Advertisement

  • Loading metrics

Open Access

Essay

Towards dynamic evaluations: Assessing the politics of Stavanger’s free public transport experiment

Abstract

In this essay we argue for what we term ‘dynamic evaluation’ of transport policies. In a time of increasing political polarization of public transport, dynamic evaluations involve a broader scope of factors and indicators, including political processes and justice, compared to traditional evaluations typically centered on ridership. We illustrate this perspective by discussing the experiment of free public transport provision in Stavanger, Norway during July-December 2023. Planned and budgeted for a year shortly before local elections, the intervention was repealed halfway through, as the new governing political coalition held that this was a sub-optimal means to deliver public transport policies. We contrast a traditional commissioned evaluation of how the experiment modestly increased ridership with a dynamic evaluation involving broader political and justice processes, to illustrate that the experiment contributed to politicizing public transport in potentially transformative ways.

Citation: Sareen S, Meelen T, Haarstad H, Biswas S (2025) Towards dynamic evaluations: Assessing the politics of Stavanger’s free public transport experiment. PLOS Clim 4(4): e0000604. https://doi.org/10.1371/journal.pclm.0000604

Editor: William Usher, KTH Royal Institute of Technology: Kungliga Tekniska Hogskolan, SWEDEN

Published: April 30, 2025

Copyright: © 2025 Sareen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by JPI Urban Europe (ENERGY4ALL to SS) and the Research Council of Norway (349994 to SS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

In recent years public transport has come to be understood as something more than simply bringing people from one point to another (see, e.g., [1]). Advocates for thinking about public transport in justice terms have gained popularity, for example Bogota’s former mayor Enrique Peñalosa, whose TED Talk “Why buses represent democracy in action” garnered over a million views. Yet there is no consensus on how to evaluate and assess justice in public transport or mobility more broadly [2–3]. In other words, there is a research gap in evaluating the effectiveness of actions taken to improve environmental and social justice in transportation.

In this essay, we contribute to filling this research gap by elaborating on what we term ‘dynamic evaluation’ of transport policies. In order to understand the broader political, social and cultural effects of public transport, there is a need to go beyond traditional evaluations centered on quantitative measures of ridership, rider satisfaction, costs and modal preferences. This approach is bolstered by the recognition that policy measures are not necessarily informed by pre-existing evaluation [4], and that in a time of increasing political polarization of public transport, dynamic evaluations involve a broader scope of factors and indicators, including political processes and justice. This is particularly relevant given strong links between dominant and highly space- and energy-consumptive automobility and meta-narratives of individual freedom that are deeply entrenched in society [5].

We illustrate this argument by discussing an experiment of providing free public transport in Stavanger, Norway, with the purpose of contrasting a traditional commissioned evaluation with a broader evaluation that includes political and justice aspects of the same experiment in this traditional oil capital city with entrenched automobility reliance [6]. In May 2023, the leaders of six political parties declared that public transport would be free for the next twelve months. Immediately, the main opposition party labelled the subsidy of 200 million NOK (Norwegian kroner, about USD 20 million) an election campaign ploy ahead of the September 2023 local elections. The governing coalition in May 2023 did manage to operationalize their policy into practice by July 2023, on a very short timescale. After power shifted to the right-wing coalition post-election, the new municipal government announced that the free public transport scheme for residents would expire after December 2023, having run for half a year. During this period, it cost about 25 million kroner a month on average, thus using up approximately three-quarters of the originally allocated budget, meaning it could have run for two more months. Newspaper op-eds brimmed with debate and opinions from across the political spectrum.

There are multiple ways to evaluate this withdrawn policy intervention. The public transport provider Kolumbus commissioned a study of the effects of the intervention through a tender won by the Oslo-based Institute of Transport Economics. In their report [7], the evaluators found that the use of public transportation increased as a result of the introduction of free transportation. The evaluation also found some decrease in the use of cars, but these effects were uncertain. The report was based on counts of on- and off-boarding, as well as wider user surveys, with data compared to another Norwegian region, Drammen. The report notes the significant methodological weaknesses and the uncertainties of the results, and points to findings from evaluations of similar experiments elsewhere that the effects on reducing car use (often a key aim with such policies) are ambiguous [8].

Our point is not that these traditional evaluations are not useful – to the contrary, they serve an important purpose. Rather, we argue that interventions in transport also need to be evaluated in broader political and justice terms, which cannot be reduced to quantitative analyses of a limited set of factors. We argue that there is over-reliance on quantitative measures – part of a broader societal predicament famously critiqued as ‘trust in numbers’ [9] – that exclude situated analyses of such interventions in transport research, even though qualitative studies could foreground aspects related to justice implications that are hard and time-intensive to capture. For instance, the formal evaluation of Stavanger’s experiment was not equipped to articulate the role of polarized politics in shaping public opinion and perceptions. Our analysis demonstrates that the electoral politics of the time period coinciding with the free transport pilot were a decisive factor in driving the perception of success and failure, prior to any quantitative measure being available (except perhaps the budget allocation). In our usage, the concept of “dynamic evaluation” is a way to capture a broader evaluative frame for transport policy. To Feinstein [10], dynamic evaluation is a type of evaluation that tries to contribute to transformative change. Soui and co-authors [11] contrast dynamic evaluation, which they see as evaluating the learning capacity of a system, to static evaluation, which is guided by a set of rules defined by the evaluator.

Drawing on these, as well as literature on sustainability transitions discussed below, we understand dynamic evaluations for the purpose of this essay as an assessment of the systemic learning for the purpose of shifts towards sustainable, just transitions in transport. These shifts are urgent, given growing urban populations and the climate crisis [12]. The aim of a dynamic evaluation should be to identify if the policy brought about any key transformative processes in different societal domains, as well as potentially transformative interactions between these domains. This can serve as a useful complement to quantitative study.

