Abstract
This article presents findings from conjoint and adversarially co-produced narrative survey data from 577 adults within England’s North East Combined Authority collected in the run up to the first Mayoral election in 2024. This is the first instance of conjoint experimental analysis of priorities for regional transport policy. As might be expected, we found that cheaper, quicker, more reliable journeys were always preferred to slower, more expensive or less reliable journeys. This was true of journeys to London and those within the North East. Reduced deaths caused by air pollution, congestion and transport poverty significantly increased support for transport policies while greater congestion, increased car reliance and air pollution related deaths all reduce preference. Funding options such as council and income tax increases were negatively received, whereas taxes on corporations were comparatively favoured. Ownership and cycling rates were the least influential factors in shaping public preference toward transport policy.
Introduction
The English Devolution White Paper expanded transport powers to metro mayors, including greater control over local rail and road networks and integrated transport systems, aimed at addressing regional disparities in transport delivery (UK Government, 2024). Research highlighting the regional disparities in transport connectivity across the UK calls for greater devolution to localise decision making on improving transport services in a way that serves the needs of local communities (Marioni 2024: p.14). The previous UK Government’s Levelling Up (Department for Levelling Up, Housing and Communities, 2022) report highlighted the need for local transport plans as a key means of addressing regional disparities. Liverpool City Region Combined Authority is an example outside of London where devolution has improved transport connectivity. Through its local transport plan, the city region is investing millions in an integrated transport network serving both communities and the local economy (Liverpool City Region Combined Authority, 2025; Whitham, 2025a).
The Mayor for the North East Combined Authority (NECA), Kim McGuinness, oversees seven constituent authorities: Durham, Gateshead, Newcastle, North Tyneside, Northumberland, South Tyneside and Sunderland. It represents almost two million residents in the region, which has an economy of £40 billion consisting of 5500 businesses providing 820,000 jobs (NECA, 2024b). NECA has control over a directly allocated budget of £4.2bn from a combination of Government Levelling Up and other funds (Brooke-Battersby, 2024). As such, the NECA Mayor holds significant, but bounded transport powers.
Under the Bus Services Act (2017) and the English Devolution settlement, the Mayor has the power to introduce bus franchising, bringing routes, timetables and fares under public control on the model pioneered in London and more recently Greater Manchester (UK Government, 2024). This represents a substantial potential lever, given that bus travel accounts for by far the largest share of public transport journeys in the region (NECA, 2024b). The Mayor also has strategic oversight of the Tyne and Wear Metro through the Passenger Transport Executive (Nexus), authority over local transport planning through the North East Local Transport Plan and access to a City Region Sustainable Transport Settlement that provides multi-year capital funding for integrated transport investment (NECA, 2024c). NECA has therefore expressed optimism about the potential for improvement in transport and attendant outcomes in health, employment and climate change: The North East Local Transport Plan is a statutory plan which sets out our region’s transport priorities up to 2040. The plan is centred on creating a better transport network that acts as the yardstick on which all other networks are judged. This involves creating a green, integrated transport – network that works for all. We believe this will make sustainable travel options more attractive, convenient, and safer, enabling more people and freight to make greener journeys. If successfully delivered, the projects and policies will help to enable inclusive economic growth, give people the skills to succeed, achieve better health outcomes, protect our environment, and tackle climate change by providing attractive, seamless, safer, sustainable transport for people and freight across our region (NECA, 2024a).
However, these powers are limited in important respects. The North East has historically received far fewer resources than the South East, placing pressure on services and affecting the region’s economic and social activity (Department for Levelling Up, Housing and Communities, 2022). Strategic rail services remain outside Mayoral authority, under the control of national bodies. Currently Network Rail, the nascent Great British Railways and the Department for Transport (DfT) retain authority over major rail infrastructure investment and intercity connectivity. The dualling of strategic roads such as the A1, which is critical to the region’s economic connectivity, falls within the remit of National Highways rather than NECA, and securing investment requires sustained lobbying of central government. Even in the areas where devolved powers formally exist, practical constraints remain acute. The bus franchising process is a complex statutory procedure legally required to take in excess of 2 years from initiation to implementation, reflecting both legislative requirements and the region’s distinctive urban, rural and coastal geography (NECA, 2024c). Quangos and arms-length bodies operating in the transport space such as Nexus, the statutory operator of the Metro, retain significant operational autonomy even within the Mayoral framework, and the pace of improvement is often constrained by infrastructure condition, staffing and the legacy of decades of underinvestment rather than by policy choices alone. This provides an institutional context in which Mayoral manifesto commitments are developed and via which their performance is assessed.
Given this complexity, little is known about how residents understand NECA and its role in delivering transport and responding to the myriad social, economic and cultural challenges facing the region. This is a critical knowledge deficit for public administration, given the crucial role that NECA will play in navigating a series of crises that have accumulated over several decades. With disengagement and alienation from national politics high among groups disproportionately represented within the region, there is need for understanding citizens’ views of what they believe NECA can and, importantly, should do. This is essential to understanding what is electorally feasible for a Mayor to pursue as part of a coherent transport plan. As such, it is important to understand how residents assess transport policy with regard to policy attributes of ownership, management, funding and outcomes and how they prioritise those different attributes, even if Mayoral performance depends in part on factors and bodies over which they have less clear control.
To address this gap in public understanding, we conducted an experimental survey of 577 NECA residents between 5th and 19th February 2024, around 3 months prior to the establishment of the authority and the inaugural Mayoral Election. We presented participants with a combination of a description of prospective NECA transport policy from the two key Mayoral candidates’ Manifestos, along with five narrative justifications for the policy’s implementation which were adversarially co-produced with opponents of the policy identified in a separate screening survey (Ardron et al., 2025) to assess levels of support overall and in response to further argumentation. We also used a conjoint experiment to assess specific priorities and preferences within the policy’s formulation. A conjoint survey establishes participants’ preferences between randomised discrete policy feature options, the trade-offs people are willing to make and an optimal policy version.
