Road transport is associated with several types of negative externalities, including traffic accidents, congestion, and local pollution that causes increased mortality. Additionally, car ownership in cities may lead to inefficient land use. On average, private cars have low usage rates (they are parked 96% of the time), an inefficiency that is exacerbated by the increasing single-occupancy vehicle commuting rates.

Several policies are being implemented to mitigate these externalities, including congestion pricing, electric vehicle purchase subsidies, and low-emission zones. To improve the effectiveness of these policies, a deeper understanding of their effects is required on two fronts: a) the benefits and costs of the various policy options and b) their distributional implications (i.e., whether benefits/costs are accrued by lower or higher income individuals).

This study analyses the characteristics of users and usage patterns of a relatively new service: free-floating car-sharing (FFCS). In part thanks to advances in communications and tracking technologies, as well as the widespread adoption of smartphones, companies are now able to offer car rentals by the minute, allowing users to pick up the vehicle and drop it off at any place within a given service area. Importantly, the companies providing these services often do so with electric vehicles.

FFCS can potentially affect congestion and environmental externalities in opposing directions, depending on which mode of transport it is substituting. On the one hand, FFCS may alleviate urban congestion and pollution if it reduces car ownership (car shedding) or the number of trips with private polluting cars. This possibility is supported by shared vehicles being mostly electric and having much higher utilisation rates than private vehicles. On the other hand, car-sharing may increase congestion if used as a substitute for public transport, walking, or cycling. Whether one effect or the other dominates is an empirical question related to the usage and substitution patterns of FFCS.

The research drew on a unique proprietary database that contains the universe of car-sharing trips in 2019 carried out by one of the leading FFCS companies in Madrid. That information was combined with neighbourhood-level data on population, income, car ownership, and public transport network to correlate car-sharing usage patterns with socio-demographic characteristics. Additionally, daily road traffic data was compared against the seasonality of car-sharing trips.

Due to privacy constraints, the residential address of the car-sharing members was not provided by the FFCS company. To overcome this limitation, users’ neighbourhood of residence was imputed by exploiting the frequency with which the same member starts and ends a trip in the same location, which was treated as a proxy for neighbourhood. To achieve the preferred specification, different frequencies, starting times, and days of the week were explored. This made it possible to identify users living in 95 out of the 96 neighbourhoods covered by the FFCS service in Madrid. Furthermore, the hourly frequency of repeated origins and destinations made it possible to infer whether the purpose of each trip was leisure or commuting.

 To provide insight on the impacts of FFCS, car-sharing usage was linked with car ownership and the public transport network. In downtown Madrid, car ownership rates are relatively low, which is likely explained by low parking availability and abundant public transport options. This correlated positively with a low usage of FFCS, with only 7.4% of all car-sharing trips starting or ending in this district. On the contrary, the periphery generally exhibits a high rate of car ownership and a lower coverage of metro and bus stations. Most loyal FFCS customers tend to live in these suburban neighbourhoods. Hence, residents in these areas seem to own a car to counter insufficient public transport options.

This evidence is consistent with FFCS in Madrid being used as a substitute for private vehicles and as a complement for public transport. Given that the car-sharing fleet tends to be fully electric, if car-sharing customers reduce usage of their private car, FFCS could lead to an overall reduction of vehicles in cities and, consequently, to decreased road traffic and improved air quality. Moreover, FFCS and private vehicles differ substantially in usage rates (i.e., the share of hours that the vehicles are actually on the road). The average usage rate of FFCS vehicles is 14.6%, thus considerably higher than that for private vehicles (only 4%). This implies that car-sharing allows for a more efficient use of parking space.

 The seasonality of car-sharing trips closely matched the trends of other motorised vehicles. However, car-sharing peaked earlier. Moreover, while road traffic decreases during summer, car-sharing trips were highest in July and September. Hence, FFCS smoothed road traffic, contributing to reducing overall congestion. A potential reason for this finding is that the car-sharing service is paid for by the minute, encouraging users to avoid periods of intense road congestion

Furthermore, the car-sharing service experienced greater usage during workdays, particularly on Fridays. Nonetheless, most loyal customers were more likely to use the service during the weekend, for leisure purposes, and less likely to use it for commuting. Importantly, most loyal members lived in neighbourhoods with fewer public transport options, which are scarcer during weekends. Even if these users are unable or unwilling to pay for car-sharing for their regular commutes, the car-sharing service is a valuable option in the absence of public transport alternatives.

The FFCS service is only available in high- and middle-income neighbourhoods, reducing the possibility of usage by low-income individuals. Indeed, the average annual net income of car-sharing users in the study sample was close to €20,000, thus higher than the average for the entire municipality of Madrid (€16,700). FFCS might be more affordable than owning a car, but it is still more expensive than public transport. This means that this service might not be accessible or attractive to some income groups.

Moreover, customers who used the service more intensively lived in middle-income neighbourhoods that, as previously mentioned, are characterised by high rates of car ownership and fewer public transport options. In fact, loyalty was a much stronger predictor than income for the probability of taking a car-sharing trip, which suggests that car-sharing usage intensity may not be strongly associated with income but rather with other factors, notably the lack of mobility alternatives.

The patterns of use and the characteristics of the users of FCCS in Madrid depend on several variables that might be time and location specific. For instance, usage depends on the availability of car-sharing vehicles in the city, which directly impacts willingness to use car-sharing. Usage also depends on the availability of other modes of transport, and their associated costs, including the monetary costs and the time involved, which in turn depend on the average traffic conditions and the availability of parking. Therefore, some of these findings might not extend to other municipalities or other times.

Beyond increasing mobility options, FFCS can reduce congestion and emissions in cities as they promote higher utilisation rates of green vehicles. However, to unlock these benefits, users must see FFCS as a complement to public transport that allows them to move away from private vehicles. The present analysis of usage patterns in Madrid has shown descriptive evidence consistent with this hypothesis. Moreover, while not available in lower-income areas, car-sharing benefits users in middle-income areas, who tend to use the service more frequently, probably due to scarcer public transport alternatives. Furthermore, the usage patterns of car-sharing do not perfectly correlate with overall traffic, suggesting that they can smooth congestion.

FABRA, N., C. PINTASSILGO, y M. SOUZA (2023). «Observed Patterns of Use and Users of Free-Floating Car-Sharing», EnergyEcoLab Working Paper 117, Universidad Carlos III de Madrid.