Journal of Student Research 2018

Journal Student Research

48

A map of the markets represented in this study can be found in Figure 2. The data was separated into four regions: West, Midwest, South, and Northeast based on the geographic regions given by the U.S. Census. The number of observations in each region are presented in Table 1.

Region

NE South West MW

# of Observations

38 102 91 34

Table 1: Geographic breakdown of taxi markets Table 1: Geographic breakdown of taxi markets

The official Uber app was used to identify if Uber was present in each taxi market. 3 Each of the 265 observations was manually inputted into the “Request a Ride” search bar in the Uber app. From these search results, I coded an indicator of Uber presence in each taxi market. The market exhibits a “1” in the Uber category if the search results showed an available driver willing to offer a ride. The omitted category is if the search results showed “Unfortunately, Uber is currently unavailable in your area.” A total of 40 taxi markets did not have Uber available, and the remaining 225 taxi markets had Uber available. The Uber and non-Uber markets exhibited many similar characteristics. For example, the average fares for one-mile, five-mile, and 10 mile rides between the two groups do not statistically differ. 4 The differences of means for the two groups are displayed in Table 2. Overall, the two groups reflect similar values for variables, except for 2010 population. The non-Uber group on average had lower population density and a lower population in the observations. It should be noted that the Uber group had several more observations than the non-Uber group, which is shown in Table 2.

no Uber

Uber

Mean

Standard Error Mean

Standard Error

1-mile 5-mile 10-mile

$5.996 $18.07 $32.30 23,370

0.0706 0.1718 0.9889

$5.02

0.0751 0.1814 0.3248

$17.80 $31.42 27,460

3,069

1,292

2010 Population

Table 2: Difference of means test for the Uber and non-Uber groups Table 2: Difference of means test for the Uber and non-Uber groups

3 It should be acknowledged that the presence of Uber could also impact alternate trans portation methods, such as limos and other public transportation services, but this paper focuses solely on the impact on taxis. 4 Although this univariate analysis does not show a difference in fares, I am able to show a difference in fares between the groups when including control variables.