Assessing the feasibility of using place-based health information in alcohol licensing: case studies from seven local authorities in England

ABSTRACT As in most other countries, England has no explicit alcohol licensing objective around health, so objections to applications tend to focus on the traditional concerns of crime and public disorder. We examined the practicalities of using health-related information in local licensing decisions and the prospects for a dedicated health-associated licensing objective. Seven local authority pilot areas were purposively selected and provided with a compendium of health information (Public Health England Toolkit), including data-access agreements and mapping software. A series of ‘mock licensing hearings’ explored practical challenges in using health data. Key informants were interviewed at baseline and 10–12 weeks after receiving the Toolkit. Access to localised health information was problematic, and there was a mismatch between a ‘data-orientated approach’ and the need for contextualised evidence. Perceived difficulty in proving that a new licence would damage health discouraged challenges on health grounds. Constraints in using health information in alcohol licensing are not restricted to the absence of a dedicated health-associated licensing objective. While the latter may enhance the legitimacy of public health participation, improved access to localised health information, stronger collaborative working and training in how to contextualise evidence, will all be critical to better alcohol harm reduction through licensing decisions.


Introduction
The current rationale for statutory restrictions around the sale of alcohol throughout the UK is, as for most other countries, focused on public safety, child protection and the prevention of crime and disorder (Light 2010). Structural changes to public health in England from 2012 onwards have begun to encourage the consideration of a broader concept of harm as part of the process of reducing the burdens of premature mortality and ill health attributable to alcohol (Nicholls 2015), as well as its broader societal costs. In brief, local public health departments which since the mid-1970s have been located within the NHS, have been relocated to local government. Their new co-location alongside licensing, social work and environmental health, has also coincided with Directors of Public Health becoming one of the statutory responsible authorities that need to be consulted on the process of granting new licenses to sell alcohol (LGA, Alcohol Research UK 2012). While such a change in principle should facilitate increased cognisance of the chronic health burdens of alcohol in addition to the traditional focus on acute harms, this is not reflected at present in England in the four existing licensing objectives of the Licensing Act (Home Office 2012): (i) promoting the prevention of crime and disorder; (ii) promoting public safety; (iii) promoting the prevention of public nuisance and (iv) promoting the protection of children from harm. A fifth licensing objective aimed at 'protecting and improving health' was implemented in Scotland in 2009 (Gillan et al. 2014), although in Scotland public health departments themselves remain within regional health boards and have not been transferred to local authorities. A fifth such health objective has not been considered previously for England, though it does command widespread support among Council Directors of Public Health (Morris 2016, Somerville et al. 2020. While a focus on the acute health effects of alcohol can have a favourable impact on crime and disorder (de Vocht et al. 2017b) as well as on adverse health outcomes (de Vocht et al. 2017a) there are self-evident public health risks associated with overlooking the longer-term chronic health effects of alcohol, which account for substantial burdens of ill health and premature mortality, particularly in less socially advantaged areas of the UK (Katikireddi et al. 2017).
The desire to more effectively address chronic health effects of alcohol and to better realise the potential advantages of local public health expertise in reducing population-level alcohol harms, has prompted Public Health England and the UK Home Office to explore mechanisms to support the better deployment of health-related information as a more integral part of the licensing process. While there is a broad consensus that this would be a worthy objective (particularly among the public health community (Nicholls 2015)), the best means by which to achieve this remains uncertain. Should, for example, the introduction of a new fifth objective around health be considered or would it be better to integrate health information as part of the considerations under the existing four objectives? The introduction of cumulative impact zones (Woodhouse 2017), (which can designate that the existing total adverse impacts of alcohol consumption in a defined area or locality are such that new licensing applications would be rejected unless that can demonstrate that they do not further undermine the licensing objectives) may also offer opportunities to incorporate local information on health harms as part of the consideration of the overall adverse health and social impacts of alcohol on local communities (Home Office 2012, Grace et al. 2016 To gain better insight into the practicalities of answering these questions, the UK Home Office in collaboration with Public Health England, conducted some detailed exploratory surveys with English local authorities at the more severe end of the alcohol harm impacts burden (Home Office 2016) (in terms of alcohol attributable deaths and acute and chronic health impacts). Insights from this work were subsequently used to inform local alcohol harm reduction strategies (Middlesbrough Council 2016).
One significant barrier to the better utilization of health information by all 'responsible authorities' for scrutinizing licence applications was highlighted as the large degree of variability in reliable and timely access to such information, whether national or local (Humphreys and Smith 2013). Alcohol-related hospital admissions and hospital diagnosed illnesses related to alcohol for instance, can be problematic to obtain for a variety of reasons, from local recording practices to variations in data-sharing arrangements between local stakeholders (Mooney et al. 2017. In response, Public Health England sought to develop a comprehensive 'licensing-relevant data repository tool' to facilitate access to local and national data sources (Public Health England 2017b). Envisaged as a resource which local areas would be able to supplement with their own local routine data sources and/or specialised surveys, the data compendium or 'analytical support package' (ASP), was piloted in the current investigation with local authorities (Public Health England 2017b). The ASP (see Appendix A) was supported by templates for data sharing agreements, wholesale access to centrally available hospital admissions data and software and baseline background data (at middle level superoutput area (MSOA) (Office for National Statistics 2016)) to be able to undertake local geographic mapping of locations of interest. Our over-riding objective was to examine the extent to which this would assist local authorities in the appraisal of local alcohol licensing applications and in formulating plans and strategies tailored to tackling local alcohol harms.
In the current evaluation, we explore three aspects of how to improve the use of health-related information based on a three-component strategy with seven purposefully sampled LA councils. The LA selection criteria included a demonstrated commitment to examining practical mechanisms for incorporating health-related data while representing a broad geographical and socio-demographic spread reflective of England as a whole (Public Health England 2017b). The three aims of our investigation were to examine: (i) The utility of the analytical support package in facilitating access to information that would be useful to incorporate health-related concerns into evidence that may challenge the acceptability and format of new licence applications or justify the inclusion of new conditions of acceptance (e.g. around early closing, etc.) (ii) The perceived potential benefits of a new and dedicated Health-Associated Licensing Objective (HALO), and the extent to which that might enhance the legitimacy of public health as a responsible authority (iii) The wider role of public health in alcohol licensing and the development of local government policy responses to the challenges of hazardous and harmful drinking

