 100 = 1.30 × 100 = 130.0.

I_Q2 = 115.0, I_Q3 = 107.5, I_Q4 = 130.0
3
Step 3 — Interpret the Index Series

Q2 revenue was 15% above the base, Q3 dipped to only 7.5% above the base, and Q4 surged to 30% above the base. The index tells us that Q3 experienced a relative decline compared to Q2 (even though Q3 revenue still exceeded Q1), while Q4 represented the strongest quarter.
4
Step 4 — Compute Quarter-over-Quarter Percent Changes

Using % Change = ((V_new − V_old) / V_old) × 100: Q1→Q2 = (($920K − $800K) / $800K) × 100 = (120 / 800) × 100 = +15.0%. Q2→Q3 = (($860K − $920K) / $920K) × 100 = (−60 / 920) × 100 = −6.52%. Q3→Q4 = (($1,040K − $860K) / $860K) × 100 = (180 / 860) × 100 = +20.93%.

Q1→Q2: +15.0% | Q2→Q3: −6.52% | Q3→Q4: +20.93%
5
Step 5 — Cross-Check with Index Values

Verify Q2→Q3 percent change using index values: ((107.5 − 115.0) / 115.0) × 100 = (−7.5 / 115.0) × 100 = −6.52%. This confirms that percent change can be computed directly from index values — a technique that saves time when raw data is unavailable.

Cross-check confirmed: −6.52% ✓

SECTION 7

Strengths & Limitations

Like every analytical tool, index numbers and percent change calculations have well-defined strengths and important limitations. Understanding both sides helps business analysts deploy these metrics appropriately and avoid common pitfalls.

Strengths	Limitations
Unit-free: enables comparison across variables with different measurement scales (dollars vs. tons vs. hours).	Base-period dependence: choice of base year can skew perception — an anomalous base year distorts the entire series.
Intuitive interpretation: deviations from 100 immediately convey the direction and magnitude of change.	Simple indices ignore composition: a price index for one item says nothing about the broader market basket.
Percent change is universally understood by non-technical stakeholders, facilitating executive reporting.	Percent change is asymmetric — equal gains and losses do not offset, which can mislead if not understood.
Easily computed and visualized, requiring no advanced statistical software.	Percent change on small denominators produces exaggerated values (e.g., going from 1 to 3 = +200%).
Index series can be rebased, allowing temporal comparison across different reporting frameworks.	Does not indicate statistical significance; a 5% change could be noise or signal without further testing.

Comparative strengths and limitations of simple index numbers and percent change

✦

KEY TAKEAWAY

Think of index numbers as a map projection — they faithfully represent relative proportions but necessarily sacrifice some information (absolute scale). Just as a Mercator projection distorts area near the poles, an index series can distort interpretation if the base period is atypical. The antidote is to choose a representative base period and to always examine the raw data alongside the indexed series when making high-stakes decisions.

SECTION 8

Connection to Weighted Indices & Advanced Methods

The simple index number you have learned in this lesson is the building block for more sophisticated methods. In practice, analysts rarely track a single item in isolation — they need aggregate indices that combine multiple items with appropriate weights reflecting economic importance. The table below previews how this introductory concept connects to the weighted approaches you will encounter next.

Feature	Simple Index (This Lesson)	Weighted Composite Index (Next)
Number of items	One commodity, product, or metric	Basket of multiple items (e.g., CPI basket of 80,000 items)
Weighting	None — each item implicitly has weight of 1	Expenditure shares, market cap, or quantities as weights
Formula complexity	Iₜ = (Vₜ / V₀) × 100	Σ(Pₜ × Q₀) / Σ(P₀ × Q₀) × 100 (Laspeyres)
Substitution bias	Not applicable — only one item tracked	Present in fixed-weight indices; addressed via chaining
Typical applications	Single KPI tracking, simple price monitoring	CPI, PPI, stock market indices, GDP deflator

How the simple index relates to weighted composite indices

As you advance through this course, you will also encounter chain-linked indices, which update weights periodically to reflect changing consumption patterns, and deflation techniques that use price indices to convert nominal values into real (inflation-adjusted) values. Both build directly on the ratio-based logic introduced here. Mastering the simple index and percent change formula is therefore not just an academic exercise — it is the conceptual foundation for every downstream analytical method in descriptive analytics.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A company reports that its revenue index (base year 2019 = 100) stood at 145 in 2024. A manager claims that 'revenue grew by 145% since 2019.' Is this interpretation correct? Explain the error, if any, and provide the accurate statement.

Show answer

The manager's statement is incorrect. An index value of 145 means revenue in 2024 is 145% of the base-year level, not 145% above it. The correct interpretation is that revenue grew by 45% since 2019. This is because the percent change from base to current is ((145 − 100) / 100) × 100 = 45%. A common misconception is to conflate the index value itself with the percent change, but the index already embeds the base at 100, so only the deviation above 100 represents growth.

Question 2: BASIC CALCULATION

The average price of a gallon of regular gasoline was $3.20 in January and $3.68 in June. (a) Compute the simple price index for June using January as the base. (b) Compute the percent change in price from January to June.

