1 Introduction

The automotive industry will have to navigate through a number of difficult challenges in the upcoming years. With advancing climate change, the pressure to reduce CO2 emissions is rising. Due to urbanisation and growing population, cities around the world are fighting with rising traffic, declining air quality and limited space availability (Gross 2019). Stricter regulations regarding emissions, restrictions for private car use in cities and limited parking options are only some of the consequences. As a result, private car ownership becomes less attractive in urban areas. In addition, greater health awareness and concern for the environment support this movement away from the personal car (Morrison and Beer 2017). Considering that more than 50% of the world’s population live in cities (United Nations 2011), the automotive industry is seeking alternative business models to address the challenges that result from urbanisation. Numerous new passenger transportation options, collectively called mobility services, have been developed over the past 15 years. These mobility services have been identified as one possible option to address the current challenges (Gould et al. 2015). Mobility services are user-centric and responsive to the needs, habits and preferences of travellers and society. They enable users to have access to transportation services for short- and long-term periods and on demand. Mobility services often blur the lines between public transportation and private ownership (Kamargianni et al. 2016). Car-sharing, ride-hailing, ride-sharing, microtransit and bike-sharing are some examples of mobility services currently being developed. Each has its own underlying service characteristics and business model (Spulber et al. 2016). Hietanen (2014) gave one of the first comprehensive definitions of mobility as a service (MaaS). He describes MaaS as a mobility distribution model that delivers users’ transport needs through a single interface of a service provider. It combines different transport modes to offer a tailored mobility package. Burrows et al. (2015) defines MaaS as a new way to provide transport, which facilitates the users to get from A to B by combining available mobility options and presenting them in a completely integrated manner. The most cited definition for MaaS, however, is from the European MaaS Alliance, which defines the concept as “the integration of various forms of transport services into a single mobility service accessible on demand” (Durand et al. 2018). Despite the novelty of the MaaS concept and the challenge to describe its nature, a set of core characteristics can be derived from literature review, such as personalisation, customisation, tariff options, real-time information, trip planning, booking and ticketing as well as employed technologies like GPS, E-ticket and E-payment (Jittrapirom et al. 2017). According to these definitions, the unifying feature of any MaaS implementation is the integration of multiple mobility services. This means that car-sharing, bike-sharing, ride-hailing, etc., are not themselves MaaS without meaningful multi-modal integration. Car manufacturers (such as Daimler, BMW or Volkswagen) identified mobility services and MaaS as additional business opportunity, especially in urban areas, and are involved in car-sharing or ride-hailing operations. The years 2018 and 2019 were full of merges, acquisitions and new players in the field of mobility services. In the past 3 years, close to 1000 cities have added car-sharing alone, which is an increase of 47% (Philipps 2019). Further, scooter and bike-sharing services are growing in numbers. Two types of scooter-sharing systems can be distinguished: standing electric and moped-style scooters. In Taiwan, for example, the number of moped-style electric scooters increased ten times from 2016 to 2018, mainly due to battery-sharing systems (Pham et al. 2019). The standing electric scooters, e.g. operated by Bird or Lime, can be found in over 100 metropolitan areas around the globe. The major standing scooter–sharing service in the USA, Bird, launched its service in over 50 cities across Europe in 2019 (Shaheen et al. 2020). However, not only scooter- and car-sharing services are growing rapidly, also bike-sharing is on the rise. According to Moon-Miklaucic et al. (2019), approximately 95% of the 1600 bike-sharing operators in 2017 were launched since 2007, with more than 200 in 2017 alone.

As different mobility services are growing in cities around the world, the question that automatically arises is whether mobility services can lead to more sustainable transportation options, thereby improving quality of life in cities. In order to answer this question, it is important to analyse sustainability impacts of these mobility services in a systematic way, considering all three dimensions of sustainability: economic, ecologic and social. This is essential to avoid burden shifting; for example, an improvement regarding ecological impacts could lead to negative social impacts. Life cycle–based methodologies have been developed over time for this purpose (Curran 1996; ISO 14040 2006; Finkbeiner et al. 2010). Although a lot of research has been done concerning economic and environmental assessment, a standardised approach for social life cycle assessment (S-LCA) has yet to be agreed on (e.g. Dubois-Iorgulescu et al. 2016). Different S-LCA indicators and impact assessment methods have been applied and tested in a number of case studies (Di Cesare et al. 2018). The use phase plays an important role for the assessment of mobility services; for example, shared cars are used in a very different way than conventional cars in private ownership. The evaluation of the use phase, however, has been underrepresented in previous S-LCA case studies (Petti et al. 2016). Therefore, it is important to focus on use phase social impacts, not only to have a consistent assessment along all three dimensions of sustainability and to avoid burden shifting but also to be able to measure positive impacts. The assessment of positive impacts is crucial when analysing the impacts of mobility services; for example, an improved access to mobility, a higher degree of security or affordability should also be captured. For a deeper understanding of the study and more background information, the methodology of social life cycle assessment is introduced (Section 1.1) before analysing mobility services in the context of S-LCA and use phase assessment (Section 1.2).

