Draft Chapter in Bruce Curtis (ed), Gambling in New Zealand (Dunmore Press, 2002) Chapter 3, p.45-58.
Keywords Regulation & Taxation
While the practice in New Zealand is to define the gambling industry as that which the Department of Internal Affairs is involved with – including casinos, the TAB, gaming machines, the Lotteries Commission and a host of raffles and housie evenings – it is not immediately obvious why the share market or the insurance industry should be excluded.(1) Both involve the purchase of entitlements with uncertain outcomes, and each involves some of the regulatory problems of ensuring the purchase being honestly managed. Moreover, much of the gambling industry uses the terms such as ‘investment’ with the implication that, say, backing a geegee is not very different from investing in GG Corp. However this chapter focuses on the DIA defined industry, perhaps because it is seen as recreational compared to insurance and equity investment, although their regulatory lessons will not be ignored.
Another industry where there are some parallels is the alcoholic beverage industry, insofar as both gambling and drinking can generate socially damaging consequences. It would be foolish to waste the lessons learned from alcohol regulation.
The economic framework of this chapter arises from taking the price system as one of our social coordinating mechanisms. If the true social cost of any transaction is reflected in (proportion to) the price involved, then the coordination system generates self-policing beneficial voluntary transactions, because a transaction will not occur unless the transactors each think each will be better off. In practice, the assumption of price reflecting social cost appears to apply for most transactions, but there are exceptions – ‘externalities’ – where the transactors do not (or cannot) take into consideration all the costs. For instance a drinker may be unconcerned that driving in an inebriated state may cause death and destruction to others. Insofar as the price of the drinker’s alcohol does not reflect that cost there is an externality. Economic policy is often concerned with minimizing such externalities, although there are other tasks, such as ensuring the transactors know the true prices.
Gambling has been associated with sin. Secular economics is inclined to abjure such judgements, leaving them – in economist’s Denis Robertson’s famous retort – to the Archbishop of Canterbury.
A variation is that gambling is essentially an economically regressive, that is it increases the inequality of wealth by transferring it from the whole population to a few winners. This is something economists should be able to address, albeit without necessarily imposing their own personal judgements on what is an acceptable wealth distribution. Capitalism involves regressive processes, in that it depends upon, and creates opportunities for, people attempting to improve their lot. Thus, Bill Gates becoming the richest man in the world was economically regressive. However a central feature of this process is that in doing so, it involves new products or cheaper processes which benefit others – in Microsoft’s case the benefits may be spread over perhaps of others. Most defenders of capitalism would say that the benefits from the innovation outweigh any wealth regressivity. Many would say that the evidence is that most often the overall outcome is a reduction in inequality – for income anyway – as competition shares the benefits of the innovation through the entire economy.
While there may be some – probably minor – offsetting gains from workers obtaining jobs in the gambling industry, generally gambling has these wealth regressive effects, without the innovation that goes with the just described capitalist process. (We should not make too great a distinction though. Keynes likened stock exchanges to a casino and, noting that it is ‘usually agreed’ that the latter ‘should, in the public interest, be inaccessible and expensive’, thought ‘perhaps the same is true of Stock Exchanges.’(2) Ambrose Bierce commented that ‘the gambling known as business looks with austere disfavour upon the business known as gambling.’(3)) It is a matter of public policy whether the regressivity from gambling is a reason for prohibiting or restricting it. The chapter returns to this issue in regard to taxation.
In practice the economics discipline simply observes people gamble, usually for pleasure. Even so it acknowledges that insofar as there are unusual and large externalities, there may be a case for specific regulation, and wonders uneasily about the wealth regressivity of outcomes. Thus the approach is to examine the gambling industry, from an economist’s perspective, in much the same framework as for any other industry, although it will be necessary to look at the peculiarities arising from the regulatory interventions. It then looks at the externalities arising from ‘problem gamblers’, and finishes by exploring some of the regulatory issues which arise as a result, including that of tax.
The Gambling Industry
It may be useful for non-economists to begin with the distinction between ‘turnover’ and ‘expenditure’. This is necessary because some proportion of gamblers’ outlays are returned. There are no comprehensive figures but in the March 1998 year, turnover seems to have been around $7b of which around four fifths was returned to the punters. (Table 1) It is not logical to compare a turnover with the usual measure of economic activity, GDP, any more that it is sensible to make a similar comparison for the high turnovers of financial markets (many of whose transactions also cancel out), or to measure a supermarket’s performance in terms of the cash handed across the counter and ignore that which is it returned in change.
