The Gini Coefficient: Measuring Inequality
Building the Foundation: The Lorenz Curve
Before we can understand the Gini coefficient, we must first meet its graphical parent: the Lorenz curve[1]. Imagine you want to visualize how fair the distribution of cookies is in your class of 10 students. You line up the students from the one with the fewest cookies to the one with the most. The Lorenz curve is plotted on a square graph.
The horizontal axis (x-axis) shows the cumulative percentage of people, starting from 0% to 100%. The vertical axis (y-axis) shows the cumulative percentage of the thing being measured, like cookies, income, or wealth, also from 0% to 100%.
Let's create a simple example. Suppose total cookies in class = 100. Here is the distribution:
| Student (from poorest to richest) | Cookies Owned | % of Cookies | Cumulative % of People | Cumulative % of Cookies |
|---|---|---|---|---|
| First 8 students | 5 each (40 total) | 5% each | 10%, 20%, ... 80% | 5%, 10%, ... 40% |
| 9th student | 20 | 20% | 90% | 60% |
| 10th student (richest) | 40 | 40% | 100% | 100% |
Now, plot the points: (0%,0%), (10%,5%), (20%,10%), (30%,15%), (40%,20%), (50%,25%), (60%,30%), (70%,35%), (80%,40%), (90%,60%), (100%,100%). The line connecting these points is the Lorenz curve. It always starts at (0,0) and ends at (100,100).
The Lorenz curve bows away from this diagonal line. The more it bows, the more unequal the distribution. The area between the diagonal (line of equality) and the Lorenz curve is the key visual clue for inequality.
From Curve to Number: Defining the Gini Coefficient
The Gini coefficient, named after the Italian statistician Corrado Gini, takes the visual idea of the Lorenz curve and calculates a single number between 0 and 1 (or 0% and 100%).
Look at the graph with the diagonal line (perfect equality) and the bowed Lorenz curve. Let's label two areas:
- A: The area between the line of equality and the Lorenz curve.
- B: The area under the Lorenz curve.
The Gini coefficient ($G$) is defined as:
$G = \frac{A}{A + B}$
Since the area of the entire triangle under the line of equality is $A + B$, the Gini coefficient is the proportion of that total area that represents inequality.
• $G = 0$: Perfect equality. The Lorenz curve is the diagonal line, so area $A = 0$.
• $G = 1$ (or 100%): Perfect inequality. One person has everything, and everyone else has nothing. The Lorenz curve runs along the bottom axis and then shoots up at the very end, making area $B = 0$ and $A$ the entire triangle.
In our cookie example, we can approximate the Gini coefficient. The area under the Lorenz curve (B) can be estimated using the trapezoid rule[2]. The total area under the line of equality (A+B) for a 100% x 100% square is 0.5 (since it's a triangle with area $\frac{1}{2}*1*1$). A detailed calculation for our data gives a Gini coefficient of approximately 0.28, or 28%.
Interpreting the Gini Coefficient in the Real World
What does a Gini coefficient of 0.28, 0.40, or 0.60 mean? Here is a general guide to interpreting these values when applied to a country's income distribution:
| Gini Range | Interpretation (Income) | Example Countries (Approx.) |
|---|---|---|
| 0.24 – 0.35 | Low inequality. Relatively even distribution. | Slovenia (0.24), Czech Republic (0.25) |
| 0.36 – 0.49 | Moderate inequality. Typical for many developed and developing nations. | Canada (0.33), United Kingdom (0.35), United States (0.41) |
| 0.50 – 0.59 | High inequality. Large gaps between rich and poor. | China (0.47), Brazil (0.53) |
| 0.60 and above | Very high inequality. Extreme concentration of resources. | South Africa (0.63) |
It's crucial to remember that a single number cannot tell the whole story. A country with a Gini of 0.40 could have a strong middle class, or it could have a vast poor population and a small ultra-rich elite. The coefficient summarizes complexity, so it should always be considered alongside other data.
Practical Application: A Classroom Experiment
Let's apply the Gini coefficient to a concrete, hands-on example you can try yourself. Imagine a teacher wants to analyze the fairness of "participation points" awarded over a month. There are 5 students in a group. The teacher records their points:
Student Points: Alex: 10, Bailey: 15, Casey: 20, Dakota: 25, Emerson: 30.
Total Points: 100.
Step 1: Create the Lorenz Curve Data. Sort students from least to most points and calculate cumulative percentages.
| Student | Points | % of Points | Cumulative % of People | Cumulative % of Points |
|---|---|---|---|---|
| Alex | 10 | 10% | 20% | 10% |
| Bailey | 15 | 15% | 40% | 25% |
| Casey | 20 | 20% | 60% | 45% |
| Dakota | 25 | 25% | 80% | 70% |
| Emerson | 30 | 30% | 100% | 100% |
Step 2: Calculate the Gini Coefficient. We can use a common computational formula that avoids directly measuring areas:
$G = \frac{\sum_{i=1}^n \sum_{j=1}^n |x_i - x_j|}{2 n^2 \bar{x}}$
Where $x_i$ are the individual points, $n$ is the number of people (5), and $\bar{x}$ is the mean points (20). Calculating all absolute differences and summing them gives a value. Plugging in the numbers:
$ \sum |x_i - x_j| = (5+10+15+20) + (5+5+10+15) + (10+5+0+5) + (15+10+5+0) + (20+15+10+5) = 200 $
$G = \frac{200}{2 * 5^2 * 20} = \frac{200}{1000} = 0.20$
So, the Gini coefficient for this classroom group is 0.20 or 20%. This indicates a relatively equal distribution of participation points, though not perfectly equal (which would be 0). The teacher might use this to see if their award system is fair or if some students are consistently left behind.
Important Questions
The Gini coefficient has several important limitations. First, it only measures relative, not absolute, inequality. A country where everyone is poor can have the same Gini (low inequality) as a country where everyone is rich. Second, it is sensitive to changes in the middle of the distribution but can be less sensitive to changes at the very top or bottom. Third, two very different Lorenz curves can theoretically produce the same Gini coefficient, a situation known as Lorenz curve crossing. Finally, it does not tell us anything about the overall wealth or income level of a society, only how it is shared.
No, under standard definitions and when using the area-based formula $G = A/(A+B)$, the coefficient is strictly between 0 and 1. Since area A cannot be larger than the total area under the line of equality (A+B), the ratio cannot exceed 1. It is often expressed as a percentage between 0% and 100%.
The Gini coefficient is versatile. It can measure inequality in the distribution of wealth, land ownership, educational attainment, access to healthcare resources, or even internet bandwidth usage among users. Any resource or characteristic that can be quantified and distributed among a population can, in principle, have its inequality summarized by a Gini coefficient.
Conclusion
Footnote
[1] Lorenz curve: A graphical representation of the distribution of income or wealth, plotting cumulative share of people against the cumulative share of the resource they hold. It was developed by the American economist Max O. Lorenz in 1905.
[2] Trapezoid rule: A mathematical method for approximating the area under a curve by dividing it into a series of trapezoids and summing their areas. Useful for estimating area $B$ under the Lorenz curve when you have data points.
Gini coefficient: A numerical measure of statistical dispersion intended to represent the income or wealth inequality within a nation or any other group of people. It is defined as $G = A/(A+B)$ based on the Lorenz curve.