1.1 Social life cycle assessment, a methodological description

S-LCA is a rather new approach compared to life cycle assessment (LCA) and life cycle costing (LCC), and it aims to address the social impacts of goods and services along their life cycle. The starting point is the established methodology of LCA according to the ISO 14040 (2006). About 26 years ago, a “social welfare impact category” was proposed in the SETAC Workshop Report (1993): “A Conceptual Framework for Life Cycle Impact Assessment”. This started the discussion on how to deal with social and socio-economic criteria in assessing a product along its life cycle. The main topic of discussion during the second half of the 1990s was to what extent a life cycle assessment of a product or a service taking into account social criteria is different from LCA. Just like LCA, it was proposed to conduct S-LCA in line with ISO 14040 (2006). A crucial differentiation point for S-LCA is that impact categories may be assessed using different indicators depending on the stakeholder group under consideration.

An important achievement in the ongoing development of S-LCA was the issuing of the UNEP/SETAC S-LCA Guidelines in 2009. According to the UNEP/SETAC Guidelines 2009 “a social and socio-economic Life Cycle Assessment (S-LCA) is a social impact (and potential impact) assessment technique that aims to assess the social and socio-economic aspects of products and their potential positive and negative impacts along their life cycle” (UNEP/SETAC 2009). The UNEP/SETAC Guidelines (also referred to as the Guidelines hereafter) provide a good framework on how S-LCA should be conducted. The framework includes five stakeholder categories as well as six impact categories with 31 subcategories, an inventory analysis and, finally, the impact assessment. The categories identified are the following: human rights, working conditions, health and safety, cultural heritage, governance and socio-economic repercussions. The five stakeholder categories are comprised of workers, local community, society, value chain actors and consumers. The five stakeholder categories have eight, eleven, three, four and five subcategories, respectively. These subcategories are characterised with the help of more than 100 inventory indicators and published in a separate document: “The methodological sheets for subcategories in social life cycle assessment” (UNEP/SETAC 2013). However, these indicators and the corresponding reference to required raw data are only suggestions. The user can select relevant indicators and data sources, e.g. generic or specific data sources.

After the publication of the Guidelines, many S-LCAs were conducted according to the given framework (Petti et al. 2016). Though, among other shortcomings, no clear methodology for the impact assessment is given in the Guidelines: “In the impact assessment step of an S-LCA the distribution and share of positive and negative impacts is important to consider. How this should be done is an issue of further research and testing in case studies” (UNEP/SETAC 2009). This is the reason why the S-LCA research community have used various methods for the impact assessment (Russo Garrido et al. 2016). In general, the impact assessment methods can be classified into two broad categories: the performance reference point methods (type I) and the impact pathway methods (type II), as shown in Fig. 1 (Parent et al. 2010; Chhipi-Shrestha et al. 2014).

Fig. 1
figure 1

Performance reference point–based methods (type I) and impact pathway–based methods (type II)

Performance reference point methods assess social impacts using performance reference points based on minimum performance levels (Franze and Ciroth 2011; Ramirez et al. 2014). For these kinds of minimum performance levels, internationally accepted standards are used such as International Labour Organization (ILO) Conventions, the ISO 26000 Guidelines and OECD Guidelines for Multinational Enterprises. Colour-coding, scoring and weighting systems are used for aggregating the inventory indicator data to impact categories (e.g. human rights). These approaches do not use cause-effect chains because the authors are of the opinion that “cause-effect relationships are not simple enough or not known with enough precision to allow quantitative cause-effect modelling” (UNEP/SETAC 2009). These approaches are included as type I impact assessment by UNEP/SETAC Guidelines.

Impact pathway methods, on the other hand, assess the social impacts by means of impact pathways as characterisation models. Midpoint indicators and/or endpoint indicators are used, comparable to LCA. These methods are based on social effects and mostly on quantitative indicators. Unlike performance reference point methods, for these methods, cause-effect chains are used (Weidema 2018). They are included in type II impact assessment by the UNEP/SETAC Guidelines. Figure 1 shows both performance reference point–based methods (type I) and impact pathway–based methods (type II).

Another important document is the Handbook for Product Social Impact Assessment (PSIA), first published in 2014 and last updated in 2018. The PSIA is a more practical approach, developed mostly by industry leaders but started from the main scientific literature. Those were analysed and compared against the company strategy to identify key social topics and stakeholder categories. The PSIA refers to four main stakeholder categories: workers, small-scale entrepreneurs, local communities and users. The PSIA defines social topics for every stakeholder group. Social topics are social areas related to stakeholder groups such as working hours, community engagement or child labour. For every social topic, a performance indicator is defined. As the name suggests, performance indicators measure the performance of each social topic, for example the number of working hours per week or minimum wage paid. The performance indicators can be quantitative or qualitative. Two different types of metrics can be used: either full quantification of all data or a reference-scale approach. For the reference-scale approach, a score ranging from − 2 to + 2 is proposed. This allows to measure positive as well as negative impacts. The impact assessment method described in the PSIA suggests aggregation of performance indicators into social topic scores, stakeholder scores and a total score (Fontes et al. 2016; Goedkoop et al. 2018).