The convention is to describe the difference between the punters’ outlays and their prizes as expenditure. This corresponds sufficiently closely to the National Account’s measure of GDP to allow the comparison that in 1998/9 expenditure on gambling in New Zealand was around 1.2 percent of GDP, or around 2.0 percent of Consumer Expenditure. It is not a big industry. The standard national accounts ‘contribution to GDP by industry’ table lists 30 industries. Were gambling to be separated out – instead of being in the cultural and recreational services industry – it would have been 27th.
Industry growth seems to have largely been due to new products, typically in new establishments. The nominal expenditure on ‘racing’ betting has hardly changed over the last decade, despite the introduction of betting on other sports. Nor has the expenditure on the New Zealand Lottery Commission products grown much over the period (after a boom in the late 1980s following the introduction of lotto). Growth in gambling outlays has been on gaming machines introduced in 1990 and casinos introduced in 1994. At least some of that growth of outlays on the new products may have displaced the more traditional forms of gambling. However, some may have come from increased tourist outlays, or from New Zealanders using domestic outlets rather than going overseas. (There seems to be no information on internet gambling, which technically occurs offshore). But assuredly a major source of the growth has been the attraction of new products.
It is easy to argue that this growth represented an economic benefit, problem gambling aside, because of the jobs and economic activity created. Economics is cautious about the magnitude of the gain, because the additional expenditures on gambling usually reflect less spending elsewhere in the economy and hence fewer jobs and less economic activity there. Insofar as some of the additional gambling is by foreign tourists (and especially when gambling opportunities in New Zealand are a reason for their coming to New Zealand), or where it results in New Zealanders staying here rather than go overseas to gamble, there are net gains to the economy. However, almost certainly the main gain is the additional pleasure New Zealanders obtain from the opportunities that arise from the new products. The economics discipline tends to be relatively unreflective about the exact source of the pleasure. The gambling may be incidental, as when one goes out with one’s mates for a flutter at housie or a casino. It may enhance some the pleasure on another activity, such as following horse racing. Or the individual may obtain greater pleasure risking some money on a gaming machine than spending it, say, at McDonalds. Economics simply acknowledges that people generally deem themselves better off because they take advantage of the new gambling products.
The gambling industry operates under specific statutes (as well as the general law), the effect of which is often to create barriers to entry for new businesses and impediments to innovation. Economists have an a priori expectation that the barriers to entry would discourage efficiency and innovation, both of which would reduce consumer welfare. However that has to be proven, and the alternative of an unregulated industry might be more inefficient and less welfare inducing than one with appropriate regulation. The reasons, for and consequences of, this regulation is the central concern of this chapter. Were there not such regulation, or need for it, the gambling industry would be one of numerous small industries which make a contribution to economic welfare by employing workers, resources and savings, and providing people with activities they appreciate.
The Economics of Problem Gambling
According to the 1999 New Zealand Gambling Survey, around 1 percent of New Zealanders were then problem gamblers (and they accounted for 20 percent of reported gambling expenditure, so each spends about 25 times as much as the average for the rest of the public). Other chapters address the details of this addiction. This chapter merely accept that there is an economic problem, in the sense that some people’s gambling behaviour generates economic externalities. The problem of such externalities has been best explored in the case of alcohol misuse, where around 5 percent of the population are deemed problem drinkers (drinking about 40 percent of the total alcohol consumed, or about five times as much as the average for the rest of the public).
In drawing parallels between problem drinking and problem gambling, the intention is not to suggest they are identical, nor that gambling addicts cause the same level of social and health havoc as alcoholics. However it is useful to reflect on the various ways that the community has tried to deal with the drink problem. In particular until the late 1980s, industry control was seen as the main way of dealing with alcohol problems. Since then there has been a switch to great emphasis on the normalcy of drinking, while policy instruments have been increasingly targeted on prevention of pathological behaviour, of both individuals and social groups. In the jargon this is a shift from ‘supply-side’ controlling of the detrimental activity, to ‘demand-side’ controls.
