The analysis was commissioned by a “special investigator” whom the government has appointed to recommend new online gambling regulations to be included in the law.
A quick search of the report, entitled Licensing system for online gambling: Which tax-rate yields both high channelization and high tax revenues?, shows no mention of online poker.
The report is extremely thorough given the limited amount of data available, but has been structured at such a high level that the impact of taxes on individual gambling activities like online poker has not been considered.
In this case, the report’s authors have ignored how online gambling taxes have a bigger impact on online poker than on other forms of gambling.
Should the Swedish government accept the recommendations of the report as it stands, the tax rate will produce a situation where online poker players will likely still turn to unlicensed operators in large numbers.
If U.S. legislators considering an exclusive online poker model, such as that proposed in California, use the report’s recommendations in determining what tax level is correct, then they will not achieve their public policy objectives of consumer protection.
The report can be read in English here.
Copenhagen Economics was given the task of analyzing gambling taxes and their effect on state revenues as well as the proportion of those revenues generated from the regulated market.
The report was designed “to provide insights into which tax-rate would yield high channeling rates, meaning that a large proportion of Swedish online gambling is done within the forthcoming licensing system, as well as high tax revenues.”
The comparative data they present from other European regulated markets is informative.
The difference between the percentage of gaming that takes place in the regulated sector and that which takes place in the black or “gray” market is startling.
The range is surprisingly wide, from a 15 percent tax rate in the U.K., which leads to 95 percent of play taking place on licensed sites to a 45 percent tax rate in France which results in only 52 percent of play taking place at licensed sites.
The figures for Denmark, which manages to get 88 percent of players to play on its licensed sites are correctly praised in the report.
Denmark has a minimally intrusive and comparatively cheap regulatory regime. It allows a wide range of games, and for online poker, players play in the global player pool rather than being segregated into a national pool.
But, and this is a big “but” that Copenhagen Economics has missed, the actual figures from the Danish Online Gambling Association (DOGA) show that 27 percent of online poker players have played on unlicensed sites.
One other notable finding by DOGA is that:
“The average monthly spend among persons who bet with unlicensed operators is 5 times higher.”
These are players that the government should want to get inside the licensed system both for their higher tax contribution and because they may have more need of consumer protection.
The overall success of the system is boosted to 88 percent (as opposed to 73 percent for online poker) because in other activities such as sports betting, there is a much higher percentage of play at regulated sites—94 percent of sports betting is at licensed sites.
The difference between channeling 94 percent of players to licensed sites and 73 percent is largely down to the punitive impact that gaming taxes have on online poker.
The danger of using a revenue based analysis is that it doesn’t capture the reality of the numbers of people who are affected.
Online poker notoriously produces lower gross gaming revenues per player than sports betting, casino games or online slots. It operates on lower margins.
When treated only in terms of revenues, online poker represents a smaller slice of the total online gambling pie. But when active player numbers are counted, poker can be seen as a much more popular activity.
A gambling tax rate as recommended by Copenhagen Economics will bring the maximum amount of revenue into the licensed sector, but it won’t bring the maximum amount of players.
As Copenhagen Economics points out, online gambling obeys the laws of supply and demand, where increases in price reduce demand.
To an extent this applies to other gambling activities too, but poker players behave as if they are more price sensitive that those in other areas. There are several explanations for this, but there are also mitigating factors.
The lower margins in online poker mean that operators cannot leave their prices the same and accept lower profits. Some if not all of the taxes are passed on to the players either through higher rake, or lower VIP benefits.
In countries like the U.K., where players play in the global player pool, players generally receive lower VIP benefits (paywall) than those at the dot-com or dot-EU sites. In France, Italy and Spain, where the player pool is segregated, players pay higher rake as well as receiving lower benefits.
The big operators like PokerStars know that they can increase rake by a certain amount and fully understand the likely impact on their player pool. The vast majority of players carry on playing, or perhaps play slightly less, but on the whole they don’t desert the site.
Rake and VIP benefits at a sophisticated operator like PokerStars are not just whims of the week, or plucked as random figures from the air. The company has a vast amount of data and experience on which it can construct a balanced offer of benefits and rake to create its desired poker ecology.
Assuming that PokerStars has determined the optimal rake and benefit package for its dot-com customers, then that is the optimal marketing position for the company. If new gaming taxes mean rake increases or benefit reductions, then that delicate balance is upset. It is inevitably sub-optimal.
Higher rake affects cash games proportionately more than tournaments.
In a normal tournament players pay a fee to enter that is around 10 percent of the total buy-in. If the poker operator increases rake by 10 percent to offset higher gaming taxes—which is roughly what happened in the French market, the impact is not devastating.
For example in a $100 tournament, $9 goes towards the tournament fee with the remaining $91 going into the prize pool. If the rake changes so that $10 goes to the fee with $90 going to the prize pool, the impact on player behavior is small.
In cash games the impact of a similar rake increase, say from 5 percent to 5.5 percent can be seen if the change is presented in terms of big blinds per hundred hands played (BB/100).
An increase of only half a percent makes a difference of approximately 1 BB/100 to the total amount of rake paid. For example, at a cash game on PokerStars.com, the rake at stakes of 50c/$1 amounts to around 4.5 BB/100—figures calculated by Rakeback.com.
At PokerStars.it where high taxes have led to increased rake, the rake adds up to 5.5 BB/100.
It is no surprise that cash game revenues in the high tax segregated markets of France, Italy and Spain have declined much faster than online tournament revenues.
This has been the case even taking into account the impact of lottery style sit and gos such as the PokerStars Spin & Go. The decline was evident well before (paywall) the new games were introduced.
In an interview with PokerNews, the former President of ARJEL, Jean François Vilotte summarized the problem:
“It is true that as far as numbers are concerned we can see a decrease in the number of players in France, but this is due to the fact that big players are either moving out of the country or playing on illegal websites.”
His successor, Charles Coppolani has frequently called for French taxes to be lowered because of their impact on poker. So far, the French parliament has ignored him.
The Copenhagen report makes much of its graph showing the tax rate at which the maximum tax is collected, together with a curve showing forecast gross participation rate.
It identifies the 20 percent tax rate as the rate which will result in peak gaming taxes.
A report by Price Waterhouse Coopers (PwC) in 2012, entitled Regulation of online gambling in Sweden, evaluating tax scenarios in order to define the best regulatory model for the regulation of online gambling, recommended that a 10 percent tax rate would bring in the most tax.
“All other things being equal, in the long term we would expect a 10% tax on GGR to lead to the greatest tax revenues as this rate will ultimately sustain the highest proportion of the market being regulated with a Swedish license and encourage higher levels of growth than other higher tax scenarios.
Even though the Copenhagen report has the benefit of four years more data to work with, the difference between the two reports is surprising.
The PwC report went so far as to calculate the impact of France’s high taxes on the country’s tax revenues and participation rate.
“We estimate that in the two full years of regulation ending June 2012 the state has forfeited a cumulative €35 -€40 million in tax revenues by not regulating online casino games and has held back absorption which we estimate could have risen to c72% by 2011.”
It is unfair to be too critical of the Copenhagen Economics report. There is some superb data and analysis in there, and for the political level, it’s a good piece of work.
The problem for online poker is that the political level is just too high. Online poker needs a slightly different tax treatment than other online gambling to achieve the same results.
The optimal tax rate for both tax collection and consumer protection is simply lower for online poker.
U.S. states looking at an online poker-only model need to bear this in mind.