Free to gamble: The roles of the gambling industry and policy in a modern Australian society

EXECUTIVE SUMMARY

• Gambling is a pastime that has been recorded throughout human history, and holds an important place in Australia's economy, society and culture.
• Individuals derive many benefits from gambling, including potential winnings and a sense of enjoyment and fun.  Empirical estimates suggest that the existence of gambling delivers substantial welfare gains to consumers.
• Gambling businesses across Australia deliver a range of benefits to the economy, such as the provision of a desired product for consumers, investment, technological development, employment, and spin-off benefits for associated industries such as tourism and accommodation.
• The freedom of individuals to spend their own money, including on gambling, is fundamental to a free, open society.
• At the same time as governments have opened up new opportunities for people to gamble, assorted anti-gambling lobbies have advocated a host of restrictions on the industry.
• The extent of "problem gambling" is a major issue in Australian policy discussions, and a source of regulatory growth for the gambling industries.
• Empirical estimates of problem gambling prevalence are hampered by methodological and statistical issues, however every indicator suggests that problem gambling is small and in trend decline.
• The gambling industries have been very proactive in tackling problem gambling, and where possible these initiatives should be supported on subsidiarity grounds.
• Poorly targeted one-size-fits-all government regulations, such as gaming machine caps and bans on ATM withdrawals from gambling venues, can adversely affect the interests of all gamblers.
• State governments also levy a range of taxes, levies and license fees specific to the gambling industries.  Despite accusations of gambling revenue "addiction", the relative importance of these taxes has declined over the past decade.
• Anti-gambling lobbies accuse the states of using gambling taxes as a way to prey on vulnerable problem gamblers.  However, it is not possible to establish the true extent to which revenues are acquired from problem gamblers.
• Recent proposals to increase gambling taxes are likely to harm consumers, including those on low incomes, gambling businesses and the state government revenue base.
• On economic, social and liberty grounds, restrictive government policies against gambling should be resisted.  The capacity of individuals to be free to gamble should be the benchmark against which government policies towards gambling consumers and providers will be judged

THE NATURE AND BENEFITS OF GAMBLING

While definitions vary, gambling can be described as the wagering of money or goods on an event with an uncertain outcome, with intent to win additional money or goods.  This definition is used for the purpose of this paper.  The term gambling encompasses both wagering and gaming activities.

TYPES OF GAMBLING PURSUITS

Games of chance and skill have been played by people for all of recorded human history.  Implements associated with the practice of gambling have been found in ancient China dating back to about 2,300 BC, as well as in India, Egypt and Rome.  People of all creeds have participated in gambling.  For example, in 1020 AD King Olaf of Norway and King Olaf of Sweden resolved a dispute over territory by rolling two dice.

Despite the prevalence of gambling in human activity, there is also early evidence of the state attempting to constrain the gambling behaviour of its citizens.  Various English monarchs outlawed gambling from the 14th century onwards as it, along with the playing of certain games, was seen as diverting men from practicing archery.  In the twentieth century, one of the first steps of the totalitarian governments of Mussolini, Hitler, Mao and Castro was to ban almost all forms of gambling.

Modern democracies have also regularly targeted gambling.  However, in almost all cases anti-gambling laws have not banned gambling per se.  These laws have tended to be class-based, allowing gambling for the select few, but not for the masses.

In Australia, for the first six decades of the twentieth century the main issue of concern was illegal off-track gambling on horse racing, generally conducted by SP bookmakers.  While wagering was seen as legitimate when conducted by the upper classes on-course at Randwick or Flemington, or in an exclusive city club, it was not when undertaken by the working classes at pony and trotting tracks, or at John Wren's tote.  Religious ministers, conservative politicians and other "wowsers" tolerated the former, yet regularly lambasted the latter. ⁽¹⁾

Since the arrival of legalised off-course betting in the early 1960s, there has been a general liberalisation of the rules around wagering.  Currently, wagering activities on racing and sporting events comprise legal betting on horse and greyhound races with bookmakers and totalisators, both on race courses and off-course through Totalisator Agency Boards (TABs).  Sports betting involves wagering on local, national and international sporting events, whether on or off-course, in person, by telephone or through the internet.  These forms of gambling exist in all States and Territories.

There has also been a growth in gaming options in Australia in recent decades.  Current gaming activities can be categorised as follows: ⁽²⁾

• Lotteries:  Conducted by both government and commercial operators, an individual whose ticket is selected in the lottery to win a prize based on the total amounts wagered after deduction of a set percentage by the operator to cover costs.
• Keno:  Keno is a game in which a player wagers that their chosen numbers match any of twenty numbers randomly selected from a group of eighty numbers.
• Pools:  Pools is a game in which the winning numbers are based on the results of United Kingdom or Australian soccer matches.
• Gaming (or poker) machines:  Machines are available where people pay to play games where they must match symbols, either on mechanical reels that spin and stop to reveal the symbols or, in the case of an electronic gaming machine (EGM), on a video screen.

Gaming is supplied by a variety of entities.  Lottery and pools tickets are provided by outlets such as newsagents, while people can take bets on table games, gaming machines and Keno systems at 13 casinos located across Australia.  Gaming machines and Keno systems also exist in clubs and hotels across most States and Territories. ⁽³⁾

While poker machines were introduced in New South Wales in the 1950s, it was their arrival in several other jurisdictions in the 1990s that has seen poker machines become the SP bookmakers of the modern era -- the form of gambling that attracts the greatest demands for further regulatory intervention to protect some citizens from the claimed deleterious effects of this mode of gambling, in particular.

INDIVIDUAL BENEFITS FROM GAMBLING

Gambling is one of the many leisure activities available to the Australian population.  Like any other good or service, the demand for gambling products is derived from the subjective preferences of consuming individuals.  These preferences may be shaped by a range of factors including:

• Economic considerations, including price, quality, income level and availability of substitutable goods and services.
• Non-economic motivational forces such as a sense of excitement and fun when consuming gambling services, the desire to "be in it to win it" the anticipation of a win, or testing one's luck or skill.

As will be further described below, many providers bundle their gambling products and services with food, beverage and accommodation services, music and other entertainment.  For a good many Australians who benefit from these bundled outputs, gambling is part and parcel of an enjoyable day or night out "on the town".  Indeed, gambling is not unlike any other form of entertainment, such as attending a cinema or theatre or watching a cricket or football match, in that people willingly expend part of their disposable income to obtain a form of enjoyment.

The key benefit for those engaging in gambling is the enjoyment, satisfaction or utility they acquire from it, compared with the gains obtained from their next preferred form of expenditure and activity.  This gain can be measured with reference to a measure called the "consumer surplus" -- or the difference in dollars between the amount the consumer pays for a good or service and the maximum amount the consumer would be prepared to pay for it.

Quantification of the consumer surplus derived from gambling is by no means a straightforward task.  As the 1999 Productivity Commission Inquiry into Australia's Gambling Industries noted, consumer surplus estimates are contingent on variables such as price and income elasticities of demand, and the budget shares of gambling expenditure.  Some of these variables are difficult to observe directly and therefore assumptions need to be made about their respective magnitudes.

