Discrimination in the Sharing Economy

The sharing economy has largely grown as a secotr of economic activity. Since it fosters economic interactions between peers based on the sharing of individual information it has also opened up potential new avenues for discrimination to appear. The aim of this project is to attempt to measure such discriminatory patterns and to look at how they interact with the social circumstances surrounding them.

The Economic Penalty of Terrorism: Increase in Discrimination Against Arabs and Muslims after Paris Attack

Sander Wagner and Ivaylo Petev

Working Paper

Elevator Pitch

We analyze discriminatory patterns against hosts with typically Arab/Muslim (A/M) names on the online rental market Airbnb. To do so we look at at monthly panel data, summarizing all rental transactions that took place in Paris, the biggest urban Airbnb market of the world, between November 2014 and March 2018. In total we analyze 1,044,022 monthly observations for 72,991 distinct properties in those 41 months.

Our results show consistent discrimination on price with Arab/Muslim hosts asking for lower prices throughout the entire time we observe. In addition to this discrimination on occupancy rates (percentage of days a listing is rented out) emerges directly after the November 2015 terror attacks in Paris.

First we establish the average penalty in terms of price and occupancy rate that hosts pay for having a typically A/M name compared to a typically French name.

Listings of A/M hosts have a price that is on average about 13% lower than French hosts and an occupancy rate that is around 4.5% lower. Once we control for the fact that their listings have characteristics associated with lower prices and are often in less expensive neighbourhoods a price penalty of 6% and an occupancy rate penalty of 3.5% for A/M hosts remains.

Next we looked at how prices and occupancy rates developed over time

The descriptive trends show prices being higher for French than for A/M hosts at all times and dropping for both at a similar magnitude after the Nov. 2015 terror attacks. On Occupancy Rate there is no difference between French and A/M hosts before the attacks, but A/M hosts start renting out their listings less than their French counterparts right after the attack

The coefficients showing the effect of a host being A/M after controlling for the listing characteristics in each month tell a similar story: There is a persistent price penalty for A/M hosts that changes little over time, an occupancy rate penalty is added on to it right after the November 2015 attacks.

Finally, we calculated the monetary loss A/M hosts incurred due to lower prices and occupancy rates.

Our Models show that an A/M host lost about 100-120$ in revenue per month prior to the Nov.2015 attacks and about 180-220$ in revenue per month after the Nov.2015.

Overall we calculate that Arab/Muslim names cost hosts over 6 million $ in revenue in the three and a half years we studied.