Show answer

(a) Using Iₜ = (Vₜ / V₀) × 100: I_June = ($3.68 / $3.20) × 100 = 1.15 × 100 = 115.0. (b) Using percent change = ((V_new − V_old) / V_old) × 100: (($3.68 − $3.20) / $3.20) × 100 = ($0.48 / $3.20) × 100 = 15.0%. Both results are consistent: an index of 115.0 confirms a 15% increase relative to the base period. Note that you can always derive the percent change from the index by subtracting 100.

Question 3: INTERMEDIATE

A firm's monthly website traffic (in thousands of unique visitors) was: Jan = 200, Feb = 240, Mar = 228, Apr = 270. (a) Construct an index series using January as the base. (b) Compute the month-over-month percent change for each transition. (c) Verify the Feb→Mar percent change using index values instead of raw data.

Show answer

(a) Index series: Jan = (200/200)×100 = 100.0; Feb = (240/200)×100 = 120.0; Mar = (228/200)×100 = 114.0; Apr = (270/200)×100 = 135.0. (b) Percent changes: Jan→Feb = ((240−200)/200)×100 = +20.0%; Feb→Mar = ((228−240)/240)×100 = −5.0%; Mar→Apr = ((270−228)/228)×100 = +18.42%. (c) Using index values for Feb→Mar: ((114.0−120.0)/120.0)×100 = (−6.0/120.0)×100 = −5.0%. This matches exactly, confirming that percent change computed from index values is equivalent to percent change from raw data, since the base-period value cancels out algebraically in the ratio.

Question 4: APPLIED

A supply chain analyst has a cost index for raw materials published with a 2018 base year (= 100). The index values are: 2018 = 100, 2019 = 108, 2020 = 95, 2021 = 122, 2022 = 131. Management wants the series rebased to 2020 = 100 because that year represents the pandemic trough. (a) Rebase the entire series to 2020 = 100. (b) What is the percent increase in costs from the new base to 2022?

Show answer

(a) Rebasing formula: I'ₜ = (Iₜ / I_new_base) × 100. The new base is 2020, where the old index = 95. So: 2018 = (100/95)×100 = 105.26; 2019 = (108/95)×100 = 113.68; 2020 = (95/95)×100 = 100.00; 2021 = (122/95)×100 = 128.42; 2022 = (131/95)×100 = 137.89. (b) Percent increase from 2020 to 2022 in the rebased series: ((137.89 − 100) / 100) × 100 = 37.89%. This means raw material costs rose approximately 37.9% from the pandemic trough to 2022. Note that the percent change between any two years is the same regardless of which base is used — rebasing changes index levels but not growth rates between fixed time points.

Question 5: CRITICAL THINKING

A stock drops from $80 to $60 in Year 1 (a 25% decline), then rises from $60 to $80 in Year 2 (a 33.3% gain). An investor constructs a simple index with the original $80 as the base. The index falls to 75.0 at end of Year 1 and returns to 100.0 at end of Year 2. The investor argues: 'The average annual percent change is (−25% + 33.3%) / 2 = +4.17%, so the stock actually grew on average.' Critically evaluate this claim. What is the correct measure of average annual growth, and what does it reveal?

Show answer

The investor's arithmetic mean of percent changes (+4.17%) is misleading because percent changes are multiplicative, not additive. The correct approach is to use the geometric mean return. Over two periods, the cumulative return factor is ($80/$80) = 1.00, meaning zero cumulative gain. The geometric mean annual return is

(1.00)^{1/2}

− 1 = 0%, which accurately reflects that the stock ended exactly where it started. The arithmetic mean overstates performance because it weights the 33.3% gain equally with the 25% loss, ignoring that the gain was computed on a smaller base ($60 vs. $80). This exemplifies the asymmetry of percent changes discussed in the lesson: a 25% loss requires a 33.3% gain to recover, not a 25% gain. For multi-period return analysis, the geometric mean (or equivalently, the compound annual growth rate, CAGR) is the appropriate summary statistic.

SUMMARY

Lesson Summary

Summary

This lesson introduced two foundational tools of descriptive analytics: index numbers and percent change. A simple index number re-expresses any data value as a ratio of a designated base period value, scaled to 100, producing a unit-free metric ideal for cross-series comparison. The formula Iₜ = (Vₜ / V₀) × 100 converts raw magnitudes into relative positions on a standardized scale. Percent change — computed as ((V_new − V_old) / V_old) × 100 — quantifies the proportional shift between any two values, whether raw or indexed.

Key concepts to carry forward include the critical role of base-period selection, the asymmetry of percent changes (a 50% decline requires a 100% gain to recover), and the ability to rebase an index series by dividing all values by the new base's index and multiplying by 100. These simple tools form the conceptual foundation for weighted composite indices (Laspeyres, Paasche, Fisher) and deflation techniques covered in subsequent lessons, where the same ratio logic extends to multi-item baskets with economic weights.