Methods
Eight local authorities (LAs) were initially shortlisted by a panel made up of Public Health England (PHE), the Home Office, Department of Health, the Local Government Association (LGA), and 'Alcohol Change UK' based on geographical spread, urban and rural mix, and capacity to engage in the project (Public Health England 2017b). One LA subsequently had to withdraw for internal reasons leaving seven in the pilot. More by coincidence than design, all of the participating LAs were single tier authorities, which in England accounts for around a third of all LAs and means that they control all local government functions as opposed to being arranged across two tiers. Information gathering from participating LA licensing teams began with five baseline pre-ASP semistructured interviews, to confirm our understanding of the processes for dealing with licence applications prior to the release of the ASP, which was being introduced over a two-week timescale. There was one such interview in each of the five participating regions and at least two informants in each session, one of whom was always a public health team member. Other forms of early engagement with the two remaining LAs were sufficient to reassure the research team that practices did not significantly diverge. Other methods of data capture were three mock 'licensing hearing' type scenarios which dealt with hypothetical licence applications in three very different parts of the country. In addition, two shared learning events were held involving representatives of all participating local authority teams with directed discussions related to the three components outlined above (after one month and then at project conclusion) and from which roundtable discussion summaries were included in the analysis. Four end-of-project focus groups with a total of 31 responsible authority members across the seven sites were also included in the protocol. An overview is shown in Table 1. The three 'mock licensing' hearings conducted by participating LAs were tailored (as would be the case in practice) to local circumstances and according to time and staff resources. Each practice scenario therefore had a slightly different setup, as shown in Table 2.
Given the practical purpose and aims of the mock scenarios (namely to explore the possibilities for tabling health-related information within hypothetical licensing application hearings), these hearings were organised to explore the practicalities of public health contributing meaningfully to alcohol licensing decisions and any broader implications there may be for the development of local alcohol policies.
Interview transcripts and focus group discussions were recorded and transcribed verbatim and summary observer notes from a research team member/observer were analysed from the learning events and the mock hearings. Observer notes were shared between two researchers (ZS and JM) and any observations that were not convergent were discussed with a third member of the team (JL) before thematic analysis of all texts using NVivo version 11 (Stern and Porr 2011). Ethical permission was obtained from the ethics committee of the University of Sunderland.

Findings
A total of 28 preliminary topics were grouped into three overarching themes, based on the original objectives of the evaluation and their allocation was agreed between two authors (ZS and JM). The three overarching themes, corresponding to the original objectives were: (1) Utility and acceptability of ASP as an information resource in the licensing and alcohol policy evaluation setting (2) The potential added value as well as any likely drawbacks of introducing a new licensing objective based around health (3) Current status of the public health contribution to local authority alcohol licensing and recommendations for how the role might be further developed.
Quotes from participants are labelled according to their job role and from which of the three settings the inserted quote arose (INT = interview; MLH = Mock Licence Hearing and FOC = focus group).

Utility and acceptability of the ASP
In principle, the ASP was seen a welcome and convenient resource especially as it could serve as a single repository for all licensing relevant information:  The basis of the challenge was a broadly defined health and well-being objective Area 3 (Southwest) One nightclub and one off-licence Two Previously approved applications re-evaluated ± a hypothetical health objective I think we're grateful that we've got this, because what we don't want is people having to go to five different sources to get what they need . . .

LA Public Health Specialist: INT
There was clearly a large variability in the types of data available (due to access issues and technical capacity restrictions), as well as resolution (e.g. in relation to the granularity of the data): [We] then overlaid a number of different health indicators at MSOA (middle level super output areas 1 ) level, because that was the most robust level that we can go to . . . But there might only be four MSOAs in that area. So when you're really trying to drill down to, oh, is this premises having an impact on street a, street b, street c. MSOA is never going to be robust enough.

LA Information Analyst: INT
In the mock hearings setting, participants felt there were dangers in a heavily data-reliant approach for two reasons: the first being the difficulty of attribution which industry legal advisors have been keen to highlight in relation to individual premises applications: Overall, however, there was broad support for the need to improve data accessibility and all responsible authorities in alcohol licensing understood the need for public health to have access to appropriate information with which to make their arguments: One of the things that I've heard nationally is that public health colleagues have kind of felt that they've joined the party late by being, you know, one of the last people to be a responsible authority. So, it's a lot of catching up to do. And if as a result of having an available data tool you feel as if you can present your information in a way which is up to speed with everybody else, then that itself is a valid outcome.

LA licensing officer: INT
As the usefulness of local health data lies in its local specificity, problems in obtaining health data of a sufficient granularity to be of assistance with individual licence applications remains a source of frustration for licensing teams: Pharmacy-based prescribing of emergency contraception is now being more consistently collated, and perhaps not surprisingly peaks at the weekends . . .

LA (out of hours) pharmacy commissioning lead: FOC
While the ASP data compendium was characterised therefore by a wide regional variability in terms of implementation, technical capacity issues and a shorttime period for appraising its effectiveness, for the most part it integrated well with existing data collation systems and prompted their enhancement with locally available data sources, as had been the intention. As health-related information is already being used to support the existing four licensing objectives in England, public heath teams could potentially glean much from other responsible authorities on what might be the best ways to present information to licensing hearings.

The potential benefits (and drawbacks) of a dedicated 5 th Objective (HALO)
Experiences such as that noted above, in which a responsible authority RA encountered difficulty raising an objection under a particular licensing objective without evident support from the lead agency involved, would support the contention that a dedicated 'health objective' is likely to substantially enhance the credibility of public health-led representations: The problem that we've got at the moment with public health is that the licensing regime has been inaccessible, nationally, to public health teams because of the way the thing is being presented . . . [i.e. with a predominant focus on acute harms and crime and disorder and specifically excluding health -investigators' interpretation].

