In the sharing economy (dynamic services for consumers for ride sharing, resource sharing for paid part-time work, and so on), the physical geographical area where the consumer resides or uses the service plays a vital role in how the service is accessible to the customer, as the authors of this paper show. The paper creates an appreciable measure of structural inequality in our cities and localities; the conclusions are oftentimes intuitive to understand, but not so easy to quantify. In that respect, the authors have done a wonderful job in scientifically quantifying this implicit understanding. They use example data from Uber and TaskRabbit and analyze its effectiveness for consumers using four key principles from human geography. These measures include: (1) distance decay (when the distance between two locations increases, interaction tends to decrease); (2) structured variations in population density across different metropolitan areas; (3) mental maps; and (4) spatial homophily or residential clustering (also known as big sort).

Through their in-depth analysis, the authors demonstrate that the sharing economy clearly imposes structural geographic biases on service availability, quality, and affordability. It is revealed that these are more effective and successful in some areas (often higher socioeconomic status areas) and are not so successful in those with lower socioeconomic status, or those areas with low population density. They show that in Chicago (for which the data was available), people in poor neighborhoods and outer-ring suburbs wait longer for Uber cars and will often have a harder time finding work help through TaskRabbit. They corroborate that often those areas with lower socioeconomic status correlate strongly with membership in certain protected classes, such as those defined by race and ethnicity, and show correlations that impact underserved populations more negatively.

In addition, the authors also reaffirm their findings through three other approaches, including controlled experiments, qualitative survey response analysis, and the use of spatial Durbin modeling, an advanced geostatistical technique. Through these methods, they also afford a window of revelations into some possible solutions that governments can possibly try to mitigate these biases for their citizenry. For example, they observe that very few TaskRabbit workers (they term them “micro entrepreneurs”) live in low socioeconomic status neighborhoods and hence the long distances to job locations contribute to a higher price point and a decreased willingness by these workers to accept job requests arising from these neighborhoods.

In summary, the paper reaffirms some intuitive generalizations on the cost, accessibility, and affordability of these services from the sharing economy, and its implicit biases to serving those from lower socioeconomic areas. They also note their use of an available targeted dataset, and hence need to study this in the context of other cities across the globe and more generally validate these findings accordingly for those areas.

Finally, this study, although not revealing, confirms the intuitive understanding of how a customer-oriented business operates and scales. The results of the paper must provide more impetus for government agencies to fill the voids created by such unfortunate divisions. Technology, unfortunately, is still not an equalizer for some of mankind’s long-existing problems.