Conjoint analysis has an established and well-validated record in public policy research. Originally developed in mathematical psychology and marketing science (Green and Rao, 1971; Luce and Tukey, 1964), the method was formally extended to causal inference in experimental settings by Hainmueller et al. whose framework for estimating Average Marginal Component Effects (AMCEs) from stated preference experiments has been widely adopted in political science, welfare research and public administration (Hainmueller et al., 2014). Conjoint designs have since been used to study preferences over immigration policy (Hainmueller and Hopkins, 2015), social policy and taxation (Nettle et al., 2025), healthcare systems (Béland et al., 2019), energy policy (Thew et al., 2026) and, increasingly, transport infrastructure outside the UK (Aryal et al., 2022; Mertens et al., 2016). The method’s key advantages for public policy research are well-established in the literature: it forces respondents to make realistic trade-offs between competing attributes rather than expressing unconstrained approval of all desirable outcomes; it is resistant to social desirability bias because preferences are revealed through choices rather than direct attitudinal questions; and it permits estimation of the relative importance of individual policy attributes at the population level and across sub-groups, providing actionable evidence for policy design. The key means of analysing conjoint analyses are Average Marginal Component Effects (AMCEs), which measure how much the probability of selecting a particular binary choice changes when one specific attribute changes, while averaging out the effect of all other attributes The validity and reliability of AMCEs derived from online convenience samples has been extensively tested against nationally representative data, with findings consistently showing that treatment effect heterogeneity is moderate and that estimates replicate well across sample types (Coppock, 2019; Peer et al., 2022).
Adversarial narrative co-production is a newer but increasingly well evidenced method for testing the resilience of policy support under conditions of realistic argumentative challenge. The approach builds on evolutionary accounts of reason, which suggest that opponents of positions are more able to identify salient persuasive arguments in favour of those policies than supporters (Johnson et al., 2024). This ensures that the challenges to which respondents’ support is exposed are ecologically valid: they represent the most compelling opposition that real-world opponents are capable of generating, rather than straw man arguments that understate the force of the case against reform. The method has been validated in a series of studies examining welfare reform (Johnson et al., 2022), housing policy (Cooper et al., 2026) and various other aspects of public policy.
This article deploys these methods in analysing both general support and specific trade-offs within prospective NECA transport policy. We find that residents prioritise reduction in avoidable deaths by pollution, funding methods that do not increase tax burdens on most individuals directly and improve affordability of regional and national travel. We also find that there is scope for tying the highest priority (reducing avoidable deaths) to the lowest priority and least impactful narrative (environmental impact). Together, this constitutes the first and most comprehensive assessment of public opinion on UK Combined Mayoral Authority transport policy, filling a significant gap in both the transport policy literature and the literature on devolved governance and public legitimacy. The study provides the first systematic, population-based evidence on what NECA residents want from the new authority’s transport powers. This is of direct relevance to assessment of electoral feasibility for NECA transport policy at a time of flux. While the findings draw on transport policy preferences of NECA residents in the period immediately preceding the establishment of the new authority and the election of its first Mayor, findings support those elsewhere: that participants endorse policies that reduce poverty, inequality and financial insecurity (Common Sense Policy Group, 2025). The transport policy prospectus presented to respondents was derived from the plans that had been developed at that stage by the candidates and the prospective Combined Authority, including those set out in the prospective North East Local Transport Plan and Mayoral manifestos. These were appropriate plans to assess, given that residents were being asked to vote in part on the basis of their being formal commitments. The timing also ensures that responses were not conditioned by prior experience of NECA governance or anchored to announced decisions, providing a baseline measurement of preferences against which subsequent attitudinal research can be compared.
We begin by reviewing existing research to contextualise these findings within broader debates on devolution, transport equity, public legitimacy and redistribution.
Transport reform and public preferences
Once relatively prosperous in the wake of post-1945 reforms, many areas of the North East of England have suffered disproportionately from the loss of industry and the lack of investment in infrastructure and public transport. It has long been recognised that the region’s transport challenges are marked by isolation from London and the Westminster Government, competition between cities and communities and a relatively sparse population (VIGAR, 2006). The consequence, in many parts of the region, is young people forced to move away in search of work and opportunity and an ageing population confronted by increasing antisocial behaviour and chronic pressure on policing and health and care services.
Thus far, measures to address the slide have been ad hoc, with communities continuing to see their interests suffer across Conservative, New Labour, Coalition and Conservative Governments. Those measures have focused on the symptoms, not the causes, of the decline. The 2019–2024 Conservative Government’s Levelling Up agenda recognises the need for substantive consideration of causes and transport is often one of the most fundamental of those (Department for Levelling Up, Housing and Communities, 2022). The establishment of NECA in May 2024 marked a significant development in the governance of transport across the region. The North East Local Transport Plan 2024–2040 sets out an integrated strategic vision for transport in the region, oriented around four objectives: enabling inclusive economic growth; improving public health and wellbeing; protecting the environment and tackling climate change; and achieving seamless, affordable and accessible connectivity for all residents (NECA, 2024a). These objectives situate NECA’s transport remit squarely within broader debates about regional inequality, public health and environmental sustainability and give the authority a potentially transformative role in addressing the cumulative disadvantages that have shaped life in the North East over several decades.
Comparative analysis of combined authority transport governance across England highlights both the potential and the limitations of devolved transport powers. In Greater Manchester, the establishment of the Bee Network, integrating buses, Metrolink tram services and cycling infrastructure under public control, is widely regarded as a model of the kind of integrated, publicly accountable transport system that devolution can enable (Whitham, 2025a). Liverpool City Region has similarly used its devolved transport powers to invest in an integrated network aimed at improving both connectivity and equity of access (Liverpool City Region Combined Authority, 2025). For NECA, the challenge is compounded by the region’s relative economic disadvantage, its geographically dispersed settlement pattern and the condition of its inherited transport infrastructure, not least the Tyne and Wear Metro, which has operated with limited renewal since its construction in the 1980s. Marioni (2024) notes that transport connectivity disparities between London and the northern English regions persist despite sustained policy attention, and that meaningful improvement requires long-term, sustained capital investment rather than the time-limited, project-specific funding characteristic of levelling up settlements. Understanding what residents want from NECA’s transport remit is therefore not only an exercise in democratic legitimacy but also a prerequisite for designing transport policy that commands public support and delivers durable improvements in a context of acute and historically entrenched disadvantage.
3.3 million people in the North of England suffer from transport-related social exclusion (Transport for the North, 2022a). Securing affordable, effective and reliable transport is essential to economic activity, particularly for those on low incomes (Sun and Thakuriah, 2021; Zhang and Pryce, 2020) and, with it, a number of outcomes in health, education, crime and culture. Spending on public transport in the region is significantly lower than in London and other parts of the country. Over the decade to 2022/23, transport spending per person averaged just over one-third of that in London (IPPR, 2025). If there is to be Levelling Up, spending on transport is key. The impacts of transport reform are diffuse.