The presented social topics in the PSIA are often overlapping with the subcategories of the Guidelines. The respective indicators, however, differ greatly and target different issues, as can be seen in Table 1. For example, regarding the stakeholder group local community, both the Guidelines and the PSIA suggest indicators to assess impacts on local employment. Nevertheless, the Guidelines, in addition, include indicators that measure access to material and immaterial resources, which are missing in the PSIA. On the other hand, the PSIA includes indicators regarding skill development, which is not included in the Guidelines. As the stakeholder group consumers, respectively users, is of importance for use phase assessment of mobility services, a focus lies on this stakeholder group. According to the Guidelines, the stakeholder group consumer covers “end-consumers as well as the consumers who are part of each step of the supply chain” (UNEP/SETAC 2009). The PSIA, however, is referring to the term ‘user’ and makes a distinction between products developed for consumers and products developed for workers, which can be seen as professional users. In that way, a distinction between primary, secondary and passive users is made. In the example of a public bus transport, the differentiation of these three kinds of users is explained. A person taking the bus from A to B is a primary user. However, the driver or the person who cleans the bus can be seen as a secondary user. Further, people who do not use, drive or maintain the bus may also be impacted, for example by exposure to potential noise or congestion. This last group of stakeholders is referred to as passive users. Therefore, the definition by PSIA is broader as it includes consumers, workers and passive users (Goedkoop et al. 2018). When analysing the stakeholder group consumers (Guidelines) and users (PSIA) in detail, some similarities as well as major differences can be found. It is notable that in the Guidelines and in the PSIA, ‘privacy’ is one of the topics that overlap. The suggested indicators, however, vary to a great extent. Whereas the Guidelines suggest indicators focusing on country rankings or strength of policies regarding privacy, the PSIA lies the focus on company implementations concerning privacy. The Guidelines, in addition, include the number of consumer complaints or complaints by regulatory bodies related to privacy as quantitative indicator, whereas the PSIA includes scale-based indicators assessing the company’s performance, without stating any quantitative indicators. Other categories show similar variations. ‘Health and safety’ are combined into one category in the Guidelines, whereas the PSIA separates this category into two social topics. The Guidelines include the subcategories ‘Feedback mechanism’, ‘Transparency’ and ‘End-of-life responsibility’, which are missing in the PSIA. The PSIA, though, includes ‘Responsible communication’, ‘Inclusiveness’ and ‘Effectiveness and comfort’, which are not included in the Guidelines. Consequently, the stated indicators for those subcategories and social topics are not overlapping and assess different aspects. In Table 2, a full list of suggested indicators is presented according to the Guidelines and the PSIA, which reveals the different indicators in detail. The major differences in the presented categories and indicators underline that no standardised approach has been defined for category and indicator selection and highlights the necessity of further research regarding the selection of categories and indicators for the assessment of mobility services.

Table 1 Subcategories and social topics as presented in the Guidelines and in the PSIA
Table 2 Full list of suggested indicators for the stakeholder groups consumer (Guidelines) and user (PSIA)

1.2 Mobility services in the context of S-LCA and use phase assessment

Social welfare is considered one of the main development goals of modern society. Understanding and assessing what could improve or undermine well-being is a key element in public policies. In the last years, increasing concern about urban mobility can be observed through scientific and non-scientific literature. Sustainable urban mobility has become one of the main challenges. According to Gould et al. (2015), mobility services have the potential to improve quality of life in cities by reducing the use of private cars and encouraging the diffusion of electric vehicles within cities. Further, mobility services lead to increased mobility options that may result in time saved by users or higher comfort due to, for example, closer pick-up locations (Karlsson et al. 2019). Ride-hailing services, like Uber or Lyft, claim higher safety for their users compared to traditional taxi services due to online tracking and user review options. On the other hand, rising numbers of accidents with shared electric scooters suggest higher health and safety risks for its users (Blomberg et al. 2019).

In the automotive industry, strict health and safety requirements exist, which are monitored and controlled with respective simulations and crash tests. Interior emissions, for example, are tracked and measured to ensure drivers’ and passengers’ health, not to mention emissions from the internal combustion engine. However, mobility services take social impacts to a new level, as the stated examples emphasise. To be able to better understand the implications of mobility services and to improve quality of life in cities, it is therefore essential to evaluate the social impacts in a systematic way, including all related stakeholder groups. In doing so, it is necessary to be able to measure potential positive impacts, for example time saved or increased comfort.

Although the UNEP/SETAC Guidelines as well as the PSIA are applicable for both products and services, most applications focus on products. Petti et al. (2016) systematically investigated S-LCA case studies that have been conducted according to the UNEP/SETAC Guidelines between 2010 and 2015. Among the analysed case studies, 32% used system boundaries from ‘cradle to grave’ and 3% from ‘gate to grave’, making it only 35% of the studies that consider use phase impacts. In addition, only 7% of the studies conducted according to the UNEP/SETAC Guidelines consider the stakeholder group ‘consumers’. Zanchi et al. (2018) analysed S-LCA applications in the automotive sector. Out of the thirteen identified S-LCA applications in the automotive sector, only four considered the stakeholder group ‘local community’ and none of the analysed S-LCA applications considered the stakeholder group consumers. Tarne et al. (2017) focused on LCSA in the automotive industry and thereby analysed S-LCA studies. A comparatively low maturity of S-LCA in the automotive sector was identified. Whereas most of the applications in the automotive sector focus on materials and automotive parts, none of the analysed studies applied the S-LCA framework to mobility services (Zanchi et al. 2018). Consequently, the question is whether the suggested indicators in the Guidelines and the PSIA are adequate for the assessment of mobility services, especially the ones recommended for the stakeholder group consumers and users (Table 2).