On this basis, we approach the puzzle of Stavanger’s free public transport experiment with the aim of pointing to some of the broader systemic learning that may result from the free public transport provision. Proponents argued that it was an effective means to justly allocate municipal funds saved in the pandemic during 2020–2022, by making them broadly available to lower the burden on household budgets during a period of high inflation and rising interest rates. Conversely, opponents argued that it was an expensive, ineffective and wasteful way to fund public transport, and that increasing routes would have been more effective. These debates played out throughout the period, most prominently in the national newspaper anchored in the city, namely Stavanger Aftenblad. Underlying these discussions are divergent ideologies of the role of the public sector in transport provision, of what justice means in this context, and what types of systemic shifts are necessary and desirable.

Our dynamic evaluation of the experiment points, in particular, to the political effects of the intervention, which achieved a politicization of public transport but also a clear political backlash. While systemic effects are too early to assess and require robust follow-up, we point to some potential systemic processes that may result from this experiment. Our key contribution to the literature is to illustrate the contrast between traditional, quantitative evaluations based on a set of pre-determined parameters on the one hand, and the wider systemic learning processes in focus for dynamic evaluation on the other.

The essay proceeds as follows. We review literature on mission-oriented measures, social equity, and transport policy evaluation (Section 2). This leads into our analytical framework of a systemic-processual evaluation framework for transport policy towards general welfare (Section 3), followed by a brief overview of methods to apply this (Section 4). We then present our empirical analysis (Section 5), followed by a discussion which reflects upon the implications of our dynamic evaluation (Section 6), and the conclusion (Section 7).

2. Literature review: Evaluating transport policy for social equity

2.1. Social inclusion and equity

There is an emerging discussion on how to evaluate justice in public transport [2–3]. There is some consensus that transport justice and equity assessments require a fuller picture of accessibility, accounting for the multiple peoples and effects of transportation systems in society [13,14]. Allocations of transportation services to fulfil expressed mobility needs, overcome past deficits in accessing socio-economic opportunities through transport services, and open new forms of socio-economic opportunities are essential features of just and equitable transport ecosystems [15,16]. Thus, from the standpoint of diverse transport users, just and equitable outcomes would be defined by measuring if they are mobile enough to reach places of value without being unfairly burdened by detrimental trade-offs [17].

However, public transport system should not just be considered for the effects on the individual. Public transport systems play a systemic role in society, being a key policy instrument of governance for welfare redistribution in societies [18]. Yet the predominant technical-rational approach focused on transport infrastructure and its operation is largely disengaged from contextual and comprehensive policy processes more common in other domains [19]. This may be attributed to public transportation issues being largely ensconced in the urban planning paradigm [20], whereas the complexity and system-wide implications of public transport make it much more of a public policy issue that local urban governance is ill-equipped to tackle [21]. The intertwining of public transport with broader policy and politics transcending urban planning boundaries can be seen during times of financial crisis, when austerity in national policies is instantly reflected in urban transport systems that usually function independently [22]. Thus, assessments of cross-societal benefits, burdens or trade-offs and their mitigation, public participation and adaptive learning in policymaking are critical measures of socially-relevant transport policies.

Public transportation as a sector further transcends local boundaries due to the emissions it creates, forming uneasy networks among international institutional frameworks to tackle climate change, corporate interests and political positions [23]. Effects spanning individual to global scales and multiple levels of actors and processes strategically or tacitly involved in public transportation systems make these systems resemble a contemporary common pool resource of society [24]. This conceptual thread of transportation systems as a common pool resource connects the individual or societal perspective of justice in transportation and the state policy perspective [25]. This enables conceptual convergence between aggregate and usually quantified planning and policy parameters (e.g., equity, distribution of benefits and costs, harms, and risks associated with particular decisions) with more qualitative and experiential narratives of benefits and costs distribution, recognition and public goods [3]. The degree of convergence in turn can be characterized by participation and inclusion of diverse actors as stakeholders. A challenge here is the rift between elite actors with heavy identification with automobility and considerable policy influence, and marginalized population groups who often bear the burden of limited prioritization of public transport services [26].

In sum, despite considerable work on mobility justice, there is a methodological gap in relation to assessing justice implications in and of transport systems, and specifically a need for evaluative frameworks that go beyond individual measures to consider the broader, systemic effects of transport policies.

2.2. Evaluating free public transport experiments policy

The number of academic studies that review free public transport policies is relatively limited (but see, e.g., [8,27], considering the vast amount of research conducted on more widespread solutions for just urban transport transitions. Brussels introduced free public transport for students in 2003–2004, and de Witte and co-authors [28] conducted a quantitative survey and in-depth interviews to evaluate the effects of the study on transport behavior. They found that there was an increase in public transport use among the students that benefited from the measure, but that there were significant variations between groups. Overall, travel patterns do not appear to change significantly when public transport is made free. Fearnley [8] examines a handful of free public transport initiatives across European cities, including Stavanger’s previous attempt with such a policy in 2011. The study finds that the main effect of the initiatives is to significantly increase patronage. At the same time, the general picture observed in most cities is that the increased patronage is overwhelmingly people who alternatively would have walked, cycled or not travelled at all [8].

We know less about the wider systemic effects of free public transport policies. Kęblowski [29], in a review of free public transport policies worldwide, argues that existing assessments of fare abolitions have been mostly of a techno-economic nature, which ignore the social, political, and power dynamics in which fare-free experiments are embedded. Most existing evaluations are survey-based and report on modal shares resulting from free transport experiments [8,27]. Although they provide specific insights on the effects of free transport policies, these surveys do not systematically assess if the policy contributes to a more comprehensive shift away from car mobility across societal domains.

Thus, there is limited engagement in evaluative studies with the qualitative aspects of urban transport transitions; a lacuna that seems significant given the stilted and polarized nature of transport debates characterized by commitment to elitist automobility in transport planning legacies [5] and the neglected need for effective public transport services that have disproportionate adverse impacts on marginalized population groups who cannot buy their way out [30]. We argue that a qualitative and dynamic approach to evaluation can bring in important contextual aspects that matter a great deal for understanding the systemic and user-centered impact in a splintered sectoral political economy.