The concept of transport poverty has emerged as an important analytical framework for understanding the relationship between inadequate access to transport and the reproduction of social and economic disadvantage. Luce and Tukey (1964) identify transport exclusion as a multi-dimensional phenomenon shaped by the intersection of income poverty, geographic isolation and transport systems oriented primarily towards private car users. Where public transport is unreliable, infrequent or unaffordable, those without access to private vehicles – disproportionately women, older people, disabled people and those on lower incomes – face material barriers to employment, healthcare, education and social participation (Church et al., 2000; Preston and Rajé, 2007). Such barriers reproduce and deepen existing inequalities, compounding the disadvantages already associated with poverty. In the North East, where both car ownership rates and public transport quality lag behind national averages, the consequences of transport poverty are particularly acute (Sun and Thakuriah, 2021; Transport for the North, 2022b). The region’s spatial structure made of dispersed collections of towns, former pit villages and rural communities linked by an incomplete and ageing public transport network means that access to economic opportunities is heavily mediated by mobility and that those without cars face systematic disadvantage in accessing jobs, services and amenities.
UK survey evidence on public preferences with regard to transport and inequality suggests that these material realities are well understood by residents. Research consistently finds that concerns about affordability and equity are central to how the public evaluates transport policy, with low-income and rural communities reporting the strongest sense of being underserved (Annous, 2025; Frost and Singer Hobbs, 2024). One in five people worry about affording the transport they need, with low-income and rural communities especially underserved (Frost and Singer Hobbs, 2024). Nearly two-thirds feel they have limited influence over transport decisions, with rural residents (43%) feeling more excluded than those in urban areas (24%) (p.13). Insight from More in Common (Annous, 2025) based on five focus groups across Britain shows that people perceive transport as a practical issue rather than strictly a political one. Irrespective of area or political affiliation, participants consistently called for more reliable, affordable and accessible public transport, less congestion and safer roads (Annous, 2025). These findings echo earlier literature on public attitudes towards transport policy which demonstrate near unanimous support for improving public transport and a broad consensus for policies that prioritise spending towards public transport over roads, favour buses and pedestrians and reward clean car use (Goodwin and Lyons, 2010a, p.8).
There is broader public recognition that the costs of an inadequate transport system fall disproportionately on those least able to bear them. Support for transport reform that reduces transport poverty is correspondingly high, and the evidence suggests that this concern for distributional outcomes can offset, at least partially, aversion to the tax increases that such reform would require (Goodwin and Lyons, 2010b). The political salience of transport poverty has grown considerably in the context of the cost-of-living crisis, which has intensified pressure on household budgets and made the cost of getting to work, education or healthcare more acutely felt. For North East residents, for whom transport spending has historically been among the lowest in England, this context of accumulated underinvestment and financial insecurity provides an important backdrop for understanding how they evaluate and prioritise transport policy options.
The cost-of-living crisis has highlighted the role of fuel poverty on poverty and inequality overall as well as health. The increased cost of energy affects personal finances so dramatically as a result of ‘double energy vulnerability (DEV)’, which ‘is the increased likelihood of negative impacts upon well-being, owing to the intersection of domestic energy poverty (DEP) and transport energy poverty (TEP)’ (Robinson and Mattioli, 2020). Increasing concern over costs and reliability of privatised public transport mean that people may elect to rely on cars which, themselves, are highly expensive to run. In 2021, owning and running a car cost on average £3406.80 without finance and £5744.40 with finance (Ellmore, 2021). There is real reason to believe, given that poverty is a recognised determinant of a range of outcomes, that reform to transport can produce improvements across the region.
The public health burden of transport-related air pollution in the United Kingdom is substantial and well-documented. The former Director of Public Health, Newcastle, Eugene Milne, has emphasised the impossibility of addressing large public health crises through individualised approaches. He highlights public health crises averted in Ancient Rome from banning carts and 20th Century Britain from replacing horses with cars, before highlighting the ongoing crisis of pollution and climate change caused by cars. He cites three aspects of public health policy advanced in recent years with regard to transport: health protection has been improved via legislation banning drinking and driving; although the least important aspect of public health, health care services have improved via vehicular development; health improvement has been affected by reduction in active travel, such as walking and cycling, and increase in pollution (Milne, 2012). The Committee on the Medical Effects of Air Pollutants (2010) (COMEAP) estimated that long-term exposure to fine particulate matter was responsible for the equivalent of approximately 29,000 premature deaths in the UK annually, with road traffic constituting a major source of these emissions. This burden is unequally distributed: lower-income communities, children, older adults and those with pre-existing respiratory and cardiovascular conditions bear a disproportionate share of the harm (Department for Environment, Food and Rural Affairs, 2019). Woodcock et al. (2009) demonstrated through comparative modelling of urban transport scenarios that policies promoting a substantial shift from private vehicle use to active travel and public transport would generate very large public health gains in the near term, principally through reductions in physical inactivity and exposure to air pollution (Woodcock et al., 2009). The health co-benefits of such policies substantially exceeded those attributable to reduced carbon emissions alone, suggesting that the public health rationale for transport reform is both independently compelling and reinforces the climate rationale. There is, moreover, growing evidence that active travel is positively associated with mental as well as physical health outcomes across the life course (Mytton et al., 2016), with the benefits of walking and cycling extending beyond the individual journeys they facilitate to encompass broader wellbeing gains for communities and neighbourhoods. These findings make a strong case for integrated transport reform that prioritises active travel infrastructure alongside improvements in public transport, not least in a region such as the North East, where the public health burden of poor air quality and physical inactivity is compounded by broader socioeconomic disadvantage.
