1. Demonstration: Airport Concentration

The methodology used in the exercise is explained in Chapter 4, Method 1. To illustrate the Gini coefficient we will examine the degree of concentration of air traffic among the top 10 airports in the world and among the top 10 airports of Canada.

Source: ACI Airport Statistics. Note values are rounded up to 2 digits.

The Dissimilarity Index (DI) is the sum of the differences between the share of airports and share of passengers (X-Y) multiplied by 0.5, so DI = 0.23*0.5 = 0.115. The Gini Coefficient (GC; calculated in this spreadsheet) is very similar, with a value of 0.138. The distribution has a somewhat uniform Lorenz curve, which indicates no specific concentration of traffic among the largest airports, a reasonable assumption since each is a hub servicing its own market area and is not competing with other large airports. However, an analysis at the national level may reveal a different pattern. In this example, we determine the extent of concentration of traffic at Canada’s top ten airports.

Source: Statistic Canada, 1996.

The Dissimilarity Index is 0.405 (0.81 * 0.5) and the Gini Coefficient is 0.514 with a steeper Lorenz curve. The results of the analysis produce very different coefficients. For the top 10 airports in the world, the Gini Coefficient of 0.138 indicates a very low level of concentration, a result that is confirm visually by the graph. On the other hand there is much more concentration in the Canadian airport system, with a Gini coefficient of 0.514 with a much steeper graph tracing. This is explained by the fact that Toronto, as Canada's major hub, accounts for more than a third of all passenger traffic among the top 10 airports.

2. Exercise: Port Concentration

Consider the following two tables, the first ranking the world's largest container ports and the second ranking North America's largest container ports.

The following tasks are to be completed:

• Using the traffic information, calculate the Dissimilarity Index and the Gini Coefficient the world and North America (results can be found in this spreadsheet).
• Plot the Lorenz Curve for both distributions.
• Using the DI and GC, discuss the level of traffic concentration.
• Discuss the differences (or the similarities) in the traffic concentration for both distributions.
• This exercise can be expanded to include the analysis of several years, a larger number of container ports and/or other regions (such as Western Europe or Pacific Asia). A complete dataset about container ports is available in this spreadsheet.