Gambling in Suburban Australia

Research Report – March 2019

10. Conclusions

This study focused on exploring the influence of environmental factors on the uptake of gambling and the harms emanating from gambling in two local areas. In the following section we:

  • discuss the strengths and limitations of the study
  • lay out the study's main findings
  • present our recommendations for further actions.

10.1 Strengths and limitations of the study

The use of multiple methods generated rich data that provided insights in a number of areas that were not the initial focus of the research. This includes data describing the connection between family violence and gambling as well as the experience of migrants in each area. However, the exploratory nature of this research means that it has not exhaustively studied all issues that have been raised in this report. More research is needed to fully understand these complex phenomena.

Those who participated in this study were a self-selected sample who responded to invitations to participate through advertisements and/or referrals from professional services. This may mean that the sample reflects residents with strong views about gambling and may not be representative of the wider local population in each area.

Challenges with recruiting participants from a small local area meant there was an imbalance of genders between the two sites, with only four female gamblers participating in Site 2, compared to 16 in Site 1. The difficulty in recruiting more females who gambled in Site 2 may suggest comparatively low gambling activity among this demographic in this area. However, it is also possible that, despite the multiple forms of recruitment employed, the study was not successful in recruiting a sufficient representation of women from Site 2. Similarly, a significantly lower number of community professionals were interviewed in Site 2. However, triangulation of the data from multiple sources indicated a lower level of professional services relating to social support and welfare in Site 2, suggesting that this reflects the nature of service provision in this area.

The study population sample was uneven between sites overall, with 108 total participants in Site 1 and 51 participants in Site 2. The bulk of these differences in numbers relate to the large sample of Vietnamese focus group participants in Site 1, which was not replicated with a non-English speaking community in Site 2 due to lack of resources.

The core group of participants in this study, with relatively similar numbers between sites, were people who reported being harmed by their own gambling (24 vs 20 in Site 1 and 2 respectively) and significant others (12 in Site 1 and 8 in Site 2). Aside from the relatively small number of female people who gambled and experienced harm in Site 2, we were able to recruit a relatively diverse group of participants in both sites (in terms of age, country of birth and time living in the area).

While the data are based on self-report, and participants self-selected to take part in the research and therefore may not be representative of the population who live in each area or gamble at harmful levels. The patterns of gambling and harm reported by people in these sites are indicative of trends requiring further investigation. Further, these data were supplemented by researcher observations in venues, reports from professional key informants and local resident community members, and VCGLR and ABS census secondary data. This triangulation is a key strength of the study.

Researcher venue observations totalling more than 34 hours (2,060 minutes) were conducted. This involved multiple observations per venue, enabling the emergence of clear patterns of gambling behaviour within and across the venues in the study. In interviews, many observational inferences were confirmed by participants' descriptions of their own and others' gambling behaviours; for example, multiple withdrawals of cash. However, observations may be limited in representativeness as researchers were not fully immersed in the venue environment for extended periods of time, particularly overnight or in the very early morning.

The PGSI was developed to measure the likelihood of experiencing gambling problems and adverse consequences and, as recommended by those who designed this instrument, was only administered to survey participants who reported having gambled in the past 12 months (Ferris & Wynne, 2001). However, during interviews with participants who were in recovery and who had not gambled in the previous 12 months, it became apparent that many long-term effects of gambling harm were still being felt, even many years later. These 'legacy' harms were not captured with the use of the PGSI scale administered in this way. In future studies, it would be beneficial to employ a gambling harm measure capable of assessing lifetime or legacy harms, in order to overcome this methodological issue and provide additional quantitative data.

10.2 A lack of alternative social spaces

Our study found that high-intensity EGM gambling was easily accessible in both sites, but especially in Site 1, the area of higher disadvantage.

Evidence presented in section 3.1 and section 3.2 demonstrated a very high level of geographic and social gambling accessibility in Site 1, as measured through the number of venues and machines, and as reported by study participants. Participants also described this as being compounded by a lack of alternative social spaces in this area.

Residents in Site 1 in particular, including people who gambled, significant others and local residents, asserted that gambling was too accessible locally (section 3.2). Many found that they inadvertently ended up at gambling venues when they were undertaking routine daily activities. They also described reluctantly attending venues with their children because of the convenience and indirect promotions that made these venues seem more affordable compared to available alternatives (chapter 3).

