Dockless bike sharing in Singapore: perceptions of stakeholders, tensions of management, and implications for the public space

ABSTRACT

Dockless bike sharing has experienced remarkable expansion, with significant impacts on public spaces and intense social debates. This paper presents a case study of Singapore to analyze perceptions of different stakeholders about dockless bike sharing within the public domain. A content analysis of 229 news articles is conducted, where the entire process and focused topics are summarized, and perceptions of four stakeholders are mapped and compared. Debates among stakeholders focused on its role in the transport system, severity and causes of indiscriminate parking, and the strictness of regulatory requirements. The core was managing its usage of public space, involving tensions between under and over management. Two dilemmas are highlighted, including oversupply against low bike density, and flexible parking against high expectations of orderliness and cleanliness. Nevertheless, it is argued that dockless bike sharing can be a catalyst for promoting cycling and reinventing public space, with advocating three interconnected opportunities.

KEYWORDS:

• Dockless bike sharing
• stakeholder perception
• public space management
• Singapore

Introduction

In the past decade, dockless bike sharing has experienced remarkable expansion in many countries. Unlike the previous generation of docked bike sharing system that were typically managed by municipalities as a public good, dockless bike sharing is generally operated by the private sector with tremendous capital investments. It has been widely perceived as a low-carbon mode of transport that helps bridge gaps of existing urban transport system (Duran-Rodas, Villeneuve, and Wulfhorst 2020; Yin, Qian, and Shen 2019). However, the rapid expansion of dockless bike sharing in many cities has led to problems, including oversupply, a lack of proper management, and misbehaviors such as vandalism and indiscriminate parking that heavily impact the public domain (Janmaat 2019; Laa and Emberger 2020; Ma et al. 2018).

Existing research discussing dockless bike sharing from the public domain has focused on several key aspects, including user behavior, sustainability, public reflection, and regulation. First, most studies have analyzed ridership data to examine spatial patterns of user behavior, evaluated their social-environmental factors, and identified the role of dockless bike sharing in transport networks. For example, in Singapore, dockless bike sharing contributes to first-and-last-mile travel, but has problems with low utilization and uneven distribution (Shen, Zhang, and Zhao 2018; Xu et al. 2019; Zhu et al. 2020). Second, researchers have assessed and projected the environmental effects of dockless bike sharing, as it claims to be an environment-friendly mode of transport. While improvements in bike utilization and resource conservation have been found, some works have suggested that more greenhouse gases may be released due to bike rebalancing and early dumping (Luo et al. 2019; Sun and Ertz 2021). Third, public discussions surrounding dockless bike sharing have been mapped using social media or news data, where both positive and negative opinions are prevalent (Duran-Rodas, Villeneuve, and Wulfhorst 2020; Pojani et al. 2020; Rahim Taleqani, Hough, and Nygard 2019; Yin, Qian, and Shen 2019). Supportive discussions include affordable last-mile connections, perceived sustainability, and technological innovation. There are concerns around unregulated market relying on venture capital, oversupply flooding public space, and vandalism causing early dumping. Finally, researchers have pointed out a lack of recognition and integration of new social actors, such as user groups, and have called for the establishment of regulatory frameworks and standards through multi-stakeholder (or government-business-society) collaborative regimes (Laa and Emberger 2020; Ma et al. 2018; Yin, Qian, and Shen 2019).

There are certain gaps within existing research that need to be addressed. First, establishing multi-stakeholder collaboration requires identifying the stakeholders or actors and understanding their perceptions. However, most existing studies only summarized focused topics and positive or negative opinions in general, without comparison between stakeholders relating to dockless bike sharing in various ways. For instance, Pojani et al. (2020) extracted five groups of public attitudes, including cycling advocates, technology enthusiasts, anti-technology luddites, sharing economy supporters, and anti-capitalists. Second, despite the heavy impact of dockless bike sharing on public spaces, the implications have not received enough attention. Direct sharing of bikes has to presume the indirect sharing of public spaces (Chan and Zhang 2018), so the management challenges of both are intertwined. Furthermore, tensions between stakeholders are a reoccurring theme for both public space management and dockless bike sharing management. For the former, researchers have argued that the public in public space is composed of different social groups with conflicts of interests, and there is a trend to better engage stakeholders to produce collectively agreed policy (Carmona 2010; De Magalhães and Carmona 2009). For the latter, a major friction identified by Janmaat (2019) was between private interests (operators maximize their freedom to use and commercialize public space) and public values (municipalities control public space quality). In our view, the tensions between stakeholders may be more complex than pro- or anti- dockless bike sharing opinions and binary public-private conflicts. Such gaps underscore the need to understand the perceptions of various stakeholders within the public domain.