LA licensing officer: INT
Indeed, there seemed to be a broad consensus among all responsible authorities that the introduction of health as a fifth licensing objective would enhance the attention and credence given to health-related information in the context of licensing applications and local licensing policy. It was difficult to separate this, however, from the difficulties noted in the previous section on data resolution aspects of the ASP, since it is not clear whether higher resolution data would have addressed the shortcomings. A separate participating LA was also exploring the potential of adding health to enhance the evidence case for existing cumulative impact areas: From the perspective of existing cumulative impact areas, the introduction of a health objective would enable them to be strengthened to take account of small area clusters of alcohol-related ill health. Since CIP areas need to be reviewed regularly to maintain their relevance, adding health impact would be relatively straightforward in principle, thereby getting around the issue of single premises attribution . . .

LA legal advisor for licensing: FOC
Despite significant practical considerations therefore, the very act of making such a data compendium tool available and inviting LAs to consider how they might best use it, stimulated a number of innovative responses which contributed to the broader case for the inclusion of health considerations.

Public health roles in alcohol licensing and harm reduction
The Public Health teams in this current pilot were all keen to explore ways of engaging more effectively with the licensing process. Many of the issues which arose would be in keeping with any large structural reorganisation, as has occurred in relation to public health in local authorities. There is no doubt, however, that when it comes to issues such as licensing, where other council departments have a long history and have established ways of working which suit the stakeholders they work with, there are clear differences in general approach, particularly with regard to handling data and evidence, which is acknowledged to be a significant part of the public health skill set: So, the toolkit itself has very much been left to public health to see what they can pull out of it, rather than anybody else.

LA Public Health Specialist: FOC
From the lessons learned during the mock scenarios and areas of consensus from the end of project focus groups, all three sub-topics developed under this theme could be considered together as pre-requisites for public health being able to make a successful contribution to local alcohol licensing policy, these being: The criticality of defining what is meant by a public health objective is particularly pertinent to those charged with its enforcement or implementation, so it was clearly inappropriate to assume that all parties agreed about the definition. A suggestion made by an LA licensing officer, was that a common wording which fulfils the Scottish fifth objective might be adopted: Applicants will be expected to demonstrate in their operating plan that suitable and sufficient measures have been identified and will be implemented and maintained to protect patrons' health. This will include such measures as making available information with regard to sensible drinking, the effects of excessive alcohol consumption and contact points for assistance to address problem drinking.

Format and strategy for building successful representations
Related to the need for clarification of what is meant by the health objective, several licensing teams echoed concerns around how the focus for using health data needed to move away from being restricted to single premises considerations: Unless any new objective comes with an increased weighting -that [lets] you take account of an area, rather than a specific premises, it's going to be meaningless

Environmental health officer: INT
Licensing representations with more than one supporter (and therefore drawing upon more than one of the four objectives) were perceived to have a greater chance of success, so it was perhaps unsurprising that the mock-case built on health impact data alone was not felt to be sufficient: The way that we interpreted what we needed to do in this mock hearing . . . was that we could go solely for public health data to see what impact that could have. And it wasn't enough, basically . . .

LA public health specialist: MLH
As part of the same mock hearing, but covering a second example, one of the public health teams sought to deploy crime and disorder information without the backing from the lead agency for this objective: . . . the committee would say, well, if the police aren't making a representation, why are you talking about crime and disorder, saying it's a problem?

Police licensing specialist (follow up from MLH)
Non-participation of one responsible authority can therefore detract from the credibility of an objection on specific grounds highlighting the importance of 'joined-up representations' where possible. In a separate mock scenario, the committee were persuaded of the adverse health impacts from likely noise levels (for local residents) and danger from traffic in the early hours of the morning in a restricted road space. Ostensibly this drew on the public disorder objective, though acute health impacts (from chronic stress/sleep disruption) were cited. This particular scenario was the same setting where the addition of health to an existing CIZ was proposed for the next renewal of evidence for CIZ justification, as outlined above. Perhaps the most viable way forward in terms of dealing with the likely defences to the deployment of health data or the implementation of a fifth objective on health was embodied in the third mock scenario, where a broad interpretation of 'health and well-being' was adopted as a target objective. This had the dual benefit of helping direct the focus away from single premises, as well as broadening the definition of health impacts at the local area level: OK, we get it that it's not about these individual premises, but we're going to actually take an area . . . . and look at it as a hot spot area [for health effects] .then actually look closely at everything that comes through . . .

LA public health alcohol lead: FOC
Ultimately, of course, this allows for the factoring of pre-existing higher risks or vulnerabilities, thereby starting to embed a true public health approach: This area has got high deprivation. We know that that will mean that people in this area have an increased level of harm. Therefore additional [opening] hours, effectively extending provision in this area, we believe, is likely to increase the potential harm for these individuals.

Training and skills needs
In terms of the components of the support package and the capacity to use linked datasets and build maps, there was a degree of expertise in each participating LA and a few were already fairly advanced in their thinking and practice (even if it was seen as a specialised task, see 3 rd quote below): I think from the feedback we've had, most of that has been fairly readily accessible and easy enough to interpret. And we've then been able to develop the hotspot maps [showing areas of high alcohol-related harm] and whatever else . . .

LA public health alcohol lead: FOC
We were quite fortunate because we had something to build on. Because we already had our mapping. I think we were already some way into realising the benefits of looking at local data sets for a specific geography. And we'd already used that kind of data in relation to some of the previous licensing applications.

LA information analyst: FOC
So I counted the benefits of making data easily accessible, but then you need the training for the public health practitioners who would be doing that, because they weren't supported by analysts.

LA information analyst: MLH
Information aspects aside, while non-trivial barriers to solve such as data access issues were reported, these were seen to be less important within the setting of the mock hearings than basic lack of familiarity with the format and processes of the committee hearings: Indeed, the two previous statements can be recognised as summing up the complexity of the overall approach required and convey a degree of confidence that there is now a real appreciation, at least in these participating LAs, of the 'paradigm shift' required in how evidence is used and presented.

