Hypothesis 1: increased resources

The first hypothesis, as articulated by those designing, implementing and delivering the intervention, was that the LNL would increase the resources available to police and manage the NTE, which would be used for street-based community safety and policing and additional street cleaning services. These services, would, in turn, lead to a number of positive impacts for residents, visitors and commercial actors (figure 1):

This will produce additional funding for the council and police to use to address the impacts and strains on local services that occur between midnight and 6am in [LA]. […] we believe that the LNL can be used to reduce the instance of crime, disorder and anti-social behaviour during the levy hours as well as improve the local environment (LNL Consultation, 2013).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Hypothesis 1. Orange bubble=national variable; yellow bubbles=local variables; green bubbles=immediate theorised impacts stemming from levy introduction. Solid line: positive relationship between variables; dashed line: inverse relationship between variables; dotted green lines: theorised impacts stemming from the levy introduction. ASB, anti-social behaviour; NTE, night-time economy.

Some residents and visitors further described the mechanism by which such change would occur, placing an emphasis on the additional police and community safety presence; for example, one woman we interviewed in a pub believed more police on the street ‘means I can walk home safely at 2:00’ Others described police as a deterrent for anti-social behaviour and noise:

I live on a side street of a late licensed premise and am woken up between 2 am and 7 am regularly every Saturday and Sunday morning. I don’t think they realise the noise they’re making so if there was a police presence I don’t think they’d be as boisterous (consultation response, resident).

It was theorised that a safer NTE would enhance the overall reputation of the area, driving up visitor numbers and encouraging individuals to spend more money in local establishments. Implicit in this argument was that the levy fee would be easily offset by increased customer numbers, thereby making the levy beneficial for both commercial actors and the LA:

the money’s going to pay for more policing, and [licensees] get more policing and that is beneficial for them because the safer an area is, the more people that come to the area and the more money that gets spent and the more money they make (interview, community safety officer).

Hypothesis 2: reduced support for PPP schemes

A second hypothesis, articulated by businesses, was that licensees would be disinclined to continue to support PPP schemes because the resources do so would be redirected to paying the levy:

If operators do choose to pay the levy then it will impact on funding they can provide for partnership initiatives such as BIDs, Pubwatch and Best Bar None which the Council should look to support and promote in preference to a levy (consultation response, Pub company).

Many licensees, in particular, expressed concern about the LNL’s impact on the BID, which funds dedicated police officers and cleaning services. In consultation responses, licensees agued they would vote against the BID when it came up for renewal, which in turn, would cause the BID to fail: (figure 2): ‘As a BID payer if the levy were to come into force I would be voting no the next time the bid comes up for tender’ (consultation response, licensee).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Hypothesis 2. Orange bubble=national variable; yellow bubbles=local variables; green bubbles=immediate theorised impacts stemming from levy introduction. Solid line: positive relationship between variables; dashed line: inverse relationship between variables; dotted green lines: theorised impacts stemming from the levy introduction. LA, local authority; NTE, night-time economy; PPP, public–private partnership.

If the BID were to fail as a result of the LNL, licensees theorised that there would be a range of unintended consequences in the local system. These included reducing the overall resources available to manage and police the local area and damaging economic impacts because the BID is intended to create an environment that encourages residents and visitors to the area:

It is a possibility that nearly 40 licensed premises in the [local area] BID area will not vote for the BID again if this means that they pay two levies instead of only one. A BID needs a majority by numbers and also rateable value to succeed. A failure to achieve either one of these would, therefore, jeopardise the provision the BID makes for policing and cleaning as well as what we do to ensure a good shopping environment for local people, Christmas lights, hanging baskets, support for community events and much more (consultation response, Pub manager).

The local BID Board also perceived that any negative impact of the levy on the BID could generate impacts beyond the boundary of the local system. These system stakeholders argued that this hypothetical impact should be considered within the Mayor of London’s goal to increase the number of BIDs throughout London:

BIDs are burgeoning in London and the Mayor has set a target for a number of additional BIDs by 2015. It would be a loss, not just to [LA] but to London as a whole should [BID name] not get re-elected and become the first BID in London to fail (consultation response, BID Board).

