Responsible gambling codes of conduct: lack of harm minimisation intervention in the context of venue self-regulation

Abstract Responsible Gambling Codes of Conduct (CoC) are used around the world to describe electronic gambling machine (EGM) operator commitments to reducing harm from gambling. In addition to the provision of passive product information and warnings, CoC describe how venues should assist EGM users displaying signs of problematic gambling. The focus in this paper is on venue adherence to the active strategies described in these documents relating to supporting ‘responsible gambling’ and discouraging harmful, intensive and extended gambling. The paper triangulates data from aspirational statements by EGM operators published in CoC documents; structured, unannounced observations by the research team in 11 EGM venues; and interviews and focus groups conducted with 40 gamblers and 20 professionals in Melbourne, Australia. Results showed only isolated evidence of supportive interactions between staff and gamblers to address gambling harm. The weight of evidence demonstrated that venues often fail to respond to signs of gambling problems and instead encourage continued gambling in contradiction of their CoC responsibilities. Signs of gambling problems are a normalised feature of EGM use in these venues. To genuinely address this public health and public policy challenge, improved consumer protection for gamblers may be achieved through legislation requiring venues to respond to signs of gambling problems. This may include a range of measures such as banning food and beverage service at machines and limiting withdrawals of cash by gamblers, as well as using behavioural tracking algorithms to identify problematic gambling patterns and binding universal pre-commitment systems to complement supportive interventions by venue staff.


Introduction
Community hotel and club-based electronic gambling machines (EGMs) comprise the largest share (51%) of Australia's $AU22.7 billion gambling losses (Queensland Government Statistician's Office 2016). Around 75% of people experiencing gambling problems report EGMs as the primary source of this harm (Productivity Commission 2010). EGMs have been shown to promote misconceptions among regular users through reinforcement that encourages continued gambling despite substantial losses. Such features include celebratory effects when gamblers bet on multiple lines and experience a 'loss disguised as a win' 1 (Harrigan et al. 2014).
The concept of 'responsible gambling' was developed by the gambling industry in response to community concern about its products (Blaszczynski et al. 2011). A recent study from Finland (Selin 2016) described responsible gambling as a form of industry self-regulation that lacks credibility. The term has been criticised as being elastic, non-specific and open to interpretation (Livingstone & Woolley 2007;Livingstone et al. 2014). Responsible gambling shares common origins with the term responsible drinking, which has been described as strategically ambiguous (Babor 2010) and problematic for emphasising individual responsibility over that of the government and the industry. A further concern is that promoting responsible gambling messages may also lead gamblers to believe that there is a way to use EGMs safely. However, a recent study reported a dose-response relationship between EGM expenditure and harm, concluding that there is no evidence that lower EGM expenditure avoids the risk of harm (Markham et al. 2015).
It is possible to prevent and reduce harm through upstream 2 measures such as the modification of EGMs; for instance, by adjusting game features and changing machine characteristics such as introducing universal, binding precommitment technology. Norway has the most advanced EGM harm reduction regulatory system with demonstrated reductions in expenditure and gambling problems. These reforms involved a temporary ban on machines from 2006-2009, followed by the re-introduction of modified machines that removed banknote acceptors, reduced maximum bets and introduced universal weekly and monthly maximum loss limits and a full pre-commitment system (Rossow & Hansen 2015). Few other jurisdictions operate universal pre-commitment systems for EGMs.
Considerable investments have been made in Responsible Gambling Codes of Conduct (CoC). EGM venues around the world frequently use variations of these documents (Hing 2003; The Association of British Bookmakers 2015; Ontario Lottery and Gaming Corporation 2016). These documents outline the venue's duty of care and specific harm reduction measures venues should follow to minimise harm associated with the products they provide. CoC documents must be available in venues and typically appear on venue websites serving to promote their corporate social responsibility efforts (Hing 2003;Bondy et al. 2008). It has been suggested that the existence of CoC may also serve to mitigate operator risks of potential future litigation by gamblers who have been harmed on their premises (Sch€ ull 2012).
In Victoria, CoC are a condition of licencing under the Gambling Regulation Act 2003 (GRA 2003) and the content of the CoC must be approved by the state regulatory body, the Victorian Commission for Liquor and Gambling Regulation (VCGLR). The CoC describe processes by which gamblers may obtain cash and collect winnings and provide information about gamblers help services and self-exclusion programmes. A venue's CoC is often considered by the VCGLR when granting new or additional licences to EGM operators and these documents are given considerable weight in regulatory settings (Francis et al. 2017). In these hearings venues frequently cite their CoC to demonstrate their competence and capacity to ensure the provision of responsible gambling on their premises. However, there is limited evidence of efficacy for many of the components of the CoC, including self-exclusion. For instance, self-exclusion often relies on self-enforcement by gamblers and manual recognition by venue staff and has been found to be prone to breaches ). These shortcomings highlight the importance of conducting research that investigates the practice of self-regulation for these potentially harmful products.
Studies have demonstrated that there are visible indicators of gambling problems and, with training, staff can observe these (Delfabbro et al. 2007;Delfabbro et al. 2016). A key responsible gambling practice found in all CoC is that staff actively respond to any signs of gambling problems. A recent study identified over 30 behavioural indicators that may be displayed by EGM gamblers experiencing harm. The degree of certainty about the presence of a gambling problem is increased with the observation of multiple behaviours; however, even a small number of indicators (3)(4)(5) is likely to be highly indicative of gambling problems (Delfabbro et al. 2016).
In Australia, EGM venues are commercial operations managed by the private sector. Hotel and club venues that operate EGMs pay a tax to the state government that is indexed according to the monthly average aggregate user loss on machines at their venue. The VCGLR regulates the provision of EGMs and is required to detect non-compliance in EGM venues. The Act (GRA 2003, s. 3.4.25 (g)) provides for the VCGLR to take disciplinary action against venues who repeatedly breach their CoC, which may involve a suspension or loss of licence and/or fines. The VCGLR employs inspectors responsible for ensuring compliance with relevant regulations; however, the resources allocated to this have been described as inadequate (Willingham 2014). Further, to our knowledge, there is currently no formal system by which the VCGLR verifies compliance with aspirational claims in CoC regarding staff responses to patrons exhibiting signs of gambling problems. This may partly explain why there were no reported formal warnings issued to the more than 500 EGM operators in Victoria in 2015-16 (VCGLR 2016b).

