Gambling and the EU Internal Market: An in-depth Analysis
"Britain, like many other nations, is also facing a possible explosion in internet-based
gambling, particularly remote poker. Reliable statistics are hard to get, but the anecdotal
evidence one gets from around the world is hard to ignore. Explosion may indeed be the
right word..."
Speech given by Robert Young, Principal, Europe Economics, London, to a
conference on The Future of Gambling in the Internal Market organised by
the Academy of European Law, Trier, Germany, February 8-9 2007
Gambling in Britain
In Britain, the most frequently-used form of gambling is the National Lottery. This was launched in
1994, after a prohibition on lotteries which had lasted about 170 years. The National
Lottery proved to be immensely popular from the start, and brought a lot of people into
gambling for the first time, especially women and higher socio-economic groups. The
state-licensed lottery operator, Camelot, also runs virtually the entire British scratch card
market, with a share of about 96 per cent. Our next most popular forms of gambling are
bingo, which appeals mainly to women, and betting, which appeals mainly to men,
followed by fruit machines, which are everywhere.
We have a new Gambling Act on the statute books. Although passed in 2005 it does not
become fully operational until September 2007. At that point our new regulator, the
Gambling Commission, already appointed, acquires all its powers. The government’s
main aims in creating the new Act were to sweep away much old and outdated gambling
legislation, to create a new and fairer regulatory régime and, so far as possible, to keep
gambling free of under-age access, criminality, and problem gambling. It is hard to
argue against any of these aims.
There are other important aspects to the new Act, too. One is that it allows for the
licensing of 17 new casinos in Britain – 8 so-called small casinos, 8 large ones, and one
regional casino, also referred to as a resort casino or super-casino. As you may know,
at the end of last month, Manchester was recommended to the government by an
independent panel as the site of the first regional casino. As my firm advised
Manchester on its proposal, I am very pleased about this. However, the process of
approval is not quite complete: the Secretary of State has yet to bring before Parliament
orders which allow the licensing of the 17 new casinos, so there remains the possibility
of a challenge, whether in Parliament by MPs or through the courts by aggrieved losing
bidders.
Once the new Act is fully in force, the government would like UK operators of remote
gaming to bring these operations to Britain. Hitherto it has been illegal to offer remote
gaming (though not remote betting) on a British site, so large companies like Ladbrokes,
William Hill and Coral have had to operate their remote gaming offshore. They have
located in, for example, Gibraltar and Curaçao, where they enjoy very low tax rates. If
they bring remote gaming to Britain they face VAT at 17.5 per cent and corporation tax
at over 30 per cent, so it remains to be seen what they do.
Britain, like many other nations, is also facing a possible explosion in internet-based
gambling, particularly remote poker. Reliable statistics are hard to get, but the anecdotal
evidence one gets from around the world is hard to ignore. Explosion may indeed be the
right word.
Finally, while gambling opportunities and gambling participation increase, there is quite a
lot of public noise which objects to gambling. It is certainly audible in Parliament, from
the government’s own side as well as from the Opposition, and there is much to read in
the Press. One senses too that there is opposition among a sizeable minority of
ordinary citizens. The National Lottery is less well regarded than it was because of
suspicion that the “good causes” it helps to finance have become substitutes for things
that the government itself should do or are projects that the New Labour government’s
friends want. If large amounts of Lottery money are sucked into the 2012 Olympics in
London, resentment could worsen. Consistently with such mounting scepticism, some
suspicion is also falling on one of the government’s main reasons for the new Regional
Casino: the urban regeneration benefits it is intended to bring.
To round off the British gambling scene, you may like to know that we Brits spend more
on gambling than any other EU25 Member State. We stake over €80 billion on gambling
each year. This is about half what we spend on food, about the same as the
government spends on education, and much more than it spends on defence. Of the
€80 billion, just over 60 per cent goes on betting.2 Of course, amounts staked are not
the same thing as amounts lost, but the figures may still be of interest. I have some
Member State comparisons of net expenditure – stakes minus winnings – to give you in
a few moments.