Indicators help to transfer complex issues into understandable figures and allow to monitor goals as well as to measure targets. They can establish a common language to communicate impacts and support in decision-making. The choice of indicators is therefore essential as it directly affects management and decision-making. Nevertheless, choosing the right indicators is challenging, as hundreds of indicator systems are available, typically each developed for a specific purpose (Huovila et al. 2019). Considerable efforts have been made to develop indicators that can be used to assess urban mobility, often in the context of measuring the sustainability of a city as a whole. However, these indicators do not fulfil the specific requirements to measure the impact of mobility services. For this reason, a systematic literature review was carried out with a focus on social indicators that allow the assessment of the use phase impacts of mobility services. The indicators were analysed and allocated to stakeholder groups in order to identify hotspots as well as weaknesses. Finally, the indicators are compared to the ones suggested in the Guidelines and the PSIA in order to lead the way towards a comprehensive and inclusive set of indicators for the assessment of mobility services.

2 Methods

For the systematic literature review, Web of Science, ScienceDirect and Springer Link were used. As search strings, multiple notations of ‘life cycle assessment’, ‘S-LCA’, ‘sustainable urban mobility’, ‘social sustainability’ or ‘sustainable urban transportation’ were used in combination with the following keywords:

  • (LCA OR ‘Life Cycle Assessment’ OR ‘Life-cycle assessment’ OR ‘Life-cycle-assessment’) AND ‘Mobility’ AND ‘Mobility Service*’

  • (LCA OR ‘Life Cycle Assessment’ OR ‘Life-cycle assessment’ OR ‘Life-cycle-assessment’) AND ‘Sustainable City’

  • (LCA OR ‘Life Cycle Assessment’ OR ‘Life-cycle assessment’ OR ‘Life-cycle-assessment’) AND ‘City’ AND ‘Indicator*’

  • (‘Social life cycle assessment’ OR ‘Social LCA’ OR ‘S-LCA’) AND ‘Mobility’ AND ‘Mobility Service*’

  • (‘Social life cycle assessment’ OR ‘Social LCA’ OR ‘S-LCA’) AND ‘Sustainable City’

  • (‘Social life cycle assessment’ OR ‘Social LCA’ OR ‘S-LCA’) AND ‘City’ AND ‘Indicator*’

  • (‘Sustainable urban mobility’) AND (‘Social indicator*’)

  • (‘Social sustainability’) AND (‘transportation systems’)

  • (‘Sustainable urban transportation’) AND (‘social indicator*’)

Excluded from the results were all secondary studies, duplicate studies, primary studies not written in English or grey literature. Only articles from peer-reviewed journals written in English were selected, published between 2011 and 2019. The selection process can be described in four phases. In the first phase, only publications that include social indicators were filtered. In the second phase, the social indicators were categorised according to associated stakeholder groups and clustered in an analytical grid. The selection of the stakeholder groups was done in accordance with the Guidelines (UNEP/SETAC 2009) and the corresponding methodological sheets (UNEP/SETAC 2013): workers, local community, society, value chain actors and consumers. While identifying the stakeholder group, it was also analysed whether the indicators are of quantitative (q), semi-quantitative (s) or qualitative/descriptive (d) nature. In some cases, the indicator type was not stated and was therefore not identified (n.i.). In the third phase, indicators were selected based on their relevance for mobility services. For this purpose, ‘relevance’ was defined as suggested by Laprise et al. (2015): the indicator reflects the performance in relation to a given criterion. In this case, the criterion is sustainable urban mobility. This allows to select, for example, an indicator assessing impacts on space occupancy and, at the same time, allows to not select, for example, an indicator assessing the number of beds of medical institutions per capita, as this does not measure the performance of a mobility service in relation to sustainable urban mobility. In a subsequent step, measurability as well as data availability were examined. For the selection based on measurability, the following criteria were fixed: (1) The indicator is already being measured or can be measured with little extra effort, (2) the measurement will show change over a year and (3) the measurement can be compared with other mobility services (Kunstler et al. 2016). In the fourth phase, the selected indicators were supplemented by missing aspects presented in the Guidelines and the PSIA. In Fig. 2, the four phases are illustrated. After the systematic literature research in the stated databases and identification of all publications that include relevant social indicators, a total of 51 papers were selected, as shown in Table 3.

Fig. 2
figure 2

Method for the definition of set of indicators to assess social sustainability of mobility services

Table 3 Summary table of the analysed publications

3 Results and discussion

Based on the search strings and the focus on urban mobility, the selected publications focus on sustainability assessment of cities (47%), transportation systems (21%), neighbourhoods (16%) or infrastructure/building projects (8%). Only a few studies matched the search criteria with a different focus, for example social aspects in the mining sector or domestic water reuse (8% others) (see Fig. 3).