Aims and objectives
The aim of this paper is to explore whether practices relating to responsible gamblingas reflected in Victorian gambling venues' mandatory CoCare actually implemented in local club and hotel venues. The paper particularly focuses on staff intervention with gamblers displaying signs of problematic gambling.
The objectives of this paper are firstly to compare expected and observed staff interaction with gamblers according to commitments outlined in CoC and secondly to triangulate these observations with the experiences of gamblers and professionals.

Methodology
The data analysed for this paper was part of a larger mixedmethods study Gambling in Suburban Australia (GISA) that explores the role of place in gambling consumption (report forthcoming). The study was conducted in two distinct local neighbourhoods of Melbourne. Data collection occurred over the year from February 2015. The current paper triangulates qualitative findings from observations conducted at all 11 venues located within the two areas ( Table 1), analysis of the 8 CoC documents published by these venues, and semistructured in-depth interviews and focus groups from 40 gamblers and 20 professionals who respectively lived or worked in the prescribed areas.

Recruitment
Adult gamblers (18 yearsþ) were recruited into the larger GISA study via a short English language survey. The key inclusion criteria for participating in this study was that the gambler lived in one of the local sites studied. All gamblers reported in this paper had visited at least one of 11 local venues (5 clubs and 6 hotels) in the study area. The survey (not reported in full here) collected information about participants' demographics, use of local recreational facilities, venue visitation and gambling attitudes and behaviour, including the Problem Gambling Severity Index (PGSI). Recruitment for interview from the survey was purposive with participants selected based on the experience of past or current gambling problems and the use of EGMs in local hotel or club venues within the two study sites. Local gambling and community help services in each site were purposively contacted and invited to take part in interviews or focus groups.

