Archive for the ‘casinos’ category

Betting on Gambling Assumptions

March 5, 2013

The gambling debate in NH is as hot as it has ever been as the NH Senate just passed a casino gambling bill.  Since I have no dog in the fight (or more appropriate to the debate – no pony in the race) I’ll use this blog to add my $.02.   I think the issue will be decided largely on the basis of something other than the impact casino gambling would  have on state revenues, but to the extent that fiscal impacts are a part of  policymaker’s decision process I’d like to see them have access to the best information and tools with which to make their decision.   Public policy analysis is not physics, there aren’t formulas with constants that govern  behaviors today the same way they did one million years ago.   Policy research is mostly social science that relies on a combination of disciplines like economics, sociology, and demography, and others.  The goal of policy analysis isn’t to prove anything or have something published in an academic journal (a fact usually lost on academics)  it is to improve the information in the debate and to marginally improve the decision-making process.  Policy research is best when it not only provides information, but also when it increases policy maker’s understanding of the issue and how even small changes in policy proposals might affect the ultimate impact of a proposal.   A lot of lobbyists want to provide the one “answer” to what will be the impact of this or that proposal or what will be its fiscal costs or benefits.  A lot of lawmakers want a single “point estimate” of impacts as well, when in fact there is always a range of likely impacts (some more likely than others) and usually they depend on a set of assumptions.   I’ve done a lot of policy research and I never assume anyone will agree with any of the assumptions I include in my policy models so I always design them for policy makers to insert their own assumptions in order to calculate the impacts of policy proposals.  That both increases the confidence policy makers have in their decision-making by helping them understand the sensitivities of estimates to different assumptions and the key determinants or levers that produce different impacts, and it reduces concerns that my analyses are using unrealistic assumption or “cooking” the numbers.  But what it really does is provide a ‘tool” for policy makers to use rather than giving them my “answer” to any policy question.  I usually do have my preferred answer but it doesn’t do any good unless lawmakers can see that it isn’t just “my preferred answer” bu the result of some pretty sound empirical analysis, even when it can be interpreted differently.  Invariably I offer to make my models available to policy makers but to date at least, only a few have every taken my up on it.

With that long preface I’d like to suggest that all sides of the gambling debate make their models and assumptions available and allow policy makers to get a better understanding of the sensitivities of their estimates to different assumptions.  My friend Dennis Delay at the NH Center for Public Policy Studies  is about the best there is at shooting straight and trying to develop the best estimates possible but while he does the analysis I don’t think he does all of the report writing so I would like to see more attention in their analysis of gambling to demonstrating likely impacts under a range of assumptions because it seems that not everyone agrees with theirs.  I haven’t seen any detailed analyses by gambling proponents and think they need to provide their assumptions and models for estimating revenues and impacts as well if they want lawmakers to adopt their proposals (they may have done this I just haven’t seen any analysis).

To demonstrate how important assumptions are in estimating revenues I developed a small model of gambling revenue in NH (not including any social costs).  The model results presented below assume one casino in Southern NH with 5,000 slot machines (or video lottery terminals) but any number of slots can be entered as a variable.  The base for state revenues (on which a state tax would be applied) is estimated using  per slot machine revenue data from Connecticut Casinos for the most recent year (2012).  This seems like the most similar market but again, a different per slot figure can be entered into the model to yield different results.  In addition, different tax rates and different impacts from Massachusetts casinos can be entered into the model.  I’m not trying to estimate revenues here but I am trying to highlight just how important model assumptions can be in determining fiscal impacts and until all sides show how their estimates are affected by their assumptions I think it is hard for lawmakers to make reasoned decisions based on fiscal impacts.  As the chart below shows, estimated state revenues vary greatly when even a few assumptions change.  The “Y” or left, vertical  axis shows estimated state revenues, and the “X”  or bottom, horizontal axis shows increasing tax rates from left to right.  Each colored line on the graph shows estimated state revenue at each tax rate and each  colored line represents a different assumption about the impact on revenues depending on how much casinos in Massachusetts affect casino revenues in NH.

Sensitivity of Revenue Estimates

Finally, I would like someone to articulate and provide some data on how  casinos in NH would perform in a competitive market depending on the type of experience they provide.  That seems to me to be a question best answered by the industry.  I think it is important in understanding the impacts of an increasingly competitive gambling market and the data I have looked at suggest that, at least in Nevada (see below), casinos have derived an increasing share of their revenues from rooms, meals, beverages, retail and shows.  Entertainment seems to play a larger role in the business models of casinos in that state and I wonder if that will be true in NH or in Massachusetts and what are the implications if it isn’t in either, both, or if it is in just one state.

Sources of Casino Revenue

Can We Be Different Like Everyone Else?