This paper aims to understand perceptions of different stakeholders on dockless bike sharing within the public domain, identify tensions in its management, and discuss implications on public space. The paper proceeds with explanation on research design, including the case study (Singapore), data collection (media news article), and data analysis (content analysis). The subsequent section presents the results on focused topics of public discussion and intertwined perceptions of specific stakeholders. The results are discussed through the lens of a framework of public space management.

Methodology

The case study: Singapore

Singapore is selected as the case study because of the full-cycle wave of dockless bike sharing it experienced and the uniqueness of its public space management. Singapore’s dockless bike sharing market experienced a full-cycle boom and retreat from January 2017 to April 2019 (Toh 2019). The market included three large operators, Ofo, Mobike, and oBike, as well as some smaller ones. It provided a clearer and more complete picture of dockless bike sharing and social discussions than other cities. By April 2019, the three major operators withdrew, and bike supply and ridership collapsed around the same time (Song et al. 2021). In addition, Singapore’s low cycling rate (less than 2% in 2019) is also typical in many big cities worldwide and has potential to be promoted (Choo 2019). The climate (e.g. heat, rain) may be a factor deterring cycling, but it unnecessarily causes such a low rate (Bean, Pojani, and Corcoran 2021; Lee and Pojani 2019).

Given the scarcity and high value of urban land in Singapore, the management of public space, as a precious resource, is unique with cleanliness and orderliness (Hee and Ooi 2003). Thus, the impact of dockless shared bikes, exceeding 100,000 at the peak, was particularly apparent, which caused significant public attention and intense social discussions (CNA 2018b). Being “far too prone to vandalism” and dumping has been regarded as prevalent problems and major reasons for the failure of dockless bike sharing in many big cities (Pojani et al. 2020). Singapore, featured by common abidance of strict laws in public space, is hence an ideal case to observe the dockless bike sharing experiment and its implications for public space, as less dominance of vandalism exposes other problems or tensions.

Data: media news

Media news provide opportunities to understand the status-quo and stakeholders’ perceptions of dockless bike sharing. Previous researchers have relied on methods such as survey, interview, focus groups, and social media to capture stakeholders’ opinions (Fishman, Washington, and Haworth 2012). Media news is also an effective means of understanding these opinions, as news platforms perform several crucial functions (Cheng and Edwards 2019; Gössling 2020; Leung, Xue, and Wen 2019). First, disseminating information on social affairs and public policies. Second, identifying important issues that merit attention and setting the agenda for the public. Third, highlighting public concerns and revealing how people talk about social issues. A few researchers have applied media news data to study various sectors of sharing economy, including attitudes of interest groups in mixed sectors (Leung, Xue, and Wen 2019), public attitudes of dockless bike sharing (Pojani et al. 2020), overall public concerns on e-scooter sharing (Gössling 2020), and focused topics in tourism and sharing hospitality (Cheng and Edwards 2019). Distinct from these related works, this research emphasizes on public domain instead of business or market, reveals various perceptions of stakeholders, and dedicates to one city, exploring tensions in-depth under a specific urban context.

For this research, news articles from Singapore’s two major English news media companies, Channel News Asia (CNA) and Straits Times, were collected online. These two companies have long been considered Singapore’s top English news media companies. The collection of news articles was conducted in Google Advanced Search in February 2020, using keywords “bike,” “Singapore,” and “sharing OR shared,” with site restrictions on channelnewsasia.com and straitstimes.com, and date restrictions from January 2017 to May 2019. The search yielded 300 articles from Straits Times and 180 articles from CNA. Articles unrelated to dockless bike sharing were excluded through skimming of titles and texts (e.g. scooter sharing, general cycling collisions, electric bikes, and cases in other cities). As a result, 158 news articles from Straits Times and 71 articles from CNA were used for content analysis.

Data analysis: content analysis

The manually coding content analysis was conducted using MaxQDA. Each news article was coded for basic information, such as the date, source, and title, and for a framework consist of two major parts. The first part was developed using an inductive approach, which yielded 22 focus topics that identified “what is going on” with dockless bike sharing. The word counts of these focus topics showed “what are important” in news. The coded segments in this part covered the entire body of text in all news articles without overlap, except for those irrelevant to bike sharing. The second part of the framework identified opinions from four categories of stakeholders – government, operators, users (cyclists) and residents. By comparing keyword frequencies and extracting their contexts, we captured opinions within each stakeholder group and mapped major focuses and tensions between them.