3. A systemic-processual evaluation framework for transport policy

To assess if a transport policy really brings about systemic changes, we need a perspective able to comprehensively capture its impacts across societal domains. Our concept of “dynamic evaluation” draws on previous uses of the term by Feinstein [10], who sees it as a type of evaluation that tries to contribute to transformative change, and Soui and co-authors [11], who see it as evaluating the learning capacity of a system. Drawing on these, for the purpose of this essay, we understand dynamic evaluations as an assessment of the systemic learning for the purpose of shifts towards sustainable, just transitions in transport.

We lean on the sustainability transitions perspective to understand the system shifts towards sustainability. The various frameworks developed in the socio-technical transitions literature (e.g., Multi-level Perspective, Social Practice Theory) together provide such a perspective, as they have identified key sets of processes that mark fundamental societal transformations [31]. For a traditional summative policy evaluation, it would be needed to assess all policy outcomes on socio-technical systems, actor-networks and institutions and then assess to what extent these are in line with the formal objectives of a policy. However, as observed by Bergek and Haddad [32], in earlier stages of policy implementation it makes more sense to analyze if a policy led to some key intermediate processes indicating transformative change.

In a dynamic evaluation, the aim should be to identify if the policy brought about any key transformative processes in different societal domains, as well as potentially transformative interactions between these domains. During transitions socio-technical change processes take places in various environments, where different struggles take place against incumbent actors and institutions. Enabling real-time assessment of an ongoing process, our evaluation strategy is theory-based, and we identified key transformation processes for each societal domain as described in previous literature The transition literature usefully identifies multiple environments in which societal embedding of innovations such as free transport takes place (i.e., business, policy, culture and users) [33,34]. In each of these societal domains, distinct struggles between policy proponents and incumbent actors can be identified, that together shape implementation. We adapt and specify these domains for the public transport sector (Table 1). We use the category of “public administration and public transport organizations” instead of “business”, given the specific organizational structure of public transport. The category of “users and user practices” captures a contextualized understanding of users and their travel behaviors [35]. (Non)-changes to different transport modes should be understood in relation to user’s daily lives. Given the highly politized nature of transport policies these days [4,36] we explicitly identify the domains of “politics” and “culture and public debate” in our evaluation framework. To be able to assess the directionality of the policy, we selected processes in line with our normative evaluation criteria of individual, institutional and transboundary justice as described in 2.2. For example, we assess to what extent outsider actors were able to voice their concerns in the public debate.

thumbnail
Download:
Table 1. Key transformative change processes in different societal domains.

https://doi.org/10.1371/journal.pclm.0000604.t001

One of the key points of the transition literature is its emphasis of non-linear, systemic forms of causality and careful process-tracing of system interactions [37]. Particular attention is paid to positive feedback loops between processes in heterogenous elements, which have been referred to as “cumulative causation” or “motors” across different literature strands [38]. For instance, the idea of “transition motors” refers to positive feedback loops that cross different societal domains. When an introduced policy leads to changes in infrastructure (e.g., more charge points), that then leads to different use practices (more electric vehicle driving), further stimulating infrastructure investment.

To summarize, we carefully study the emergence and impact of the policy of free public transport in the context of the urban transport system in Stavanger. First, using a process-tracing approach we map the emergence as well as intended and unintended outcomes of the policy of free transport in Stavanger, for different time periods (done in section 4: empirical analysis). Second, we analyze whether the policy brought about any key transformation processes across different societal domains (Table 1) (done in section 5: discussion). We pay particular attention to potential positive feedback loops between changes in different domains, which might indicate that the policy-induced changes are gaining momentum. Herewith, we reflect on the systemic impacts of the policy at the urban level, capturing its system-level additionality [32]. This normative and systematic approach informed by rich contextual awareness characterizes the dynamic evaluation that we argue for. Table 1 also indicates related work on evaluation in transition studies that our approach builds upon.

The purpose of the empirical analysis, in the following section, is to illustrate the broader systemic processes at play, which are not captured by the commissioned quantitative evaluation, in order to contrast the different approaches to transport policy evaluation.

4. Empirical analysis

We draw upon data from several sources for this essay. As part of a larger project, one of the authors conducted a workshop with a wide range of 13 transport sector decision-makers on 11 November 2022, and supervised numerous Master theses and internships on urban transport related themes in and around Stavanger during 2020–2023. This enabled familiarity with sectoral debates and developments, as well as discussions with experts at the municipal and regional level during events with an urban transport transition theme in Stavanger. Since the intervention was first announced in May 2023, we tracked main media debates until December 2023. Here, we focus on the most relevant of those published in a national newspaper headquartered in Stavanger, namely Stavanger Aftenblad, based on simple use of the search function on its website, for instance for (‘gratis transport’) free transport. We limited scope to articles from the intervention announcement onwards until the period ended. We combined this with desk study of peer-reviewed literature on just transport transitions in Stavanger and similar urban settings, in addition to the themes covered in the literature review. We also consulted supplementary online resources, such as media coverage of transport related controversies in Stavanger going back to 2018, when the Climate and Environmental Plan 2018–2030 was formally adopted. Notably, this includes a period of heightened debate on road tolls, led by a single-issue political party that emerged from Stavanger and gained significant traction in the 2019 local elections, only to fade over the years. On this robust basis, we analyze three phases leading up to, during, and in the immediate aftermath of the short-lived yet substantial free transport intervention in Stavanger.