This contrasted with evidence from Site 2 where the reported relative abundance of alternative social spaces and activities were combined with fewer EGM venues (section 3.3). Nevertheless, despite the lower level of EGM access, several people who gambled in this area also described the convenience of these venues as problematic. In this context, it should be noted that the comparison to Site 2 is a relative measure, and that local accessibility of high-intensity EGM gambling even in Site 2 remains unusual internationally, given very few other jurisdictions allow EGMs to operate in local communities, particularly when combined with access to alcohol.

Venues encourage the use of EGMs through the promotion of other activities

It is apparent both through researcher observations and self-report from participants in both sites that many EGM venues utilise the crossover between the use of non-gambling facilities in their venues, such as sporting facilities, the bistro and bar, and the EGM area (section 4.1). For instance, activities or promotions such as 'morning melodies' for seniors, subsidised lunches, 'model waitresses', and/or exercise classes attract a wide cross-section of local residents, enabling venues to portray themselves as local hubs for a range of activities, and to enmesh themselves into the fabric of the local community.

Participants frequently reported using EGMs in venues that they had initially attended for other purposes such as dining or socialising. Further, offers such as free coffee and tea provided in the EGM area also appear to be part of a strategy to encourage use of this space in the venue.

EGM venues divert money from other businesses in the local area

EGM operators are able to offset losses on food and beverages, for example, against high-level profits made on EGMs through the taxation system. Other published evidence has shown that a consequence of these subsidies available to EGM venues is that other businesses subsequently experience lower demand for their goods and services. This was demonstrated by participants in this study who described that they would restrict spending on discretionary - and, at times, essential items - in order to gamble (see sections 3.4 and 6.2). This affects the viability of businesses that do not operate EGMs, which may be seen as a form of market distortion, providing EGM venues with a significant market advantage while effectively limiting competition (Carmignani & Eslake, 2018).

Recent research has demonstrated that low- and moderate-risk gamblers account for 85% of the burden of harm at the population level (Browne et al., 2016). Chapters 3, 4 and 5 highlight how limited community amenity and high levels of social stress, combined with seemingly attractive EGM facilities and promotions, leads to higher levels of EGM use.

10.3 Harms magnified in disadvantaged areas

Gambling did harm in both sites but the harm was intensified in the already-disadvantaged site.

A 2015 study (Markham et al., 2015) demonstrated that there is no 'safe' level of EGM use. That is, harm increases with increased EGM use. As summarised in the previous section, in some communities, particularly those that are socially and/or economically disadvantaged or otherwise stressed, it can be difficult to avoid EGMs in social settings, such as Site 1. Markham and colleagues' (2015) findings demonstrate that gambling can do harm to anyone, anywhere and may affect any community. However, harms are more pronounced and prevalent in communities that have a higher density of EGMs (see chapters 5 and 6).

Participants across both study areas reported substantial financial, health, relationship and emotional harms from gambling (chapters 6, 7 and 8). Gambling harms were often enduring and affected everyday life, including the ability to regularly maintain essential expenditure such as food, housing, transport and education. Participants who gambled would frequently experience strained or damaged relationships with immediate and extended family members, often as a consequence of unpaid loans or other financial impositions and/or distrust. The ongoing and long-term effects of such harm were described by some participants as severe and resulting in distrust, relationship dysfunction, family violence, other crimes, mental illness and suicide.

In Site 1, where many participants were already under considerable financial and social stress, the severity of harms among those who gambled reached crisis levels relatively quickly. Recovery, on the other hand, could be slow and difficult, creating an enduring legacy. This included difficulties such as struggling to make mortgage or rental payments, having a house repossessed, bankruptcy and/or homelessness (section 6.2 and section 6.3). In several cases, these crises amplified existing chronic life stressors, such as pre-existing poverty, family violence and/or intensive caring responsibilities. Stressful life events, such as the death of a family member or divorce, put further strain on many households, which in some cases led to an escalation in gambling (chapter 5). The consequences of such life events could be ruinous for families with limited resources.