In Singapore, stakeholders directly related to the management of dockless bike sharing were clear. Singapore has a centralized planning system and strong spatial control practices (Hee and Ooi 2003). Operators and users (cyclists) of dockless bike sharing were obviously two main stakeholders. The third stakeholder was the Land Transport Authority (LTA), the primary regulator responsible for transport planning and management, including dockless bike sharing. Another stakeholder was the residents or the community, referring to individuals who barely used dockless bike sharing but were directly or indirectly affected by its impact on public space. Other interest groups identified by related works did not align with the aims of this research. For instance, partners and competitors (Leung, Xue, and Wen 2019) were not included, as we emphasized on public domain.

Results

In this section, we present the results in two parts. First, we summarize the entire process of dockless bike sharing wave in Singapore by analyzing the news content distribution over time and among coded focus topics (Figures 1 and 2). Second, we interpret the perceptions of the four stakeholders by comparing keyword frequencies and examining keyword contexts in respective coded segments of stakeholders (Figure 3).

Figure 1. News word counts by month with milestones.

Figure 2. News word counts of coded focus topics.

Figure 3. Term frequencies divided by total word counts of respective stakeholder for 20 most frequent terms of each stakeholder.

Entire process and focus topics

The early news content focused on observing the phenomenon of dockless bike sharing and associated problems like indiscriminate parking. Although the city had a public docked bike sharing plan, when private operators entered the market, the government put that plan aside and hoped to see how private-managed bike sharing worked (Lin 2019). When dockless bike sharing entered the city, the news focused on observing this new phenomenon (rank 5th among all focus topics). From January to March 2017, the three major operators, Ofo, oBike and Mobike, launched in Singapore, introducing a few thousand bikes. The first peak of news content appeared in mid-2017, when problems like vandalism and indiscriminate parking arose (Figure 1). The number of shared bike grew rapidly to over 100,000 (Toh 2019). The most covered problems were vandalism (rank 4th) and indiscriminate parking (rank 8th). Although vandalism received more news coverage than indiscriminate parking, it had a “very low rate” (“less than one percent” of bikes) in Singapore compared to other cities (Lam 2017; Sivaram 2017). Vandalizing behaviors were rare and eye-catching, so that it attracted much attention, especially in Singapore’s well-ordered public space. Residents identified “errant parking” as “a major issue” (Varakantham 2018). Correspondingly, the subsequent regulations and measures mostly focused on indiscriminate parking.

Operators and the government took various measures to address the above issues related to dockless bike sharing. Operators implemented technologies such as geo-fencing (rank 9th) and credit systems (rank 10th) to manage user behavior. In May 2017, LTA introduced a policy that required operators to remove indiscriminately parked bikes within a half day upon receiving a report, failing which the bikes would be impounded with a fine of S$140, namely, “half-day removal or fine and impound” policy (rank 17th). Additionally, the government developed an app function for residents to report indiscriminate parking (CNA 2017c 2017b). In October 2017, operators, government departments (including LTA), and town councils signed a memorandum of understanding (MOU) for “responsible use” of dockless bike sharing “in public spaces” (CNA 2017c 2017b), primarily against indiscriminate parking (rank 15th). As part of the MOU, LTA committed to provide more parking spaces and cycling paths (rank 6th). However, operators failed to comply with the half-day removal policy, leading to LTA issuing over 2100 removal notices and impounding 341 bikes by February 2018 (J. S. Ng 2018).

In March 2018, LTA announced new regulations, including a license scheme and a QR-based designated parking rule. The new regulations, still focusing on “regulating the indiscriminate parking,” were the most influential regulations and marked a new stage of dockless bike sharing (CNA 2019d). The license scheme (rank 2nd) controlled operators’ fleet numbers and levied a license fee based on the fleet size. After implementing the license scheme, only 40,500 bikes were legal (in contrast to 100,000 before). The QR parking system (rank 3rd) regulated users by requiring them to park bikes in designated parking zones (boxes) and scan the QR code to finish parking. This made the system more or less equal to a docked system, as users were no longer allowed to park anywhere. Failing to park correctly incurred S$5 fine, and after three times led to up to one-year ban, which included all operators’ shared bikes. License applications started in July 2018, while QR parking system was launched in January 2019. Discussions about the new regulations formed two highest peaks of news content quantities (Figure 1).