4.1. Phase I: 2018–2023

In the five years preceding the intervention, Stavanger Municipality had committed to acting upon its Climate and Environmental Plan 2018–2030. This envisaged an ambitious 80 percent reduction in car usage and 70 percent of trips completed using active and collective transport by 2030. The trend was not going in this direction at all approaching the required rate by March 2020 when the Covid-19 pandemic disrupted usual urban transport practices significantly until September 2021. During this 18-month period, personal automobile use increased across Norway, as people were encouraged to avoid exposure to other people during any necessary travel, and Stavanger witnessed a similar trend. Tellingly, annual trips using public transport per 1,000 persons, which had traditionally been in the mid-50,000s but gone up to 62,532 in 2019, fell to 43,834 in 2020 and recovered only partially to 50,093 in 2021. (Based on Statistics Norway data on key figures for transportation in municipalities and counties, last accessed on 11 April 2024 here: https://www.ssb.no/statbank/table/11822) This makes it unreasonable to compare performance during this period with the required transition, given the unexpected and debilitating societal impact of the pandemic.

Yet by late 2021, Stavanger Municipality was faced with the very real dilemma of how to get back on track towards achieving its ambitious goals for a time-bound mobility transition. While registered petrol and diesel cars in Stavanger declined from 25,702 and 23,958 in 2019–21,975 and 21,873 in 2021 respectively (a combined reduction of 5,812 cars), electric and hybrid cars rose from 8,707 and 4,704 in 2019–14,188 and 7,365 in 2021 respectively (a combined increase of 8,142 cars), marking an overall increase of over 3.5%: 2,330 cars, or 2,320 cars if including gas cars which decreased from 26 in 2019–16 in 2021. (Based on Statistics Norway data on registered vehicles by region, fuel type, statistical variables and years, last accessed on 11 April 2024 here: https://www.ssb.no/statbank/table/11823/) With only nine years left until 2030, the municipality did not see momentum shift back in favor of public transport in 2022 compared to pre-pandemic usage levels. Various feedback loops between different societal domains further locked in the car-centered transport system [41].

People had become used to using cars, and those who had acquired them recently were tied into a period of ownership that tends to run several years for cars. Low usage during the pandemic had also meant reduced service provision requirements for Kolumbus to fulfil, based on specifications of Rogaland County. These were gradually upped towards their previous levels. Even though revenue from ticketing had fallen massively during the pandemic period, and did not immediately recover to past levels in 2022, Stavanger Municipality (and other municipalities in Rogaland County) still saved significantly on public transport provision costs, given that ticketing revenue makes up only a small fraction of this total cost.

Hence, the decision in May 2023 can be understood as a political response to three factors. First, a sense of urgency to make more rapid progress aligned with the Climate and Environmental Plan, with only seven years (instead of 12 originally) left to achieve massive reductions in car usage and promote over two-thirds of passenger trips using public and active transport. Second, an intent to distribute funds saved during the pandemic period in a manner with socially equitable impact. This was emphasized by the political proponents who devised the scheme, and saw it as a means to distribute the municipal savings to those in need in a widely-accessible manner. Those using public transport in Stavanger – among the richest municipalities in Norway – were among the more needy of its residents, or so they argued.

Third, with elections coming up, there was some truth to the political opposition’s charge that this was an election campaign ploy – the political leadership of Stavanger Municipality did need some sort of major success story to boost its chances. The intervention certainly captured public attention. Politically, it was a gamble seen as worth taking for the potential payoff, with the risk being some critique but nonetheless the rapid use of unspent funds towards a purpose that would undoubtedly deliver some public value, by offering residents one essential service for free over an extended period.

Further back, it is worth noting that Stavanger has lavished funds on many transport sector innovations, from a free public transport scheme back in 2011, to adopting the first electric buses in Norway in 2015 (then more in 2017), and trialing autonomous vehicles in recent years. These have been argued to be pilots that avoid countering the car-centric transport model [41]. Hence, more than its predecessors, the free-transport scheme was a break from the car-centered political ideology dominant in local politics. Such a radical intervention is quite rare in a sector typified by experimentation that rarely challenges the hegemony of automobility [41], hence the debate it stirred can be considered quite predictable.

4.2. Phase II: June-November 2023

The free public transport policy announcement in May 2023 took everyone – including the regional transport operator Kolumbus – by surprise, emerging from a coalition of six political parties whose leaders announced it overnight. Immediate cries of an election ploy ensued. This pushback became more extended in vigorous media debates both for and against the measure. A key point of contention was the very feasibility of implementing the planned intervention, with the opposition claiming it was not doable and would be far more expensive. However, promoters did manage to formulate a scheme that Stavanger Municipality and Kolumbus were able to implement with a month’s notice, on 3 July 2023.

The part that deserves most credit is the smooth rollout. Despite critics, Stavanger Municipality ensured relatively clear communication about how the scheme worked. It tied in closely with existing practices of different user groups. Residents were able to log on using their mobile phone number, linked to their national identity login on the municipal website – something every legal resident of Norway possesses – or were able to assist friends or family members who were eligible as residents of the municipality but unable to navigate the digital interface themselves. Those without a mobile phone had the option of going to Kolumbus’ head office in central Stavanger next to the bus terminus (a very accessible hub) and getting a free monthly ticket placed on their physical Kolumbus card upon confirming their residential status in Stavanger municipality. Less than 1 percent of users availed of this option during the scheme period, indicating that users were sufficiently digitally literate and that even the small proportion that did require alternate options were afforded scheme access, which was manageable without overloading the manual service.

A key question in initial months was what the intervention would cost in practice. The number for July was about 26 million NOK, which led to some speculation that the costs would shoot up to run through the 200 million NOK budget very quickly once people returned from summer holidays to regular work rhythms in August. Yet the demand remained relatively constant around 25 million NOK a month, varying by a couple of million either way, and with a retrospective reduction of 15 million NOK in cost for month passes issued to users who subsequently used them negligibly (for details, see https://www.stavanger.kommune.no/vei-og-trafikk/gratis-kollektivreiser/). This meant that the original budget allocation would have sufficed for about nine months – less than the year expected by scheme proponents, but three-quarters of the way to the target. Since the scheme was rolled back after six months, a third of the budget allocation (66.5 million NOK) remained unspent, and was freed up for the coalition led by the political opposition – voted into power in September 2023 – through transfer to the municipal climate and environmental fund for relevant measures.