Notwithstanding these estimation difficulties, consumer surplus remains one of the best indicators available to quantify the extent of consumer welfare from gambling.  For recreational (non-problem) gamblers, the Productivity Commission considered that "the presumption that the surplus represents a genuine addition to the welfare of consumers is a reasonable one". ⁽⁴⁾  The surplus attained by these consumers, as estimated by the Commission, was in the range of between $2.7 billion and $4.5 billion in 1997-98.

Later Australian studies also reveal that consumers benefit from the availability of gambling products and services.  A 2006 study commissioned by the Northern Territory Treasury found that gambling consumers enjoyed a surplus of between $42 million and $91 million. ⁽⁵⁾  A revision of a 2001 study undertaken by ACIL Tasman found that regulatory restrictions on gaming in the Ballarat region would reduce net consumer surplus by between $8 million and $348 million per annum, depending on the type of restriction imposed. ⁽⁶⁾  In a study of the economic and social impacts of gambling in Tasmania, it is estimated that a net benefit of between $72 and $120 million is derived from consumers' enjoyment of gambling. ⁽⁷⁾

These studies suggest that, at least for the overwhelming majority of people, gambling delivers significant value for money and hence well-being.  The increase in gambling expenditure from 1990-91 to 1999-2000, and the significant level of expenditure continuing to this day, reflects the free choices of individuals to satisfy their demand for activities -- such as the use of gaming machines -- that were previously either highly restricted or illegal.

ECONOMY-WIDE BENEFITS FROM GAMBLING

Producers provide a range of gambling products and services, thus catering for consumer demand, because it provides a greater return on their resources than the next best alternatives.  Indeed, recent profiling of the gambling industries reveals that significant benefits are being accrued by the Australian economy because of these leisure activities.

According to the Australasian Gaming Council 2008-09 Database on Australia's Gambling Industry, there were over 15,200 gambling businesses in 2005-06 (Table 1).  These included 13 casinos, 4,652 outlets providing TAB betting services, 4,756 lottery outlets, and 5,833 hotel and club gaming machine venues.

Table 1:  Number of gambling businesses in Australia, type and jurisdiction, 2005-06a

	Casinos	TABs				Lottery outlets	Gaming machine venues
		Outlets b	Hotels	Clubs	Other c		Hotels	Clubs
NSW	1	313	1,100	716	179d	1,570	1,784	1,352
Vic	1	108	419	67	99c	780	249	274
Qld	4	148	427	116	68	1,117	770	573
WA	1	91	176	13	5	564	na	na
SA	1	57	273	16	26	535	499	79
Tas	2	31	74	14	8	83	94	10
ACT	1	16	4	30	5	73	14	65
NT	2	10	18	18	7	34	38	32
Australia	13	774	2,491	990	397	4,756	3,448	2,385

^a Estimates provided in this table refer to the number of gambling forms available within venues.

^b Comprising standalone stores.

^c Includes on-course, at casinos, sporting events, newsagents and mobile outlets.

^d Tabcorp also operates on-course facilities across NSW and Victoria in line with the racing events calendar.  The numbers above are fixed locations at Randwick (NSW) and Flemington (Vic).

Source:  Australasian Gaming Council, 2008,
A Database on Australia's Gambling Industry 2008/09.

About 46 per cent of the total number of gaming businesses was located in NSW, followed by Queensland (21 per cent), Victoria (13 per cent), South Australia (10 per cent) and Western Australia (six per cent).

A useful measure of the economic contribution of gambling industries to the national economy is value added.  Information obtained from industry consultants IBISWorld by the Australasian Gaming Council shows that clubs (hospitality) and pubs, taverns and bars contributed about $10,420 million to the Australian economy. ⁽⁸⁾  Together with the estimated value-added of $2.9 billion by casinos in 2006-07, this was equivalent to about 1.2 per cent of GDP. ⁽⁹⁾

The gambling industries are linked to other sectors of the economy.  For example, tourists from interstate or overseas attend gambling venues, with flow-on benefits for local economies.  In 2006-07, there were approximately 2.1 million international visitors to casinos in Australia.  Expenditure of $135 million at casinos was reported by these visitors in 2006. ⁽¹⁰⁾  This information excludes gamblers known as "high-rollers", who gamble significant amounts of money and receive incentives from competitor casinos to play in their gaming rooms.

Gambling venues also cater for interstate visitors within Australia.  According to the Australasian Gaming Council, there were about 4.2 million interstate visitors to casinos in 2006-07 alone. ⁽¹¹⁾  Of all interstate tourists, $241 million was spent on gambling products and services in 2005.

The Australian gambling industry also has extensive links with gaming machine manufacturing and technology.  According to research undertaken by the Centre for International Economics, revenues from the manufacture of gaming machines in Australia was $610 million in 2006-07, including $310 million in export revenues. ⁽¹²⁾

The gambling industries also make a substantial contribution to employment in Australia, with a diversity of occupations available to suitably qualified people from licensed gaming staff at casinos and trainers and jockeys in thoroughbred racing to food preparation staff and cleaners in various venues.

In 2004-05, it was estimated that approximately 151,300 people were working in gambling businesses at hotels, clubs, casinos, lottery agencies and thoroughbred horse racing venues. ⁽¹³⁾  At least to some extent, the revenue proceeds from gambling products and services support the engagement of people in these industries.

Changes in employment and wages effectively represent another indicator of economic welfare derived from gambling industries.  Data from the ABS reveals that employment in hospitality clubs, pubs, taverns and bars with gaming facilities increased by about 10,300 people from 1997-98 to 2004-05 (Figure 1). ⁽¹⁴⁾  In addition to this, employment in the casino industry increased from about 19,500 in 2003-04 to about 19,900 in 2006-07. ⁽¹⁵⁾

Figure 1:  Employment in selected venues with gaming facilitiesSource:  ABS, various years, Clubs, Pubs, Taverns and Bars, Australia, cat. no. 8687.0.

According to IBISWorld data, wages expenditure per employee in the hospitality clubs industry increased from about $24,900 in 2002-03 to about $31,400 in 2006-07 whilst wages grew in pubs, taverns and bars from about $17,800 to $24,000 over the same period (Figure 2). ⁽¹⁶⁾  For the casino industry, wages per employee rose from about $36,600 in 2002-03 to about $51,000 in 2006-07. ⁽¹⁷⁾

The relatively high rate of casualisation in these industries was reflected in wages being below the national average (ordinary time) wage of $55,124 in 2006-07. ⁽¹⁸⁾

Figure 2:  Wage expenses per employee in selected venues with gaming facilitiesSource:  Australian Casino Association, 2008, The Australian Casino Industry Economic Report 2006-07;  IBISWorld, 2008, Industry Report Clubs (Hospitality) in Australia, February;  IBISWorld, 2008, Industry Report Pubs, Taverns and Bars in Australia, April.

GAMBLING IN A FREE SOCIETY, AND
THE PROBLEM GAMBLER QUESTION

ON THE FREEDOM TO GAMBLE, AND ANTI-GAMBLING SENTIMENTS

The freedom of the individual to spend his or her money as he or she sees fit is fundamental both to economic welfare and to individual liberty.