Discussion
The long-established dominance of crime and public disorder as the main rationale for statutory restrictions on alcohol availability (Jeffs andSaunders 1983, Nicholls 2015) informs the basis of much of the legislative environment around alcohol in many parts of the world (Alcohol and Public Policy Group 2010, Babor et al. 2010). As part of structural changes to the English public health landscape in 2012/13, local public health teams moved from the National Health Service (NHS) back to their original pre-1974 location in local government, while at the same time acquiring 'statutory consultee' status with respect to new alcohol licence applications (LGA, Alcohol Research UK 2012, Nicholls 2015. These developments raised the prospect of public health teams being able to make more pro-active use of health information within the decision-making processes around alcohol availability and licensing (Gillan et al. 2014). Our primary focus in the current study was to appraise the usefulness and adaptability of a localised data repository for information designed to help local authorities make decisions on whether to grant new applications for alcohol licenses for on-trade or off-trade. In addition to improving data availability, we also aimed to establish if there were other resources, knowledge gaps or skills that might affect the ease with which public health teams can engage with the licensing application process and the broader context of local alcohol harm reduction policies in general. The difficulties faced by local authority public health teams in gaining access to health and social care data, which helped inform the current study, are not confined to issues around alcohol. In the words of Professor John Newton, then Chief Knowledge Officer for Public Health England, speaking to the House of Commons Health Committee in late 2016: (House of Commons 2016) 'The situation we have is that the information -the data -is in one place, and the people who have the capacity and capability to use it are in another.' In the case of alcohol licensing, the new 'responsible authority' status of public health brought the issue into focus, as public health was not in possession of the information they needed (Humphreys and Smith 2013). Additionally, even where the information access issue could be resolved, public health teams were without the executive (legislative) authority to make use of it, as health itself is not considered as a legitimate licensing objective within the English system (LGA, Alcohol Research UK 2012). Even where a fifth objective around health has been introduced in Scotland, difficulties around evidence gathering and the localised attribution of harms have meant that its implementation has been less than straightforward (Fitzgerald et al. 2017). With problems of data access, compounded by issues of legitimacy in a licence decision-making setting, published confirmations that proactive alcohol policies by local councils are not only associated with a reduced incidence of alcohol related crime (de Vocht et al. 2017b), but also an increased likelihood of real reductions in alcohol related hospital admissions (de Vocht et al. 2016), have presented public health teams with the impetus and encouragement to negotiate these difficulties more effectively. Since these associations were still being promoted against a backdrop in which public health expertise can be effectively disenfranchised from the decision-making process for the reasons just highlighted , Somerville et al. 2020, the potential for further gain should these difficulties be addressed must yet be considerable. Data access issues were not universal as they are solvable, often through informal close working relationships, individuals in shared posts or through intelligence sharing networks dedicated to joint undertakings (Phillips andGreen 2015, Public Health England 2017b). In other evidence developments, a recently published selection of natural experiments has also demonstrated the feasibility of evaluating changes to the local alcohol environment down to the level of single premises (de Vocht et al. 2020).
The barriers to accessing health data noted above were not the only challenge in bringing robust health-related intelligence into the licensing arena. Differences were also clearly recognisable in the information/data preferences of different stakeholders along the lines that have already been documented, both in ethnographic studies with local government in London and across different policy translation settings (Frost et al. 2012, Phillips and Green 2015, Rushmer et al. 2015. In brief, the 'evidence-based medicine' paradigm for presenting evidence, often accompanied by restrictive caveats concerning its interpretation and application, is not an approach that lends itself well to what can be very polarised and principle-driven debates in the licensing chamber or in other local government settings (Phillips andGreen 2015, Atkins et al. 2017). Additionally, in keeping with the findings and perceptions from a recent survey of London Local Authorities , the approaches which worked best in the mock scenarios were those brought forward by more than one responsible authority. Such joint approaches helped facilitate the incorporation of complementary strands of evidence tailored to each small local area in a process with broad similarities to place-based approaches in public health (Randle and Anderson 2017)] and reminiscent of whole system policy frameworks that are adopted for other 'wicked' public health problems amenable to local and national planning such as obesity (Mooney et al. 2015). Ultimately, such a place-based format was the ideal platform to incorporate the use of clear narratives and case studies as a method of bringing data to life in a relevant and understandable way for the locality in question (Phillips andGreen 2015, Public Health England 2017b). A similar reasoning seems very much to inform a home office impact assessment on the introduction of health into the licensing application process, which concluded with a recommendation that the most viable option was to incorporate health evidence as an integral component of the case for a cumulative impact policy (Home Office 2012).
In this context, the very broad definition of a fifth objective around 'Health and Wellbeing' adopted by two participating LAs, which calls for an appropriate 'weighting' of places characterised by a complex mix of disadvantage, is in keeping with what Cummings and colleagues call for 'recognising that there is a mutually reinforcing and reciprocal relationship between people and place' (Cummins et al. 2007). Evidence from Scotland of higher rates of alcohol-related deaths and hospitalisations in less advantaged areas (Richardson et al. 2015), together with the recently published confirmation that more socially disadvantaged population groups experience greater levels of harm, even after adjusting for levels of alcohol consumption (Katikireddi et al. 2017), may add support to the case for focusing alcohol availability restrictions in areas where it causes the greatest harm, as a means of reducing social inequalities in health outcomes. Recent evidence from London which found an apparently limited impact of cumulative impact policies in reducing alcohol sales (Pliakas et al. 2018) might also suggest that CIPs themselves need to be more carefully targeted or expanded, possibly by incorporating socioeconomic profiling with retail sales information.
We found a high perceived need for further training across all aspects of the alcohol licensing application process by local authority public health teams. PHE had effectively already partly anticipated this by publishing comprehensive guidance for public health teams for their participation in licensing (Public Health England 2017a) which aims to set out how best to accommodate health within the existing four licensing objectives, several of which are acknowledged to be underused (e.g. protection of children from harm, when there are inadequate safeguards to prevent underage sales). Experience in the mock hearings however confirms previously observed concerns that there is a mismatch in preferred types of evidence (Humphreys andSmith 2013, Phillips andGreen 2015) and that council public health teams are often more inclined to locally tailor their practice than to adhere to national guidelines (Martineau et al. 2014). Also, the suggestions from some respondents in this evaluation, that training be extended to other members of licensing committees and to councillors, would resonate very much with a recent House of Lords Select Committee report highly critical of the conduct and processes in licensing committees (House of Lords 2017). In the words of the committee chairman, Baroness McIntosh of Pickering: The Committee was shocked by some of the evidence it received on hearings before licensing committees. Their decisions have been described as 'something of a lottery,' 'lacking formality,' and 'indifferent,' with some 'scandalous misuses of the powers of elected local councillors.' Their report recommends the abolition of dedicated licensing committees and for licensing applications to go to planning committees, given their long experience in being responsive to the views of local populations and balancing different commercial and other interests.
As this type of reform was largely outside the scope of the current research and not pro-actively pursued, it would not be helpful to dwell upon it here, but it does at least concur with the observed wide variations in practice encountered even within this relatively small sample of English Las and further highlights, the extent to which decision making around alcohol licensing remains a 'contested space' (Richardson et al. 2015) for public health and public health evidence.