Britain compared
Let me now cover some of the main differences between gambling in Britain and
gambling in other EU Member States. Our National Lottery is not state-owned (as it is in
a number of other Member States) but it is state-licensed, and it is of course licensed as
a monopoly. We also have a large betting sector, with some 8,500 betting shops. Most
of our betting is at fixed odds, with relatively little pari-mutuel (pool) betting, and in that
respect we differ from most of the rest of the EU except Ireland. You can bet on a very
wide variety of sporting events, of which horse-racing is still the most popular, though
betting on football is now overtaking greyhound racing into second place. Within the last
three to four years bookmakers have also done well out of fixed-odds betting terminals.
These are machines on which you can play a variety of virtual games, though roulette
accounts for well over 90 per cent of usage.
Let me also put some perspective on casinos. We have about 120 casinos in operation
now, but they have had to operate as clubs, and most of them are small.
Even the eight new “small” casinos to be licensed under the 2005 Act will, typically, be bigger. But,
conversely, the single Regional Casino that so worries the anti-gambling lobby will be
much smaller than any you might see in Las Vegas.
Gambling is not, of course, static. Information technology alone makes for change
bordering on turbulence. In the last few years it has brought British (and other) gamblers
fixed-odds betting terminals, on-line betting exchanges and remote gaming. IT changes
the economics of gambling supply and continues to open up new possibilities and new
expectations for gamblers. It may bring unwanted side effects too.
It is notoriously difficult for national governments to keep pace with technical change and
to devise regulatory interventions which contain adverse effects while not spoiling the
enjoyment of the majority – though one may wonder if, collectively, the EU will do any
better. That is one of the things we are here to discuss, of course.
I would summarise by saying that in Britain our attitudes towards gambling are a strange
mix of greed, Puritanism and laissez faire. And although, as we shall see shortly,
gambling preferences may differ between EU Member States, I suspect that the mix of
attitudes we show in Britain is not so unusual.
Who gambles?
Let me now start to say something about the socio-economics of gambling. Who
gambles? And, importantly in the minds of many, who gambles too much?
The first thing to say is that there are not enough good recent studies of gambling round
the world. The US is quite well documented, and, size for size, so are Canada,
Australia, New Zealand and South Africa. In the UK we have a substantial Gambling
Prevalence Study that dates from 2000, and a repeat of that is due this year.3 My own
firm has published two reports on betting terminals which also take into full account the
wider gambling context in Britain.4 The literature situation is getting better in the rest of
the EU too.
I know we have a number of speakers from the Swiss Institute of
Comparative Law here today and I also know that their huge study for the EU has
attracted a measure of criticism.5 But my own view is that it is an immensely valuable
report, and I congratulate them on it.
If we ask who gambles most in the EU25, then the answer in absolute terms,
unsurprisingly, is the Member States with the largest populations.
Britain tops the Big
Five, with Gross Gambling Revenues (GGR), that is to say amounts staked less money
returned to players, of €11 billion (a measure of net expenditure as mentioned a moment
ago). Germany, despite having the largest population, comes second at €8.4 billion,
then, in order, France at €7.6 billion, Italy at €6.2 billion and Spain at €4.9 billion.
But size of population does not have much to do with propensity to gamble. Only one of
the biggest five member states by population makes it into the top five for gambling
expenditure per capita – what I have called the “Happy Five” – and that is Britain, which
occupies fourth position. Ireland comes top by some way, at €279 per person, followed
by Finland at €239, Luxembourg at €194, Britain at €181, and Sweden at €176.6 All of
these except Britain have small to medium size populations among the EU25. I should
add that the figures for Luxembourg may be larger than shown, since the Swiss
Institute’s report, the source of these figures, indicates that the returns for Luxembourg
are probably not complete.
If we look at where the Big Five come in the EU25, then, measured by gambling per
capita, France and Spain are 9th equal, Italy 12th and Germany 13th.
Whether you look at the Big Five or the “Happy Five”, gambling preferences are not
uniform. Lotteries are fairly popular throughout but the mix of other things varies. Table
1 below shows the most favoured forms of gambling, measured by GGR, in each of the
Big Five and the “Happy Five”.
The spread of gambling expenditure per capita is even wider than might be implied by the statistics.
The lowest recorded gambling expenditure in the EU25 is that of Poland, at €11 per
annum per person, equivalent to less than 4 per cent that of Ireland; and the lowest in
the former EU15 is that of Belgium, at €65 per annum, under one quarter that of Ireland.
With such diversity of expenditure and preferences, one must question whether there is
any real prospect of creating a single market for all gambling in the EU. One might also
ask whether there would even be serious interest in creating a single gambling market.
Perhaps if internet-based gambling becomes explosively popular in all member states,
that will create a desire for common treatment – but of course internet-based gambling is
recognised as perhaps the hardest of all modes of gambling to regulate.
It is one thing to get data about money spent on gambling, or taxes raised from it, but it
is altogether more difficult to get details of who plays and who has problems with
gambling. Two quotes from the Swiss Institute’s EU study confirm this quite forcibly: As
the Institute states:
“There has…been no uniformity in the way the governments of EU Member States have
addressed the issue of negative social impacts, including problem gambling.” (Executive
Summary, p.xlii)
“…most EU Member States have neither carried out prevalence studies nor put into place
explicit strategies for developing a greater understanding of the causal or contributing
factors to problem and pathological gambling within their borders.” (p.1443)
To try to develop some statistics I will give you some information about Britain. I do not
claim that our data will be typical of other Member States, but they may serve to show
you what economists think is important to capture.
We have first to identify which gamblers we want to study.
The earlier British study I
referred to, the Gambling Prevalence Study published in 2000, considered two groups –
those who gambled at all during the past week and those who gambled during the past
twelve months. In the work that Europe Economics were commissioned to do, the focus
was rather different, and we researched those who gambled at least once per month. A
further difference was that for our clients we were interested only in people aged 18 or
over, whereas the Prevalence Study considered those aged 16 or over.7 There can be
big differences arising from this.
For the Europe Economics studies we used large sample sizes – 12,000 in each of two
years, compared with just under 7,000 in the Prevalence Study. We found that almost
half of British adults, 42 per cent, gamble regularly, and a very large proportion of those,
89 per cent, regularly play the National Lottery. Scratch cards come next, with 12 per
cent of adults buying them. That is not surprising because you buy scratch cards at
exactly the same place as lottery tickets, and they come from the same operator, who no
doubt has a great deal of information about the purchasing practices of customers.
After
that, betting and bingo attract approximately equal percentages of the population, at 8
per cent, mostly women, for bingo and 7 per cent, mostly men, for betting. Five percent
play fruit machines. Casinos in their present form in Britain are very much a minority
interest, with under 1 per cent of adults visiting them.
Problem gambling
Problem gambling has been defined in a number of different ways. Two American
psychologists, Lesieur and Blume, have described it as “gambling to a degree that
compromises, disrupts or damages family, personal or recreational pursuits”, and this is
the definition we adopted for our own research.
There are two very well-established
psychological tests for identifying problem gamblers, namely the DSM-IV test and the
South Oaks Gambling Screen (SOGS). Of these two we chose the DSM-IV for two
reasons. It is shorter for use in mass surveys – 10 questions rather than 20 – and
psychology experts say it gives more reliable results in mass surveys than the SOGS.
However, new research is going on. Some commentators already favour the new
Canadian Problem Gambling Index, and a DSM-V test is due to become available in
2011.
If we are involved in future studies we will certainly take advice on the possibility
of using the newer tests.
We found in the 2005 research an overall prevalence of problem gambling in Britain of
0.4 per cent, somewhat lower than the 0.6 per cent that the Prevalence Study had found.
But there are two caveats to apply. First, the 95 per cent confidence interval applicable
to our study would give a range of 0.2 to 0.6 per cent, so gambling prevalence might or
might not be lower than in the earlier study. Secondly, we did not obtain our sample in
the same way as the researchers did in 1999. Nevertheless we still feel confident in
saying that levels of problem gambling probably did not increase in the five years
between the two sets of studies. Thus, it does not necessarily follow that problem
gambling always rises irresistibly, nor that problem gambling increases with new
gambling opportunities. Our calculation is that there are between 250,000 and 300,000
problem gamblers in Great Britain.
Now let me come back to the Prevalence Study, to look at what forms of gambling might
– and I emphasise might – be associated with problem gambling. In that study, casino
gambling emerged as the most problematical, since 5.6 per cent of casino users were
found to be problem gamblers. Conversely, the National Lottery appeared to be the
least problematical, since only 0.7 per cent of lottery players were found to be problem
gamblers.
But if you look at the absolute numbers rather than the percentages, the
picture is very different. We calculate that there are about 60,000 to 70,000 problem
gamblers who visit casinos compared with 250,000 to 300,000 who play the National
Lottery. These absolute numbers ought to matter to policy-makers. Maybe it is time to
look harder at problem gamblers for whom the Lottery is prominent in their gambling
habits.
However, one obvious conundrum about comparing, say, casino problem gamblers and
Lottery problem gamblers is that they could be the same people (or, strictly speaking,
overlapping groups). The British Gambling Prevalence Study and our own research
show that problem gamblers generally participate in multiple forms of gambling, so it is
very likely that they do overlap. So if, say, all casinos were closed or the Lottery
abolished, would the rate of problem gambling fall? Probably not: the addiction would
predictably be transferred to one or more other forms of gambling. And if all gambling
were banned, then, so psychologist friends tell me, the addiction would not be eliminated
but might well be transformed into some other form of addiction.
In any event we at Europe Economics have doubts about whether any particular form of
gambling either causes or aggravates problem gambling. The logit modelling we did in
the second of our studies suggests that two socio-demographic features – the age at
which a person starts to gamble, and marital status – are the two factors most positively
correlated with the presence of problem gambling. In the case of betting shop
customers, the frequency of visits to the betting shop appeared to be more positively
correlated than any particular betting pursuit available in the betting shop.
Indeed, no
particular form of gambling showed any correlation with problem gambling.
If we are right in concluding that the form of gambling is in effect immaterial, it makes
sense to concentrate problem gambling research on problem gamblers themselves. Our
own government has never funded or officially encouraged research of this kind. We
had only a resounding silence when we suggested it in our second report.
Gambling and governments
I move on, then, to gambling and governments, which marks the start of what I have to
say on the micro-economics of gambling. Here I see three major issues to talk about –
purely economic issues, then moral issues and public health issues. These last two both
have economic consequences.
On balance we can say that, for governments, gambling both is and is not an economic
activity like any other. Generally speaking, they want the gambling sector to thrive, and
some governments actively promote national gambling interests against competing
national interests, and in those respects EU governments appear to regard gambling like
any other industry.
On the other hand, in some respects they also bend the rules for
gambling: they sanction lottery monopolies, they devise special tax deals, and they
allocate percentages of gambling revenues to specific objectives. In these respects
gambling is unlike other industrial sectors.
The tax picture in particular is very varied, as the Swiss Institute’s comparative study has
confirmed. Some quotes from p.1057 serve to illustrate:
“[tax is]…sometimes expressed as a percentage rate and sometimes as a fixed sum of
money payable annually or monthly.”
“…the percentage rate may be referable to…net revenue…[or] gross revenue…[or]
winnings.”
“…complications are introduced by national legislation in the form of reduced rates or
exemptions from tax liability.”
Is the economic significance of gambling tax very great? Other forms of tax generate far
higher revenues, in both absolute and relative terms. In Britain, gambling duties (as distinct from VAT and corporation tax) come to only around 1 per cent of all duties, and
gambling tax is a tiny percentage of all tax.8
On the other hand, tax régimes are, as I have said, sometimes used as a basis of
competition between rival jurisdictions aiming to attract gambling operators. When some
countries offer zero or close to zero rates (for example, Curaçao and the Mohawk
Territory of Kahnawake offer zero per cent, Malta 0.5 per cent and Gibraltar 1 per cent)
you understand why some operators choose to base their operations there.
Where tax is levied on gamblers, tax rates can affect their behaviour too. If taxes are
levied at too high a rate on gamblers themselves, it depresses demand. But when tax is
taken off the gambler and put on the supplier, the results can be electrifying. This is
what we saw in Britain, with a change in taxes on betting in 2001. Before then the
régime imposed levies on bettors that, at their highest, required bookmakers to deduct 9
per cent from amounts staked. In 2001 the tax on bettors was removed and tax at 15
per cent was imposed instead on the gross profits of bookmakers. The result was that
between 2002 and 2004 betting volumes tripled.
One last point on tax. Our own government believes that UK operators who currently
have to run remote gaming offshore will come back to the UK when they can lawfully do
so after September 2007 because British regulation is so much better than elsewhere.
Should we really believe that tax is a price worth paying for British regulation? Why
should countries with low tax régimes not develop good regulatory régimes? We British
used to believe that nobody could make ships or motor bikes as well as we did, and I
wonder what went wrong there.
If I try to sum up for the EU as a whole, it must surely be true that a necessary precondition
for a single market is a harmonised tax régime? And that a necessary
precondition for that would be a harmonised tax rationale? By harmonised I do not
mean necessarily identical tax rates, but I do mean consistent as to the objectives and
basis of taxation.
As regards moral stance towards gambling, it is hard to see consistency among
governments. In some Member States gambling is prohibited except where allowed and
in others allowed except where prohibited. Consumer welfare is arguably an important
component of moral stance, and on this point I refer you to an excellent paper by Dr.
David Forrest of the University of Salford on the apparent disregard that our own
government has shown towards consumer welfare in its treatment of gambling regulation
– the tendency it shows to attach more importance to the ½ per cent or so who are
problem gamblers than to the 99½ per cent who are not.9 And I am impressed by the
arguments run by Charles Whitebread, Professor of Law at the University of Southern California, to the effect that when governments get up on a moral high horse and start
regulating from there, the effects are uneven, and often bear hardest upon the least
influential, least articulate members of society.10
Next I turn to problem gambling as a public health issue. Here again, there is a shortage
of good research material conducted or commissioned by governments. Within the
EU25 it looks as though only Spain and Sweden recognise problem gambling as a public
health issue which deserves attention by their national health services. The Swiss
Institute records that:
“In some Spanish regions, problem gambling specialists are routinely included in the local
Mental Health Departments” (p.1338)
and that:
“the Swedish National Institute of Public Health (Folkhälsoinstitutet) proposes various
measures to control gaming addiction”. (p.1358)
All other EU Member State governments rely on third party charitable organisations to
research and treat gambling addiction.
That said, it is worth remembering that gambling addiction – for the moment at any rate
– appears to be a problem of a much smaller scale than, say, alcohol or drug addiction.
I have only UK numbers to show you, and they suggest that compared with about
300,000 problem gamblers in Great Britain we have probably twice as many drug
addicts11 and at least ten times as many (3 to 4 million people) who are alcoholdependent.
12 These are not detailed statistics, but you get the general picture. The only
defence I can see for what the majority of governments do is that, if problem gambling
were formally made a public health issue, the effects of the marginal burden falling on
already stretched national health services could be severe.