The three largest operators retreated from the city, citing difficulties in complying with the new regulations, but the problems they left behind resulted in widespread criticism of their business model (rank 1st). In June 2018, oBike announced that it would quit due to “difficulties foreseen … to fulfil the new requirements” (CNA 2018e 2018l). In early 2019, Ofo “failed to reduce its … fleet number” to meet the stipulated maximum of license and implement QR parking system (CNA 2019e). LTA suspended Ofo's license in February and canceled it in April. Later, Mobike applied to surrender its license, as a part of the “plan to rationalise … operations in Southeast Asia” (Toh 2019). Retreat of the three biggest operators marked the end of the first wave of dockless bike sharing in Singapore (Co 2019). Many discussions on this topic focused on operators failing to refund users’ deposits and clear bikes from public spaces. For example, oBike owed town councils S$118,000 in fines and owed users S$8.9 million in deposits (CNA 2018k). Besides, many people criticized operators’ business model, which relied on venture capital investment and oversupplied bikes at lower-than-cost prices (rank 7th).

Comparison of perceptions from four stakeholders

To compare perceptions of the four stakeholders, we summarized most frequent terms and their typical contexts in respective coded segments. The total word counts of stakeholders’ perception segments are as follows: government (4411), operator (7242), user (3568), resident (12795). We selected the 20 most frequent terms (with synonyms combined) for each stakeholder, resulting in a total of 41 words. We divided each word’s frequency by the total word count of the stakeholder to enable cross-stakeholder comparison of term frequencies, which we named Relative Frequency (see Figure 3). We examined the contexts of highest and second highest Relative Frequency terms (compared between stakeholders) considering the overall discourse of each stakeholder.

Within the public domain, the perceptions of the four stakeholders regarding dockless bike sharing showed tensions mainly in three aspects. First, for this phenomenon itself, controversy laid over the contribution to the transport system, the effectiveness of business model, and the support from public sector. The second aspect involved debates over the seriousness and causes of indiscriminate parking, which was the major issue. The third aspect was different opinions on strength of regulatory requirements, particularly the license scheme and the QR designated parking rule.

Whether dockless bike sharing was beneficial and who should operate it

Operators, users, the government, and some residents recognized dockless bike sharing as a supplement to Singapore’s transport network. Operators marketed dockless bike sharing as an “environmentally friendly and efficient” mode that satisfied “users’ commuting” demands. They “strongly believe(d) in” its benefits (CNA 2018i 2018a; H. Ng 2018b). Users praised the dockless bike sharing “service” for being “convenient” and “useful,” as it bridged the gap between office or home and “MRT station” or “bus stop” (CNA 2018b; CNA; 2017a). In short distance, “cycling” was found to be “only slightly slower” than bus (Toh 2019), and users could “park it anywhere” that was “as near to … destinations as possible” (CNA 2018b). Some users were attracted by the low price. For example, a typical “cost” of “30-day pass” was initially only S$6 (CNA 2018i 2018a), as opposed to “between S$50 and S$100 per month” if “taking the bus for short distances” (CNA 2018c 2018j). Even after the price of “180-day pass” increased to S$50, some users still felt “reasonable,” while others did not because “the whole selling point” was “low prices” for them (CNA 2018c, 2018j). The government and some residents acknowledged that dockless bike sharing provided “a convenient and healthy option for first-and-last-mile journeys” (CNA 2018m 2018i). These opinions roughly corresponded to what Pojani et al. (2020) called “cycling advocates” and “technology enthusiasts,” who saw dockless bike sharing as a flexible, affordable and sustainable mobility empowered by technology.

However, a major controversy was whether dockless bike sharing is “part of our public transportation ecosystem” (Lau 2018), in other words, to what extent the public sector should be involved in supporting dockless bike sharing. First, a major critique from many residents is that dockless bike sharing, operated by private companies, took “advantage of … public spaces” to make money (Delios 2017). Meanwhile, operators priced the service “below costs” and oversupplied bikes for grabbing a large “market share,” which disrupted the market and was not profitable (CNA 2018g 2018d). They opposed public investment in bike infrastructures, as they believed it would only “benefit … private companies” (Delios 2017). Second, on the other hand, operators claimed that dockless bike sharing “supplement(ed) the public transport network” (CNA 2018a 2018f). The government somewhat tended to encourage cycling through dockless bike sharing. If such “business” is not profitable, intervention of public sectors is required, such as providing “incentives” proposed by some residents (Varakantham 2018). Third, residents held different opinions on who should operate dockless bike sharing. Some residents advocated having “two to three big companies” (Varakantham 2018). Some suggested that bike sharing should be fully operated by public sector (Lin 2019). Some others even did not regard cycling as a necessary last-mile commute option because “public transport and covered walkways” already covered public housing towns, which constituted 80% population (Lim 2017a). Many of these debates were also seen in “luddites,” “sharing economy supporters,” and “anti-capitalists” in Pojani et al. (2020).