Yet although alcohol industry regulation before 1990 focussed on prevention of pathological behaviour by restrictions on industry activities, there is little evidence that it was effective. Despite limiting access to alcohol by restricting outlets by location, and by hours of opening (not to mention a host of petty restrictions on what may be purchased from, or consumed at, an outlet) alcoholism was rife, and death and injury from heavy drinking evident.
Following the 1989 Sale of Liquor Act, which was largely aimed to regulate the liquor industry much like any other, and specifically states that the only reason for a specific restriction is to reduce liquor abuse, there has been no upsurge of drunkenness. (Indeed alcohol consumption has continued to fall, although this probably reflects long term trends in social behaviour and attitudes, changes in the tax regime, and successful public health promotion policies, rather than the legislative change per se. More recently there may have been a surge in teenage drunkenness, for reasons possibly attributable to wider liquor advertising or the lowering of the drinking age.) Of course the industry remains subject to the ordinary requirements of industry legislation (such as outlets requiring consents under the RMA) but generally the production and distribution side of the industry is as unregulated as any other. There are some requirements in regard to customers, including minimum drinking and purchasing ages, and that those who inebriated should not be served. There has also been a vigorous public education campaign targeting alcohol abuse (especially at prone groups), rigorous measures to curb drunk driving, while the taxation has been raised and rationalised.
The point here is not to argue that exactly the same policies should apply to gambling (and it is certainly not to imply that gambling addicts should be treated with alcohol and drug addicts). The point is that a shift similar to that which occurred in the alcoholic beverages industry is occurring in the gambling industry, although perhaps less coherently because there has been no community-industry based equivalent of the Alcohol Advisory Council of New Zealand. It seems likely this trend to increasingly treating the gambling industry much like any other will continue, although there are some production side issues peculiar to it.
In summary, while not denying there is a problem of pathological gambling, it is not obvious that this should be the sole focus of regulating the gambling industry. Excessive zeal at the expense of normal gambling will result in welfare losses to the population as a whole, through reduced opportunities and inefficient uses of resources.
Production Side Issues
There is an association between gambling outlets and criminals, which must come partly because of legal restrictions on an activity for which there is a popular demand. Even if these restrictions were minimized there would still be an attraction for criminal elements, in part because gambling provides an opportunity to launder illegal acquired funds, and also because of the opportunity to bias outcomes against the apparent odds.(4) New Zealand has dealt with criminal potential by a licensing regime which involves the licensees to not have a criminal record. This seems eminently sensible, for it is a form of quality licencing not a substantial barrier to entry. Similar provisions apply to the car hire industry, but there is no shortage of crime-free taxi drivers.
The issue of the possibility of a business organizing its gambling operation in order to make an ‘unfair’ profit by biassing outcomes to favour the business, is more complicated. It is an old problem. The life assurance industry required statutory regulation from the nineteenth century for a similar reason. It probably cannot be simply covered by existing legislation, such as the Fair Trading Act, because of the complexity of the transaction (although of course existing consumer legislation applies to gambling and there may be usefully extended). In recent years the general approach in financial markets has been to provide investors with as much information as seems reasonable and then leave them to judge. However requiring, say, a casino to provide each customer with a prospectus would seem clumsy, while the consumer information provisions of the Fair Trading Act may not be sufficiently robust. It is not always obvious what information a punter requires. In the case of fixed odds betting, the true odds may be all that punters need to know. But it is hard to identify a meaningful set of information for, say, a lotto transaction. The finance industry and government are currently struggling with the required information for consumer credit transactions. There is no single parameter which really reflects the ‘true cost of credit’, although it may be possible to legislate for one which gives the borrower a rough indication of the return they face. Much the same complexities apply to gambling where there are not fixed odds.
Were there a demand, there would be a case for providing the odds for other fixed odds transactions, such as in casinos or gaming machines and housie. The remainder of the industry would then be dependent upon the trust of the punters, as for racing and the lotteries commission activities, underpinned by an annual report and parliamentary scrutiny. Raffles would require some other provision. This suggests that consumer information is not a justification for restrictions or barriers to entry where there are clear fixed odds, although there may be a role for licensing operations to eliminate criminal activity, and perhaps a case for some prudential supervision to ensure that the offered odds are true, where the punter cannot verify them.
