Airport Accessibility and Detroit Passengers

The effect of decreased accessibility on passenger usage of the Detroit airport is investigated using techniques of statistical inference. The purpose is to demonstrate the usefulness, even necessity, of these techniques for the proper analysis of many of the data that arise in transportation engineering problems. The great variability inherent in growth rates for individual city-pairs makes isolation of the accessibility effect at Detroit difficult. However, after certain adjustments to the data, an analysis of variance strongly indicates the presence of an accessibility effect on 1946 to 1947 growth. Conversely, the data do not offer any evidence of further accessibility effects on growth rates in later years. A point estimate of the percentage of passengers lost because of relocation is 28.7%, and 95% confidence limits are 12.4% to 41.9%.