What is the role of carsharing toward a more sustainable transport behavior? Analysis of data from 80 major German cities

Figures & data

Figure 1. Cities investigated by number of inhabitants and classification as metropolis according to RegioStaR7 (BMVI, 2020c).

Table 1. User groups used as dependent variable categories.

User groups	Corresponding share of environmentally friendly modes (walking, cycling, public transit)	Group of sample 2013	Group of sample 2015	Group of sample 2017
Reference group	First quartile (0%–25%)	0%–22.5%	0%–13.2%	0%–15.5%
Low eco	Second quartile (25%–50%)	22.5%–55.1%	13.2%–45.9%	15.5%–48.3%
Medium eco	Third quartile (50%–75%)	55.1%–91.5%	45.9%–87.1%	48.3%–85.9%
High eco	Fourth quartile (75%–100%)	91.5%–100%	87.1%–100%	85.9%–100%

Table 2. Summary of the independent variables used for the analysis.

Explaining variables	Description	Characteristics	Source
Sociodemographic characteristics
Age	Age of the respondent in years	Cardinally scaled variable	BMVI (2020a)
Gender	Gender of the respondent as dummy variable	1 = male, 0 = female	BMVI (2020a)
Income	Household income of the respondent, reference group is the middle one	Low < 2,000 , Middle ≥ 2,000–< 4,000, High ≥ 4,000	BMVI (2020a)
Education	Educational background of the respondent	Low: main school or middle school Middle: Higher education entrance or vocational training High: University degree	BMVI (2020a)
Individual mobility resources
Driver’s license	Respondent owns a driver’s license	1 = yes, 0 = no	BMVI (2020a)
Number of cars in household	Numbers of privately owned cars in household	Cardinally scaled variable	BMVI (2020a)
Public transit pass	Respondent owns a transit pass	1 = yes, 0 = no	BMVI (2020a)
Bicycle ownership	Respondent owns a bicycle	1 = yes, 0 = no	BMVI (2020a)
Carsharing membership	Respondent is carsharing member	1 = yes, 0 = no	BMVI (2020a)
City/supply characteristics
Train stop	Train stop in walkable distance	1 = yes, 0 = no	BMVI (2020a)
Bus stop	Bus stop in walkable distance	1 = yes, 0 = no	BMVI (2020a)
Tram stop	Tram stop in walkable distance	1 = yes, 0 = no	BMVI (2020a)
Subway stop	Subway stop in walkable distance	1 = yes, 0 = no	BMVI (2020a)
Urban railway stop	Urban railway stop in walkable distance	1 = yes, 0 = no	BMVI (2020a)
Carsharing supply density	Number of carsharing cars per 1,000 inhabitants	Cardinally scaled variable	BCS (2013, 2015, 2017)
Metropolis	City is classified as metropolis according to RegioStaR7	1 = yes, 0 = no	BMVI (2020c)

Table 3. Five highest supply densities of station-based carsharing in 2019 (Data from 2013 and 2019).

City	Shared vehicles per 1,000 inhabitants 2019 (2013 in brackets)	City	Shared vehicles per km² 2019 (2013 in brackets)
Karlsruhe	2.96 (1.93)	Karlsruhe	5.35 (3.31)
Freiburg	1.59 (0.50)	Freiburg	2.38 (0.98)
Heidelberg	1.04 (0)	Munich	2.36 (1.94)
Darmstadt	0.89 (0)	Stuttgart	2.01 (1.91)
Göttingen	0.88 (0)	Frankfurt/Main	1.84 (1.43)

Table 4. Five highest supply densities of free-floating carsharing in 2019 (Data from 2013 and 2019).

City	Shared vehicles per 1,000 inhabitants 2019 (2013 in brackets)	City	Shared vehicles per km² 2019 (2013 in brackets)
Munich	1.63 (0.44)	Munich	7.72 (1.93)
Berlin	1.43 (0.67)	Berlin	5.84 (2.65)
Hamburg	1.40 (0.39)	Hamburg	3.41 (0.93)
Düsseldorf	0.97 (0.16)	Düsseldorf	2.76 (2.3)
Cologne	0.82 (0.69)	Stuttgart	2.27 (2.17)

Figure 2. Total number of station-based and free-floating carsharing vehicles in Germany (own display based on BCS, 2013, 2015, 2017, 2019).