Hypotheses 3 and 4: premises will (3) vary hours or (4) close due to unwillingness or inability to pay the levy

In response to the Council’s consultation, 42% of businesses reported they would voluntarily change their permitted licensing hours in response to the introduction of the levy. A smaller number argued that the levy would force some businesses to close as they became economically unviable. Consultation submissions from business hypothesised that these possible responses could lead to a range of unintended consequences, including undermining the levy, re-introducing a ‘terminal hour,’ reducing the diversity of late-night provision and ultimately generating negative economic consequences to the local area in the form of reduced employment and local investment (figure 3). Nearly all of these claims about potential impacts were made without reference to research evidence (and in fact, as will be raised in the Discussion, there is evidence that challenges these impact claims). However, one pub company’s consultation submission did cite a study before going on to develop unsupported claims about the lack of a need for, and impact of, the levy. The study referenced57 58 had found changes from fixed to staggered closing times were not associated with changes in overall violence but ‘may have contributed to additional problems by spreading violence into the early hours of the morning’ (Humphreys58 p8). The pub company’s submission referred to these study findings and then claimed they should be interpreted as meaning that more policing was not needed for the NTE and that the levy would not lead to extra policing:

Overall the net effect should be that same level of policing is required as previously. Cynics may suggest the Late Night Levy is a tax-raising measure to pay for the same level of policing that has always been provided (consultation response, Pub company).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Hypotheses 3 and 4. Orange bubble=national variable; yellow bubbles=local variables; green bubbles=immediate theorised impacts stemming from levy introduction. Solid line: positive relationship between variables; dashed line: inverse relationship between variables; dotted green lines: theorised impacts stemming from the levy introduction. ASB, anti-social behaviour; NTE, night-time economy.

Only two consultation responses, submitted by Public Health and the Clinical Commissioning Group, considered the LNL in terms of its ability to reduce alcohol consumption and associated harms by restricting the availability of alcohol. In contrast, all other system actors discussed the levy in terms of addressing the harms associated with acute intoxication, focusing primarily on disturbance, anti-social behaviour, crime and to a far lesser degree on health-related indicators such as ambulance call-outs or hospital admissions. In this sense, discourses around reducing or preventing alcohol consumption (primary prevention) were largely absent, with a focus instead on making the NTE a safer space for consumption and the possible economic and cultural impacts of the levy.

As businesses shut early, or closed entirely in response to the levy, business actors hypothesised that the LA would fail to generate sufficient revenue to provide the new proposed services: ‘We remain to be convinced that the LNL will raise the amounts of money anticipated, as a significant number of permissions within (LA) are likely to be withdrawn, by way of the free minor variation procedure’. (consultation letter, Pub company and brewer). This represented an example of a perceived negative feedback loop; it was hypothesised that as fewer businesses remained to contribute to the levy through late-night provision, the ability for the levy to continue as an intervention would be jeopardised.

If businesses varied their operating hours to avoid the levy, some in the licensed trade argued, without reference to research evidence, that this could effectively reintroduce a ‘terminal hour’, whereby many premises close at the same time, which would lead to an increase in crime and anti-social behaviour:

If a number of premises reduce their hours as a result of the levy, this could potentially create anti-social behaviour issues with a large number of premises closing at the same time and a return to the spike of crime, disorder and nuisance and midnight observed across the country prior to the introduction of the Licensing Act 2003 (consultation response, trade organisation representing on-licence premises).

Some system actors expressed concern that smaller, independent businesses as well as those which are not alcohol-led, would be most affected by the levy, leaving a less diverse NTE dominated by pub chains and clubs. A reduction in diversity was theorised to make the LA less attractive, which, in turn, was hypothesised to have the potential of leading to negative economic impacts as customers choose to go elsewhere, moving beyond the boundaries of the LA:

Many operators will have to curtail their hours irrespective of the economic consequences, thereby reducing the number of post-midnight premises in the borough. […] visitors to the Borough’s late night economy [would be] choosing other areas of London where no such restrictions apply with obvious economic consequences for [LA]’s late night economy and the businesses that rely on it (consultation response, operator of managed pubs).

In underscoring how elements of the system are interconnected, a number of businesses suggested that the LNL would have negative economic impacts that affect more than just late-night alcohol retailers, making the LA a less appealing area to operate a business:

I am currently looking at sites in the borough; I run a high end food and drink offer, if this levy is introduced I would have to look if the operation could still be viable. My venues do not run beyond midnight but I understand that early evening venues are intrinsically linked to the later venues and if these were to close or relocate it would reduce footfall in the areas affected (consultation response, prospective licensee).

Ultimately, many businesses argued that the LNL, through changes to opening hours, lower profit margins, premise closures and lack of investment in the local area, would result in negative economic consequences and job losses for the LA as a whole:

The council will further kill off the high street if they implement this levy. Pubs and bars will re-locate to other nearby locations where the levy is not in place and lose a number of job opportunities for local people. I thought the council’s major objective as to increase employment opportunities for local people, not decrease it (consultation response).