Data collection
Gamblers Semi-structured, in-depth interviews (Galletta 2013) were conducted with gamblers in English. Equal numbers of male and female gamblers (20 each) were interviewed. The majority were aged 25-64 (33 gamblers) with a smaller number 18-24 or 65-74 years (3 and 4 gamblers in each group respectively). Three quarters of gamblers spoke English as a main language at home and 39% were born in a country other than Australia. A summary of gambling activities of these participants is provided in Table 2.
A range of topics about their experiences of gambling was covered, including experiences of interaction with staff (see the question guide, Appendix A). For the purposes of this paper, data analysis focussed on staff interactions, such as supportive interventions to discourage excessive gambling. Sampling aimed to ensure a balance of male and female participants across a range of ages. It was anticipated that around 40 interviews would result in saturation for the broader GISA study that sought to distinguish differences between two sites. Previous qualitative research has found that 12 interviews are usually sufficient to reach saturation (Guest et al. 2006). Forty-four interviews were conducted for the wider GISA study in total; however, four of these were excluded from this analysis as these participants did not use EGMs locally. Saturation was achieved with no new  Percentages are based on available data and may not add to 100.0% due to rounding. Data presented was obtained from completion of the survey with additional information obtained at interview. a. Participants completed the PGSI and were also asked if they considered they had ever had a lifetime gambling problem when completing the survey. Four participants who answered no to this question had PGSI scores in the moderate risk range.
information emerging from the data (Fusch & Ness 2015). Interviews ranged in length from 30-159 minutes with an average length of 58 minutes. These participants were each provided with a $AU50 grocery voucher in recognition of their time. Quotations from gamblers are reported in the results with a 'G' followed by an identification number.

Gambling professionals
Selected professionals, including venue, treatment and policy and regulation professionals, participated in semi-structured, in-depth interviews to enable more detailed documentation of their knowledge and experiences. Focus groups (Wilkinson 2015) were also used to capture discussion and debate from a larger number of participants in a relatively short period of time. All professional participants were asked to complete a short demographic form at the outset of the interview or focus group. Two mixed-profession focus groups (67 and 99 minutes) comprised 15 professionals (5 venue and 10 treatment professionals), and 8 individual interviews with venue (3), treatment (3) and policy and regulation (2) professionals (range 57-94 minutes) were conducted with a total of 20 professionals. 3 Professional experience averaged 9.1 years (range 0.1-33 years). Quotations from professionals are reported in the results with a 'P' followed by their study identification number. Professional interviews and focus groups asked about the effects of gambling-related harm on their clients. Professionals were also asked what they thought encouraged or discouraged gambling in venues. Question guides for focus groups and interviews with professionals are provided at Appendices B-D.

Venue observations
The CoC cover very similar topics and each venue linked to a CoC on their website; some venues used identical CoC resulting in 8 versions of CoC analysed in total. Observations were conducted multiple times at each of the 11 venues. Venues were each visited between 2 and 5 times. Observations within venues were unannounced and involved researchers participating as 'detached insiders' (Li 2008). In this study, this meant that interactions with other patrons were avoided and researchers used EGMs periodically to maintain an unobtrusive presence in the venue.
Observations were between 20-100 minutes in length, with an average length of 45 minutes, and were predominantly conducted on weekdays between 9 am and 8 pm. Venue EGM numbers and losses as reported by the regulator (Table 1) provide an overall indication of venue scale and activity, with annual gambler losses at the venue ranging from $AU516,534 to $AU13.6 million. Observations of individual gamblers were generally of short duration partly because it was difficult to undertake longer observations unobtrusively. Voice memos were taken during observations to improve the accuracy of observation notes.
Researchers familiarised themselves with the venue CoC documents and together with a gambling behavioural checklist of validated signs of gambling problems (Thomas et al. 2014), developed an understanding of common signs that should warrant a supportive interaction from staff. In addition, a series of prompts about the venue environment (such as who is at the venue, activities they were engaging in, venue promotions, presence of children, betting styles) were prepared prior to the commencement of fieldwork to assist observations.

Data analysis and interpretation
Quantitative demographic data from all focus group and interview participants and basic descriptive statistics of 40 gamblers were analysed using STATA 14 TM . The data from both sites are analysed in aggregate here. All interviews and focus groups were digitally recorded and professionally transcribed with consent from participants. Transcripts, along with CoC documents and summary observation notes, were uploaded into NVivo 11 TM software. Documents were initially thematically coded to nodes by the first and second authors. Codes were refined, sorted and clustered as analysis progressed (Miles et al. 2013;Saldaña 2015). Coding was cross-checked and validated between these two authors. The first two authors held frequent discussions about the themes to test observations and insights that were emerging from the data.
Ethical approval for the study was provided by the Australian Institute of Family Studies HREC (ref 14/27).

Results and discussion
The eight CoC publications provided information about the responsible manner in which gambling is expected to be supplied by venues to their customers. They also form the basis of the interventions researchers expected to observe in venues where signs of gambling problems were present.