In the autumn, Stavanger Municipality in cooperation with Kolumbus launched a tender inviting bids to evaluate the intervention. Yet the results of this study became available in December 2023, meaning it would only shed retrospective light on the scheme. The tender was awarded to a bid led from the Oslo-based Institute of Transport Economics, which also meant it would be informed by quantitative information about the scheme impact far more than by everyday lived experience of its unfolding. Our parallel examination of not only media debates but the everyday rhythms of public transport usage during the course of this intervention suggests that there is value in also contributing this perspective, without the quantitative basis in fine-grained numbers on scheme impact of the official evaluation. A key point for us is that throughout the intervention, it remained clouded in uncertainty about its timespan, taking away much of the potential of lasting changes to users’ practices associated with a major transport intervention with a sizable span of even a year, as this one was originally announced to have.

4.3. Phase III: Immediately after December 2023

The formal evaluation commissioned by the transport provider Kolumbus evaluated the free public bus experiment by (1) using automatic passenger counts to study the change in the total number of boardings in public transportation in Stavanger municipality compared to the rest of Rogaland county, and (2) using surveys to study the changes in travel habits on public transportation, cars, and bicycles [7]. The evaluation considered data until October 2023. It found an increase of 11 percent per day in number of boardings, and a reduction of car use of 17 percent when comparing the city (where the free buses operated) with the wider region. The report notes several caveats and measurement weaknesses behind these numbers. The evaluation concludes overall that, in line with similar experiments elsewhere, there is an increase in the use of public transportation trips as a result of the free provision, and that there is a decrease in car use, while other effects are uncertain [7].

In this sense, one might say that the evaluation shows that the free public transport experiment was successful – although the achievements must be seen in relation to the costs and the alternative investments the municipality could have made with the same funds. Judging from the media debate during the implementation period, which died down thereafter, critics hardly accepted these numbers as proof that their objections had been misguided. Rather, as we saw from the media debates, numbers were used selectively and strategically to support widely divergent claims while the spotlight was on a radical sectoral intervention. Thereafter, the pendulum swung back to business-as-usual approaches.

In terms of academic assessment and debate, the traditional evaluation certainly has merit and value. However, in the perspective we advance here in this essay, we need a broader type of evaluation to understand the systemic effects on justice and draw attention to the wider political economy of sectoral decision-making that thwarts efforts to support popularization of public transport in ways that can actually become credible alternatives to entrenched automobility practices and infrastructures. The design question for planners and policymakers would be one of profiling and addressing comprehensive costs and benefits for the diverse demographics of Stavanger. Expanding public value of transport infrastructure [15] through not only cost reduction but through greater opportunities and increased welfare will be critical to broader public support and participation. Similarly, managing risks and developing incentives and sanctions for long-term policy actions will be critical, in particular addressing the risk of backlash politics in transport policy [4]. In short, there are many key questions that are not answered by this evaluation, particularly regarding justice.

5. Discussion: Towards a dynamic evaluation?

Our analysis above illustrates that quantitative evidence of increased ridership does little to alter the political rhetoric and public debate on the effectiveness of a free public transport, during and after the experiment. Also, the isolation of the quantitative and qualitative aspects does not reveal any temporary and lasting societal impacts of this experiment, particularly in terms of social justice. Therefore, a dynamic evaluation of free public transport in Stavanger should evaluate not only ridership and modal preferences, but look more broadly at (1) whether it combatted injustice in transport provision, and (2) whether it enhanced progress towards the sustainability mission. Here, we discuss these two points in more detail, to suggest what dynamic evaluations would look like. Our discussion is broadly attentive to key transformation processes in the societal domains we emphasise in Table 1: politics; users and practices; public administration and public transport organizations; and culture and public debate.

5.1. Did the intervention combat injustice in transportation?

There is no data on usage among specific user demographics, but in principle, the free provision combatted injustice by making transport accessible for those who could not otherwise have afforded it. Unlike many socially regressive interventions like the national subsidies and incentives for electric cars that primarily benefitted wealthy households [26], the free public transport intervention primarily benefitted users of public transport, a demographic group that maps on to lower-income and environmentally responsible users. In this respect, the intervention was a break with the dominant car-centered ideology, and socially progressive, like other recent, more modest measures by Stavanger Municipality including subsidies for electric bicycles for families and businesses. Moreover, the qualitative analysis of the program’s rollout highlights the importance of coordination among co-existing societal infrastructures for equitable delivery of public services. The existing data infrastructure at the disposal of the city was leveraged to maximize reach to the public, and the quality of user experience with the mobile application was a key factor in participation. The degree of media coverage can also be considered a mobilizing factor of public opinion throughout the spectrum of transport users, bringing attention to the practicality of addressing public transport as a climate and social welfare policy question, in addition to it being an element of the Stavanger climate and environmental action plan. This pilot can be expected to serve as a public and institutional memory reference point, leading to more systemically aware plans and actions. This inference is obviously based on the secondary information the authors have mined from varied sources. A formal dynamic evaluation with the appropriate data strategy would have been able to investigate and establish these correlations with greater rigor.

As we argue above, politics, culture and societal debate (Table 1) are key elements of a dynamic evaluation. A critical factor is whether a policy intervention sparks backlash, shifting the overall system in the direction of more car dependence. Thus, alongside drawing the public’s attention, the Stavanger experiment clearly politicized public transport, turning it into one of the main election issues ahead of an election that saw a shift towards car-friendly parties to the right on the political spectrum. These parties - the People’s Campaign No to Toll Roads and the Industry party, have campaigned heavily in the past against green policies in transport. This has had considerable appeal among suburban voters, even though its main proponents included affluent dwellers with big houses and automobiles on urban peripheries. In Stavanger, the People’s Campaign once pulled a publicity stunt by setting up a fake toll station on a bicycle lane, arguing that bicyclists should pay to finance transport infrastructure they benefit from.

