Polling Pennsylvanians Poorly

Much has been made of recently released “survey” results by our good friends at CSIG about the attitudes of the citizens of the Keystone state as to online gambling.  And they ain’t pretty.  The results, for us, that is.  The overall take away from the results suggests that Pennsylvanians are unfavorable towards the idea of online gambling, and once presented with selected “messages” are even more inclined to be unfavorable.  This is considered “bad news” for us poker advocates and for more of the legislature as it steamrolls towards becoming the fourth U.S. state to legalize and regulate online poker.

And it would be bad news if in fact the results accurately measured the feelings of the state’s citizens.  Which I do not believe it does.  Many poker advocates have suggested reasons why this survey does not hold water.  I agree, but not for many of the reasons that have been suggested.

First, some background.  Most of you know me as an ice cream entrepreneur, but my previous occupation (of many) was Marketing Research.  For eight years in Idaho I did field work, crafted questionnaires, did analysis, made presentations and reports.  From this perspective I speak, rather than a poker advocate.  Ergo, this kind of thing is in my wheelhouse.

And in short…this survey seems shoddy.

First, let’s discuss the sample of respondents.  It’s been suggested that 513 respondents is inadequate to measure the entire state.  That’s wrong – 513 can be an adequate sample, provided certain precautions are taken (learn all about sample sizes here – but be warned – it’s boring stuff).  While this poll was of “registered voters,” no other information is given as to how they were selected.  I assume it was a “random” sample, and that there were attempts to mirror the population at large, but given that none of this is shown in the results…I am skeptical.  Yes, 500 is enough, but only if it’s done right.  Without any additional information (a shortage that many have rightly criticized), I remain skeptical.

The margin of error shown for the survey was given at +/-4.33%.  This would be about correct for a sample of 513, but only for the ENTIRE survey results.  The individual group breakouts (Republican/Democrat, Liberal/Conservative, age demos) would be much larger.  In looking at ago breakouts in particular is where I get very skeptical.

Take this question about whether online gambling is very different from casino gambling.  Here are the results for the “very different” responses, and the percent of the population each age demo represents:

Overall	18-39	40-54	55-65	66+
68%	57%	63%	74%	72%
% of population	34.6%	33.3%	11.5%	20.5%

Note that the “younger” demos – basically all those 54 and younger – are less likely to think online is very different than the older crowd, yet…overall percentage of 68% is much closer to what the oldies think.  I think this means the respondents skew older (disproportionately) which, of course, biases the results.  Even though Harper Polling said it was conducted on both land line phones and cell phones, it seems off to me.

And on the subject of Harper – they are notorious as both a “robo-caller” and a “push poller” (not really seeking information as much as trying to persuade) for the Republican Party.  This of itself doesn’t discredit them to provide accurate results, but…in reading the report, you see this:

The State lottery has the best gaming image in the eyes on Pennsylvanians, at 84% favorable. Horse racing tracks (60% favorable) and “casinos with Las Vegas style gambling” (59% favorable) are also viewed favorably by most Pennsylvanians. By comparison, online gambling has a remarkably bad reputation with 72% of voters holding an unfavorable opinion of it (21% favorable).

Respondents said they had an "unfavorable" opinion of online gaming. Yet the report writers said "online gambling has a remarkably bad reputation."  Those are not synonymous statements, and any market research company that writes crap like that should NEVER be trusted.

The rest of the objections that others have brought up also make me question the results.  There is no indication as to what, if any, screening questions were asked.  Certainly, self-qualification as a register voter vs. actually using registered voter records is one thing I’d like to know about (many people will say they are registered even if they aren’t because “it’s the right thing to say.”  Seriously.).  But did they exclude anyone else…like anyone who gambled?  We don’t know.  Was there any statements given about gambling or online gambling prior to the questions they show in the survey?  I’d love to see the ENTIRE questionnaire just to be sure.

Finally, it’s clear to me that the purpose of the survey wasn’t so much as to measure respondents’ attitudes but to test messages for a potential battle in the state for the legalization of online gambling.  More specifically, to use as a “threat” to legislators who dare pass pro-online gambling regulation.  We have the report itself as evidence.  It states:

See Table 1 for a breakdown of the message tests against online gambling ranked by effectiveness.  All of the messages were highly effective, making at least 74% of those surveyed less likely to favor legalizing online gambling. The most effective argument was about the potential for children to be exposed, which appeals to a wide range of demographic groups. Even the two messages provided as “pro-online gambling” arguments earn high ‘less likely’ numbers, reflecting the fact that voters are so strongly predisposed to disliking online gambling regardless of what arguments they hear in support of it.

I have to guess the two “pro-online gambling” messages were not shown in Table 1, or I was unaware  that phrases like “Online gambling is a job killer” and “the FBI told Congress last  year, ‘online casinos are vulnerable to a wide array of criminal schemes…” were pro-online gambling.  Who knew?

If this truly was the intent of the poll, and I suspect it was, I could have saved Sheldon thousands of dollars by telling him what he already knew.  I would have said: “Keep telling the same lies to people about children, money laundering, and killing jobs – just be sure you call it “information.”  And remember – all the lies seem to work about the same.”

Asshat.