Whether indiscriminate parking was severe

Both the government and residents perceived “rampant indiscriminate parking” as a major issue, but there were debates on its severity. Indiscriminate parking created problems in two ways. First, indiscriminately parked bikes conflicted with other street users such as pedestrians and “users of wheelchairs” (Lim and Hong 2017). Residents complained that bikes were “along the roads, corridors, on staircases and even blocking fire escapes,” becoming “obstruction” and posing a “danger” (CNA 2017a). During this time, Singapore started allowing cycling on sidewalks, many of which were less than 2 meters wide, causing residents to worry about the safety of “shared pathways.” Second, indiscriminately parked bikes “degrade(d)” public space (CNA 2018m 2018i). Some residents saw them as “eyesore,” making public space “untidy” (P. S. K. Tan 2018). In contrast, some residents argued that automobiles were “not considered ‘clutter’ even though they (took) up much more space than bicycles” (G. Tan 2018). Residents also pointed out that indiscriminate parking was normal “in any global … city … with dockless … system,” implying Singapore residents might be overly sensitive to it (Lim 2018a).

Stakeholders had different opinions about the cause of indiscriminate parking. First, for users, there were insufficient “parking spaces” for bikes (e.g. at transit stations). Some users parked bikes randomly near an MRT station, aware that their parking spots were “not proper,” but they did so “for convenience” and to “save time” (CNA 2018d 2018c). Some residents also said the bike parking spaces were yet to be “visible” and “well positioned” (Lim 2018a). Second, the government assessed the cause to be “aggressive … fleet” expansion by operators, leading to oversupply beyond their management capacity (CNA 2018m 2018i). Indiscriminate parking persisted regardless operators “commit(ted)” for “responsible” parking through user education, technology applications, and credit systems (Lam 2017). Third, some residents believed that a lack of “a proper government regulatory framework” was the cause, as “it would be difficult to require the operators to enforce strict rules on … customers” (CNA 2017b 2017a). Under the free market, if one operator strictly penalizes improper user behavior, it would just lose customers, as the user can simply shift to other operators (CNA 2018i 2018a).

Whether regulatory requirements were too strict

Regulatory policies were changed from loose to strict, as the government observed that operators were “confident” and “commit” a lot but often failed to fulfill. Initially, the government only kept “monitor” (Abdullah 2017b). With persistence of indiscriminate parking and residents’ complaints, the government started to “take strict enforcement action” and “tighten regulations,” in order to “ensure” operators’ “responsible manner” (CNA 2018m 2018i; Abdullah 2017c, 2018a). Actions ranged from half-day removal, app for reporting, to the license scheme and QR designated parking rule. Despite the strict regulation, operators left behind a mess after their closure, such as failing to “ensure that the bicycles do not clutter public spaces” (CNA 2019c). LTA pointed out that oBike “reneged on its earlier commitment” on “responsibility to … refund customer’s deposits” and clear bikes “from public areas” (CNA 2018h 2018e). During operators’ closure, users complained about fail “deposit refund” and no response of “customer service” (CNA 2018f 2018g; CNA 2019a).

The policy changes provoked intense debate among other stakeholders. Under half-day removal policy, it was difficult for operators “to locate every single errant bike” within a half day (Abdullah 2017a). Regarding the license scheme and QR designated parking rule, operators complained that they were neither “business friendly,” nor in line with “a car-lite society” that “government … committed” (CNA 2018h 2018e; CNA 2018i 2018a). A user said that the new regulations was “harsh” and “could put users off”(Abdullah and Ng 2018). Users believed “bike-sharing is the future,” and the city should not lose “such a convenience,” but they understood “the need for such regulations” (Wong and Ng 2018). However, some residents said that “tough legislation (was) needed” (P. S. K. Tan 2018). They requested to “fine (and) ban … errant users” (Lim 2018b) or return to a “docked system” for reducing “illegal parking” (Lim 2018a).