As discussed in our submission to the 1999 Inquiry into Australia's Gambling Industries, the fundamental approach to assessing gambling activities that take place is to appreciate that they represent the preferences that people express in their spending decisions.  Where gambling is prevented or deterred in any way, consumer expenditures are being redirected towards goods and services from which they would expect to enjoy less satisfaction.

As noted above, gambling is an activity that has taken place since recorded human history, with no evidence to suggest that such pursuits have destroyed the fabric of society over time.  It could even be argued that, to some extent, the liberalisation of Australia's gambling industries since the 1980s has been consistent with a growing acceptance of gambling as a legitimate consumption choice in a free and modern society.  It follows that the liberalisation of gambling is also congruent with the inherent value of people being able to spend their own money as they please.

Paradoxically, as gambling has grown in Australia there has also emerged an assortment of groups demanding various restrictions on such activities.  Some groups hold moral objections against gambling, subscribing to the belief that it is a "sinful" activity.  For example, groups have linked their criticisms of gambling to Biblical strictures against covetousness (the process of obtaining money or goods without returning "fair value" in exchange).  Alternatively, as the Geelong Catholic Social Justice Committee has stated in various submissions to government, gambling is a sin if "it raises the pursuit of riches to the status of an idol". ⁽¹⁹⁾

More recently, a number of well-organised individuals -- including self-styled "no pokies" politicians and media-savvy campaigners -- and lobby groups have sought to oppose gambling on other grounds.  They have moved from value-laden notions that gambling is "bad" per se toward more subtle argumentation about the risks of gambling.

In particular, these point to the desirability of government regulation to eliminate the adverse financial and social consequences of pathological (or problem) gambling.  Seizing on the emerging "behavioural economics" literature, for example, they argue for restrictions on gambling to counter cognitive or behavioural biases held by consumers (Box 1).

Box 1:  Behavioural economics and government intervention:  A critical perspective Behavioural economics is the application of cognitive and psychological research to economic behaviour.  It has received widespread publicity in recent years, partly due to Daniel Kahneman, a leading researcher in the field, receiving the Nobel Prize in Economics in 2002, and the appointment of another high-profile researcher, Cass Sunstein, to a top advisory position in the Obama US Administration. In contrast to the neoclassical economic paradigm that individuals behave in a rational manner -- characterised by optimising behaviour with consistent, stable preferences -- behavioural economics insists that people are less likely to pursue choices in their own best interests.  This is because individuals are afflicted with a range of cognitive and behavioural biases affecting their economic choices.  These biases could include self-control problems, information processing failures, and an inability to accurately establish the costs and benefits of action. Behavioural economists insist that governments have a role to address these errors in decision making.  Some advocate a policy stance of "libertarian paternalism", a blend of actions to "improve" and "guide" choices without coercion or the elimination of freedoms. For example, to balance the right of gamblers to continue gambling against the desire to "nudge" them toward less gambling, a suite of "sin" taxes are recommended.  This would raise the cost of gambling, and this extra "pain of the purse" should give gamblers a greater sense of willpower to quit gambling.  Another proposal supported by behavioural economists is to establish laws enabling problem gamblers to put themselves on a list banning them from entering gambling venues or collecting winnings from gambling. Such ideas have been subject to extensive criticisms by economists specialising in public choice theory and Austrian economics.  By employing the standard public choice assumption of symmetrical behaviour by private and public sector agents, Glaeser suggested that government officials are unlikely to set appropriate policies as they are also affected by cognitive and behavioural biases.  Further, successful rent-seeking behaviour by interest groups could detract from the design and implementation of policy. Rizzo and Whitman draw on F.A. Hayek's famous "knowledge problem" to raise doubts about whether governments can accurately take account of the idiosyncratic circumstances and preferences of a multitude of individuals.  Government interventions are based only on the limited knowledge of, and information available to, the government agent, and so it is unlikely that desired policies to correct individual behavioural biases would eventuate. Whitman and Rizzo point out that many of the policy arguments put forward by behavioural economists are conducive to the "slippery slope" effects of more prescriptive policies in the future.  They also point out that, where the effects of errors made by private individuals are relatively localised, policy errors by governments are spread across society and even on future generations. It cannot be denied that individual behaviour is often inconsistent with the formalistic neoclassical ideals.  Nonetheless, the criticisms expressed above provide an important "health warning" against the unquestioning acceptance of behavioural economics by policymakers.  The burden of proof that gambling restrictions will provide net benefits entirely rest with those advocating such policies. Source:  Edward Glaeser, 2006, "Paternalism and Psychology", University of Chicago Law Review 73 (1) 133-156.  Mario J. Rizzo and Douglas Glen Whitman, 2008, "The Knowledge Problem of New Paternalism", New York University School of Law, Working Paper No. 08-60;  Douglas Glen Whitman and Mario J. Rizzo, 2007, "Paternalist Slopes", New York University School of Law, Working Paper No. 07-08.

Despite the sophisticated rhetoric that modern anti-gambling groups typically use, their opposition to gambling appears to rest on the highly moralistic precepts more commonplace in earlier periods.  They are often prone to using colourful, vivid language (for example, their claim that "gambling preys on the weak and vulnerable").  This is used to castigate legitimate businesses that prove gambling products and services, and emboldening policymakers to impose increasingly prescriptive controls over gambling.

It is tempting to dismiss the sometimes emotive criticisms of anti-gambling lobbies in a policymaking context.  However, given the success of these groups in persuading governments to restrict the freedoms of individuals to engage in legal gambling pursuits, and of businesses to provide gambling products and services, it is necessary to critically assess some of the key issues they raise.

THE DIMENSIONS, AND EXTENT, OF PROBLEM GAMBLING IN AUSTRALIA

A major issue raised by anti-gambling groups relates to the prevalence of problem gambling in Australia.  Yet, despite the regular use of the term "problem gambling" in the public lexicon, there is no agreed definition, let alone measure, of this phenomenon.  This has been previously acknowledged by the Productivity Commission, when it stated in its 1999 Inquiry Report that there was a "lack of precision in the definition of problem gambling [that] poses difficulties for the identification of those affected". ⁽²⁰⁾

There are no direct physiological indicators to distinguish problem gamblers from the rest.  As a consequence of this, the psychology literature proceeds to conceptualise problem gambling as a mental health issue.  The Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) of the American Psychiatric Association characterises gambling as an impulse control disorder.  It also sets a range of diagnostic criteria to establish whether an individual displays symptoms of a persistent and maladaptive gambling pattern causing financial, personal and vocational difficulties.