Hypothesis1: increased resources

The key hypothesis as described by those who designed and implemented the levy was that it would bring in additional resources to manage and police the NTE (figure 1). In the first year, the levy raised £397 278 and in the second year £377 122 (Council LNL Year 1 and 2 Reports). While these figures were lower than the Council’s projected £450 000, the Council described these as sufficient to plug ‘an identified gap’ in managing and policing the NTE (Council, LNL Year 1 and 2 Reports). The additional resources were used to fund an NTE-specific police team and a four-person community safety patrol, delivered by a police-accredited, private company, that worked Thursday–Sunday nights from approximately 20:00–8:00. The new community safety service is the primary focus of phase 2 of this process evaluation; an overview of the structure of the new service is provided in table 5, along with output data from the service provider’s annual reports.

A key component of the new patrol service, which significantly increased from year 1 to year 2, was engagement with users of the NTE to ensure their welfare and to intervene early in anti-social behaviour, disturbance and nuisance to prevent its escalation:

So not only are they there to deal with the response side of it, but it’s to try and prevent that happening in the first place, so to deal with those people who potentially would go on and commit further offences because they’ve started shouting and swearing and causing problems with someone up this end of the street. By the time you get down the other end, they’ve stopped in five other pubs on the other way, not been challenged, not been highlighted to anybody on the way down, although their behaviour’s getting more and more rowdy. Then they go in, have a fight or cause a disturbance and need for police action further down the road (interview, police officer).

The welfare aspect of the service, which included community-safety officers helping members of the public, was also considered a critical component of the service and as shown in table 5, increased significantly from year 1 to 2. The officers also provided medical care and the medical service represented an evolution of the service. While it was always within their remit to have a first-aid trained officer, they expanded this provision shortly after starting the service and purchased additional medical equipment. In addition to supporting members of the public, the medical side of the service was seen as a low-cost mechanism to reduce the burden on the London Ambulance Service and NHS. In the first 2 years, the service reported preventing or cancelling the dispatch of 54 and 57 ambulances, respectively, which they calculated as savings of £16 200 and £14 478.

The Council reported a 17% reduction in alcohol-related crime between midnight and 8:00 and a 14.4% reduction in alcohol-associated violence compared with the previous 12 months, although they assumed that this was not all attributable to the levy. They also reported a large increase (29–30%) in calls to the police and anti-social behaviour line about alcohol-related incidents, which they further argued justified the need for the levy funding (Council, LNL Year 1 Report). In year 2, the council reported a 21% reduction in alcohol-related crimes compared with the previous 12 months, and a 24% decrease in anti-social behaviour calls (Council, LNL Year 2 Report).

Engagement with the licensed trade

While hypothesis 1 emphasised the resources to police and manage NTE users in the area, the new patrol service also sought to develop relationships with local actors. Notably, they tried to develop relationships directly with the licensed trade—to monitor and support licensed operators to encourage safer business practices aimed at minimising anti-social behaviour within and outside the premises. In the first year of the levy funding, the patrol provided an introductory visit to 251 of the venues on the levy, which they argued was as an important mechanism to overcome hostility towards the levy and its funded patrols.

Outside of the initial visits, the patrol worked to develop relationships and trust with venue staff through the repeated interactions; a key element of this, which they contrasted with the police, was the deployment of the same officers every night, particularly in the first year of the service:

One of the things you absolutely have when you’re any form of policing, really, you’ve got to have that consistency. You’ve got to have the relationships. That comes from, you know, repetition. It’s from meeting the DPSs [designated premise supervisors] on a regular basis, building up a trust and an understanding of what you’re there to do […]. Well if you’re on rotation you can’t possibly know. You wouldn’t even know who that person is and you certainly wouldn’t be able to kind of build a balanced intelligence picture (interview, community safety officer).

The Council in its report on the levy following the first year of implementation similarly underscored that the patrol was ‘resourced by regular officers’ and highlighted the relationships they developed with businesses:

Not only have the Nightsafe Patrol Officers have developed a good working relationship with licence holders and their door staff the team have acquired excellent working knowledge or the night-time economy in Islington and made a significant contribution to information gathered by the police and Local Authority (council, LNL Year 1 Report).

When probed, community safety and police officers described an evolution of the relationships such that many licensees began to engage with the service, overcoming their initial resistance:

We came up initially against a lot of unhappiness because it’s another tax effectively, a levy on these premises. They don’t want to pay it. They’re already paying ridiculously high rates and other business taxes and stuff. So, but, you know, I get that. But we’re seeing a change now, you know. A year, 18 months down the road, they can see a benefit to it, so […], if they need help they’ll get help. You know, they’ll prevent stuff happening and hopefully make their business more attractive (interview, police officer).

Others licensees, however, remained what the officers referred to as ‘hostile venues,’ continuing to oppose the levy and its associated services. Officers put this down to a misunderstanding of the service’s remit: ‘they (the licensees) see it as an enforcement role instead of a support role’. (excerpt from fieldnotes).