A key effect of the experiment, in our assessment, has been the politicization of public transport. This is of course difficult to measure quantitively, and the effects are highly dynamic in the sense that they change in a non-linear way over time. In the short term, it seems to have contributed to a shift of political power to more car-friendly political parties – or at least, it was not sufficient to halt this shift ahead of the 2023 local elections. Transport policies became a hot topic in the media, and the public were exposed to strong argumentation pro et contra free public transport, and debates about the role of public transport in society in general. The effects of this over time may be to highlight what is lost if public transport is undermined and make more publicly visible which parts of the citizenry rely on public transport. In this sense, the long-term effects of politicization may be the legitimation of a more expansive and more just transport system – or the contrary. The key point is that the politicization of changes to public transport (or any public infrastructure, for that matter), illustrates why decision making should be cognizant of and pursuant of parameters beyond singular objectives like increases in bus ridership. This also relates to the potentially evolving role of public administration and public transport organizations, since these could develop important new competencies through such interventions; ensuring access to the scheme for all users including the sub-1% who were not on digital platforms is a case in point here, that others can take instruction from as transport systems rapidly digitalize.

5.2. Did the intervention enhance transformational progress towards sustainability?

The quantitative evaluation found what we can call a modest increase in public transport ridership. To understand the effects in terms of progress towards sustainability, there are several caveats to this way of evaluating the experiment.

First, the impacts vary by individual circumstance to an extent that any evaluation will have trouble unpacking in full. One could build up some archetypes of households, but the complex factors that configure transport needs and preferences would make this too impractical a typology to implement if one took complexity adequately on board. Second, the intervention was always shrouded in uncertainty, with the year-long duration criticized by sceptics from the outset, and eventually shortened to six months on the heels of the local election results. This meant that residents in effect never adjusted their practices as if there was a sizable period of free public transport on offer, just making the most of it while it lasted. This limits the impact one can expect or read into the intervention, hence researchers should take great care not to make any broad generalizations dismissing other such interventions based on this case.

Third, it is unclear what can be concluded regarding the changing demand for transport services. Since the uncertainty surrounding the intervention is unlikely to have caused residents to make long-term changes to their commuting practices (those who had cars would continue having them, those who did not would likely continue without them at least while free public transport was being offered), this result is in line with what one would expect. Changes to practices and behaviors around technological systems at the individual-level are heavily influenced by the sentiments of certainty, durability and stability of the public infrastructure ecosystem those technologies are part of [42,43], otherwise known as public trust.

A key lesson for the governance and political process of policymaking in transboundary systems can be drawn from the polarizing effect this sudden and opportunistic intervention evoked. There was insufficient articulation of and engagement on the cost-benefit proposition of the free transport measure in terms of climate, social welfare, new recreational and economic opportunity. As a result, the public debate was confined to beliefs and assumptions of commentators. Strengthening the civic debate on public transportation through a nuanced presentation of the multiple attributes associated with the history and future prospects of transport sector in Stavanger could have contributed to a constructive politicization of the issue. This is an important aspect of the contribution of radical interventions: an opportunity to reframe public attitudes for better alignment with ambitious urban climate and environmental plans that require rapid shifts to collective systems of transport provision. Dynamic evaluation that integrates the climate policy context and informs public debate about transport and urban planning can complement a systematic quantitative evaluation by contextualizing such key local aspects.

In sum, the intervention changed little in terms of progressing towards a sustainable transport system in the short run. But it would be harsh to judge it as a failure due to this. Did the intervention fail to enhance this progress, or did other systemic factors that hold back this progress simply stay in place and remain out of reach of this intervention, much as they have been out of reach of other, smaller progressive transport schemes?

Using our analytical scheme of change processes in different societal domains, we observe transformative change in the domain of politics. This intervention did have strong backing by six political parties, and constituted a break with the car-centered political ideology. Hence it can be considered a potentially transformative political effort – combined with the opportunism of having cash available and the timing of the local elections, certainly. Nevertheless, it was an intervention that broke with most of the mainstream policies usually considered, and arguably shifted the terms of the public debate on transport. However, the intervention failed to bring about change processes in other societal domains, as well as positive feedback (transition motors) between them. In the domain of user practices, although the rollout on the whole was quite smooth, the constant uncertainty regarding the scheme likely prevented meaningful changes in practices bundled together with commuting, such as dwelling, working and groceries (another key societal domain highlighted in Table 1).

Implementing a dynamic evaluation underscores the value of framing an evaluation within its broader context, rather than in a narrowly focused manner. If additional data was available, with fine-grained attributes of users along demographic lines, or usage time intervals, it would be informative to run quantitative analyses of those patterns simply due to data availability and to thus identify usage trends. However, simply describing use patterns and demographic in more fine-grained way would not necessarily help advance a transformation of the transport system in a sustainable direction.

Given that five of wealthy Norway’s ten richest municipalities are in the Greater Stavanger Region, it is hardly due to the allocation of 200 million NOK to free public transport that better service provision does not exist. It is an artefact of a car-centric culture where the wealthy buy out of the public provision system by owning and using personal automobiles, of which the most elite – electric vehicles – even receive subsidies and incentives from tax exemptions to reduced tolls and parking fees and until recently even permission to drive in bus lanes. Meanwhile, public transport is arguably not sufficiently prioritized. There is a lack of public transport to the main destinations for leisure daytrips, such as to attractive sandy beaches, the Rogaland County arboretum, and popular hiking areas in the nearby mountains [6]. There are far lower public transport offerings on the weekends, rather than running special services to recreational spots as many other mid-sized cities do. This makes public transport a limiting choice [44]. Making it free is not enough to shift practices of car owners, who are bound by a system of mobility that offers far more perks to car use than public transport users can avail.

Acknowledging that pilot programs should be treated in the spirit of learning for implementation insights and probing the systemic expanse of the problem-solution sphere [45], the free transportation pilot can potentially offer several such lessons beyond success or failure judgements. However, it must also be stressed that post-hoc evaluations are usually less effective in fulfilling this learning purpose [46]. Stavanger’s pilot can thus be criticized for not commissioning a parallel evaluative learning program at planning and rollout stages. The smooth rollout of the program could otherwise have been an early lesson and might have informed adaptations to other aspects of the pilot, for instance, a stronger tracking of the anticipated climate goals, user behaviors and shifts in mobility pattern at different phases of the pilot. By 2025, one sees little learning from the 2023 intervention taken forward.

A single experimental intervention can likely not work by itself as a transformative transport policy measure – it needs to operate a transition motor in conjunction with other measures and processes. Even for the political proponents, the intervention cannot realistically have been seen as a measure to move Stavanger Municipality towards achieving its ambitious 2030 goal of an 80 percent reduction in car usage with 70 percent passenger trips being made by public and active transport. Achieving such a target without massive, systemic changes in Stavanger’s car-centric system of mobility is, arguably, unachievable. The intervention did not change systems of mobility, just the financial incentive to participate more heavily in public transport modes within the existing system.

6. Conclusion

The purpose of this essay has been to argue for evaluating transport interventions in a systemic scope, what we termed dynamic evaluation. There is a need to go beyond the traditional evaluations centered on quantitative measures of ridership, satisfaction, and modal preference, and take account of the broader political, social and cultural effects of transport measures in light of the longer-running sectoral political economy and local planning context. We have discussed Stavanger’s short-lived policy of providing free public transport in this light, and pointed to broader processes and effects that would not be captured by traditional evaluative frameworks. The broader systemic effects of transport policies include the way that they politicize the role of transport in society, for example by making the societal effects of the politics visible in new ways; how they spur infrastructural reconfigurations that shift transport practices in the longer term; or how they change the cultural meanings attached to mobility and movement.

Set in this context, it is relevant to question how the free public transport policy in Stavanger, for example, brought transport injustices into public view in new ways. The fact that public transport became hotly debated in the local media, where both proponents and opponents of the measure debated public transport policies in terms of justice, distribution and societal effects, can potentially change the position of public transport in public discourse – shifting it from a technical administrative issue to a key issue of ideological contestation. This does not necessarily mean that there will be more social justice-oriented policies, but it likely means that transport becomes a justice-type issue, where arguments have to relate to the wider societal effects of measures that are implemented. This is an important and necessary step given the dominance of automobility, whose entrenchment in sectoral decision-making holds back necessary just urban transport transitions. Indeed, radical interventions may play into the backlash politics currently characterizing the transport sector, stimulating various types of populist, right-wing responses to green policies [4]. This politicization of public transport can be a “transition motor” that sparks positive feedback loops across the societal system.

These wider systemic effects are what the concept of dynamic evaluation attempts to capture. Change is non-linear and occurs in interaction between different societal domains, as illustrated by Table 1 above. The traditional evaluations of ridership, satisfaction, modal preference, and so on have an important function. But we need to be able to offer analyses of the way concrete policies influence and create positive feedback loops across societal domains, sparking broader systemic transformations of the transport system.