In particular, the license scheme partially addressed some issues, such as oversupply and free-riding public space, but it resulted in higher costs for operators and less accessibility of bikes for users. Operators claimed that “no other market would levy such a fee,” and the “license fee” exceeded “engineering and manpower costs,” making dockless bike sharing “more expensive” (CNA 2018i 2018a). Furthermore, the “awarded fleet size” was less than half the number of bikes that operators already had at that time, leading to concerns about “harder … access” to bikes for “users” (CNA 2018l 2018h). Indeed, users did “struggle … to find bicycles,” which used to be “everywhere” (CNA 2019a). Users also expected price increases due to the “higher operating costs after … license scheme” (Abdullah 2018c). The government explained that fleets would be allowed to grow if an operator was “able to manage indiscriminate parking and ensure good utilisation” (Abdullah 2018b). “The license fee” was for covering “regulatory costs,” so that such costs were “not passed to taxpayers” (CNA 2019d).

Introducing QR designated parking and associated penalties aimed to standardize the rules for dockless bike sharing users in public space, but it reduced users’ parking flexibility and could affect operators’ profits. The government’s intention was to make “efficient use of limited public spaces” (Abdullah 2018d). However, users had to walk farther before and after cycling, especially when they felt that designated parking spaces were insufficient and difficult to find (CNA 2019b). Operators also mentioned that “users may … find inconvenience,” and hoped LTA could provide more “parking spaces” (CNA 2018j 2018m). Users believed the S$5 fine and possible one-year ban would “deter … improper parking” (Wong and Ng 2018). Users were afraid of being fined and banned, as many of them purchased seasonal or monthly pass (Wong and Ng 2018). “To avoid the risk,” some users said they “may switch to buying … own bicycle” (Abdullah and Ng 2018). If users were pushed out, operators’ profit would be affected – that was why they were concerned about users’ convenience.

In sum, debates among stakeholders focused on whether dockless bike sharing overused the public space, to which extent it should be allowed to use public space, and whether its usage of public space was over-managed. The case of dockless bike sharing in Singapore was unique in some aspects, including widespread complaints about indiscriminate parking, strict regulations compared to most big cities, and concentration of disagreement over usage of public space and public resource. Why did controversies concentrate on these particular topics, what specific urban context contributed to these debates, and what are the implications to the city’s public space? In the following part, we will discuss them in-depth.

Discussion

The management of dockless bike sharing and public space are intertwined and both relate to the tension between under and over management. For public space, the spectrum of under and over management is a critical framework to analyze management status (Atkinson 2003; Carmona 2010). Under management leads to distressed, neglected and unsafe public spaces, which can alienate residents and contribute to a decline in public culture. Over management brings surveillance, restriction, and commodification, resulting in the exclusion of “undesirable” activities and people (often from deprived groups). This “domestication” process can change public spaces to cater to middle-class and suburban culture. Cultural change, in turn, legitimizes the under- or over-management policies.

The management of dockless bike sharing in Singapore underwent both under-management and over-management stages. During the under-management stage, operators took advantage of public space by oversupplying dockless bikes and pricing below cost. Operators were unable to manage users’ improper behavior, so that indiscriminate parking became a frequent complaint by residents. Without coordination between operators and government intervention, operators were not able to establish standardized rules and penalties. Operators’ poor management was also demonstrated by their irresponsible closure, failure to refund deposits, and inability to clear bikes from public spaces. During the over-management stage, the “half-day removal” policy was proved difficult for operators to comply with. It was followed by the new regulations, including the license scheme and QR designated parking rule. The former addressed problems of free-riding and oversupply, while the latter clearly confined the space that bikes can be parked, and standardized rules towards users. However, the new regulations became a significant contributor to the decline in ridership and the collapse of major operators. Through the lens of under- and over- management, we identify two dilemmas.

Dilemmas between under and over management

The first dilemma was between oversupply of bikes causing mess and low utilization, against low density of bikes resulting in even less usage. Singapore’s low cycling rate (less than 2%) was a major reason, which could be attributed to insufficient cycling infrastructure (Choo 2019). Researchers suggested that better cycling infrastructure is likely to encourage people to ride longer distances (Shen, Zhang, and Zhao 2018). On the under-management stage, operators set lower-than-cost prices and oversupplied bikes to let users access bikes conveniently and increase demand for dockless bike sharing. Researchers in Singapore found that larger fleets were associated with higher usage (Shen, Zhang, and Zhao 2018). Additionally, relying on low prices and venture capital investment to gain market share is a common problem for many “sharing economy” platforms (Lim 2017b; Pojani et al. 2020; Toh 2019). The consequences were twofold: first, the average utilization of shared bikes was lower than what is needed to be profitable (Shen, Zhang, and Zhao 2018); and second, the fleet number exceeded operators’ management capabilities. It was inevitable that shared bikes caused some mess in public spaces.