In which case what is the justification for restrictions such as on the number of casino outlets and the prohibition of private bookmaking? Undoubtedly some of the public opposition is exaggerated fears of problem gambling – a reversion to the past liquor strategy of attempting to control outlets as a resolution. One suspects some of the opposition comes from puritanism – H.L. Mencken described it as being haunted that someone, somewhere, at sometime may be having fun – an externality the economics profession tends to dismiss, but which remains politically important. There is also a widespread belief that a casino (in particular) damages local interest in ways that resource consents do not fully take into consideration – but that is an issue for reforming the RMA. (As also in the case of the liquor industry another source of litigation is incumbent competitors try to use the regulatory environment to restrict entry of competitors.)
Whatever the personal view an economist, the discipline draws attention to the consequences of the barriers to entry, which are likely to lower efficiency and slow product innovation. In addition, in the case of casinos those regions which have them may benefit from their presence via the attraction of New Zealand tourist from outside the regions the expense of other regions.
Where it is harder to provide the consumer with meaningful probabilities – such as the lotteries and totalisers – there may be a case for closer supervision of those activities. A government-owned monopoly is one such means of doing so (which also enables the government to appropriate any monopoly profits). Even so, the economist would worry about inefficiency and reluctance to innovate. A government monopoly would provide evidence to deny both worries, but the economist is likely to be happier with a market test.
Another issue is the gambling business which does not (or cannot) pay its debts – the myth includes the (illegal) bookmaker who skips when the long odds horse wins and the raffle whose prizes are eaten up in costs. These are but examples of business failure, as occurs in other industries, although the likelihood of it happening may be higher because of the gap between the payment (bet) and the return (winnings). It is not obvious that this should be managed in any other way than through the usual market arrangements of the supplier’s reputation, the criminal law, and caveat emptor. It may be, as in the case of the financial securities and insurance industries, some not overly restrictive requirements could add to the effectiveness of the market information flows in gambling transactions too.
In summary, there are peculiarities of (parts of) the production process of the gambling industry which may require regulation (outside the issue of problem gambling). However the oddities are not unique, for they are found in different combinations and to different extents in other industries. Their experience suggests that there may not be an ideal regulatory regime and the regulator ends up being faced with uncomfortable tradeoffs.
It is not obvious why the gambling industry should be subject to an industry specific taxation regime. (Of course, where applicable, the agency will pay generic taxes such as GST.) Even so, much of it has been. The explanation may be the historical notion of gambling being sinful, with the payment of taxation as a penance. But there is also the practical difficulty that most of the institutions providing gambling activities are not corporations (excepting casinos), and therefore do not pay taxation on profits via corporate taxes. Some tax may be to offset this loophole.
Whatever, the result has been a series of ad hoc gaming duties:
Totaliser duty is 20 percent of the betting profits of a racing club or totaliser agency board;
Lottery duty is 5.5 percent on all tickets (and so on) sold on behalf of the Lottery Commission;
Gaming machine duty is 20 percent on the profits of dutiable games;
Casino duty is 4 percent on the ‘casino win’.
Most gambling activities also make a ‘community contribution’.
– The TAB contributes to the costs of the racing industry and makes grants to other sports. (It could be argued that without these transfers the racing industry would collapse, to the detriment of the punter. Those who follow sport without punting are also beneficiaries of these payments.)
– More than a fifth of the outlays to Lotteries Commission products are distributed by the New Zealand Grants Board to community causes such as arts, sports, charitable and community causes.
– A smaller proportion of turnover from the outlays to private gambling businesses is disbursed in a similar manner.
Since the punter has no control over these grants, and since they are not a cost of running the activity, they may be thought of as a ‘voluntary’ tax. (Economic analysis might suggest that if there was open entry to the gambling industry, these grants would be would be driven down to near zero, so they might be thought of as an element of monopoly profit.) But there also seems to be an acceptance by punters that they should make a community contribution. This may seem irrational – or a penance for the guilt of sinful gambling – but this public willingness to pay a voluntary taxation reduces the dependence on compulsory taxation.
Taxing Gambling Per Se
Insofar as the taxation is on gambling per se, rather than as a substitute for a tax on operator profits, consider the specific taxation on alcohol. Again the justification for special taxation because of a view of the sinfulness of alcohol has been a part of the story but, practically, excise duties have been relatively easy to collect, because of the restricted numbers of points of wholesaling (although that is less true for wine, and there are leakages such as smuggling and home brewing). In these secular days sinfulness is less attractive as a justification for taxation. Today there are two broad reasons for taxation on alcohol.
First, it remains relatively easy to collect. (There is the added attraction that the deadweight loss
from the additional tax is relatively small, so there is not a loss of efficiency although that may still be equity issues. I propose not to expand this orthodox economic argument.) Given that all tax revenue is difficult to raise while conversely it is easy to spend on worthwhile projects, there is some merit is taxing something where the task as simple as it is for alcohol.
Second, the taxation is justified on the basis of the externalities that are involved with alcohol misuse. Some drinking involves social costs which are not captured in the private market cost. These include mortality and morbidity, and associated treatment costs. Importantly, the costs are not all borne by the drinker but by others both from general taxation and in personal costs of injury, death and heartache as well (as private property damage).
Imposing a specific tax on alcohol consumption operates in two ways. By signalling to the purchaser that the costs to the community are higher than the private costs of the production of the drink, excessive drinking is discouraged. Where it does not, the excise revenue contributes to community funds to cover (some of) the social costs incurred.
One major difficulty is that it is not possible to signal precisely the additional costs of drinking since that varies with circumstances. (Not all drinking involves alcohol misuse, and the same drinking situation can generate different levels of social cost – e.g. a driver and non-driver in a bar.) Thus the excise duty on alcohol is only a crude approximation as to true social cost. (Indeed the policy advice tends to be to align all the other policy instruments as efficiently as one can in order to minimize alcohol misuse, and then use an excise duty in an attempt to reduce it further.)
A second major difficulty is that it is not clear what should be the appropriate level of excise duty. In practice the revenue exceeds the known fiscal costs of alcohol misuse, but is substantially less than the estimated total social costs, including non-government costs. (While the argument in the previous sentence is put in terms of average costs, optimal policy uses marginal costs, which in the case of alcohol misuse are higher than average costs, implying that tax rates should be higher than what the averages would suggest).
Applying these principles to gambling is helpful, but still leaves many gaps. If the purpose of the special tax is to reduce externalities, there are no estimates of the social costs of gambling, so there are not even orders of magnitude. If the ‘sinfulness of gambling’ argument still predominates, in which case economics analysis provides little guidance.
Gaming Duties as a Profits Tax
An alternative approach might be to see gaming duties are proxies for the business taxation that non-corporation operators would pay were they businesses. (But casinos pay corporation taxes and a gaming duty.) If this were a concern, then it might be easier to require the various operators holding licences to set themselves up as corporations and tax them accordingly or, where that is impractical, impute to impute a tax as if they were corporations.
Gaming Duties as a Tax on Monopoly Profits
Another justification for gaming duties is that because of licensing restrictions on entry some operators may make monopoly profit. A tax pulls that profit back to the community. This shifts the question to why the monopoly in the first place. Even if that were justified, it might be more economically efficient to auction the right to the monopoly. (Possibly a levy rate on turnover may be the tender variable, the franchise going to the highest offer.)
Gaming Duties because Gambling is Wealth Regressive.
Suppose the justification for specifically taxing of gambling is to reduce its impact on the wealth distribution. That, presumably, would involve a rate which progresses by the size of the winnings. As far as I know there has been no attempt to tune the tax rate for such a purpose (although it is noticeable that the highest effective tax rates apply to the Lotteries Commission, which offers the greatest prizes). In any case, it may be that it would be more efficient to moderate any redistributive effects via a comprehensive lifetime capital transfers tax which would also cover capital gains, gifts and inheritance. There is a gifts tax, apparently as a mechanism to reduce income tax avoidance, but since there is no public enthusiasm for such a tax regime, it is not obvious that gambling should be separated out for such taxation treatment.
In summary there are a number of reasons which might justify a gaming duty. In fact there are too many, and until that is clearer which are the reasons for intervention, it is not obvious what the duty rate if any should be. It is not even obvious what should be taxed: turnover, expenditure, business profits (perhaps defined by analogy as if the operator was a corporation), or punters’ winnings? (A parallel puzzle in the finance sector, means that GST is still not levied on its output.) The danger is the wrong tax regime could generate the deadweight costs of economic inefficiency. For instance the tax could be so high it could discourage some forms of gambling altogether without, say, reducing gambling addiction (if that is the purpose of the tax).
