Did anyone else notice that the much-discussed Globe poll is not of “likely voters,” or even “registered voters,” but rather, “randomly selected Massachusetts adults”? How seriously should we take poll results if they don’t reflect what would happen in an election held today?
Also, I noticed this curiosity in the breakdown of the poll’s 545 respondents (see, e.g., p. 20, the breakdown of the “right direction/wrong track” question):
Registered Democrat: 188
Registered Independent/Unaffiliated: 265
Registered Republican: 73Democrat: 261
Independent: 103
Republican: 173
I couldn’t find an explanation for the difference between those sets of figures. Maybe it’s obvious; perhaps someone could enlighten me.
As for how Governor Patrick should campaign: I’ll repeat a piece of advice I once heard a successful pol give to Barack Obama in 2008: campaign like you’re willing to lose. The Gov should be forthright about what he’s accomplished, and also about where he’s fallen short. He should take (and give) credit where credit is due; and he should acknowledge that there are things he could have done better. He should stick to the high road, as he did against Kerry Healey even when she was throwing all kinds of garbage his way.
And then let the voters decide. Far better to be a one-term governor with a record you are proud of than to achieve reelection by abandoning the principles that got you into politics in the first place. (Just to be clear: I don’t see any sign that Patrick is “abandoning” anything. I’m just saying that I hope that continues to be the case, regardless of where the poll numbers go.)
So considering Massachusetts has a population of 6.497 million, of which there are about 4.098 million are registered voters, and that 3.042 million voted in 2008 (about 47% of the total) can we conclude that the opinion of 50% of the respondents was meaningless as a matter of practical politics, and thus the survey has a +- 50%, or at least 25%, margin of error? đŸ˜‰
<
p>Statisticians, to your keyboards and enlighten us.
“Margin or Error” is a measure of statistical error. Basically it guarantees the poll will be at least that close to the true result 95% of the time, assuming your sampling is perfect. Because sampling methods are usually flawed, the true error can be significantly larger and much more difficult to calculate.
<
p>In order to calculate the effect of polling “Adults” rather than “Likely Voters” you’d have to come up with a model of how likely voters differ from the population as a whole. I imagine it increases the error somewhat, but not by nearly as much as you suggest that it would.
Margin of error is an entirely different concept than you suggest. Sampling just adults (and not likely voters) means the poll only tells us that this is the opinion of a random sampling of adults and tells us nothing about the opinion of likely voters. But nothing to do with margin of error.
<
p>This unusual sampling methodology does raise a lot of questions about the sub-sample of “likely voters” polled here or what a “real poll” would show about how people might vote if the election were held today.
<
p>I tend to agree, unfortunately, with Jay Severin who has been theorizing this afternoon that a poll of “likely voters” would probably show even less support for Governor Patrick. Compared to adults, “likely voters” tend to be white, more conservative, wealthier, more educated and older. More Baker, Mihos and Cahill voters than Patrick.
Interesting. And thanks to you too Marcus.
Lost me on the third. That was very, very selective in describing likely voters. “Likely voters” also tend to be more educated people, who also tend to vote for more liberal candidates. Furthermore, less knowledgeable voters usually vote at a far smaller rate — and those would include the third-rail types that have been particularly angry with Patrick as of late. For example, people were railing about immigration issues in 2006 and many claimed it would be the downfall of Patrick… except the anti-immigration third railers don’t vote.
<
p>Because the poll didn’t include “likely voters” we have little to no idea of how that poll would have been changed had the Globe created a valid poll. To say anything else is pure conjecture. The fact that Jay Severin thinks it so does not make it so… far from it, in fact, for Jay Severin is widely known to be an off-balanced wingnut who knows little about the very things he rails against…
That is probably a truer breakdown of how some of the unenrolled voters classify themselves. I know many Republicans and Democrats who register unenrolled to give themselves a “choice” in which primary to vote. So if I asked you what do you consider yourself to be you’d say Democrat. However if you were registered unenrolled you’d give a truthful answer to how are you registered.
<
p>That probably explains the difference.
but add up the two columns. They differ by 11 people. A 1 person difference could be a round off error, but 11?
There are 11 more total in the cross-tabs than registered. This could account for the difference in adults and registered. If you were asked how you were registered to vote and you weren’t you would answer not registered and not show up in the registration numbers.