It is also, clearly, critical to climate and the importance of changing behaviour to move people from cars to other forms of transport has long been recognised (Kingham et al., 2001). Transport remains the largest single source of domestic greenhouse gas emissions in the United Kingdom, accounting for approximately a quarter of total emissions, with road vehicles representing the dominant contributor within the sector (Department for Energy Security and Net Zero, 2024). While the contribution of transport to climate change is a global issue, NECA’s capacity to control transport grants it regional capacity for impact. In this respect, NECA also has capacity to affect urban design and development which, clearly, has an impact on sustainability (Aditjandra et al., 2013). Transport behaviour is shaped by a combination of personal, social and environmental factors and people are more likely to change their behaviour when the benefits are immediate and personally relevant, such as saving money, improving the local environment or their child’s health (Richardson et al., 2007). It is therefore essential to understand how residents in the North East evaluate and prioritise the potential benefits and trade-offs associated with transport policy interventions, particularly those aimed at reducing car dependency through measures such as fare reductions, congestion mitigation and improvements in air quality.
Given that policies require a trade-off between different levels and forms of both investment and taxation, it is important to establish how and in what ways North East residents prioritise potential impacts and means of funding. One means of achieving this has been through conjoint analysis, which has long been used in market research, but it is increasingly used to elicit participants’ preferences with regard to competing policy elements (Bremer and Bürgisser, 2024; Hainmueller et al., 2014). A conjoint survey asks participants repeatedly to select their preferred policy design from two randomly produced policy programmes. The policy configurations compared include randomly allocated versions of attributes, such as price of travel and cost to the taxpayer. This allows researchers to estimate the average impact of any particular feature attribute-value on preference and strength of preference for the policy using comparable scales.
There have been precedents in conjoint analysis of transport policy. A Netherlands study examined preferences for cycling, with respondents preferring models that increase safety and separate bicycles from motorised traffic (Mertens et al., 2016). A study in Nepal examined demand and preferred design for a mass rapid transport (MRT) system using mode, waiting time, one-way fare per km, commute time per km and payment method as attributes (Aryal et al., 2022). They found that integration between different aspects of the system, cost, waiting time, commuting time and ease of payment was important in promoting shifts to public transport.
The broader stated preference literature in transport research has consistently found that fare affordability, service reliability and journey time are the dominant drivers of preference for public transport, with environmental outcomes and ownership structures typically lower in salience unless made directly and personally relevant to respondents (Hensher and Rose, 2007; Louviere et al., 2000). De Witte et al. (2013) identify a consistent tension in the transport preference literature between the outcomes that residents state they value most highly, principally affordability, reliability and access to opportunities and the infrastructure and governance choices that would most effectively deliver those outcomes, which are often more complex, less visible and less immediately legible to the public (De Witte et al., 2013). This suggests that conjoint methods, which make the connection between policy attributes and outcomes explicit and legible, may produce a more accurate and policy-relevant mapping of preferences than conventional survey instruments that ask residents to rank options in the abstract. Within the broader literature on stated preferences for public spending and redistribution, conjoint studies have demonstrated that support for progressive taxation to fund public services is more robust than conventional surveys suggest, particularly when the distributional consequences of both spending and funding choices are made concrete (Barnes et al., 2022; Bremer and Bürgisser, 2024; Rincon, 2023). These findings have important implications for the design of transport policy, suggesting that the public may be more receptive to ambitious, well-funded programmes of transport reform than policymakers typically assume, provided that the benefits are made explicit and credible, particularly for those on lower incomes.
For this study, we designed the conjoint experiment in line with similar categories of funding mechanisms, ownership and outcomes to examine the broader relevance of these preferences with regard to transport and specifically within the North East region. In using conjoint experimental and adversarially co-produced narrative analysis of public preferences for regional transport policy within a UK combined authority context for the first time, and thereby contributing both methodologically and substantively to an emerging field of research at the intersection of transport studies, public opinion research and devolved governance.
Methods
Participants and recruitment
We recruited 577 NECA resident adults through the online survey platform Prolific. They completed the study online. Prolific and related services provide convenience samples, in that participation is limited to those who have decided to sign up and respond to the study call, but their demographic diversity is fairly broad. Research using Prolific has been validated by comparison with other sampling methods for a number of known findings in psychological and political science (Coppock, 2019; Peer et al., 2022). The Prolific pool over-represents younger and more educated people compared to the UK adult population (Radkani et al., 2023). Relative to the UK population, our sample contained an over-representation of people who intend to vote for the Labour party at the 2024 general election. However, the primary analyses of interest in a conjoint study are not point estimates of levels of support but average marginal component effects – the marginal impact of individual attribute levels on the probability of policy choice, averaged across respondents. AMCEs derived from non-representative convenience samples are reliable estimators of population-level preferences, provided that treatment effect heterogeneity is moderate (Hainmueller et al., 2014). The subgroup analyses reported below show that, while some statistically significant interactions exist, the direction of effects is broadly consistent across demographic groups: no subgroup preferred more expensive fares, lower reliability or regressive funding. The fundamental preference ordering identified here is thus unlikely to be an artefact of sample composition. The sample size exceeds that required for stable AMCE estimation in conjoint designs of this complexity. Power analyses for conjoint experiments indicate that stable estimates can be obtained with substantially smaller samples when the number of choice tasks is sufficient and attribute levels are approximately balanced in their randomisation (De la Cuesta et al., 2022). With participants completing multiple paired choice tasks, the effective number of observations for estimation purposes considerably exceeds the number of participants, providing adequate statistical power for both main effects and subgroup comparisons. Prolific samples have been shown to respond to narrative and argumentative stimuli in ways that are broadly consistent with the responses of more representative samples, and are particularly suitable for experiments that require careful reading and engagement with extended text, given that the platform’s quality controls and attention checks are more rigorous than those available through many other online research panels (Peer et al., 2022).
Survey
Attributes and their levels for the conjoint experiment.
Participants were then presented with a description of a series of areas of transport reform along with the impacts such reforms evidence indicates will follow from their implementation (see Supplemental File Section 1). Participants were asked to rate their opposition or support to those policies on a scale of 0–100. They were then shown one of the five randomised adversarially co-produced narratives in Supplemental File Section 2 and asked to rate its persuasiveness on a scale of 0–100 and then to rate their opposition or support for the policy again on a scale of 0–100. The five narratives were co-produced with NECA resident opponents (≤30 levels of support for transport reform) of the policy via a separate survey, the results of which are published elsewhere (Ardron et al., 2025; Stark et al., 2025). Participants produced written narratives (minimum 150 words) that we standardised for language, style and length (150 words +10% max) and organised around five specific justificatory elements: absolute gains; relative gains; security; economic benefit; environmental benefit.