While often used as a basis for measuring the extent of problem gambling, the DSM-IV approach has been criticised in that it was designed for use in a non-Australian clinical environment, lacks product specificity, largely focuses on the act of gambling rather than its consumption, and could be misapplied when used in surveys. ⁽²¹⁾

As noted by Abbott, pathological gambling, as a formal diagnostic entity, was "developed" in 1980;  the very same year that homosexuality as a mental disorder was erased from DSM-III.  He also concluded that "pathological gambling is one of a number of "sunrise disorders" likely to have increased in recent years.  Other disorders placed in this category included major depression among young men, youth suicide, eating disorders and some firms of chronic organic brain syndrome". ⁽²²⁾

The criticisms of the medical disorder/mental health approach have led others to take a broader approach to defining problem gambling.  Some see it as an activity-based spectrum ranging from recreational gambling through to problem gambling leading to adverse consequences to the individual and close relations, friends and colleagues, through to pathological gambling where the adverse consequences tend to be more severe and the gambler is identified against some diagnostic criteria. ⁽²³⁾

On the basis of these competing definitions, several instruments have been devised to measure the prevalence of problem gambling in Australia.  As noted in the PC Issues Paper, these include the South Oaks Gambling Screen (SOGS) and the Canadian Problem Gambling Index (CPGI).  Recent estimates tend to suggest that the extent of problem gambling is small, affecting two per cent or less of the adult Australian population (Table 2).  For comparison, available estimates on the number of individuals who avail themselves (either voluntarily or as part of court orders) for gambling counselling are also presented.

Table 2:  Problem gambling prevalence estimates, selected studies

	SOGS 5+	CPGI +8	Self-referential
NSW	2.55 (1999)	0.8 (2007)	0.01 (2004)
Vic	1.12 (2003)	0.97 (2003)	0.07 (2001-02)
Qld	1.88 (1999)	0.47 (2006-07)	na
WA	0.7 (1999)	na	na
SA	1.9 (2005)	0.40 (2005)	0.08 (2003-04)
Tas	1.41 (2005)	0.54 (2008)	0.1 (2004-05)
ACT	1.91 (2001)	na	0.08 (2002-03)
NT	1.06 (2006)	0.64 (2006)	na

^a Self-referential problem gamblers refer to number of individual counselling clients who self-identify their own gambling problem.  Excluding families and other counselling participants with a relationship to primary individual receiving counselling.

^b Self-referential prevalence estimated as a proportion of state population.

Source:  Australasian Gaming Council, 2008, A Database on Australia's Gambling Industry 2008/09;  M. Walker et al, 2004, "Problem Gamblers Receiving Counselling or Treatment in New South Wales";  M. Dickerson, 2004, "Analysis of clients presenting to problem gambling counselling services, July 2001 to June 2002";  South Australian Department of Families and Communities, 2005, "Break even problem gambling services:  client data (South Australia):  analysis of clients using services 1 January 2003-31 December 2004";  KPMG, 2007, "Break Even Gambling Services Client Information Fact Sheet for the period 1 July 2000 to 30 June 2005";  J. McMillen et al, 2004, "Help-Seeking by Problem Gamblers, Friends and Families:  A Focus on Gender and Cultural Groups".

There remains great controversy over the most appropriate methodology to quantify problem gambling prevalence.  Measurements based on surveys are subject to a number of well-known limitations, such as sampling errors and non-responses.  For some survey respondents, there may be a tendency to confound the direct problems associated with problem gambling, and other issues in their personal lives (for example, relationship strains) that are, at least, not directly caused by gambling consumption.  This may also bias available survey results.

While the number of individuals receiving counselling for gambling problems might be a more direct indicator, prevalence estimates must also be treated with caution.  This is because some gamblers are unaware, or choose not to avail themselves, of such services.  Other gamblers who reach subjectively defined "excessive" levels of gambling involvement can discontinue gambling without any professional assistance required.

The debate on the number of problem gamblers is hampered by the paucity of Australian longitudinal data on this issue.  A comparison of point estimates over time, for example, conceals the switching behaviour that can occur as some problem gamblers become non-problem, or recreational, gamblers.  Some of these people may switch over time as a consequence of assistance from family or counsellors, while others can do so through their own decisions to reduce gambling consumption.

Problem gambling is far from being a static phenomenon.  Even the most severely addicted problem gamblers can remain sensitive to changes in economic parameters relevant to gambling pursuits, such as alterations in the implicit price of games. ⁽²⁴⁾

Furthermore, from an economic perspective it is even less clear to what extent problem gambling is, in fact, a problem for the individuals concerned.  Excessive behaviour is a major rationale for limiting access to gambling, yet excessive behaviour patterns are common to other pastimes and, indeed, many human activities in general.

The Becker-Murphy "rational addiction" model suggests that an individual might rationally choose to consume an addictive good if the present value of doing so is sufficiently high so as to outweigh the expected costs of addiction. ⁽²⁵⁾  As Douglas Walker explains in his book, The Economics of Casino Gambling, "the decision whether or not to gamble is a rational choice. ... The implication here is that even pathological gambling is a condition that results from rational behaviour. ... risking addiction is not inconsistent with rationality. ... it [rational addiction theory] implies that the costs of pathological gambling are actually private in nature". ⁽²⁶⁾

The conditions that cause over-consumption for some people are serious for the individuals and for society.  But, unless they undermine some basic foundations of society, their resolution is best left to the individuals most directly concerned to rectify.  This rests on the well founded premise of a democratic society that the individual is primarily responsible for his or her actions.  On the basis of the best available evidence, as preliminary as these may be, problem gambling does not appear to be of a sufficient magnitude to justify denying the population as a whole of their liberties to consume.

In summary, the best available evidence suggests that the prevalence of problem gambling is small on a population-wide basis, and has declined over the last ten years.  However, more empirical research in this area is required, including the utilisation of longitudinal techniques in order to ascertain the dynamics of gambling consumption behaviour.

INITIATIVES TO ADDRESS PROBLEM GAMBLING

Role of public and private sectors in addressing problem gambling

Governments have applied increasingly restrictive regulations and policies to minimise the extent of problem gambling in Australian society.  Prior to assessing the efficacy of current regulations in this area, it is appropriate to consider the appropriate division of roles by the public and private sectors in addressing this issue.

If problem gambling is pervasive across the population, then there may be a legitimate role for government to address what constitute a collective action problem.  If, as the available evidence suggests is the case, problem gambling is of a small scale and, in parts, is localised by nature, then the appropriateness of government involvement is more questionable.

As noted previously, governments often lack sufficient knowledge to identify problem gamblers (particularly if they adjust their gambling consumption patterns over time), and have difficulties developing targeted and customised solutions to a complex issue that affects relevant individuals differentially.

Regulations and other policies by government that restricts the overall availability of legal gambling also has the potential to reduce the welfare of recreational gamblers, as well as negatively affect the viability of gambling industries.  Other issues, such as the burden upon taxpayers in funding the administration and other costs of government regulations, are also relevant.

The potential problems of government regulation were neatly summed up by Clubs Australia Chairman Peter Newell in June 2008:

"Governments and clubs are committed to ensuring that appropriate and effective measures are in place to minimise the number of people experiencing problems as a result of their excessive gambling.  Clubs are also committed to intervening as early as possible to assist people who have recognised that they have a problem and therefore are likely to benefit from treatment.

The history of harm minimisation measures taken by governments across Australia has been somewhat chequered in that too often governments have implemented measures that may appear politically attractive but in practice are minimally effective in achieving their intended objective of reducing the incidence of problem gamblers or gambling related harm.  A feature of this approach has been the "leap frogging" that has taken place between States to have the "toughest anti gaming laws in the country".

The common feature of these type of "populist", reactive measures is that they are not supported by solid research to indicate they have a chance of real success or will make a real difference". ⁽²⁷⁾

Given the prevalence and nature of problem gambling, various non-governmental entities -- including gambling businesses, not-for-profit counselling services and even family members -- all have an important role to play in ameliorating problem gambling behaviours.  An advantage of actions by these groups is that it allows for bottom-up regulatory experimentation, enabling society to discover the best standards and procedures to address given problems.

A diverse array of self-regulatory initiatives have been occurring for at least the past ten years.  For example, many gambling businesses and industry associations have established their own codes of practice to promote responsible gambling use by consumers (Table 3).

For example, the Australasian Gaming Council's Strategy for Responsible Gambling includes an emphasis on information provision to gamblers, staff training on responsible gambling policies and procedures, and targeted strategies catering for consumers who seek assistance or indicate they have a problem with their gambling.

At present there are some 40 codes of practice in operation across Australia, with some developed or maintained as a requirement of government policy.  Given the extensive coverage of industry codes to address problem gambling, and the policies and procedures nested within each, it would appear that such self-regulation would be sufficient to appropriately address the incidence and prevalence of this issue.

Table 3:  Selected industry responsible gaming codes

Australia	NSW	Vic	Qld	WA	SA	Tas
Australasian Gaming Council (Strategy for Responsible Gambling) Australian Leisure & Hospitality Group Ltd (Responsible Gambling Ambassador Program) Australian Bookmakers (National code of practice for the responsible provision of wagering services)	Tabcorp (Responsible Gambling Code of Practice) Australian Hotels Association NSW (Responsible Conduct of Gambling) Clubs NSW (Code of Practice / Advertising Code of Practice / ClubSafe Code of Practice) BetSafe (Paul Symond Consultancy Hotel Program)	Crown Melbourne Ltd (Code of Practice and Responsible Gaming Support Centre) Tabcorp (Responsible Gambling Code of Practice) Greyhound Racing Victoria (Responsible Wagering Code of Practice)	Clubs Queensland (Responsible Gaming Policy) Tabcorp (Responsible Gambling Code of Practice) UNiTAB (Responsible Gambling Policy) Reef Casino (Responsible Service of Gambling Program)	Burswood Entertainment Complex (Responsible Gambling Code of Practice)	Adelaide Casino (Responsible Gaming Code of Practice / Advertising Code of Practice) SKYCITY Adelaide (Host Responsibility Program) Australian Hotels Association SA (Gaming Care Program) Licensed Racing Club (Code of Practice / Advertising Code of Practice)	Wrest Point Casino (Responsible Service of Gaming Policy) Launceston Country Club and Casino (Responsible Service of Gaming Policy) Betfair (Responsible Gaming Policy) Tasmanian Gambling Industry Group (Code of Practice / Advertising Code of Ethics)

Source:  Australasian Gaming Council, Responsible Gambling in Australia,
Fact Sheet No. AGC FS 22/08.

Electronic gaming machine caps

As indicated, governments have also progressively applied a range of regulatory restrictions on gambling activities, in an attempt to address problem gambling.  A particular, but not exclusive, focus of these policies have related to the availability of gaming machines in casinos, pubs and clubs.  In particular, state governments have progressively applied caps on the number of new machines to be installed in gaming venues (Table 4).

Table 4:  Number of electronic gaming machines and new machine caps

Jurisdiction	Number of gaming machines a	Cap applicable on gaming machine numbers b
NSW	19.0	99,000
Vic	7.5	27,500 c
Qld	11.4	24,705
WA	1.0	523 d
SA	11.1	14,000
Tas	9.9	3,860
ACT	20.4	5,200
NT	12.6	1,190

^a Number of EGMs expressed per 1,000 adult population.

^b Incorporates new policy announcements, but not yet implemented.

^c Excludes regional cap requirement of 10 machines per 1,000 adults in selected non-metropolitan areas.

^d Poker machine numbers restricted to one casino site, and not permitted in clubs or hotels.

Source:  Australasian Gaming Council, Gaming in Australia, Fact Sheet No. AGC FS 03/08;  various state gambling authority websites.

A major rationale for government EGM caps is to tackle problem gambling prevalence by minimising the availability of gaming machines to the overall population.  The 1999 Productivity Commission gambling inquiry supported the notion that problem gambling prevalence rates were generally higher in states and territories with greater accessibility, however this idea has been challenged by a number of experts on gambling behaviour.

A research paper by M.W. Abbott outlines an alternative "adaptation theory" which suggests that an increase in problem gambling prevalence following the initial exposure to EGMs levels off as people adapt and modify their behaviour:

"In New Zealand and Australia, despite substantial increases in EGM availability and expenditure, current national prevalence estimates are between a third to a half what they were 15 years ago.  In Australia there are indications that the relationship between availability and prevalence breaks down somewhere between six to 10 EGMs per 1,000 adults. ... exposure to the agent gambling is multidimensional and the effects of exposure are complex". ⁽²⁸⁾

Increased public awareness of problem gambling, the availability of treatment services, and the capacity of people to help themselves and adjust when, in their view, they are consuming too much gambling product, all help to explain the adaptation hypothesis.  The efforts of the gambling industries, in identifying and encouraging treatment for problem gambling, also appear instrumental in enabling the Australian population as a whole to adapt successfully to the introduction of EGMs since the 1980s.

Former University of Queensland researcher Rohan Miller also stated that the oft-assumed relationship between EGMs and problem gamblers was contentious.  In a study of the Gold Coast region, it was found that the rate of increase in customers who nominated EGMs as their preferred form of gambling declined over time, even as the number of EGMs in the region was increasing. ⁽²⁹⁾

Adaptation to an increasing number of EGMs also appears to have been realised in some overseas countries.  Shaffer suggested that "observations about gambling-related problems in Nevada provide support for the adaptation hypothesis of addiction.  That is, after the novelty of initial exposure, people gradually adjust to the risks and hazards associated with potential objects of addiction". ⁽³⁰⁾

Surveys in the US have also shown that some regions with relatively high gambling expenditure have low problem gambling prevalence.  Norway, with a relatively high density of gaming machines (including in petrol stations and supermarkets) also has relatively low prevalence rates.

It is questionable whether or not EGM reductions and the introduction of machine caps have had any significant effect on problem gambling prevalence.  An analysis of Victoria's regional caps showed no clear evidence of a reduction in gaming expenditure in "capped" regions compared to others. ⁽³¹⁾

If anything, however, it is likely that a cap on EGMs would harm recreational gamblers;  an ACIL Tasman report in 2001 found that residents in the Ballarat region of Victoria would be made worse off were a regional cap to be imposed.  In a revised study published in 2006, it was found that a stricter cap effectively halving gaming expenditure in Ballarat would reduce net consumer welfare by between $8 million and $29 million per annum.  Even tighter restrictions would cause greater losses in consumer well-being, including as recreational gamblers compete for playing time on a capped number of EGMs. ⁽³²⁾

Caps would also have the effect of constraining potential future revenue gains in casinos, pubs and clubs, with subsequent implications for investment and employment growth in these industries.

ATMs in gambling venues

Another issue subject to current debate relate to government regulations preventing individuals from withdrawing cash from automatic telling machines (ATMs) at gambling venues, presumably to prevent problem gamblers from expending in gambling venues.

According to the commonwealth government's National Snapshot of Harm Minimisation Strategies, all states and territories have in place measures relating to ATM access in pubs, clubs and casinos.  Specifically, ATMs and EFTPOS facilities are not permitted in gaming areas, with Tasmania currently the only state prohibiting ATMs in hotels and club gambling venues. ⁽³³⁾

This area of gambling regulation is also increasingly subject to commonwealth involvement, despite gambling policy being the primary responsibility of the states.  The Ministerial Council on Gambling (comprising commonwealth and state ministers) agreed in July 2008 on new priority work areas, including issues relating to access to cash and pre-commitment technologies.  This was followed by independent Senator Nick Xenophon introducing into the Senate in September 2008 the ATMs and Cash Facilities in Licensed Venues Bill to prevent ATMs from being installed, owned or operated in a licensed gambling venue.

It is unclear whether additional ATM restrictions have any significant effect on reducing the extent of problem gambling.  In a submission to the Senate inquiry on the Bill, Professor Jan McMillen stated that recent studies on the relationship between problem gambling and withdrawals from ATMs or EFTPOS machines were inconclusive.  Indeed, ATM restrictions are not only unlikely to reduce problem gambling but impose costs on recreational gamblers.  In the context of gambling in the Australian Capital Territory, McMillen noted that "removal of ATMs from gaming venues would inconvenience a significant proportion of ACT club patrons, recreational gamblers and non-gamblers". ⁽³⁴⁾

As McMillen noted, non-gamblers would be significantly inconvenienced by further ATM restrictions in gambling venues.  In a public hearing on the Senate inquiry on the Bill, the Australian Hotels Association estimated that 60-70 per cent of sales in their venues were unrelated to gambling, including sales of food, beverage and non-gambling entertainment.  In some regional centres and country towns, licensed gambling venues are the other sites containing ATMs at least within reasonable walking distance.  It is therefore unclear how further curtailing the freedom of non-gamblers to access their own money would reduce problem gambling.

TAXATION ARRANGEMENTS

GAMBLING TAXATION IN AUSTRALIA

Apart from the benefits of gambling accruing to consumers and the broader economy, state and territory governments also benefit from the transfer of revenues collected from gambling industries in the form of taxes, levies and license fees.

In practice the states levy a range of taxes on different forms of gambling, which are generally based on the gross value of gambling activity.  These include gaming machine taxes, taxes on government and/or private lotteries, taxes on casinos, race betting and other gambling such as internet gaming, keno and betting exchanges.  Other levies, such as mandatory contributions to specific community programs or funds, or revenue obtained by the distribution of government-owned gambling providers, exist.

Gambling taxes raised vary between the different forms of gambling (Figure 3).  From 1998-99 to 2008-09, it is estimated that total gambling tax revenue increased from $4.1 billion to $5.1 billion -- an average increase of two per cent per annum.  Gaming machine taxes increased at an average annual rate of three per cent, above the increase in total revenue.  However, total gaming machine tax growth has been flat since 2006-07.

The growth in taxes on lotteries and casinos were below the average annual growth rate for total gambling taxes.  Race betting taxes declined over the period, while other taxes increased by 15 per cent on average each year, albeit off a low base.

Figure 3:  Total gambling taxation revenue, by gambling tax sourceSource:  ABS, Taxation Revenue, Australia, cat. no. 5506.0;  State and Territory Budget Papers.

In 2008-09, it is estimated that taxes on gaming machines will comprise 59 per cent of the total gambling tax revenues acquired by the states and territories, followed by lottery taxes (22 per cent) and casino taxes (ten per cent).

Gambling tax burdens also vary considerably across the states and territories (Table 5).  For example, it is estimated that the proportion of gaming machine revenues to total gambling taxation varies from zero in Western Australia to 74 per cent in South Australia.  Lottery tax burdens are reasonably similar across jurisdictions, with the exception of WA with no gaming machine taxation.

Table 5:  Gambling tax sources, proportion of total gambling tax revenue, 2008-09 (per cent)

	Gaming machines	Lotteries	Casinos	Race betting	Other
NSW	64.0	18.4	6.6	10.2	0.7
Vic	62.0	21.5	8.2	7.8	0.5
Qld	60.8	23.9	6.4	4.0	-
WA	-	49.6	35.2	15.2	0.4
SA	73.8	18.7	5.0	2.0	0.5
Tas	-	27.7	65.5	6.7	-
ACT	68.7	25.6	3.8	1.9	-
NT	3.7	17.2	54.0	15.2	9.9

Source:  State and Territory Budget Papers

Casino taxes are relatively more important to Tasmania, Northern Territory and Western Australia, underlining the importance of gambling revenues attained from one physical location (viz. their casinos).  Race betting taxes generally account for a relatively small share of revenue, even though these taxes are important to the WA and Northern Territory gambling tax mix.

Another way to look at the relative burden of gambling taxes is to consider tax capacity and effort ratios calculated by the Commonwealth Grants Commission. ⁽³⁵⁾  The latest estimates by the CGC indicate that the Northern Territory, Victoria and South Australia have effort ratios well above the Australian average and their own gambling revenue raising capacities.  In other words, these jurisdictions tend to overtax their gambling revenue base.  Conversely, Western Australia, ACT, Tasmania and Queensland all have effort ratios below the Australian average as well as their respective capacity ratios.

Figure 4:  Gambling taxation revenue capacity and effort ratios, 2006-07Source:  Commonwealth Grants Commission, Relative Fiscal Capacities of the States 2008.

A FISCAL FEDERALISM PERSPECTIVE ON GAMBLING TAXATION

The imposition of gambling taxes in Australia should be considered within the broader context of commonwealth-state financial relationships, and in particular the significant changes that have occurred over the past decade.

Due to constitutional, political and historical developments, state and territory governments retain a set of narrowly-based taxes that yield insufficient revenue to finance their expenditure functions (for example, health, education, transport, and law and order).  The advent of the GST reforms, including the abolition of a number of state own-taxes such as financial transactions taxes, has reinforced this trend. ⁽³⁶⁾

The gradual erosion of state own-taxation has led to anti-gambling critics suggesting that state governments have accordingly become reliant on gambling taxation, and is unprepared to place further restrictions on gambling industries for fear of risking the growth of a productive base for future taxation.

However, this contention is not borne by the evidence.  Gambling taxation represents less than 12 per cent of state and territory tax revenue, and only about four per cent of total general government revenue.  This is not a share of revenue consistent with any reasonable definition of "reliance".  In addition, gambling taxes as a proportion of state own-taxation and total revenue have both slightly declined since 1998-99 (Figure 5).  The extent of this levelling off in tax reliance is even more significant when compared against the growth of gambling taxation during the 1980s and 1990s following industry liberalisation.

Figure 5:  Gambling taxation as proportion of taxation revenue and total revenue

^a Figures for general government sector.

Source:  ABS, Taxation Revenue, Australia, cat. no. 5506.0;  State and Territory Budget Papers.

TAXATION REVENUE FROM PROBLEM GAMBLERS

Some groups have suggested that governments receive a significant proportion of their gambling tax collections from individuals with gambling problems. ⁽³⁷⁾  For example, the Australian Institute of Criminology has claimed that "problem gamblers are the main profit generators for the gambling industry and, hence, contribute revenue to governments", ⁽³⁸⁾ while VCOSS suggested that nearly half of net gaming machine revenue comes from gambling by problem gamblers. ⁽³⁹⁾

At the outset, it must be stated that it is difficult, if not impossible, to obtain a direct evidence base for the amount of taxes paid by individuals.  For most gamblers, the structure and rates of gambling taxes that they pay would be largely unknown.  Taxes tend to be are based variously on the gross revenue, profit or losses attained from the games available.  Taxes vary by state, and in accordance with the actual games.

The consequent inability of most, if not all, individual gambling taxpayers to accurately recount their tax liabilities, or estimate their expected future liabilities, is known in the public finance literature as "fiscal illusion".  Indeed, it could be reasonably argued that governments would have an incentive to reduce the visibility of their taxes to the gambling public.

Furthermore, due to the need to respect the confidentiality of individual gamblers, as well as the prohibitive administrative costs associated with tracking actual spending with individuals, it is not possible to directly allocate tax shares to individual (problem and recreational) gamblers.

Researchers have sought to indirectly estimate the share of state and territory gambling taxes borne by groups of gamblers.  The Productivity Commission's 1999 gambling inquiry estimated that 33 per cent of all gambling expenditure was attributable to problem gamblers.  It then uses this to estimate that problem gamblers raised $1.4 billion in taxation (or 33 per cent of the total tax share) for state governments in 1997-98.

However, there are a number of issues about the validity of these estimates.  The Commission scaled up the gambling expenditure figures obtained from a survey, based on its judgment that reported spending by gamblers were likely to be understated.  While it mentioned issues such as problem gamblers concealing their magnitude of spending, as well as potential "recall biases" of gamblers regarding their actual losses, little detail was given to explain the basis for the adjustments it made to reported gambler spending.

Other questions were also raised about the Commission's results, such as the validity of extrapolating the results from a low sample (approximately 4,000 people) across the population.

To be sure, the amount of revenue transferred from problem gamblers to governments is non-zero.  However, the expenditure shares estimated by the Commission -- a basis upon which to (indirectly) impute tax burdens -- are statistically crude and should not be relied upon until more authoritative results are made available sometime in the future.

POTENTIAL IMPACTS OF GAMBLING TAX CHANGES

A number of anti-gambling groups have suggested that taxes on gambling be significantly raised to deter problem gamblers from consuming gambling products.  For example, the independent Senator Steve Fielding has proposed a new commonwealth tax "on individual poker machines in pubs and clubs which gradually increases, eventually rendering them non-profitable". ⁽⁴⁰⁾

The Centre for International Economics has recently modelled the impact of additional gambling taxes similar to that proposed by Senator Fielding.  In the short run, economic output is expected to contract by $6.6 billion (or 0.6 per cent), with consumption expected to reduce by one per cent.  Employment in the gambling, and associated, industries would fall by some 74,000 people.  Most of these negative impacts are expected to flow through into the long run.

Of the regional economies, both NSW and the ACT would be affected disproportionately by higher gambling taxes.  Output in both economies would be expected to decline by about one per cent.

In practice, the gambling industries are already highly taxed relative to most other sectors of the economy.  Not only are they subject to sector-specific taxes on gambling products, but they are also liable to pay general business taxes such as payroll tax and stamp duties.

According to the Australasian Gaming Council 2008-09 Database, hotels and clubs with gambling facilities paid over $2 billion in gambling taxes and $225 million in other commonwealth, state and local taxes in 2004-05.  In 2006-07, casino operators paid $403 million in gaming-specific taxes and $604 million in commonwealth, state and local non-gaming taxation.

As noted above, the intention of the tax policy put forward by Senator Fielding is to render the use of poker machines as an unviable business proposition.  This will not only cause a rationalisation of economic activities in pubs, clubs, hotels and casinos, but the eventual elimination of gambling as a tax revenue source will encourage states to increase other taxes.  This eventuality is highly likely to impose its own distortionary effects on the market economy.

Therefore, any increases in gambling taxation are likely to have a deleterious impact on the viability of gambling operators, particularly in the current environment of a slowing national economy, and will reduce the economic welfare of the gambling population.

The equity implications of an increase in gambling taxes should also be considered.  It is well known that the incidence of taxes on gambling tend to fall particularly heavily on poorer people.  This outcome is reflective of the fact that, as a share of income, individuals on lower incomes spend more on gambling products.  Therefore, an increase in gambling taxation will have a disproportionate impact on those less well-off in society.

CONCLUSION

In his seminal work, Capitalism, Socialism and Democracy, Joseph A. Schumpeter described how economic growth is spurred by the introduction of new products and processes:

"The essential point to grasp is that in dealing with capitalism we are dealing with an evolutionary process ... The fundamental impulse that sets and keeps the capitalist engine in motion comes from the new consumer goods, the new methods of production, or transportation, the new forms of industrial organization that capitalist enterprise creates". ⁽⁴¹⁾

The policy deregulation of the gambling sector since the 1980s is consistent with Schumpetarian growth, as it encouraged new forms of gambling and the expansion of existing services.  As described in detail above, consumers and businesses alike have both largely enjoyed the fruits of these policies.

Governments have at the same time seen the gambling industries as an instant source of revenue.  They have also sought to acquire economic rents through such measures as auctions for gambling licenses and, in the case of Western Australia, giving exclusivity to its sole casino to operate poker machines.  While these trends would have the effect of increasing the effective price of gambling, until the last half of the last decade individuals were generally accorded greater freedoms to gamble.

Over time, concerns over the social consequences of gambling have occupied the minds of public sector policymakers.  Partly as a consequence of the Productivity Commission 1999 inquiry into gambling, as well as the lobbying efforts of groups opposed to gambling activities, a crescendo of government interventions has been aimed at ameliorating the incidence of problem gambling in particular.

The available evidence, even if hampered by methodological and statistical limitations, all points to problem gambling afflicting a very small, and possibly declining, proportion of the population.  However, government measures appear to lack a sense of real perspective about the extent of the problem.  In particular, new regulations and tax measures effectively restrict the freedom of all gamblers -- problem and recreational -- from enjoying one of their favourite pastimes.

If such trends towards restrictive government actions in gambling continue then:  consumers risk seeing more of their welfare diluted;  businesses will have a lesser capacity to grow, invest and employ people;  and the capacity for bottom-up self-regulatory solutions to genuine problem gambling issues will be diminished.  There are no clear winners, except for greater numbers of government regulatory agents and the self-satisfaction of anti-gambling groups, from draconian intrusions into the gambling sector.

Therefore, in the present and for the future, the capacity of individuals to be free to gamble will be the benchmark against which government policies towards gambling consumers and providers will be judged.

---

ENDNOTES

1.  Richard Allsop, 2008, "Class and Casinos", Review 60 (4):  10-15.

2.  Australasian Gaming Council, 2008, A Database on Australia's Gambling Industry 2008/09, p. 4-14.

3.  In Western Australia, gaming machines are only permitted in its sole casino.

4.  Productivity Commission, 1999, Australia's Gambling Industries, Report No. 10, AusInfo, Canberra, p. 5.17.

5.  ACIL Tasman, 2006, The Economic Impact of Gambling on the Northern Territory, Final report to Charles Sturt University, June.

6.  ACIL Tasman, 2006, The Impact of Gaming on Ballarat Revised, Summary report, April.

7.  South Australian Centre for Economic Studies, 2008, Social and Economic Impact Study into Gambling in Tasmania, Report commissioned by the Tasmanian Department of Treasury and Finance, June.

8.  Hospitality clubs are clubs or associations for sporting, social or recreational purposes (such as bowls, football, surf lifesaving and RSL clubs) licensed to sell alcoholic beverages, and include clubs with licenses for gaming machines.

9.  This total provides information on the contribution to value added of businesses that provide gambling products and services.  It excludes the shares of value added from special TAB and lottery outlets, and is not adjusted for the share of gambling in value added.

10.  Australasian Gaming Council, 2008, A Database on Australia's Gambling Industry 2008/09, p. 118-119.

11.  Australasian Gaming Council, 2008, p. 117.

12.  Centre for International Economics, 2008, Gambling with policy:  The economic impacts of removing gaming machines from clubs and pubs, Briefing paper prepared for Gaming Technologies Association, November.

13.  Australasian Gaming Council, 2008, A Database on Australia's Gambling Industry 2008/09, p. 96-98.

14.  Australian Bureau of Statistics (ABS), various years, Clubs, Pubs, Taverns and Bars, Australia, cat. no. 8687.0.

15.  Australian Casino Association, 2008, The Australian Casino Industry Economic Report 2006-2007, April, p. 18.

16.  IBISWorld, 2008, Industry Report Pubs, Taverns and Bars in Australia, April.

17.  Australian Casino Association, 2008, The Australian Casino Industry Economic Report 2006-2007, p. 16.

18.  ABS, Average Weekly Earnings, Australia, cat. no. 6302.0.

19.  Geelong Catholic Social Justice Committee, 2005, Submission to Victorian Government Review on Regional Electronic Gaming Machine Caps, http://www.justice.vic.gov.au/wps/wcm/connect/DOJ+Internet/resources/file/eb9322439ec66be/Geelong_Catholic_SocialJustice_Committee.pdf.

20.  Productivity Commission, 1999, Australia's Gambling Industries, p. 19.

21.  Australian Hotels Association NSW, 2003, Supplementary Submission to the Independent Pricing and Regulatory Tribunal Gambling Harm Minimisation Review, November.

22.  M.W. Abbott, 2006, "Do EGMs and Problem Gambling go together like a Horse and Carriage?", http://www.nceta.flinders.edu.au/events/documents/AbbottInPress.pdf, p. 4.

23.  SA Centre for Economic Studies and University of Adelaide Department of Psychology, 2005, Problem Gambling and Harm:  Towards a National Definition, p. vi.

24.  In a report on the economic impact of gambling in the Northern Territory, ACIL Tasman also noted that "problem gamblers do not always experience a problem whenever they gamble.  Rather they only do so in certain circumstances and over time they tend to teach themselves how to anticipate and avoid the adverse circumstances".  ACIL Tasman, 2006, The Economic Impact of Gambling on the Northern Territory, Report for Charles Sturt University, p. x.

25.  The seminar paper explaining the rational addiction theory is Gary Becker and Kevin Murphy, 1988, "A Theory of Rational Addiction", Journal of Political Economy 96 (4):  675-700.

26.  Douglas M. Walker, 2007, The Economics of Casino Gambling, Springer, New York, p. 118-119.

27.  Peter Newell OAM, 2008, "Problem Gambling in Australia -- A Way Forward", June.

28.  M.W. Abbott, 2006, "Do EGMs and Problem Gambling go together like a Horse and Carriage?", p. 1.

29.  University of Queensland news online, 2001, "Problem gambling a complex issue, says UQ researcher", http://www.uq.edu.au/news/?article=2619.

30.  Howard J. Shaffer, 2005, "From disabling to enabling the public interest:  Natural transitions from gambling exposure to adaptation and self-regulation", Addiction 100 p. 1228.

31.  Max Abbott and Rachel Volberg, 2006, "Gambling Exposure and Problem Gambling:  Are they related?", Presentation to 13th International Conference on Gambling and Risk-taking, May.

32.  ACIL Tasman, 2006, The Impact of Gaming on Ballarat Revised, p. 9-10.

33.  Victoria has recently announced legislation to take effect from 2012 that will also remove ATMs from gambling venues.  In addition, Victoria, South Australia and the Northern Territory impose limits on the amount of cash that can be withdrawn from gambling venue ATMs, while Tasmania imposes an EFTPOS limit of one transaction for gaming each day.

34.  Jan McMillen, 2008, Submission to Senate Community Affairs Committee Inquiry into ATMs and Cash Facilities in Licensed Venues Bill 2008, p. 7.

35.  Revenue raising capacity is a ratio indicating the capacity of a jurisdiction to raise revenue relative to the Australian average.  It is measured by dividing assessed revenue per capita by average revenue per capita.  A ratio greater than 100 indicates that a state or territory has a capacity to raise revenue above the Australian average.  Revenue raising effort is a ratio indicating the actual effort made by a jurisdiction to raise revenue relative to the Australian average effort.  It is measured by dividing actual per capita revenue by assessed revenue per capita.  A ratio greater than 100 indicates that a state or territory is making an effort to raise revenue above the Australian average.

36.  Under the tax reform arrangements, the states agreed to adjust their gambling tax arrangements to take account of the impact of the GST on gambling operators from 1 July 2000.  This entailed a reduction in some gambling tax rates.

37.  In a similar vein, the Anglican Diocese of Melbourne Social Responsibilities Committee, "state and territory governments have become dependent on the tax revenue raised by poker machines, making it increasingly difficult for them to deal with problem gambling issues objectively".  Anglican Diocese of Melbourne Social Responsibilities Committee, 2008, Submission to the Senate Inquiry on the Poker Machine Harm Reduction Tax (Administration) Bill 2008, p. 1.

38.  Yuka Sakurai and Russell G. Smith, 2003, Gambling as a motivation for the commission of financial crime, Australian Institute of Criminology, Trends & Issues in Crime and Criminal Justice Paper No. 256.

39.  Victorian Council of Social Service, 2007, Submission to Victorian Legislative Council Select Committee on Gaming Licensing, p. 2.

40.  Senator Steve Fielding, 2008, Poker Machine Harm Reduction Tax (Administration) Bill, Second Reading Speech, 14 February.

41.  Joseph Allois Schumpeter, 1942, Capitalism, Socialism and Democracy, p. 82.