January 4, 2013

I was surprised to see the number of states that have allowed casino gambling.  In a prior post I focused on what I thought were the states that are perhaps most identified with casino gambling (Nevada, New Jersey, and Connecticut).  Twenty three (23) states and five since 2005 (if you count Massachusetts) now allow some type of casino gambling.  As the map below shows, the Northeast region of the country is the king of casinos.  I don’t know what that says about the Northeast but Vermont and New Hampshire are now the only states in the region that do not have some form of casino gambling. (a note about the data in the charts below:  I have taken reasonable steps in the limited time I allocate to this blog to provide accurate information – if anything appears inaccurate please let me know).

Note: Map is Updated thanks Curtis!

 Competitive Casino Map

I think whether or not to become more like other states in the region is an important and ongoing debate in New Hampshire, whether it be about our revenue structure, which stands out in the region, or our political, legislative, and regulatory structures which to a lesser degree do as well.  I’ve long argued that the state was able to buck the region’s unfavorable demographic and  economic trends because it was somewhat unique in the region.  Some who disagree with me on that argue that the state should, in the case of casino gambling, refuse to become more like the rest of the Northeast region.  While others who agree with me on the benefits of NH’s uniqueness are arguing that NH should have casino gambling because other states in the region are doing it.  Consistency isn’t what it used to be or perhaps I just confuse consistency with rigidity.  It is also possible that I am misreading the whole consistency and change aspect of the debate.  Could it be that gambling is consistent with NH’s fiscal traditions but inconsistent with its uniqueness in the region?  I don’t expect there will be a lot of testimony on that at any public hearings on casino proposals.  For those more interested in the pedestrian issue of how much state revenue we can expect, below is a chart that shows how much states currently take in from casinos (in very broad categories).  Interesting to see that Pennsylvania is now the champion in terms of state revenues from casinos.  That state is, in large part, responsible for the decline in revenues in New Jersey.  Things are definitely changing in NH and the upcoming debates over whether or not to allow casino gambling will, I think, tell us a lot about the direction of that change.

State Revenue from Casinos

The Odds on Gambling Have Gone Up?

December 4, 2012

(Update:  An astute reader has suggested that if gambling is more likely then, in fact, the “odds have gone down”  – I’ve added a question mark to the title to reflect my uncertainty.)

With a newly elected governor open to this possibility of introducing casino gambling in NH and a state senate that has been similarly inclined in recent years, by all accounts casino gambling will be a hot topic in the legislature this year.    Because this blog attempts to address timely issues, this is the first of what will be a number of posts that take an empirical (as opposed to moral or ideological) look at the fiscal and economic impacts of casino gambling in NH.  First, a disclosure:  I am not now nor have I ever been involved in debates regarding expanded gambling in New Hampshire or anywhere, either professionally or personally.  I only point this out because so much of the debate is framed by advocates or opponents, the information from each is often discounted by some percentage because of their advocacy.  There may be other reasons to discount the information I offer here but my position as an advocate or opponent ought not be one.  Most of what I will be posting here are aspects of casino gambling that interest me, and thus may not be the most important or most relevant from a public policy perspective.   If there are issues you think I ought to examine,  feel free to let me know.

casino tax revenues

Looking at the basics, casino gambling provides substantial revenues to the three states I examined for this post.  The chart above shows the state government revenues provided by casino gambling revenues of CT, NJ, and Nevada.  For Nevada, the  revenues do not include casino gambling related revenues such as taxes on rooms or meals and for each state the figures do not include any licensing revenues.  Revenues in Connecticut are the “cleanest” to interpret, they represent the state’s 25% share of the “win” from slot machines at both Foxwoods and Mohegan Sun.  I don’t know (yet) the revenue formula for Nevada or New Jersey but I do know that slots are important regardless.  In Nevada, slot machines represent about 65% of casino revenue (61% at publicly traded – larger- casinos).  A key question from the chart is whether the decline in state revenues is cyclical, or structural.  Cyclical involves changes resulting from economic conditions, while structural involves issues such as competition (the increase in gambling locations) and perhaps the 800 pound gorilla, internet gambling.  Either way, the decline in gambling revenue has been especially dramatic in New Jersey, which likely reflects a combination of increased competition (especially from new Pennsylvania casinos) as well as economic conditions.  It will be interesting to see how CT and NJ fare with the impending opening of casinos in Massachusetts.

Casino gambling revenues come from the discretionary spending of consumers.  Discretionary income suffered significantly during the recent recession.  In NH, the revenue source probably most dependent on discretionary spending is the state’s meals and rooms tax.  A significant percentage of the meals and rooms tax is related to entertainment and recreation spending, similar to casino gambling (although the meals and rooms tax is also a function of business conditions and expenditures).  A quick comparison of gambling and meals and rooms tax revenues shows that the meals and rooms tax (adjusted for rate changes) declined during the recession but has been a more stable source of revenue for NH than has casino gambling revenues has been for other states.  Another interesting fact is that revenue from NH’s meals and rooms tax was about $237 million in FY 2012, while casino gambling revenues in NJ were about $238 million.Casino vs meals and rooms

%d bloggers like this: