Thursday, November 6, 2014

Massachusetts gubernatorial polls were less biased than national midterm polls

A look at the polling average and a comparison of each pollster's last poll to the final result

While a comprehensive look at the polls of the 2014 midterm races show that the polls were biased by several points towards Democrats, that did not seem to be the case for the totality of the Massachusetts gubernatorial polls of the race between Martha Coakley and Charlie Baker. The final Mass. Numbers time-weighted polling average showed Baker up by two points, essentially equalling the unofficial election result of a 1.88 point Baker win, and a simple average of the last poll from each pollster shows a similar result of Baker +2.3.

While the average was very close to the election result, the final polls of the Massachusetts gubernatorial race showed a reasonably wide range from a 4 point lead for Martha Coakley, to a 7 point lead for Charlie Baker.

The following table gives details of each pollster's final poll sorted by difference from the election result. The polling methods included internet panel, live phone interviews, automated phone calls (IVR), and IVR supplemented with an internet panel to reach non-landline respondents.

MA Governor polls vs. result (chart)

The YouGov poll showing Coakley up 4 was the biggest miss of the final polls with a difference of -5.88 points from the actual result. The Globe/SocialSphere poll had Baker up 7 (a 5.12 point difference), the UMass poll missed by -4.88 points, and the Emerson College Polling Society poll missed by 4.02 points.

The closest poll to the final result was the Rasmussen IVR/Internet poll 22 days before the election which had an almost perfect 2 point Baker margin. MassINC Polling was also close to perfect with Baker +1, closely followed by Suffolk with Baker +3. UMass Lowell was just over 2 points off (Baker +4), and WNE had Baker +5 for a miss of 3.12 points.

While some of the polls of the Massachusetts gubernatorial race were remarkably accurate and others missed the mark by 5 or 6 points, the overall picture provided by the polls in the form of an average gave us a good idea of the eventual result.

Wednesday, November 5, 2014

Money from Massachusetts GOP shadow group fails to unseat targeted state representatives

Massachusetts Fiscal Alliance spending on immigrants-over-veterans mailings had little affect on race outcomes

An outside spending political action committee run by Republican State Committee member Rick Green and called the Massachusetts Fiscal Alliance portrayed itself as an issues-based committee concerned with fiscal discipline. But mailings undertaken by the committee against a group of 19 Democratic State Representatives made it clear that the goal of the group was to campaign against those house members and support of GOP challengers using a debunked immigrants-over-veterans attack.

The mailings seem to have failed as 17 of the 19 representatives were re-elected, several in very GOP-leaning red districts. A loss by Rep. Rhonda Nyman to GOP challenger David DeCoste was in the 5th Plymouth District where President Obama lost by 8 points, Senator Warren lost by 26 points, and is ranked the third most GOP-leaning Rep. district in the Commonwealth. This is a hard district for any Democrat to hold.

The other loss came in the more Democratic 2nd Franklin District. Rep. Denise Andrews of Orange lost to two-time challenger Susannah Whips Lee. Lee outraised Rep. Andrews in the pre-primary and pre-election reporting periods, raising $44,698 compared to $16,960 for Andrews.

Election results for MFA targets

The success of Democratic incumbents Josh Cutler, Colleen Garry, Jim Cantwell, Ted Speliotis, and Jim Arciero in Republican-leaning districts shows that the PAC mailings were not an effective campaign tool for the GOP.

Monday, November 3, 2014

A data supplement to legislative races to watch article

District partisan lean and pre-election fundraising data for competitive legislative races

Gintautas Dumcius and Mike Deehan put together a great overview of some of the interesting Massachusetts races for Massachusetts State Senate and Representative available on WBUR Poll Vault.

While Dumcius and Deehan touched on fundraising and district partisanship in the article, I put together a table that has each candidates fundraising total for the OCPF pre-election period, and the Obama and Warren margins for each district for your analyzing pleasure.

Legislative races to watch

Here is an online spreadsheet with the same data.

Not all survey methods are equal in 2014 midterm race polling

Many midterm races have seen a similar method-based bias as the Massachusetts gubernatorial race (internet+Dem, IVR+GOP), but others show the opposite

There have been significant and consistent differences between IVR, live operator, and internet polls of the race for Massachusetts governor with internet polls showing a Coakley +4 lead, live operator polls showing Baker +2, and automated phone (IVR) polls showing Baker +5. But is this internet+Dem/IVR+GOP bias seen across other 2014 midterm polling?

In order to answer this question I downloaded all of the October polls for a large sample of reasonably competitive midterm Governor and Senate races from the Huffington Post Pollster site, computed a polling average for each race, and then compared the race-wide average to the average for each survey method, recording a difference as positive number for Democratic bias and a negative for a Republican bias.

The results are in the following table with polling methods including live phone calls, automated phone calls (IVR by itself, or with supplements live phone calls or internet panels to reach non-landline respondents), and internet-based survey panels.

2014 Midterm survey bias by method (chart)

The results show an internet+Dem/IVR+GOP bias similar to MA-Gov in the Colorado, Maryland, Maine, and Michigan gubernatorial races, and also in the Colorado and New Hampshire Senate races.

However, some races show exactly the opposite, including Alaska, Georgia, and Michigan Senate, and Arkansas, Arizona, Florida, and Kansas Governor. Michigan is particularly strange with the Governor and Senate race on opposite sides.

A look at the distribution of differences for all of the races for each survey method provides some insight into the race-by-race biases.

2014 Midterm survey bias distribution by method (graph)

The histogram for live phone polling sows a nicely shaped normal distribution centered at the average margin, with a similar probability of differences above and below the mean.

Internet polls are skewed to the Democratic side, but shifted to the left past the mean of zero, possibly a result of the sample weighting. On the other side, IVR polls are somewhat skewed to the GOP side, but shifted to the right past the mean of zero, possible due to weighting.

The IVR/Online polls look closer to a normal distribution, but the IVR/Live Phone polls have very irregular-looking distribution shape.

A possible explanation differences in bias for some of the races could be weighting overcompensation for Democratic bias in the internet polling and for GOP bias in the IVR polling.

After Tuesday's election results it will be possible to determine which survey method type was most accurate in the midterm polling and look for explanations of the affect of survey methods on poll bias.