References

  1. 1. Martens K., 2016. Transport justice: Designing fair transportation systems. Abingdon: Routledge.
    • 2. Holden E, Banister D, Gössling S, Gilpin G, Linnerud K. Grand Narratives for sustainable mobility: A conceptual review. Energy Research & Social Science. 2020;65:101454.
    • 3. Verlinghieri E, Schwanen T. Transport and mobility justice: Evolving discussions. J Transp Geogr. 2020;87:102798. pmid:32834675
    • 4. Wanvik TI, Haarstad H. Populism, instability, and rupture in sustainability transformations. Ann American Association of Geographers. 2021;111(7):2096–111.
    • 5. Sovacool BK, Axsen J. Functional, symbolic and societal frames for automobility: Implications for sustainability transitions. Transportation Research Part A: Policy and Practice. 2018;118:730–46.
    • 6. King K, Tråsavik H, Sareen S. Take a hike: Spatializing allemannsretten and transportation accessibility for outdoor recreation in the Greater Stavanger Region, Norway. Norsk Geografisk Tidsskrift-Norwegian J Geography. 2024;78(5):324–37.
    • 7. Gregersen FA, Hartveit KJL, Christiansen P. Evaluering av gratis kollektivtransport. 2023.
      • 8. Fearnley N. Free Fares Policies: Impact on Public Transport Mode Share and Other Transport Policy Goals. IJT. 2013;1(1):75–90.
      • 9. Porter T. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. Princeton: Princeton University Press;1995.
        • 10. Feinstein O. 2019. Dynamic evaluation for transformational change. In: van den Berg R; Magro C, Salinas Mulder S. (Editors). Evaluation for transformational change: Opportunities and challenges for the sustainable development goals. IDEAS: Exeter. 17-31.
          • 11. Soui M, Abed M, kolski C, Ghedira K. Towards a dynamic evaluation of personalized information system based on multi-agents approach: Application in intelligent transport. 2011 4th International Conference on Logistics. 2011:68–75.
          • 12. Hickman R. and Banister D. Transport, climate change and the city. Abingdon: Routledge; 2014
            • 13. Pereira RHM, Schwanen T, Banister D. Distributive justice and equity in transportation. Transport Reviews. 2016;37(2):170–91.
            • 14. Sheller M, Urry J. The New Mobilities Paradigm. Environ Plan A. 2006;38(2):207–26.
            • 15. Hananel R, Berechman J. Justice and transportation decision-making: The capabilities approach. Transport Policy. 2016;49:78–85.
            • 16. Stead D. Effectiveness and Acceptability of Urban Transport Policies in Europe. International Journal of Sustainable Transportation. 2008;2(1):3–18.
            • 17. Oswald Beiler M, Mohammed M. Exploring transportation equity: Development and application of a transportation justice framework. Transportation Research Part D: Transport and Environment. 2016;47:285–98.
            • 18. Beyazit E. Evaluating Social Justice in Transport: Lessons to be Learned from the Capability Approach. Transport Reviews. 2011;31(1):117–34.
            • 19. Marsden G, Reardon L. Questions of governance: Rethinking the study of transportation policy. Transportation Research Part A: Policy and Practice. 2017;101:238–51.
            • 20. Rode P. Urban planning and transport policy integration: The role of governance hierarchies and networks in London and Berlin. Journal of Urban Affairs. 2017;41(1):39–63.
            • 21. Bai X, McAllister RR, Beaty RM, Taylor B. Urban policy and governance in a global environment: complex systems, scale mismatches and public participation. Current Opinion in Environmental Sustainability. 2010;2(3):129–35.
            • 22. Veeneman W, Augustin K, Enoch M, Faivre d’Arcier B, Malpezzi S, Wijmenga N. Austerity in public transport in Europe: The influence of governance. Research in Transportation Economics. 2015;51:31–9.
            • 23. Marsden G, Rye T. The governance of transport and climate change. Journal of Transport Geography. 2010;18(6):669–78.
            • 24. Šestáková A, Plichtová J. Contemporary commons: Sharing and managing common-pool resources in the 21st century. Human Affairs. 2019;29(1):74–86.
            • 25. Nikolaeva A, Adey P, Cresswell T, Lee JY, Nóvoa A, Temenos C. Commoning mobility: Towards a new politics of mobility transitions. Trans Inst British Geog. 2019;44(2):346–60.
            • 26. Remme D, Sareen S, Haarstad H. Who benefits from sustainable mobility transitions? Social inclusion, populist resistance and elite capture in Bergen, Norway. Journal of Transport Geography. 2022;105:103475.
            • 27. Cats O, Susilo YO, Reimal T. The prospects of fare-free public transport: evidence from Tallinn. Transportation. 2017;44(pp.):1083–104.
            • 28. De Witte A, Macharis C, Lannoy P, Polain C, Steenberghen T, Van de Walle S. The impact of “free” public transport: The case of Brussels. Transportation Research Part A: Policy and Practice. 2006;40(8):671–89.
            • 29. Kębłowski W. Why (not) abolish fares? Exploring the global geography of fare-free public transport. Transportation. 2019;47(6):2807–35.
            • 30. Mattioli G, Roberts C, Steinberger JK, Brown A. The political economy of car dependence: A systems of provision approach. Energy Research & Social Science. 2020;66:101486.
            • 31. Köhler J, Geels FW, Kern F, Markard J, Onsongo E, Wieczorek A, et al. An agenda for sustainability transitions research: State of the art and future directions. Environmental Innovation and Societal Transitions. 2019;31:1–32.
            • 32. Bergek A., Haddad C.R. Evaluating Transformative Innovation Policy Outcomes as Unfolding Processes Of Change In Socio-Technical Configurations. In: Velez Cuartas G, Romero Goyeneche O, (Editors). Transformative Metrics. Fondo Editorial FCSH, Universidad de Antioquia. 2022.
              • 33. Jasper Deuten J, Rip A, Jelsma J. Societal embedding and product creation management. Technology Analysis & Strategic Management. 1997;9(2):131–48.
              • 34. Mylan J, Morris C, Beech E, Geels FW. Rage against the regime: Niche-regime interactions in the societal embedding of plant-based milk. Environmental Innovation and Societal Transitions. 2019;31:233–47.
              • 35. Watson M. How theories of practice can inform transition to a decarbonised transport system. J Transport Geography. 2012;24:488–96.
              • 36. Wilson A, Mitra R. Implementing cycling infrastructure in a politicized space: Lessons from Toronto, Canada. Journal of Transport Geography. 2020;86:102760.
              • 37. Geels FW. Causality and explanation in socio-technical transitions research: Mobilising epistemological insights from the wider social sciences. Research Policy. 2022;51(6):104537.
              • 38. Suurs RAA, Hekkert MP. Cumulative causation in the formation of a technological innovation system: The case of biofuels in the Netherlands. Technological Forecasting and Social Change. 2009;76(8):1003–20.
              • 39. Meelen T, Farla J. Towards an integrated framework for analysing sustainable innovation policy. Technology Analysis & Strategic Management. 2013;25(8):957–70.
              • 40. Braams RB, Wesseling JH, Meijer AJ, Hekkert MP. Legitimizing transformative government. Environmental Innovation and Societal Transitions. 2021;39:191–205.
              • 41. Haarstad H, Sareen S, Kandt J, Coenen L, Cook M. Beyond automobility? Lock-in of past failures in low-carbon urban mobility innovations. Energy Policy. 2022;166:113002.
              • 42. Antonopoulos CA, Fuentes TL, McCord KH, Rackley ALS, Biswas S. Regional assessment of household energy decision-making and technology adoption in the United States. Energy Policy. 2024;185:113940.
              • 43. Biswas S, Fuentes TL, McCord KH, Rackley ALS, Antonopoulos CA. Decisions and decision-makers: Mapping the sociotechnical cognition behind home energy upgrades in the United States. Energy Research & Social Science. 2024;109:103411.
              • 44. Tråsavik HS, Loe MR, King K, Sareen S. Leisure mobility: Situating emotional geographies of friluftsliv in urban mobility transitions. Emotion, Space and Society. 2024;50:101003.
              • 45. Smajgl A, Ward J. Evaluating participatory research: Framework, methods and implementation results. J Environ Manage. 2015;157:311–9. pmid:25929196
              • 46. Hatakeyama T. Sustainable development indicators: Conceptual frameworks of comparative indicators sets for local administrations in Japan. Sustainable Development. 2018;26(6):683–90.