Many participants in Site 2 were in a position to draw down on their own assets, or the wealth of extended family, to endure the damaging effects gambling had on their finances, relationships, health and/or ability to work (section 6.2). This reflects the findings (section 3.1) of the secondary data, showing higher average socio-economic status in Site 2. Many participants reported diverting rent or mortgage payments for gambling. However, those in severe and ongoing crises in Site 2 were in the relative minority.

For those participants in Site 2 with access to more significant social and financial capital, the diversion of funds into gambling meant that they were not able to allocate spending to other non-gambling activities or discretionary items. The erosion of assets, such as home ownership or superannuation, resulted in a reduction in socio-economic status and loss of lifestyle amenity both immediately and in the longer term (chapter 6). While life events had inevitably affected people in this area, many reported having strategies in place that would support their recovery. However, these harms should not be considered less important. They represent substantial opportunity costs, and impose real and often enduring costs on people who gamble, families, communities and society.

Ripple effects of harms are widespread and significant

Participants in this study reported a range of gambling harms experienced not only by those who gambled but also affecting significant others, workplaces and employers and the wider community (chapters 5-8). A recent study reported that for every 'problem gambler', an average of six others are directly affected (Goodwin et al., 2017). In communities where individuals and households experience significant underlying financial and social stress, it is plausible and likely that the number of affected others will be higher. These harms also extend beyond the household and affect the local economy and community via the erosion of social and financial capital, the effects of implicit subsidies on local expenditure patterns and the associated diversion of spending into less employment-intensive industry sectors (South Australian Centre for Economic Studies, 2005).

People who gamble who experience an addiction may be unable to make rational decisions about their gambling expenditure (Yücel et al., 2018). In line with this, research participants reported the need to obtain emergency food supplies, or forgo meals, and re-mortgage their homes in order to continue to gamble (chapter 6). This suggests that money intended for essential household spending has in these cases been diverted into profits for gambling operators.

In the current study, participants reported concern about reliance on government revenue from gambling, as well as policies that allow high levels of EGM availability in low socio-economic areas (section 9.1). This is an essential policy-relevant consideration with ethical implications and requires further exploration of upstream population-level measures to support better informed debate about the consequences of current EGM gambling arrangements.

These findings highlight the need for further consideration of EGM licensing and regulation, such as the current location and number of community-based EGMs and manner in which they are provided. This was a perspective expressed by many participants in this study (section 9.1). New policy settings may be required to address this significant transfer of resources from families and communities to corporate interests (and their shareholders). Further, the significant effective regressivity of existing gambling tax arrangements requires careful consideration.

Specific recommendations are provided in section 10.4.

Social isolation can be a risk factor for use of EGMs

Social isolation can also be a result of gambling harm.

Social integration (a form of social capital) has been shown to improve health status (Berkman & Glass, 2000). Many participants who gambled reported dislocation from their families and social networks (chapter 5, particularly section 5.4). In some cases, those who already experienced isolation and loneliness began gambling as a way to address this situation. Participants described the apparently 'non-threatening' environment of the EGM venue, in which lone attendance is common and where staff seem friendly and welcoming (section 4.2). In other cases, isolation was a consequence of harmful gambling behaviour that strained relationships. For these participants, it was very difficult to rebuild trust with friends and family, even after they had reduced or ceased gambling (section 8.2).

Under these circumstances, gambling-related harm is significantly magnified and intensified in areas already experiencing considerable socio-economic disadvantage (Site 1), when compared with a less disadvantaged area (Site 2). This is unsurprising, given households with fewer resources struggle to absorb continued losses of large sums of money.

The relationship between disadvantage and EGM density and expenditure has been well established in other studies (Doran & Young, 2010; Livingstone, 2001; Marshall & Baker, 2001; Rintoul et al., 2013). The 2013 study by Rintoul and colleagues described a regression model, where 40% of the apparent effect of disadvantage was accounted for by the density of gambling opportunities. This indicates that an oversupply of gambling opportunities drives high losses, rather than consumer demand. This observation is supported by the views of residents who described major concerns about the ubiquity of EGMs in their community, arguing that there were too many opportunities to gamble, particularly in Site 1 (see section 9.1).

10.4 Recommendations

The relationship between life stressors and gambling harms is reflexive, with no single pathway of causation. This does not mean that the issue is too complex to address (Petticrew et al., 2017). Rather, this reflexivity should be considered in the development of appropriate responses, acknowledging that multiple strategies at multiple levels of government and regulation will be required to reform harmful gambling arrangements.

The range and magnitude of gambling-related harms described by many participants in this study, combined with their overwhelmingly negative attitudes towards gambling, mirror findings from other population surveys in Australia. These other studies have similarly found current industry operations and government regulation are not in line with community expectations. Further, the overwhelming majority of the population are in favour of EGM reform (McAllister, 2014).

When considering what reforms might be possible, it is useful to explore lessons from other contexts. In 2006, in response to widespread public protests, the Norwegian Government removed their monopoly-operated EGMs from the community completely. A study of this experiment demonstrated a decline in gambling problems following this significant change. Eventually, in 2008, the Norwegian Government re-introduced newly designed gambling machines incorporating a range of harm minimisation features including a universal pre-commitment system, statutory loss limits and account-based operation (Rossow & Hansen, 2016).

While gambling has a range of other substantial and significant negative externalities that are still not well understood, there are some obvious options that could prevent and reduce gambling-related harm, thereby increasing the net benefits of gambling. Exploration of policy options to significantly reduce harm from EGMs is required within the Australian context. This could include, among other things, consideration of the appropriateness of the widespread availability of EGMs in local communities, the modification of EGM structural characteristics, and features such as a reduction in maximum bet sizes and universal pre-commitment systems (Livingstone, 2017; Rintoul & Thomas, 2017). A range of options are possible and include:

  • Restrict the distribution and level of EGM availability in local communities.
    • This could range from the removal of non-casino-based EGMs, as in Norway and as has been recently proposed in Tasmania (Carmignani & Eslake, 2018), through to restricting access to EGMs in local neighbourhoods (e.g. reducing opening hours, the number of venues and number of machines in venues).
  • Provide less harmful gambling machines by introducing harm reduction measures such as:
    • universal pre-commitment systems
    • an effective electronic and universal self-exclusion system
    • reduction in maximum bets
    • modification of machine characteristics known to increase reinforcement (e.g. free spins, losses disguised as wins).14
  • Separate alcohol from ambient gambling, as in the UK.
  • Create alternative non-gambling spaces, including those open during the day, and at night,15 for local residents to meet and socialise.
  • Restrict indirect venue promotions, including 'family-friendly' subsidies and activities for families and children.
  • Increase resources to police and regulators to ensure EGM venues comply with existing laws and regulations, including active enforcement of the provision of responsible gambling and improved codes of conduct (Rintoul et al., 2017).
  • Review tax concessions to 'not-for-profit' clubs who operate EGMs and reform 'community benefit' schemes.
  • Implement improved protections by venues and the financial and banking sector. For instance, this may include:
    • preventing the use of credit cards and the provision of other high-interest loans to fund gambling
    • preventing the use of superannuation funds and reverse mortgages to fund gambling
    • limiting EFTPOS transactions in venues
    • reviewing withdrawals of large sums of money and checking with joint account holders to prevent financial abuse in families.
  • Require venues across Australia to provide, to regulators and public interest researchers as a condition of licensing, detailed data about EGM use at venues (e.g. individual bet size per spin, machine utilisation, session length).
  • Invest further in research to facilitate the continued development of the evidence base to inform policies that will support the prevention and reduction of gambling-related harm. This should include independent research to:
    • understand the nature and dynamics of gambling and family violence, including financial abuse in the context of gambling
    • understand the biological mechanisms between social stress and gambling
    • undertake analysis of detailed machine-level data from EGM venues (when available)
    • assess the policy implications of the long-standing regulatory requirements such as the practice that requires EGM operators to hold a liquor licence.
  • Develop and implement a National Gambling Strategy to provide coordinated direction and support to the prevention and reduction of gambling harm in communities across Australia.

14 For further information about the nature of these and other EGM features see Livingstone, C. (2017). How electronic gambling machines work: Structural characteristics. Melbourne: AIFS.

15 For example, the Libraries After Dark Program, see