The community safety service was tasked with collecting intelligence to help the police or inform licensing decisions. Key to this intelligence-gather strategy was developing cooperative, rather than adversarial relationships with venue managers and staff, as described above. Information gathering and sharing among police, patrol and licensed venue operators was reciprocal, or in system terms, represented a positive feedback loop. Closer relationships among these three groups of actors appeared to emerge as a consequence, along with an ‘othering’ of certain venues who remained outside of this information sharing subsystem. Furthermore, the information gained was used to inform licensing decisions:

And [the community safety officers] assist us as well. Not just us as licensing officers, but the police on the whole, because within our briefings we can say to them, just little things that have happened, that you wouldn’t normally get a chance to deal with, can you go and check on this and this, this, this, and just have a look and even in terms of where new applications are coming in and people are asking to do various different things in their licence, and we’re thinking, not sure you could do that, but we need to check the place out. […]. And they report back to us, and then that assists us in saying whether someone can or can’t have a licence. It’s invaluable, really (interview, police licensing officer).

Through the mechanisms described above, the Council, the police and the community safety officers reported that more venues were operating in a ‘responsible manner’ following the implementation of the levy. Hence, while the initial hypothesis around extra resourcing focused on policing and management of NTE users, by the second year of LNL’s implementation a new mechanism for impact had emerged through information sharing and relationship building between NTE operators and the agents that patrolled and policed the NTE:

Interviewer: do you think it (LNL and other licensing policies) has changed kind of how people consume alcohol in the borough?

Respondent (Police licensing officer): I don’t think it’s changed how people consume their alcohol in the borough. I think it’s changed how operators operate.

Taken holistically, the new service was perceived to have changed how actors within the system behaved and interacted with one another, disrupting previous patterns of behaviour as system elements responded and adapted to the new services.

Hypothesis 2: reduced support for PPP schemes

The second key hypothesis, articulated by businesses, was that if businesses were liable to pay the levy, they would no longer support PPP schemes, particularly the local BID (figure 2). This initial hypothesis did not accurately theorise how the system would adapt in the first 2 years of the intervention. Instead, in October 2016, members of the BID ‘again voted resoundingly for us to continue’ (BID website) and the BID expanded to cover a larger geographical area. Following the introduction of the levy, the BID reported a key priority for safety in their area was: ‘achieving 24-hour security at (BID area) through co-ordinated working with street patrol (LNL-funded service)’ (BID Annual Report, 2015/2016) and a licensee described a reliance on both BID-funded and LNL-funded patrols:

The night time economy is a major contributor to the wealth of the [BID area]. Making sure the environment is fun yet safe is a huge undertaking, not only for us licensees but also for the police and [LA] Council. [BID name] makes sure we are all working together. Not only do we have the [BID-funded] Police Team at our disposal but can also rely on [LNL-funded service] (BID Annual Report, 2016/17, Bar Owner).

Prior to the levy’s implementation, members of the licensed trade argued that the BID-funded services addressed their policing and safety needs. However, as the LNL-funded community safety patrol became embedded in the local system, some members of the BID came to see the community safety patrols as a complement to their own funded services and promoted collaboration between the two services, leading to greater resources for managing the local environment.

Hypotheses 3 and 4: premises will (3) vary hours or (4) close due to unwillingness or inability to pay the levy

The final hypotheses were that a large number of premises would vary their hours in response to the introduction of the LNL, or in some cases close completely, which would lead to unanticipated consequences (figure 3). The data reported by the council showed that approximately one quarter of all premises who were initially liable to pay the levy either varied their licence hours or closed prior to the implementation date, which was lower than the 42% who indicated they would during the consultation period. The majority of these businesses varied their hours, rather than permanently closing their doors.

The majority of premises that were identified as being liable to pay the levy continued to operate after midnight and the LA did not see a re-introduction of a ‘defacto terminal hour’. However, there remained clusters of bars and pubs that closed at similar time, which the community safety officers would refer to informally as ‘kick-out times’.

Members of the licensed trade and some residents and visitors theorised the levy would create an NTE that lacked diversity, which in turn would drive down visitor numbers. During the course of fieldwork, we observed a busy NTE with bustling streets and busy venues. All the users of the NTE we spoke with during the course of fieldwork in the second year of the levy described numerous and diverse places to go out in the LA:

[Name] was talking about how there used to be only one place really to go (The Name—which she says is a great pub), but now there are so many options. The places to go out don’t just include alcohol: “It used to be that there were just three places to eat … [she lists their names] and now there are so many to choose from (excerpt from fieldnotes).

Cumulatively, these data show that some premises did vary their hours in response to the introduction of the levy, but the levy remained viable and that an insufficient number of premises closed at midnight to reintroduce a ‘terminal hour’. The LA maintained a reputation for providing a diverse and busy NTE following the implementation of the intervention.