Table 5. Sociodemographic characteristics for the groups of carsharing members and nonmembers (Data from 2017).

Age
		N (individuals)	18–25 y/o	26–35 y/o	36–50 y/o	51–60 y/o	61–70 y/o	>70 y/o
MiD	Members	4,142	6.6%	33.4%	36.6%	15.7%	6.3%	1.4%
	Nonmembers	39,619	7.5%	13.3%	23.0%	20.9%	17.8%	17.5%
	p value		0.015*	0.000***	0.000***	0.000***	0.000***	0.000***
MOP	Members	39	7.7%	41.0%	23.1%	20.5%	5.1%	2.6%
	Nonmembers	779	3.0%	9.1%	20.5%	27.1%	21.7%	18.6%
	p value		0.122	0.000***	0.415	0.240	0.006**	0.003**
Monthly household net income
		N (households)	<900	900–<2,000	2,000–<3,000	3,000–<4,000	4,000–<5,000	>5,000
MiD	Members	3,717	1.0%	6.3%	17.2%	21.4%	18.8%	35.3%
	Nonmembers	30,920	2.5%	13.6%	24.7%	22.0%	15.6%	21.5%
	p-Value		0.000***	0.000***	0.000***	0.187	0.000***	0.000***
MOP
	Members	32	3.1%	9.4%	31.3%	18.8%	15.6%	21.9%
	Nonmembers	543	6.1%	23.0%	24.5%	23.2%	10.9%	12.3%
	p-Value		0.421	0.048*	0.253	0.369	0.278	0.103
Education^†
		N (individuals)	No certificate	Primary school	Middle school	Higher education entrance	University degree
MiD	Members	4,142	0.1%	1.3%	7.4%	17.6%	72.6%
	Nonmembers	39,619	1.0%	11.8%	20.3%	19.8%	45.4%
	p-Value		0.000***	0.000***	0.000***	0.000***	0.000***
MOP	Members	39	0.0%	5.1%	10.3%	15.4%	69.2%
	Nonmembers	779	0.0%	16.9%	24.6%	16.0%	41.5%
	p-Value		–	0.032*	0.025*	0.563	0.001***
Residence in RegioStaR7 Category Metropolis
				N (individuals)			Metropolises [71]
MiD	Members			4,142			80.3
	Nonmembers			39,619			53.6
	p-Value						0.000***
MOP	Members			39			71.8
	Nonmembers			779			56.1
	p-Value						0.037*

Asterisks indicating significance: p < 0.001***, p < 0.01**, p < 0.05*.

^†Remainder to 100% is “other/not specified”-category.

Table 6. Undertaken trips by carsharing members and nonmembers including distribution across transport modes (Data from 2017).

		N (trips)	By foot	Bicycle	MIT	Public Transit
MiD	Members	13,716	23.7%	23.8%	30.9 %	21.7 %
	Nonmembers	116,990	23.0%	15.2 %	48.4 %	13.3 %
	p-Value		0.0847	0.0000***	0.0000***	0.0000***
MOP	Members	998	28.7%	24.4 %	23.4 %	23.2 %
	Nonmembers	18,096	22.9%	13.0 %	49.2 %	14.7 %
	p-Value		0.0000***	0.0000***	0.0000***	0.0000***

Asterisks indicating significance: p < 0.001***, p < 0.01**, p < 0.05*. Remainder to 100% is “other/not specified”-category.

Table 7. Model results and odds-ratios of multinomial logit regression.