Fig. 3
figure 3

Identified focus areas from indicators analysed

In total, 579 social indicators were identified (phase 1). The allocation of the indicators to associated stakeholder groups and the identification of the indicator type (phase 2) demonstrate that out of the identified indicators, 209 (36%) assess social impacts that affect the stakeholder group local community. One hundred seventy (29%) of the indicators are societal or institutional, whereas 159 (28%) target the stakeholder group ‘consumer’. Thirty-seven (6%) of the analysed indicators assess social impacts related to the stakeholder group ‘worker’ and only 4 (1%) ‘value chain actors’, which can be seen in Fig. 4.

Fig. 4
figure 4

Social indicators categorised by stakeholder groups from the 51 reviewed publications

The fact that the stakeholder group worker is underrepresented might be unexpected. This result differs from other S-LCA case study reviews where 32% considered workers (Petti et al. 2016). However, this outcome can be explained by the focus of this literature review on the assessment of use phase impacts. Consequently, indicators assessing impacts regarding the local community, society or consumers become more prominent, whereas workers and value chain actors get less attention.

Out of the total amount of 579 social indicators, 365 (63%) are quantitative, whereas 74 (13%) are semi-quantitative. Sixty-one indicators (10%) are of qualitative or descriptive nature, and for 79 (14%), the assessment method was not stated and therefore the type of indicator could not be identified (see Fig. 5). An overview of the results of phase 2 can be seen in Table 4.

Fig. 5
figure 5

Social indicators categorised by the type of indicator

Table 4 Analysis of the indicators and allocation to stakeholder group (count and type of indicator)

In the subsequent step, for each stakeholder group, all indicators that fulfil the defined precondition regarding relevance for the assessment of mobility services were filtered and grouped into categories. For every indicator, data availability was analysed. For this purpose, different possible data sources were examined, including geographic information systems, publicly available data from mobility service providers as well as the social hotspots database (SHDB) and data collection possibilities described by the methodological sheets (UNEP/SETAC 2013). For the analysis of data availability, three classifications were used in line with Litman and Burwell (2006): (1) limited, may require special data collection; (2) often available but not standardised; and (3) usually available in standardised form. This is necessary, as some indicators require data that may be difficult to obtain or evaluate. The costs of data collection and ease of use should be taken into consideration when selecting indicators. Nevertheless, indicators should not be selected only based on data availability, as important aspects may be missed (Science for Environment Policy 2018).

The analysis of the filtered and grouped indicators (phase 3, see Fig. 2) reveals that most of the indicators of the stakeholder group local community assess impacts on public space (29%). Many different indicators can be found in this category. However, most of them measure green space or park area in square metres, often in relation to the total population of the study area. Although most of the indicators in this category are quantitative, few qualitative or semi-quantitative indicators can also be found, for example the assessment of harmony with the surroundings. Twenty-three percent of the indicators for the stakeholder group local community assess impacts on air quality. Most of the indicators in this category are quantitative, measuring air pollutant emissions. Nevertheless, mortality effects of air pollutants or actions to reduce air pollutants are additional measurements suggested in the reviewed literature. Indicators that evaluate impacts on local employment take third place with 16%. In this category, the variety of indicators is much smaller, often measuring the same aspect formulated in different ways. The indicators primarily target job opportunities, job availabilities or unemployment rate. Indicators assessing noise pollution (12%) and community engagement (8%) follow. All of the indicators stated to measure noise pollution are quantitative; however, the indicators themselves differ a lot. Some measure the area in square metres, others the percentage of population that is affected by noise pollution and yet others suggest measuring noise complaint cases. The indicators that assess impacts on community engagement are mostly qualitative, targeting, for example, the degree of population participation, the degree of information access or the existence of a response system. Seven percent of the indicators target effects on space occupancy. Here, again a great variety of indicators was found in the reviewed literature. Some indicators measure land consumption or green space destruction, and others focus on per capita area of paved roads or the proportion of land paved for transport facilities. Only 2% of the indicators focus on citizens’ satisfaction measured either qualitatively (degree of social acceptance) or quantitatively (number of complaints or number of citizens satisfied with their local area). Three percent of the indicators measure other aspects not previously stated, such as light emissions or the impact on non-motorised transportation.

Regarding the stakeholder group consumers, the majority of indicators target accessibility (41%). Here, the indicators partly assess the expansion of infrastructure, e.g. measuring the length of mass transport network or road density, and partly the transportation options itself, e.g. the number of transport points or the number of passengers per transport mode. Indicators assessing safety take the second position (25%), followed by indicators assessing convenience (11%). The indicators assessing safety can also be classified into indicators measuring the safety of the infrastructure, e.g. safe pedestrian pathways, and indicators assessing safety of the transport mode itself, e.g. mortality rate or accidents rate. The analysed indicators assessing convenience are mostly qualitative or semi-qualitative. They measure many different aspects ranging from thermal comfort and ventilation potential to supply the reliability and punctuality of deliveries. Indicators targeting inclusiveness and affordability take the last place with 7%, respectively. All of the reviewed indicators measuring inclusiveness are either semi-quantitative or qualitative. They evaluate, for example, inclusive design or the degree of universal access. Indicators assessing affordability, however, are mostly quantitative and measure e.g. household expenses on transportation or costs for public transport services. Nine percent of the indicators that assess impacts on consumers target other aspects, for example the average duration of travel to work or intermodal terminals.