CoC in relation to responsible gambling
The analysed venue CoC shared many common features. Broadly, these included: a requirement to display information about the venue's commitment to responsible gambling and where gamblers can seek help and information about the venue's selfexclusion programme; a description of venue protocols: for obtaining winnings, in the event of staff developing a gambling problem, for revising the code of conduct document itself, and lodging complaints about the venue; descriptions of signs that indicate a gambler may be experiencing problems, and behaviours deemed inappropriate and warranting action from staff. These included some but not all items on the behavioural checklist. Signs of problem gambling emphasised inappropriate social behaviours and emotional responses (signs of 'aggression', 3 Three professionals attended a focus group and an interview. 'shouting at the machine or other people in the venue', 'appearing extremely sad or depressed', being 'withdrawn', 'emotional' or 'sweating excessively whilst using an EGM'). There was also reference to signs of intensity and duration ('extended gambling') and money seeking ('continuing to gamble with proceeds of large wins'). Reference to losing control in the context of alcohol consumption was made in four of these documents.
A notable difference was that in some venues staff were permitted to use the EGMs after their shift, whilst in others they were not.
The following section presents results of the thematic analysis, triangulating data sources (observations, interviews, focus groups). In each theme, data is compared to content from the CoC. Consistency and divergence between data sources is discussed. The analysis revealed themes that demonstrated a contradiction between stated responsible gambling aspirations of the CoC and actual existing practices in five areas: an overall lack of staff interaction in the context of apparent gambling problems, staff encouraging gambling, a lack of intervention when gamblers were accessing cash, gamblers using multiple machines and breaches of self-exclusion.

Staff interaction
It was clear from our observations that venues were generally compliant with the passive responsible gambling strategies described in CoCs, such as the display of signage, provision of information about the venue self-exclusion programme and where to seek help. However, compliance with active strategies involving a response to signs of gambling problems was frequently lacking.
In 34 hours of observations, signs of probable gambling problems were almost always present at all venues. This included numerous occasions when researchers observed behaviours such as gambling very fast and intensely, betting above $3 a spin on a machine, gambling through a usual mealtime and withdrawing cash multiple times. However, there were no observed approaches by staff to the gamblers to interrupt EGM use (e.g., suggesting that they slow down their gambling or take a break).
Gambler interview data supported these observations, with gamblers recalling only rare and exceptional instances of staff-initiated interactions related to their gambling. Of the 36 gamblers who were asked about this specifically at interview, four reported relatively minor efforts to discourage excessive gambling. One gambler for instance described her appreciation of the single occasion in her many years of gambling when a staff member provided empathetic advice: "Well the one that did it [provide support] to me. She looked like a friend. 'Cause I went up to her and I got really upset and [she said], 'Wendy 4 go home. Go home, Wendy … What are you doing out this late?' You know, like 1 o'clock in the morning … That one time it happened she wasshe was lovely 'cause she cared." -G286 This contrasts sharply with the general gambling experiences recalled, particularly those of the remaining 32 gamblers who reported never having been approached by staff about their gambling. This lack of interaction was, for some, despite a desire for additional support: "There's been many, many times where I would've loved for someone to stop by and say, 'Do you need someone to speak to?' … I'm sitting there and promising myself, 'Okay, this is the last $50 … ' And you see the money go down quite quickly and you know that that money is needed for so many other things in your life and yet you can't walk out, so maybe just having that someone come up to you and say, 'Do you need assistance?', would have just been enough to get you out even that one time to be able to … have that money for something else." -G364 A lack of intervention was apparent even when gamblers were demonstrating clear signs of problems.
"I've never been offered the support. Which is quite funny 'cause I often say to them things like, even last fortnight I said to the girl, 'I've got a list here of things that I have to do and a note on the bottom saying "do not spend this money it is not spare"', you know. I said 'so much for that note' and she [staff member] went 'Oh, you know' [oh well]" -G1134 However, this view was not shared by all gamblers. Several expressed surprise and concern that an intervention might be expected of staff: "No … I'd be so negative about it [staff offering support], yeah. I think I'd be a bit upset with them, a bit spare with them, thinking, ' … Do they know that I'm a problematic gambler or something?' … I think that would be quite the embarrassment." -G251 Staff were observed regularly in the EGM area: servicing machines, providing book pays, 5 delivering food and refreshments and sometimes engaging in friendly small talk with gamblers. This indicates that opportunities to provide assistance in line with CoC were available but not utilised.