Monopolies and the market
The Swiss Institute reports that EU governments give a variety of reasons for wanting to
own or directly control their own national lotteries and to organise them as monopolies:
for example, the preservation of public order, protection against gambling addiction, and
the prevention of fraud and money laundering. The good causes element is referred to
by some, though this is not an obviously compelling reason for state control. In Britain at
least, good causes come low down on the reasons gamblers give for playing the Lottery
– under 1 per cent in our 2005 study. In round terms, government reasons for retaining
complete control of lotteries can be reduced to a simple assertion – “we know best” –
yet it is not at all clear why “we” do.
regards economic structure, the traditional approach of governments has been that in
any one state – however we define “state” – there should be only one state-scale lottery.
This is because the normal economic aim imposed on a lottery operator is to maximise
revenue, i.e. ticket sales, and conventional wisdom is that only big prizes attract big
sales and that a monopoly lottery will produce bigger jackpot prizes than competing
lotteries. Apart from this, there is no obvious natural monopoly element about a lottery –
lotteries are not like gas pipelines or electricity grids, where economic efficiency would
fall with duplication. So in lotteries as in other sectors you may well get greater
innovation and efficiency under competition.
The problem, however, is that nobody has looked seriously at this. Again I am indebted
to David Forrest for suggesting that a good starting point would be more research into
the way lotteries actually work.13 He has pointed out that, although big prizes are
assumed to attract big ticket sales, the UK’s Lotto Extra (which offered only one big
prize) failed, possibly because it produced too few winners. Since lottery players
generally play multiple lottery games, Dr. Forrest argues that we ought to know more
about how the games interact with each other, and about the impacts that different prize
structures would have. Such research might shed new light on whether monopoly is the
right economic structure for maximising lottery revenue.
On the whole, one sees more innovation, greater cost efficiency and better value for
money in competitive forms of gambling than in state-licensed monopolies. Lottery
returns to players are generally a good deal lower than in any other form of gambling. In
Britain, about 45 per cent of money staked on the National Lottery is returned to players.
In betting, the figure is closer to 80 per cent, and with fixed-odds betting terminals 97 per
cent.
But even in forms of gambling subject to competition, governments intervene as and
when they think they should. Let me give you an example. When fixed-odds betting
terminals started to become popular in Britain, the government seriously contemplated a
ban on them because one Minister had seen what he regarded as similar machines
(“pokey” machines) being addictively played in Australia. No matter that the machine
specifications as between Australia and Britain were different, or that Australia allows
gambling and alcohol to be mixed in a way that Britain does not, the government’s kneejerk
reaction here was that fixed-odds betting terminals had to be stopped before they
caused a flood of gambling proliferation and addiction.
After a flurry of exchanges
between the government and bookmakers calmer counsels prevailed and systematic
research was carried out (by Europe Economics). Yet even now, after five years, and
with a Code of Practice in place, betting terminals are still regarded as effectively on
probation by a government that has nevertheless happily sanctioned 1,250 unlimited
prize slot machines in Britain’s first Regional Casino.
If there is any good news to point to it is that governments are perhaps – and I stress
perhaps – becoming just a bit less sure about their traditional assumptions of moral
rectitude. New attitudes towards alcohol licensing have resulted in regulatory changes that narrow the long-standing gap between UK practice and other EU Member States.
Against such a background it is likely that attitudes towards gambling will continue to
change too, a process that in Britain began when the National Lottery proved to be so
popular. From a consumer welfare point of view, the open-minded re-examination of
gambling, even if it results in no changes, can only be a good thing.
Conclusions?
To conclude, what would an economist say about gambling and the internal market? At
this stage, not very much. I have no nuclear flash of new insight to offer you.
EU governments would surely have to begin by agreeing what gambling is and what role
it should play in our society.. But if they could agree on what role in society is served by
gambling, they would next need to agree on whether a single gambling market is
feasible, and whether it should be made to apply across all gambling forms.
It might be
easier with some than with others, so maybe an acceptance of gradual progress would
be wiser than an expectation of instant success.
And even if the EU could devise some form of internal harmony what about the border
between gambling in the EU and gambling in the rest of the world? Does it need to be
policed, can it be policed, and if so how?
|
Blog 