In the over-management stage, the license scheme addressed oversupply, but reduced bike accessibility to users, resulting in a decline in usage. The license scheme, controlling the fleet numbers, was essentially a restriction on the amount of public spaces that dockless bike sharing could occupy. The license fee meant that operators could no longer free-ride the public space, thereby increasing operators’ costs. Although the supply met the demand in number, this number of bikes was insufficient to provide a convenient access for users. Prices also increased due to operators’ higher costs. As a result, the usage of shared bikes declined, and operators became even less profitable than the under-management stage, contributing to their collapse.

The second dilemma was between society’s emphasis on orderliness and cleanliness, against users’ requirement for parking flexibility. In the under-management stage, residents perceived indiscriminate parking to be a big issue, but in fact it was not as serious as in other big cities worldwide (Lim 2018a; H. Ng 2018a). Given the low cycling rate, the amount of public space occupied by shared bikes was not substantial. This suggests that residents had a lower tolerance for disorder and unplanned spillover in public space, such as disorganized bike parking, than other big cities. The reason was what researchers have described – under strong capacity of planning system, public space in Singapore were homogenized, clear in hierarchy, strong in boundary, and sanitized or sterilized (Hee and Ooi 2003; Pomeroy 2011). Cycling was often considered not desirable in residential public spaces, and most residents did not cycle (Pomeroy 2011). Moreover, dockless bike sharing was operated by private companies and did cause conflicts with other street users, which left residents with a negative impression as an “invader” of public space. Most streets in Singapore did not have designated space (e.g. bike lane) for cycling, forcing cyclists to either use sidewalks (often less than two-meter width) or vehicle moving lanes. This situation increased the possibility of conflicts, and “illegitimated” cyclists among street users.

In the over-management stage, dockless bike sharing lost its primary feature, parking flexibility, leading to an even lower cycling rate, while many residents supported the over-managed measures. Under QR designated parking rule, restricted and insufficient parking zones made bike sharing resemble a docked system. Before and after each ride, the user needed to walk a long distance to collect a bike, and look for a parking zone. Reduced parking flexibility made dockless bike sharing less convenient, since cycling is often the last-mile travel. Such inconvenience was exacerbated by lower bike density under the license scheme. Besides, many users purchased monthly or seasonal passes, leading to their fear of the possible-one-year ban penalty. The result was a decline in usage of dockless bike sharing and even lower cycling rate. On the other hand, some residents “fully support(ed) LTA’s action and call(ed) for tougher penalties,” in order to “bring back order” (P. S. K. Tan 2018). Residents legitimated over management because these measures were aligned with Singapore’s long-standing strict management of public space.

In sum, it was difficult to balance oversupply and low bike density. The former caused mess, poor management, and low utilization, while the latter resulted in inconvenience to access bikes and even deterred cycling. Concurrently, it seemed convenient and flexible parking was essential for cyclists, but inevitably resulted in some mess, which was a situation that residents are hard to tolerate.

Future opportunities

So, in Singapore, is it possible to develop dockless bike sharing and make it co-exist peacefully with other public space users? We argue that this is achievable. Retaining street life and promoting cycling are both Singapore’s goals and, more importantly, are mutually reinforcing. In Singapore, community’s public space relevant to everyday life has been removed from wider urban public spaces such as street networks (Hee and Ooi 2003). “Singapore is globally renowned for its orderly urban planning … But too much of a preoccupation with order and discipline can stymie creative thinking and shackle innovation” (Agarwal 2018). The city is reimagining public spaces as vibrant places that nurture public life (Urban Redevelopment Authority 2024). It is important to retain a vibrant street life, while cycling contributes to this goal. Encouraging cycling contributes to a larger mode share of active transport that is both sustainable and healthy. A significant portion of transit travels is short-distance, which has the potential to be substitute by cycling (Agarwal 2018; Kumar, Nguyen, and Teo 2016). Singapore is constructing cycling paths, aiming to raise cycling rate from less than 2% to 4–6% (Choo 2019; Land Transport Authority Singapore 2019). In this process, dockless bike sharing can “serve as a catalyst” (Cheong, 2017; Menkhoff, 2017). Dockless bike sharing encourages more residents to cycle, and creates more on-street interactions between cyclists.