Encouragement of gambling
Researchers observed multiple instances where staff effectively encouraged gamblers to continue gambling by delivering food and drinks to patrons at EGMs. These were often provided free of charge. Gamblers also widely reported these practices in venues, consistent with our observations. "The only reason they'd approach you, … 'Would you like another drink, sir?' You know, that's it. Or, 'Look, we've got a tray of sausage rolls by the by. You're welcome to go grab a couple'." -G1094 "No they'd encourage it [continued gambling]. They'd give you a coffee or a Coke or -the longer you played there, the more they'd give you … when you're winning, they'd attend to you. But if you're losing or anything -they wouldn't come near you." -G258 Another gambler reported more general forms of encouragement from staff: "No I've never had anyone talk to me about it, nah. They probably encourage you more. Like, 'Oh you know, … I saw this 4 Name changed. 5 A book pay allows a gambler to cash out of the machine without collecting $1 coins paid at the machine. It involves requesting a venue staff member to authorise a paper docket, which can be exchanged at the cashier station for notes or a cheque. bloke winning like $2,000 on this machine yesterday.' Like, 'Oh, really? Okay, I'll have a go maybe'." -G254 Further encouragement to continue gambling was also personally experienced by a researcher during an observation. In this case the researcher (who was gambling) requested a 'book pay' following a $AU60 win. However, the gaming attendant declined this request, on the basis that it was venue protocol to only provide a book pay for values over $200. The attendant went on to suggest that the gambler should continue using the machine as she might experience another win and accumulate sufficient funds to later warrant a book pay.

Lack of intervention when accessing cash
There is consistent evidence that people experiencing harm from gambling are much more likely to withdraw money from venue cash facilities compared to those not experiencing harm (Thomas et al. 2013; Victorian Responsible Gambling Foundation 2015). Removing access to automatic teller machines (ATMs) in venues in Victoria was a significant measure introduced in 2012 aimed at preventing and reducing harms associated with EGMs (Thomas et al. 2013). There was a 9% decline in real EGM spending across the state in the period after the removal of ATMs ), initially indicating there may be some effects with this change.
Previous to this, a $AU400 daily limit was mandated on ATM cash withdrawals per venue in Victoria. However, this was replaced with a $AU200 limit on EFTPOS 6 withdrawals per transaction with no in-venue daily limit on withdrawals.
A key argument for the retention of EFTPOS facilities after the removal of ATMs revolved around the requirement that gamblers had to interact with a staff member to make a withdrawal. It was argued that this would act as a natural deterrent to multiple withdrawals and would provide staff with a point of intervention.
The current study found that the deterrent effect of staff contact had subsided.
"I don't think it's made a difference at all … Maybe it made it a little more inconvenient but that's it." -G1126 "Oh yeah there's nobody says 'y'know ay, hold on, this is the third time you've come to me and y'know I just think you're spending too much'" -G1026 Gamblers and professionals interviewed were in agreement that removing ATMs but retaining EFTPOS without a daily limit had done little to curb access to cash for many gamblers at risk of gambling harm, and may have done more harm to some: " … when they had them and you could only get $400, after that you had to leave the venue and go somewhere else. Um, which gave you time to think and break, you know. Nowadays you can go up to the bar but you can only get 200 at a time. However, you can do that as many times as you like. So I've gone through a $1,000 [a session] that way" -G1134.
Observational data also showed that staff facilitated cash withdrawals in instances where this should not have been encouraged. For example, a researcher's card was declined when attempting to withdraw cash at one venue. Instead of encouraging this customer to complete her gambling session, the staff member suggested she try withdrawing a slightly smaller amount. This may lead a gambler who is struggling with controlling their spending to exhaust all available funds, an industry practice documented by Sch€ ull (2012) as 'playing to extinction'.

