To test this I did a very quick LexisNexis search. On April 7, 2003, The Boston Globe published an article written by Frank Phillips on Mitt Romney’s first 100 days. Here’s an excerpt:
The survey – the first comprehensive poll taken since Romney assumed office – was conducted by KRC Communications Research and has a margin of error of plus or minus 5 percentage points.
It shows that Romney is relatively popular – with 55 percent of respondents expressing a positive view of him, and 32 percent a negative one.
Huh. 63% approve of Deval–a full 8% higher than Romney’s favorable ratings at this point in time.
The article continues:
But only a bare majority, 51 percent, approves of his [Romney’s] handling of the budget, while 40 percent disapprove.
And that concern is affecting his overall job performance rating, with 55 percent expressing approval and 39 percent disapproval.
Mitt Romney’s disapproval ratings were a full 6% higher than Patrick’s at this time, yet the Globe says that Deval’s disapproval numbers are “a relatively high number for a governor’s honeymoon period.” I’m not a student of political science so I have a basic question: relative to what?
Also, disapproval ratings were a full 6% higher than Patrick’s at this time, yet job approval rating is a full 7% higher. How can this be?
I’m not exactly sure, but I think it has something to do with expectations. Deval’s campaign raised our collective expectation of what a governor can do – he asked us to hold him to a higher standard. It was an extraordinary campaign. We now find ourselves not reading the poetry of the campaigns but winding our way through the complex prose of governing. Yes, the negative press wasn’t great, but frankly I don’t think most of the state cares about the Caddy, and the poll says as much. The reason people still have a highly favorable view of Patrick while a less favorable view of his job as governor is, I think, that his Administration still represents potential. Stumbling out of the gate distracted us from that potential. Even though he’s done great work on the bond bill, stem cells, Cape Wind, RGGI, and more, those things don’t make the splash that tone-deaf missteps do.
P.S. Want a blast from the past? Check out the Phoenix’s skewering of Romney’s first 100 days.
because this is a well though out and well written post but you’re forgetting that Democrats get rated on a scale. It’s basically scaled to a number < 1.
<
p>
I had a professor like that in college, he died mysteriously a few years back.
The post is here.
<
p>
I remembered some of the early problems, and was reminded of some by the invaluable archive, Romney is a Fraud – available on the BMG blogroll. (BTW – how’s about taking down the anti-Healey sites? Isn’t that kicking someone while they’re down just a little?).
<
p>
My conclusion – same number/type of gaffes. Difference – attitude.
Back in the day, when real journalists roamed the earth, no poll with with a margin of error greater than +/-3% would ever see ink…
<
p>
… you post is reasoned well, but on shaky premises: one poll has a MOE of +/-4.4% and the other an MOE or +/-5% which, I’m sorry to say, invalidates any comparison; and neither poll (as I aver above) should be taken seriously because of the wide margins wherein error might reside.
Yes. I agree. 5% is a bad MOE.
<
p>
But: both polls have similar MOE. More importantly, the point is not that individual data points within one single poll is a great way to tell us how well the Governor is doing. The point is that people seemed to make a fuss over that (perhaps flawed) data point, but it’s in fact a whole lot better than Romney’s metrics at this time in his Administration.
<
p>
I agree with you — a 5% MOE is pretty bad. And if that’s true, people should not definitively be saying that the Patrick Administration has,
“more to worry about than to brag about.”. With that said, the only poll I can find (can you find better?) that is historically similar has Romney fairing worse in terms of approval ratings. So people should calm down over the minor stuff and focus on the business of governing.
I don’t think so.
<
p>
Let’s say (hypothetical) Governor M had a disapproval rating of 100% and (hypothetical) Governor P had a disapproval rating of 0%. How likely is it that this difference is statistically significant?
<
p>
With that kind of spread, it would be hard to find any sample size that would suggest the confidence level is anything less than 100% that this is significant.
<
p>
Let’s take the numbers at hand: Guv P has 33% and Guv M has 39%. What is the probability that this 6-point spread is indistinguishable from zero? According to my back-of-the-envelope calculations, there is about a 95% chance that this result is different from zero.
<
p>
So, about a 5% chance that the comparison is invalid. Or, about a 5% chance that you could be right.