Participants were finally asked to provide basic demographic data and socioeconomic data, including income, self-rated risk of destitution on a 100-point scale, financial management, satisfaction with income, MacArthur ladder of subjective socioeconomic status (Adler et al., 2000), health status, including Depression PHQ-8 (Kroenke et al., 2001), Anxiety GAD7 (Spitzer et al., 2006), political affiliation, voting from 2019 General Election and intention for NECA Mayoral Election and faith in politicians established by six items in prior project iterations (Johnson et al., 2023). Participants received £3.00 in remuneration. Raw data are available at https://doi.org/10.17605/OSF.IO/3UX4M.
Data analysis of conjoint experiment
First, we used the patterns of participant choice to compute two quantities. The first was the utility for the average participant of each value of attribute. This is the marginal impact of the attribute being at that level on the choice of the policy. It can be interpreted as the value of that attribute level to the average participant. Thus, for example, if the utility of the single journey fare being £3 is negative, this means that a policy including a single journey fare of £3 makes that policy less likely to be chosen. The other quantity was the attribute importance. This is a way of comparing between attributes (rather than between the different levels of the same attribute). It can be interpreted as the relative importance of each attribute in determining the overall preference for the policy.
For the main analysis, we computed Average Marginal Component Effects (AMCEs) (Hainmueller et al., 2014) from linear probability models. The AMCE for a given level of an attribute can be interpreted as the marginal effect on the probability of choice of the attribute being at that level compared to the reference level, averaging across the possible levels of all other attributes. For subgroup analysis, we used linear mixed models to account for multiple responses from each participant. We formally tested by Type III ANOVA, whether including interactions between the subgroup identity and the attribute improved model fit. The Type III F-statistic was used to assess the significance of these interactions. We then followed up with pairwise comparisons of the estimated effects (AMCEs) for each attribute level between subgroups, using z-tests to identify statistically significant differences in how each group responded to the level of attribute. All analyses were carried out using the conjoint R package (Hainmueller et al., 2014). Continuous and categorical variables were treated as factors, with relevant categories defined for analysis. All p-values are two-sided. In addition to expecting strong support for a quicker, cheaper and more reliable transport network, our prior assumption was that younger individuals, women and those finding it difficult to manage financially would be more likely to favour a publicly owned transport system that is funded through progressive taxation and reduces transport poverty and avoidable deaths from air pollution. The rest of the analyses are considered exploratory.
To examine heterogeneity, we subset the data along five dimensions of respondents: men (n = 192) versus women (n = 314); white (n = 461) versus BAME (n = 45); those aged 45 and over (n = 152) versus those aged 18 to 44 (n = 354); a grouping based on self-reported financial position; a grouping based on income satisfaction and a grouping based on mayoral election voting preference. For financial position, we used responses to the question ‘how well would you say you are managing financially these days?’, grouping responses ‘living comfortably’ and ‘doing alright’ as ‘not difficult’ (n = 219), and ‘just about getting by’, ‘finding it difficult’ and ‘finding it very difficult’ as ‘difficult’ (n = 287). For income satisfaction, we categorised responses to the question ‘how dissatisfied or satisfied you are with the income of your household’ into two groups: ‘Satisfied’, (n = 267) and ‘Dissatisfied’ (n = 239). For mayoral election voting preference, we used the candidate individuals intended to vote for in the North East Mayoral Election on 2 May 2024. For statistical power reasons, we restricted this comparison to those who voted for Jamie Driscoll and Kim McGuinness (n = 153) versus those who don’t know or don’t intend to vote (n = 269).
Results
Participant demographic characteristics
The sample included 61% female, 38% male and 1% who described themselves in another way (Stark, 2025). 88% identified as white, slightly higher than in the 2021 England and Wales Census (81.7%), while the median age was 36 (mean 38.55, SD 12.67), lower than in the 2021 England and Wales Census (Office for National Statistics, 2022). The median annual non-equivalised household income was £40,000, higher than the national median income for the year ending 2023 of £32,500 (Department for Work and Pension, 2024). Participants reported a mean score of 28.41 and median score of 20 (SD 26.38) out of 100 for risk of destitution, with 0 representing extremely low risk and 100 extremely high risk. The average MacArthur ladder score was mean 5.35 and median 5.00 (SD 1.56), with 1 representing the worst off in society and 10 the best off.
The sample included 13% voted Conservative in the 2019 General Election, 56% Labour, 5% Lib Dem, 2% Green, 3% Brexit Party or other and 22% did not vote or refused to reveal preferences. For the 2024 NECA Mayoral Election, 2% intended to vote for the Conservative candidate, 19% for Labour, 10% for Jamie Driscoll, 2% for the Liberal Democrats, 2% for the Greens and 5% for Brexit or Reform, while 61% did not know, did not wish to reveal their preferences or did not intend to vote. Percentages may not total 100 due to rounding. Participants rated their left-right political affiliation on a 100-point scale as mean 37.43 median 40.00 (SD 19.48), indicating they identified as being in the centre-left of the political spectrum, which is reflective of voting patterns in the NECA region.
Levels of support
As Supplemental Table 1 indicates, the overall, initial level of support for transport reform was high (mean 79.35, SD 16.88). A large proportion of respondents, termed ‘lovers’, expressed strong support (≥70, 69.87%), including 16.32% who assigned the maximum of 100. A small proportion expressed strong opposition (≤30, 2.08%), termed ‘haters’. As the regression data provided in Supplemental Table 3 indicates, there were statistically significant differences in pre-treatment support by voting intention. Relative to those who reported being female, white British, not working and intending to vote Conservative, Labour (p < .05), Driscoll (p < .01) and Green (p < .01) voting intention was associated with higher levels of support for transport reform. Reporting higher MacArthur Ladder (p < .05) and being satisfied with income (p < .05) was associated with lower levels of pre-treatment support, while low MacArthur score (≤3) was associated with higher levels of support (p < .01). As Supplemental Table 2 outlines, participants overall scored the relative gains narrative highest as 77.96 mean and 82.00 median, with ‘lovers’ scoring the security narrative highest 79.79 mean and 83.50 median and ‘haters’ the economic benefit narrative 42.50 mean and 46.00 median – there were insufficient ‘hater’ ratings of environmental benefit to calculate a mean and median. As Supplemental Table 1 outlines, the effect of narrative treatment was a non-significant reduction in support overall (−0.29, p = .789), but a significant reduction among ‘lovers’ (−2.00, p = .019) and a non-significant increase among ‘haters’ (15.58, p = .091). Supplemental Table 4 shows that, in comparison to being shown the ‘absolute gains’ narrative, there was a statistically significant reduction in pre-treatment support among those shown the environmental benefit narrative −2.531 (p < .05) and among owner occupiers −2.986. The proportion of ‘lovers’ reduced slightly post-treatment to 73.78%, and the proportion of haters remained the same at 2.08%. The proportion of 0% and 100% scores remained the same. Compared to female respondents, there was a significant increase in support among male respondents, but this still constituted a net reduction in support (4.162).