In particular, we highlight three interconnected opportunities, multi-stakeholder management, smart technologies, and spatial redistribution and redesign. First, a public–private partnership may be more effective in providing and operating dockless bike sharing. Given the low cycling rate, the private sector alone is unable to make dockless bike sharing profitable, and thus cannot maintain public space well. In short term, effective management requires financial support and coordination from the public sector. Engagement of public sector also helps monitor the usage of users’ data, as commodification or insecurity of users’ data is a concern (Pojani et al. 2020). To seek a balance between under and over management, different stakeholders should engage in setting prices, rules, and players’ duties, as well as conducting public education and activities. For example, how flexible can users park their bikes (or how dockless is it)? There could be either enough designated parking space or planning no-parking zones. From the scope of public space, local communities can be better empowered by exploring more effective ways to involve various stakeholders in a variety of roles of public space management, including policy design, implementation, and service provision (Atkinson 2003, De Magalhães and Carmona, 2009; Hee and Ooi 2003). A multi-stakeholder platform for management of dockless bike sharing, along with public education and activities, will be a part of this reinvention of public space. It will also inform similar approaches to managing other urban affairs.

Second, smart technologies provide new empowerment to better balance under and over management of dockless bike sharing. In addition to the issue of indiscriminate parking, another challenge is the clustering of bikes at transit stations during the morning rush hours, leaving a shortage of bikes in residential areas. Repositioning bikes is costly in manpower and energy (Luo et al. 2019; Pojani et al. 2020; Zhu et al. 2020). By implementing rebalancing algorithms and providing incentives, the system can encourage cyclists to find and use underused bikes or those parked indiscriminately. Participation of cyclists and residents in the rebalancing process is also a form of public engagement in maintenance of public space. Additionally, technologies also enable the analysis of travel patterns for allocating parking spaces (or no-parking zones) and infrastructures.

Third, the current conflicts between cyclists and other public space users are mainly due to the lack of cycling infrastructures, particularly on streets. Constructing cycling infrastructure not only allocates spaces for cycling, which reduces conflicts, but also reminds other users that cycling is legitimate and even encouraged as a mode of transport. Redistributing and redesigning street space requires the engagement of different stakeholders. This is further facilitated by the assessment of travel and rebalancing behaviors based on analyzing data generated by the system. Through these three interrelated approaches, residents’ attitude and tolerance on cycling-related behaviors and public space discipline could change gradually. A balanced management of dockless bike sharing and the reinvention of public space are anticipated.

Conclusion

This paper analyzes different stakeholders’ perceptions of dockless bike sharing within the public domain, tensions in its management, and implications for public spaces based on a content analysis of news articles in Singapore. The findings suggest that debates among stakeholders focused on three main aspects. The first focused aspect was surrounded by whether dockless bike sharing is beneficial to Singapore’s transport system, who should operate it, and whether private operators’ business models is profitable. The second focused topic was around the major issue, indiscriminate parking, debating whether it was a serious problem and what the causes were. The third major disagreement was whether regulatory requirements were too strict and led to collapse of the major operators. These debates essentially revolve around whether dockless bike sharing overuses public space, what extent should its usage be allowed, and whether it is over-managed.

The paper further discusses tensions in the management of dockless bike sharing and public space, based on under-versus-over management framework. It appeared difficult to achieve a balance between under and over management, especially facing two dilemmas. First, because of the low cycling rate, the supply of bikes resulted in either oversupply or a low bike density. The former led to low utilization, poor management, and mess in public spaces, while the latter led to inconvenience to access bikes and even less usage. Second, the flexibility of parking was essential for dockless bike sharing, but this feature caused some unavoidable mess, which was difficult to be tolerated by residents, who had high expectations for cleanliness and orderliness. Nevertheless, we argue that promoting cycling and retaining vibrant street life are both Singapore’s goals and are mutually reinforcing, where dockless bike sharing can serve as a catalyst. We pointed out several interconnected opportunities, such as provision of dockless bike sharing through public–private partnerships, multi-stakeholder management of dockless bike sharing and public space, technology application for rebalancing bikes, as well as public space redistribution and redesign.

More research is needed to gain a better understanding of stakeholders’ opinions and policy contexts in different cities. This paper primarily relies on media news as the data source, but there are debates about whether news represent objectively “true” facts and discourses, as news items could be driven by interests, emotions, or personal preferences of the journalist in some cases (Gössling 2020). To understand more specific insights from different stakeholders that are operational in practice, surveys, interviews, and focus groups could be conducted (Fishman, Washington, and Haworth 2012). In addition, although Singapore may be unique in public space discipline and state capacity, it is not the only city facing over-management issues in both public space and dockless bike sharing. Many cities encounter challenges in managing disruptive sharing mobility platforms (e.g. bike, scooter). Policy assessment and comparison can provide cross-city references for better regulatory development.

Disclosure statement

No potential conflict of interest was reported by the author(s).