Stronger regulation of access to cash needed
The difficulty of limiting access to cash in the absence of clear regulation regarding daily limits is demonstrated by this venue professional: "Even when going to the EFTPOS and that's probably one of the biggest issues for staff, when someone goes 8, 9, 10 times. And then the card's declined. So it's very embarrassing for the person concerned and also the staff member. Um, theywe teach them to have a conversation and just say, 'Look, how are you doing?' on the, say, the third visit to an EFTPOS … or 'You told me not to give you any more, I'm just reminding you.' Which a lot of punters do … but of course, like, it's their money. If they want to pull it out, they have to be allowed to." -P540 Of particular concern was an instance where a venue staff member described being actively discouraged by their manager from fulfilling obligations under the CoC: " … You get your responsible service of gaming license and you [learn you] can legally tell them [gamblers] 'no you've had enough go home, stop, cool down … but you get [to work] and your boss is like 'you do that, you're out of here' … well, he didn't say it directly but that was what he was hinting at … I had one guy come in and get like $200 cash out three times in an hour … the kind of guy who had dirty work boots, you know, ripped up clothes and he was just really desperate. And it was like my third shift there and I knew I should've said something like, 'Dude, slow down' but then the manager's just like … 'No, no, no, give him another one [cash withdrawal]'" -P1063 Frequent observations of multiple withdrawals by gamblers, without intervention, indicates the need for stronger regulation of support staff intervention when a request is made to withdraw cash.
"I can't fathom on what planet we thought taking away that $400 limit was okay … there's no need for it to be more than $400 a day. I don't care whether you're addicted to playing or you're a social gambler, $400 for anyone is a lot of money." -P1703 A policy and regulation professional concurred with the assessments of other professionals and gamblers stating that the lack of daily limits on EFTPOS was "a policy area I think that could do with attention by government."-P547 These findings indicate restricting access to cash may need to be mandated by law as barriers to intervention by staff extend beyond confidence and training. 6 EFTPOS is the acronym for 'electronic funds transfer at point of sale' that involves the use of a debit or credit card at a payment terminal that authorises a purchase or withdrawal of cash.

Use of multiple machines
Using 2 or more machines at the same time is a sign of gambling problems (Thomas et al. 2014) as it intensifies the speed at which losses occur. Researchers observed gamblers using two machines simultaneously on many occasions. Observations in one venue noted a sign advising patrons "Only 2 machines maximum at any time can be played by 1 person continuously." Staff interventions in response to multiple machine use were described by one venue professional as inconsistently applied and possibly driven more by a desire to reduce queues and provide a greater number of patrons access to a machine than to reduce harm: "Staff won't talk to people [to get off the second machine] if it's during the day and it's quiet, but they will at night time if it's busy and people are waiting to get on machines." -P1703 Disallowing the use of more than one machine at a time may improve practices in venues.

Breaches of self-exclusion programmes
Self-exclusion programmes (SEP) are a form of pre-commitment where gamblers can elect to enter into an agreement with gambling venue(s) to prevent them from entering the gambling area of a venue. As such, SEP are a downstream measure designed to support gamblers who have an established gambling problem. SEP are required to operate in all EGM venues in Victoria. The system requires gamblers to register and sign a deed that they agree to be refused access to the nominated venue(s) gambling area for between 6 and 24 months. Venue staff are required to prevent selfexcluded gamblers from using machines and penalties may apply for venues that are not compliant with the selfexclusion orders.
Despite being recognised by staff and known to be selfexcluded, some gamblers expressed frustration at breaches of self-exclusion by venue staff when they were allowed to use EGMs: "Interviewee: … they don't really care because I did have myself on the … exclusion list … it only lasted three months, if that, because one of the security guards on the door … said, 'Oh well, I haven't seen you come into today so I don't know what's going on.' Interviewer: So he let you in?
Interviewee: He let me in, yes." -G123 Another gambler described the lax approach to selfexclusion where she phoned a venue after returning home from a gambling session to request that they comply with her order:

Strengths and limitations
In the current study, our observations were unannounced. Direct observation by researchers was likely to have captured 'real world' behaviours and responses. This avoids problems associated with the Hawthorne effect (Wickstr€ om & Bendix 2000), allowing a more natural reflection of actual behaviours within venues for both staff and gamblers. This evidence was strengthened by the triangulation of evidence from gamblers and professionals. Accordingly, researchers frequently observed behaviour that suggested gamblers were experiencing problems, without intervention by venue staff on the floor. Researchers made repeated efforts to identify evidence that was inconsistent with this conclusion (Miles et al. 2013) and the very limited contrary evidence has been reported here.
Remaining unobtrusive in the venue meant that extended observation of individual gamblers was not possible. The limited nature of our role as researchersunlike staffmeant that not all behaviours on the behavioural checklist (Thomas et al. 2014) were observable (such as extended and/ or frequent sessions, increases in spending, gambling until closing time). It is also possible that interventions were made to some patrons outside the periods of observation. Further, most observations were undertaken during business hours or early evening, with no observations in the middle of the night (e.g., at closing time, which is commonly 4:00 am). Daily, weekly and seasonal variation in EGM use is likely. Despite these limitations, researchers were able to observe many different behaviours indicative of gambling problems including using multiple machines, withdrawing cash multiple times, becoming agitated or frustrated with the machines, and betting more than $3 per spin. Improved understanding of results could have been achieved had researchers been able to further triangulate findings with detailed machine data to examine betting behaviours at different times across venues (e.g., the load up of cash on machine, denomination of machine, bet size, session duration). Unfortunately, these data are not currently available.
Interviews and focus groups were conducted in English, which limited participation by non-English speaking gamblers. All venues were invited to participate in the research; however, only 2 venues agreed to participate and did so by displaying recruitment flyers. Two venue staff also participated in this research; greater participation from staff directly employed by EGM operators could have provided additional valuable information. The remaining 9 venues either did not respond or refused to participate.