The vast majority of indicators that assess impacts regarding the stakeholder group worker assess health and safety issues (53%). Different aspects are suggested to measure in the reviewed literature for this category, mainly of quantitative nature, such as the number of fatal and non-fatal accidents or compensated occupational problems. However, semi-quantitative indicators are also stated, for the assessment of work environment or appropriate working equipment. Thirteen percent of the indicators for workers assess fair salary. Here, not only the compensation itself is suggested to measure but also whether the workers are paid regularly. Nine percent of the indicators evaluate training and education, including the existence of educational programmes for self-development and the number of school absences of children. To evaluate discrimination, the suggested indicators (6%) assess equality and diversity as well as whether formal policies against discrimination exist. Only small numbers (3%) of indicators are found in the reviewed literature that evaluate child labour, freedom of association and collective bargaining as well as work-life balance. At this point, it should be highlighted that due to the selected search terms and the focus on mobility services, the stakeholder groups worker and value chain actors are underrepresented, as stated above (Fig. 4). Concerning the stakeholder group value chain actors, all of the identified indicators assess supplier relationships. It can therefore be assumed that other categories and indicators may need to be added for a holistic assessment. For the stakeholder group society, 63% of the analysed indicators assess health impacts, mainly in the form of CO2 emissions or other greenhouse gases. Twenty-five percent of the indicators focus on urban development and evaluate the existence of an urban development plan or compatibility with local urban mobility policies. Finally, 13% of the reviewed indicators of the stakeholder group society target tax income.

In Table 5, the filtered and grouped indicators are shown for the respective stakeholder groups, including information regarding indicator type and data availability (results of phase 3). For reasons of clarity and comprehensibility, indicators that are mentioned several times or evaluate the same aspect are only stated once. Overall, Table 5 not only demonstrates a huge variety and diversity of indicators intending to measure the same aspect. Many indicators are, in addition, not clearly defined, which makes them difficult to measure. The lack of concrete calculation methods in combination with a lack of data constitutes a major challenge and leads to the necessity for experts and decision-makers to select a calculation method of their own.

Table 5 Most frequently assessed categories and indicators used for the respective stakeholder groups

Therefore, on the basis of the above-stated analysis, a core set of indicators is proposed, including a description and concrete measurement for each indicator. For this core set of indicators, data availability and consultations of experts in the field of urban mobility were taken into account. For a holistic set of indicators, the results from the literature review were compared to the social topics, subcategories and indicators in the Guidelines and in the PSIA. Missing aspects, especially for the stakeholder groups workers and value chain actors, were added for a comprehensive assessment. Here again, only missing aspects that fulfil the defined precondition regarding relevance for the assessment of mobility services were added. This core set of indicators is presented in Table 6 (results of phase 4). For all core set indicators, a 5-point scale from − 2 to + 2 is defined, including underlying performance reference points. In addition to Table 6, the reference scale and performance reference points can be found in Table 7 (quantitative indicators) and Table 8 (qualitative indicators).

Table 6 Core set of indicators for the assessment of mobility services
Table 7 Reference scale and performance reference points for all quantitative indicators
Table 8 Defined reference scale and performance indicators for all qualitative indicators that are not part of the PSIA

For the assessment of impacts regarding public space, the majority of indicators focus on the assessment of green areas or public open space (see Table 5). Mobility services, just like other transportation systems, require infrastructure, which may lead to green space destruction. At the same time, higher efficiency or occupancy rate may also lead to less space occupancy and, consequently, to new possibilities for green or open space. Therefore, an indicator evaluating these aspects is essential and included in the core set. For the evaluation of air quality, relevant air pollutants, as suggested in the reviewed literature, are selected and measured per passenger kilometre. For the effects on employment, two indicators are suggested for the core set. One measures the percentage of employees hired and the other the percentage of employees hired locally during the study period. In that way, general job creation as well as local effects can be measured. For the assessment of noise pollution, two main assessment methods were identified. One targets the populated area in square metres that is exposed to noise pollution (Mansourianfar and Haghshenas 2018; Fouda and Elkhazendar 2019), and the other measures the hindrance of population by traffic noise based on a weighting factor for population density (Oses et al. 2017). Due to the difficult data situation for noise pollution, both indicators were adopted for the core set and can be applied according to data availability. For the evaluation of community engagement, though, the literature review presents many different qualitative indicators and, often, a clear assessment method is not stated in the respective publications (Mapar et al. 2017; Laprise et al. 2018; Ameen and Mourshed 2019). Therefore, it is proposed to use the qualitative, scale-based assessment method according to the PSIA, which evaluates the degree of population participation. This international handbook offers a transparent and clearly defined 5-point scale from − 2 to + 2, including underlying performance indicators, that is already internationally recognised for the assessment of community engagement (Goedkoop et al. 2018). This type of assessment method is part of type I impact assessment (see Fig. 1). A major advantage and reason for the selection of this type of assessment approach is the possibility to measure positive as well as negative impacts.

As previously mentioned, the literature review revealed various indicators for the assessment of space occupancy, often with a focus on a specific transportation mode. For the core set, however, the indicators stated in the literature are summarised and generalised. That way, three indicators are suggested for the core set, namely direct and indirect space occupancy for the different mobility modes in relation to passenger kilometre and in relation to the total study area. Direct space occupancy refers to e.g. roads, cycle lanes or railways, whereas indirect space occupancy refers to open or private parking, stations, service areas or petrol stations (WBCSD 2015). As data availability might be limited for this indicator, space occupancy (excluding infrastructure) in relation to green and open space is also included. Hereby, the different indicators can be applied according to data availability. In the reviewed literature, 2% of the indicators evaluate citizens’ satisfaction. However, the indicator evaluating community engagement already includes citizens’ opinion and the possibility to give feedback, which is why this aspect is not included additionally in the core set.