Conjoint preferences
Feature importance
Figure 1 shows the attribute importance. The most important attributes were deaths from air pollution and funding (as this included possible tax rises). These were closely followed by fares, both within the North East and to London. The least important attribute was the proportion of journeys by bike. Average utilities for each level of the eleven attributes varied.
Average marginal component effects (AMCEs)
Figure 2 displays how each attribute level influences the probability of choosing a transport policy. Reliability and fare prices, both within the North East and from Newcastle to London, had the strongest influence, showing the largest effect sizes and highest statistical significance. Longer commuting times of 26 minutes more (−0.104, p < .001) decreased support, while shorter times of 26 mins less (0.040, p < .01) increased support compared to no change in commuting times. As expected, compared to a £100 average single fare from Newcastle to London, a £125 fare (−0.033, p < .01) reduced preference while a £25 fare (0.123, p < .001) increased preference. A £3.00 single fare (−0.170, p < .001) within the North East transport network significantly decreased support compared to a £1.00 baseline fare. Less reliable transport including 60% on time (−0.146, p < .001) negatively influenced preference compared to 100% reliability. Average Marginal Component effects (AMCEs) for each level of each attribute, whole dataset. Dots represent central estimates, and horizontal lines 95% confidence intervals.
Changes in avoidable deaths from air pollution notably influenced policy preferences, with a 100% reduction in avoidable deaths (0.054, p < .001) significantly increasing support, whereas a 100% increase (−0.098, p < .001) led to a substantial drop in support compared with unchanged death levels. A 50% decrease in transport poverty (0.025, p < .05) positively increased support, whereas a 50% increase in transport poverty (−0.055, p < .001) had significant negative influence on choice compared to unchanged levels of transport poverty. A 50% reduction in congestion (0.032, p < .05) significantly increased support, while a 50% increase in congestion (−0.046, p < .001) led to a decline in preference, compared to unchanged congestion levels.
Each of the funding options negatively influenced choice preference compared to an annual wealth tax on individuals with more than £2m, with council tax increases (−0.145, p < .001) having the biggest negative impact on choice. Compared to North East Combined Authority (NECA) ownership of the regional transport network, private corporation ownership was significantly dispreferred (−0.063, p < .001), while UK government ownership also showed a small, significant negative effect (−0.025, p < .05).
Only 10% of journeys made by bike (−0.00, p = .996) showed small, negative, non-significant effects on choice preference. The remaining levels, higher proportions such as 15% and 20%, exhibited positive effects; however, these too were non-significant, suggesting that changes in the proportion of journeys by bike had minimal influence on choice preference. In contrast, a 50% decrease in reliance on cars (0.045, p < .001) significantly increased favourability, while a 50% increase in reliance on cars (−0.119, p < .001) significantly decreased favourability, compared to unchanged levels of car reliance.
Heterogeneity
Type III ANOVA tests showed no significant main effects for age (F (1, 15090) = .83, p = .359), managing financially (F (1, 15090) = 1.53, p = .214), mayoral vote (F (1, 12570) = 1.08, p = .362), income satisfaction F (1, 15090) = .15, p = .693) and MacArthur ladder (F (1, 15090) = 1.11, p = .298). While overall preferences were consistent, some subgroups showed notable sensitivities to specific transport attributes. Significant interaction effects were found between managing financially and ownership of the regional transport network (F (3, 15090) = 4.32, p < .01) and average (peak and off-peak) single fare from Newcastle to London (F (4, 15090) = 4.33, p < .01). There were significant effects between MacArthur ladder and ownership of the regional transport network (F (3, 15090) = 5.78, p < .001) and average (peak and off-peak) single fare from Newcastle to London F (4, 15090) = 2.50, p < .05). There was also a significant effect between mayor voting intention and funding options (F (5, 12570) = 3.00, p < .05) and between income satisfaction and funding options F (5, 15090) = 3.10, p < .01).
Ethnicity (F (1, 15090) = 4.78, p < .05) and gender (F (1, 15090) = 5.35, p < .05) showed significant main effects. Ethnicity moderated preferences for car reliance (F (4, 15090) = 2.49, p < .05), while gender moderated preferences for avoidable deaths from air pollution (F (4, 15090) = 2.68, p < .05), indicating that both subgroups were not only independently linked to choice preferences but also influenced how specific policy attributes were evaluated.
We used z-tests to compare estimated AMCEs across age, gender, managing financially and income satisfaction, identifying statistically significant subgroup differences in responses to avoidable deaths, transport poverty, ownership and funding options.
Avoidable deaths from air pollution and transport poverty
Those finding it difficult (−0.066) and not difficult (−0.003) to manage financially both dispreferred a 50% increase in avoidable deaths from air pollution. Those finding it difficult expressed stronger opposition (p < .01). Both male and female subgroups significantly dispreferred this level of avoidable deaths (p < .05), with females (−0.059) expressing greater dispreference than males (−0.004). In response to a 100% increase in avoidable deaths, females (−0.122) showed a stronger opposition than males (−0.058), and the difference was statistically significant (p < .05). Dispreference for increased transport poverty was observed across financial situation and age groups. A 25% increase was more strongly opposed by those struggling financially (−0.051) than those not struggling (−0.004, p < .05). Similarly, a 50% increase was dispreferred by all age groups, with a statistically stronger opposition among those aged 45+ (−0.102) compared to 18 to 44 (−0.033, p < .01).
Ownership and funding
Those not finding it difficult to manage financially expressed a preference for transport ownership by not for profit, non-government body (0.034), whereas those finding it difficult did not (−0.036) and the difference was statistically significant (p < .01). Both income groups opposed a council tax increase to fund the transport system, with satisfied (−0.190) groups expressing statistically stronger disapproval compared to dissatisfied (−0.095, p < .01). Females (0.029) showed greater support than males (−0.029) for funding the transport policy through corporate carbon taxes with a statistically significant difference (p < .05).