	2013 (n = 480)			2015 (n = 688)			2017 (n = 763)
Group (% of sample)	Low eco (22.5%–55.1%)	Medium eco (55.1%–91.5%)	High eco (91.5%–100%)	Low eco (13.2%–45.9%)	Medium eco (45.9%–87.1%)	High eco (87.1%–100%)	Low eco (15.5%–48.3%)	Medium eco (48.3%–85.9%)	High eco (85.9%–100%)
Intercept	2.046	14.326	24.251	0.415	2.485	12.133	0.178	1.948	9.782
Sociodemographic data
Age	1.008	1.021	1.023	1.005	0.990	0.987	0.999	1.009	0.999
Male	0.874	0.611	0.573	1.103	0.769	0.644	1.001	0.678	0.658
Income low	1.073	1.087	1.939	0.775	0.861	1.397	0.574	0.556	0.639
Income high	0.866	0.776	1.199	1.619	1.629	0.892	0.894	1.339	1.730
Education low	0.541	0.396*	0.467	0.608	0.900	0.532	2.223*	1.062	1.593
Education high	0.649	0.660	0.633	0.654	1.005	0.764	2.169*	1.768	2.352
Mobility tools
Driver’s license	0.761	0.082*	0.076*	0.882	0.540	0.207	1.224	0.094*	0.055
Number of cars in household	0. 648*	0.254***	0.025***	0.597**	0.221***	0.025***	0.993	0.516***	0.023***
Public transit pass	4.129***	18.727***	13.669***	6.529***	24.525***	19.798***	2.567*	6.603***	8.221***
Bicycle ownership	2.500**	1.921	3.384*	1.286	2.918***	5.553***	1.286	2.438**	2.706**
Carsharing membership	/	/	/	/	/	/	2.727	2.753	1.085
City/Supply characteristics
Train stop	0.800	1.345	1.537	1.270	1.415	1.892	0.996	0.937	2.093*
Bus stop	0.598	0.821	0.847	3.186*	1.398	1.612	2.023	1.680	2.003
Tram stop	0.722	0.778	0.657	1.294	0.852	1.240	1.378	1.099	2.292*
Subway stop	0.912	1.823	2.091	0.953	0.568	1.103	0.953	1.511	1.699
Urban railway stop	1.499	1.689	1.518	1.937*	2.268**	2.115***	1.348	1.823*	1.807
Carsharing supply density (1,000 inhab.)	1.258	0.877	0.792	1.568*	2.337	1.601	0.790	1.459	1.505
Metropolis	0.814	0.672	0.792	0.523*	0.775	1.096	1.015	0.684	0.680
	McFadden’s Pseudo-R^2: 0.306			McFadden’s Pseudo-R^2: 0.314			McFadden’s Pseudo-R^2: 0.285
	LL-null: −665.42			LL-null: −953.77			LL-null: −1057.74
	LL: −461.65			LL: −654.67			LL: −755.69
	LLR: 407.54***			LLR: 589.20***			LLR: 604.09***

Asterisks indicating significance: p < 0.001***, p < 0.01**, p < 0.05*; LL-null: Value of constant-only log-likelihood, LL: Log-likelihood of the model, LLR: Likelihood-ratio Chi²-statistic.

LLR p-value: Chi²-probability of getting a log-likelihood statistics greater than LLR.

BMVI. (2020c). Regionalstatistische Raumtypologie (RegioStaR). Retrieved Accessed August 27, 2020, from https://www.bmvi.de/SharedDocs/DE/Artikel/G/regionalstatistische-raumtypologie.html.

 Google Scholar

BMVI. (2020a). German Mobility Panel (MOP): Longitudinal study on the travel behavior of the population. Federal Ministry of Transport and Digital Infrastructure (BMVI). Retrieved August 28, 2020, from http://mobilitaetspanel.ifv.kit.edu/english/index.php. Data repository: https://daten.clearingstelle-verkehr.de/192/.

 Google Scholar

BCS. (2013). CarSharing Städteranking 2013. Bundesverband Carsharing.

 Google Scholar

BCS. (2015). CarSharing Städteranking 2015. Bundesverband Carsharing.

 Google Scholar

BCS. (2017). CarSharing Städteranking 2017. Bundesverband Carsharing.

 Google Scholar

BCS. (2019). CarSharing Städteranking 2019. Bundesverband Carsharing.

 Google Scholar