Improving the transparency of the sharing economy

The idealistic beginnings of the sharing economy made ways to an entrenched battle to win over the public opinion and for law makers to appreciate its benefits and its risks. The stakes are high as the success of services like Airbnb reveals that under-utilized assets (e.g. spare rooms or apartments left vacant) can be efficiently matched to individual demands to generate a significant surplus to their owners. Rules and regulation, which are increasingly felt as necessary by many communities, also create friction over the best way to leverage these opportunities for growth. To make things worse, the sharing economy is complex and poorly documented: Three recent reports from public institutions and lobbying groups arrived at opposite conclusions with seemingly contradictory facts about the occupancy distribution. In this paper, we show how to overcome this opacity by offering the first large-scale, reproducible study of Airbnb’s supply and transactions. We devised and deployed frequently repeated crawls using no proprietary data. We show that these can be used to accurately estimate not only the supply of available rooms, but the effective transactions, occupancy, and revenue of hosts. Our results provide the first complete view of the occupancy and the distribution of revenue, revealing important trends that generalize previous observations. In particular we found that previous observations that seemed at odds are all explained by a variant of the “inspection paradox”. We also found from our detailed data that enforcing a maximum occupancy of 90 nights a year would greatly reduce most concerns raised by various advocacy groups, while affecting only marginally the justifying claims that Airbnb quotes to argue for its beneficial impact.