Eleven people could be unregistered. Only about 2% unregistered is very low but possible. I would guess very few unregistered adults would agree to answer a political survey given over the phone.
Responding to the original question regarding how the Governor should campaign I would say David is right, Deval should avoid the mistakes many around here were making when they presumed he was invulnerable and the MA voters can’t possibly be so “stupid” as to vote for Cahill or Baker, and instead campaign as if his life depended on it and his back was against the wall. Clearly many BMGers are out of touch with the average MA voter when they presume they dismiss with a wave of their hand the strong challenges that Cahill and Baker/Mihos will make in the general election and why these should be taken seriously.
<
p>I have said this before, Deval is great on substance issues, he is probably one of the few governors who gets public policy and can be really wonkish on the details. Unfortunately while actual governing is about substance, politics is all about style-specifically the importance of symbols. For far too long Deval has been either too naive or too arrogant to realize that some of his actions, while materially non-substantial, like the drapes, like Marian Walsh, like Wilkerson, like the book deal, etc., are incredibly important symbols of corruption, out of touch elitism, and arrogance that have now become associated with Deval. The brand of Deval that was once successfully outside, progressive, reformer is now tainted as insider, tax and spend liberal, and corrupt.
<
p>I would argue that Deval should instead radically untie himself from Beacon Hill connections. He should can the Walsh promotion, distance himself from Wilkerson, and start going on the stump to cities and towns across MA condemning the ‘do nothing legislature’ and the ‘beacon hill culture’. He should put his impressive grassroots army to good use by getting them to campaign against DINOs and hacks that impede his agenda and by running progressive alternatives in every district. Sciortino and Elridge are proof that every now and then the good guy can beat out the hack, unfortunately they did it solo-we need a statewide effort in every district that has a hack to get rid of them.
<
p>The Gov should also put good government first by getting ballot initiatives term limits, salary caps, campaign finance reform all come to mind that would really start to put a nail in the coffin to the entrenched interests that have been running Beacon Hill for far too long. Paint Cahill as one of the hacks, paint Baker as a service cutter not a reformer, and really give em hell.
<
p>Someone said Devals coalition is untenable-I agree. It is time he severs his alliance with big labor and with corrupt political bosses on Beacon Hill and instead take his cause to the people. Harry Truman lost his southern supports because he was pro-Civil Rights, he lost his northern union support when he drafted the train workers, but in spite of a challenge to his left and right he still emerged victorious because he just campaigned as if he already lost, threw polite political considerations out the window, and was just honest with the people about what needed to be done to save the country. If Deval can do the same thing then he has a fighting chance to survive.
Because of seriously flawed polling data.
In 1948:
Surveys were conducted by telephone. Who had telephones? People who could afford them.
If i may stereotype for a moment. Republicans could afford telephones, democrats could not.
Therefore the polling data was seriously flawed, which besides the fact that women hated Dewey’s mustache, and women weren’t allowed to get their housecoats wrinkled or dirty in politics, so women and/or poor americans were not included in the polling.
because he thought women wouldn’t (in general) vote for a mustachioed man, but is there any evidence that the mustache impacted the way women voted?
I like this idea. What should we get on there?
* abolition of the state senate and adoption of unicamerialism (as in NE)
* abolition of county sheriffs
* part-time legislature
* consolidation of “constitutional officers” — treasurer and auditor; attorney general and sec of state
* abolition of the Governor’s Council
I think the Gov needs to seriously turn on the legislature just like Truman turned on the ‘do nothing’ Congress. Its the only choice he has to distance himself from the culture of corruption on Beacon Hill and possibly position himself as an outsider AND its the only chance he can actually govern if he does win. I see Deval living to fight another day if his opposition is split and due to the overwhelming advantage any democrat has in this state. That said, any win would be by a slim margin and cripple his ability to govern-if he doesn’t think outside of the box.
<
p>I think unicameral and eliminating state offices and the governors council would require constitutional amendments and couldn’t be done by the ballot BUT the Governor could put on advisory referendum on those topics so the people can at least make their voices known and associate that populism with Deval.
<
p>Realistically term limits, abolishing county government (albeit ballots at the county level), capping state legislator pay, all of those reforms could be done on the ballot and I think could convince independents as well as progressives to give Deval another shot.