Figure 3 presents the net average marginal component effects (AMCEs), the estimated impacts of individual attribute levels on support for two key policy attributes: ‘Average single fare from Newcastle to London’ and ‘Ownership of the regional transport network’, averaged across those finding it difficult and those not finding it difficult to manage financially and interpreted relative to the baseline levels £100 and North East Combined Authority (NECA). The reference level is set at 0, and all other levels are interpreted relative to it. Subtle but relevant differences are presented between groups. For those finding it difficult, the most favoured scenario featured North East Combined Authority ownership of the regional transport network and a £25 average single fare from Newcastle to London. In contrast, those not finding it difficult preferred a not-for-profit, non-government ownership paired with a £25 single fare. Notably, both groups expressed opposition to private ownership and higher fare options, with those not struggling showing stronger opposition. This suggests that even financially secure groups may resist transport policy associated with profit making. Visualisation of trade-off between average (peak and off-peak) single fare from Newcastle to London and ownership of the regional transport network, separated by those finding it difficult and those not finding it difficult to manage financially. The shading represents levels of preference, with darker red indicating stronger dispreference and darker green indicating stronger preference compared to the baseline (North East Combined Authority and £100 single fare). Lighter shades represent combinations closer to the baseline.
Figure 4 illustrates support levels for two key policy attributes, funding options and transport poverty, comparing responses from those aged 18–44 and those aged 45 and above. Those aged 45 and above are strongly opposed to income and council tax increases regardless of the impact on the levels of transport poverty, demonstrating only minimal levels of support for carbon taxes on corporations and an annual wealth tax alongside decreased transport poverty by 25%. In contrast, respondents aged 18–44 show a stronger preference for carbon taxes and an annual wealth tax when paired with a 50% reduction in transport poverty. These funding mechanisms are also favourably viewed when associated with a 25% reduction. Additionally, this age group appears receptive to private investment, provided it is accompanied by a 50% decrease in transport poverty. This suggests that younger people are more receptive to funding options if there are tangible social outcomes while older individuals appear more resistant to funding options regardless of the social benefit. Visualisation of trade-off between the different funding options and transport poverty, separated by those aged 18–44 and 45 and over. The shading represents levels of preference, with darker red indicating stronger dispreference and darker green indicating stronger preference compared to the baseline (unchanged poverty and an annual wealth tax on individuals with more than £2m). Lighter shades represent combinations closer to the baseline.
Discussion
The findings presented here need to be understood in the context of limitations in sample size and over-representation of 2019 General Election Labour voters and younger participants overall. The findings should be understood as indicative rather than definitively representative of the full NECA resident population and should be interpreted alongside findings reported elsewhere (Annous, 2025; Frost and Singer Hobbs, 2024). However, the preferences identified appear to reflect broader trends in terms of concern for financial insecurity and belief in pathways to improvement in service delivery. There is a high level of support for comprehensive, integrated transport reform overall. People prefer a public transport system that mitigates its contribution to avoidable deaths from air pollution and one that is affordable and reliable. Although not all subgroup main effects reached statistical significance, interaction analyses and pairwise comparisons confirmed that gender, age and managing financially did influence support for public ownership, reduced fares, progressive funding mechanisms and reductions in transport poverty, partially confirming our prior assumptions.
While AMCEs show that it among the least influential attributes in determining policy preferences, ownership of the transport system can have a facilitative role in achieving these outcomes. Among both those finding it difficult and those not finding it difficult to manage financially, there was strong opposition to private ownership indicating a broader consensus against profit-driven control of the transport network. In contrast, support for ownership by NECA remained consistently high across both groups, suggesting a preference for regionally accountable, publicly governed transport solutions. This may reflect a broader trend in terms of ownership to control and control to service quality and affordability seen in support for other programmes of nationalisation (E Johnson et al., 2025; YouGov, 2025). The ambitions set out in the NECA’s North East Local Transport Plan outline the Mayor’s blueprint for a transport network that is simple, affordable and built to last.
Despite a £1.85 billion funding package for North East transport, including a Metro line expansion to Washington, accessibility improvements, station upgrades and walking and cycling routes (NECA, 2025a), Mayor McGuinness has emphasised that fixing the region’s broken transport system requires further investment, particularly for key infrastructure upgrades like dualling the A1 and modernising the metro signalling network (Holland, 2025). Local leaders are in broad agreement that national government needs to view investments in terms of their impact on the economic potential of future transport development in the North East region. The reopening of the Ashington line, which welcomed over 300,000 passengers in its first 6 months following a 60-year closure, has been cited as ‘proof of the public’s desire for public transport that is reliable and value for money’ (Holland, 2025).
The findings presented here need to be interpreted in light of the institutional constraints on Mayoral action outlined above. Preference for public ownership of the transport network is striking in its strength and consistency, but the form of public ownership that residents are likely envisaging, namely, direct control over routes, fares and service standards rather than asset ownership per se, may map most closely onto the bus franchising model that the Mayor is pursuing through the Angel Network programme (NECA, 2025b). The strong preference for affordability across the North East is consistent with the Mayor’s stated commitments to fare capping but raises important questions about the limits of what can be achieved through locally funded subsidies alone, given the gap between public preferences and the revenue required to sustain a fully integrated network. While the Mayor has intervened to use NECA discretionary funds to cap bus and Metro fares, including in light of the closure of a flyover on the main arterial road into Newcastle from the south due to literally crumbling infrastructure (NECA, 2024c), it is not clear that those resources are sustainable.
The findings on funding preferences are instructive here: residents’ strong rejection of council tax and income tax increases as mechanisms for funding transport improvement, combined with relative support for taxes on corporations and wealth, point toward a political economy of transport finance that is oriented toward redistribution from capital to labour rather than toward cross-subsidisation within the resident population (EA Johnson et al., 2025). This has direct implications for the Mayor’s advocacy towards national government. The finding that residents prefer progressive national funding mechanisms over local tax instruments provides an evidence base for demanding that Westminster, rather than the Combined Authority, bears a greater share of the cost of transformative transport investment. There is clear public support for greater national funding for regional capital investment.