Strengths and limitations
The current project set an ambitious agenda in trying to understand the current practices and rationale for alcohol licensing decisions, alongside appraising the value of a data-compendium toolkit (ASP) to help with the process of increasing the integration of health data where possible. Since public health teams had been back in local government for less than 5 years in 2017 and the exercise represented a very short time frame to evaluate a complex resource, it was difficult to draw firm conclusions about the prospects for increased use of health data in licensing, whether mediated through a new dedicated licensing objective or incorporating it into the existing four. It is also important to point out that this was a purposefully selected sample of single tier local authorities, who had expressed interest in making better use of health information in licensing. While in principle, a single tier authority is more straightforward since functions are not spit between an upper and a lower tier, in practice alcohol licenses would sit with planning in all LAs, so this distinction is not likely to influence the transferability of our findings. The fact that the participating LA's in this evaluation were selected from those expressing an interest in improving licensing decision making, means that there is a strong component of self-selection and previous work in this area has shown a wide variability in the strategies and underlying drivers which determine an individual LAs overall approach to alcohol harm reduction (Mooney et al. 2017).
In brief, this project was able to demonstrate the potential value of a data repository to help with licensing and perhaps more importantly it helped to highlight what additional measures would be helpful in order to better integrate health information and concerns, most notably some dedicated training for licence team members (not necessarily restricted to public health), and the adoption of a broad concept of health harms at a small area level that circumvents the necessity to attribute concerns to a single premises.

Conclusions
While changes to public health structures and responsibilities in the UK, and England in particular, have the potential to facilitate a more health-orientated approach towards alcohol licensing policies and applications, the changes themselves also throw up some major challenges in terms of the practicalities of achieving this. In light of some of these difficulties, there are signs that the appetite for taking a more comprehensive view of alcohol harms (and by extension on solutions) has been diminishing since 2016, when the UK Government's crime strategy included alcohol as one of six principle drivers of crime (Home Office 2016) and the more recent UK Government's 'Beating crime plan' including proposals for the alcohol tagging of offenders (Home Office 2021). As many of the policy levers and organisational structural components are now in place for a fuller integration of health information and health intelligence with targeted local alcohol policies, this would be a significant lost opportunity. The pioneering work undertaken by some of the most innovative LAs in England needs to be learnt from and built upon including the multistranded place-based approaches to licensing in this report as well as broad-based interpretations of the licensing objectives which emphasise the stated ethos of 'promoting' the aims inherent in those objectives (Cammiss andManchester 2012, Somerville et al. 2020). Multiagency representations will always be more persuasive than single agency, and concerns about needing absolute proof linking a premises to potential problems can be overcome if approached thoughtfully regarding the data used and how they are presented within licensing committees. Ultimately, therefore, in order to better incorporate the use of health information as part of the licensing decision-making process, efforts should ideally be directed at contextualising the evidence presented and framing any adverse health impacts in a small, localised area or defined small-area population level, as opposed to trying to attribute health harms to an individual premise or outlet. If health indicators can be an integral part of the common language to bring this about, then public health and the health of the less affluent sectors of society that are most blighted by the damaging impact of alcohol, will be the main benefactors.