As already stated, two main aspects were identified in the reviewed literature to evaluate accessibility for the stakeholder group consumers. One is the expansion of infrastructure, e.g. the length of mass transport network or road density and the other transportation options itself, e.g. the number of transport points or the number of passengers per transport mode. For the core set, the latter is selected, as the expansion of infrastructure is difficult to capture with one single indicator that is suitable for the different mobility options (Robati et al. 2015; Li and Li 2017; Fouda and Elkhazendar 2019). Out of the indicators that were suggested in the literature to assess safety of consumers, the number of fatal and non-fatal accidents in relation to passenger kilometre was adopted as a quantitative indicator, for data availability reasons. As previously mentioned, diverse indicators were found in the literature to measure convenience, most of them of qualitative or semi-qualitative nature. However, with the aim of quantifying convenience and for data availability reasons, punctuality of deliveries was selected for the core set, measuring the number of punctual trips in relation to the total number of trips. For mobility options, which can be used on demand without waiting time, zero delayed trips can be assumed. This indicator supplements the indicator selected for accessibility, as the number of transport points within the study area also indirectly affects convenience. The higher the number of e.g. car-sharing cars or electric scooters within the study area, the lower the necessary walking time to the nearest transport point, which, in turn, leads to higher convenience. For the assessment of inclusiveness, all of the indicators in the reviewed literature were either semi-quantitative or qualitative or the assessment method was not stated (Verseckiene et al. 2017; Jasti and Ram 2018). This emphasises the difficulty to quantify inclusiveness. Therefore, the scale-based indicator as stated in the PSIA with the already mentioned clearly defined 5-point scale with corresponding performance indicators is suggested for the core set (Goedkoop et al. 2018). For the assessment of affordability, however, a quantitative indicator is selected, evaluating the trip fare for different transport modes in relation to average income. This way, local income differences can be considered. Although no indicators were found in the reviewed literature that measure data privacy or evaluate consumers’ feedback, these two indicators were stated in the Guidelines and in the PSIA (see also Table 2). Therefore, these two indicators were included in the core set. For the assessment of data privacy, the scale-based approach was adopted according to the PSIA (Goedkoop et al. 2018), whereas to measure consumer satisfaction, the number of consumer complaints was selected based on the Guidelines (UNEP/SETAC 2013).

The reviewed literature revealed that the majority of indicators assessing impacts on the stakeholder group worker evaluate health and safety issues. Therefore, the indicator that measures the number of fatal and non-fatal injuries was selected for the core set as a quantitative indicator with good data availability (Onat et al. 2014; Bui et al. 2017). For the evaluation of remuneration, though, two indicators were included in the core set of indicators. The first indicator measures a combination of wages and social benefits received by workers, a scale-based indicator as proposed by the PSIA. This indicator was included, as it not only assesses the wages for workers but also comprises social benefits. The second indicator measures the percentage of workers whose wages meet at least legal or industry minimum standards. This indicator was added to the core set of indicators, especially for the application in countries where violation of the law regarding minimum wage is a problem. This indicator might not be of relevance in every country. Hence, it can be used depending on the respective legal environment. For the assessment of discrimination, no clearly defined assessment method was stated in the reviewed literature (Aparcana and Salhofer 2013; Ameen and Mourshed 2019). As the PSIA suggests a scale-based indicator to evaluate prevention of discrimination, this assessment method was adopted. Only a few indicators were found in the reviewed literature regarding child labour, freedom of association and collective bargaining as well as work-life balance with again no clearly stated assessment method (Aparcana and Salhofer 2013). This is why, for the assessment of these aspects, the suggested scale-based indicators of the PSIA were adopted again. When comparing the indicators that were found in the reviewed literature to the indicators that are stated in the Guidelines as well as in the PSIA (Table 1), one can see that forced labour is suggested for evaluation in both the Guidelines and the PSIA, which, however, was not found in the reviewed literature. This is why the assessment of forced labour is included in the core set. For the assessment method, the scale-based indicator as stated in the PSIA was selected for consistency reasons. The literature review revealed that, in addition, 9% of the indicators evaluate training and education possibilities. Though, one of the two selected indicators for the evaluation of remuneration measures a combination of wages and social benefits. The social benefits comprise training and education possibilities. This is why no indicator evaluating training and education is included additionally in the core set.

Similar to the approach for the stakeholder group workers, the indicators found for the stakeholder group value chain actors were compared to the indicators stated in the Guidelines and in the PSIA and missing aspects were added to the core set. In the case of value chain actors, only supplier relationships were evaluated in the reviewed literature, as already mentioned. Therefore, the assessment of fair competition and intellectual property rights, as well as promoting social responsibility, was added for a holistic evaluation (Table 1). Here again, the adoption of a 5-point scale from − 2 to + 2 is proposed in order to be consistent. The proposed 5-point scale including underlying performance indicators can be found in Table 8. For the evaluation of social responsibility promotion, a quantitative indicator assessing the percentage of audited suppliers is additionally included in the core set, as suggested by the Guidelines (UNEP/SETAC 2013). Depending on data availability, one or the other indicator can be selected.