The means of funding a largescale programme of transport reform are tied to concern for impacts on poverty insofar as there are real reasons to believe that council tax and income tax increases are perceived as squeezing resources among those who are already under significant financial strain as a result of the cost-of-living crisis and longer-term effects of the Global Financial Crisis and austerity measures (Common Sense Policy Group, 2024). The measures that are regarded as plausible need to be understood in light of other preferences. Wealth taxes and carbon taxes are the two most popular means of funding improvements via reduction in transport poverty among those ages 18–44. Wealth tax is an increasingly popular reform insofar as there is increased public awareness of highly unequally distribution of wealth and the likelihood that wealth taxes are likely to affect those who are financially secure and able to pay (H Reed et al., 2025). Carbon tax on carbon-intensive corporations may be tied to strong preference to avoid deaths from pollution. Carbon production is increasingly viewed as being in contrast to the interests of people in light of the climate crisis (Ardron et al., 2025). Both taxes are national in nature, and increased concern about the need for increased Westminster funding for local and regional authorities indicates possible pathways to renewed devolution settlements. It may be that reformed, progressive proportional property taxes, such as those in Northern Ireland, can replace existing Council Tax to achieve similar impacts to wealth taxes but within the authority area (NIDirect, 2016).
It is particularly important to recognise the priority that residents place on the effects of pollution to dealing with various environmental responsibilities. People strongly reject transport policy that leads to increased deaths from air pollution and strongly favours reducing such impacts. While they regard active travel (e.g. proportion of journeys by bike) as being the lowest priority, it may be possible to link the need for active travel to reduction in pollution as means of promoting walking and cycling. This is important in the context of meeting environmental targets (Thew et al., 2024).
Assessment of the narratives adds weight to this analysis. Environmental benefit was the lowest-rated narrative and actually reduced levels of support post-treatment. Participants appeared not to link environmental benefit to the issue that they did prioritise: avoidable deaths. This reflects a broader issue in directing discussion of environmental policy directly toward non-human ends and interests. There is scope for policymakers to develop narratives on the environment that directly link environmental impacts to human interests, most keenly in this instance, in reducing avoidable deaths from pollution. Public preference for taxation on carbon-producing businesses indicates that there is some recognition of the impact of environmentally damaging activity.
Conclusion
This first conjoint and adversarial co-produced narrative analysis of Combined Mayoral Authority devolved transport policy suggests a series of public preferences that may appear, at first, anodyne: cheaper, quicker, more reliable journeys were always preferred to slower, more expensive or less reliable journeys. This was true of journeys to London as well as those within the North East. However, the findings suggest that residents view public control and ownership as the effective means of achieving this. That reflects a broader trend away from belief in the capacity of the private sector to deliver essential services in ways that reduce burdens on low-to-middle-income residents. There is a clear preference, overall, for transport policy that frees up resources for those on lower incomes by taxing those with the ability to pay. The preference to reduce congestion and a strong preference to reduce deaths caused by pollution provide policymakers with means of tying together parts of policy that are seen as lower priorities, such as active travel and preferences for wealth and carbon tax. These redistributive preferences reflect broader trends in support among a population increasingly exposed to financial insecurity (HR Reed et al., 2025; 2025; Nettle et al., 2024, 2025).
The original contributions of this study to the academic literature are substantial and multi-layered. Methodologically, it demonstrates for the first time that conjoint experimental analysis is both feasible and highly informative when applied to the specific policy choices facing UK Mayoral Combined Authorities, providing a template that can be replicated across the growing number of devolved regions in England. The combination of conjoint analysis with adversarially co-produced narrative treatment reveals not only the hierarchy of residents’ preferences among specific transport attributes but also the conditions under which those preferences are stable or susceptible to narrative challenge. The finding that support for comprehensive transport reform is highly robust against well-constructed opposition narratives, remaining largely unaffected by adversarial framing across the full range of narrative types tested, is itself a substantive contribution to the political economy of transport reform, suggesting that the electorate is less easily mobilised against ambitious public transport investment than political caution might assume. Substantively, the study makes a novel contribution to the literature on transport preferences and inequality by demonstrating that residents in a region of acute and historically entrenched transport disadvantage hold preferences that are clearly oriented toward redistribution, public accountability and the reduction of avoidable harm – preferences that cohere across demographic groups and are not confined to those with immediate material interest in transport reform. This extends and enriches the existing evidence base on UK public attitudes to transport.
These contributions open several avenues for further research. First, and most directly, longitudinal follow-up studies are needed to assess how public preferences for NECA transport policy evolve as the authority’s programme is implemented. Second, replication studies across other Mayoral Combined Authorities, including Greater Manchester, West Yorkshire, Liverpool City Region and the West Midlands, would enable systematic comparative analysis of whether the preferences identified here are specific to the North East’s particular history, economic conditions and political culture or represent a more general pattern of public preference for redistributive, publicly accountable transport systems across the devolved regions of England. Such comparative work would contribute significantly to the emerging literature on devolution and public legitimacy (Liddle et al., 2023; Whitham, 2025b). Together, these avenues of research would build on the foundations established in this article to produce a richer, more comprehensive and more democratically grounded evidence base for transport policy in the North East and beyond.
Supplemental Material
Supplemental Material - What do residents want from a North East CombinedAuthority transport policy? Evidence from conjointexperimental and adversarially co-produced narrativesurvey data
Supplemental Material for Reading cities: What do residents want from a North East CombinedAuthority transport policy? Evidence from conjointexperimental and adversarially co-produced narrativesurvey data by Anna Thew, David Littlefair, Graham Stark, Howard Reed, Matthew T. Johnson, Elliott A. Johnson in Local Economy.
Footnotes
Acknowledgements
The authors would like to thank Professor Daniel Nettle, Northumbria University, for his insightful, thoughtful and valuable feedback on the methods in this paper.
Ethical considerations
This study has been approved by the Faculty of Health and Life Sciences Ethics Committee, Northumbria University (5814). This committee contains members who are internal to the Faculty. This study was reviewed by members of the committee, who must provide impartial advice and avoid significant conflicts of interests.
Funding
This study was funded by Insights North East. Insights North East had no input into the study design or analysis.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
All data are cited in text and are publicly available.
Disclosure statement
The authors are all members of the Common Sense Policy Group. Howard Reed is the Director of Landman Economics. Graham Stark is the Director of Virtual Worlds.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