Conclusions
Stigma and shame associated with gambling often leads gamblers to hide problems from family and friends (Hing et al. 2016). Therefore, venue staff are likely to be the first to see evidence of gambling problems, and could play an important role in early intervention and prevention, by offering support to gamblers who may be experiencing harm. For this reason, EGM venues are an important frontline site of intervention to minimise harm to gamblers. Unfortunately, the results of this study show that venues do not routinely follow regulations or their venue's CoC by supporting gamblers showing signs of gambling problems. Nor, at times, is support provided to those actively seeking assistance.
It should be recognised that responsible gambling measures articulated through the CoC are downstream measures that depend upon the identification of gamblers already displaying signs of a problem. Although staff training has been found to encourage interaction with gamblers, under the current arrangements the CoC are not an effective response to reducing or preventing harm. Improved safety for gamblers may be achieved if regulatory frameworks mandate the interruption of EGM use where indications of gambling problems are apparent. This may include, for instance, making it an offence to serve food and beverages at machines and limiting cash withdrawals.
It would appear that the business model currently operating in Australia and many other jurisdictions does not incentivise the enforcement of the CoC and other responsible gambling initiatives due to the interest of operators in maximising profit. Despite much evidence obtained in the course of our research demonstrating that breaches were common, there was no reported formal action taken by the VCGLR against EGM venues during 2015-2016 (VCGLR 2016b). Current arrangements effectively allow EGM operators to self-regulate compliance with CoCs through their annual reporting mechanisms. Formalising specific offences such as encouragement of continued gambling by those experiencing problems, using more than one machine at a time and restricting access to cash is warranted. Resources to monitor compliance with this should also be made available to regulators.
This study also raises important questions about the adequacy of existing regulations focussed on in-venue assistance to gamblers and the utility of CoCs. The significant weight these documents carry in outlining how venue operators should deliver responsible service of gambling in Victoria and elsewhere may currently be overvalued. This is consistent with self-regulation in analogous areas such as alcohol control, which has found self-regulatory codes are 'under-interpreted and under-enforced' (Babor 2010).
However, it may be unrealistic to expect staff intervene in some gambling sessions, given the known difficulty of such confrontations for both staff and gamblers. The implications of these findings therefore extend beyond interactions between staff and gamblers. User tracking technology could be better utilised to reduceand in some cases circumvent the need for such interactions. Behaviour tracking technology is likely to be most effective if provided in the context of a binding, universal pre-commitment system that (a) requires gamblers to set a loss limit before commencing a session, and (b) prevents further gambling once the loss limit is reached (Williams et al. 2007). In addition, algorithms could be applied to detect emerging gambling patterns indicative of harm (Schellinck et al. 2015) and, once detected, automated messages could alert gamblers to the potential for harm and ways to avoid this. The introduction of these systems could achieve significant improvements to current consumer protection and harm minimisation efforts.
(probes: social isolation, family conflict, violence [perpetrator or victim?], diversion of household money, e.g. food, housing, education, health) 21. What impact do you think your gambling problems will have on your future? And on the future of your family? 22. Who manages the money in your household?
If partner knows about gambling problems, has this changed since they discovered you had a problem with gambling? 23. What is the most significant change that has happened to you since you have been gambling? 24. Why would you choose this as the most significant amongst all of these changes that you have already described?

Support services
27. Are you in contact with other services (e.g., Gamblers Help, financial counsellors) to help with your gambling problems? Have they been helpful or not? 28. If not, are there services that you would find particularly useful?
What? Why/why not?