For the assessment of impacts on the stakeholder group society, the indicators that were found in the reviewed literature mainly evaluate health impacts in the form of greenhouse gases and other emissions that affect society as a whole (63%). Here, the suggestions from the reviewed literature were followed and indicators assessing global warming potential, acidification potential and eutrophication potential were included in the core set. Another important aspect revealed by the literature review is the existence of an urban development plan and the extent to which the mobility company is engaging with city authorities for the promotion of urban development. Hence, this indicator is included in the core set with the proposed 5-point scale (see Table 8). The last indicator found in the reviewed literature for the assessment of impacts regarding the stakeholder group society targets tax income. This indicator is also included in the core set, evaluating taxes paid per passenger kilometre. The stakeholder group society is included in the Guidelines; however, it is not listed in the PSIA (Table 1). Therefore, the same approach as for the stakeholder groups worker and value chain actors could not be applied. This is why in this case, no additional indicators from the two documents were included in the core set. Further, the number of social aspects in the core set of indicators should be kept to a reasonable amount for applicability reasons. This is also why all indicators that were mentioned in less than 2% of the reviewed literature (summarised under the category ‘Other’ in Table 5), were not taken into consideration for the core set.

4 Conclusion and recommendations

The UNEP/SETAC Guidelines as well as the PSIA handbook substantially contributed to the progress in S-LCA. Whereas different S-LCA indicators and impact assessment methods have been applied and tested in a number of case studies, the use phase has been underrepresented in previous S-LCA case studies and there is still uncertainty regarding use phase evaluation. Use phase impacts, however, play an important role for the assessment of mobility services. Both the Guidelines and the PSIA present different subcategories and social topics for the stakeholder group consumers or users, but still they are not enough to assess products and services where the use phase plays a prevalent role, such as mobility services or a building. This gap is analysed in this paper by starting with a critical literature review. The focus of the literature review on the use phase aims to supplement other life cycle stages of mobility, for which the UNEP/SETAC Guidelines as well as the PSIA handbook have already been applied and tested (Ekener-Petersen et al. 2014; Chang et al. 2015; Reuter 2016; Zanchi et al. 2018).

The literature review was necessary to categorise and systematically analyse indicators that were used for the assessment of urban mobility in previous studies. In total, 579 indicators were identified, analysed and grouped according to the stakeholder groups mentioned in the Guidelines. On the one hand, it was revealed that the stakeholder groups worker and value chain actors were underrepresented in the reviewed literature, and on the other hand, many indicators could be directly used from the Guidelines and the PSIA handbook. Furthermore, the indicators that were found in the reviewed literature present a huge variety and diversity, and this makes it difficult to define a consistent assessment method. However, based on the reviewed literature and supplemented by missing aspects from the Guidelines and the PSIA, a holistic set of 39 indicators for the assessment of mobility services is proposed, including an assessment method for every indicator. Out of the total set of indicators, 25 are of quantitative and 14 of qualitative nature. For the stakeholder group local community, 13 indicators are proposed, out of which only one is a qualitative indicator. Out of the eight indicators suggested for the stakeholder group consumers, two are qualitative. However, the stakeholder group workers also comprises eight indicators from which six are qualitative. For the stakeholder group value chain actors, five indicators are proposed, with four qualitative indicators and one quantitative indicator. Contrary, for the stakeholder group society, five indicators are proposed in total, out of which only one is qualitative.

It is the first time that indicators from previous studies in the field of urban mobility were systematically analysed, evaluated and allocated to stakeholder groups in order to find suitable indicators for social sustainability of mobility services. Thus, this systematic approach as well as the resulting insights constitute not only a novelty but also a major strength of this study. For the literature review, three major databases were selected and search terms were defined in order to find suitable publications with relevant social indicators. The selected search terms needed to be both general enough to find social indicators and specific enough to find social indicators that meet the particular needs of mobility services. To overcome this challenge, general search terms like ‘Social LCA’ were combined with specific search terms like ‘Mobility Services’ or ‘Transportation Systems’. The selection of search terms influences the results, which is why the selected databases as well as the defined search terms can be seen as a limitation of this study. Additional keyword combinations in more databases could improve the results.

To validate the applicability of the suggested indicators and the proposed assessment method, further research and implementation to real case studies are necessary. The obvious next step is therefore the application of the proposed set of indicators to mobility service case studies. This application and the subsequent evaluation are regarded as essential for the validation. In a further step, the results of the social impact assessment can be combined with an environmental impact assessment, for a holistic approach. Special attention should be paid to social indicators that might already be included in the environmental dimension, for example indicators assessing air quality. These indicators should only be included once, either in the social or in the environmental dimension, to avoid double counting. The combined assessment of social and environmental impacts of mobility services is considered as an important part of future research objectives. In addition, further research concerning the development of impact pathways helps to better understand the implications of mobility services and can help to facilitate the application of indicators. In that way, the proposed set of indicators may help to answer the frequently asked question whether mobility services can improve quality of life in cities.