A particular issue is whether the levy on gambling should be at the same rate and on the same base for all activities. Excise duties on alcohol are on absolute alcohol content (although there is still no alignment between the rates on spirits on the one hand, and beer and wine on the other). Of all the available practical candidates for a levy, absolute alcohol is probably the closest to proportional with social costs. An analogue for gambling is far from obvious. The existing hybrid probably reflects a historical mixture of ad hoc decisions and practical reality, rather than any rational overview.
(An issue of concern to Treasury is whether the various ‘community contributions’ should be absorbed in the overall tax, and the revenues distributed via the public purse. A particular anxiety is that these are defacto taxes which are not subject to standard constitutional arrangements. The advantage of the current regime is that there is a decentralisation of the disbursement, a useful feature of any government decision making and especially valuable in a democracy. On the other hand some funds may go to projects not which are not considered appropriate – donations to a political party are an often cited example – although this may be resolved by tighter legislation.)
In summary, there may well be a case for rationalisation of specific taxation on gambling, but until the purpose of that taxation (other than to raise revenue for public and community spending) is clear, the implementation of any rationalisation is likely to be confused.
Direct Demand Management
Gambling is unlike alcohol consumption in that while both have addicts, alcohol misuse includes nonaddictive behaviour such as binge drinking with the potential to damage others by violence or accidents. There is not a comparable public generated externality in binge gambling. Its costs are largely internalised to the individual (and her or his family).
The issue of gambling addiction is considered much more thoroughly in other chapters, but the economists’ point is to ask to what extent it is any more of a problem than, say, eating too much or loving too much. The point is that these may be proper issues for the health services to treat, but they are hardly matters for economic measures targeting eating and loving. There may be, of course, a case for education on eating, loving and gambling, but it is not obvious in economic terms that there is a need for any further industry specific measures.
There would be some justice in levying the industry to fund treating problem gambling and the public promotion of prudent gambling. (The liquor industry with its specific levy to fund the Alcohol Liquor Advisory Council provides a model, but it might be thought inconsistent with the constitutional principles of taxation, insofar as it is a tagged tax.).
In summary, it is not obvious that there is a need for an active program of demand management as there has been for alcohol (and tobacco), other than health services for the problem gamblers and some public education.
While some parts of the gambling industry were established many years ago, the economic liberalisation of the late 1980s has created new gambling activities, which have not been integrated with the older ones. As the change in alcohol policy at the time indicates, there was also a change in the view of social regulation of activities which had previously been treated as ‘sinful’. But the logic of this has not yet been systematised in relation to the gambling activities, nor has the logic of market principles been applied to ensure that there is maximum welfare from the resources used.
Among the difficulties the policy analyst faces are that a country’s gambling industries and activities reflect the particularities of its history and culture. Given the rapid changes that have occurred in the last decade, it is difficult to predict what (quasi-)rational policies are practical in the public agenda and which are likely to be adopted. In the interim policy advocacy will dominate, often based on far from fully articulated views or analysis and, even more often, on self interest.
Table 1: Licensed Gaming Activity: Turnover and Utilisation
Estimates for March year 1998 ($m or %)
Gaming Activity – Turnover – Winnings – Winnings
(% of turnover) – Gaming Duties Revenue – Tax (% of turnover)
Casino 3060 2815 92.0 10 0.3
Clubs & Pubs /Trusts 2600 2140 82.3 58 2.2
Housie 45 30 66.7 0 0.0
Lotteries Commission 639 351 54.9 35
(+CC =135) 5.4
TAB-Racing 999 809 81.0 37 3.7
TAB -Sports 53 46 86.8 0? 0.0
Other (raffles etc) 20 10 50.0 0 0.0
TOTAL 6916 5969 86.3 140 2.0
Sources: First two columns B – Curtis estimates. Column 4 – Treasury. (CC community contribution).
1. I am grateful to Bruce Curtis and John Yeabsley for comments on an earlier draft.
2. J. M. Keynes, The General Theory of Employment Interest and Money, 1936, p.159
3. A. Bierce, The Devil’s Dictionary, 1911.
4. Criminals may also have a(n illegal) comparative advantage in collecting gamblers’ debts